Seino Van Breugel-A Grammar of Atong PDF

A Grammar of Atong
Submitted by
Jonkheer Egbert Joost Seino Clifford Kocq van Breugel, M.A.
a.k.a. Seino van Breugel

A thesis submitted in total fulfilment
of the requirements for the degree of
Doctor of Philosophy

Research Centre for Linguistic Typology

La Trobe University
Bundoora, Victoria 3086
Australia

16 December 2008

v
Table of contents
_____________________________________________________________________

LIST OF TABLES ..................................................................................................... XVI
LIST OF MAPS ........................................................................................................ XIX
LIST OF FIGURES ................................................................................................... XIX
SUMMARY ...............................................................................................................XX
STATEMENT OF AUTHORSHIP .............................................................................. XXIII
ACKNOWLEDGMENTS ........................................................................................... XXV
DAKANGGABA KATHA ......................................................................................... XXIX
BADRI KHUCHUKSANG ..................................................................................... XXIX
SIJYW KHUCHUKSANG ..................................................................................... XXIX
FOREWORD IN ENGLISH ...................................................................................... XXX
LIST OF ABBREVIATIONS AND SYMBOLS .............................................................. XXXI
CHAPTER 1 THE ATONG LANGUAGE AND ITS SPEAKERS .................... 1
1.1 LOCATION OF THE LANGUAGE AND NUMBER OF SPEAKERS ....................... 1
1.2 NAMES AND ALLONYMS .......................................................................... 6
1.2.1 Language names ................................................................................. 6
1.2.2 Remarks on some toponyms on Map 3 ................................................ 6
1.3 THE ATONG PEOPLE ................................................................................ 8
1.3.1 Ethnic affiliation ................................................................................. 8
1.3.2 Social organisation............................................................................ 10
1.3.3 Living environment: the compound................................................... 11
1.3.4 Living environment: the jungle ......................................................... 12
1.3.5 Ceremonies and festivals .................................................................. 13
1.3.6 Contact with others ........................................................................... 15
1.3.7 Economy .......................................................................................... 16
1.4 LINGUISTIC ENVIRONMENT, LANGUAGE STATUS AND LANGUAGE USE ..... 17
1.5 THE ATONG SPELLING SYSTEM .............................................................. 21
1.6 DIALECTAL VARIATION ......................................................................... 22
1.7 LINGUISTIC AFFILIATION ....................................................................... 24
1.8 PREVIOUS WORK ON ATONG ................................................................. 34
1.9 FIELDWORK .......................................................................................... 37
1.9.1 Data collection .................................................................................. 37
1.9.2 Recording equipment ........................................................................ 41
CHAPTER 2 PHONOLOGY ............................................................................. 43
2.1 SYLLABLE STRUCTURE ......................................................................... 43
2.2 CONSONANTS ....................................................................................... 44
2.2.1 Stops ................................................................................................. 44
2.2.2 Fricatives .......................................................................................... 45
2.2.3 Affricates .......................................................................................... 46
TABLE OF CONTENTS vi
2.2.4 The tap or trill and the oral continuant ............................................... 46
2.2.5 Nasal continuants .............................................................................. 47
2.2.6 Glides ............................................................................................... 47
2.3 THE MORPHOPHONOLOGICAL PROCESS OF FUSION .................................. 49
2.4 VOWELS ............................................................................................... 51
2.5 VOWEL DEVOICING AND ELISION ........................................................... 54
2.6 VOWEL ASSIMILATION .......................................................................... 55
2.7 VOWEL PHONOTACTICS ......................................................................... 59
2.8 MORPHOPHONOLOGICAL VOWEL ASSIMILATION .................................... 60
2.9 CONSONANT LENGTH ............................................................................ 61
2.10 VOWEL LENGTH .................................................................................... 63
2.11 AMBISYLLABIC CONSONANTS ............................................................... 63
2.12 GLOTTALISATION ................................................................................. 65
2.12.1 Alternative analyses against glottal prosody. ..................................... 68
i The glottal stop as a phoneme ........................................................ 68
ii Glottalised continuants ................................................................... 69
2.12.2 Conclusion ........................................................................................ 70
2.13 THE ATONG WORD ................................................................................ 70
2.14 ACCENTUATION, STRESS AND PROSODY ................................................. 71
2.15 PHONOLOGICALLY ABERRANT WORDS ................................................... 77
2.16 THE PHONOLOGY OF LOAN WORDS ........................................................ 78
2.16.1 Vowels ............................................................................................. 78
2.16.2 Consonants ....................................................................................... 80
i Loans from English ........................................................................ 80
ii Loans from Indic languages ........................................................... 81
iii Loans from Garo ............................................................................ 82
CHAPTER 3 WORD CLASSES: AN OVERVIEW ......................................... 83
CHAPTER 4 VERBS ......................................................................................... 85
4.1 CLAUSAL PROPERTIES ........................................................................... 85
4.2 PHRASAL PROPERTIES ........................................................................... 86
4.3 MORPHOLOGICAL PROPERTIES .............................................................. 86
4.4 SEMANTIC PROPERTIES ......................................................................... 86
4.5 SUBCLASSES OF VERBS ......................................................................... 87
4.5.1 Primary-A verbs ............................................................................... 88
i Intransitive verbs ........................................................................... 88
ii Verbs of emotion and interaction ................................................... 88
iii Verbs that take arguments which are obligatory unmarked for case 89
iv The copula and the locative/existential verbs .................................. 89
v Transitive verbs ............................................................................. 93
vi Extended transitive verbs ............................................................... 94
vii The interrogative verb atak ............................................................ 94
viii Verbs denoting natural phenomena ................................................ 96
TABLE OF CONTENTS vii
4.5.2 Primary-B and Secondary verbs ........................................................ 99
4.5.3 The Secondary speech-verb ............................................................ 100
4.5.4 Phasal verbs .................................................................................... 101
4.6 INTRANSITIVE-TRANSITIVE LEXICAL PAIRS .......................................... 101
CHAPTER 5 ADJECTIVES ............................................................................ 104
5.1 TYPE 1 ADJECTIVES ............................................................................ 105
5.2 TYPE 2 ADJECTIVES ............................................................................ 108
5.2.1 Clausal properties ........................................................................... 109
5.2.2 Phrasal properties ........................................................................... 109
5.2.3 Morphological properties ................................................................ 109
5.2.4 Semantic properties ........................................................................ 109
5.3 REMARKS ON CERTAIN ADJECTIVES ..................................................... 111
CHAPTER 6 NOUNS ....................................................................................... 114
6.1 CLAUSAL PROPERTIES ......................................................................... 114
6.2 PHRASAL PROPERTIES ......................................................................... 114
6.3 MORPHOLOGICAL PROPERTIES ............................................................ 114
6.4 SEMANTIC PROPERTIES ....................................................................... 115
6.5 SUBCLASSES OF NOUNS ....................................................................... 115
6.5.1 Common nouns ............................................................................... 115
6.5.2 Nouns denoting persons and proper names ...................................... 116
6.5.3 Inherently locational nouns ............................................................. 117
6.5.4 Mass nouns ..................................................................................... 118
6.5.5 Gender sensitive nouns ................................................................... 118
6.6 JUXTAPOSITION OF NOUNS .................................................................. 119
6.6.1 Addition interpretation .................................................................... 119
6.6.2 Modifying interpretation ................................................................. 120
6.6.3 Different-NP interpretation ............................................................. 121
CHAPTER 7 KINSHIP TERMS ..................................................................... 122
7.1 MORPHOLOGY-BASED DIVISION OF KINSHIP TERMS: THE ENCLITIC
<=gaba ~ =ga> .................................................................................. 123
7.2 SEMANTIC DIVISION OF KINSHIP TERMS ............................................... 126
7.2.1 Classificatory versus descriptive kinship terms ............................... 126
7.2.2 Reciprocal versus non-reciprocal kinship terms .............................. 127
7.2.3 Reference versus address kinship terms .......................................... 128
7.3 ADDRESS TERMS ................................................................................. 129
7.4 THE CONSANGUINEAL FAMILY FROM THE PERSPECTIVE OF a me ...... 133
7.5 THE IN-LAW FAMILY ........................................................................... 137
7.6 FAMILY LOSS ...................................................................................... 139
7.7 HOW TO ADDRESS PEOPLE WHO ARE NOT KIN ....................................... 140
7.7.1 Addressee is younger than the speaker ............................................ 141
7.7.2 Addressee is older than the speaker ................................................. 141
TABLE OF CONTENTS viii
CHAPTER 8 DEMONSTRATIVES ................................................................ 142
8.1 DEICTIC PROPERTIES ........................................................................... 142
8.1.1 Purely deictic use ............................................................................ 142
8.1.2 Anaphora ........................................................................................ 144
8.2 CLAUSAL PROPERTIES ......................................................................... 145
8.3 PROPERTIES AS HEAD OF A PREDICATE ................................................. 146
8.4 PHRASAL PROPERTIES ......................................................................... 147
8.5 MOROPHOLOGICAL PROPERTIES .......................................................... 149
8.6 OTHER FUNCTIONS OF THE DEMONSTRATIVES ...................................... 149
8.7 THE ADVERBIAL DEMONSTRATIVE otokoy ............................................ 150
8.8 DEICTIC-ONLY DEMONSTRATIVES ....................................................... 152
CHAPTER 9 INTERROGATIVES ................................................................. 154
9.1 PROPERTIES OF INTERROGATIVES ........................................................ 155
9.2 ca who .......................................................................................... 156
9.3 ato what ........................................................................................ 157
9.4 atotokoy why, how come ............................................................... 158
9.5 atakna ~ atana why ......................................................................... 158
9.6 atomay'na why ............................................................................. 159
9.7 atakay ~ atokoy how ........................................................................ 159
9.8 bie ~ bi which, where ...................................................................... 159
9.9 biskon AND boysok how much/many ................................................ 160
9.10 biba when, in whatever place .......................................................... 162
9.11 bitokoy by which way? .................................................................... 162
9.12 bici where ....................................................................................... 163
9.13 bisa to/from where AND bisami from where .............................. 163
9.14 bimi ~ bimo (from) where .............................................................. 164
9.15 biga ~ bigaba which ........................................................................ 165
CHAPTER 10 INDEFINITE PROFORMS .................................................. 166
10.1 THE INDEFINITE PROFORM je any, whichever, whatever .................... 166
10.2 DERIVATIONS FROM je any, whichever, whatever ............................. 167
10.3 caba, atoba, biciba, bisaba AND bimiba ....................................... 168
10.4 cagaba whoever ............................................................................ 170
10.5 daraba anybody............................................................................. 171
10.6 gumuksa everywhere ..................................................................... 171
CHAPTER 11 NUMERALS ......................................................................... 174
11.1 TYPES OF ATONG NUMERALS .............................................................. 175
11.1.1 Unit numerals ................................................................................. 179
11.1.2 Round-Number numerals and the use of different paradigms .......... 182
11.2 BORROWED NUMERALS ....................................................................... 186
11.2.1 English loans .................................................................................. 186
11.2.2 Hindi loans ..................................................................................... 187
TABLE OF CONTENTS ix
11.3 WHAT IS QUANTIFIED WITH WHICH NUMERALS? .................................. 188
11.4 THE POSITION OF THE CLASSIFIER ........................................................ 190
11.5 SYNTACTIC AND MORPHOLOGICAL PROPERTIES OF NUMERALS ............. 192
11.6 ORDINAL NUMBERS ............................................................................ 197
11.7 THE NUMERAL sa one: ITS DIFFERENT FUNCTIONS AND
GRAMMATICALISATIONS ..................................................................... 199
CHAPTER 12 CLASSIFIERS ...................................................................... 204
12.1 THE SYNTACTIC AND SEMANTIC PROPERTIES OF CLASSIFIERS ............... 204
12.2 CATEGORIES AND TYPES OF CLASSIFIERS AND THEIR USE ..................... 207
12.2.1 Sortal classifiers .............................................................................. 208
12.2.2 Repeater classifiers ......................................................................... 209
12.2.3 Mensural classifiers ........................................................................ 211
12.2.4 The relationship between noun and classifier .................................. 211
12.3 AUTO-CLASSIFIERS ............................................................................. 212
12.4 MEASURE NOUNS ............................................................................... 215
12.5 THE ORIGIN OF CLASSIFIERS IN ATONG ................................................ 216
CHAPTER 13 POSTPOSITIONS ................................................................ 226
13.1 THE POSTPOSITION daka ................................................................... 226
13.2 THE POSTPOSITION konsa .................................................................. 227
13.3 THE POSTPOSITION gomon ................................................................... 227
13.4 THE LIMITATIVE POSTPOSITION dabat ................................................. 228
13.5 THE LIMITATIVE POSTPOSITION thol' ................................................... 231
CHAPTER 14 TIME WORDS ...................................................................... 234
14.1 THE PROPERTIES OF TIME WORDS ........................................................ 234
i Clausal properties ........................................................................ 234
ii Phrasal properties ......................................................................... 235
iii Morphological properties ............................................................. 235
iv Semantic properties ...................................................................... 237
14.2 THE WORD daka ............................................................................... 237
14.2.1 As time word .................................................................................. 238
14.2.2 As a genitive-marked Possessor ...................................................... 239
14.2.3 With the attributive suffix <-gaba ~ -ga> ........................................ 239
14.2.4 With the adverbialising suffix <-gaba ~ -ga> .................................. 239
14.2.5 As underived adverb ....................................................................... 240
CHAPTER 15 ADVERBS ............................................................................. 242
CHAPTER 16 DISCOURSE CONNECTIVES ............................................ 246
16.1 TYPE 1 DISCOURSE CONNECTIVES ........................................................ 246
16.1.1 The origin of Type 1 discourse connectives .................................... 248
16.1.2 otokoymo and its allomorphs .......................................................... 249
TABLE OF CONTENTS x
16.1.3 otokoysa .......................................................................................... 252
16.1.4 otokcido .......................................................................................... 254
16.1.5 otokciba and otokma'ciba ................................................................ 255
16.2 TYPE 2 DISCOURSE CONNECTIVES ....................................................... 257
16.2.1 uciba ............................................................................................... 257
16.2.2 umigomonci ~ umogomonci............................................................ 258
16.2.3 una ................................................................................................. 258
CHAPTER 17 OTHER WORD CLASSES .................................................. 262
17.1 THE ADDITIVE CONJUNCTION aro and ............................................... 262
17.2 PERSONAL PRONOUNS ......................................................................... 263
17.3 THE GENERIC PRONOUN ...................................................................... 267
17.4 PROCLAUSES ...................................................................................... 268
17.5 ONOMATOPOEIA ................................................................................. 273
17.6 INTERJECTIONS ................................................................................... 274
CHAPTER 18 WORD-CLASS-CHANGING DERIVATION ..................... 276
18.1 TYPES OF DERIVATION ........................................................................ 276
18.2 DENOMINAL VERBS OR DEVERBAL NOUNS, ZERO DERIVATION .............. 276
18.3 DE-ADJECTIVAL NOUNS OR DENOMINAL ADJECTIVES: ZERO DERIVATION
.......................................................................................................... 280
18.4 DE-ADJECTIVAL VERBS ....................................................................... 281
18.5 MAKING A NOUN MORE VERB-LIKE ...................................................... 282
18.6 DEVERBAL AND DE-ADJECTIVAL ADVERBS BY REDUPLICATION ............ 282
18.7 DEVERBAL ADVERBS BY ZERO DERIVATION ......................................... 284
18.8 DENOMINAL ADVERBS ........................................................................ 285
18.9 NOMINALISATION ............................................................................... 287
CHAPTER 19 PHRASAL ENCLITICS ....................................................... 290
19.1 THE POSSESSIVE ENCLITIC <=tha> ..................................................... 291
19.2 THE RECIPROCAL ENCLITIC <=maran> ................................................ 292
19.3 THE PLURAL ENCLITIC <=dara ~ =dora> ......................................... 294
19.4 THE QUANTIFIER ENCLITIC <=gumuk> ................................................ 297
19.5 THE DISTRIBUTIVE ENCLITIC <=pek> ................................................... 298
19.6 THE EXCLUSIVE ENCLITIC <=tara> .................................................. 299
19.7 THE PRIVATIVE ENCLITICS <=noy ~ =ni> AND <=ri> ............................ 300
19.8 THE ENCLITIC <=rara> ....................................................................... 300
19.9 THE ASSOCIATIVE ENCLITIC <=para> .................................................. 301
19.10 THE ALTERNATIVE ENCLITIC <=sega ~ =siga> ................................. 302
19.11 THE ADDITIVE/EMPHATIC ENCLITIC <=ba> .......................................... 303
19.11.1 Addition ......................................................................................... 304
19.11.2 Emphasis ........................................................................................ 304
19.11.3 Marker of speaker ........................................................................... 305
19.12 THE FOCUS/IDENTIFIER ENCLITIC <=an>.............................................. 305
TABLE OF CONTENTS xi
19.13 THE TOPIC ENCLITIC <=do> ................................................................ 311
19.14 THE FOCUS ENCLITIC <=e> ................................................................. 313
CHAPTER 20 CASE MARKING ................................................................. 316
20.1 ZERO MARKING .................................................................................. 320
20.2 THE MOBILITATIVE/LOCATIVE/INSTRUMENTAL CASE MARKER <=sa> . 322
20.2.1 Mobilitative interpretation .............................................................. 322
20.2.2 Locative interpretation .................................................................... 325
20.2.3 Instrumental interpretation .............................................................. 326
20.3 THE LOCATIVE CASE MARKER <=ci> (LOC) ........................................ 327
20.4 THE GENITIVE/ABLATIVE/NOMINALISER CASE MARKER <=mi ~ =mo> . 329
20.4.1 Indication of the relationship between nouns within an NP ............. 330
20.4.2 Marker of a Source ......................................................................... 331
20.4.3 Marking of the standard of comparison in equative clauses ............. 332
20.4.4 Nominalisation ............................................................................... 332
20.4.5 Repeated genitive case marking ...................................................... 333
20.5 THE COMITATIVE CASE MARKER <=mu ~ =mu ~ =mo> ..................... 334
20.6 THE DATIVE/ALLATIVE CASE MARKER <=na> ...................................... 336
20.7 REPEATED DATIVE CASE MARKING ...................................................... 337
20.8 THE ACCUSATIVE CASE <=aw ~ =taw> ................................................ 338
20.8.1 The marking of O arguments .......................................................... 339
20.8.2 Marking of material of which something is made ............................ 344
20.8.3 Purely referential/individuating/definiteness usage of the morpheme
<=aw ~ =taw> ................................................................................ 345
20.8.4 The morpheme <=aw ~ =taw> on clause initial topical S arguments 347
20.8.5 Repeated accusative case marking .................................................. 349
20.8.6 More than one accusative marked NP in a clause ............................ 350
20.9 THE HOMOPHONOUS MARKERS <=tokoy> (VIA) FOR THE PERLATIVE AND
<=tokoy> (LIKE) FOR THE SIMILATIVE ................................................ 351
20.10 MULTIPLE CASE MARKING .................................................................. 355
20.10.1 Local/Direction marking + marking of clausal function .................. 355
i Location and O ............................................................................ 355
ii Marking a Location as a Goal ...................................................... 356
20.10.2 Local marking + local marking: Direction and Source..................... 357
20.10.3 Local marking + local marking + clausal function: Direction, Source
and O .............................................................................................. 358
20.10.4 Marking of clausal function first and then of phrasal function ......... 358
20.10.5 Stem-forming genitive governed by <gomon> reason, about ......... 359
20.11 REPEATED CASE MARKING SUMMARY.................................................. 359
TABLE OF CONTENTS xii
CHAPTER 21 TRANSITIVITY ................................................................... 362
21.1 NO FORMAL DISTINCTION BETWEEN CORE SYNTACTIC ROLES ............... 363
21.2 OPTIONALITY OF COMPLEMENTS AND S=A AMBITRANSITIVITY ............ 365
21.3 S=O AMBITRANSITIVITY ..................................................................... 367
21.4 PIVOTS ............................................................................................... 370
CHAPTER 22 THE PREDICATE ................................................................ 372
22.1 DEFINING THE PREDICATE AND THE PREDICATE HEAD .......................... 372
22.2 THE MORPHOLOGICAL STRUCTURE OF THE PREDICATE HEAD ................ 373
22.3 THE VERBAL PREDICATE ..................................................................... 376
22.4 THE TYPE 2 ADJECTIVAL PREDICATE .................................................... 377
22.5 THE NOMINAL PREDICATE ................................................................... 377
22.5.1 Main clause nominal predicates ...................................................... 378
22.5.2 Subordinate clause nominal predicates ............................................ 379
22.5.3 Not only nouns ............................................................................... 380
22.6 COMPLEX PREDICATES ........................................................................ 380
22.6.1 Complex predicates with identical verbs or Type 2 adjectives ......... 380
22.6.2 Type 2-adjective-plus-support-verb compounds .............................. 382
22.7 COMPLEX PREDICATES WITH INCORPORATED NOUNS ............................ 384
22.7.1 The predicate with a prototypically associated noun ........................ 384
22.7.2 The noun-plus-support-verb predicate ............................................. 390
i The support verbs kha'- ~ kha- to do, make and tak- to do ...... 391
ii The support verb ra'- to take, get .............................................. 392
iii The copula as support verb ........................................................... 393
CHAPTER 23 PREDICATE HEAD SUFFIXES ......................................... 396
23.1 THE CAUSATIVE SUFFIX <-et> ............................................................. 396
23.2 THE CAUSATIVE ON TRANSITIVE VERBS ............................................... 396
23.3 THE RECIPROCAL SUFFIX <-ruk> ......................................................... 399
23.4 THE COMPARATIVE/SUPERLATIVE SUFFIX <-khal> ............................... 401
23.5 THE EXCESSIVE SUFFIX <-duga> ......................................................... 402
23.6 THE SIMPLICITIVE ASPECT SUFFIX <-ari> ............................................. 403
23.7 THE INCOMPLETIVE ASPECT SUFFIX <-khu> ......................................... 404
23.8 THE CUSTOMARY ASPECT SUFFIX <-a> ................................................ 405
23.9 THE DESIDERATIVE SUFFIX <-na> ....................................................... 405
23.10 THE FUTURE MODALITIES ................................................................... 407
23.10.1 The imperious future suffix <-naka ~ -ka> ...................................... 408
23.10.2 The future suffix <-ni> .................................................................... 410
23.11 THE REFERENTIAL SUFFIX <-an> ........................................................ 413
23.12 THE NEGATIVE SUFFIX <-ca> .............................................................. 415
23.13 THE CHANGE OF STATE SUFFIX <-ok ~ -ak ~ -k> ................................... 417
23.13.1 On verbal predicates ....................................................................... 418
23.13.2 On Type 1 adjectival predicates ...................................................... 420
TABLE OF CONTENTS xiii
23.13.3 On nominal predicate heads ............................................................ 421
23.13.4 On other types of predicates ............................................................ 421
23.13.5 On negated predicates ..................................................................... 421
23.14 THE PROGRESSIVE/DURATIVE ASPECT SUFFIX ...................................... 422
CHAPTER 24 THE FACTITIVE SUFFIX .................................................. 426
24.1 FACTITIVE-MARKED MAIN CLAUSE PREDICATES ................................... 426
24.2 THE FACTITIVE ON TYPE 1 ADJECTIVES................................................ 432
24.3 FACTITIVE-MARKED COMPLEMENT CLAUSES ....................................... 433
24.3.1 Factitive-marked object complement clauses and nominalisation .... 433
24.3.2 Factitive-marked subject complement clauses ................................. 439
24.3.3 The syntactic status of factitive-marked complement clauses .......... 440
24.4 FACTITIVE-MARKED CLAUSES WITH DATIVE AND LOCATIVE CASE-
MARKING ........................................................................................... 441
24.4.1 Factitive-marked Standard of comparison and Comparee clauses .... 442
24.4.2 Factitive-marked adjunct clauses with the dative case ..................... 443
24.4.3 Factitive-marked adjunct clauses with the locative case .................. 443
24.4.4 Factitive-marked adjunct clauses with the similative case ............... 443
24.5 FACTITIVE-MARKED COMPLEMENT CLAUSE OF POSTPOSITION .............. 444
24.6 SUMMARY OF PROPERTIES OF FACTITIVE-MARKED CLAUSES................. 445
24.7 DIACHRONIC NOTE ............................................................................. 446
CHAPTER 25 EVENT SPECIFIERS........................................................... 448
25.1 THE FUNCTION OF EVENT SPECIFIERS ................................................... 448
25.2 ORIGIN AND MEANING DIFFERENTIATION ............................................. 449
25.3 CATEGORIES ...................................................................................... 450
25.4 STRIKING PHONETIC FEATURE ............................................................. 450
25.5 OVERVIEW AND SOME COMMENTS ....................................................... 451
CHAPTER 26 CLAUSE TYPES .................................................................. 458
GENERAL PROPERTIES OF INDEPENDENT AND DEPENDENT CLAUSES ...................... 460
26.1 INTERROGATIVE CLAUSES ................................................................... 461
26.1.1 Content questions ........................................................................... 461
26.1.2 Predicateless focus content question clauses ................................... 462
26.1.3 Clauses with interrogatives as predicate head .................................. 463
26.1.4 Marked and unmarked polar questions ............................................ 464
26.1.5 Alternative question sentences ........................................................ 466
26.2 IMPERATIVE CLAUSES ......................................................................... 467
26.2.1 Politeness ....................................................................................... 468
i The bare imperative ..................................................................... 468
ii The imperative with <=bo> .......................................................... 469
iii The imperative with <-khu> ......................................................... 470
26.2.2 The prohibitive with <=bay> .......................................................... 472
26.2.3 The prohibitive with <ta> ............................................................... 474
TABLE OF CONTENTS xiv
26.2.4 The optative .................................................................................... 476
26.2.5 The hortative strategy ..................................................................... 476
26.3 DECLARATIVE CLAUSES AND IDENTITY/EQUATION CLAUSES ................ 477
26.4 THE PRESENTATIVE CLAUSE ................................................................ 478
26.5 COPULA CLAUSES ............................................................................... 479
26.6 QUOTATIVE CLAUSES .......................................................................... 481
26.7 REACTIONS TO INDEPENDENT CLAUSES INVOLVING PROCLAUSES ......... 482
26.7.1 The agree/disagree system .............................................................. 482
26.7.2 The yes/no system ........................................................................... 483
26.7.3 The echo system ............................................................................. 485
26.8 THE IRREALIS ENCLITIC <=com> ......................................................... 486
26.8.1 Supposition interpretation ............................................................... 487
26.8.2 Irresultative interpretation ............................................................... 488
26.8.3 Frustrative interpretation ................................................................. 489
26.8.4 Implicative interpreation ................................................................. 490
26.9 THE SPECULATIVE ENCLITIC <=khon> ................................................. 491
CHAPTER 27 DATIVE- AND LOCATIVE-MARKED CLAUSES ........... 494
27.1 DATIVE MARKING ON INFLECTED PREDICATES ..................................... 495
27.1.1 Reason clauses ................................................................................ 495
27.1.2 The standard of comparison clause .................................................. 499
27.2 DATIVE MARKING ON VERBAL ROOTS OR STEMS .................................. 501
27.2.1 Dative-marked complement clauses ................................................ 501
27.2.2 Dative-marked subject complement clauses .................................... 505
27.2.3 Purpose adjunct clauses .................................................................. 506
27.3 DATIVE-MARKED CLAUSES AS COMPLEMENT OF POSTPOSITION ............ 509
27.4 SUMMARY OF DATIVE-MARKED CLAUSES ............................................ 509
27.5 LOCATIVE-MARKED CLAUSES .............................................................. 510
27.6 THE CONCOMITANT ACTION SUFFIX ..................................................... 516
27.6.1 Temporal Location adjunct clauses ................................................. 517
27.6.2 Temporal attributive clauses ........................................................... 520
CHAPTER 28 ADVERBIAL AND SEQUENTIAL CLAUSES .................. 524
28.1 ADVERBIAL CLAUSES .......................................................................... 524
28.2 SEQUENTIAL CLAUSES ........................................................................ 527
CHAPTER 29 ATTRIBUTIVE CLAUSES .................................................. 534
29.1 TERMINOLOGICAL PRELIMINARIES ...................................................... 534
29.2 NO COMMON ARGUMENT .................................................................... 540
29.3 NO GAPPING AND NO OBLIGATORY SEMANTIC RELATIONSHIP ............... 548
29.4 PRE- AND POST-HEAD ATTRIBUTIVE CLAUSES ...................................... 550
29.5 ARCH NPS WITH POST-HEAD ATTRIBUTIVE CLAUSES ............................ 551
29.6 GENITIVE-MARKED A ARGUMENT OR POSSESSOR? / NO INTERNAL HEAD
.......................................................................................................... 554
TABLE OF CONTENTS xv
29.7 VARIATION CONSTRAINTS IN THE POSITION OF THE ATTRIBUTIVE CLAUSE
.......................................................................................................... 558
29.8 ATTESTED ATTRIBUTIVISATIONS ......................................................... 561
29.9 ARCH NPS AS PREDICATES OF VERBLESS CLAUSES ............................... 563
29.10 HEADLESS ARCH NPS ......................................................................... 565
29.11 LEXICALISATIONS ............................................................................... 567
29.12 THE MORPHEME <-gaba ~ -ga> AS ATTRIBUTIVE SUFFIX ...................... 568
i Numerals ..................................................................................... 569
ii The bound interrogative formative ............................................... 569
iii The time word daka ................................................................... 570
29.13 THE NOMINALISATION ~ RELATIVISATION ~ GENITIVISATION SYNCRETISM
.......................................................................................................... 570
29.14 CONCLUSION ...................................................................................... 571
APPENDIX 1 TEXTS ....................................................................................... 574
TEXT 1 Saduthamaran mo'tham PART 1 ..................................................... 574
TEXT 2 Saduthamaran mo'tham PART 2 ..................................................... 580
TEXT 3 Way khuruta ...................................................................................... 589
TEXT 4 Ca'masami way ............................................................................... 596
TEXT 5 Alsia Raja .......................................................................................... 599
APPENDIX 2 ATONG-ENGLISH DICTIONARY ........................................ 616
REFERENCES .................................................................................................... 712

xvi
List of tables
_____________________________________________________________________

Table 1 List of pairs of last names that represent the same blood lineage for Garo
and Atong speakers but that have a different pronunciation in both
languages. .............................................................................................. 9
Table 2 The relationship between the phonemes of Atong and the way they are
written in the orthography developed for the language. ........................ 22
Table 3 Illustration of the variation in lexemes and grammatical morphemes in
the dialects of Badri and Sijyw............................................................. 23
Table 4 The classification of Boro-Garo languages according to Jacquesson
(2006: 294) including Atong, until now correctly suspected to be closest
to Boro (see Jacquesson (2006: 293, quoted above). ............................ 30
Table 5 The reflexes of Proto-Boro-Garo */kr, gr, kl/ in Garo, Rabha, Boro,
according to Jacquesson (2006: 285) with the addition of Atong. ......... 32
Table 6 The reflexes of Proto-Boro-Garo */r/ and, */l/ in Boro-Garo languages,
according to Jacquesson (2006: 285) with the addition of Atong. ......... 33
Table 7 List of texts collected during fieldwork ................................................ 40
Table 8 Atong consonant inventory .................................................................. 44
Table 9 Evidence for aspiration and voicing opposition in stops ....................... 45
Table 10 Evidence for the phonemic contrast of the two fricatives /s/ and /h/ ..... 46
Table 11 Evidence of the phonemic contrasts of the nasal continuants ................ 47
Table 12 The possible combinations of vowels plus glide in Atong .................... 48
Table 13 Syllable final consonants ..................................................................... 49
Table 14 Vowels ................................................................................................ 51
Table 15 Evidence for vowel quality contrast ..................................................... 52
Table 16 Minimal pairs of syllables with and without glottal stop ....................... 67
Table 17 Loanvowels ......................................................................................... 79
Table 18 Minimal and near-minimal pairs of words with and without loanvowels
............................................................................................................ 79
Table 19 List of word classes ............................................................................. 83
Table 20 Some salient general tendencies of verbs, Type 1 and Type 2 adjectives
and nouns ............................................................................................ 84
LIST OF TABLES xvii
Table 21 Types of verbs ..................................................................................... 87
Table 22 Verbs denoting natural phenomena and their corresponding nouns ...... 98
Table 23 List of Primary-B and Secondary verbs (not exhaustive) .................... 100
Table 24 Phasal verbs ....................................................................................... 101
Table 25 Transitive and intransitive verb pairs ................................................. 102
Table 26 List of adjectives sorted by semantic category and class ..................... 107
Table 27 Gender sensitive nouns ...................................................................... 119
Table 28 Nouns occurring with the morpheme <-gaba ~ -ga> (RELATIONAL) ... 126
Table 29 Kinship terms Type 1: (a) Consanguineal kinship terms ..................... 130
Table 30 Type 2 kinship terms, consanguineal and affinal ................................ 132
Table 31 My blood relations. ............................................................................ 135
Table 32 Spouses of aunts, uncles and siblings, their children and grand children
and their relation to me ...................................................................... 136
Table 33 Reference terms uncles and aunts use for me. .................................... 137
Table 34 My in-laws, me being masculine. ....................................................... 138
Table 35 My in-laws, me being feminine. ......................................................... 138
Table 36 Address terms that my in-laws use for me and my siblings................. 139
Table 37 Address terms that my brothers- and sisters-in-law use for me. .......... 139
Table 38 List of interrogatives .......................................................................... 154
Table 39 List of indefinite proforms ................................................................. 166
Table 40 Counting in Atong. ............................................................................ 175
Table 41 Morphemes participating in the formation of Unit numerals .............. 180
Table 42 Round-Number numerals ................................................................... 182
Table 43 English numerals borrowed into Atong .............................................. 187
Table 44 Numerals borrowed into Atong from Hindi ........................................ 188
Table 45 Examples of loans from English (probably through an Indic language)
and Indic languages with their classifiers. This is not an exhaustive list.
.......................................................................................................... 190
Table 46 The categorisation of Atong classifiers .............................................. 208
Table 47 List of classifiers ............................................................................... 218
Table 48 List of time words .............................................................................. 234
Table 49 List of adverbs and intensifiers .......................................................... 243
Table 50 List of discourse connectives and their historical make up ................. 247
LIST OF TABLES xviii
Table 51 Personal pronouns.............................................................................. 263
Table 52 List of proclauses ............................................................................... 269
Table 53 List of interjections ............................................................................ 274
Table 54 Nouns that also occur as verbal predicate heads. ................................ 277
Table 55 The properties of denominal adverbs compared to those of adverbs and
nouns ................................................................................................. 285
Table 56 Overview of NP enclitics ................................................................... 290
Table 57 Marked and unmarked syntactic and semantic argument types ........... 317
Table 58 The Atong case markers and the types of NPs they can mark. ............ 320
Table 59 Pragmatic conditions for accusative case-marking of O ..................... 339
Table 60 The grammaticalisation path of the case marker <=tokoy> (VIA/LIKE)
.......................................................................................................... 353
Table 61 What cases are found repeated and why. ............................................ 359
Table 62 Properties of different types of predicate depending on the head ........ 373
Table 63 Predicate head suffixes in their respective slots. ................................. 374
Table 64 Prototypically associated nouns with their verbs. ............................... 386
Table 65 Elements incorporated into predicates with the support verb
ra'- to take ...................................................................................... 393
Table 66 Suffixes and clausal enclitics indicating a modality ............................ 427
Table 67 The functions of the factitive suffix <-wa> (FACT) ........................... 446
Table 68 Event specifiers ................................................................................. 453
Table 69 Clause types in Atong ........................................................................ 458
Table 70 Clausal enclitics ................................................................................. 459
Table 71 General properties of independent and dependent clauses .................. 461
Table 72 The structure of the fully inflected imperative predicate head ............ 467
Table 73 The structure of the prohibitive with <=bay> ..................................... 473
Table 74 The effects of the dative case enclitic on clauses ................................ 510
Table 75 Examples of lexicalised attributivised verbs ....................................... 568
written in the orthography. ................................................................. 616

xix
List of Maps
_____________________________________________________________________

Map 1 The location of the state of Meghalaya within India ............................... 4
Map 2 The Atong language area within Meghalaya .......................................... 4
Map 3 The Atong speaking area in South Garo Hills District ............................ 4

List of figures
_____________________________________________________________________

Figure 1 Schematic chart of Sino-Tibetan Groups, from Benedict (1972: 6) ....... 26
Figure 2 Relationships among the Bodo-Konyak-Jinghpaw languages, from
Burling (2003 a: 175) ........................................................................... 27
Figure 3 The classification of Boro-Garo languages according to Franois
Jacquesson (2006: p. 293) .................................................................... 28
Figure 4 The first few lines of the Parable of the Prodigal Son by the Rev. E.G.
Philips, taken from Grierson (1901: 86). .............................................. 35

xx
Summary
_____________________________________________________________________

This thesis is a grammar of Atong, a Tibeto-Burman language spoken in the South
Garo Hills district of Meghalaya State in Northeast India. The grammar is based
primarily on data collected during a total of twelve months of fieldwork, spread out
over two trips, between 2005 and 2007, in the villages of Badri Maidugytym and Siju.
I will summarise a few important typological features of the language. Atong is an
analytic and mildly polysynthetic language with suffixes and phrasal and clausal
enclitics. There are no prefixes or proclitics. Several phrasal enclitics can also
function as clausal enclitics. Constituent order in a clause is pragmatically determined.
Atong makes extensive use of zero anaphora, i.e. referents that are retrievable from
the context (co-textual as well as real-world context) are usually omitted; no NP has
to be obligatorily expressed in any clause. Semantic role marking of NPs is
determined by different pragmatic and lexical factors. Clauses with multiple
unmarked NPs are common and their semantic role has to be inferred from the
context. Boundaries between word classes in Atong can be fuzzy, as not only verbs,
but also nouns and members of several other word classes can function as predicate
head.
As for the phonology, Atong does not have phonological tone, but exhibits
glottalisation, a prosodic feature that operates on the level of the syllable. Depending
on the syllable type, glottalisation manifests itself in different ways, all of which
involve the occurrence of a glottal stop in the phonetic realisation of the syllable.
Moreover, Atong does not allow any consonant clusters in word-initial syllables,
except in loanwords. In non-initial syllables, only clusters with /r/ as second member
are allowed, but a schwa can always be inserted in these cases, as is frequently done
by native speakers.
The grammar consists of 29 chapters. The first is a general introduction to the
Atong language, its speakers and also discusses its possible historical affiliations.
Chapter 2 describes the phonology and discusses the phenomenon of glottalisation.
Chapter 3 gives an overview of the word classes. The different word classes are
discussed in detail in Chapters 4 to 17. Chapter 18 describes the different types of
word-class-changing derivation. Chapters 19 and 20 describe the phrasal morphology
SUMMARY xxi
of the language. Chapter 21 treats the subject of transitivity, which plays only a minor
role in the language. Chapters 22 to 25 describe the predicate and predicate
morphology. Chapter 26 gives an overview of the different clause types, some of
which are treated in more detail in that chapter, while others are treated in chapters 27
to 29.
There are two appendices to this thesis. The first appendix contains five Atong
texts of different genres. Four of them are fully glossed and translated, one serves as
an example of the seemingly dying practice of spirit incantation and cannot be
translated. The second appendix is an Atong-English dictionary. The Atong entries
and examples in this dictionary are written in the orthography I designed for the
language.
xxii
xxiii
Statement of authorship
_____________________________________________________________________

Except where referenced in the text of the thesis, this thesis contains no material
published elsewhere or extracted in whole or in part from a thesis submitted for the
award of any other degree or diploma.
No other persons work has been used without due acknowledgement in the main
text of the thesis.
The thesis has not been submitted for the award of any degree or diploma in any
other tertiary institution.
The author, Jonkheer Egbert Joost Seino Clifford Kocq van Breugel, has
published and will publish under the name Seino van Breugel. All references to Seino
van Breugel in this thesis refer to the author of this thesis.

__________________________________________________
Jonkheer Egbert Joost Seino Clifford Kocq van Breugel, M.A.
xxiv
xxv
Acknowledgments
_____________________________________________________________________

First and foremost I want to thank all the Atong people who helped and supported me
during my fieldwork. Thank you for your trust, your good care, your enthusiasm and
your patience with me while I was learning your language and culture. Thank you for
speaking with me and letting me record your beautiful language. I will only have good
memories of all of you.
I am particularly thankful for the hospitality of the people of the villages of Badri
Maidugytym and Siju. The family of Susil S Marak (Nisawa) and his wife Kelbish M
Sangma (Nisajyw) took me into their house during the entire stay in Badri
Maidugytym. Latith and Janita M Sangma and their family took care of me like one of
their own family. Dalcheng M Sangma taught me how to drive my motorbike, which
was essential for my work in the field. Plindar R Marak negotiated my stay in Siju.
While I stayed in Siju I found good homes with the following very kind and
hospitable families: Peslar R Marak, his wife Golaphy R Sangma and their family;
Elsina R Sangma (AB Fernandajyw), her mother Monjila, her brother Dilseng and
their family; Kalison R Sangma (Dambewa), his wife Jontol D Sira (Dambejyw),
their daughter Radia, Bairik D Sira and their family. Thank you all from the bottom of
my heart for your hospitality and your good care. I will never forget your kindness.
I want to thank all the people who taught me Atong and helped me with the
transcription and translation of their language. There are some people to whom I am
particularly grateful: Salseng R Sangma, who was my main and best translator, who I
could always rely on and who explained the most difficult aspects of his language to
me with great patience. He was always there for me when I needed him. Ranus M
Sangma from Badri, who was my first Atong language teacher who helped me to
understand the people when I first arrived in Badri and did not speak a word of Atong
yet. He also helped me to write down and translate the first stories I recorded. Sandish
M Sangma, who transcribed many many many texts for me with great skill, dedication
and remarkable perseverance. In addition to his writing, he showed me the beautiful
jungle around Badri and taught me how to fish. Samrat N Marak from Siju, who was
by my side most of the time while I was in Siju. He was a good friend, a good teacher
and a great guide in the jungle. I could not have coped so well without
ACKNOWLEDGEMENTS xxvi
him. Inden R Sangma of Siju, who learned how to write his language with surprising
ease and who never stopped writing until all texts were completely finished. Shyam R
Marak and Plindar R Marak, who also helped me a great deal with translation.
Nikseng S Marak, who not only helped me translate one of the most lengthy and
difficult texts, but also corrected and improved the drafts of the dictionary and the
story book.
When I first arrived in Tura in 2004 on a pilot project funded by the Leiden
University Fund, it was Father V.A. Cyriac, principal of Don Bosco College, who got
me started on my journey of discovery by introducing me to my first guide to the
Atong area, Sanggra A Sangma. I thank Fr. Cyriac for his support throughout my
fieldwork and for comforting my family back home when they were worried about
me. My gratitude also goes to Dr. Caroline R. Marak, former head of the Garo
Department of the Tura branch of the North Eastern Hill University (NEHU), who
greatly facilitated my first stay in Tura by setting my up in the NEHU guest house.
When I came back to Tura for my first fieldwork trip for RCLT in 2005, Dr. Caroline
R. Marak introduced me to her brother S.R Marak (Chaki) whose idea it was to let me
stay in Badri Maidugytym and who introduced me to my aforementioned host Susil S
Marak. During my first journey from Tura to Badri, it was the family of Kroshnil D
Sangma (Winchipa) in Williamnagar, who was so kind to give me shelter during the
incessant rain that prevented me from travelling for days.
I thank the many people whose company and hospitality I enjoyed during my
recurrent trips to Tura during my fieldwork. My special thanks go out to Mobbin J
Sangma, my first Garo friend, his sister Barul, cousin Ditu and brother Savior, who
took great care of me and made me feel at home in both their student houses. I thank
the family of Teroth Ch Momin and the family of Barul and Tengcheng: Shyangtho D
Sangma (Nikchengpa) and Medhina Ch Momim (Nikchengma) for their hospitality.
Thanks also to the family of Mr. Alphonse A Sangma their hospitality and care during
my last stay in Tura when my hand was injured (and Pijas foot).
I want to thank Madhumita Barbora, her brother Rohit Barbora (Raja) and their
family very much for their help and support, hospitality and friendship, which made
my trips to Assam so enjoyable. I am also very grateful to Tondra Barbora (Bunbun)
and her late husband Lanu for their help and hospitality.
ACKNOWLEDGEMENTS xxvii
I am particularly grateful to Sasha Aikhenvald and Bob Dixon for having me as a
PhD student at the RCLT, which is a truly inspiring place to learn about writing a
grammar. Sasha Aikhenvald, my main supervisor, read through two or three drafts of
most chapters of my thesis and gave many insightful comments. Her extraordinary
knowledge of linguistic literature, linguistic typology and her ideas on grammatical
analysis profoundly influenced my thesis. The many in-depth talks, seminars and
workshops organised at the RCLT were a constant source of inspiration throughout
my PhD. Participating in these events gave me the opportunity to practice my
presentational skills and to disseminate the discoveries I made while writing my thesis
to specialised linguistic audiences, whose feedback always broadened my knowledge
and understanding and helped me to improve my writings.
I would like to thank Randy LaPolla very much for his support during the final
part of my candidature. He commented on several drafts of chapters and read through
the entire thesis before I submitted it, despite his ever busy schedule. His comments
were constructive and illuminating. I could not have benefitted more from his vast
knowledge on Sino-Tibetan/Tibeto-Burman languages and linguistics.
Siew-Peng Condon truly is an organisational and logistic miracle without whose
help and support nothing would have been possible. I thank her very much for
everything she has done for me, especially for her efforts to provide me with a place
to stay on campus throughout my PhD. I would also like to thank May Tan for her
help while Siew-Peng was on maternity leave.
All members of RCLT were sources of inspiration and information. Gerd
Jendraschek was a great sounding board for my ideas throughout my PhD. He was
always ready to discuss theoretical approaches and insights. I benefited immensely
from his wide linguistic and knowledge in combination with an open mind and
fabulous analytical ability. Sheena Van Der Mark, besides offering lots of moral
support, helped me enormously by teaching me how to deal with the technical side of
linguistics, i.e. digitalising my recordings, transcribing them, the make up of a Word
document etc. I benefitted greatly from the linguistic expertise and swift reading skills
of Stephen Morey. Special thanks also to Rosemary Beam de Azcona, Birgit Hellwig,
Rene Lambert-Bretire, Simon Overall, Roger Wales, Alec Coupe, Cindy Schneider,
Rik de Busser, Jingyi Du and Roberto Zariquiey.
ACKNOWLEDGEMENTS xxviii
Certain linguists outside of the RCLT also supported me. I would like to thank
Robins Burling for his comments on Chapter 2. Thanks also to George van Driem,
whose support helped me to start my PhD.
Thanks to my loving mom, who was supportive and very courageous during my
trips to India, made many beautiful books for me and sent me Sinterklaas packages
every year. Thanks to my darling sister Claartje, who did not forget me while I was
far away for a long time. Thanks to Tin Lee for being such a good friend.
xxix
Dakanggaba katha
_____________________________________________________________________

Badri khuchuksang
Ian dakanggaba Atongkhuchukmyng gremyr. Dakangdo Atongkhuchuk saina
manchachym, tedo diksyneriba golpholekhaba ganangok. Ie golpholekhaba
diksyneriba panga morotdyrangna taksakwamyng dongni noai ang
khadonga. Ytykyimyngdo ie lekhaaw saina taksakwamyng gymyn ang
Atongmorotdyrangaw ang bajudyrangaw dyngthangmancha mythela.
Nangtym angna Atongkhuchuk nemai sykiok. Ytykyimyng saigabasangba
atongba sangwalwamyngba katha perengchagabadarang ganangchido, angaw
khema khaphabo. Ang ie khuchukaw bibyrokhon saina raiathirini. Umyng
gesepchian ang nangtymna khapakni. Anga nangtymaw sung rakhamni.
Ytykyimyng Atongmorotdyrangaw sangwalchawa.
Mythela,

Seino

Sijyw khuchuksang
Ian dakanggaba Atongkhuchukmi gremyr. Dakangdo Atongkhuchuk saina
manchachym, teewdo diksyneriba golpholekhaba ganangok. Ie
golpholekhaba diksyneriba panga morotdyrangna taksakwami dongni noai
ang khadonga. Ytykyimudo ie lekhaaw saina taksakwami gymyn ang
Atongmorotdyrangaw ang bajudyrangaw dyngthangmancha mythela.
Nangtym angna Atongkhuchuk nemai sykiok. Ytykyimu saigabasangba
atongba awanwamiba katha perengchagabadarang ganangchido, angaw khema
khaphabo. Ang ie khuchukaw bibyrokhon saina raiathirini. Umi gesepchian
ang nangtymna khapakni. Anga nangtymaw sung rakhamni. Ytykyimu
Atongmorotdyrangaw awanchawa.
Mythela,

Seino
DAKANGGABA KATHA xxx
Foreword in English
This is the first grammar of the Atong language. Before, it was not possible to write
Atong, but this has now changed. Now there is a dictionary and a story book. I hope
that the dictionary and the story book will be useful for many people. I thank the
Atong people and especially my friends for helping me to write these books. You
have taught me your language well. Please forgive me for any mistakes and
misconceptions in these writings. I will come back one day to study your language
again. In the mean time I will miss you. I will keep remembering you. I will not forget
the Atong people.
Thank you,

Seino
xxxi
List of abbreviations and symbols
_____________________________________________________________________

Abbreviations

&co associative
A transitive subject
AC attributive clause
ACC accusative
ADD additive and, also
ADV adverbial
ALL allative
ALT alternative
ATTR attributive
CAUS causative
CC copula complement
CLF classifier
COM comitative
CONF confirmative
CONJ conjunctive
COS change of state
CP comparative
CS copula subject
CUST customary aspect
DAT dative
DCL declarative
DESI desiderative
DIS distributive
DLIM delimitative
DREF definite &
referential
DREL derelational
DST distal demonstrative
DUR durative
e exclusive
E third argument of an
extended transitive
verb
EMPH emphatic
FACT factitive, reification
FC focus marker
FC/ID focus/identifier
FUT future
GEN genitive
i inclusive
IE.be identity/equation
copula
IFT imperious future
IMP imperative
IMPEMPH imperative
emphasiser
INCEPT inceptive
INCOM incompletive
INDEF indefinite
INSTR instrumental
interj interjection
IRR irrealis
LIKE similative
Lit. literally
LOC locative
MIR mirative
MOB mobilitative
Name persons name
NEG negative
NP noun phrase
NR nominaliser
O transitive object
p plural
1

PARTRED partial reduplication
pe plural exclusive
pi plural inclusive
Pname place name
POS emphatic positive
ppp personal pronoun
plural
PRIV privative
PROG progressive
PROH prohibitive
PRX proximate
demonstrative
PUR purposive
Q interrogative suffix

1
Also as superscript to disambiguate English
glosses.
LIST OF ABBREVIATIONS AND SYMBOLS xxxii
QF interrogative
formative
QUOT quotative
RC reciprocal
REF referential suffix
REM remote
demonstrative
REMEMPH emphatic remote
demonstrative
Rname river name
S intransitive subject
s singular
SEQ sequential
SIMP simplicitive
Sname1 first surname
Sname2 second surname
SPEC speculative modality
TAG affirmation seeking
marker
TOP topic
VIA perlative
WHILE concomitant action
XS excessive

Symbols

- morpheme boundary
= enclitic boundary
+ morpheme boundary in compound
// phoneme
<> morpheme
{} predicate boundaries
[] In Atong example sentences the square brackets indicate phrases.
In the translation or gloss square brackets indicate that the words
inside must be inferred from the context.
|| clause boundaries
< comes from
* this form is ungrammatical or reconstructed
1 first person
2 second person
3 third person

1
Chapter 1 The Atong language and its
speakers
_____________________________________________________________________

In the most beautiful part of the jungle-clad South Garo Hills, around the Mountain of
the Great Spirit and in the stream area of the scenic Symsang river live the Atong
people, who speak a language with the same name. The literature provides us with
different ways of spelling this name and there is no formal spelling for most of their
villages, as we will see in section 1.2. Section 1.3 will reveal the ethnic affiliation of
the Atongs, and relates about their way of life as I observed it during my fieldwork,
complemented by citations and references from the relevant literature. Where they
live and how many of them there are will be pointed out and commented upon in
section 1.1. Atong speakers are not the only inhabitants of the South Garo Hills. The
linguistic environment in which the language is spoken is one of the topics of section
1.4, where language status and use will also be discussed. Contrary to prior beliefs of
the speakers themselves, the Atong language can be written, and even has an
orthography especially designed for the language by the author of this grammar.
Section 1.5 introduces and explains the orthography to the reader. There is not one
way to speak Atong but several. The phenomenon of dialectal variation is treated in
section 1.6. Section 1.7 on linguistic affiliation contains evidence for the idea that
Atong is a Central Boro-Garo language, more closely related to Boro than to Rabha
and Koch. However, before the reader will be presented with this evidence, (s)he will
get an overview of different genetic classifications of Atong in the literature, from the
earliest sources to the most recent ones. An overview of the research that has been
done on Atong in the past is given in section 1.8. Finally, section 1.9 gives the bare
facts about the fieldwork that was conducted to write this grammar.
1.1 Location of the language and number of speakers
Atong is a Tibeto-Burman language spoken in the stream area of the Symsang
[semsa] river in the south Garo Hills District of Meghalaya in Northeast India, and in
adjacent areas in the West Khasi Hills and, according to my Atong friends, the
Mymensingh district of Bangladesh. Map 1 shows the location of Meghalaya within
1 THE ATONG LANGUAGE AND ITS SPEAKERS 2
India. Map 2 depicts the area where Atong is spoken within Meghalaya and Map 3 is
a close-up of the language area in which Nangwalbibra (locally called Nongal [n:al])
demarcates the northern border of the language area and Baghmara the southern one.
In Nongal and Jadi [dzadi], Atong, Garo and Indic speakers live side by side. These
places are market towns where different tradespeople from all over North India have
settled to set up shop. The main languages spoken in these places are Garo, Atong,
Bengali, Punjabi and Nepali; there are also a few settlers from Rajasthan and Bihar.
Code switching and mixing is commonplace here. Atong people from the villages
along the main road up to Badri frequently come to Jadi and Nongal to shop and
work.
Between Jadi, Raiwak, Rongsu and Waimong Mountain lies the heart of the
Atong speaking area. Between Raiwak and Baghmara lie many villages where Garo is
spoken and a few where Atong is spoken alongside Garo. As one approaches
Baghmara the number of Atong speakers dwindles considerably. In Baghmara there
are many Atong speakers, but they seldom or never speak their language outside their
homes. Many people in Baghmara were Atong speakers in their childhood but have
now switched to Garo completely.
There are a few Atong villages in the West Khasi Hills district of Meghalaya
which are not on the map due to lack of information about their exact location.
According to some of my consultants, there are many Atong people living on the
Bangladeshi side of the border, south and southeast of Baghmara. Some say that these
Atong still speak Atong, while others claim that they are now speaking Bengali or
Garo. I was not able to travel to Bangladesh and verify the existence of an Atong
language community there. Exploring the Bangladeshi side of the Meghalaya-
Bangladesh border will be a matter for future fieldwork investigations.
Many Atong have migrated to Tura, the capital of the Garo Hills (see Map 2), in
search of work or for their education. There are at least several hundred Atong
speakers in the city, many of which are still in regular contact with family members in
the Atong speaking area in the South Garo Hills. Atong people in Tura do not speak
their language in public, but use Garo instead. Atong is spoken in the home, amongst
family or friends, when there are no Garo speakers present. There are also many
people, sometimes whole families, of Atong background who have completely given
up their language and now speak only Garo. This language shift takes place under
pressure of Garo as prestige language in the region.
There is no official account of the number of Atong speakers. Grierson (1902: 85)
mentions some fifteen thousand Atong speakers, while Van Driem (2001: 541) speaks
about a few thousand. Even after twelve months of fieldwork in the area, it was
impossible to estimate the number of speakers. New Atong-speaking villages are
reportedly still being built, while at the same time Garo, a closely related and
regionally important language, encroaches rapidly on a lot of existing Atong-speaking
villages.
Map 1 The location of the state of Meghalaya within India

Map 2 The Atong language area within Meghalaya

Map 3 The Atong speaking area in South Garo Hills District

The boundaries on maps 1, 2, and 3 are not necessarily authoritative.
1.2 Names and allonyms
1.2.1 Language names
The Atongs call themselves Atong [at:] or Atong morot
2
[at: mor:t] Atong
person and call their language Atong [at:] or Atong khuchuk [at: k
h
utcuk]. The
origins of the name of the language are not known. There is an interrogative, or
question word, in the Atong language with the same pronunciation, viz. /ato/ [at:]
what?. Atong is also known under the following names in the literature: Grierson
(1902): tong, Kuchu, or ting. Remark: The word kuchu must be related to the
ethnonym Koch. Playfair (1975 :21) remarks: To nearly all other Garos the Atongs
are known as Kochu, though they themselves prefer the former appellation. Today,
where the Atongs are generally known as Atongs. Atong speakers refer to the Garo
language and its speakers (see 1.4) as Hachyk [ha?cek]. Jacquesson (2006), Shafer
(1953 and 1974), Benedict (1972), Burling (1959 and 1963) and Playfair (1909) write
Atong. Van Driem (2001) and Burling (2003) and Gordon (2005 a) write Atong.
Burling (2004) writes Atong, with a glottal stop represented by the raised dot . I
write the name of the language as Atong, because this orthography reflects the way
the Atong people pronounce the name of their language.
3

1.2.2 Remarks on some toponyms on Map 3
The spelling of place names (toponyms) in Map 3 is, for the most part,
unconventional. Place names in the Atong language area find their origin in different
languages, viz. Atong, Garo, Indic (Hindi, Bengali or Assamese) and English.
Baghmara, the headquarters of the South Garo Hills district, Siju and Nangwalbibra
are spellings found on maps of the region. Baghmara is also spelled Bghmra on
some maps but is pronounced [bak
`
mara] by the Atongs. The market place of
Nangwalbibra is locally known as Nongal [n:al]. According to the Atongs, Siju

2
The Atong word morot person, man is an Indic loan, cf Hindi /mard/ man.
3
There is also a Niger-Congo language in the African country of Cameroon which is called Atong. This
language also goes under the alternate name of Etoh (see Gordon (2005 b).
[sidzu] is the Garo pronunciation of the name of the village; the Atong pronunciation
is [sidzew]. The Garo pronunciation is widely used among the Atongs nowadays, and
Siju is the conventional way to write the name of this village. However, because I
think that the Atong pronunciation should not be forgotten, I give both the Garo
spelling, Siju, and the Atong spelling, Sijyw, on Map 3.
The places Badri Maidugytym [badri majdugetem], also recorded with an
aspirated /t/ as Badri Maidugythym [badri majdugethem], Badri Rongdyng Hawai
[badri rode ha?waj], and Badri Rongsa Hawai [badri rosa ha?waj] are also found
written in Garo as Badri Maidugittim, Badri Rongding Awe and Badri Rongsawe
respectively. Raiwak is spelled Rewak in Garo and there is no conventionalised
pronunciation or spelling for the place Artika, also recorded pronounced as [areteka ~
aretika ~ arektika].
Waimong mountain, Waimong habyri [wajm: ha?beri] in Atong, is the most
significant landmark in the region. Its summit is at an altitude of 1026 metres (3367
feet). As one drives from Baghmara to Siju, its flat peak changes its shape from
triangular to rectangular. A traditional story tells how a giant took the mountain out of
the ground in Balphakram (conventional spelling Balpakram) and carried it on its
back to its current location. This explains why there is such a huge, deep gorge in
Balphakram, the land of the spirits, now a national park, about ninety kilometres east
of Baghmara. In Garo the mountain is known as Chutmang [tcut
`
ma]. The Atong
name Waimong is made up of two elements, viz. <way> spirit and <mo> main,
the second of which we also see in the word so-mo (village-main) main village.
The name of the mountain could be freely translated as Mountain of the Great Spirit.
Finally, the main river in the region goes by two names, an Indic one: Someswari, and
a local one, of which the pronunciation is [semsa] in both Garo and Atong.
Although the Garo spelling, found on maps, is Simsang, I have given the spelling in
Atong, which is Symsang.
1.3 The Atong people
The finest physique is to be met with among the inhabitants of the higher ranges, and
among the Atongs of the Someswari valley. (Playfair, 1975: 2)
1.3.1 Ethnic affiliation
Although Atong and Garo are not mutually intelligible, the Atongs consider
themselves and are considered by the surrounding populations to be ethnically Garo
(see also Burling, 2004: 11, 2003 a: 176, 2003 b: 387, 1961: 80, 1963: 390-4, 1959:
437, Playfair, 1975: 62). Their cultural traits and customs, as far as I have been able to
observe, are mostly the same as those of the Garos, a fact which has also been
observed by Burling (1963: 390). Functionally equivalent cultural activities may have
different names in both languages. The biggest cultural festival, for instance, during
which the people drink, dance, sing and tell epic stories, is called Chywgyn /cowgon/
[tcewgen] in Atong but Wanggala in Garo.
Garos and Atongs intermarry and share the same surnames. A surname is made up
of two components that refer to groups of people which Burling (1963: 22-23) refers
to as sib and moiety. I will refer to the components of the surname as first and
second surname. In the surname Mongsrang Sangma, for example, the component
Mongsrang is the first surname and Sangma the second surname. The second surname
represents what Burling (1963: 22) refers to as a matrilineal descent group. There are
five of those groups within the Garo ethnic community which, ideally, should be
completely exogamous (idem). The biggest groups are Sangma [sama] and Marak
[marok]; Momyn [momen] (written Momin in Garo) is a smaller one, Sira [sira] is
very small and there are reportedly no Atong people that belong to the group called
Areng [arc]. Burling writes the following on the groups of people associated with the
first surname or sib:

Each moiety is divided into numerous named divisions which can be called
sibs. These name groups frequently include many thousands of people and
extend over too wide an area for all the members to cooperate or even know of
each others existence. Since the moieties are exogamous, no evident function
is left to the sib, though Garos do feel more strongly about a breach of sib
exogamy than about the necessity for moiety exogamy. The sib amounts to
little more than a name group, but by virtue of the name it is one of the few
kin groups with completely unambiguous membership. (1963: 22-23)

Playfair (1975: 155-6) lists 138 names of some subdivisions of the Garo exogamous
septs, i.e. first surnames for the groups Momin, Marak and Sangma, but not for Sira
and Areng. People always inherit the whole surname, i.e. both components together,
from their mother, a matter which will be discussed further below. Some of the first
surnames are different in Atong and Garo. Table 1 lists pairs of surnames that
represent the same blood lineage for Garo and Atong speakers but have different
pronunciations in the two languages.

Table 1 List of pairs of last names that represent the same blood lineage for Garo and
Atong speakers but that have a different pronunciation in both languages.
4

Atong Garo
Geneng [gcnc] Sangma Dawa [dawa] Sangma
Sinthang [sintha ~ sentha] Marak Sinthang [senthang ~ sintha] Marak
Dicham [ditcom] Sangma Chisim [tcisem ~ tcisim] Sangma
Raicyl [rajtcel] Marak Rechil [retcel ~ retcil] Marak

One of the most important cultural traits shared by speakers of Garo and Atong is
the fact that they are matrilineal and matrilocal. Matrilineal means that the blood
lineage, indicated by someones surname, is inherited through the mother. The
matrilocal aspect of the culture refers to the practice that when a man marries, he will
move to the house of his wife when she is the heiress of her family. When a man
marries a woman who is not the heiress of her family, then the young couple will have
to build a new house somewhere else in the village.

4
The Atong varieties are written in the Atong spelling as explained above, the Garo varieties are
written in their Garo spelling. Pronunciation of first surnames is given in IPA.
Of all the Tibeto-Burman speaking populations in Northeast India, only the Garo
(i.e. those belonging to the Garo ethnicity, which includes Atong speakers), Rabha
and Koch are matrilineal and matrilocal (see Jacquesson 2006: 281). Interestingly
these populations live next to the Khasis, who are also matrilineal. The Khasis, who
speak Mon-Khmer languages, are believed to be the remnant of the oldest population
in the area which lived there long before the arrival of the speakers of Tibeto-Burman
languages. Before the arrival of the Tibeto-Burmans, the Khasis must have occupied a
much larger area than today, and almost certainly formed part of one continuous Mon-
Khmer language area stretching out from Vietnam, through Southeast Asia, all the
way to the state of Orissa in India, where the Munda languages are still spoken today
(Jacquesson 2006: 279-80, see also van Driem 2001: 411-17 for a more detailed
discussion). Matrilineage of speakers of Atong, Garo, Rabha and Koch, although
unique amongst Tibeto-Burman people, might not be so strange when seen in the light
of their adjacency to the Khasis. Jacquesson (2006: 281) states that this remarkable
fact of matrilineage sexpliquerait soit par linfluence profonde et ancienne des Khasi
sur ces gens qui occupent comme eux le Meghalaya et ses abords, soit mme par leur
changement de langue, si lon fait lhypothse quune partie de ces gens ont t
ethniquement des Khasi : ils auraient abandonn leurs parlers mon-khmer dautrefois
pour passer aux parlers tibto-birmans des nouveaux venus plus influents, mais
auraient conserv une part dcisive de leur organisation sociale.
5

1.3.2 Social organisation
Despite the Atongs being matrilineal, the society is patriarchal. The men with the
most power in making decisions concerning family matters are the older brothers of a

5
English translation of quote: Jacquesson (idem) states that this remarkable fact can be explained
either by the profound and ancient influence of the Khasis on the peoples who, like them, occupied the
Meghalaya and adjacent areas, or by the fact that they swopped languages, when we hypothesise that
part of the population was once ethnically Khasi: they would have given up their Mon-Khmer
languages that they used to speak before and started speaking the Tibeto-Burman languages of the
more influential newcomers, however, they preserved an important part of their social organisation.
married woman. These older brothers are collectively known as the chara and the
eldest of a married womans brothers is called the charamong.
The society is layered according to age, generation and marital status. The
younger ones must have more respect for the elder members of society and married
couples are higher on the social hierarchy than unmarried boys and girls. The groups
of which the society consists are the children (sagyrai child), the unmarried men
(banthai bachelor) and women (nawmyl marriageable girl), the married men
(meapha married man) and women (meama married woman), elderly men (achu
grandfather) and women (awyi, abu grandmother). Although there is a specific
word for marriageable girl, viz. nawmyl, there is no specific word for marriageable
boy. When talking about marriageable boys and girls, the Atongs use the expression
bipha nawmyl, where bipha means male, man, boy and nawmyl marriageable girl.
The ways in which different members of society are addressed is treated in Chapter 1.
Like the Garos, the Atongs practice cross cousin marriage. Cross cousins are the
children of mothers brother or fathers sister (see Table 31). As was mentioned
above, a marriage between a couple with different first and second surnames is
preferred.
1.3.3 Living environment: the compound
Most Atongs live in houses made of wood and bamboo, called nok, with roofs of
corrugated iron, as reed, which is used for thatch, has become very scarce and
therefore very expensive. Only very rich people can afford to build a cement house,
called bilding (from English building). All houses are built on a piece of land that
has been made completely level, called nok+hap (house+place). Every traditional
Atong household lives in a compound consisting of at least two, but often more
structures. There is a main house in which the married couple, their small children and
sometimes the parents of the wife sleep. The main house has one big central room to
receive guests and smaller side rooms which are bedrooms. There are separate
bedrooms for the married couple, the wifes parents and the male and female children.
Many main houses are built on a wooden frame that is lifted about 1 to 1 metre
from the ground by big rocks fixed in the ground. Cooking is not done in the main
house. There is a separate kitchen house (babylsi) opposite the main house on the
compound. Meals are prepared and eaten in the kitchen. Toilets (called letrin, toilyt,
toilet, dikyntyk or paikhana) are also separated from the main house and are
constructed outside the compound.
When young men become old enough, they build separate rooms either as an
attachment to the main house, when it is not elevated, or as separate structure on the
compound. These separate structures are called nokbanthai (nok house + banthai
bachelor) bachelors house. There the young, unmarried men live until they marry
and move to their wifes house.
A little away from the compound, there may be a place where water comes out
continuously from the end of a pipe, lifted two or three feet off the ground with a stick
or pole. This is the place where people wash themselves, their clothes and the dishes.
Usually more than one household makes use of a single pipe. The pipes get their
water from the nearest river, in which they are fixed with stones. Other households
wash in a nearby river or stream.
Those rich families that live in cement houses usually have only this one building
in their compound, which contains different rooms for the functions that traditional
families spread out over different structures.
Many households keep domestic animals like dogs, chicken, pigs and sometimes
cows. These either run around freely or are kept in separate enclosures away from the
compound. Usually meat is bought at the market from butchers. Usually households
slaughter their domestic animals only on special occasions, such as weddings,
funerals, Christmas and Easter, although dogs make a tasty meal all year round, but
not everyone eats them. Sometimes domestic animals are kept to fatten or get
pregnant, after which either the pregnant animals or the offspring are sold.
Around the compound is usually a stretch of land where edible fruit trees grow,
like banana, jackfruit, coconut, betel nut, lychee, star fruit, mango and sometimes also
pineapple and useful species of bamboo. These fruits and plants are usually eaten by
the household themselves, but can also be sold, e.g. betel nut is a big source of income
for many households.
1.3.4 Living environment: the jungle
The jungle of the South Garo hills is thick and quasi-impenetrable, with many steep
slopes and many streams that flow in beds invariably sewn with rocks. Though much
of the jungle has been cut at some time or other to give way to rice fields, there are
probably still many untouched areas. These areas contain wild animals that can be
hunted for food, and the Atongs eat every one of them, and some plants that are used
as medicine or food, e.g. bamboo shoots. The jungle also provides timber to build
houses. Almost all edible fruits, on the other hand, are grown in the village in peoples
gardens or on plantations, as are all the useful bamboo species, i.e. those used for the
construction of houses and the making of baskets and other artefacts.
The most dangerous animals in the jungle are the elephants, which abound in the
South Garo Hills. Rice fields but also villages can be destroyed by these powerful
animals that inspire such awe into the hearts of the Atong that they cannot call the
animal by its name, mungma or mongma, while walking in the jungle. Instead they
call the animal achu, which means grandfather, so as to not attract its wrath. Other
animals that we can find in the jungle are squirrels, several species of deer,
porcupines, wild pigs, gibbons and macaques, many species of snake and wild cats.
Wild birds, as far as I can tell, are not eaten, but young boys love to shoot them with
their slingshots.
The scenery in many parts of the hills is very fine, the finest being that on the
Someswari river, which flows through a very narrow valley between high and
precipitous hills. (Playfair, 1975: 6) There is an incredible multitude of streams and
rivers in the Garo Hills, which is not surprising since it is one of the areas on earth
with the most rainfall per year, all of which falls within several months during the
monsoon, which begins in June and ends mid-October. There might be some
occasional light rain in March, April and May. During the dry season, from mid-
October until the end of February, not a drop of rain falls from the sky. Every village
is close to one or several streams or has streams running through it. The rivers and
streams are the source for drinking water and provide the Atongs with snails, shrimp,
fish, eels, frogs and crabs to eat and a place to wash.
1.3.5 Ceremonies and festivals
Unfortunately I did not witness any traditional religious ceremonies during my
fieldwork. This is probably due to the fact that most of the Atong in the villages where
I conducted my fieldwork are Christians. The Atong probably abandoned their old
ceremonial practices after conversion. It might also be that they have been able to hide
them from me very successfully, but this I cannot know. On a trip to the village of
Dajong, I once saw a non-Christian religious bamboo object, which indicates that the
non-Christian religious practices are still surviving. Unfortunately nobody wanted to
explain to me what the altar was for. The conversion of the Atongs began between
forty and fifty years ago. Old people can still remember how life was before the
conversion, but are reluctant to talk about it.
Before the advent of Christianity, the Atong believed in gods, myte [mete], spirits,
wai [waj], and ghosts memang [me
?
ma] or mimang [mi
?
ma]. The word for
thunder, goira [gjra], for example, comes from the name of the god of thunder,
Goira. The supreme God was Babyra [babera], and there were the gods of the sun,
Saljong [saldz:], and the moon, of which I did not record the name, and maybe
many others whose names are not mentioned any more or are forgotten, at least in the
villages where I conducted my fieldwork. There were priests, kamal in Atong, in each
village, who performed incantations to summon spirits and cure the sick. How this is
done is told in Text 3, which was told by a man who had been a kamal before he
converted to Christianity a few years ago. I recorded from him part of the incantation
of the chamasangmi wai [ca
?
masami waj] downstream spirit, which is
untranslatable according to my Atong friends. This incantation is presented as Text 4.
I was not able to find any source that relates to the spiritual life of the Atong speaking
population in pre-Christian times. To get an idea of what pre-Christian spiritual life in
a Garo village must have been, I refer the reader to Burling (1963: 54 ff), who
describes it in much detail.
The wedding and funeral ceremonies I witnessed were conducted much like in
Western countries, and were presided over by members of the Christian church.
Funeral ceremonies retain a very pleasant aspect of the old days: after the death of the
person, a wake is held at the house of the family of the deceased which lasts two days
and one night, during which people come and go. The visitors eat, drink and smoke
and play cards to distract the family of the diseased and make them happy. Marriage
and funeral ceremonies are important family gatherings at which the boys and girls
get to know their marriageable cousins (see Chapter 1).
I witnessed the new rice festival, called maidan syla toka [majdan selaj t:ka]
(rice+new beautiful beat) in the village of Badri Maidugytym. This festival is
celebrated at the end of October or the beginning of November. The richest men in the
village slaughtered a cow, some pigs and a lot of chickens to be cooked and eaten, and
invited the neighbouring villagers to come and join in the celebration, which lasted
two days. The whole village helped in preparing the food.
The village of Siju has the tradition to organise the yearly waribula festival on the
Symsang river and its banks at Dabatwari, the place on the Symsang river where the
river from the Bat Cave (Tawpakkhal [tawpakhal]) comes out into the Symsang, and
where Siju has its origins as a village. During the festival people try to catch as many
big fish as possible while others compete in wrestling. The festival takes place in
January or February on a day that the weather is favourable. Unfortunately, due to
pollution of the Symsang river, the catch gets smaller and more disappointing every
year.
The festival of chywgyn [cewgen], equivalent to the wanggala festival of the
Garos, is not celebrated any more in the villages where I did my fieldwork, since it is
considered a heathen festival. It is, however, still held in other villages, but
unfortunately, I was unable to attend the festival. According to my Atong friends,
during chywgyn people drink, dance, sing and tell epic stories. The festival lasts for
more than a week, it is said. Chywgyn is celebrated after the harvest, between
November and January. Each village has its own festival on a different date. For a
description of the wanggala festival of the Garos, see Burling (1963: 63 ff).
1.3.6 Contact with others
The Atong people are a very mobile lot. People frequently travel for miles on foot or
by motorised transport to visit friends, relatives or to go to the market or to school, or
to play sport competitions in other villages etc. This means that people from different
places are in frequent contact with each other and news travels fast. There are a few
bus services that connect Bari and Siju, as well as the other villages along the main
road, to Williamnagar, headquarters of the East Garo Hills district, Baghmara,
headquarters of the South Garo Hills district, and Tura, the headquarters of the West
Garo Hills district. People also travel by private, shared taxis and in the dry season
there are many coal and stone trucks on which young people can hitch a ride. Only
some rich people have their own car or motorbike.
Some Atongs have a radio, although, apart from an English language station
broadcasting from Shillong and a few stations that broadcast in Indic languages, there
is little that can interest them, as Russian and Chinese speaking stations fill the rest of
the ether. Some Atong people, even in remote villages, have a TV with a satellite dish
and can watch not only the news in Garo and several other Indian languages, but also
CNN and BBC World, and much more. They also have the opportunity, for instance,
to take notice of modern Western trends on house decoration and holidays, but this is
usually not a favoured way to pass their time in front of the TV. Many Atong that
have a TV dont have a satellite dish but do have a VCD player. The most favoured
thing everybody likes to watch are Hollywood action movies, Christian movies and
Bollywood movies, which can be bought on the market.
1.3.7 Economy
As far as their economic situation is concerned, until today most of the Atongs
practice the traditional slash-and-burn agriculture. They cut part of the jungle on the
slope of a hill, and use it for one year to cultivate dry rice and a variety of vegetables
and tubers. Very few places are suitable for the cultivation of wet rice. Although the
clearing of the jungle is an activity organised by almost the whole village, the clearing
is divided into several plots, one for each household. The borders between the plots
are indicated by partly burned, black tree branches. Each household has its own rice
field house where those who work in the field can eat and sleep. In the villages where
I stayed during my fieldwork, in the Badri area and Sijyw, this form of cultivation is
never enough to support a whole village, or even one household for the course of a
year, so that rice grown elsewhere in India and other food supplies have to be bought
on the market.
The Atong living in the Badri area, Badri Maidugytym, Badri Rongdyng Hawai,
earn most of their money from the exploitation of their coal mines. These, however,
are only operable during the dry season. During the rainy season money and work are
scarce for most of the male part of the population. On the rice fields, though, work is
abundant during the monsoon. While clearing and burning the jungle and the
construction of rice field houses is a mans job, the maintenance of the fields and the
harvest is done mostly by the women. Some young Badri men earn money by driving
and repairing taxis and coal trucks. Meat is obtained mostly through hunting and
fishing in the jungle, which is a mans job. Meat can also be bought on the market.
The people in Sijyw earn their money doing many diverse jobs, but mainly by the
sale of betel nut. The village, when looked at from above, can hardly be discerned
because most of the houses are invisible under the leaves of the betel nut palms. Apart
from that, money is made by the cutting of stones, which the road contractors buy, the
sale of sand, bamboo, dry and fermented fish and other products on the market and
fishing. A few Sijyw people have jobs in one of the schools in the village or in the
government administration in Baghmara.
Other Atong villages that I visited on the western side of the Symsang river also
sell betel nut and stones. The villages on the eastern side of the Symsang river,
especially those at the foot of the Waimong mountain, grow and sell oranges and
some sell reed, which is used to thatch roofs.
1.4 Linguistic environment, language status and language use
The Atong are surrounded by speakers of different Garo dialects on all sides of their
language area. Garo is at the same time the name for a group of dialects spoken in the
Garo Hills and adjacent areas in the Khasi Hills, Assam and Bangladesh, the name for
the speakers of these dialects and an ethnonym comprising the speakers of the Garo
dialects, the Ruga people, who live in the East Garo Hills and who have lost their
Ruga language, having all converted to speaking Garo, and Atong speakers (see
below and 1.3.1 for more details about ethnic affiliation, and Burling 2004:9 for
more information about the term Garo). Burling (2004: 9) reports that there are more
than half a million Garos in India and well over a hundred thousand in Bangladesh.
The Garo language is written and has a standardised form often referred to as Achik.
The standard form is based mainly on the Awe dialect of Garo spoken in the northern
part of the Garo hills (see Burling 2004: 11). Together, the Garo dialects have more
speakers than all other languages spoken in the Garo Hills.
Only in the western-most part of the Atong language area is there direct contact
with speakers of Khasi languages, which belong to the Mon-Khmer language family.
There does not seem to be any influence of the Khasi languages on Atong, as far as I
am aware; however, the influence of Garo on Atong is rather great. Unfortunately,
due to the fact that Atong and Garo are closely related, although not mutually
intelligible, it is not always possible to tell whether identical lexical items are
borrowed or not and in which direction the borrowing went. There are many cases in
which it is possible to identify Garo loans, mainly on phonological criteria, e.g. the
retention of initial clusters in the pronunciation of Garo loans. The influence of Garo
on Atong is particularly noticeable in the use of numerals, at least in the villages
where I conducted my fieldwork. The Atongs in those villages, especially the younger
speakers, count almost exclusively in Garo (see Chapter 1).
Almost all Atong speakers in the places where I did fieldwork, on the western side
of the Symsang river, are bilingual in Garo from a very early age. Only in some of the
more isolated villages on the eastern side of the river is the situation different in
favour of Atong: the proficiency of people in Garo is much less developed there. Garo
replaces Atong in several domains in daily life. First of all, the Atong do not speak
their language to strangers. If a stranger visits the village, they will first speak Garo
until another suitable language of communication is found. The Atong also do not
speak their language when they are in the company of Garo speakers. In market places
like Jadi and Nangwalbibra (see Map 3), when an Atong speaker addresses an
unknown sales person, they will always speak Garo, even if the sales person reveals
herself or himself to be Atong. The Atongs have a rather negative image of their own
language and are not comfortable speaking it in front of strangers and non-Atong
speakers, especially Garos and especially in Tura, since Atong speakers there are
often ridiculed by some Garo speakers who say that the Atong are backward savages.
Secondly, Garo is the language of the Church. Almost the entire Atong-speaking
population on the west side of the Symsang river are Christians. On the other side, in
the more isolated villages, the amount of converts is, as far as I was able to make out,
much less, although this situation is changing rapidly as Christian proselytisation
continues relentlessly. In all churches, Baptist, Catholic and United Church
Association alike, services are held entirely in Garo. Even personal prayers to God are
in Garo. The same is true for private services held in peoples houses. When I asked
an Atong friend whether it was possible to pray in Atong, he said no, they have to
pray in Garo. When I asked if God did not understand Atong, my friend thought for a
while and then said that of course God did understand Atong, since He understands all
languages. He was visibly surprised by my question and then remarked that he did not
fully understand why they had to pray in Garo. The main reason that Garo is the
language of the Church is, I think, because the bible has not been translated into
Atong, while there is a bible in Garo. Moreover, all mission work is carried out in
Garo. The local Catholic missionaries, for example, who run a nunnery and an
English school in Siju, all Indians from the state of Kerala, do not learn to speak the
language of their congregation, but learn to speak Garo instead. Even church
functionaries who are native Atong speak Garo as soon they are in church, and often
even switch to Garo when speaking about church related matters when they are
outside the church.
Thirdly, Garo is the language of education. This is true for all villages where there
is a school. The only exceptions are the English school in Gangga, Siju, behind the
football field, and the Catholic mission school, where the medium of education is a
local variety of English. Many Atong students who are still in school by the age of 15
seem to go to places like Williamnagar and Tura (see Map 2) for their education,
where Garo is the main language.
On top of the use of Garo in certain domains of life, the fact that Atong speakers
hold Garo to be a prestigious language can be seen in the remark that one of my
friends once made about his brothers, i.e. that they like to show off their proficiency
in Garo. A contrary experience also needs to be documented. While we were driving
to Balphakram, land of the spirits and national park, an Atong friend told a Garo
speaker that he knew a beautiful story about a certain subject, but that he would not be
able to tell the story in Garo. Although Atong is under a lot of pressure from Garo, the
language is still learnt by most children in villages where Atong is the dominant
language and shows no signs of obsolescence. It is impossible for me to say
objectively what the situation of Atong is at the moment. Given the enormous
percentage of bilingual speakers and the prestige of Garo, the majority of the Atong
language community could simply decide to abandon their language overnight. I have
no idea how long this situation has already existed and how long it will still continue
to exist. It seems to me that the language will not just disappear within a few
generations. When last I left the Garo Hills, in September 2007, some of the most
popular tunes of the moment in the Garo Hills were pop music songs sung in Atong.
These songs are distributed on cassettes through shops. Even Garo speakers were
trying to sing them without knowing the meaning of words.
Indic languages have been of great influence on Atong allegedly for many
centuries. Atong is heavily infiltrated with Indic loans, some of which look like Hindi,
and others are probably Bengali or Assamese. Because of the similarity of the
languages, no attempt will be made in this thesis to distinguish between loans from
Assamese, Bengali and Hindi. Examples of Indic loans in Atong are duk sorrow,
sadness, bia wedding, tas game of cards, gari vehicle, baji hour and taka
money. Meghalaya is surrounded by speakers of Assamese in the west, north and
east, and of Bengali in the west and south. As has been noted above, speakers of Indic
languages from all over north India come to market places like Jadi and Nangwalbibra
for commerce. Although non-tribal people in Meghalaya cannot possess land, they
can rent property, and thus settle among the Tibeto-Burman population. The police
force consists mainly of people who speak Indic languages. There are also a lot of
Bengali immigrants, mostly economic refugees, who continuously pour into
Meghalaya and settle there, often illegally. Road workers and workers in the coal
mines of the Badri area are also mostly Bangladeshis. During the dry season there is a
great influx of seasonal workers, mainly Northwest Indians working in the transport
sector and coal trade. Many Punjabis, for instance, use their huge trucks to transport
fruit in summer and coal in winter.
English also has an impact on the Atong language. Many words for modern
objects, notions and practices that are newly introduced into the culture are borrowed
form English, e.g. skul school, edres address, tibi TV, sendel flip flops, sandals,
gilas ~ golas glass, rens wrench and igec engage (see also 1.11.2i and van
Breugel, 2009 a). For lack of native English speakers in their direct environment, we
have to assume that these words are all borrowed indirectly through the various
people who sell these new modern objects and through the schools and churches
where people learn about new notions and practices associated with a modern way of
living. As was said above, there are two schools in the Atong language area that
provide education in English, both of them in Siju. No doubt loans find their way into
society through education. Another source of English loans is the administration. The
official language of government administration in Meghalaya is English. My
perception was that English is mainly used as a written language and that the
administrative staff speak whatever language is most convenient with the person they
are talking to. Village meetings on the other hand are held in Atong or Garo. For a
historical overview and references on colonial historical description I refer the reader
to van Driem (2001: 528 ff). All I will say here is that, despite the fact that the British
colonial administration gained control over the Garo Hills in the 1870s, the Atong
speaking area was reportedly still almost inaccessible in the 1950s for lack of a road.
Atong and Garo speakers intermarry, being of the same ethnicity and having the
same surnames. As a result, in most villages Atong and Garo people live side by side.
In some villages that are said to have been Atong speaking in the past, the presence of
Garo speakers has led to the complete disappearance of the Atong language from the
village. It is impossible for me to say whether the number of Atong speakers is on the
rise or not at the moment.
Until 2006 Atong was an unwritten language. When I arrived in the area, the
people even thought that their language could not be written. Luckily, they were
wrong. I have developed an orthography for the language based on the Roman
alphabet, which will be explained in the next section. Today, the first editions of an
Atong-English dictionary and a book with stories collected during my fieldwork are
being written (van Breugel, 2009 a and b), while draft editions of the same books have
already been distributed in the language community. With some of my Atong friends I
now maintain correspondences in writing in Atong via mail and sms messages. Time
will tell if literacy in the language catches on and becomes more wide spread.
1.5 The Atong spelling system
The way in which the Roman or Latin alphabet is used to write Atong is represented
in Table 2 below. An overview of the phonology can be found in Chapter 2. Most of
the writing system is self evident. Only a few remarks are in place.
As we see in Table 2, the grapheme is used to represent both the vowel
phoneme /i/ as well as the glide /y/. This choice was made because the Roman letter y
is used to represent /e/, like in Welsh. As a consequence, the orthography makes
Atong look like it has diphthongs, e.g. kyi /koy'/ dog, askhui /askhuy/ star, and
mai /may/ rice, while phonologically, diphthongs do not exist in the language, but
are in fact sequences of vowels and the off-glide /y/ or /w/, as is discussed in 2.2.6.
Because literate Atong alive today have already learned how to spell in Garo, I
preserved the tradition of the Garo writing system by representing the phoneme /c/
with the digraph ch and // with the digraph <ng>. It is not necessary to write /c/ as a
digraph, since the letter <c> is not used anywhere else in the language, but this is the
written in the orthography developed for the language.
Phonemes Graphemes Phonemes Graphemes Phonemes Graphemes
ph

ph m m i i
th

th n n e e
kh

kh ng a a
p p r r o o
t t l l u u
k k s [c
h
~ c] s y
b b c [tc] ch ii
d d j [dz] j ee
g g h h oo
w w y [j] i aa
glottalisation [?]

or

way my consultants insisted on writing their language, and they are right that it is
convenient when one is already accustomed to writing the same phoneme in Garo like
that. It goes without saying that sequences of // and /g/ are spelt <ngg> in the
orthography, e.g. hanggal /hagal/ charcoal.
The raised dot <> to represent the glottal stop that is the result of glottalisation of
the syllable (see 2.12) is also a perpetuation of Garo spelling and can also be written
with an apostrophe, which is readily available on all typewriter and computer
keyboards.
I do not propose a standardised spelling for Atong. The language is spoken in
different ways and the spelling system can represent all the dialectal differences that I
came across during my fieldwork. The word for big knife, for example, is chawkyi
/caw'koy/ in the Badri dialect and changkui /ca'kuy/ in Siju and both varieties can
be spelt. For more detailed information on the spelling I refer the reader to the
introductions of van Breugel (2009 a and b).
1.6 Dialectal variation
Although I was not able to conduct a dialectal survey throughout the whole of the
Atong speaking area, I was told by various speakers that there were four major
dialectal areas in Atong, viz. Badri, Rongsu, Siju and Baghmara. From my encounters
with people from those places, I know that all these dialects are completely mutually
intelligible and that the main differences between them are purely lexical and
morphological, morphological differences being very small. In addition to there is a
difference in pronunciation: Atong speakers from Siju and Baghmara speak softer and
more melodic, while people from Badri and Rongsu speak loud and more
monotonous. I have the impression that the Rongsu dialect is closer to that of Badri
and the Baghmara dialect closer to that of Siju.
This grammar is based on the dialects of Badri and Siju. The Atong foreword to
this grammar is an illustration of how different the two dialects are. Table 3 presents a
list of some of the most important lexical and morphological differences between the
two dialects. It has to be noted that the sound correspondences shown in the table are
not regular.

Table 3 Illustration of the variation in lexemes and grammatical morphemes in the
dialects of Badri and Sijyw
BADRI SIJYW
gloss
lexeme/morpheme lexeme/morpheme
sangwal- awan- to forget
awoy abu grandmother
khugri koksi small basket
ha'cepcep gukcepcep cricket
cengkuy caw'koy big knife
gukmadom gukmatom grasshopper
kririp koyrop type of edible plant
taw'toy taw'ti egg
badol patol slingshot, catapult
=mo =mi genitive/ablative case
=mo =mu ~ =mu comitative case
=mo =mu ~ =mu ~ =muna~ =muna sequential clausal enclitic

There is also significant difference in elocution between the two dialects. Badri
speakers insert more glottal stops when vowels meet across morpheme boundaries,
while Sijyw speakers tend to fuse adjacent vowels into one, within the same word,
which makes their language more difficult to understand.
Both Siju and Badri people think of themselves as speaking what they call pure
Atong. To my amusement they always wanted to demonstrate this by accusing each
other of using a certain word for shirt: In Siju they say chola, but the real Atong
word is jama!, the Badri people would say, and vice versa. Unfortunately, both chola
and jama are not Atong at all, but loanwords from some Indic language.
6
Given the
huge number of loan words in the language, there is no such thing as pure Atong,
there are just different varieties of it.
1.7 Linguistic affiliation
Most authors who have worked on the linguistic history of the Tibeto-Burman
languages spoken in the valley of the Brahmaputra in Assam remark that they have
many words in common, so that it is easy to see that they are historically related.
Nevertheless, because of the scantiness of the data on most languages, often not more
than small lists of vocabulary transcribed in disparate ways, there appear to be many
different genetic classifications or groupings of the languages of Assam, Meghalaya
and the adjacent areas.
The available documentation on languages generally held to belong to the same
group as Atong (the Bodo-Garo or Bodo-Koch group) is short. For Garo we have a
grammatical description by Burling (1961, 2004). Rabha is described by Joseph in his
grammar (2007) and Rabha-English dictionary (2000). The earliest documentation of

6
The Atong word chola and jama could be borrowed from Bengali, Assamese or Hindi: Bengali
/cla/ bodice, corset, modesty vest, skirt, pronounced as [tcula], related to Assamese [sula]
jacket, tunic, coat and Hindi /cl/ gown, cloak, brides garment. As for the origin of Atong
jama: Bengali or Assamese /jm/ coat, shirt, blouse jacket, related to Hindi /jm/gown,
wedding robe. It hank Stephen Morey for helping me here with his knowledge of Assamese and
Bengali.
Boro (or Bodo) is by Reverent Endle (1881), who calls the language Kachari or Br.
Later descriptions are by Bhattacharya (1977) and Basumatary (2005). Deuri is
described by Jacquesson (2005).
As we will see below, it was noted as early as Grierson (1902) that Atong is
closely related to Garo, and other languages spoken in and around the Garo Hills.
However, opinions about which languages belong to the same group as Atong, the
nature of the relationship and the name of the group change with time. I will give an
overview of the different affiliations of Atong claimed by different authors in the
twentieth and twenty first century.
Grierson, in the famous Linguistic Survey of India (1902: 85), and Benedict, in his
equally famous (1972) Sino-Tibetan conspectus, consider Atong to be a dialect of
Garo, which, on a higher taxonomic level, belongs to the Bodo-Garo group or
nucleus within Tibeto-Burman (see Figure 1).
7
Benedict (1972: 6-7) states the
following in connection to the affiliation of Atong: Garo shows an interesting
division into two subtypes, which we have named Garo A (Rabha, Ruga, Atong) and
Garo B (Abeng, Achik, Awe), the latter spoken by the dominant political divisions
of the tribe. Robins Burling (2003 a: 176) reacts to Benedicts classification, saying:
It has long been clear that the languages of the Atong and Ruga are close to Rabha,
which is spoken to the north and northwest of the Garo area, and Benedict duly
included Rabha among the Garo A languages, even though no one considers the
Rabhas to be Garos. To call Atong and Ruga Garo A is to recognise common

7
The existence of Koch is not mentioned in Benedict (1972). This is interesting because the status of
Koch as a separate language, or group of dialects, is questioned in Grierson (1902: 95-96), who writes:
The name Kch, in fact, everywhere connotes a Hinduized Bodo who has abandoned his ancestral
religion for Hinduism and the ancestral Bodo language for Bengali or Assamese. There is, however, in
Dacca, the Garo Hills and Goalpara a small body of people who are known as Kch or Pni Koch, and
who still speak a language belonging to the Bodo group, and are either animistic or nominal Hindus.
Six sections have been recorded in the Garo Hills, viz., Harigay, Satpariy, Dasgay, or Banai,
Chapra, Wanng, and Tintekiy. [] These six sections used to be considered to be the only pure
Kches in existence, but it is now believed to be much more likely that they are not Kches at all, but
are Grs who have never got beyond an imperfect stage of conversion to Hinduism, involving merely
the abstinence from beef. [] Their language, so far as I can judge from the specimens which I have
seen, is a mongrel of Gr, Bengali or Assamese. For more information and references to literature on
Koch, see van Driem 2001: 534 ff.
ethnicity; to group Rabha with Atong and Ruga is to recognise linguistic similarity.
To conclude that Rabha is a kind of Garo is like calling Welsh a form of French
because it is so much like Breton, whose speakers are French. The Rabha do not count
as Garos either by ethnicity or by language, any more that the Welsh count as
French. Burling attributes the great similarity between Atong and Garo to heavy
mutual influence of the two languages on each other.

Figure 1 Schematic chart of Sino-Tibetan Groups, from Benedict (1972: 6)

Burling (1959: 437 and 1961: 80), as did Shafer (1953: 228), classified Atong as
belonging to the Koch group of languages. The Koch group, in turn, is part of a larger
Bodo group of languages, which consists of Koch, Garo and Bodo. Bodo and Garo are
somewhat more closely related to each other than to the languages of the Koch group.
In later work, viz. Burling (2003 a: 175-6), his view on the position of Atong within
Tibeto-Burman remains unchanged, but the genetic super grouping is different and
more refined: the Koch group of languages is now part of a larger group called Bodo-
Koch which is, in turn, part of the Bodo-Konyak-Jinghpaw super group of Tibeto-
Burman languages, as is represented in Figure 2.

Figure 2 Relationships among the Bodo-Konyak-Jinghpaw languages, from Burling
(2003 a: 175)

Van Driem (2001: 501-2, 534) also classifies Atong as belonging to the Koch
group of languages. This classification is based on references and not on his own
diachronic research. Probably based on his interpretation of Burling (1961), van
Driem states that Atong is a Koch dialect (2001: 541), which, as we will see below,
is not the case. The other languages of the Koch group, according to van Driem, are
Ruga, Rabha and Pani Koch. The Koch language group is a subgroup of the Bodo-
Koch languages which all belong to the Brahmaputran branch of Tibeto-Burman. On
page 501, Brahmaputran consists of Konyak, Bodo-Koch, Dhimaslish and Kachinic,
whereas on page 502, Brahmaputran consists of Bodo-Koch, Dhimaslish, Northern
Naga and Kachinic. There is no explanation for these different constituencies of
Brahmaputran.
Robert Shafer (1974) has a totally different classification of the languages within
what he calls the Sino-Tibetan language family than the authors mentioned so far.
According to Shafer Atong belongs to the South Central Branch of the Barish Section
within the Baric Subdivision of Sino-Tibetan. The languages Shafer regards as closest
linguistic relatives of Atong are: Rabha, Ruga, Konts, Tintekiya, Cooch Behar and
Kots. Shafers classification is the result of a thorough phonological comparison
between the languages of the Baric Subdivision.
In Joseph and Burling (2006: 1), Boro-Garo is presented as one of the longest
recognised and most coherent subgroups of the Tibeto-Burman family of languages.
Boro-Garo, where Boro is just a different way to spell Bodo, consists of four
subgroups, viz. Garo (several dialects), Koch (consisting of Atong, Rabha, Ruga, the
Koch languages and maybe Mandai (the existence of this language is not certain)),
Boro (Boro, Kokborok and Tiwa, Kachari and Mech) and Deuri (a language that
constitutes a branch on its own).
Jacquesson (2006) poses a number of phonological criteria on the basis of which
he divides the Boro-Garo languages in three groups: Western, Central and Eastern.
The phonological criteria are the occurrence or not of diphthongs in the language,
of nasal vowels, of consonant clusters /kr-/, /gr-/ in first syllables, and of the existence
of a phonological distinction between /l/ and /r/ in the languages. The division of
Boro-Garo languages proposed by Jacquesson is represented in Figure 3.
The most striking difference between the Central group and the two other groups
of Boro-Garo languages is the occurrence of so called diphthongs in the Central
group and their absence in the Western and Eastern groups. These diphthongs of the
Central group, written /ai, au, ei, eu/, show regular correspondences with
monophthongs in the other Boro-Garo languages, e.g. Kokborok toi water, with the
diphthong /ei/, corresponds to Garo /ci/ water, with the monophthong /i/ (see
Jacquesson 2006: 286). Only Boro and Mech have all four of the diphthongs, the
other languages present different subsets (idem: 294), e.g. Boro thoi
2
blood with the
diphthong /ei/, corresponds to Dimasa thi
2
blood (idem: 288) with a monophthong
/i/ etc.

Figure 3 The classification of Boro-Garo languages according to Franois
Jacquesson (2006: p. 293)

1. Western group (Groupe occidental)
Garo
Rabha, Koch
2. Central group (Groupe central)
Boro and Mech
Bru
Dimasa and Moran
Kokborok
3. Eastern group (Groupe oriental)
Deuri

Jacquesson (2006: 292-3) discards Burlings classification of Atong as a Koch
language, indicating at the same time why Burlings label Koch for the group of
languages that are not Bodo is infelicitous. Given that Atong has diphthongs,
Jacquesson classifies it as a Boro language. It is worth quoting Jacquesson in extenso:

Malheureusement, outre quil est fort imprudent de promouvoir le nom dune
langue (koch) qui nest plus gure atteste depuis longtemps, il se trouve que
ses propres donnes sur latong montrent que cette langue possde plusieurs
diphtongues, ce qui la classe selon nous dans le groupe du boro Tout
rcemment
8
un jeune chercheur qui tudie ce parler, S. van Breugel, nous a
aimablement confirm quil possdait quatre diphtongues analogues celle du
boro.
Il nous semble donc inutile dutiliser cette tiquette de bodo-koch,
puisque quen ralit nos meilleurs renseignements sur le koch montrent quil
sagit dun parler trs proche de celui des Rabha [] Quant latong, en
attendant les rsultats de S. van Breugel, il parat raisonnable dy voir un
parler de type boro.
9

Since historical comparison lies outside the scope of this grammar, I will limit myself
to showing that, by Jacquessons criteria, Atong is indeed a Central Boro language,
but not without first making an important remark. What Burling and Jacquesson
analyse as the diphthongs /ai, au, ei, eu/, I analyse phonologically as sequences of
vowels and off-glides. The argumentation for this analysis can be found in Chapter 2.
Moreover, Atong has seven of these vowel-plus-off-glide sequences, viz. /aw, ew, ew,

8
Tout rcemment very recently: Jacquesson means during the first conference of the Northeast
Indian Linguistics Society (NEILS), held at Guwahati on 6 & 7 February 2006, see also van Breugel
(2008).
9
English translation of quote: Unfortunately, except for the fact that it is very unwise to promote the
name of a language (Koch) that has not been attested for a long time, it turns out that his own data on
Atong show that this language has several diphthongs, which classifies this language, according to our
criteria, in the same group as Boro Very recently, a young researcher who studies this language, S.
van Breugel, has confirmed that it has four diphthongs corresponding to those in Boro. Therefore, it
seems futile to us to use the label Bodo-Koch, because, in reality, our inquiries on Koch show that
this language is very close to Rabha [] As for Atong, while we await the results of S. van Breugel, it
seems reasonable to consider it as a language of the type to which Boro also belongs.
ay, ey, oy, uy/. I would like to remind the reader that my grammar is a synchronic
description of the Atong language, and that I have analysed the sounds in the language
on the basis of their current function in the phonemic system and not on the basis of
their alleged history, however well documented.
Let us now turn to the classification of Boro-Garo languages according to the
criteria as presented in Jacquesson (2006: 294). I will list the criteria, translated into
English and slightly adapted to fit the phonology of Atong, and copy the table that
Jacquesson presents, in which he indicates how the criteria apply to the different
Boro-Garo languages, adding Atong to it (Table 4).

Table 4 The classification of Boro-Garo languages according to Jacquesson (2006:
294) including Atong, until now correctly suspected to be closest to Boro
(see Jacquesson (2006: 293, quoted above).
Criteria:
A typical diphthongs/vowel-plus-off-glide combinations
B which diphtongs/vowel-plus-off-glide combinations
C nasal vowels
D clusters /kr-/, /gr-/ in the first syllable
E distinction between /r/ and /l/
Criteria A B C D E
Western
Garo no no yes no
Rabha,
Koch
no
no
no yes
Central
Boro,
Mech
yes ai, au, ei. eu
no

Atong yes
ay, aw, y. w,
oy, uy, ew
no
no no
Bru yes ai, au, ei
no

Dimasa,
Moran
yes ai, au
no

Kokborok yes ai, ei
no

Eastern Deuri no yes no no

As we can see, according to criterion A, typical diphthongs/vowel-plus-off-glide
combinations, Atong is a language very much like Boro, since Atong has the same
diphthongs as Boro, viz. /ay, aw, ey. ew/ plus three additional ones: /oy, uy, ew/.
10

When we look at criterion D, clusters /kr-/, /gr-/ in the first syllable, the claim that
Atong is a Central Boro-Garo language seems to hold. Although he leaves the box
under D empty for Boro and Mech in Table 4, Jacquesson tells us that some languages
allow clusters only when a syllable is not the first one in a word, and that certain
languages tend to simplify the cluster by dropping the /r/. Jacquesson writes: Les
seules langues o il ny ait pas de groupes dans ces conditions sont le boro et le deuri.
En kokborok, la situation diffre selon les dialectes [] (2006: 286).
11
A simple
extract of the compared data presented by Jacquesson (idem: 285) will serve as
evidence for the classification of Atong. This evidence is the comparison of the four
lexical items shown in Table 5. We see that the box under to buy in the last column
has been left empty for Atong since there is no verb to buy in the language.
12
In
addition to the verb khop- to cover, presented in the table, I also recorded the form
gorop- to cover, in which the /r/ of the proto-form has been preserved
13
, the cluster

10
It has to be noted that according to Joseph and Burling (2006: 15) Boro has eight diphthongs, viz. /iu,
eo, ou, ao, oi, ai, ui, ui/. Not only is the amount of diphthongs given by Joseph and Burling twice as
much as the inventory presented in Table 4 for Boro, but Joseph and Burlings Boro diphthongs also
seem to be phonetically different in nature, so that they are very difficult to compare to the Atong
diphthongs. Basumatary (2005: 23) also describes eight diphthongs for Boro, but these are different
from those in to Joseph and Burling, viz. /ui, wi, wu, oi, ou, au, eu, ai/. Finally, Endle (1881: 3)
describes six diphthongs for Boro, written as /au, a, u, ai, , oi/. The description of the phonetic
realisation of these sounds is hard to understand and can therefore not be given with certainty here in
IPA symbols. Not indicated in Jacquessons table (2006: 293) and therefore not in Table 4, are the
diphthongs of Garo and Rabha, which are available from Joseph and Burling (2006), Joseph (2007)
and Burling (2004). Joseph and Burling (2006: 20) posit three diphthongs for Garo, viz. /ai, oi, au/ of
which /oi/ only occurs in loanwords (see also Burling 2004: 29-31) and eight for Rabha, viz. /ai, ao, au,
ui, ui, oi, eu, eo/ (2006: 25, see also Joseph 2007: 57-66).
11
English translation of quote: The only languages where there are no clusters in these conditions are
Boro and Deuri. In Kokborok the situation differs according to the dialect [].
12
The verb ra'- to get is used in the same situations where English uses the verb to buy. If a speaker
wants to emphasise that money was involved in obtaining something, he can express that using the
word taka money as we can see in example (429) in 20.2.
13
I also recorded another such word in which the proto */r/ or */l/ is preserved but where the insertion
of schwa prevents a cluster, viz. gorow- to shake (a fixed object). It might well be that there are word
families (see Matisoff, 1978 and 2000: 344-7) within Atong of lexical items that have lost proto*/r/ or
/l/ in clusters with initial /g/ or /k/ and which exists alongside words that have preserved proto */r/ or /l/
but have broken the cluster up with a schwa; in certain cases the schwa might have assimilated to the
has been broken up by a schwa to meet the phonological requirement that there be no
consonant clusters in initial syllables.
14

Table 5 The reflexes of Proto-Boro-Garo */kr, gr, kl/ in Garo, Rabha, Boro,
according to Jacquesson (2006: 285) with the addition of Atong.
gloss to cry sour to cover to buy
proto-Boro-Garo form *grap *krei- *klep- *brai-
Garo CC grap- krip-, kip bre-
Rabha C(C) khap- khi- khep- pri-
Boro C gab- khei- kheb- bai-
Atong C kep- khy- khp-

Finally, we examine criterion D, distinction between /r/ and /l/. As I argue in 2.2.4,
the two phonemes are in contrast in syllable-initial position. This might of course be
due to unrecognisable integrated loans in the language. Only in animal names
containing the initial syllable ruk ~ luk are /l/ and /r/ interchangeable, e.g. rukwak ~
lukwak type of frog, rupek ~ lupek type of frog. In syllable-final position Atong
only has /l/ in indigenous words (see 2.2.4). It is difficult to find possible Atong
cognates for two of the four words used as evidence in this matter in the second table
on page 285 of Jacquesson (2006). Thus to find proof, other vocabulary items should
be compared for which there are clear Atong cognates. This remains a matter for
further investigation. All I can add to the table in question found in Jacquesson (2006)
are the words for dry, ran'-, and long, raw'-, as compared to the other Boro-Garo
languages in Table 6, adopted from Jacquesson (2006). As we can see, the distinction
between /r/ and /l/ that existed in the proto language is no longer there in Atong,

vowel in the next syllable (see 2.6). Thus in Atong the verbs khop- and gorop- are allofams within
one word family, i.e. phonosemantically similar but not identical forms that can be traced back to a
single etymon (Matisoff, 2000: 344).
14
In non-initial syllables we find speakers alternating freely between clusters CrV and CerV (see
Chapter 2).
where proto */l/ has been replaced by /r/. It has to be remarked that there are many
more words starting with /r/ than with /l/ in Atong (see van Breugel, 2009 a).

Table 6 The reflexes of Proto-Boro-Garo */r/ and, */l/ in Boro-Garo languages,
according to Jacquesson (2006: 285) with the addition of Atong.
15

gloss dry long
proto form *ran
2
*lau
2
Garo ran?- ro?-
Rabha ran
2
- ro
2
-
Atong ran- raw-
Tiwa ran- luw-
Dimasa rai
2
- lao
2
-
Boro ran
2
- lao
2
-
Kokborok ran
2
lok
2
-
Deuri ran
2
- lu-

We can conclude that, when following Jacquessons criteria for the classification of
Boro-Garo languages, Atong seems to be more closely related to Boro than to Koch
and Rabha, as Jacquesson already assumed. However, despite all the surface
similarities between all the languages of the Boro-Garo group, I think that in order to
refine our knowledge about the way in which these languages are related, internal
reconstruction should precede interlingual historical comparison. This is a task that
has yet to be begun for all of the languages concerned. We should heed Matisoffs
words when he writes: The easiest proposals to dismiss as chimerical are those which
depend entirely on surface similarity among forms from modern languages, without

15
The superscript 2 indicates high tone in the languages that have tone, viz. Rabha, Dimasa, Boro
and Kokborok. Since the tones of Tiwa are not yet fully understood, tone is not indicated for this
language. Garo, Atong and Deuri have no tones. Garos glottal stop and Atongs glottalised syllables
correspond to syllables with a high tone in the other languages (see also Joseph and Burling 2006).
bothering to attempt reconstructions of proto-forms in the languages to be compared.
(2000: 357)
1.8 Previous work on Atong
The linguistic Survey of India (Grierson 1902: 85-88) mentions the existence of
Atong as a dialect of Garo and briefly presents some aspects of the grammar and a
translation of the Parable of the Prodigal Son with interlinear glosses in English,
translated by the Rev. E.G. Philips. Unfortunately there is no analysis of the
phonology of the Atong or a guide to the pronunciation of the text. However, the
parable, despite the confusing orthography, is clearly written in Atong. The
differences with modern Atong as recorded in this grammar are mostly orthographic
in nature. There is no separate symbol in the writing system for the phoneme /e/, and
therefore we must assume that the text is written with the assumption that /i/ and /e/
are allophones, as is analysed for standard Garo, where /i/ has an allophone [i] in open
and [i] in closed syllables (see Burling 2004). The verb which I recorded as /thoy/
[thej] to die is written tai in Grierson in the word taiokgitchim died (page 87 and
88). It would be far fetched to assume that Atong did not have the sound [e] around
the turn of the nineteenth century. Modern Atong, as described in this grammar, has
three series of stops, viz. plain voiceless, voiceless aspirated and voiced (see Chapter
2). The text in Grierson makes no difference in the orthography between aspirated and
non-aspirated voiceless stops.
Salient lexical differences between the Atong recorded in Grierson and that
recorded for this grammar are the verb to give, which is hun in Grierson and hon'
in my recordings, and the word for but, which is utakchiba in Grierson and otokciba
in my recordings. In Figure 4 below are the first few lines of the Parable of the
Prodigal Son by the Rev. E.G. Philips, taken from Grierson (1901: 86), followed by
the first few lines of the same parable as told to me by Kempai A Sangma in the
village of Siju in 2006, presented in examples (1)-(5). The indicated translation of
Philips text consists only of the interlinear glosses.
The most salient grammatical difference between the text in Grierson and the text
recorded by me is the occurrence of a morpheme gitchim ~ chim was. The form
chim, treated separately in Grierson on page 85 and glossed was, but in the text in
occurs most frequently as gitchim, e.g. ganangitchim were (first line in Figure 4)
and maakgitchim lost-was (Grierson, 1902: 87). This morpheme corresponds in
form and function to what I call the irrealis clausal enclitic <=com> (IRR), e.g. (1)
and (34).

Figure 4 The first few lines of the Parable of the Prodigal Son by the Rev. E.G.
Philips, taken from Grierson (1901: 86).

The following examples present the beginning of the Parable of the Prodigal Son told
to me by Kempai A Sangma in the village of Siju in 2006.

(1) so dam sa =ci morot mo' sa
village CLF:VILLAGES one =LOC person CLF:HUMANS one
man' =ay sa' -bi =gaba gana =no =com.
in.great.amounts =ADV eat -VERY =ATTR exist =QUOT =IRR
In a village supposedly lived a very rich person (lit. a person who ate in very
great amounts), it is said.

(2) ue u =ci sa' banthay mo' ni gana =no =ro,
he DST=LOC child bachelor CLF:HUMANS two exist =QUOT =EMPH
man' =ay sa' =gaba mo' sa morot =ci.
in.great.amounts =ADV eat =ATTR CLF:HUMANS one person =LOC
He supposedly had two sons there, it is said, the very rich person.

(3) otokoymu wa' =gaba =ci gam pa' -a =no.
so.then father =DREL =LOC riches many -CUST =QUOT
So then, the father had lots of riches, it is said.

(4) ue gam pa' -wa =mi gomon kam pa' -wa =mi gomon
DST riches many -FACT =GEN reason riches many -FACT =GEN reason
ge'thethe mokboruk =ay =mu: baba ha'golsak =ci
3p jealous =ADV =SEQ dad world =LOC
a =mi suk =do ni' -ok.
1s =GEN happiness =TOP not.exist -COS
Because of the wealth and riches, because they were jealous of each other:
Dad, I have no happiness left in the world.

(5) a =do dothamanca na' kha'gal =ci =do na' =mi gam' =aw
1s =TOP especially 2s love =LOC =TOP 2s =GEN riches =ACC
a =na hathi -pha =bo no =ay =mu
1s =DAT divide -IN.ADDITION =IMP say =ADV =SEQ
jo =gaba =e bal -wa =no.
younger.brother =DREL =FC say -FACT =QUOT.
If you especially love me, divide your wealth for me, he said, the younger
brother, and he spoke, it is said.

What gives the text in Grierson a particularly artificial trait is the lack of right
dislocated NPs, i.e. the fact that all clauses are predicate final, and, in addition, the
complete absence of quotative enclitics or the verb no to say, to indicate that the
storyteller got the information from someone else, e.g. (1), (2), (3) and (5). Had the
story really been recited by a native speaker, it would certainly have appeared with
left dislocated NPs and quotative constructions, as we can see in (2), where the
Location NP man'=ay sa'=gaba mo' sa morot =ci (in.great.amounts=ADV
eat=ATTR CLF:HUMANS one person=LOC) at the rich man is right dislocated after the
predicate gana exist.
Other than the material in Grierson, very little Atong language material has been
published. Playfair (1975: 167) presents a short comparative vocabulary of Aw,
Atong, Ruga, Rabha and Koch words with English translation. Playfair does not mark
glottalised syllables in his transcription. There are no occurrences of the sequence /ey/
in his list where we find them in the data collected for this grammar. Moreover,
Playfair does not distinguish between plain and aspirated voiceless bilabial and velar
stops in his transcription. Basically, the list is only understandable when one already
speaks Atong and can infer what is meant by the transcription by looking at the
English gloss.
Burling (1959) collected a list of Atong lexical items which he published in an
interesting article where he makes a convincing case for the reconstruction of Proto-
Bodo. In his transcription of Atong there are only two series of stops, one glide, /w/,
one voiceless affricate, /c/, and one lateral, /r/. As this grammar will illustrate (see
Chapter 2), Atong as I recorded it has three series of stops, viz. voiced, plain voiceless
and voiceless aspirated, two glides, /w, y/, a voiced and voiceless affricate, /j, c/ and
two laterals, /l, r/.
The latest publication of Atong language material before van Breugel 2008 is
that used in the historical comparison of the Baric languages by Shafer (1974: 426-
448). Shafers transcriptions do not always correspond with the data recoded for this
grammar, and, like the sources mentioned above, fails to mention that there are
dialectal differences within Atong, which affect the lexicon and some grammatical
morphemes (see Table 3).
1.9 Fieldwork
1.9.1 Data collection
My fieldwork was carried out in two stages, the first from 27 June 2005 to 2 May
2006 and the second from 13 June to 12 September 2007. During my first fieldtrip I
spent half the time in the village of Badri Maidugytym [bodri majdugetem] and half
the time in Sijyw [sidzew], also called Siju [sidzu] (see Map 3), where I was hosted
by very hospitable and generous Atong families. During my second fieldtrip I spent
somewhat less time in Badri and more in Siju. During both fieldtrips I also visited
several other Atong speaking villages on excursions. For the translation of the
recorded material into English, I had to take regular trips to Tura, since there was
nobody in the Atong villages who could speak English sufficiently to help me with
this task. During these trips I got acquainted with the urban Atongs.
I practised the fieldwork technique of participant observation to learn how the
Atong use the language in their daily life and as a result I learned to speak the
language. Language proficiency is of great help during the grammatical analysis. I
collected the following types of materials that form the data base for this grammatical
description of Atong:

1. Notes on the language as it was used in day to day life with descriptions of the
context, objects, animals, materials etc.
2. Audio recordings of stories and three songs. Most of these are written down,
glossed and translated with the help of consultants.
3. Video recordings of spontaneous speech of young, unmarried men from Siju (just
under six minutes), presented in this grammar as Text 1 and Text 2.
4. Written material produced by native speakers: one story and four short letters.

I recorded approximately four hours of language on tape. I recorded the speech of
both male and female speakers and of speakers of all ages. Male speakers are
represented most in the recorded materials. Although most people do not know when
they were born, and are thus unable to tell me their age, I estimate that the oldest
person I recorded was around seventy and the youngest around six years of age. I
recorded twenty six different speakers. I recorded many different genres of language
use, viz. spontaneous speech of unmarried men; stories for children told by adults and
by children themselves; epic stories, where the language is more complicated;
historical narratives etc. Most of the collected material consists of fictional narratives
(stories) of different kinds, e.g. fables and adventures, but historical narratives are also
well represented in the corpus. A smaller part of the material consists of epic stories,
narratives about cultural phenomena and activities in and around the village. Songs
and spontaneous speech are least represented in the corpus that was recorded on tape.
However, my fieldwork notes, written in the notebooks that I carried around with me
continuously while in the field, contain many spontaneous speech utterencees, some
with and some without a short description of their context. I recorded one woman
singing a Christian song in Atong, and my corpus holds two transcriptions of pop
music songs that were very popular at the time of my fieldwork.
When I set out on my fieldwork trip, I had not planned to use video recordings for
the collection of language data. The two short videos of spontaneous speech were
recorded by Samrat N Marak of Siju, who had borrowed my camera while I was out
of the village for a day. Although just under six minutes in length, these videos
provide evidence that some grammatical structures that are seldom attested in
narratives, occur frequently in spontaneous veryday speech.
A list of recorded fieldwork materials is given in Table 7. The list is ordered
alphabetically by autor, i.e. the speaker or writer of a text, and lists all the text
collected during my fieldwork, their medium of communication and the processing.
The table also indicates whether the autor is male or female, their age estimate and the
place where the text was srecorded or written. Spoken texts were recorded on tape and
are indicated as recorded in Table 7, while written texts that were not recorded on
tape are indicated as written. In a few instances whas a text first written down by
the author and then recorded; this, too, is indicated in the table.
Almost all texts are used to draw example from for this grammar, except two or
three texts, which were not usefull because they had not been transcribed and
translated or because they were written translations from Garo into Atong; in the latter
case, I was not sure whether the words and contstructions that were used were natural
language or the result of calquing from Garo.
Almost all examples in this grammar are drawn from natural speech, either from
the tape-recorded texts or from the fieldnotes. As few elicited examples were used as
possible. Unfortunately, when I started this project, because my original supervisor
advised me not to mark the source of each example, I did not indicate which were
elicited. At present it is not feasible to go back and source each example.
Table 7 List of texts collected during fieldwork
_____________________________________________________________________
Authors name. Male/Female. Age estimate. Recording place. Genre. Medium (Written/Recorded (on tape)).
Transcribed and translated?
1. anonymous. female. 30-40. Badri. Christian
song. recorded. yes
2. Aristo J Momin. male. 18-25. place unknown.
published pop song, recorded (recording no
longer in my possession). yes
3. Cheng M Sangma. male. 7-11. Badri. fable.
recorded. yes
4. Dalcheng M Sangma. male. 18-25. Badri.
adventure story. recorded. yes
5. Derus R Marak. male. 30-40. Siju. epic story.
written then recorded. yes
6. Dilseng R Sangma. male. 20-25. Siju. fable.
recorded. yes
7. Dorina A Sangma. female. 30-40. Siju. local
history. recorded. yes
8. Genda R Marak. male. 40-50. Siju. adventure
story. recorded. yes.
9. Genda R Marak. male. 40-50. Siju. historical
cultural story. recorded. yes. (See TEXT
3Way khuruta)
10. Genda R Marak. male. 40-50. Siju.
incantation. recorded. yes (See TEXT 4
Ca'masami way)
11. Gongran Ch Marak. male. 50-60. Siju.
cultural story with some singing. recorded. no
12. Gongran Ch Marak. male. 50-60. Siju.
attempt at epic story. recorded. no.
13. Jamila M Sangma. female. 30-40. Siju.
recipe. recorded. yes
14. Janita M Sangma. female. 18-25. Badri. fable.
recorded. yes
15. Jendik S Marak. male. 40-50. Badri. fable.
recorded. yes.
16. Jendik S Marak. male. 40-50. Badri. comical
cautionary tale. story. recorded. yes
recorded. yes
recorded. yes
19. Jentibirth M Sangma. male. 7-11. Badri.
adventure. recorded. yes
fable. recorded. yes
22. Johan A Sangma. male. 17-21. Siju. local
history. written then recorded. yes
23. Kempai A Sangma. male. 40-50. Siju.
parable. recorded. yes
24. Kiubirth M Sangma. male. 7-11. Badri. fable
recorded. yes
25. Kiubirth M Sangma. male. 7-11. Badri. fable.
recorded. yes
26. Kiubirth M Sangma. male. 7-11.
Badri. fable. recorded. yes
27. Limbu M Sangma. male. 14-16.
Badri. fable. recorded. no
28. Limbu M Sangma. male. 14-16.
Badri. fable. recorded. no
29. Miksrang. male. 14-16. Badri. short
letter. written. not translated
30. Miksrang. male. 14-16. Badri.
anecdote. recorded. yes
31. Monjila M Sangma. female 40-50.
Siju. local history. recorded. yes
32. Monjila M Sangma. female. 40-50.
Siju. local history. recorded. yes
33. Negverson M Sangma. male. 20-25.
Badri. cultural story. recorded. yes
Badri. cautionary story. recorded. yes
36. Nikseng S Marak. male. 18-21. Badri.
adventure story. written. translated
37. Rangsewa M Sangma. male.70-80.
Badri. love story. recorded. yes
38. Rangsewa M Sangma. male. 70-80.
Badri. local history. recorded. yes
Badri. story. recorded. yes
41. Ranus M Sangma. male. 20-25. Badri.
description of game. recorded. yes
42. Salseng R Sangma. male. 20-25.
Tura. Translation of Garo pop song.
written. translated
43. Samrat N Marak. male. 18-21. Siju.
adventure story. recorded. yes
cautionary story. recorded. yes
46. Sandish M Sangma. male. 15-17.
Badri. short letter. written. not
translated
Badri. short letter. written. not
translated
Badri. short letter. written. translated
49. Sandish M Sangma. 15-18. Badri.
cautionary story. story. recorded. yes
50. Todan M Sangma. male. 70-80.
Badri. local history. recorded.
transcribe but not translated
TEXT 4 Ca'masami way 598
51. Todan M Sangma. male. 70-80. Badri. local
history. recorded. transcribed but not
translated.
52. Todan M Sangma. male. 70-80. Badri. local
history. recorded. yes
53. Tonton M Sangma, male, 20-25 Badri.
adventure story. recorded. yes (See TEXT 1)
54. Tontonjyw. female. 40-50. Badri. counting
from 1-100. recorded. yes
55. Tontonwa. male. 40-50. Badri. counting
from 1-100. recorded. yes
56. Wilseng S Marak. 18-25. place
unknown. male. published pop song.
recorded (recording no longer in my
possession). yes
57. Samrat N Marak, Nilam R Sangma.
males. 17-21. Siju. spontaneous
speech. recorded (video). yes
58. Samrat N Marak, Nilam R Sangma,
Johan A Sangma. males. 17-21. Siju.
spontaneous speech. recorded (video).
yes
_____________________________________________________________________
1.9.2 Recording equipment
I used two analogue recorders, as was the policy of RCLT at the time of my PhD.
Most of the material was recorded on a Sony TCM-500EV in combination with a
Rde NT3 microphone. Some of the material was recorded with a Sony TCM-400DV.
All recordings are mono. For all recordings I used TDK SA 90 high position tapes.
The video recordings are made on my own Sony digital camera. Once back at the
RCLT, I digitalised my analogue recordings with the program Audacity into WAV
files and transcribed them in the program Transcriber. Ultimately the recorded
material will be archived at the Pacific And Regional Archive for Digital Sources in
Endangered Cultures (PARADISEC, see http://paradisec.org.au).
42

43
Chapter 2 Phonology
_____________________________________________________________________
2.1 Syllable structure
In Atong we find morphemes, i.e. roots, suffixes and ecnlitics, with the syllable
structures V, CV, VC and CVC. We can thus say that the canonincal syllab;e structure
in Atong is (C)V(C) and that the minimum syllable consists of only a vowel. Except
for the bound forms and of the demonstratives ue ~ u distal demonstrative
and ie ~ i proximal demonstrative and the interjections o expression of
understanding/acknowledgment and a strong attention seeking interjection, there
are no roots that consist exclusively of a vowel. There is one suffix that consists of
just one vowel and two enclitics, viz. the imperfective aspect suffix <-a> (CUST), the
focus enclitic <=e> (FOC) and the allomorph <=e> of the adverbial clausal enclitic
<=ay ~ =e> (ADV). The majority of the predicate head suffixes, other than event
specifiers (see Chapter 25), and the majority of enclitics are monosyllabic with a CV
structure, fewer suffixes and enclitics have a CVC or VC structure and other syllabic
patterns occur even less frequently. Most event specifiers and NP suffixes have a
CVC or CVCV syllable structure.
In word initial syllables consonant clusters are not allowed. On the other hand, in
non-initial syllables we find speakers alternating freely between clusters of which the
second phoneme is /r/ and the same cluster broken up by a schwa, i.e. CrV(C) and
CerV(C); for example, the noun /ha'bori/hill, mountain can be pronounced [ha?beri]
with a schwa between the /b/ and the /r/, or as [ha?bri] with a cluster of stop plus /r/,
and the noun /sukoru/ river snail can be pronounced with the schwa as [cukeru] or
without the schwa as [cukru]. Likewise, the event specifier suffix /-cora/ V totally,
etc (see Chapter 25) can be pronounced with or without the schwa, viz. [cera ~
cra]. The pronunciation with the schwa breaking up the consonant cluster is the most
usual. Consonant sequences of all sorts can occur at morpheme boundaries and schwa
is never inserted, e.g. cak-si (hand-finger) [tcaksi] finger, tok-thoning (neck-?)
[t:ktheni] neck; unless the morpheme boundary becomes obscure and then a schwa
can be inserted, as in mokren ~ mokoren eye. The morpheme mok eye (< Proto-
Tibeto-Burman *mik (see Matisoff, 2003: 346)) never occurs on its own, but
2 PHONOLOGY 44
is only found in compounds e.g. mok-somol (eye-?) eyebrow and mok-soram (eye-?)
eyelash etc.
2.2 Consonants
Atong has an inventory of nineteen consonants, presented in Table 8 below. Not all
consonants occur in all positions in the syllable. All phonemes occur syllable initially
except // and /y/. Table 13 below gives an overview of the syllable final consonants.
All phonemes will be treated separately below.

Table 8 Atong consonant inventory
Place of articulation
Manner of articulation
bilabial alveolar
alveo-
palatal
velar glottal
Stops
aspirated ph th kh
voiceless p t k
voiced b d g
Affricates
voiceless c [tc]
voiced j [d]
Fricatives
s
[c
h
~ c]
h
Tap or trill
r
[r ~ r]

Continuants
oral l
nasal m n
Glides w y [j]
2.2.1 Stops
In syllable-initial position Atong has three series of stops in bilabial, dental and velar
position, viz. aspirated /ph/, /th/, /kh/, pronounced [ph], [th], [kh] respectively, plain
voiceless /p/, /t/, /k/, pronounced [p], [t], [k] respectively, and voiced /b/ [b], /d/ [d],
/g/ [g]. All stops occur word internally. Examples of words with aspirated syllable
initial stop are phaksa long half, bangphak posts at the entrance of the bachelors
house between the floor and the horizontal beam above the entrance, thoy blood,
kha'tho heart, khaw hair of the head, rokhom type, shape. Examples of words
with syllable initial and word internal plain voiceless stops are pan tree, hapun
fishing-hook, toy water, kaltok person who never washes, koy dog, akay
mothers older sister. Words with voiced syllable initial stops are baju friend, abo
maize, dokom head, damdol woven bamboo matting used as the side of a house,
2 PHONOLOGY 45
gawi woman, girl, khagal poor person. Table 9 below shows the phonemic
contrast between plain voiceless, aspirated and voiced stops.

Table 9 Evidence for aspiration and voicing opposition in stops

ph p b
phal- share, shift of work
pal flower
bal- say, tell
th t d
thap- to beat (up)
tap time, turn
dap- to be on top, to press
kh k g
khal hole
kal horn
gal pride
2.2.2 Fricatives
Atong has two fricatives, both voiceless, viz. /s/ and /h/. The phoneme /h/ occurs only
syllable initially, whereas /s/ occurs in both syllable initial and syllable-final position. The
phoneme /s/ is an aspirated alveo-palatal fricative [c
h
] in syllable-initial position. This
phoneme has a non-aspirated allophone [c] in syllable-final position. There are very few
occurrences of syllable final /s/, e.g. ros [roc] juice and anaros [anaroc] pineapple; this
phoneme occurs mostly in syllable-initial position, for example in words like sa' [c
h
a?]
child and samal [c
h
amal] small ant. The latter word, anaros pineapple, is very likely
to be a loan. I have no evidence of the origin of the word ros juice and cannot say with
certainty that it is a loan, although one would expect it to be one on the basis of its aberrant
phonological structure, i.e. the fact that the word has final /s/. The phoneme /s/ occurs
word internally in words like konsa later and hapsan together, the same. The phoneme
/s/ is written phonemically as <s> in this grammar for two reasons, the first one being
convenience: it is easy to type, and the second reason being that it is written as <s> in the
orthography of the language (see 1.5).
The phoneme /h/ is a voiceless glottal fricative [h]. It occurs only in syllable-initial
position and mostly before the vowel /a/. Only in very few words does the phoneme /h/
occur before another vowel. The phoneme /h/ occurs in words like ho'o yes, ha'ba dry
rice and vegetable field on the slope of a mountain, hawci yonder, hokhot- to go out,
hua to swim and huraw gibbon. The phoneme /h/ occurs word internally in the
unanalysable word cokhoy bamboo fishing basket, and in the word laha resin. The
following table provides some minimal pairs with the phonemes /h/ and /s/.

2 PHONOLOGY

46
Table 10 Evidence for the phonemic contrast of the two fricatives /s/ and /h/
ha' Take this! sa' Eat!
hap place sap- to know
hok- to call loudly sok- to succeed, to hold out
2.2.3 Affricates
The opposition in the affricates is that of voiceless versus voiced. The voiceless alveo-
palatal affricate is /c/ [tc] and the voiced alveo-palatal affricate is /j/ [d]. Both phonemes
occur exclusively in syllable-initial position. The opposition between the two phonemes
/ch/ and /j/ can be proved by minimal pairs like ca tea, ja month, cok- cold, jok
spouse and cow liquor, jow- to sleep. The phonemes /c/ and /j/ occur word internally
in words such as ajot a childrens game, rajasa one hundred, ici here, macot to finish,
complete. Although affricates are phonetically built up of a stop element followed by a
fricative element, they function as single units in Atong. In Atong complex onsets do not
occur. Where complex onsets would occur, an indeterminate vowel breaks them up and
syllabifies them. No indeterminate vowel (i.e. schwa) insertion is found between the
phonetic elements of the affricates, hence they are phonological units. Moreover, although
there are three series of stops, viz. voiced, plain voiceless and voiceless aspirated, we only
find a voiced-voiceless opposition in the affricates, which is another argument in favour of
treating them as phonological units. Although affricates are phonological units, only the
stop element gets lengthened under stress, as we will see in 2.9.
2.2.4 The tap or trill and the oral continuant
The phoneme /r/ is pronounced as an alveolar tap (otherwise known as a flap) [r] and less
frequently an alveolar trill [r]. The phoneme /l/ is a voiced lateral continuant [l]. The two
phonemes /r/ and /l/ contrast in syllable-initial position in words like re'ea to go away,
leave, lela to drag ones foot and raydi turmeric, laysak type of vegetable. In words
which are truly of Atong origin, only /l/ occurs syllable finally, e.g. ol- to talk, wil- to go
down, toykhal river. The phonemes /r/ and /l/ occur word internally in, for example, dala
round bamboo mat, khokhola bold person, kara wings and ha'bori hill,
mountain. There are some words, all containing the bound root ruk ~ luk frog, in which
/l/ and /r/ are in free variation in syllable-initial position, e.g. rukwak ~ lukwak type of
frog, lukchokchok type of frog etc. In loan words the phonemes /r/ and /l/ behave
2 PHONOLOGY

47
differently. More about these phonemes will be discussed in the section on the phonology
of loan words.
2.2.5 Nasal continuants
The oral nasal phonemes /m/ [m] and /n/ [n] occur both syllable initially as well as syllable
finally and word internally. Examples of words containing these phonemes are nawa
retard, idiot, san day, manap morning, rimola slippery, sam weeds, medicine. The
velar nasal // [] occurs exclusively in syllable-final position. Word internal // is also
recorded. Examples of words with the velar nasal phoneme are: digoray fish trap,
konsa later and bobo liar. Minimal and near minimal pairs that demonstrate the
phonological opposition between /m, n, / are given in Table 11.

Table 11 Evidence of the phonemic contrasts of the nasal continuants
Syllable initially: mat-, to be sharp, nat- to clean by scrubbing
Root internally:
ama mother, anay aunt: fathers sister, aa first
person singular pronoun
Syllable finally: ran'- to be dry, ram road, ra rain
2.2.6 Glides
The phoneme /w/ is a labio-velar glide [w]. This phoneme occurs both syllable initially and
syllable finally. Examples of words with /w/ include wa bamboo, wen ~ wet classifier
for times/turns, ray'wil- to walk around something, woyset to wipe off, khaw hair of
the head, te'ew now, and jow' mother, taw' chicken, bird. The phoneme /w/ occurs
root internally in words such as haw'nokhol father-in-law and gawi girl, female and
daw'gep (bird-?) duck. The off-glide /w/ occurs only after /e/ and /a/, e.g. jow' mother,
and taw' chicken, bird, except in the words te'ew now and cew'khoy big knife (Siju
dialect), where it occurs after /e/.
The glide /y/ is a palatal oral glide [j] and occurs only syllable finally. Words like toy
water, toy' blood, hay come on!, lets go!, tay'ni today, and maynol sticky rice are
examples of words with this phoneme. The glide /y/ occurs after /e, a, o, u/, e.g. koy' dog,
may rice, cok-hoy bamboo fishing-basket, askuy ~ askhuy star. Combinations of /i/ or
/e/ followed by a glide do not occur. An overview of all possible combinations of vowels
followed by a glide is given in Table 12.
2 PHONOLOGY

48
Table 12 The possible combinations of vowels plus glide in Atong
Close uy
Close-mid w, y
Open-mid ew oy
Open aw, ay

Several reasons led me to the analysis of the glides /y, w/ as consonant phonemes
instead of the alternative analysis presenting them as diphthongs /ai, au, ei, eu, oi, ui/. The
alternative analysis as part of diphthongs would create several problems, which are solved
by the analysis of a consonant inventory with glides. First of all they cannot be followed by
a consonant in a root, e.g. *CauC, *CeiC etc. The glides also do not occur before or after
another consonant but this is in accordance with the canonical (C)V(C) syllable structure in
the language. Analysing sequences as vowel plus glide safeguards the canonical (C)V(C)
syllable structure of Atong.
Secondly, if one would analyse these phonemes as diphthongs, the second element of
these diphthongs would never receive any pitch or stress. In sentences like (6) and (7)
below, intensity and higher pitch stress falls on the bold and underlined vowel, not on the
second element of the putative diphthong, despite the fact that the general intonation
pattern of a word is from low to high with most prominence on the last vowel. Vowel
sequences occur in Atong only at morpheme boundaries, i.e. the sequence /eo/ in the
predicate of rak-bebe-ok=no (hard-TRULY-COS-QUOT) ran really hard, it is said,
where the /e/ and the /o/ belong to different morphemes, are therefore not diphthongs, and
both have the possibility to be stressed or to receive a higher pitch than the other.
However, pitch and stress patterns like (8) and (9) never occur, i.e. the final vowel of the
diphthong would never receive pitch and would never be stressed. In example (7), the
vowel is very long and very high in falsetto voice to mark the surprise and awe of the
speaker.

(6) nmi re'eb!
/nemay re'ebo/
nemay re'e =bo
well go.away =IMP
Go carefully!

(7) tow!
/atoooow/
Wow!

2 PHONOLOGY

49

(8) *nma re'eb!
*/nemai re'ebo/
nemay re'e =bo
well go.away =IMP
Go carefully!

(9) *to. !
*/atouuuu/
Wow!

All in all, it seems more natural to me to analyse /w/ and /y/ as off-glides and to posit
seven possible combinations of vowel plus off-glide than to posit diphthongs. If I write
combinations of vowels, e.g. /eo, ei, ie/, the two vowels always belong to separate
syllables.
Only a subset of the phonemes presented in Table 8 occur syllable finally. These
include the voiceless unaspirated stops, and the continuants. The possible syllable final
consonants are listed in Table 13 below. It should be noted that many more consonants
occur syllable initially than syllable finally.

Table 13 Syllable final consonants
Place of articulation
Manner of articulation
bilabial alveolar velar
Stops voiceless p t k
Fricative s
Continuants
nasal m n
oral l
glides w y [j]

The phonemes // and /y/ occur exclusively in syllable-final position, as was mentioned
above. Final stops are unreleased, i.e. /-p/ [p`], /-t/ [t
`
] and /-k/ [k
`
]. Syllable final /l/ in loans
can be pronounced [l ~ r ~ r] (see section 2.16).
2.3 The morphophonological process of fusion
When the declarative enclitic <=te> (DCL) follows the negative suffix <-ca> (NEG), the
vowel of the negative may partly assimilate to the vowel of the declarative enclitic and
rises to /i/. This /i/ becomes voiceless as it is wedged in between a voiceless affricate and a
voiceless stop. The resulting form after the fusion of the two morphemes is then <-cite>
[tci te] <NEG/DCL>. The optional assimilation of /e/ to /u/ under the influence of the
2 PHONOLOGY

50
following /w/ in the word gowa [gu.a] and the raising of /a/ to /i/ in the morpheme <-
cite> (NEG/DCL) are the only instances of regressive assimilation in Atong.
There is only one morpheme that has an allomorph consisting exclusively of a
consonant. This morpheme is the perfective suffix <-ok ~ -ak ~ -k> (COS). The allomorph
<-k> (COS) of this morpheme occurs only after the negative marker <-ca> NEG. The
negative perfect <-ca-k> (NEG-COS) will then sound like [tcak]. Forms like *[tca:k] or
*[tca?ak] which one would equally expect to occur in a combination <-ca-ak> (NEG-COS)
do not occur in Atong. For an explanation of these possible ungrammatical forms see
section 2.10. In the form <-ca-k> [cak] (-NEG-COS) the two morphemes are still clearly
distinguishable, but the negative marker <-ca> (NEG) forces the perfective morpheme <-
ok ~ -ak> (COS) to discard its vowel and both of them fuse into a single syllable, i.e. /cak/,
e.g. (10).

(10) gawigamuba olrukancakno. mama manithagamuba olrukancakno.
[gawiga] =mu =ba {ol -ruk -an -ca -k} =no
wife =COM =ADD speak -RC -REF -NEG -COS =QUOT
[mama mani] =tha =ga =mu =ba
father-in-law mother.in.law =OWN =DEREL =COM =ADD
{ol -ruk -an -ca -k} =no
speak -RC -REF -NEG -COS =QUOT
He didnt speak with his wife any more, it is said. He did not speak with his own
father and mother-in-law any more, it is said.

The factitive morpheme <-wa> (FACT) (see Chapter 1) is pronounced without its initial
consonant /w/ when it occurs after a root ending in the labial /m/ or /p/. In these cases the
factitive is homophonous with the customary aspect marker <-a> (CUST). In most cases
the context will provide a clue as to how to analyse certain forms with a suffix which
sounds like [a], e.g. (11) below. In this example we see a suffix <-a> in the verb ram-a
(cook-CUST/FACT). The presence of the locative enclitic <=ci> (LOC), marking the
clause as a Location adjunct (see 27.5), is the clue to the analysis of the morpheme <-a>
as the factitive, since customary aspect cannot occur on predicates of locative-marked
clauses.

2 PHONOLOGY

51
(11) [] gawigaba kumiri romacie, na'lame gudukoknowa.
[gawigaba kumiri] {rom -a} =ci =e [na'lam =e]
wife Pname cook -FACT =LOC =FC type.of.fish =FC
{guduk -ok} -=noa
wiggle -COS =QUOT
[] when the wife Kumiri cooked, the fish wiggled about, it is said.
2.4 Vowels
Table 14 below shows the vowel inventory of the Atong language, whereas the next table
presents minimal and near minimal pairs to demonstrate the vowel contrasts.

Table 14 Vowels
Front Central Back
Close
i
[i ~ ]

U
[u ~ u]
Mid
e
[e ~ c]

[e]
O
[: ~ o]
Open
a
[o ~ a]

2 PHONOLOGY

52
Table 15 Evidence for vowel quality contrast

a i
git music
toy-gat place in the river where
one washes, gets
drinking water and
washes clothes and
dishes
a e
cek fishing net
cak hand
a o
bo-bo liar
ba-ba empty
a u
cu- big
ca who
a
ma- classifier for animals
mo- classifier for spoken
things
e i
ne' bee
ni'- not.exist
e o
dep- to be on top of
something
dok- to take off, take apart
e u
kore bone
khoru wanting to lay an egg
thep classifier for heaps and
small packets
thup nest

e
kore bone
koro sound
i o
ri yam
ro colour
i u
ri yam
ru boat
i
ri yam
ro- to drink
di-goray fish trap
do-da alone
kirin to have a hole in it
koro a sound
cing classifier for bamboo shoots
con- to offer something to a dead
person
o u
ru boat
ro colour
o
ro colour
ro- to drink
u
tu hot
to- to know a fact/someone
2 PHONOLOGY

53
The Atong language has six vowels. The vowel phoneme /i/ is a close front unrounded
vowel [i]. In closed syllables it may be pronounced slightly more lowered and
retracted as []. This phoneme occurs in all positions in the word, i.e. word initially,
medially and finally, e.g. ici here, ri yam, tay'ni today. The only originally
Atong words which start with /i/ are the proximal demonstrative pronoun in its free
form ie (PRX) and those words formed on the bound form of the proximal
demonstrative pronoun i- (PRX), although not all of these are synchronically
analysable as such, e.g. ici here, i-thara-an (PRX-only-FC/ID) only this one, i-sa
(PRX-MOB) thither.
The vowel phoneme /e/ varies freely between a heightened mid-close front
unrounded vowel [e] and the mid-open front unrounded vowel [c]. Examples of words
containing the phoneme /e/ are: sene [scnc ~ sene] seven, khen' [khcn? ~ khen?]
river crab, era [era ~ cra] type of small river fish. Word initial /e/ is always
pronounced more closed as [e], while word final /e/ is pronounced more open as [c].
The vowel phoneme /a/ varies between the open central unrounded vowel [a] and
the open back unrounded vowel [o]. Examples are: hapsan [hopson ~ hapsan] the
same, ray'iari [roj?jari ~ raj?jari] just go!, balwa [bolwo ~ balwa] wind, ato
[at: ~ ot:] what. The phoneme /a/ may be found articulated lower and more to
the back in closed syllables but I have also occasionally heard the lower allophone in
open syllables so we can say that the allophones are in free variation.
The vowel phoneme /o/ varies between [o ~ :], i.e. the back close-mid rounded
vowel [o], with often a quite high pronunciation, and the back open-mid rounded
vowel [:] respectively. The back open mid-rounded variant has a tendency to occur
mostly in closed syllables, but due to assimilation to the vowel in the next syllable
words containing two /o/ phonemes in adjacent syllables speakers often, but not
always, pronounce both phonemes with about the same quality, e.g. ho'o [h:?: ~
ho?:] yes, ortho [ortho ~ :rtho] meaning.
The vowel phoneme /u/ is a high back unrounded or not so very rounded vowel
and may be pronounced a bit towards the mid section of the vowel triangle. Examples
of words containing the phoneme /u/ are jadu [dadu ~ dadu] magic, mudu
[mudu ~ mudu] papaya, ue [ue ~ uc] (DST).
The vowel phoneme /e/, the central unrounded vowel [e], occurs in words like
akrudogol pumpkin, otokoy like that and tho'ok- to have the hiccups, nogol
2 PHONOLOGY

54
market, rogon near, jok spouse, kon back, and dokom head. Let me emphasise
here that /e/ is NOT just a reduced vowel, but a full phoneme in Atong.
2.5 Vowel devoicing and elision
When a vowel appears between two voiceless stops or voiceless affricates or a
combination of the two, the vowel may become devoiced when the following
voiceless consonant is intervocalic. This devoicing and elision may occur in rapid and
normal, and even in emphatic speech. The two consonants that cause the devoicing
may belong to the same syllable or to different syllables within the same word. The
vowel phoneme may even disappear altogether after /s/ and /c/. When the vowel is
lost the preceding /s/ may be lengthened to fill up the gap left by the absent vowel.
The result is a phonetic cluster with long [s:]. Examples of these phenomena are given
in the list here below in phonological form accompanied by their most usual
pronunciation variants.

/sok-a/ [ceka ~ ce ka ~ c:ka] <sok-a> (want-CUST) want
/cok-a/ [tceka ~ tce ka ~ tcka] <cok-a> (cold-CUST) cold
/macot-ok/ [matcetok ~ matctok] <macot-ok> (finish-COS) finished
/caksi khol/ [tcakcikhol ~tcakci khol] <caksi khol> (finger-skin) fingernail
/disu toy/ [dicu tey] <disu toy> (piss-water) piss
/sotha/ [catha ~ cetha ~ ctha~ ce tha ~ c:tha] umbrella

Even when there is no phonetic vowel in the nucleus of the root in the second
example above, i.e. /cek-a/ [tcka] (cold-CUST), the second phonological consonant
may still be lengthened as a means of stress assignment to the first syllable, viz.
[tck:a]. This is how Dilseng R. Sangma starts his story:

(12) ado cokaydonga!
pronunciation: [ado ck:ajdoa]
[a] =do {cok -aydoa}
1s =TOP cold -DUR
I am cold!

The word /gowa/ spider may be pronounced [gawa ~ guwa] or [gwa]. The
choice between [guwa] and [gwa] seems to depend on speed, the latter being
2 PHONOLOGY

55
preferred in fast speech. Whereas /e/ seems to assimilate to the glide in the bi-syllabic
pronunciation [guwa], in the monosyllabic pronunciation [gwa] the /e/ disappears
altogether. This is the only word which I have discovered so far with a stop /w/
onset sequence in which /e/ may elide.
2.6 Vowel assimilation
In roots consisting of more than one syllable, the vowel /e/ assimilates in rounding,
height and position (front-back) to the vowel in the last syllable which I call the main
syllable. This harmonisation mechanism typically affects schwas in syllables of the
CV type in roots with a CV.CVC or CV.CV.CVC syllable structure. The vowel
harmony may be incomplete, i.e. schwa does not have to assimilate completely to the
vowel in the last syllable of the root, which means that there is some variation in
pronunciation of (Ce).Ce.CVC words. Assimilation is fullest when /e/ stands between
two stops that would violate the sonority sequencing principle if they would be
pronounced together in the same syllable as a cluster. Another way of putting it would
be that the sequence of these consonants in their position relative to the main nucleus
violates the sonority sequencing hierarchy (see Lass 1984:264 and Laver 1994: 504).
A good example of this mechanism is the word sokoru river snail. The word
sokoru river snail is most often pronounced [sukeru ~ su keru] although other
possible varieties are also heard. The sequence s-k violates the sonority sequencing
hierarchy whereas k-r does not, hence the vowel between /k/ and /r/ is normally not as
much assimilated to the /u/ of the main syllable as the vowel between /s/ and /k/. All
this may not lead us to believe that the schwas in these types of (CV).CV.CVC words
are only there to break up phonological clusters in Atong. There are no complex
onsets in the language. Moreover the schwas in the kind of words under discussion
contrast phonemically with other vowels, viz. vowels which do not assimilate to the
last syllable of the root e.g.

palo jungle
awoy grandmother (Badri
dialect)
bugorok type of vegetable
bisol coin
2 PHONOLOGY

56
There are many bi-syllabic words with identical vowels which never show any
variation or reduction of their first syllable. These words represent full vowel
harmony in both syllables always. Examples are given below.

siri snow
biri cigarette made of dombol
gasam afternoon, evening
cini sugar
gana LOC be
purun goat
para reed
sorok road
sokhop a cover
samal type of black ant
morot person, man

There are three words in which /e/ does not assimilate to the following vowel, viz.
josori brother-in-law, korotha fathers older sisters child toru- to take a bath.
Furthermore, we have to remark that schwa assimilation also occurs in compounds,
i.e. word internal CV.CVC syllable structures. Examples are given below with their
most usual pronunciation variants.

/kan.to.ra/ [kantara ~ kantera] empty, this word cannot be broken up into
smaller morphemes.
/a.son.tho.lak/ [acenthalak ~ acnthelak] type of fish, this word cannot be broken
up into smaller morphemes.
/a.ko.ru.do.gol/ [akurudegel ~ akerudegel] type of pumpkin, <akoru-dogol>
(pumpkin-?).
/ri'.ko.ran/ [ri?karan ~ ri?keran] testicles <ri-koran> (penis-wings).
/sal.no.ram/ [calnaram ~ calneram] the east, not synchronically analysable
16
.
/ray'a-thori/ [raj?jatheri ~ raj?jathiri] <ray'a-thori> (come-back) come back.
/ci-nora/ [tcinara ~ tcinera] type of lemon, not analysable.
/ga'-solek-/ [ga?cclck ~ ga?celck] to sprain ones foot, not analysable.
/gu-thoni/ [gutheni ~ guthini] pointed stick, spear, not analysable.
/tok-thoni/ [tokthini ~ toktheni] neck <tok-> neck <-thoni> ?.
/wa-dokolo/ [wadokolo ~ wadekelo] bamboo water pipe <wa'> bamboo,
but this word seems to have lost its glottal stop in this obscure
compound.
Two words which are compounds, but of which the constituents are bound
morphemes whose meaning is not clear synchronically, have allomorphs with /e/

16
The word salnoram can be partly analysed with the help of Garo: sal sun, day, no ? (does not
occur as a word in Atong), ram place. The word for east in Garo is salaram.
2 PHONOLOGY

57
which does not assimilate to the following vowel. These words are mokren ~ mokoren
eye and okna- ~ ogona- pregnant.

The bound morpheme mok- eye appears in other words having to do with facial
body parts and some other words:

mokcel- to shine in the
eyes
mokep temple
mokgul eyelid
mokhang face, front
mokjow- to almost fall
asleep
moksep corner of the eye
moksomol eyebrow
moksoram eyelash
moksu- to wash your face
moktoy tear
moksolkhare ring finger
mokthoram middle finger

Particular types of monomorphemic (Ce).Ce.CVC roots in which schwa occurs and
may assimilate to the vowel in the CVC syllable are certainly discernable. These types
are based on particular sequences of consonant phonemes in the onsets of all syllables
in the respective roots and are given below.

Type 1A Stop
vd
+ /e/ + Stop or Affricate
Type 1B Affricate
vl
+/e/ + Stop

Type 1A includes roots such as the following.

bodoy old man
botoy fat of animal, to smell
nice
bothoy porcupine
dokom head
dopow snake
dotoy fathers elder brother
doka type of vegetable
dokham very small wooden
stool
gotho unripe, raw, uncooked
gobe width, breadth
goci to make an angle, be tilted
gocu ladder, stairs
goduk almost
bothu-~
bothow- to boil (intr.) (of water)
gotum ~
gotom village which forms a specific
area with certain other villages

I also include the word gici to make an angle, be tilted in this type although its
second consonant is not a stop but the voiceless affricate. Affricates, however, do
have a stop element in them as we can see if we write the word gici to make an
angle in IPA, viz. [gitci].
2 PHONOLOGY

58
All word initial onsets in the roots above are voiced stops. It is worth noticing that
onset sequences of the type voiceless stop voiced stop or affricate onset do not occur
in Atong except in the word otokoy like that and the perlative/similative phrasal
enclitic <=tokoy> (VIA/LIKE). The word otokoy like that consists of one morpheme.

Type 1B includes roots like the following.

cobom forehead
cokhow nine
cogol type of snail
cogop to fall face down on the ground
coduk to be upside down

Type 2 of the (Ce).Ce.CVC roots consist of a sequence of two consonantal onsets, of
which the first is a continuant. Examples of words of type 2A are given here below.

romot yellow
rogon near
rodom to sprout young leaves
rophi to cover the floor with a
new layer of clay and
cow dung
soki to learn, teach
nogol market
mote deity
monok to swallow, devour
rocok picket
sothi alcoholic rice from which a
beverage is drawn by adding
water
sokop to fold and close
sotha umbrella
polak piece of split bamboo to stir with

Type 2B consists of consonantal onsets, of which the second is a continuant.

boroy four
gorow to shake
goro to meet
koran wing
khorip type of vegetable
khoru wanting to lay an egg
koro horn
joro to dissolve
jorom quietly
porow to have a hole in it
soro web
gowa spider
gowa spider
gosu to cough
tholampak tongue
corom heavy
thombolo to be damaged (of a road or
bridge)

Type 3 consists of roots consisting of three syllables, of which the first two, or all
three, contain a schwa. At least one of the syllable onsets in each root is /r/.

2 PHONOLOGY

59
gorothop type of small plant
sokoru type of river snail
goromak storage rack in the kitchen
korodol hanging root
dokoret to threaten

Note again that onset sequences of the type VOICELESS STOP-SCHWAVOICED STOP do
-vowel not occur, e.g. *tobV, *pogV. When word initial, the phoneme /s/ only occurs
before voiceless stops and /r/, e.g. sora web and soki to learn, teach.
2.7 Vowel phonotactics
The phoneme /a/ is the most versatile vowel of all. This phoneme occurs in all
positions in the root, stem and word.
As was mentioned above, the vowel /i/ occurs word initially only in the proximal
demonstrative <ie ~ i> (PRX). Root or stem final /i/ is rare in Atong. There are two
suffixes with final /i/, viz. the future suffix <-ni> (FUT) and the locative <-ci> (LOC).
The vowel /e/ seldom occurs root or stem finally. The vowel /e/ does not occur
word initially except in two fish names, viz. elo and era, and in loan words.
The vowel /o/ does not occur root or stem finally except in very few words which
are probably loans, viz. salgoro north and balgoto' orchid. The only root ending in
/o/ is no to say. There are four enclitics ending in /o/, viz. the hearsay evidential or
quotative <=no> (QUOT), the imperative emphasiser <=to> (IMPEMPH), the
imperative <=bo> (IMP) and the topic enclitic <=do> (TOP).
The only two words that begin with the vowel /u/ are the distal demonstrative
pronoun <-ue ~ -u> (DST) and u'ci ~ ukci ~ uci leech. The only two roots which
end in /u/ are the kinship terms abu grandmother and acu grandfather. One enclitic
ends in /u/, viz. the allomorph <=mu> of the sequential clausal enclitic <=mu ~ =mu
~ =mo ~ =muna> (SEQ) of which the allomorphs are in free variation. Its
distribution makes /u/ a predominantly word internal vowel phoneme.
The only recorded words that begin with /e/ are om yes, ombo bamboo flute,
ompo lopsided, omoy, an interjection of surprise, ondon in vain, for nothing, and
grammaticalised derivations of the verb otok do like this/that. The phoneme /e/ never
occurs root or stem or word finally.
The VC sequences /ay/, /aw/, /ey/, /ew/, /oy/ and /uy/ do not occur word initially,
only root or stem finally, the only exception is the interjections ayaw, aya and ayu,
which indicate surprise and grief (see 17.6). As mentioned in section 2.2.6, the
2 PHONOLOGY

60
sequences /uy/ and /oy/ are rare in Atong whereas the other vowel-plus-glide
combinations are frequent.
2.8 Morphophonological vowel assimilation
There are tree types of morphophonological assimilation in Atong. The first type is
progressive and obligatory, the second type is progressive and optional, and the third
type is regressive and optional. The first type of morphophonological assimilation
occurs when the vowel of a suffix assimilates to the immediately preceding vowel of
the stem. The perfective suffix <-ok ~ -ak ~ -k> (COS) occurs as its allomorph <-ak>
(COS) directly after stems which end in /a/ or /a?/ (glottalised syllable with /a/), e.g.
(13). If the stem ends in a consonant or another vowel the allomorph <-ok> (COS)
will occur, e.g. (14).

(13) sathiriaymuna umi caythiriciba, ba', matdam sa'akno.
{sa -thiri} =ay =muna [u] =mi {cay -thiri} =ci =ba
put.a.trap -again =ADV =SEQ DST=GEN look -AGAIN =LOC =INDEF
[ba'] [matdam] {sa -ak} =no
interj otter eat -COS =QUOT
When he had put [the fish trap] in place again, then, when he looked again,
the otters had eaten [the fish] again.

2 PHONOLOGY

61
(14) toykhal patwaci ro' rimolaymu ga'sokokaymuna kokce galatokno,
saphawba galatokno.
[toykhal] {pat -wa} =ci [ro'] {rimol} =ay =mu
river cross -FACT =LOC stone slippery =ADV =SEQ
{ga'sokok} =ay =muna [kokce] {galat -ok}=no
stumble.and.fall =ADV =SEQ basket fall -COS =QUOT
[saphaw] =ba {galat -ok} =no
rabbit =ADD fall -COS =QUOT
When they crossed the river, because the stones were slippery, they stumbled
and fell and the basket fell, it is said, and the rabbit also fell, it is said.

As already mentioned, the allomorph <-k> of the morpheme <-ok ~ -ak ~ -k> (COS)
occurs invariably after the negative <-ca>. The first syllable of the
progressive/durative marker <-aydo ~ -ayro ~ aydok ~ ayrok ~ aro ~ arok~
aydoa ~ -ayroa ~ aroa> (PROG/DUR) is pronounced [e] after a stem ending in
the vowel /i/, e.g. sandi-edoa-com (search-PROG =IRR) [they] are searching in
vain.
The second case of progressive vowel assimilation optionally occurs when the
indeterminate noun je whatever, any is marked by the focus/identifier enclitic
<=an> (FC/ID). The /a/ of the enclitic is raised to /c/, which gives the resultant form
je-en [jecn] (whatever=FC/ID) whatever, a certain. The noun and the focus/identifier
enclitic may also fuse to jen [jcn] (whatever.FC/ID) whatever, a certain.
The third type of morphophonological assimilation occurs in compounds when the
vowel of the first syllable of the compound assimilates to the vowel in the next
syllable which is the second member of the compound, e.g. nok house + -khu roof
nukkhu [nuk
h
u]roof. This assimilation is optional as the variant nokkhu
[nok
h
u] roof also occurs. Allomorphs that exist as a result of this type of
assimilation are in free variation.
2.9 Consonant length
There are no phonemic long consonants in Atong. Coda consonants can be lengthened
optionally. This lengthening is one of the ways a speaker can assign stress to a
syllable. Stops, except the glottal stop, and affricates can only be lengthened when
they are in intervocalic position, the other consonants can be lengthened in all
positions. The word jamok <jam-ok> (complete-COS) finished, was alternatively
2 PHONOLOGY

62
written as jammok or jamok by my Atong friends when it appeared at the end of a
story they transcribed for me. This word may be pronounced as [damok], [da:mok],
[damo:k] or [dam:ok], which is sometimes reflected in the writing of my Atong
friends. The same mechanism was reflected in other words like sapa <sap-a>
(to.be.skilled-INF) is skilled which was sometimes written sappa, as its occasional
pronunciation [sap:a] suggests. This form also occurs as [sapa] sapa <sap-a>
(to.be.skilled-INF) is skilled. There are of course no verbal roots *sapp- and
*jamm-; what the writing of my Atong friends reflect is optional lengthening of an
intervocalic consonant. All consonantal phonemes are subject to this optional process,
including the affricates. The word san-san (day-day) every day may be pronounced
[sa n:san] with a long /n/ when the first syllable is stressed. In this word the /n/ is in
preconsonantal position and not intervocalic.
If stress is expressed through lengthening of the consonant, it is always the coda
consonant of the stressed syllable which gets lengthened. If the syllable does not have
a consonant in the coda phonemically, the onset of the next syllable is geminated, so
that phonetically the coda slot of the stressed syllable will be filled and thus the
condition for consonantal stress is met.
When an affricate is lengthened, it affects only the stop element. The word acu
grandfather may be pronounced both [atcu] with a single [t], or [at:cu] with a long
[t]. An example with the voiced affricate is raja-sa (hundred-one) one hundred,
which may be pronounced as [radasa ~ ra:dasa] or [rad:asa]. Affricate lengthening
only occurs when the previous syllable has an empty coda because affricates only
occur syllable initially.
When two of the same consonants are juxtaposed across a syllable or morpheme
boundary, they merge into one and are not pronounced as a long consonant.
Lengthening as a result of stress assignment is optional, e.g. tan-ni (put-FUT) [tani]
will put, taw'pak-khal (bat-cave) [taw
?
pak
h
al] bat cave.
Good examples of the possibility to lengthen either the vowel or the coda
consonant of the syllable are examples (25) and (26) in 2.13 below.

2 PHONOLOGY

63
2.10 Vowel length
Vowel length is not phonologically contrastive in Atong. With some speakers, and
not at all times, vowels in a clearly closed syllable may be pronounced as shorter and
lowered or backed than vowels in a clearly open syllable. This means that sometimes
the word wak pig is pronounced as [wok ~ wak ~ wa:k] and that the word wa tooth
is always pronounced as [wa]. This means that vowel length and allophonic variation
are no indication of the syllable structure in Atong.
When two of the same vowels phonemes are juxtaposed across a morpheme
boundary, three different things may happen in the pronunciation, viz.

1. The two vowel phonemes merge into one short vowel, e.g. <okha-ak> [okhak]
(hungry-COS) very hungry.
2. The two vowels are both pronounced as one long vowel. <okha-ak> [okha:k]
(hungry-COS) very hungry.
3. A glottal stop may be inserted between the two vowel phonemes. <okha-ak>
[okha?ak] (hungry-COS) very hungry.

All these strategies appear to be pretty much in free variation. There is a tendency for
certain speakers to prefer a certain strategy. Another tendency seems to be that the
faster a person speaks, the more likely he is to merge the two vowels. Conversely, the
slower he speaks, the more likely he is to pronounce the two vowels as a long vowel
or to insert a glottal stop.
2.11 Ambisyllabic consonants
We can only know what syllable structures Atong displays when we know its
morphological structure. Most morphemes in Atong are monosyllabic. It is on the
basis of known, productive morphemes that we can know what phonemes can occur
in syllable initial and final position and in which combinations. There are, however,
some multisyllabic morphemes. Polysyllabic morphemes with intervocalic
consonants, i.e. those consonants that can appear in both syllable initial and syllable-
final position, are all ambisyllabic in the sense that the intervocalic consonant cannot
be assigned unambiguously to just one of the syllables. Since there are no morphemes
which serve as indicators of the syllabic make up of multisyllabic roots, suffixes and
2 PHONOLOGY

64
enclitics, we cannot know where their internal syllabic boundaries are. Polysyllabic
roots of this type are gawi [gowi ~ gawi] girl, gotom village, otokoy like this,
pipuk belly, womb, intestines and sene [senc ~ sene ~ scnc] seven. Examples of
polysyllabic suffixes and enclitics are the event specifier <-cora> (TOTALLY,
COMPLETELY), the concomitant action predicate head suffix <-butu> (WHILE),
and the phrasal enclitic <=maran> (TOGETHER).
Since polysyllabic morphemes cannot be divided into smaller meaningful
elements, their internal syllabic structure is not only unknowable but also irrelevant
for the synchronic description of the language. For example, since the lexeme gawi
[gawi ~ gowi] girl cannot be broken up into smaller meaningful elements and the
phoneme /w/ can occur both syllable initially and syllable finally, it is impossible to
establish whether the root internal syllable boundary is before the /w/, i.e. ga.wi or
after the /w/, i.e. gaw.i, or in between to possible /w/ phonemes, i.e. gaw.wi.
The complete consonantal phoneme inventory of Atong is displayed in Table 8.
Morphological analysis shows us that only a limited set of consonants can occur in
syllable-final position. These phonemes are given in Table 13. Thus, if a polysyllabic
morpheme contains an intervocalic phoneme that cannot occur in syllable-final
position, we can say that the syllable boundary of this morpheme lies before this
consonant. An example of such a polysyllabic morpheme is acu [atcu ~ at:cu]
grandfather. The allophone [at:cu] with the long stop element of the affricate /c/ is
the result of the rule of optional consonant lengthening explained in 2.9, where it is
stated that all intervocalic consonants can be lengthened for reasons of stress and that,
when the affricates are lengthened, it affects only the stop element. The phoneme /c/
cannot occur syllable finally, so if we for some reason had to syllabify this root, the
phonotactics as they occur on monosyllabic morphemes would be an argument to say
that the syllable boundary in the polysyllabic morpheme acu grandfather lies before
the /c/, i.e. a.cu. The same monomorpheme-based phonotactics would prompt us to
argue that the syllable boundary in the first person singular morpheme aa (1s) lies
after the //, i.e. a.a, since this morpheme cannot occur syllable initially.
It is phonotactically impossible for Atong syllables to end in a consonant cluster,
so in polymorphemic roots with intervocalic consonant clusters, it is possible to
assign the consonants to either the coda of the first syllable or the onset of the second
syllable. Examples of such morphemes are makbul bear, matdam otter and
2 PHONOLOGY

65
moks(o)ra eyebrow. In these examples, the syllable straucture has to be mak.bul
bear, mat.dam otter and mok.s(o.)ra eyebrow, where the last word can
optionally contain an extra syllable because non-first syllable consonant clusters with
/r/ as second element can be broken up by the incertion of a schwa (see 2.1).
As mentioned above, the syllabification of morphemes like this is irrelevant for
the synchronic description of Atong, since they are not dividable into smaller
meaningful elements. However, monomorpheme-based phonotactics help us to
explain why certain consonantal sequences, i.e. *VpV or *VcnV never occur in
polysyllabic morphemes, the reason being that synchronically unanalysable
polysyllabic morphemes obey the same phonotactic rules as monosyllabic ones. This
phenomenon can be explained diachronically when one argues that polysyllabic
morphemes are the product of fossilised combinations of monosyllabic ones. I will
not, however, at this time argue in favour or against this explanation, since internal
reconstruction and historical comparison of Atong lies outside the scope of this thesis.
2.12 Glottalisation
Glottalisation, or glottal prosody, in Atong is a feature that operates on the level of the
syllable, and that manifests itself as a glottal stop at the end of the syllable.
Glottalisation only affects open syllables and syllables ending in a continuant. A
prosodic feature that only effects part of the syllable is called syllable-part prosody
(Lass 1984: 244). In this grammar, glottalisation is written phonemically as a glottal
stop at the end of the affected syllable.
17

An open glottalised syllable ends in a glottal stop which is not released, e.g. ca'
[tca?] leg/foot.
18
There are two phonetic possibilities for the release of a glottalised
continuant, depending on the following phoneme.

17
In the orthography of Atong, glottalised syllables are written with a raised dot or apostrophe after the
affected syllable, as explained in 1.5.
18
Lexical morphemes ending in a glottal stop do NOT have an echo vowel after the glottal stop as in the
Achik [a?tcek] dialect of Garo (see Burling 2004: 34-35). Thus the word ca' leg/foot is pronounced
[ca?] and the bare imperative of the verb ra'- to get is ra' [ra?] Get [it]!
2 PHONOLOGY

66
If the glottalised continuant is followed by a consonant, the glottalised phoneme is
not released, i.e. man'-khu-tca [man
?
.khutca] (be.able-INCOM-NEG) is not yet
possible.
If the glottalised continuant is followed by a vowel, it is released and the release
repeats the continuant so that it can be said to act like the onset of the following
syllable, e.g. man'-ok [man
?
.nok] (be.able-COS) was able.

In a glottalised syllable with final /l/ the glottal stop usually precedes the oral closure
of the [l] when followed by another vowel, e.g. mel'-a [me?.la] (be.fat-CUST) is fat.
This phenomenon also happens, but less frequently, with syllables ending in /m/, e.g.
no'm-a [no?.ma ~ nom
?
.ma] (be.soft-CUST) is soft.
As a result of the feature of repetition of an intervocalic glottalised continuant,
some suffixes and enclitics are not phonetically differentiated on certain words the
root or stem of which ends in a glottalised consonant. This means that a word which
sounds like [tan?na] may, according to the context, be analysed morphologically as
<tan'-a> (cut-INF) cut or <tan'-na> (cut=DAT) in order to cut. Another example is
a word that sounds like [saw?wa] which may be analysed as either <saw'-a> (burn-
INF) burn or as <saw'-wa> (burn-FACT) burned.
When a stem-final glottalised continuant is followed by a suffix or enclitic
beginning with the same continuant, the phonetic effect is the same as that of a
repeated intervocalic glottalised continuant. This can be exemplified thus: /man'ni/
[man?ni] <man'-ni> (be.able-FUT) will be able.
There are two alternative analyses of the occurrence of the glottal stop, viz. a
glottal stop phoneme or a series of glottalised continuants, which will be treated
below and which are not favoured over glottalisation as a prosodic feature. An
argument in favour of the glottal prosody analysis is that it simplifies the phonological
analysis of the language. There is no need for either a glottal stop phoneme with a
restricted occurrence or for a series of glottalised continuants. Just like a series of
glottalised continuants, glottal prosody solves the problem of non-canonical syllable
final consonant clusters. Glottal prosody can also account for the phonetic behaviour
of affected continuants as described above.
Another argument in favour of glottal prosody at syllable level comes from
languages closely related to Atong, namely Tiwa, Bodo/Boro and Rabha. What has
2 PHONOLOGY

67
been observed for Garo (Joseph and Burling 2006: 21) is also valid for Atong.
Glottalised syllables in Atong correspond to syllables with a high tone in Tiwa,
Bodo/Boro and Rabha in words with similar phonetic/phonological make up and
meaning.
19
The distribution of glottalised syllables in Atong differs from Garo. In
Atong any syllable within a grammatical or prosodic word can be glottalised.
Glottalisation prosody as occurs in Atong has been described for a number of
Australian languages in the central north of the continent (Dixon 2002: 616-8). In two
sets of these languages glottal articulation is found only at the end of a syllable after
a vowel or continuant (never after a stop) (ibid.), similar to Atong.
20

Glottalisation in Atong happens mostly at morpheme boundaries but there are
some instances of glottalisation in medial position in multisyllabic unanalysable
words. Maybe there once was a morpheme boundary where the glottal prosody
occurs, before the putative compound became non-transparent, e.g. ca'ma lower
side, downstream, me'ma ghost. Table 16 below shows some minimal pairs of
syllables with and without glottal prosody.

Table 16 Minimal pairs of syllables with and without glottal stop
Plain Glottalised
si- to peel si'- to sharpen a pointy object
nepal Nepali, Nepalese, Nepal ne'kat type of bee (ne' bee + kat ?)
ca tea ca' foot, leg
na- to hear na' fish
susot- to wash (something) su' vagina
rimola slippery ri' mol-a (penis small-CUST) the penis is small
wal night wal' fire
ro colour ro' stone
man- to crawl man'- to be able, to achieve
ram- to dry ram'- to search
toy water thoy' blood
taw to go up taw' chicken, bird

19
Although no internal reconstruction of any of the languages cited has been attempted or published
yet, the plethora of correspondences between lexical items in terms of phonetic/phonological make up
and meaning is striking.
20
What is described as glottalisation in Cherokee (Iroquoian, Oklahoma and North Carolina) by
Scancarelli (1992: 139) is generally realized as either a preconsonantal glottal stop or as falling pitch
on a lengthened vowel preceding the consonant. This is a different type of glottalisation than in Atong.
2 PHONOLOGY

68
2.12.1 Alternative analyses against glottal prosody.
As was mentioned above, there are two alternative analyses that account for the
occurrence of a glottal stop in Atong, viz. the glottal stop as a phoneme, and a series
of glottalised continuants. These alternatives are not favoured over the analysis of
glottalisation as a prosodic feature, but should not be totally disregarded either.
Arguments in favour of the prosodic analysis have been given above.
i The glottal stop as a phoneme
The glottal stop could be analysed as a phoneme that can only occur in syllable-final
position after a vowel. The glottal stop would then also occur word internally in
syllable-final position as in the words co'sa [tco?sa] a little bit, co'mot [co?mot]
really, te'ew [te?ew] now, ne'+kat [ne?kat] (bee+?) type of bee. There are
several features that make the glottal stop different from all the other consonants
phonemes. As mentioned before, the glottal stop only occurs syllable finally in
postvocalic position. In case a syllable ends in a consonant other than the glottal stop,
the preceding vowel may get shortened and or lowered or is articulated more to the
back. Thus with most speakers most of the time there is a difference in pronunciation
between wa [wa:] bamboo and wak [wok] pig. The effect of the glottal stop on the
preceding vowel is not the same as that of all the other syllable final consonants.
Before a glottal stop lowering or back articulation and shortening of the preceding
vowel never occurs. The distinction between the words wa [wa] tooth and wa' [wa?]
bamboo for example, is therefore only the presence or absence of the glottal stop.
The word ne' bee is never pronounced as [nc?] but always as [ne?] and the word na'
fish always as [na?]. The difference between the previous examples and the word
wak [wok] pig is that in the word wak [wok] pig the vowel may be shorter and
articulated more to the back. Alternative articulations of vowels in closed syllables are
in free variation and differ from speaker to speaker and even from instance to instance
with the same speaker. This means that the pronunciation [wak] for wak pig has also
been recorded.
Another indication that the glottal stop is not like the other stops in the
phonological system of Atong is the fact that it does not prevent morphophonological
vowel assimilation of the vowel of the change of state suffix <-ok ~ -ak ~ -k> (COS)
to the preceding vowel. This preceding vowel is always /a/. The change of state suffix
2 PHONOLOGY

69
<-ok ~ -ak ~ -k> (COS) occurs as its allomorph <-ak> (COS) directly after stems
which end in /a/ or /a?/, e.g. taw' ra'-ak (chicken get-COS) [I] bought a chicken. If
the stem ends in any other consonant the allomorph <-ok> (COS) will occur, e.g. taw'
tokok (chicken beat-COS) [I] beat a chicken. Assimilation is possible across a
syllable boundary with an intervening glottal stop, but not across a syllable boundary
with any other consonant.
ii Glottalised continuants
If one were to analyse the glottal stop as a separate phoneme, there could be two
compelling reasons to postulate a series of glottalised consonants in addition.
Glottalised continuants have been described for the North American languages
Navaho, Nootka and Kwakiutl (Sapir, 1938/1951). The first argument is the canonical
(C)V(C) syllable structure. If one would analyse /l?, m?, n?, ?, w?, y?/ as a cluster of
consonant plus glottal stop, the odd situation occurs that there are no other syllable
final consonant clusters except those of a consonant plus glottal stop. This is of course
awkward and may be easily avoided by adding a series of glottalised segments to the
phoneme inventory.
The second and equally compelling reason to posit glottalised phonemes is the
phonetic behaviour of these glottal segments. When a glottalised consonant occurs in
between two vowels, the consonant is, as it were, doubled, continuing after it has been
stopped by the glottal stop, e.g. /tan'aribo/ [tan?naribo] <tan'-ari-bo> (cut-SIMP-
IMP) just cut!, ramaydok [ram?majdok] <ram'-ay-dok> (search=ADV-PROG)
searching, /ray'ani/ [raj?jani] <ray'a-ni> (come-FUT) will come. First there is the
phenomenon of simultaneous glottal and oral closure after which the consonant is
released from the same place of articulation into the following vowel. Thus
glottalisation affects the syllable-final consonantal segment such that it stretches over
the onset of the next syllable. The glottalised consonant /l?/ behaves phonetically
different from the other glottalised consonants in that it often happens that the glottal
closure precedes the oral one so that the oral closure acts as onset of the next syllable,
e.g. gawi mel'-a [gawi me?.la] (girl fat-CUST) the girl is fat. Glottalised continuants
that are followed by another consonant are unreleased.
2 PHONOLOGY

70
2.12.2 Conclusion
If we were to posit a separate glottal stop phoneme, it would have three
disadvantages, viz. its atypical phonetic effect on vowels in closed syllables, its
restricted occurrence, and the appearance of consonantal clusters in codas with
continuants, which violate the canonical CVC syllable structure of the language. In
order to avoid this violation, we could postulate a series of glottalised continuants.
However, these glottalised continuants would then behave phonetically different from
other consonants. Glottal prosody can account for the phonetic behaviour of affected
continuants. Glottal prosody, as seen in 2.12, solves the problem of non-canonical
syllable final consonant clusters, and it simplifies the phonological analysis of the
language. We can conclude that, of the three analyses advanced above for dealing
with the occurrence of a glottal stop in Atong, viz. a glottal prosody, a glottal stop
phoneme, and a series of glottalised continuants, the glottal prosody analysis is the
simplest and most suitable one.
2.13 The Atong word
The phonological word in Atong is usually, but not always, characterised by a low
pitch on the first syllable. The grammatical word is that form which can occur on its
own as constituent of a clause. Other than these, I have found no clearcut criteria on
which to distinguish phonological and grammatical word in Atong. Properties of the
phonological word listed in Dixon and Aikhenvald (2002: 13) do not work for Atong
or cannot yet be applied because more fieldwork is needed to find out about possible
stess patterns in the language. The verb may be in the same phonological word as the
preceding phrase, but frequently verbs have a low pitched first syllable and can thus
be said to be phonological words on their own. Classifiers followed by numerals will
be in the same phonological word as the preceding noun. Maybe sentence intonation
patterns are interacting with word intonation patterns, which is a very complicated
issue and would require a more thorrow understanding of the language and more
fieldwork research to figure out than the time which is provided to write this PhD
thesis.
As far as I am able to judge, there seems to be no syllable timing mechanism that
determines the make-up of a phonological word. In other words, I have not been able
2 PHONOLOGY

71
to discover obligatory stress patterns in the language, such as iambic or trochaic. The
minimal phonological words can be just one syllable in Atong.
2.14 Accentuation, stress and prosody
The following intonation symbols will be used in this section:

v
higher pitch than previous syllable

v lower pitch than previous syllable
v
low pitch
v falsetto voice
v same pitch as previous syllable
v rising intonation
v higher intensity or amplitude
Length symbols : are phonetic and not of phonological importance.

In Atong the realisation of a syllable may be influenced by one of the following
features:
- An increased intensity or amplitude.
- Length.
- A difference in pitch. I will distinguish low and higher pitch.
- Extra high pitch characterised by falsetto voice, indicated by the symbol .

None of these features is phonemic and all may occur optionally. Increased intensity,
length and extra high pitch are means to stress the syllable. This means that Atong
presents phonetic stress. I adopt the definition stated by Van Der Mark that phonetic
stress is a method of marking prominent syllables that may involve several acoustic
variables such as pitch, loudness, duration and vowel quality. [...] Crucially, phonetic
stress cannot be marked by pitch alone. (2003:21). In Atong vowel quality does not
change in stressed or unstressed syllables. Only in unstressed syllables containing the
vowel /e/ can this vowel assimilate to the vowel in the next syllable, as has been
discussed above. Extra high pitch or falsetto voice in Atong is always accompanied by
extra lengthening of the vowel and is a phonetic means to mark intensity. In (15) we
see an example of sentence with a falsetto syllable.

2 PHONOLOGY

72
(15) ndgiim d, drk.
{ne -duga} =ay =mu [a] =do {do -arok}
tired -XS =ADV =SEQ 1s =TOP be.enough -PROG
Because I am too tired its enough.

In (15) above, the extra high pitch characterised by falsetto voice is a means to
emphasise the fact that the speaker is really very tired. Hence the excessive suffix <-
duga> (XS) is stressed. If the speaker wants to emphasise the fact that a certain train
of events had to be completed in order for another event to take place, he can stress
the sequential clausal enclitic <=mu ~ =mo ~ =mu ~ =muna> (SEQ) as we can
see in the next example where the first occurrence of <=mo> (SEQ) is stressed.

(16) otokoymo ko
ns [phaltaaw] cnkgbw n'ym
, alsi rj :nh,
otokoy clie cliso
mcydok noymo te'ewb, jokmo jlydkn.

otokoymo [konsa] [[phaltha =aw {conuk =gaba =aw
so.then later self =ACC criticise =ATTR =ACC
{na} =ay =mo [alsia raja] [na] [aa] [otokoy] {coli =e
hear =ADV =SEQ lazy.person king excl 1s like.this succeed =TOP
coli -som -ca -aydok} {no} =ay =mo [te'ew] =ba
succeed -CERTAINLY -NEG -PROG say =ADV =SEQ now =EMPH
[jok] =mo {jal -a -ok} =no
spouse =GEN run -AWAY -COS =QUOT
So then, after hearing those who criticise him, the lazy king [said]: Nah! as
far as succeeding is concerned, I have certainly not succeeded, he said [and]
he runs away from his wives.

Differences in pitch between syllables in Atong may be small and subtle. Most often
all word classes receive a low pitch on the first syllable followed by higher pitch on
all other syllables. In a sentence, numeral-classifier phrases do not have to start with a
low pitched first syllable but may just continue on the same pitch as the last syllable
of a preceding word. Interrogatives and numeral-classifier phrases are never stressed
on the first syllable. Other word classes may just receive stress on any syllable which
the speaker finds most noteworthy. It is possible for two or more consecutive syllables
to be stressed. An example of this phenomenon is (17) below where the word
cu=gaba (big=ATTR) which is big has higher amplitude on the first and second
2 PHONOLOGY

73
syllable and the first syllable has a higher pitch than the second syllable the nucleus of
which has been lengthened to great extent.

(17) uc pholom cga:ba, pho
lgo
m cgb, d'ty tanaokno.

uci [pholgom {cu} =gaba] [pholgom {cu} =gaba]
then type.of.bird big =ATTR type.of.bird big =ATTR
{di'it} =ay {tan -a -ok} =no
defecate =ADV put -AWAY -COS =QUOT
Then a really big eagle, a big eagle, left his shit behind [on the drying
greens].

When nouns are stressed on the first syllable they will still have a low pitch. When a
word is pronounced in isolation it generally starts with stress and a low pitch on the
first syllable and then the intonation rises all through the word, which will then
generally end with much higher intonation on the last syllable, e.g. jbk curry,
thmo
l type of edible root. Any other syllable of the word may also receive stress,
which will then be marked by a high pitch, after which the intonation may go down or
even further up. It most often happens that the first syllable of a word has the low
pitch but no intensity and no length and then the second syllable has high pitch. In
summary, word intonation goes from low to high, with the first and the last syllable
being the most prominent ones, the first syllable because of its low pitch and the last
syllable because of its high pitch, except when the last syllable is the quotative clausal
enclitic <=no> (QUOT) (see below). If other than the first or the last syllable is
stressed, this stressed syllable can have a higher or lower pitch than the last syllable
but always higher than the first syllable, and moreover the stressed syllable will have
the greatest length, which will be either expressed on the vowel nucleus or on an
intervocalic consonant (see 2.5 and 2.9).
This word-prosodic pattern is also the general pattern of prosody we find in the
sentence. A sentence will start in general with a low pitch on the first syllable and
then the intonation rises all though the first word. All subsequent words, especially
non-verbals, will generally have a low pitched and stressed first syllable again. Most
of the rising intonation of the sentence occurs on the final verb which may start at the
same pitch as the last syllable of the preceding word and no stress on the first syllable
or it may have a low pitch on the first syllable and rising intonation from there. This
2 PHONOLOGY

74
intonation pattern is the same for subordinate and main clauses. This means that
generally after a stress peak on the last stressed syllable of the subordinate clause, the
following clause starts on a lower pitch.
Another possible sentence intonation pattern is that there is a rise in the first part
of the sentence, and then after the most important word according to the speaker, the
intonation goes down again to the end of the sentence This is particularly frequent in
telling events, i.e. non-quotation parts, in story telling. Low pitched stress is an
optional property of the first syllable of a word, and the high pitched stresses are
conditioned by the speaker. A good example of a stretch of speech with intonation is
given here below.

(18) c, kvnkhlthgbd san:sann dbtwrs dgry sn r'roa
noro. o
to
koym dgry sakno. ko
nsd, mn: pmi srmonmon

r'aymona, dbtwrs, dgry skno.
ucie [kynokhol] =ha =gaba =do [san san] =an
then son-in-law =OWN =ATTR =TOP day -day =FC/ID
[dabat wari] =sa [digaray] {sa} =na
Pname deep.section.in.river =MOB fish.trap set.as.trap =DAT
{re'e -ro -a} =no =ro
go.away -USUALLY -CUST =QUOT =EMPH
[otokoymu] [digaray] {sa -ak} =no}
so.then fish.trap set.as.trap -COS =QUOT
[konsa =do [manap =mi sirimonmon]
later =TOP morning =GEN break.of.dawn
{re'e} =ay =mona [dabat wari] =sa
go.away =ADV =SEQ Pname deep.section.of.river =MOB
[digaray] {sa -ak} =no
fish.trap set.as.trap -COS =QUOT
Then, the son-in-law went to Dabatwari every day to put up his fish trap. So,
he has put up his fish trap. Later, very early in the morning at the break of
dawn, having gone to Dabatwari, he managed his fish trap.

In the next example we can clearly observe how the pitch of the first syllable of the
subordinate verb py-ca-ay-mo (bear-NEG=ADV=SEQ) not bear is higher than
that of the preceding syllable and that the overall intonation of the verb goes from
high to low so that the subordinate clause ends on a low pitch. All of this may mean
2 PHONOLOGY

75
that there are no uniform prosodic criteria to distinguish between different words and
different clauses in a sentence.

(19) cs mcn mkbln mmn pycaaymo, bo
ldo
bolda jln
h'bceok.
uci =sa [maca] =na [makbul] =na [moma] =na
then =DLIM tiger =DAT bear =DAT elephant =DAT
{pay -ca} =ay =mo [boldobolda] [jal =na] [ha'bace -ok]
bear -NEG =ADV =SEQ all.over.the.place run.away =DAT begin -COS
Then the villagers did not bear the tigers, bears and elephants any more and
started running away all over the place.

Interrogatives never receive high pitch or high amplitude on the first syllable, The
name of the language under discussion is pronounced [at:] with higher pitch on
the second syllable.
The quotative enclitic <=no> (QUOT) and any following enclitic are seldom
stressed. Normally there is a sharp drop in pitch and intensity on the quotative enclitic
<=no> (QUOT). However, the most powerful stress of the whole word may be shifted
to the quotative enclitic <=no> (QUOT) or any enclitic that comes after <=no>
(QUOT) as in (20) below. In that sentence the first syllables of all the words are
pronounced with a low pitch, the rest of the word has a higher pitch that the first
syllable and the quotative enclitic is stressed and has the highest pitch.

(20) wb s'nckn. gwigmn lrknckn. mm mnithgmb
lrknckn.
[u] =aw =ba {sa' -an -ca -k} =no
DST=ACC =EMPH eat -REF -NEG -COS =QUOT
[gawiga =mu -na] {ol -ruk -an -ca -k} =no
wife =COM =DAT speak -RC -REF -NEG -COS =QUOT
[mama mani] =tha =ga =mu =ba
father-in-law mother-in-law =OWN =DREL =COM =ADD
He didnt eat that any more, it is said. He didnt speak to his wife any more, it
is said. He didnt speak to his mother-in-law any more either, it is said.
2 PHONOLOGY

76
Nouns positioned after the main verb of the sentence can occur in the same prosodic
sentence as the main clause or as an afterthought with an overall lower pitch than the
main sentence.
The question enclitic <=ma> (Q), when occurring on a predicate head, may
receive a separate high pitched stress even though the preceding syllable also has a
high pitched stress, e.g. (21) below.

(21) r'ydm.
{re'e -aydo} =ma.
go.away -PROG =Q
Are you going?

Question clauses may also lack the question enclitic <=ma> (Q). There are no special
interrogative intonations for interrogative sentences or clauses. As for interrogative
clauses without interrogative pronouns and without the question enclitic <=ma> (Q),
the fact that the clause is a question has to be deduced from the context. If a speaker is
really surprised about something, the sentence will have a higher overall pitch and
rising intonation. In (22) a son asks his mother why her cooking is so good today and
the mother replies. Then in (23) the same son asks or exclaims his surprise about the
reason why the food is so tasty, to which the mother gives a strong affirmative
answer. In (24) we observe two question sentences with level intonation on every
syllable except the last one, which is higher in pitch.

(22) :ttokoy ty'nd thwks
y ja'bek?, nokn. t doww

m?nokn. h? n'wt bb [...]
[ato] =okoy [tay'ni] =do {thaw -ok} =soy [ja'bek]
what =LIKE today =TOP tasty -COS =MIR curry
[ato] {dow -wa} [ama] [ho] {ni' -wa} =te [baba]
what add -FACT mother interj not.exist -FACT =DCL son
Why is the curry so tasty today? he said, it is said. What did you add,
mother? he said, it is said. Huh? Nothing, [my] boy!

2 PHONOLOGY

77
(23) pholgom di'itdapay tanawa?! nookno. ho'o nookno.
[pholgom] {di'it -dap} =ay {tan -a -wa} {no -ok} =no
eagle shit -ON.TOP =ADV put -AWAY -FACT say -COS =QUOT
[ho'o] {no -ok} =no
yes say -COS =QUOT
A very big bird shat [in the curry]?! he said, it is said. Yes. she said, it is
said.

(24) atkn korew? mrtmdrn?
[takna] {kore -wa} [morot] =ma' =dara =na
why fear -FACT person =interj =p =DAT
Why are you afraid? Because of the people? [the fox says to the other
animals].

Exclamatory sentences are characterised by an overall higher amplitude and
optionally extra stress on the last syllable of the sentence. The overall intonation of
the sentence may rise as in (25) but may also rise and then fall as in (26).

(25) c:'msmiw cyc!
[ca'masa] =mi =aw {cay -ce}
down.side =GEN =ACC look.at -FIRST
Look at [your] lower side first!

(26) c':msba cayok na'!
[ca'msa] =aw =ba {cay -ok} [na'a]
lower.side =ACC =EMPH look.at -COS 2s
I looked at the lower side, oh you!

The above examples (25) and (26) form a nice minimal pair to show that in a stressed
syllable it is either the vowel or coda consonant which gets lengthened. Example (26)
is also a good example to show that the glottal stop is susceptible to lengthening in
stressed syllables and that it can be lengthened in preconsonantal position.
2.15 Phonologically aberrant words
There are a number of words which show phonological anomalies. One of these
words is the negative proclause /hm?m/ [m ?m] no. This word has no vowels, is
2 PHONOLOGY

78
spoken with the mouth shut and has a voiceless nasal as first consonant, a sound
which is found nowhere else in the language. Exclamations may end in strong post
vocalic aspiration, e.g. hah! exclamation of satisfaction. No other words in Atong
end in /h/. Furthermore some exclamations end in /e/ like the mirative exclamation
ho?! Huh?!. No other words in Atong end in /e/.
Onomatopoeia are sometimes phonologically aberrant. The sound a goat makes is
pronounced with a strong and long trilled /r/ in syllable-final position. The sound is
dur:meme. The onomatopoeia denoting someone smoking vigorously is krrrrrrr with
a long syllabic trilled /r/.
As said in Table 8 the word for to have the hiccups, tho'ok-, is phonologically
aberrant. It is the only word in which the sequence /e?/ occurs. The word ho'o yes
is phonologically aberrant in that is the only word in which the sequence /o?o/ occurs
in a root. However, the form ho'o meaning yes is not at all aberrant when one takes
a cross linguistic perspective, as in Parker (1996). In this article Parker exhibits a
template or canonical pattern with the shape /he?(e)/ which he proposes to posit as
the default form for yes. The nasal // in the Atong word ho'o is explainable
through rhinoglottophilia (Matisoff 1975).
A full analysis of all phonologically aberrant words is a matter for a separate
investigation, and will therefore not be pursued in this grammar.
2.16 The phonology of loan words
This section deals with the most salient features of loan words in Atong. Atong has
many loans from Garo, English and Indic languages, viz. Bengali, Assamese and
Hindi. Since it is not always possible to establish the exact origin of an Indic word, all
loans from these languages will be termed Indic loans.
2.16.1 Vowels
Atong has six vowels occurring in indigenous as well as loan words. In addition, there
are four vowels which are only found in loanwords from English and Indic languages.
These are the so called loanvowels (see below), which are usually, but not always
pronounced longer that the indigenous vowels. The loanvowels are represented in
Table 17. In the orthography they are simply written double. Note that // (u with
macron) and schwa with macron are not attested.
2 PHONOLOGY

79
Loanvowels are usually but not always pronounced long, and when they are not
pronounced long, the difference between the loan and the indigenous words is a
matter of vowel quality. In closed syllables, where Atong vowels would be
pronounced lowered and more retracted, the loanvowels will have the same quality as
the Atong vowels in open syllables. Not all loan words that have long vowels in the
source language have long vowels in Atong, and not all loans that can be pronounced
with a long vowel in Atong have a long vowel in the source language.

Table 17 Loanvowels
Front Central Back
Close
[i: ~ i]

Mid
[e: ~ e]

[o: ~ o]
Open
[a: ~ a]

Examples of minimal pairs and near minimal pairs are given in Table 18. Although
long vowels are only found in loans, not all loans contain long vowels, as we can see
in the first minimal pair. The word tin corrugated iron is an English loan without
long vowel, which contrasts phonologically with the Indic loan tin three, which does
contain a long vowel.

Table 18 Minimal and near-minimal pairs of words with and without loanvowels
Without loanvowel With loanvowel
tin [tin ~ tin] corrugated iron tin [ti:n ~ tin] baji three oclock
pel- [pcl] to copulate pel [pe:l ~ pel] do'ok failed
mat [mot] wild animal t [a:t ~ at] baji eight oclock
(see Table 44)
ret [rct] childrens game rel [re:l ~ rel] train
ri'gol [ri?g:l] penis (as swearword) gl [go:l ~ gol] sa'ak got a goal

2 PHONOLOGY

80
I am not familiar enough with the Indic languages Hindi, Assamese and Bengali,
which are all possible sources for the Indic loans we find in Atong, to give a precise
description of the vowel changes that occur when words from these languages are
borrowed into Atong. As far as I can judge, Indic vowels undergo very little change in
the process of borrowing. Indic differences between /e/ and /c/ and /o/ and /:/ are
merged into Atong /e/ and /o/ respectively. Any Indic nasality is discarded in Atong.
Words borrowed from English undergo very intricate, sometimes seemingly
random vowel changes so that they cannot all be described in detail in this grammar.
Especially English schwa has a wide variety of pronunciations in Atong. I will just
give a few examples. Vowel harmony plays a role in the sense that English schwas
can be assimilated in Atong to a neighbouring vowel, e.g. English government >
Atong: gobormen. Some English schwas can be pronounced with different Atong
vowels in free variation, e.g. English November > Atong nobembol ~ nobembol.
Sometimes the English schwa becomes a totally different vowel in Atong, e.g. English
first > Atong: phas. The English word August is borrowed by Atong as agos, while
English October is oktobol ~ oktobol in Atong, reflecting two different results of the
English vowel /:/.
2.16.2 Consonants
Atong changes all loan words in such a way that they fit the Atong consonant
inventory and sound system. There are very few exceptions. Loans from English,
Indic languages and Garo will be treated separately. A list of English loans in Atong
can be found in the appendix of van Breugel (2009 a).
i Loans from English
The most salient sound changes that occur when English words are borrowed into
Atong are the following: (English > Atong) f > ph, v > b, 0 > t and > r ~ l. English
/s/ and /J/ collapse together in Atong into /s/. Examples of these changes are:

English Atong
English /f/ to Atong /ph ~ p/ officer > ophisor ~ opisor
English /v/ to Atong /b/ government > gobormen
English /0/ to Atong /t/ thirty > torti
2 PHONOLOGY

81
English // to Atong /r ~ l/ radio > redio
September > septembyl
English /J, tJ/ to Atong /s/ national > nesonol
wrench > rens

Moreover, word final clusters are simplified, e.g. English: licence > Atong: laysen,
English: government > Atong: gobormen. Word initial clusters are usually broken up,
e.g. English: clip > Atong: kilip ~ kylip, English: glass > Atong: gilas ~ gylas. The
English word blue is borrowed into Atong as bolu ~ blu blue with two syllables by
some speakers and with initial cluster by others. However, skul is always [skul]
school and never *[sekul]. This phenomenon may reflect different chronological
layers of loans. An exceptional, new loan word, that breaks the rules of Atong word
formation, because the affricate /c/ appears syllable finally, is the English loan ingec,
engage, which can only be used with the support verb kha'- to do (see 22.7), e.g.
ingec kha'-ak ni=do (engage do-COS 1pe=TOP) We are engaged to be married.
ii Loans from Indic languages
The most salient sound change in words of Indic origin is the collapse of the
retroflex trills and taps into one phoneme /l ~ r/ and the loss of the distinction between
dental and retroflex consonants. Aspiration is not borrowed consistently but varies
freely with non-aspirated consonants in Atong, e.g. Atong: kata ~ khata ~ katha ~
khata word, corresponding to Hindi (kath) story, talk.
In words of Indic origin the phonological distinction between /l/ and /r/ is
neutralised. Whenever a syllable final /l ~ r/ occurs it can be pronounced as [l ~ r~ r].
Some instances of /r ~ l/ in Atong come from the retroflex series in Assamese and
Bengali. I recorded the word golmal chaos, quarrel, fight, dispute pronounced as
[g:rmal], sometimes [g:lmal]. Of the following word, meaning axe, three variants
have been recorded, viz. kulal ~ kular ~ kural. The word isor ~ isol God, is an Indic
loan, cf. Sanskrit asura god, evil spirit, Hindi asur demon, evil spirit, though Garo
isol God, the meaning God being undoubtedly imposed by the missionaries that
converted the Garos to Christianity. In Garo it is invariably pronounced as [isol], but
in Atong the final phoneme is in free variation with /r/. The word papol ~ papor food
2 PHONOLOGY

82
made of wheat comes from Bengali ppod. Some speakers in the Badri area
pronounce this word with a very long and strong [r:], i.e. [papor:] maybe to accentuate
their awareness that it is a loan. Another example of this phenomenon is the word
bagal ~ bagar Bangladeshi, Indic person which is often pronounced as [bogor:]
in Badri. This strongly vibrating of /r/ can also appear word internally as in the word
hagersak everything, the world.
This phenomenon of syllable final [r ~l] variation is a good indication that a
certain word is not of Atong origin. The word ha'golsak everything, the world looks
very much Atong but is in fact a Garo loan which may be detected though the variant
hagorsak with syllable final /r/. On the contrary, some English loans ending in /r/ are
ostentatiously pronounced with final [l], e.g. mastel male teacher There are some
words in which the two allophones of final /l/ are never mixed up, viz. skul school
and sendel sandal, both English loans. An example of an English loan with /l ~ r/ in
free variation is kabar ~ kabal cover, lid. It can be presumed that early English loans
conform more to Atong phonology that later ones. Thus skul school, mastel male
teacher and sendel sandal are early English loans and kabar ~ kabal cover, lid is a
late English loan.
iii Loans from Garo
Although the Garo and Atong phonological systems are very similar, there are subtle
differences that can indicate Garo loans in the Atong lexicon. For a description of
Garo phonology I refer the reader to Burling (2004). Garo loans other than those with
/l/ do not present any problem to the Atong phonology and are taken into the language
without alternation. Even complex syllable onsets which are not split up in Garo may
be kept in tact in Atong, e.g. grok- [gr:k] and not *[ger:k] classifier of gulps, i.e.
amounts of liquid drunk at a time.

There are other modifications that loans can undergo to make them fit in to the Atong
sound system, but there are so many English and Indic loans borrowed at different
stages into Atong that describing them all in detail lies beyond the scope of this
grammar.

83
Chapter 3 Word Classes: an overview
_____________________________________________________________________

Word classes in Atong share properties. The most salient overlapping properties of the
major word classes in Atong have to do with the possibility to function as predicate
head. Sometimes the differences between word classes can be very subtle or even
fuzzy.
Atong distinguishes the word classes listed in Table 19. This table also indicates
whether a member of a word class can function as head of a predicate or not and
where the word classes are treated in this grammar. Table 20 gives an overview of
some of the salient tendencies of four major word classes, viz. Verbs, Type 1 and 2
adjectives and nouns. Type 1 adjectives are a subclass of intransitive verb.

Table 19 List of word classes
OPEN WORD CLASSES
CAN FUNCTION AS
PREDICATE HEAD?
Verbs YES
Nouns YES
Adverbs no
CLOSED WORD CLASSES
Type 1 adjectives YES
Type 2 adjectives YES
Time words no
Postpositions no
Demonstratives YES
Deictic-only demonstratives no
Interrogatives SOME
Indefinite proforms no
Discourse connectives no
Numerals no
Classifiers no
The additive conjunction no
Personal pronouns YES
The generic pronoun no
Proclauses no
Onomatopoeia no
Interjections no
The prohibitive word no
3 WORD CLASSES: AN OVERVIEW

84
Table 20 Some salient general tendencies of verbs, Type 1 and Type 2 adjectives
and nouns

ADJECTIVES
TYPE 2 NOUNS VERBS
ADJECTIVES
TYPE 1
MORPHOLOGICAL CRITERIA
can undergo valency changing derivations
cannot undergo valency
changing derivation
can take all event specifiers
difficult to use with
event specifiers when
head of a predicate
can take all predicate head suffixes
(see Table 63)
not attested with all
predicate head
suffixes
cannot take all
predicate head suffixes
can form compounds with nouns and verbs
only a few lexicalised
compounds with
nouns are attested
can form compounds
with nouns and verbs
CLAUSE LEVEL FUNCTIONS
can occur as predicates of all clause types
not attested as
predicate in all clause
types
can only occur as
predicate in one
independent/main
clause type:
identity/equation clause
and in some
subordinate clause
types
can take all argument
types
as head of a predicate can only take S argument
can function as argument in a clause only if
nominalised
can function as argument in a clause and be
case-marked
can participate in comparative and excessive constructions
PHRASE LEVEL FUNCTIONS
usually predicate head, but does not have to be
head in complex predicates (see 22.6.1)
can be predicate head usually a constituent or
head of an NP, can be
head of a predicate

can be modifier in an
NP

can be possessor and
possessed
can be morphologically
marked for plural
can be quantified
can be counted
SEMANTIC CRITERIA
The change of state has
change of state value
The change of state
has the effect of
reinforcement of the
quality denoted by
the Type 1 adjective
as well as indicating
the result of a change
of state.
The change of state indicates the result of a
change of state: has become X

85
Chapter 4 Verbs
_____________________________________________________________________

Verbs form an open class. The properties of verbs functioning as head of a predicate
are compared to those of adjectival and nominal predicate heads in Table 62 in
Chapter 22. Other properties of verbs compared to adjectives and nouns are summed
up in Table 20 above. An overview of the properties of verbs is given in sections 4.1
to 4.4. Subclasses of verbs are treated in section 4.5. The phenomenon of transitive
and intransitive verb pairs is discussed in section 4.6.
4.1 Clausal properties
Verbs can be the head of a predicate of all clause types. An important criterion of
verb-hood is the possibility to appear as the head of a predicate of an imperative
clause. However, stative verbs and most of the type 1 adjectives (a subclass of verbs
denoting qualities) seldom appear in imperative clauses. This presumably has to do
with the fact that imperative clauses are usually used to command others to do an
activity that they can control, and an interlocutor can usually not control the
occurrence of a quality. A counter-example is (27).

(27) tarakboto na'a!
{tarak} =bo =to [na'a]
fast =IMP =IMPEMPH 2s
Be fast, oh you!

This means that the imperative clause criterion is not a distinctive criterion by which
to define a verb for many members of that class. Interestingly stative verbs and type 1
adjectives can express imperious future (see 23.10.1). Verbs do not obligatorily have
to be a predicate in a clause, as a single or reduplicated bare verbal root can modify a
verbal predicate, e.g. (28) (see also Chapter 18). In the example below, the
reduplicated verbal root awan to forget functions as adverbial modifier to the
predicate wen-ok (wind-COS) wound. The adverbial phrase awan awan forgettingly
4 VERBS

86
is separated from the predicate it modifies by three NPs, viz. de'the he, i here and
dokom head.

(28) otokoymuna phaltha kangabaaw awan awan de'theba ici dokomci
wenoknoro.
otokoymuna [[phaltha] {kan} =gaba] =aw [awan awan]
CONJ self wear =ATTR =ACC forget RED
[de'theng] =ba [i] =ci [dokom] =ci {wen -ok} =no =ro
3S =EMPH PRX =LOC head =LOC wind -COS =QUOT =EMPH
So then, he forgetfully wound what he himself was wearing, here, around his
head, it is said.

Verbs can take arguments, which is one of the most salient differences to nouns,
which cannot take arguments when functioning as predicate head. However, no
argument in any clause has to be expressed when retrievable from the context, which
means that it is difficult to determine which types of arguments are obligatorily
conceptualised with certain verbs. Moreover, there are a few verbs denoting natural
phenomena which have a valency of zero.
4.2 Phrasal properties
A bare root or a reduplicated bare root can modify another verb which functions as
head of a predicate, e.g. (28), (see also Chapter 18). A non-finite predicate can modify
another verbal predicate head. A verbal predicate head can be modified by an adverb,
e.g. (265) in 6.6.
4.3 Morphological properties
Verbs can take all predicate head suffixes (see Table 63 in Chapter 22), provided the
construction is semantically felicitous. They can undergo valency changing derivation
by means of transitivisation.
4.4 Semantic properties
Verbs denote activities, processes, states and qualities. Verbs denoting qualities are
called Type 1 adjectives.
4 VERBS

87
4.5 Subclasses of verbs
The following subclasses of verbs have been identified depending on the semantics of
the verb and the types of argument that they can take. Primary-A verbs (see Dixon
2006 b) are verbs of which all arguments must be NPs or pronouns. There are very
few Primary-B verbs (see Dixon 2006 b), i.e. verbs which can take an NP or pronoun
or a complement clause as second argument apart from the subject (A). The list of
verbs that can only take complement clauses as second argument (apart from an A
argument), the so called Secondary verbs (see Dixon 2006 b), is very short. Seconday
verbs are verbs that cannot take an NP as their argument, but can only take a
complement clause as an argument. Speech act verbs are treated separately because
they can embed direct speech in a clause.

Table 21 Types of verbs

Primary-A verbs
Intransitive verbs
Type 1 adjectives
Other intransitive verbs
Verbs of emotion and interaction
Transitive verbs
Extended transitive verbs
Verbs that take arguments which are obligatory unmarked for case
The verb mo- to call someone/something a name
The interrogative verb atak to do what?
The demonstrative verb ytyk- to do like this/that.
Verbs denoting natural phenomena
Primary-B verbs
Secondary verbs
The secondary speech verb no to say
Phasal verbs

Atong has a number of intransitive-transitive lexical pairs which are discussed
separately in 4.6.
4 VERBS

88
4.5.1 Primary-A verbs
i Intransitive verbs
Intransitive verbs are verbs that cannot occur in transitive constructions and that
cannot take an O argument, unless they take transitivising morphology. Type 1
adjectives, a group of intransitive verbs denoting qualities, are treated in 5.1. The
following example illustrates the use of the intransitive verb jal- to run away. The
only NP expressed in the clause, i.e. matsa tiger, has to be interpreted as the S.

(29) matsae jalaokno.
[matsa]
S
=e {jal -a -ok} =no
tiger =FC run.away -AWAY -COS =QUOT
The tiger ran away, it is said.
ii Verbs of emotion and interaction
Verbs denoting emotions and interactions are kha'gal- to love, barat- to be
ashamed of, feel shy, canphe- to defend, pay'- to bear, sak- to bear etc. One
can argue that these verbs are intransitive because they cannot take an accusative-
marked O argument. When, besides the S argument, a second argument is expressed it
is usually a Target which is dative-marked. Examples (30) and (31) here below are
illustrative. As is mentioned in Chapter 21, no NP has to be expressed obligatorily in
any clause and it is impossible to test whether the dative-marked NPs of the verbs of
emotion and interaction are obligatorily conceptualised when they are not expressed.
Hence we cannot determine whether these NPs have core argument status or not.

(30) Boro na'na otokphin'ay kha'galano, otokciba na'na baratano.
[boro]
S
[na'] =na {otok -phin'} =ay {kha'gal -a} =no
Pname 2s =DAT do.like.this -FULLY =ADV love -CUST =QUOT
otokciba [na'} =na {barat -a} =no
but 2s =DAT be shy -CUST =QUOT
Boro loves you so much, [she] says, but she feels shy toward you, [she] says.

4 VERBS

89
(31) a randay rasanna pay'ca
[a randay]
S
[rasan] =na {pay' -ca}
1s body sun =DAT bear -NEG
My body does not support the sun.

Verbs denoting negative emotions and interactions are transitive since they can have
an accusative-marked O argument. Example (32) below is illustrative.

(32) a phulistaw ka'peta bajuau tokwana
[a] [phulis =taw] {ka'pet -a} |[baju] =aw {tok -wa}| =na
1s police =ACC be.angry -CUST friend =ACC beat -FACT =DAT
Im angry with the police because [they] beat [my] friend.
iii Verbs that take arguments which are obligatory unmarked for case
So far only two verbs have been recorded for which both core arguments are
obligatorily unmarked for case. These are the verbs plol'- to change into and the
identity/equation copula do'- ~ do- (IE.be), treated in the next section. The verb
phol'- to change into is intransitive and therefore the NP that is the undergoer of the
change is S, the second argument is a Result, which is a type of NP that is always
unmarked. Hence the clause headed by plol' to change into contains two obligatorily
unmarked arguments, e.g. (33).

(33) imi wa' juw' wak phol'wa.
[i =mi wa' jow']
S
[wak]
RESULT
{phol' -wa}
PRX =GEN father mother pig change.into -FACT
This ones father and mother have changed into pigs.
iv The copula and the locative/existential verbs
There is one copula in Atong, viz. the identity/equation copula do'- ~ do- (IE.be),
of which the allomorphs are in free variation. The copula functions as the head of the
predicate of copula clauses (see 26.5) and in support-verb constructions (see
22.7.2iii). The copula is not attested in non-finite clauses.
The identity equation copula do'- ~ do- (IE.be) supports two arguments, i.e. the
Copula Subject and the Copula Complement, e.g. (34). Both CS and CC are unmarked
for case.
4 VERBS

90
(34) ue hape cigacak te'ew kol india kolani hapan do'wacomno.
[ue hap] =e [cigacak]
CS
[te'ew] [kol india kolani
DST place =FC Pname now coal India colony
hap]
CC
-an {do' -wa} =com =no
place =FC/ID IE.be -FACT =IRR =QUOT
That place Chigachak is now supposedly the Coal India Colony place, it is
said.

The identity/equation copula do' ~ do- (IE.be) is homophonous with the intransitive
verb do' ~ do- be enough, be sufficient, which occurs in the next example.

(35) aya taka do'tawancakthay ado rocoygoksaan raariwa.
aya [taka] {do' -taw -an -ca -k} =thay} [a] =do
interj money enough -UPWARD -REF -NEG -COS =MIR 1s =TOP
[ro coygok] =sa =an {ra -ari -wa}
CLF:MONEY ten =DLIM =FC/ID bring -SIMP -FACT
Damn! To my surprise the money is not enough any more (does not reach up
to enough), I only brought ten rupees.

There is also a homophone do'- to arrive, also intransitive, shown in the next
example.

(36) otokoymo ray'akno ray'akno , nokthacina do'okno.
otokoymo {ray' -ak} =no {ray' -ak} =no
CONJ go -COS =QUOT go -COS =QUOT
[nok -tha] =ci =na {do' -ok} =no
house -OWN =LOC =ALL arrive -COS =QUOT
So then, he went and went, it is said, [and] arrived at his own house, it is
said.

4 VERBS

91
The identity/equation copula do'- ~ do- (IE.be) is used to tell the time (37).

(37) ato bajima. tin baji dook.
[ato baji] =ma [tin baji] {do -ok}
what hour =Q three hour IE.be -COS
Whats the time? Its past three.

The locative existential verb gana exist and the negative locative existential verb
ni'- not.exist are intransitive verbs. They support one core argument, i.e. S, as
illustrated here below in examples (38) (39) and (40). These verbs can express the
same grammatical categories as other verbs (see Table 63 in Chapter 22) as long as
the result is semantically possible in the context of the utterance. The
locative/existential verb gana exist cannot take the customary aspect <-a> (CUST)
except when it carries a derivative suffix (38). To express a customary aspectual
meaning the locative existential verb gana exist occurs without any suffix.
In the Badri dialect, the causative form of the negative existential and locational verb
ni'- not.exist occurs most frequently in the sense to switch off, e.g. layt ni'-et=bo
(light not.exist-CAUS=IMP) Switch of the light!.

(38) morot mo'sa ganano. uba jow'taraanokno. wa' ni'okno. ue gawicie sa'
mo'korok ganano aro dethe pipukci sa' mo'sa ganakhuano.
[morot mo' -sa]
S
{gana }=no [u]
S
=ba
person CLF:HUMAN one exist =QUOT DST =EMPH
{jow' -tara =an -ok} =no [wa']
S
{ni' -ok} =no
mother -EXCLUSIVELY =FC/ID -COS =QUOT father not.exist -COS =QUOT
[ue gawi] =ci =e [sa' mo' korok]
S
{gana} =no
DST woman =LOC =FC child CLF:HUMAN six exist =QUOT
aro [de'the pipuk] =ci [sa' mo' sa]
S
and 3d belly =LOC child CLF:HUMAN one
{gana -khu -a} =no
exist -INCOM -CUST =QUOT
There is one person, it is said. She has become only a mother [by herself], it is
said. There is no father any more, it is said. The woman has six children, it is
said and in her belly there is one more, it is said.

4 VERBS

92
(39) taka ni'ciba ganaciba a na'aw nemnuka.
[taka] {ni'} =ci =ba} {gana} =ci =ba}
money not.exist =LOC =INDEF exist. =LOC =INDEF
[a] [na'] =aw {nemnuk -a}
1s 2s =ACC like -CUST
Whether you have money or not, I like you.

(40) sandis ni'wa.
[sandis] {ni' -wa}
Name not.exist -FACT
Sandish is not [here].

Example (41) below illustrates that the negative locational/existential verbs can take
event specifier suffixes. The topic is a comparison between two Swiss army knives.
The speaker of the next example keeps pulling the different tools out of each knife
and notices that one knife has something the other does not have. So then he says:

(41) ici otokoy ni'siga
[i] =ci [otokoy] {ni' -siga}
PRX =LOC like.this not.exist -ALT
On this one [a tool] like this is not there in turn.

As an answer to all kinds of questions, the reified form ni'-wa (not.exist-FACT) can
be used with the meaning nothing with a variety of interpretations depending on the
context and the intonation of the speaker, as (42) (43) and (44) illustrate.

(42) aya ato korowa? nowacie. ay ni'wa na'a. a diphusa nowano.
aya [ato] {koro -wa} {no -wa} =ci =e}
interj what make.sound -FACT say -FACT =LOC =FC
ay {ni' -wa} [na'a] [a di'phu] =sa {no -wa} =no
interj not.exist -FACT you 1s fart =DLIM say -FACT =QUOT
When he said: Wow! what made that sound? Oh! nothing [dont worry],
oh you. Its only my fart. they/he said, it is said.

4 VERBS

93
(43) atakaro bayrik? ni'wa.
{atak -aro} [bayrik] {ni' -wa}
do.what -PROG Name not.exist -FACT
What are you doing Bairik? Nothing (This can be interpreted as really
nothing or as I dont want to tell you/its none of your business.)

(44) bisa re'ewa? ni'wa.
[bi] =sa {re'e -wa} {ni' -wa}
QF =MOB go.away -FACT not.exist -FACT
Where did you go? Nothing. (Interpretation: I dont want to tell
you/its none of your business.)

Example (45) illustrates the locative/existential verbs as head of a predicate of an
attributive clause, and (46) shows the negative locative existential verb as predicate
head of a clause nominalised by the genitive/nominaliser clausal enclitic <=mi>
(GEN/NR).

(45) garuaw ue di' ganagabaaw susetca, dowetoknoro.
[garu] =aw |[ue di' {gana} =gaba] =aw {suset -ca}
mustard =ACC DST shit exist =ATTR =ACC wash -NEG
{dow -et -ok} =no =ro
add -CAUS -COS =QUOT =EMPH
[She] did not wash the vegetables with that shit on them [but/and] added
them, it is said.

(46) momawa ni'wamian man'ay sa'cak khagal do'ok.
|[moma -wa] {ni' -wa}| =mi =an {man'} =ay {sa' -ca -k}
elephant -tooth not.exist -FACT =NR =FC/ID obtain =ADV eat -NEG -COS
[khagal] {do' -ok.}
poor.person IE.be -COS
[Because of] the non-existence of the elephant tusks, [they] were no longer
rich, [they] had become poor.
v Transitive verbs
Transitive verbs are those verbs that can occur in transitive constructions and
maximally take two core arguments, A and O (47). A is always unmarked for case and
O can either be unmarked or, only when referential and definite, accusative-marked,
4 VERBS

94
but this accusative marking is optional (see Chapter 20). The following example
illustrates the use of a transitive verb, i.e. dopolo- to flatten.

(47) gari bebolokaw depoleok
[gari]
A
[bebolok]
O
=aw {dopole -ok}
vehicle toad =ACC flatten -COS
The car has flattened the toad.
vi Extended transitive verbs
So far, only two extended transitive verb have been discovered in the language. These
are the verbs mo- to call something a name and no- to call something a name. The
obligatory extended argument, semantically the Name of a named entity, is always
unmarked for case. The named entity is the O argument and receives accusative
marking (48).

(48) ue ha'boriawe se'sotay matsa caw'koy mosigaariok.
[ue ha'bori]
O
=aw =e {se'sot} =ay [matsa caw'koy]
E/NAME
DST hill =ACC =FC abbreviate =ADV Pname
{mo -siga -ari -ok}
call.a.name -ALT -SIMP -COS
That hill is also just called Matsa Cawkyi for short.
vii The interrogative verb atak
Atong has an interrogative verb, viz. atak to do what. I will demonstrate its different
usages with the following examples. In (49) we see how the verb is used to inform
about the actions of a specific person.
4 VERBS

95
(49) ayaw! ado cokaydoa. atakwa? Te'ewmamasa toyruwa na'a.
[ayaw] [a] =do {cok -aydoa}
interj 1s =TOP cold -PROG
{atak -wa}
do.what -FACT
[te'ew -mama] =sa {toyru -wa} [na'a]
now ONLY =DLIM bathe -FACT 2s
Ow! Im cold. What have you done? I took a bath just now, oh you!

Sometimes the verb is used as a devise in discourse to create an expectation about the
following stretch of discourse. In these cases the predicate containing the interrogative
verb atak- to do what? can be translated as what happened?, as we see in (50).

(50) konsado atakoknowa? jamjolay gopca amakawe.
[onsa]=do {tak -ok} =nowa
later =TOP do.what -COS =QUOT
{jamjol} =ay {gop -ca} [amak] =aw =e
complete =ADV burry -NEG monkey =ACC =FC
What did [they] do later, it is said? alternatively What happened later, it is
said? They didnt burry [him
i
] at all, the monkey
i
.

The interrogative verb can be used to ask for a reason. In this case the clause in which
the verb appears is dative-marked (see Chapter 27) and can be translated with why?
in English, as we can see in (51). By simplification of the cross-morphemic cluster
/kn/ to /n/, the dative-marked form gave rise to a real, opaque interrogative morpheme
atana why, as described in 9.5.

(51) na?a atakna jumuaydoa ie ha'thaphoraawe? nowano
[na'a] {atak} =na {jumu -aydoa} [ie ha'thaphora] =aw =e
2s do.what =DAT collect -PROG PROX ashes =ACC =FC
{o -wa} =no
say -FACT QUOT
Whay are you collecting those aches? [they] said, it is said.
Literally: You are collecting those ashes to do what?

The interrogative verb can occur as the predicate of an attributive clause, modifying a
noun which is the head of an arch NP (see Chapter 29), as we can see in example (52).
4 VERBS

96
In that example the noun raja king is modified by the attributive clause atak=gaba
(do.what=ATTR). In these cases, the arch NP can best be translated into English as
what kind of X?, since the question is more about what kind of X it is than about
what activities the X actually does

(52) atakgaba raja na'a ana gore lapcagabaaw watetwa nookno.
----------arch NP--------
-AC-
[ |atak| =gaba raja] [na'a] [a] =na [gore lap -ca =gaba] =aw
do.what =ATTR king 2s 1s =DAT horse profit -NEG =ATTR =ACC
{watet -wa} {no -ok} =no
send -FACT say -COS =QUOT
What kind of king are you [that] you send me a good-for-nothing horse?,
[he] said, it is said. Literally: You are a king who does what?
viii The demonstrative verb tk-
There is one demonstrative verb in Atong: otok- to do/be like this/that and itek- to
do like this. By far most occurrences of this verb in the corpus collected during my
fieldwork are anaphoric, i.e. coreferential to whatever action or state a verb earlier in
the in the text referred to; only one instance of purely deictic use of this verb is
attested, which is illustrated in example (30), here repeated as (53). In this example,
the extent of Boros love has to be inferred from the context in which this sentence
was uttered, not from any preceding discourse.

(53) Boro na'na otokphin'ay kha'galano, otokciba na'na baratano.
[boro]
S
[na'] =na {otok -phin'} =ay {kha'gal -a} =no
Pname 2s =DAT do.like.this -FULLY =ADV love -CUST =QUOT
otokciba [na'} =na {barat -a} =no
but 2s =DAT be shy -CUST =QUOT
Boro loves you so much, [she] says, but she feels shy toward you, [she] says.

Example (54), taken from TEXT 1 (lines 24-28), demonstrtates the anaphoric use of
the demonstrative verb. In the last line of the example, the demonstrative verb is
coreferential with the verb thik-an-ca (be.good-REF-NEG) are not good in the
4 VERBS

97
second line, i.e. both verbs refer to the state of not being good. Other example of the
anaphoric use of this verb are (273), (587), (693) and line 40 of TEXT 2.

(54) Speaker 1: atakna watna na'a, ma?
Speaker 2: gawian thikanca.
Speaker 1: o thikanca?
Speaker 2: ho'o.
Speaker 1: otokaria, te'ewrawrawmi gawido.

atakna wat -na ma
why send.away -DESI Q
Why do you want to send her away [i.e. dump her]?
gawi =an thik -an -ca
girl =FC/ID right -REF -NEG
The girl is not decent.
o thik -an -ca
interj right -REF -NEG
Oh, not decent.
ho?o
yes
otok -ari -a te'ew -rawraw -mi gawi =do
do.like.that -SIMP -CUST now -CONTINUOUSLY -GEN girl =TOP
They do just like that, the girls of nowadays.
ix Verbs denoting natural phenomena
Verbs describing natural phenomena are a small closed class of verbs. The five
members are listed in Table 22. The verb balwa- to blow (of the wind) can only take
the cognate S balwa wind as its argument. The verbs wal- to (be) night, manap- to
(be) morning and gasam- to (be) evening have a valency of zero, i.e. they cannot
take any arguments. The only S argument said to be possible with the verb wal- to
(be) night is san day, but I have never heard it in spontaneous speech. The verb wa-
to rain is intransitive and can only have the prototypically associated noun ra rain
as its S argument. Verbs denoting natural phenomena can express the same
grammatical categories as other verbs except imperative.

4 VERBS

98
Table 22 Verbs denoting natural phenomena and their corresponding nouns
VERB VALENCY EXAMPLE NOUN
balwa- to blow (as wind) cognate S only (55) balwa wind
wal- to (be) night prototypically
associated S
only, i.e. san
day
(56) wal night
manap- to (be) morning nil - manap morning
gasam- to (be) evening nil - gasam evening
wa- to rain prototypically
associated S
only, i.e. ra
rain
(58) --

(55) tay'ni balwa tha'rakay balwaaok.
[tay'ni] [balwa] {tha'rak} =ay {balwa -a -ok}
today wind strong =ADV wind -AWAY -COS
The wind blew strong today.

(56) te'ewdo walnaka.
[te'ew] =do {wal -naka}
now =TOP night -IFT
It will soon become night now.

(57) konsa golphook golphook golphook. golpho kha'wacie walaaydok.
[kosa] {golpho -ok} {golpho -ok} {golpho -ok}
later.on talk.extensively -COS talk.extensively -COS talk.extensively -COS
[golpho] {kha' -wa} =ci =e {wal -a -aydok}
story do -FACT =LOC =FC night -AWAY -PROG
Later on they talked and talked and talked extensively. When they talk/talked,
it is/was becoming night. Literally: When they did story it is/was nighting
away.

In the next example the prototypically associated noun ra rain might be
incorporated into the predicate because the noun is unmarked, non-referential and
together with the verb it can denote one recognisable unitary concept (see Mithun
1984). It is, however, possible to separate the noun ra rain and the verb wa- to
rain as (59) illustrates. Moreover the verb can be used without the noun as in (60).
4 VERBS

99
(58) asikatimo somayci nemen ra waakno.
[asikati =mo somay] =ci [nemen] {ra wa -ak} =no
August.September =GEN time =LOC very rain rain -COS =QUOT
In August and September it rained very hard, it is said.

(59) rang pa'ay waaydok
[ra] [pa'} =ay {wa -aydok}
rain much =ADV rain -PROG
Rain is falling heavily.

(60) rasan raboromaydoa, waynikhon
[rasan] {raborom -aydoa} {wa -ay -ni} =khon
sun be.shrouded.in.clouds -PROG rain -TOWARDS -FUT =SPEC
The sun is blocked by the clouds, it might rain.

The only attested verb denoting a natural phenomenon functioning as the predicate of
an attributive clause is given here below.

(61) a ra wagaaw nemnuka
[a] |[ra] {wa} =ga] =aw {nemnuk -a}
1s rain rain =ATTR =ACC like -CUST
I like the rain that falls.
4.5.2 Primary-B and Secondary verbs
All arguments of Primary-B verbs can be NPs or pronouns but one argument can
alternatively be a clause. The complement clause fulfils the function of O argument in
the matrix the clause. Most Primary-B verbs take dative clauses as complements; only
one verb has been discovered that can take either a dative- or a factitive-marked
clause as complement, viz. the verb ga'ak- to be compelled to. Secondary verbs
cannot take an O argument but have to take a dative-marked clause as a complement.
The Primary-A and Secondary verbs are listed in Table 23 below. Factitive clauses
are treated in Chapter 1 and dative-marked clauses are treated in Chapter 27.

4 VERBS

100
Table 23 List of Primary-B and Secondary verbs (not exhaustive)
VERB MEANING
Can take NP
as O argument
Type of clausal
complement
Primary B verbs
na- to need, have to yes
dative clause
man'- to be able; to obtain yes
tak- to do yes
canci- to think about yes
ha'bace- to begin yes
Secondary verbs
pa'- to dare no
dative clause
sok- to want no
sap- to know a skill no
neng'- to lack, to fail to no
mykso- to plan, to intend no
ga'ak- to be compelled to no dative or
factitive clause
4.5.3 The Secondary speech-verb
The verb no- say can participate in only two constructions. First, it can take a quote,
i.e. a stretch of direct speech as its complement, in order to form a grammatical
clause. Thus it can embed a direct speech report into a higher clause. This is the most
common usage of the verb, e.g. (62). In this example the direct speech report is
underlined and is the complement of the verb no- to say, which is subordinate to the
main verb bal- to speak, say, tell. The topic marker <=do> (TOP) does not mark the
reported speech, but only the phrase tay'ni today.

(62) otokciba pherue hm'm, kakay sa'arini na'aw tay'nido noaymo
balariano.
otokciba [pheru]
A
=e [hm'm] {kak} =ay {-sa' -ari -ni} [na' =aw
but fox =FC no bite =ADV eat -SIMP -FUT 2s =ACC
[tay'ni] =do {no} =ay =mo {bal -ari -ok} =no
today =TOP say =ADV =SEQ speak -SIMP -COS =QUOT
But the fox having said: No, I will just bite and eat you today just spoke, it
is said.

Second, it can take a dative-marked complement clause, of which there is only one
recorded example (755), which is presented in 27.2.1. The use of this verb is
compulsory in clauses in which direct speech is embedded, to signal the direct speech.
4 VERBS

101
4.5.4 Phasal verbs
Phasal verbs indicate beginning or completion. So far only three phasal verbs have
been discovered in Atong. They cannot all take NPs as O argument but they can all
take complement clauses of which the predicate is factitive-marked. These
complement clauses never occur accusative-marked. Table 24 below lists the phasal
verbs in Atong.

Table 24 Phasal verbs
VERB MEANING Can take noun
as O argument
Type of clausal
complement
jam- to finish, complete yes
factitive clause macot- to finish yes
da'- to enter a state of no
4.6 Intransitive-transitive lexical pairs
Atong displays a few vestiges of an old, now unproductive, system of verbal
transitivisation. These remnants have resulted in lexical pairs of intransitive and
transitive verbs. The transitive verbs show reflexes of what is presumably the proto
Tibeto-Burman causative prefix *s-. There are two systems.
In one system, to which most pairs adhere, the transitive verbs are derived from
the intransitive ones with an unproductive prefix <tho- ~ do-> (TRANSITIVE) The
schwa of the unproductive causative prefix assimilates completely to the vowel of the
root. The distribution of the allomorphs of the unproductive Atong transitive prefix is
as follows. The allomorph <-do> occurs before plain voiceless and aspirated initial
stops while the allomorph <-tho> occurs before voiced initials and /s/. Transitivised
verbs with initial affricate /c/ or /j/ have not been recorded. Surprisingly similar to the
fossilised Atong prefix in phonetic make up and distribution are the palatalised
allomorphs <o- ~ do-> of the causative prefix <s- ~ o- ~ do-> in Jingpho (see
Matisoff 2003: 101) or Kachin, as the language is called by Benedict (1972: 105).
4 VERBS

102
The proto Tibeto-Burman causative prefix *s- is reflected as devoicing and
aspiration of initial stop on one pair of verbs, viz. bay to break (intransitive) and
phay' to break (transitive), to translate
21
, in which the consonant initial varies
between a voiced stop for the intransitive verbs and a voiceless aspirated stop for the
transitive ones. Limbu, a Kiranti language of Nepal (van Driem 1987: 245 sq.) and
Burmese (Okell 1969: 42), have many similar verb pairs that reflect Proto-Tibeto-
Burman *s-. The intransitive and transitive verbal pairs are listed here below in Table
25.

Table 25 Transitive and intransitive verb pairs
intransitive transitive
System 1
barat- to be ashamed thabarata to make ashamed
bejaw- to tickle thebajaw- to experience the sensation of
being tickled
gal'- to fall thagal'- to loose
ku- to be dammed up a circle of
stones (of water in the river
as a technique to catch fish)
duku- to dam up water by making a
circle of stones in the river(as a
technique to catch fish)
mot- to extinguish thomot- to put out (of fire)
mimi- to smile thimimi- to make someone smile
myn- to be ripe thymyn- to ripen (after being picked)
nuk- to see thunuk to show
sa- to wake up thasa- to wake somebody up
khop-
~khup-
to put clothes on, to cover, dokhop- ~
dukhup-
to dress someone
koro- to make noise dokoro- to make noise on purpose
koroy- to fear dokoroy- to threaten
thoy- to die dothoy- to kill (for ceremonial
purpose)
kirin torn (of clothes, paper etc.) dikirin- to tear (clothes, paper etc)
phi- to be full diphi- to fill
-not recoded- dopolo- to flatten
System 2
bay- to break phay'- to break, to translate

21
Interestingly, whereas in Atong only the transitive verb phay' to break is glottalised, in Burmese
both the transitive and the intransitive member of the pair have a final glottal stop, viz. hpya
'
cut,
break and pya
'
be cut, broken; snap (Okell 1969:42).

103

104
Chapter 5 Adjectives
_____________________________________________________________________

There are two small closed classes of adjectives in Atong which I will call Type 1 and
Type 2. The criteria for distinguishing these two types can be found summarised in
Table 20. Type 1 adjectives form a subclass of intransitive verbs, they denote a
quality and are used primarily predicatively. For type 1 adjectives to be used
attributively to an NP, an attributive form needs to be derived with the attributive
clausal enclitic <=gaba ~ =ga> (ATTR), the function of which is extensively treated
in Chapter 29. Type 2 adjectives can be used either predicatively or attributively
without any derivational process.
Approximately forty five members of the Type 1 adjective class have been
recorded indicating the semantic types (see Dixon, 2004 a) of dimension, value,
colour, physical property, position, quantification. Approximately fifty Type 2
adjectives have been recorded. Type 2 adjectives denote age, colour, physical
property, position, speed, quantification, possessiveness and similarity. There are
some adjectives that have a semantic extension, viz. Type 1 rak- means hard as
physical property and difficult and the Type 1 adjective ga'- good as a value can
be extended to human propensity meaning good hearted. It is noteworthy that there
are no other adjectives signifying human propensity in Atong. Table 26 below lists the
adjectives of both classes recorded to date.
It has to be noted that the property of being able to participate in comparative,
superlative and excessive constructions is not a property that is characteristic only of
adjectives but also of verbs (see Table 20 above). Both word classes can express the
comparative (morethan)/superlative <-khal> (CP/SUP) and the excessive <-duga>
(XS) categories. This is illustrated in the next examples with adjectives of Type 1
(63), (65) and verbs (64), (66).

(63) cak raw'khalay suletkhubo.
[cak] {raw' -khal} =ay {sulet -khu} =bo
arm/hand long -CP =ADV stretch -INCOM =IMP
Stretch your arm longer.
5 ADJECTIVES

105
(64) ge'the ana dayay sa'khala.
[ge'the] [a] =na {day} =ay {sa' -khal -a}
3s 1s =DAT be.bigger=ADV eat -CP -CUST
He eats more than me.

(65) tibimi koronggaba rakdugabutungci cangba nietok
[tibi =mi koro =gaba] {rak -duga -butung} =ci
television =GEN sound =ATTR strong -XS -WHILE =LOC
[ca] =ba {ni -et -ok}
who =INDEF not.exist -CAUS -COS
When the sound of the TV was too loud, someone turned it off.

(66) ge'thengdo may sa'dugaak.
[ge'theng] =do [may] {sa' -duga -ak}
3s =TOP rice eat -XS -COS
He has eaten too much rice.

An alternative interpretation, with may rice incorporated into the predicate
(see Chapter 22):

[ge'theng] =do {may sa' -duga -ak}
3s =TOP rice eat -XS -COS
He has eaten too much. Literally: he has rice-eaten too much.
5.1 Type 1 adjectives
Type 1 adjectives are a subclass of intransitive verbs, i.e. they are stative verbs
denoting a quality. They are distinguished from other intransitive verbs only on the
basis of their semantics and the semantic effect that the change of state suffix <-ok ~ -
ak> (COS) has on these adjectives. The change of state marker can produce an
intensifying effect on Type 1 adjectives, e.g. sol- beautiful sol-ok (beautiful-COS)
very beautiful, thaw- tasty thaw-ok (tasty-COS) very tasty, ga'su- splendid,
great, cool ga'su-ok (cool-COS) totally cool, very cool. However, in the right
context a change of state morpheme on a Type 1 adjective can also be interpreted as
asserting a change of state (67).

5 ADJECTIVES

106
(67) golpho lekha catok.
[golpho lekha] {cat -ok}
story book thick -COS
The story book has become thick. or The story book is very thick.

The different properties of Type 1 adjectives as predicate head are compared with
those of other verbs, Type 2 adjectives and nouns in Table 62 in Chapter 22.
Type 1 adjectives are intrinsically predicating (68), but as head of a predicate of
an attributive clause (see also Chapter 29) they can be used as modifiers of nouns
(69).

(68) ue gawi sola.
[ue gawi] [sol -a]
DST girl beautiful -DCL
That girl is beautiful

(69) toi ga'gaba wari thow'gaba toysamci hap solgabaci mo'ni
bay'sigathamaran ca torowceonaka noaydoano.
[toy {ga'} =gaba] [wari {tow'} =gaba] [toysam] =ci
water good =ATTR deep.place.in.river deep =ATTR water.edge =LOC
[hap {sol} =gaba] =ci [mo' ni bay'siga -tha -maran]
place beautiful =ATTR LOC CLF:HUMAN two friend -OWN -RC
|{ca torow -ce -naka}| {no -aydoa} =no
who take.a.bath -FIRST -IFT say -PROG =QUOT
At the waterside of a place in the river where there was nice and deep water,
in a beautiful place, the two friends are arguing who will take a bath first, it is
said.

In the next example we see how a Type 2 adjective, in this case abun other,
different, does not need to be attributivised before it can function attributively to a
noun, in this case the noun khu'cuk language.

5 ADJECTIVES

107
(70) khu'cuk abunsa balcido blepayri noay balna man'nicom.
[khu'cuk abun] =sa {bal} =ci =do
language other =INSTR speak =LOC =TOP
[blepayri] {no} =ay {bal} =na {man' -ni} =com
blank.firing say =ADV say =DAT be.able -FUT =IRR
If [you] say it in another language, [you] could say blank firing. Literally:
you could sayingly say blepayri.

Table 26 List of adjectives sorted by semantic category and class
DIMENSION Type 1 VALUE Type 1
cu- big nem- good
mol- small ga'- good
thow'- deep ga'su- sensational, cool
cow'- high hanse- beautiful
raw'- long (time, body, thing) thaw- tasty
su- short (time, body, thing) rak- hard, strong, difficult
damrak- expensive, costly
AGE Type 2 sol- pretty, beautiful
pidan new (thing) som- sweet
picam old (thing) botom- nice smelling
bodoy old (person) manam- foul smelling
bylak- strong (of persons)
COLOUR Type 1 mothel- thankful
bok-a white, unripe
nak-a black PHYSICAL PROPERTY Type 1
sak-a red ka'- bitter
koy- sour
COLOUR Type 2 rak- hard, strong, difficult
pi'bok white, unripe ga- erect
pinak black toysi- wet
pi'sak red, blond cek- cold
kecek green, blue bolak- strong
romot yellow, orange tu- warm
bolu blue kanjot- skinny
khengsorok dark green manak- dark
mel'- fat
nom'- soft, tender
ne'- tired
corom- heavy
demdo- weak, soft
su'ut- damp
ce- light, not heavy
co'- bright

5 ADJECTIVES

108
Table 26 continued
PHYSICAL PROPERTY Type 2 POSITION Type 2
gotho ~ sul next, neighbouring
githing ~ cu'ret stuck
githyng unripe, uncooked, raw cega upright
doda alone bathan lying on its/his back
bokphol inside out borop lying on its/his belly
phangphyl upside down goci nearby, slant, aslant
domdam naked abun following, next,
raw're long and slender neighbouring, other,
baba empty someone else,
karam poisonous different
khuru wanting to lay an egg
kirin torn (paper, clothes) SPEED Type 2
nagok deaf khasin slow
toykaran thirsty
bukala to have a hole in it SIMILARITY Type 2
(clothes) hapsan the same
thombolo to have holes in it, gapsan the same
damaged (road, bridge) dotha different
porow to have a hole in it (walls) alaga other.
mo main baybay the same
bromborom multicoloured
thokborom multicoloured QUANTIFICATION Type 1
(toy)karan thirsty pa'- many
ja'jot narrow in the middle
ompong lopsided QUANTIFICATION Type 2
thowkho bulging bay'dam some
khingceng aslant, slant mama only, exclusively
cing'pheng aslant, slant
pylang flat TEMPORAL Type 1
ja'raw- long
VALUE Type 2 khengkhang eternal
ramram ordinary, normal
POSSESSIVE Type 2
POSITION Type 1 goga having, with
jan'- far
neka- near

5.2 Type 2 adjectives
An overview of properties is listed here below followed by examples. Type 2
adjectives share some nominal and some verbal properties. The properties of Type 2
adjectives functioning as predicate head are compared to verbs and nouns in Table 63
5 ADJECTIVES

109
in Chapter 22. In Table 20 in this chapter other properties of Type 2 adjectives are
compared with those of verbs and nouns.
5.2.1 Clausal properties
Type 2 adjectives can function as head of a predicate of an identity/equation clause,
just like Type 1 adjectives.
5.2.2 Phrasal properties
Type 2 adjectives can modify a noun within an NP in post or pre-head position
without any difference in meaning; a Type 1 one adjective needs to be attributivised
before it can do this.
Type 2 adjectives can modify a verb as head of a predicate of an adverbial clause
without being adverbialised with the adverbial clausal enclitic <=ay> (ADV), but also
occur with this adverbial enclitic (more fieldwork is needed to determine the factors
that determine this choice); Type 1 adjectives need to be adverbialised.
5.2.3 Morphological properties
Type 2 adjectives can be reduplicated or partially reduplicated to express greater
intensity, which Type 1 adjectives cannot.
5.2.4 Semantic properties
Type 2 adjectives express age, colour, physical property, position, speed,
quantification, value, possessiveness and similarity. Some of these categories overlap
with those of Type 1 adjectives (see above).
Type 2 adjectives can modify nouns in unmarked form and usually in post head
position, e.g. (70) above, (71) and (75) below. A Type 2 adjective can take case
marking and other nominal morphology, e.g. (71), even when it appears without its
head noun, e.g. (72), (73) and (74).

(71) ge'theo nok picamthaaw pay'gorumok.
[ge'the] =do [nok picam] =tha =aw {pay' -gorum -ok}
3s TOP house old =OWN =ACC break -collapse -COS
He has destroyed his own old house.
5 ADJECTIVES

110
(72) thomboloaw ray'cawa.
[thombolo] =aw {ray' -ca -wa}
with.holes.in.it =ACC go -NEG -FACT
I will not take the one with holes in it. (Speaking about a road.)

To use a Type 1 adjective in the same construction as (72), we would have to use an
attributive clause in a headless arch NP (see Chapter 29), viz. [{nem}=gaba]=aw
ray'-naka (good=ATTR=ACC go-IFT) [I] will take the good [one (i.e. road)].

(73) bay'dam Rodo toykhal ha'waycina jalaok.
[bay'dam] [Rodo toykhal ha'way] =ci =na {jal -a -ok}
some Pname river plain =LOC =ALL run -AWAY -COS
Some [people] run away to the Rongdyng river plain.

(74) dakami picammi kamdoarado
[daka] =mi [picam] =mi [kam] =dora =do
in.the.past =GEN old =GEN activities =p =TOP
As for how things went in the old days

Type 2 Adjectives can function as head of a predicate of identity/equation clauses,
where it shares properties with both nominal and verbal predicate heads (75). The
properties of Type 2 adjectival predicate heads are discussed in 22.4.

(75) nok romot picamanca.
[nok romot] {picam -an -ca}
house yellow old -REF -NEG
The yellow house is not old.

Type 2 adjectives can take the adverbial clausal enclitic <=ay> (ADV) to modify a
following predicate (76), (77) or be reduplicated (78), (79).

(76) ge'the pinakay sola.
[ge'the] {pinak} =ay {sol -a}
3s black =ADV beautiful -CUST
She is black and beautiful. (lit. blackly beautiful)
5 ADJECTIVES

111
(77) boropay jowbo.
{borop} =ay} {jow} =bo
lying.on.belly =ADV lie.down =IMP
Lie down on your belly.

(78) khasin khasin re'enbo.
[khasin khasin] re'e =bo
slow RED go.away =IMP
Walk slowly!

(79) na'na doda doda hapsan golgolni.
[na'na] [doda -doda] [hapsan] {golgolni}
we
i
alone -RED together roam
We
i
will both roam alone.
5.3 Remarks on certain adjectives
The word bodoy old (persons), Type 2, also occurs as a noun meaning old man.
The adjective bodoy also occurs as a bound morpheme suffixed to the word jow?
mother, where it yields the meaning jow'bodoy old/married woman. The
compounding of adjectives with nouns is not a productive process.
The word alaga other, Type 2, also occurs as a noun meaning someone else.
Compare the following examples where it is used as adjective in (80) and as noun in
(81). The adjective alaga other cannot modify verbs.

(80) nok alaga
nok alaga
house other
another house

(81) alagami nok
alaga =mi nok
someone.else =GEN house
someone elses house

The prefix <pi-> ? on the Type 2 colour and age adjectives is unproductive. The
adjectives pidan new and picam old also occur as bound morphemes without the
5 ADJECTIVES

112
unproductive prefix pi- ? in the words maydan new rice, socam former village
and wa'cam post-harvest rice stalk. As was said above, suffixing adjectives to nouns
is an unproductive process.
The Type 2 adjective khasin slow is the only adjective of its type that has been
recorded as a command: Khasin! slowly!, and with the imperative clausal enclitic
<=bo> (IMP) khasin=bo! (slow=IMP) Slowly!.
The Type 2 adjective hapsan ~ gapsan same, together (allomorphs in free
variation) can be used as pre-head or post head modifier. As post head modifier this
adjective can function as head of a predicate, e.g. (83), whereas in pre-head position it
cannot, e.g. (82). The adjective hapsan ~ gapsan same can modify other adjectives
of both types as in the equative construction in (84).

(82) ie hapsan nok
[ie] {hapsan nok}
PRX same house
This is the same house

(83) ie nok hapsan
[ie] hok {hapsan}
PRX house same
This house is the same.

(84) a na'mi hapsan cua.
[a] [na'] -mi [hapsan] {cu -a}
1s 2s =GEN same big -DCL
I am as big as you.

In (84) above we see an example of the equative construction in Atong. The structure
of the equative construction is Comparee Standard hapsan Parameter. The
standard takes the genitive case and hence functions as second object of the
predicative adjective. Hapsan functions as a modifier of the predicate just like in (79)
and (85) here below. It is a property of Type 2 adjectives that they can modify verbs.

(85) na'na hapsan sa'nine.
[na'na] [hapsan] {sa' -ni} =ne
1pi together eat -FUT =TAG
We will eat together, OK?

113

114
Chapter 6 Nouns
_____________________________________________________________________

Nouns are an open class. We will start with an overview of the properties of this word
class in sections 6.1 to 6.4. Nominal subclasses are treated in section 6.5. In section
6.6 we will examine the different interpretations of juxtaposed nouns.
Nouns can function as core or peripheral argument (adjunct/oblique argument). Nouns
can also function as predicate head. Nominal predicates are treated in 22.5.
The noun is the head of the NP. The NP can take NP enclitics such as case-marking
(see Chapter 20) and other phrasal enclitics (see Chapter 19). Demonstratives always
occur in the left-most position in the NP, before the head. Other modifiers within the
NP can precede or follow the head without any apparent difference in meaning.
whether the variation of the modifiers within the NP results in subtle pragmatic
differences will have to be determined by more fieldwork.
Other phrasal properties of nouns are that they
can be incorporated into the predicate (see 22.7),
can modify another noun in juxtaposition (see 6.6 below),
can be modified by Type 2 adjectives (see 5.2) and attributive clauses (see
Chapter 29),
can be possessed
6.3 Morphological properties
Nouns can be pluralised and quantified (see Chapter 1 and Chapter 12). Mass nouns
can only be quantified with measure nouns (see 12.4). The noun determines the
choice of classifier and reduces the scope of reference of classifiers (see Chapter 12).
6 NOUNS

115
6.4 Semantic properties
Nouns denote activities, places, abstract notions, animals, artefacts, body parts, food
items, geographic, geological or natural phenomena, kinship terms, locations, measure
terms, names of persons and locations, plants, qualities, quantities, shapes, substances
and time.

Reduplication of a noun has an adverbialising function (see 18.8). The properties of
nouns functioning as head of a predicate compared to verbs and adjectives are given
in Table 62 in Chapter 22.
6.5 Subclasses of nouns
On the basis of their syntactic, phrasal and morphological properties, Atong presents
the following subclasses of nouns:

1. Common nouns
2. Nouns denoting proper names and persons
3. Kinship terms (treated in Chapter 1)
4. Inherently locational nouns
5. Mass nouns
6. Gender sensitive nouns
7. Auto-classifiers (treated in 12.3)
8. Measure nouns (treated in 12.4)

The first four subclasses will be treated below in this chapter, the three remaining
subclasses are treated in other chapters, as indicated above.
6.5.1 Common nouns
The class of common nouns consists of all nouns that do not belong to any of the
other subgroups. Common nouns denote concepts in the following semantic domains:
abstract nouns, human activities, results of or circumstances related to human
activities, animals, artefacts, body parts of humans and animals, diseases, food items,
6 NOUNS

116
ingredients used for food, geographical, geological or natural phenomena, plants,
parts of plants, qualities, quantities, and shapes.
6.5.2 Nouns denoting persons and proper names
Proper names (86) and nouns denoting persons (87), (88) are the only ones that can
take the associative plural <=para> (&co). This agrees with the generalisation pointed
out in Moravcsik (2003: 472) that the focal referent of a group referred to by the
associative plural must be a definite human individual. Nouns denoting animals are
also eligible to take the associative plural, but only in stories, when they talk and act
like humans (89). Kinship terms also denote persons and are treated separately in
Chapter 1. Examples (87) and (88) contain a kinship term marked by the associative
plural.

Proper name:
(86) aa letitparamo nokci sa'ni.
[aa] [letit =para =mo nok] =ci {sa' -ni}
1s Pname =&co =GEN house =LOC eat -FUT
I will eat at the house of Latith and company.

Noun denoting close human relationship (subset of kinship terms):
(87) na' bajupara ca'masa na' punna re'ewa.
[na' baju] =para [ca'ma] =sa {na' pun =na} {re'e -wa}
2s friend =&co down.side =MOB fish catch.fish =PUR go.away -FACT
Your friend and his company went to the river side to catch fish.

Kinship term:
(88) ah baba, na' dadaparado usa sikal ramna re'eok.
ah baba [na' dada] =para =do [u] =sa
interj grandson 2s older.brother =&co =TOP DST=MOB
[sikal] {ram =na} {ra'e -ok}
hunting search =DAT go.away -COS
Oh grandson, the group of your elder brothers went that way to try to hunt.

6 NOUNS

117
Animal in story:
(89) amakparae bodoy baday re'eaydoknowa.
[amak] =para =e [bodoy baday] {re'e -ok} =no -wa
monkey =&co =FC old couple go.away -COS =QUOT -FACT
The monkeys in each others company, the old couple, went away, it is said.

Of all the subclasses of nouns, the personal names subclass is the most open of all
because new lexical items are being added continuously. In the culture of the Atong,
just as in that of the Garos (Burling, 2004: 228-9), parents go to considerable lengths
to find unique names for their children. Hence it is usually only some weeks after its
birth that a child is named. Names usually vary from two to five syllables in length
and can consist of entirely invented sequences of sounds as long as they produce a
pleasant sound. Alternatively they can be taken from other languages (outside the
tribal communities of Meghalaya). Some names consist of a mix of invented syllables
as well as existing lexemes and grammatical morphemes to give the name a symbolic
value. Every name conforms to the patterns of Atong phonology in speech. The way
in which names are written may be quite different from the way they are pronounced,
especially when the name is or is supposed to resemble a Western name, e.g.
Holybirth [olibot], Fernanda [pananda]. There are certainly whims of fashion in
syllables used for naming children. There are a considerable number young males
now having a name ending in -bat (< English: birth, written as birth but pronounced
[bot]) and a considerable number of slightly older males having names ending in -se.
Proper names behave like prototypical nouns: they can take the full set of case
markers and other phrasal enclitics, can function as head of a predicate and can be
modified, can be used as argument of a verb and can even be possessed and
quantified.
6.5.3 Inherently locational nouns
Inherently locational nouns denote places and names of places (proper names).
Inherently locational nouns do not have to be marked with the mobilitative enclitic
<=sa> (MOB) when they function as Direction adjunct, e.g. (90), (91).

6 NOUNS

118
(90) nogol re'ewa
[nogol]
DIRECTION
{re'e -wa}
market go.away -FACT
[I] went to the market.

(91) hanep bakmara re'eni
[hanep] [bakmara]
DIRECTION
{re'e -ni}
tomorrow Pname go.away -FUT
Tomorrow [we] will go to Baghmara.
6.5.4 Mass nouns
Mass nouns denote substances and distinguish themselves from other nouns in that
they can only be quantified with measure nouns (see 12.4). Among the members of
this subclass of nouns are ja'bek curry, may rice, toy water, thoy' blood, som'
salt, wal' fire, etc. These nouns can take the plural phrasal enclitic <=dora> (p)
which will then indicate a large quantity of the substance denoted by the noun, e.g.
(92). However, depending on the semantics of the noun, sometimes clitisisation of the
plural denotes more than one unit of the substance, e.g. (93).

(92) a ie maydoraaw sa'cawa.
[a] [ie may] =dora =aw {sa' -ca -wa}
1s PRX rice =p =ACC eat -NEG -FACT
I will not eat so much rice.

(93) wal'doraaw nukca imi
[wal'] =dora =aw {nuk -ca} [i] =mi]
fire =p =ACC see -NEG PRX =GEN
[We] dont see the fires from here.
6.5.5 Gender sensitive nouns
There are only a few nouns in Atong which change their phonological shape to agree
with the sex of their referent. These nouns are all Indic loans and are borrowed in
masculine/feminine pairs. The masculine form ends in /a/ and the feminine form in /i/,
All the gender sensitive nouns recorded so far appear in Table 27.

6 NOUNS

119
Table 27 Gender sensitive nouns
MASCULINE gloss FEMININE gloss
boba crazy man bobi crazy woman
raja king rani queen
harata lazy man harari lazy woman
6.6 Juxtaposition of nouns
The interpretation of two or more unmarked juxtaposed nouns depends on the context.
The following interpretations are possible:
1. The nouns belong to the same NP and do not modify each other, the interpretation
is one of addition.
2. The nouns belong to the same NP and one modifies the other.
3. The nouns belong to different NPs with different argument statuses.
6.6.1 Addition interpretation
In case both nouns belong to the same NP but do not modify each other, they share
the headedness of the construction. Examples are given in (94) and (95). In (94) both
nouns, pan wood and wa' bamboo, form the head of the arch NP (an NP
comprising a head and an attributive clause). This head is modified by the attributive
clause (AC) of which the verb mat- to cut is the predicate (see Chapter 29). The
nouns do not modify each other and the relation between them is additional, hence the
translation with an added and in English.

(94) ------------arch NP-----------------
-AC-
[pan wa' {mat} =gaba] =tho =na =an
wood bamboo cut =ATTR =CONTINUOUSLY =DAT =FC/ID
because of continuously cutting wood [and] bamboo (the person got many
wounds on his body).

Another example of an unmarked additional relationship between two nouns within
the same NP is (95). The two nouns are both the O argument of the predicate thik
kha'- (exactly do) to agree upon. Again, their relationship is one of addition.

6 NOUNS

120
(95) takrukna san somay thik kha'wacom.
{takruk} =na [san somay]
O
{thik kha' -wa} =com
fight =DAT day time exactly do -FACT =IRR
to fight, [they] had supposedly agreed upon a day [and] time.

In example (607) in 23.12 we see three unmarked nouns, simen tota tin cement,
plank, corrugated iron in the same NP in an additive relationship. The NP is in O
function in the adverbial clause of which the verb pirin to mix is the predicate.
6.6.2 Modifying interpretation
When, as was given as the second option, the nouns belong to the same NP and one
modifies the other, one noun will be the head and the other the dependent of a
compound noun. The modifier always precedes the head, e.g. macha nokbanthay
(tiger house+bachelor) the tigers bachelors house, where the head is nokbanthay
bachelors house. Other examples are: taw' nok (chicken house) chicken coop,
ma'su khol (cow skin) cow skin, taw' sagoray (chicken child) chick, pan bothon
(tree shadow) the trees shadow, rupek bisi (frog poison) frogs poison muma wa
(elephant tooth) elephants tusk, dawa mahari (Dawa family) the Dawa family and
sol daray (iron sword) iron sword This construction is also used in toponyms (place
names): the name will precede the head which indicates the kind of place it is, e.g. siju
so (Siju village) the village of Siju, dabat wari (Dabat deep.place.in.the.river) the
wari called Dabat, rongdo toykhal (Rongdyng river) Rongdyng river, waymo
ha'bori (Waimong mountain) Waymong mountain and taw'pak khal (bat hole) the
bat cave. Compounds of more than one noun, although infrequent, do occur, e.g.
(633) in 24.1 where we see the quantified compound ha' no' tanran (earth inside
layer) layers of the earths inside.
The compounds we have seen above are all very transparent. Some compounds
developed meaning extensions that are not obviously predictable from the meanings
of the components. The compound di' khal (shit hole) arse, bottom, for example,
has extended its meaning to bottoms of receptacles. Some combinations have lost
their transparency altogether and have to be treated as unanalysable lexical items. The
word toykhal, for instance, can be dissected into the components toy water and khal
hole but the meaning is that of river.
6 NOUNS

121
The classifier that will be used to count compounds like the ones we have just
seen depends on the referent of the compound as a whole, not only on the meaning of
the head. The words sa'goray means human child when it is not modified and is
counted with the classifier for humans, mo'. The compound taw' sagoray (chicken
child) chick refers to an animal and is therefore counted with the classifier for
animals.
Only one compound has been recorded where the head precedes the modifier, viz.
in the word nokbanthay, which can be analysed as nok+banthay (house+bachelor)
bachelors house. This word has the allomorphs nokphandai and nokphande, which
sound more like Garo word /nokpante/ [nokphanthe] with the same meaning.
6.6.3 Different-NP interpretation
The third interpretation of juxtaposed nouns can be that they belong to different NPs
and have different argument statuses. An example of this is given in (823), of which
we present the relevant part here in (96), where we see that the compound moma wa'
(elephant tooth) elephants tusk is the head of an arch NP (an NP that comprises a
head noun and an attributive clause, see Chapter 29) which functions as O in a matrix
clause, hence the accusative enclitic. The quantified noun dora weight measure of
about 5 kg) is the only argument, the copula complement, of the
identificational/equational copula do'- (IE.be), which is the predicate of the
attributive clause (AC, between vertical lines) that modifies the head of the arch NP.

(96) moma wa' dora boroy do'gabaaw
--------------------------arch NP--------------------------
-------------AC-------------
[moma wa' |[dora boroy] {do'}| =gaba] =aw
elephant tooth weight four IE.be =ATTR =ACC
an elephant tusk that weighs four dora

122
Chapter 7 Kinship terms
_____________________________________________________________________

Key to the symbols used in tables this chapter
male _ female x married to

The Atongs have a classificatory Kinship system.
22
The system is typical for Tibeto-
Burman languages, as it distinguishes cross cousins and parallel cousins. Cross
cousins are the children of mothers brother or fathers sister. Parallel cousins are the
children of mothers sister or fathers brother. Cross cousins are considered real
cousins and marriages between such cousins are allowed and encouraged, while
parallel cousins are considered brothers and sisters and are not marriageable.
Furthermore, like the other Tibeto-Burman languages, the Atongs use the same
address terms for mothers brother and father-in-law (mama), and for fathers sister
and mother-in-law (mani). Atong also has some purely descriptive kinship terms,
such as some reciprocal kinship terms, terms to distinguish between maternal and
paternal aunts and uncles and to refer to in-laws and some specific relations between
in-laws. In addition, Atong distinguishes kinship relative age, like most of the
languages in Asia.
To elucidate the use of the term descriptive used above: an Atong speaker can
refer to his/her father-in-law as mama and address him as mama, which is the same
term that is used to refer to and address mothers brother. However if a speaker wants
to be very specific, the term haw'nokhol can be used, which is a referential term that
only denotes father-in-law, and cannot be ambiguous.
In section 1 all kinship terms will be divided into two classes, based on their
morphological properties. Kinship terms can also be divided on the basis of their

22
The Atong classificatory kinship system has a lot in common with that of the Australian Aborigines
as described in Elkin, 1970: 51-58.
7 KINSHIP TERMS

123
semantic oppositions, as is described in section 2. Section 3 describes the address
terms that occur in Atong. Section 4 gives an overview of the terminological
organisation of the consanguineal relations (blood relations). The affinal relations
(relations by marriage, i.e. in-laws) are treated in section 5. Section 6 tells us about the
terminology used when people lose relatives. Finally, in section 7 we will look at how
kinship terms are used to address people who are not kin.
7.1 Morphology-based division of kinship terms: the enclitic <=gaba ~ =ga>
Kinship terms are a subclass of noun and include some nouns which do not denote
kinship relations, e.g. baju friend and bay'siga ~ bay'sega friend. Although these
nouns do not denote kinship relations, but close human relationships, they behave
morphologically the same as kin-denoting nouns, and therefore I consider them to be
part of the subclass of kinship terms
On morphological grounds, kinship terms are divided into two subclasses of
which the members are listed below.

1. Type 1, which cannot attach the derelational enclitic <=gaba ~ =ga> (DREL)
directly to the root. (This type includes the synonymous nouns baju friend and
bay'siga friend.)
2. Type 2, which can enclitisise the derelational enclitic <=gaba ~ =ga> (DREL)
directly to the root. (This type includes the noun jok spouse.)

All kinship nouns can take the possessive derivational enclitic <=tha> (OWN). After
attaching the possessive enclitic <=tha> (OWN) the derelational enclitic <=gaba ~
=ga> (DREL) can be attached, i.e. the form *dotoy=gaba (uncle=DREL) of the Type
1 kinship term dotoy uncle: fathers elder brother is not possible, but
dotoy=tha=gaba (uncle=OWN=DREL) own uncle: fathers elder brother is
grammatically correct.
The data that have been recorded so far make the use of the derelational enclitic
seem almost completely optional. However, there are some principles:

7 KINSHIP TERMS

124
1. Any kinship term used as an address term cannot take the enclitic when you
address your own kin. Thus one cannot address ones elder brother like this: * o
jo=gaba! (interj younger.brother=DREL). The grammatically correct way would
be like this: o jo! (interj younger.brother) Hey younger brother!.

2. However, the enclitic can be used optionally when you address somebody elses
kin, e.g. (97).

(97) o samratmi haw'(gaba)!
o [samrat =mi haw'] (=gaba)
interj Pname =GEN uncle:mothers.brother =DREL
Hey Samrats uncle!

Thus, kinship terms can only be used with <=gaba> (DREL) when speaking about
someone elses kin or friend and not when you speak about your own kin or friend.
It seems that the enclitic <=gaba ~ =ga> marks derelationalisation when certain
inherently relational nouns denoting kinship and close human relationships are used in
less- or non-prototypical situations. (Shin 2004: 67, 68 is insightful on this matter.)
The prototypical situation is one where an inherently relational noun occurs with its
prototypical possessor, a speech act participant. In this prototypical situation the
inherently relational noun is most likely to be unmarked. Example (98) is illustrative.
In this example we see the kinship term sa'banthay son with a prototypical
possessor: a, the first person singular.

(98) o ie a sa'banthay co'motan bebe nookno.
o [ie] {a sa'banthay} [comot] =an [bebe] {no -ok} =no
interj this 1s son really =FC/ID truly say -COS =QUOT
O, this [is] really and truly my son, he said, it is said.

Less prototypical situations, in which a third person (i.e. non-speech act participant)
occurs with an inherently relational noun are more likely to be marked, e.g. (99), and
most likely to be marked are non-prototypical situations in which the inherently
relational term is used without any possessor, as in (100). In other words, the
derelational marking allows certain inherently relational nouns to function in non-
7 KINSHIP TERMS

125
prototypical situations. This is similar to what Lehmann (2003:73-75) calls
absolutivisation in Yucatec (Mayan, Yucatan Peninsula, Mexico).
In example (99) we see how the kinship term sa'banthay is marked with the
derelational enclitic, while it occurs as the Possessum of a third person. The Possessor
is implied and is coreferential with the A argument of the clause, morot man.

(99) morot sa'banthaigabaaw konci baaymu daw're kawwano.
[morot]
A
[sa'banthay]
O
=gaba =aw [kon] =ci {ba} =ay =mu
man son =DREL =ACC back =LOC carry.a.child=ADV =SEQ
[daw're] {kaw -wa} =no
eagle shoot -FACT =QUOT
A man carrying [his] son on [his] back shot the eagle, it is said.

In example (100), the kinship term jogaba younger brother occurs without even an
implied possessor, and is therefore marked with the derelational enclitic.

(100) otokma'ciba ue jogabae katha ra'cano.
otokma'ciba [ue jo] =gaba =e [katha] {ra' -ca} =no
but that younger.brother =DEREL =FC word get -NEG =QUOT
But the younger brother did not obey [his fathers] words, it is said.

Since derelational marking is not obligatory, we also find counterexamples in Atong
of the marking mechanism described above, e.g. (101), where the classificatory
kinship term naw younger sister is unmarked. In the example below, the kinship
term naw younger sister does not carry the derelational enclitic despite the fact that
it occurs with a third person possessor.

7 KINSHIP TERMS

126
(101) ge'hemi nawaw ado khomthirini.
[ge'the =mi naw] =aw [a] =do {khom -thiri -ni}
3s =GEN younger.sister =ACC 1s =TOP marry -AGAIN -FUT
I will again marry his younger sister.

Conversely, the same morpheme <gaba ~ ga> is attested as relational morpheme on a
few other nouns, not belonging to the subclass denoting close human relationships and
not inherently relational. The morpheme <gaba- ~ ga> (RELATIONAL) has changed
the meaning of these nouns to relational nouns. Examples are given in Table 28
below. Since attachment of the relational morpheme on nouns is not productive we
have to consider these formations as fossilised. Synchronically, nouns with the
relational morpheme are morphologically non-transparent lexical items.

Table 28 Nouns occurring with the morpheme <-gaba ~ -ga> (RELATIONAL)
lexical item gloss of parts meaning
gawi-gaba female-RELATIONAL wife
bipha-gaba male-RELATIONAL husband
nok-gaba house-RELATIONAL landlord, house owner
7.2 Semantic division of kinship terms
There are three different types of semantic opposition between kinship terms in
Atong.

1. classificatory versus descriptive
2. reciprocal versus non-reciprocal
3. reference terms versus address terms

These different semantic oppositions will be commented on one by one below.
7.2.1 Classificatory versus descriptive kinship terms
When a descriptive term is used, it can only represent one type of relationship
between two people, while a classificatory term represents one of many different
types of relationships. Table 29 and Table 30 give an overview of all kinship terms.
The different referents of classificatory terms and those of the descriptive ones are
given in Table 29 and Table 30. All classificatory kinship terms can be used as
7 KINSHIP TERMS

127
address terms, except for the purely referential terms sa'moncok daughter and
sa'banthay son.
23

7.2.2 Reciprocal versus non-reciprocal kinship terms
Reciprocal kinship nouns denote a relation between two or more people, whereas non-
reciprocal kinship nouns denote the unidirectional relation of one person to another.
The term gumi brother-in-law: elder sisters husband, for instance, denotes the
relation of a person to his elder sisters husband, but not the inverse relation and is
thus a non-reciprocal kinship term. The term used to indicate the relation of
someones elder sisters husband to this someone is josori brother-in-law: wifes
younger brother. The relation of men who married women that are sisters is called
sadu, which is a reciprocal kinship term. These men will call each other sadu and will
call each others elder brother sadu cuga(ba) (sadu big=ATTR) big sadu and each
others younger brother sadu molga(ba) (sadu small=ATTR) small sadu. Note that the
relative enclitic has two allomorphs, viz. <=gaba ~ =ga> (ATTR).
The terms baju, bay'siga ~ bay'sega, camay ~ came, sadu, bono, mawsa ~ mosa,
(see Table 29 for glosses) are real reciprocal terms, in that they only denote
reciprocal relations. The term ja'cu the relation of a man and his wifes elder sister,
or of a woman and her younger sisters husband or sister-in-law: elder brothers
wife is confusing because speakers in neither Badri nor Siju agreed on its meaning.
Therefore it is represented twice in Table 29 (a) and (c). Some speakers thought that
ja'cu and boci are synonyms, while others said that boci is actually Garo. According
some speakers ja'cu is a reciprocal kinship term, and according to others it is not.
There are two reciprocal kinship terms that also denote non-reciprocal relations.
The terms acu grandfather and baba father are only used to indicate the
relationship between grandfather and grandchild and between father and son, when

23
The kinship term sa'banthay can be split up into two existing lexemes, viz. sa' offspring, child and
banthay bachelor. We recognise the same morpheme, sa' offspring, child in the kinship term
sa'moncok daughter. The second element of this term, moncok female is not a lexeme in Atong at
present, but is cognate with the Garo word me'cik female.
7 KINSHIP TERMS

128
they are used as address terms. This means that the grandchild can call his/her
grandfather acu and the grandfather can call his grandchild acu back. It is reported
that the reciprocal use of baba father is falling into disuse. Hence the following
recorded comment.

(102) wa'gabaawba baba noariate dakami casodo. ami wa'aw baba nohiba
aawba baba noaria.
[wa'] =gaba =aw =ba [baba] {no -ari -a} =te
father =DREL =ACC =EMPH father say -SIMP -CUST =DCL
[daka =mi caso] =do
before =GEN generation/era =TOP
[a] =mi [wa'] =aw =ba [baba] {no} =ci =ba
1s =GEN father =ACC =ADD father say =LOC =INDEF
[a] =aw =ba [baba] {no -ari -a}
1s =ACC =ADD father say -SIMP -CUST
[We] just said baba to [our] father, Im telling you, (in) the past
generation/era. Whenever [I] would say baba to my father, [he] would just say
baba to me too.

The reciprocal kinship terms all belong to Type 1, except for jok spouse, which
denotes the relationship of a married couple.
7.2.3 Reference versus address kinship terms
Reference terms are used to talk about someone, whereas address terms are used to
get someones attention. Whereas all address terms can also be used as reference
terms, the inverse is not the case. Address terms are treated in more detail in the next
section.
Table 29 gives an overview of Type 1 kinship nouns, i.e. those kinship nouns that
cannot attach the derelational enclitic <=gaba ~ =ga> (DREL) directly to the root.
This table is divided in three sections (a) reciprocal terms, (b) consanguineal terms
(terms of blood relations) and (c) affinal terms (relations by marriage, in-laws). Table
30 gives an overview of Type 2 kinship terms, i.e. those that can attach the
derelational enclitic directly after the root. As we can see in these tables, some kinship
terms have extended meanings, and some are used to address persons that are not kin.
7 KINSHIP TERMS

129
7.3 Address terms
An address term is used to address someone or call someone to get this persons
attention. All address terms can also be used referentially, while not all referential
terms can be used as address terms. The noun jow' mother, for instance, cannot be
used as an address term. To address his/her mother, an Atong speaker uses the noun
ama mother. Table 29 and Table 30 indicate whether a kinship term can be used as
address term or not. There are several remarks to be made about the information in
these tables.
All kinship terms that can be used reciprocally can be used as address terms. All
kinship terms beginning with /a/ can be used as address terms. Some of these, all core
family terms, have a corresponding referential form without the initial /a/, viz. woy
awoy grandmother, wa' awa father (note the absence of glottal stop in the
address term), soy asoy aunt: fathers younger sister, wa awa uncle: mothers
younger brother, nay anay aunt: mothers sister and no ano aunt: mothers
sister. Both the forms with and without initial /a/ can be used referentially, but only
the forms with the /a/ can be used as address terms. This has led me to believe that the
/a/, in these cases, is a fossilised vocative prefix that has now become part of the stem.
Matisoff (1982: 65) describes a prefix a- for Lahu that forms vocatives of kinship
terms, and remarks that [t]his kin-prefix is extremely widespread in Tibeto-Burman.
7 KINSHIP TERMS

130
Table 29 Kinship terms Type 1: (a) Consanguineal kinship terms
r
e
c
i
p
r
o
c
a
l

Term Meaning as reference term
Can be
used as
address
term
Extended
meaning
Other uses
or referents
baju friend yes
bay'siga ~
bay'sega
friend yes
camay ~
came
1. female cross cousin
2 the relation of female cross
cousins
3. the relation of the parents of a
married couple
yes
girlfriend
(in the sense
of lover or
sweetheart)

sadu
brother-in-law: the relation of men
who married women that are
sisters
yes
bono
brother-in-law: the relation of a
man and his younger sisters
husband or a man and his wifes
elder bother
yes
mawsa ~
mosa
1. male cross cousin
2. the relation of male cousins
from intermarriageable families
yes
a male
friend
belonging to
a
marriageable
family

ja'cu
1. (Siju and Badri dialects) the
relation of a man and his wifes
elder sister, or of a woman and her
younger sisters husband
2. (Badri dialect) sister-in-law:
elder sister of ones wife or elder
brothers wife
yes
acu grandfather yes
the relation
of
grandfather
and
grandson,
only when
used as
address term
to address
an elderly
man
baba father yes
the relation
of father and
son, only
when used
as address
term
to refer to or
address an
uncle

7 KINSHIP TERMS

131
Kinship terms Type 1 continued: (b) Consanguineal kinship terms
c
o
r
e

f
a
m
i
l
y

Term Meaning as reference term
Can be
used as
address
term
Other uses or referents
abu grandmother yes
to address an unrelated elderly woman
awoy (archaic) grandmother yes
akay aunt: mothers elder sister yes
to address an unrelated married woman
older than the speaker
asoy ~ asi aunt: mothers younger sister yes
mama uncle: mothers brother yes
to address an unrelated man older than
the speaker, to address ones father-in-
law
ama mother yes
to talk about or address an aunt, used
by a mother to address her daughter.
awa (archaic) father yes
abi elder sister yes mothers-in-law can call each other abi
nono younger sister yes
to talk about or address a related
younger female of your generation:
cousin, to address a young unrelated
female person younger than the speaker
dada elder brother yes
to speak about or address a related
older male of your own generation:
cousin, or to address an unrelated man
older than you
jojo younger brother yes
to talk about or address a related
younger male of your own generation:
cousin, to address a young unrelated
male person younger than the speaker
ano aunt: fathers sister yes
sister-in-law: husbands elder sister
no aunt: fathers sister yes
anay aunt: fathers sister yes
address term for mother-in-law
mani aunt: fathers sister yes
dotoy uncle: fathers elder brother yes address term for father-in-law
awa
uncle: fathers younger
brother
yes children call their stepfather awa
ade stepmother yes
khoroytha nephew yes
namcok niece yes
namgaba niece no

7 KINSHIP TERMS

132
Kinship terms Type 1 continued: (c) Affinal kinship terms
i
n
-
l
a
w
s

Term Meaning ass reference term
Can be used as
address term
Address term
gumi
brother-in-law: elder sisters husband or
husbands elder brother
yes
biawtha brother-in-law: wifes elder brother yes
boci ~ ja'cu
sister-in-law: elder brothers wife or
elder sister of ones wife
yes
nawsori sister-in-law: younger brothers wife yes
josori
brother-in-law: spouses younger
brother
yes
nay'nokhol mother-in-law no mani/anai
haw'nokhol father-in-law no mama
konokhol son-in-law yes
namnokhol daughter-in-law yes

Table 30 Type 2 kinship terms, consanguineal and affinal
Term
Meaning as
reference term
Can be
used as
address
term
address term Other uses or referents
jok spouse yes
jow' mother no ama
wa' father no awa (archaic), baba
soy
uncle: mothers
younger brother
no asoy ~ asi
haw'
uncle: mothers
brother
no mama
nay'
aunt: fathers
sister
no ano, anay, mani, no
wa
uncle: fathers
younger brother
no awa
children can refer to their
stepfather with wa
naw younger sister yes
(alternatively, and more
respectfully) nono
to address a younger female cousin
or an unrelated woman younger
than the speaker
jo
younger
brother
yes
(alternatively and more
respectfully) jojong
to address a younger male cousin
or an unrelated man younger than
the speaker
ja'naw elder sister yes
respectfully) abi
to address an older female cousin
or an unrelated woman older than
the speaker
paw'jo elder brother yes
respectfully) dada
to address an older male cousin or
an unrelated man older than the
speaker

7 KINSHIP TERMS

133
Type 2 kinship terms continued
Term
Meaning as
reference term
Can be
used as
address
term
address term Other uses or referents
sow' grandchild no
cucu grandchild, acu
the relation of
grandfather and
grandson
(no information on other reference)
sa'banthay son no
the name of the child,
baba the relation of
father and son
to refer to the children of my elder
or younger brother
sa'moncok daughter no
the name of the child,
ama the relation of
mother and daughter
to refer to the children of my elder
or younger brother
7.4 The consanguineal family from the perspective of a me
The consanguineal family, or blood relations, are the kin without the in-laws, or those
relatives that are related by birth and not by marriage. Table 31 below is a family
diagram with reference and address terms from the perspective of a me, i.e. the
table shows how I would call the members of my blood relations. The diagram
organises the kinship terms by generation. Within my generation and my parents
generation there is a subdivision into older and younger siblings. The underlined
lexical items can be used as address terms. As we can see, there are specific reference
terms for cross cousins, viz. mawsa ~ mosa male cross cousin and camay ~ came
female cross cousins whereas the parallel cousins are treated as younger siblings in
both reference and address terms. The cross cousins, which are marriageable to a
me, are the children of anay fathers sister and mama mothers brother. The
children of akay mothers elder sister, dotoy fathers elder brother are
unmarriageable cousins, asoy mothers younger sister and awa fathers younger
brother are parallel cousins and are not marriageable to a me. The word camay ~
came has an additional meaning sweetheart, lover, girlfriend. People usually address
their marriageable cousins as an older or younger sister or brother. As was mentioned
above, the word for marriageable male cousin, mawsa ~ mosa, is also used to address
unrelated men of a marriageable family in a familiar way, like friends.
Small children can be addressed with their name, or, when it is a boy, with babu,
and when it is a girl with rni [rani], which is an Indic loan, related to Hindi
(rn) queen. I have not witnessed a parent addressing their children with sa'goray
child, except when they are angry. Small boys can be addressed by their parents
7 KINSHIP TERMS

134
with baba and by other people with babu. I have no information about special address
terms for small girls. When the children grow up and are not very young any more,
their parents will call them exclusively by their name.
The people in the generation above me, i.e. parents, aunts and uncles, refer to me
as sa'banthay son or sa'moncok daughter, except for mothers brothers, mama,
and fathers sisters, nay ~ no, who refer to me as khoroytha nephew and namgaba
niece. Table 33 gives a detailed overview of this information and of how cousins
from marriageable families address each other. A person can speak about his or her
aunts and uncles as if they were his or her parents. Thus it can occur that the
fieldworker walks around the village with a friend and is amazed at the large number
of fathers and mothers this friend has. The children of akay aunt: mothers elder
sister and dotoy uncle: fathers elder brother are treated as my siblings, as we expect
in a Tibeto-Burman classificatory kinship system.
Table 32 gives an overview of the kinship terms of spouses of aunts, uncles and
siblings and how their grandchildren and/or children call me. In that table the address
terms are underlined. Remember that all address terms can also be used referentially.
The table should be read as follows. My asoy is married to a man I call awa or wa.
Their children are my mawsa ~ mosa and camay ~ came and their grandchildren are
my nono and jojo. The grandchildren call me ama when I am a female and baba
when I am a male.
7 KINSHIP TERMS

135
Table 31 My blood relations.
The underlined forms can be used as address terms. All address terms can
also be used referentially. The non-underlined forms can only be used as
reference terms. The colours match the cousins to their parents.
g
r
a
n
d
p
a
r
e
n
t
s

abu, awoy, woy'gaba
grandmother
_ acu
grandfather
m
y

p
a
r
e
n
t
s

g
e
n
e
r
a
t
i
o
n

o
l
d
e
r

s
i
b
l
i
n
g
s

maternal paternal
akay
aunt
_ haw', haw'gaba
mama
uncle
nay', nay'gaba,
no
anay, ano, mani
aunt
_ dotoy
uncle
p
a
r
e
n
t
s

jow', jow'gaba, ama,
mother
_ wa', wagaba, baba, awa
father
y
o
u
n
g
e
r

s
i
b
l
i
n
g
s

soygaba,
asoy ~ asi
aunt
_ haw',
haw'gaba,
mama
uncle
nay', nay'gaba,
no
anay, ano, mani
aunt
_ wa, wagaba,
awa
uncle
m
y

g
e
n
e
r
a
t
i
o
n

o
l
d
e
r

s
i
b
l
i
n
g
s

ja'naw, ja'nawgaba,
abi
elder sister
_
a
me
_ dada, phaw'jo, phaw'jogaba
elder brother
c
o
u
s
i
n
s

camay ~
came
cousin
_ mawsa
~ mosa
cousin
naw, nawgaba
nono
younger sister
_jo, jogaba,
jojo
younger brother
marriageable cross cousins unmarriageable parallel cousins
y
o
u
n
g
e
r

s
i
b
l
i
n
g
s

naw, nawgaba, nono
younger sister
_ jo, jogaba jojo
younger brother
m
y

c
h
i
l
d
r
e
n

samoncok, sa'moncokgaba
daughter
_ sabanthay, sa'banthaygaba
baba
son
m
y

g
r
a
n
d
c
h
i
l
d
r
e
n

_ sow', sow'gaba, cucu

The Atong are matrilineal. They inherit their family names, and therefore their clan
membership, from their mother. This is discussed in detail in Chapter 1.
7 KINSHIP TERMS

136

Table 32 Spouses of aunts, uncles and siblings, their children and grand children and
their relation to me

my aunt.uncle, brother or
sister
term for husband or wife
The parents are my
asoy, soy
aunt: mothers younger sister
x
_ awa, wa
uncle: father;s younger brother
Their children and their
spouse are my
_ mawsa ~ mosa
marriageable male cousin

camay ~ came
marriegeable female cousin
Their grandchildren are
my
naw, nono
younger sister

_ jo, jojo
younger brother
Their grandchildren call
me
ama mother, _ baba father
The parents are my
_ mama
uncle: mother;s brother
x
anay, nay'
aunt: fathers sister
spouse are my
_ mawsa ~ mosa
marriageable male cousin

camay ~ came
marriegeable female cousin
Their grandchildren are
my
naw, nono
younger sister

_ jo, jojo
younger brother
Their grandchildren call
me
ama mother, _ baba father
The parents are my
ja'naw, abi
elder sister
x
_ gumi
brother-in-law: elder sisters husband
spouse are my
namgaba, namcok
niece

_ khoroytha
nephew
These children call me _ mama uncle: mothers brother asoy ~ asi aunt:mothers younger sister
The parents are my
_ phaw'jo, dada
elder brother
x
boci
sister-in-law: elder brothers wife
spouse are my
sa'moncok
daughter

_ sa'banthay
son
These children call me
_ wa, awa uncle: fathers younger brother
nay', noanay, ano, mani aunt: fathers sister
The parents are my
naw, nono
younger sister
x
_ bono
brother-in-law: younger sister;s
husband
spouse are my
namgaba namcok
niece

_ khoroytha
nephew
These children call me _ mama, uncle: mothers brother akay aunt: mothers elder sister
The parents are my
_ jo, jojo
younger brother
x
nawsori
sister-in-law: younger brothers wife
spouse are my
sa'monhcok,
daughter

_ sa'banthay
son
These children call me
_ dotoy uncle: fathers elder brother
nay', noanay, ano, mani aunt: fathers sister
7 KINSHIP TERMS

137
Table 33 Reference terms uncles and aunts use for me.
The underlined forms can be used as address terms or reference terms,
whereas the non-underlined forms can only be used as reference terms.
These people refer to a me as
m
y

p
a
r
e
n
t
s

g
e
n
e
r
a
t
i
o
n

soy, asoy
aunt: mothers younger sisiter

sa'moncok
daughter
_
sa'banthay
son
wa, awa
uncle: fathers younger brother
akay
aunt: mothers elder sisiter
dotoy
uncle: fathers elder brother
nay, no, anay, ano, mani
aunt: fathers brother

namgaba
niece
_
khoroytha
nephew
mama
uncle: mothers brother
m
y

g
e
n
e
r
a
t
i
o
n

mawsa ~mosa
1. marriageable male cousin 2 the relation of
cousins from intermarriageable families
came _ mosa
camay ~ came
1. marriageable female cousin,
2 the relation of female cousins from
marriageable families
came _ mosa
7.5 The in-law family
Table 34 gives an overview of the referential terms for in-laws from the perspective of
a male spouse, whereas Table 35 does the same from the perspective of a female
spouse. Some terms have to be added here, which I will introduce from the
perspective of a me. My parents and the parents of my spouse address each other
as camay ~ came the relation of the parents of a married couple. The referential term
of this relation is nokcama the relation of parents of a married couple. My anay
mother-in-law and my ama mother can also call each other abi elder sister. In the
dialect of Siju, my wifes elder sister and me call each other ja'cu the relation of
man and his wifes elder sister or of a woman and her younger sisters husband. The
word ja'cu in the Badri dialect means either sister-in-law: elder sister of ones wife
or sister-in-law: brothers elder sister. Elder brothers wife is referred to and
addressed as boci in the Siju dialect.
7 KINSHIP TERMS

138
Table 34 My in-laws, me being masculine.

nay'nokhol, anay, mani
mother-in-law
x
_ haw'nokhol,
mama
father-in-law
wifes parents
ja'cu
the relation of man and his wifes
elder sister or of a woman and her
younger sisters husband

_ biawtha
brother-in-law
wifes older siblings
_ a
me
x
gawigaba
wife

nawsori
sister-in-law

_ josori
brother-in-law
wifes younger
siblings

Table 35 My in-laws, me being feminine.
husbands parents
nay'nokhol, ,
nay, anay, mani
mother-in-law
x
_ haw'nokhol,
mama
father-in-law
husbands older siblings
ano ~ no
sister-in-law

_ gumi
brother-in-law

_ biphagaba
husband
x
a
me
husbands younger
siblings
nawsori
sister-in-law

_ josori
brother-in-law

The referential terms that the parents-in-law use to indicate me and my siblings are
given in Table 36. According to some Atong speakers, the terms konokhol son-in-
law and namnokhol daughter-in-law are used to refer to and address both the
son/daughter-in-law and his/her siblings. Other Atong speakers say that the siblings
are referred to and addressed by different terms, viz. khoroytha nephew and
namgaba/namcok niece.

7 KINSHIP TERMS

139
Table 36 Address terms that my in-laws use for me and my siblings.
The underlined terms can be used as address terms, the non-underlind
terms are only referential.
These persons

use
this referential/address
term
for these referents.
nay'nokhol
mother-in-law,
haw'nokhol
father-in-law
A
c
c
o
u
n
t

1

konokhol
son-in-law
a _ me, dada [my] elder
brother, jojo [my] elder
sister

namnokhol
daughter-in-law
a me, abi [my] elder
sister, nono [my] younger
sister
A
c
c
o
u
n
t

2

konokhol
son-in-law
a _

namnokhol
daughter-in-law
a
khoroytha
dada [my] elder brother,
jojo [my] elder sister
namgaba, namcok
abi [my] elder sister
nono [my] younger sister

The address terms that brothers- and sisters-in-law use for a me are the same as the
reference terms and are represented in Table 37.

Table 37 Address terms that my brothers- and sisters-in-law use for me.
(All address terms can also be used referentially).
These people call me (a)
gumi
brother-in-law: elder sisiters husband

josori _
brother-in-law: wifes
younger brother
nono
younger sisiter
boci (Siju dialect), ja'cu (Badri dialect)
sisiter-in-law: elder brothers wife or elder
sister of ones wife
bono
brother-in-law: younger sisters husband

bono _
brother-in-law: wifes
elder brother
nono
younger sisiter
nawsori
sister-in-law: younger brothers wife

gumi _
brother-in-law:
husbands elder brother
boci
sister-in-law:
husbands elder sister
7.6 Family loss
The reference term for widow and widower is jokoroy. When a widower marries
again, his children will call his new wife ade stepmother. When a widow marries
again, her new husband will be addressed as awa fathers younger brother by her
children. A stepmother/stepfather refers to her/his stepchildren as sa'moncok
daughter and sa'banthay son or just sa' child or sa'gyray child. The word for
stepchild, used to refer to someone elses stepchildren, is sa'thora. A child who lost
7 KINSHIP TERMS

140
his mother is called jow'=ri (mother=LOST). A child who has lost his father is called
wa'=ri (father=LOST). When a child lost both his parents, an orphan, it is referred to
as wa'=ri jow'=ri (father=LOST mother=LOST).
24

7.7 How to address people who are not kin
As Table 29 and Table 30 show us, some address terms can only be used to address a
person the speaker is related to, i.e. kin. The term bono brother-in-law: the relation
of a man and his younger sisters husband can only be used to address your bono
and the term namcok niece, can only be used to address your namcok. Other kinship
terms can also be used to address persons to whom the speaker is not related, i.e. non-
kin. The choice of kinship term to address an unrelated person depends on the
following factors:
1. age of the speaker relative to the addressee, older/younger
2. sex of the addressee, male/female,
3. level of familiarity and respect, not respectful/respectful
Scenarios where the addressee is younger than the speaker are described in 7.7.1,
and scenarios where the addressee is older in 7.7.2.
Some address terms are only used between people who are approximately the
same age. The terms mawsa ~ mosa male cousin of marriageable family, can be
used by men to address other, non-related, men in a friendly and familiar way.
Usually it is used to address someone from a family that the family of the speaker can
intermarry with. There are two words for friend, the more intimate baju, which I
heard women use as well as men, and the less intimate bay'siga ~ bay'sega, which I
only heard men use.

24
The NP wa'ri jow'ri consists of two suffixed nouns in coordination.
7 KINSHIP TERMS

141
7.7.1 Addressee is younger than the speaker
When the addressee is a small child, male or female, it can be addressed as sa'goray
child. When it is a boy, it can be addressed as babu, and when it is a girl you can call
it rni [rani], which is an Indic loan, similar to Hindi (rn) queen. Remember
that the address term baba is only used to address the speakers own small male
children. A group of children can also be addressed as sa'goray, when some of them
are the speakers kin and some are not. Small children can also be addressed with
their name. When the addressee is female and not a small child any more, she can be
addressed as naw younger sister or, more respectfully, nono younger sister. When
the addressee is a younger male, the term of address would be jo younger brother
or, more respectfully jojo, younger brother People of the same age or younger than
the speaker can also be addressed with their name, if it is known to the speaker. Using
proper names is an intimate way to address someone.
7.7.2 Addressee is older than the speaker
An older female can be addressed as ja'naw elder sister, or more respectfully as abi
elder sister, when she is just a little older than the speaker. When she is lot older
than the speaker, but could not yet be the speakers grandmother, she can be
addressed as akay aunt: mothers elder sister. This term is used especially when
unmarried boys and girls address an unrelated woman who is most probably married.
Old women are addressed as abu grandmother. An older male can be addressed as
phawjo elder brother or more respectfully as dada older brother or very
respectfully as mama uncle: mothers brother. Old men are addressed as acu
grandfather. It is offensive to call people older than you by their first name or to talk
about them using their first name. If you want to talk about someone older than
yourself who is not related to yourself or (one of) the persons you talk to, you either
call them by the name of their first child, e.g. dambe wa' (Dambe father) Dambes
father or dambe jow' (Dambe mother) Dambes mother, or you talk about them in
terms of their kinship relation to someone else.

142
Chapter 8 Demonstratives
_____________________________________________________________________

Demonstratives are a closed class containing two members, viz.

ie ~ i proximal demonstrative (PRX) (first deictic degree),
ue ~ u distal demonstrative (DST) (second deictic degree).

The bound forms (PRX) and (DST) of the demonstratives are used when
enclitics immediately follow the demonstratives. The free forms are used when there
are no enclitics immediately following the demonstratives. A demonstrative can take
all case markers and can occur as all possible types of argument (see Table 58). Other
properties of the demonstratives are treated here below. The deictic-only
demonstratives are a separate word class, described in 8.8.
8.1 Deictic properties
No matter what syntactic function it has, a demonstrative can be used anaphorically or
purely deictically. Moreover, the demonstratives can be used for both substitution
anaphora and textual anaphora (cf. Dixon 2003). I have no clear cut examples of
demonstratives used for cataphora in the language. For textual cataphora Atong uses
the adverbial demonstrative otokoy treated below in 8.7.
8.1.1 Purely deictic use
Here are some examples of deictically used demonstratives with various case
markings.

Locational deixis:
(103) na' bayk ici tanbo [speaker points with his head].
[na' bayk] [i] =ci {tan} =bo
2s motorcycle PRX =LOC put =IMP
Put your bike here [speaker points with his head].

The demonstrative can always be followed by another locative-marked NP in
apposition to the demonstrative phrase, e.g. (104)
8 DEMONSTRATIVES

143
(104) na bayk ici bothonci tanbo.
[na' bayk] [i] =ci [bothon] =ci {tan} =bo
2s motorcycle PRX =LOC shade =LOC put =IMP
Put your bike here in the shade.

Goal deixis:
(105) a icina se'khalay ray'ana naacom.
[a] [[i] =ci =na [se' -khal} =ay {ray'} =na
1s PRX =LOC =ALL early-CP =ADV go =DAT
{na -a} =com
need -CUST =IRR
I should have come here earlier

Source deixis:
(106) toy umi jokaydok.
[toy] [u] =mi {jok -aydok}
water DST=ABL escape -PROG
The water comes out from there.

(107) usangmi ray'acwa. dolong nosto don'ok.
[u =sa =mi {ray'a -ca -wa} [dolo] {nosto do' -ok}
DST=MOB =ABL come -NEG -FACT bridge damage IE.be -COS
[They] will not come from there. The bridge is damaged.

Pathway deixis:
(108) utokoy ray'na man'ca. moma pa'a.
[u] =tokoy {ray' =na} {man' -ca} [moma] {pa -a}
DST=VIA go =DAT be.able -NEG elephant many -CUST
[We] cant go like that/via that way. There are many elephants.

Comitative adjunct deixis:
(109) umo ray'cawa.
[u]=mo {ray' -ca -wa}
DST=COM go -NEG -FACT
[I] will not go with him.
8 DEMONSTRATIVES

144
8.1.2 Anaphora
The demonstrative functions as a pronoun. The next example illustrates how the distal
demonstrative u (DST) refers back to the location, Badri, mentioned in the previous
clause.

(110) [] gandruawsa badri mocewano. uci mu'butu somayci badri nemen
man'ay sa'ano.
[gandru] =aw =sa [badri] {mo -ce -wa} =no
Pname =ACC =DLIM Pname call.a.name -first -FACT =QUOT
[u] =ci mu' -butu +somay} =ci
DEM =LOC stay -WHILE +time =LOC
[badri] [nemen] {man'} =ay {sa' -a} =no
Pname very in.great.amounts =ADV eat -CUST =QUOT
[] Gandrung was the first [village] to be called Badri, it is said. At that time
when [they] were living there, Badri was very rich, it is said.

In the following example we see how the distal demonstrative refers back to the NP
te'ewrawrawmi gawi in the previous clause.

(111) Speaker N: otokaria, te'ewrawrawmi gawido.
Speaker S: unan symsakna naaro.
{otok -ari -a} [te'ew =rawraw =mi gawi =do
do.like.that-SIMP -CUST now =CONTINUOUSLY =GEN girl =TOP
[u] =na =an {som -sak} =na {na -a} =ro
DST=DAT =FC/ID follow -APPROPRIATELY =DAT must -CUST =EMPH
Speaker N: They just do like that, the girls from now on. (i.e. the girls of
today)
Speaker S: For those ones in particular you have to be careful. (Lit. you
have to follow appropriately for them.)

(112) hay sigyret hyn'etsora na'a uaw.
[hay] [sigoret] {hon' -et -sora} [na'a] [u]=aw
come.on cigarette give -CAUSE -TOTALLY 2s DST=ACC
Come on, give the cigarettes, oh you, those!
8 DEMONSTRATIVES

145
Demonstratives
can be head of a predicate of an identity/equation clause, in which case
it has to be marked by the focus/identifier enclitic <=an> (FC/ID), e.g.
[daba] {i=an} (coconut PRX=FC/ID) this is a coconut.
can be an argument, core or oblique.

In the next example we see the proximal demonstrative as an oblique argument
(peripheral argument/oblique), viz. a Facsimile, hence marked by the similative
enclitic <=tokoy> (LIKE). The demonstrative in this example, the last sentence of a
story, refers back to the story that has just been told.

(113) ado itokoy balaymu tanarinaka.
[a =do [i] =tokoy] {bal} =ay =mu {tan -ari -naka}
1s =TOP PRX =LIKE tell =ADV =SEQ put -SIMP -IFT
Having told like this, I will now just stop.

In the next example we see the proximal demonstrative i=tokoy (PRX=LIKE) like
this used deictically, referring to a trampling movement with the feet that the speaker
makes while saying the sentence.

(114) utokoyimu uaw do'rego wadacoawdo acu ambido tawnaan do'rego
wa'daco jatram saphayram noaymu samaw ca'aw itokoy [gestures] tokano.
utokoymu [u =aw do'rego wa'daco] =aw =do [acu ambi] =do
CONJ DST=ACC Pname =ACC =TOP grandpa grandma =TOP
{taw} =na =an [do'rego wa'daco] [jatram saphayram] {no }=ay
go.up =DAT =FC/ID Pname type.of.plant type.of.plant say =ADV
=mu [sam] =aw [ca'] =aw [i] =tokoy {tok -a} =no
=SEQ medicine =ACC foot/leg =ACC PRX =LIKE beat -CUST =QUOT
As for Dorenggo Wadachong, in order to go up on Dorenggo Wadachong,
our ancestors beat so called jatram and saphayram medicinal plants with their
feet like this [gestures], it is said.
8 DEMONSTRATIVES

146
8.3 Properties as head of a predicate
Demonstratives are attested with a limited number of predicate marking categories
and clausal enclitics, viz. negative polarity <-ca> (NEG), irrealis <=com> (IRR),
speculative modality <=khon> (SPEC), examples (115) and (116) below are
illustrative. No evidence exists that demonstratives can take aspectual suffixes. Given
that a demonstrative predicate head can be negated, I predict that it can also take the
change of state suffix <-k> (COS), which is the allomorph that follows the negative
suffix <-ca> (NEG).
In the following example we see, in the last clause, how the proximal
demonstrative functions as head of the predicate and is marked by the phrasal
focus/identifier enclitic <=an> (FC/ID) and the irrealis clausal enclitic <=com>
(IRR).

(115) na'nami garohils ie indiami no'ci nembatgaba ha'kogore, nembatgaba
ha'saldo iancom.
[na'na] =mi garohils] {ie india =mi no' =ci] {nem -bat} =gaba
1pi =GEN Pname PRX Pname =GEN inside =LOC good -MOST =ATTR
ha'kogore] [{nem -bat} =gaba ha'sal] =do
area that falls under one headman good -MOST =ATTR fertile land =TOP
{[i] =an} =com
PRX =FC/ID =IRR
Our Garo Hills, this [is] the best area which falls under one headman in India,
this was, but is not any more [because the indigenous population did not know
how to make good use of it], the most fertile land.

In the example below, we see that the proximal demonstrative functions as head of the
predicate of two main clauses. In the first clause the demonstrative is marked with the
phrasal focus/identifier enclitic <=an> (FC/ID) followed by the quotative and
speculative clausal enclitics <=no> (QUOT) and <=khon> (SPEC), in their typical
order (see 26.9). In the second clause, we see the demonstrative marked by the
phrasal focus/identifier enclitic <=an> (FC/ID) and the speculative and declarative
clausal enclitics <=khon> (SPEC) and <=te> (DCL) respectively.

8 DEMONSTRATIVES

147
(116) iannokhon. amami garu ramgabaci de'tgaba iankhonte ie.
{[i] =an} =no =khon [[ama =mi garu {ram} =gaba] =ci]
PRX =FC/ID =QUOT =SPEC mother =GEN mustard.leaves dry =ATTR =LOC
{de'et} =gaba] {[i] =an} =khon =te [ie]
shit =ATTR PRX =FC/ID =SPEC =DCL PRX
This might be the one, it is said. This might be the one which shat in mothers
drying mustard leaves, Im telling you, this one!
Demonstratives
can make up a complete NP on their own, and this occurs frequently in
Atong. This is the so called demonstrative pronoun function (see
Dixon 2003:65).
can modify a noun, e.g. (269), (118), (119), (120), (121).
cannot co-occur with a personal pronoun, except when the personal
pronoun possessively modifies another noun (117).

(117) ie a mola hongabaaw roancoy
[[ie a mola] {hon} =gaba] =aw {ro -an -coy}
PRX 1s tobacco give =ATTR =ACC drink -REF -TRY
Try to smoke this my tobacco which I give.

Other phrasal properties of demonstratives are that they
cannot modify a verb,
cannot be modified,
have a tendency to attract the case and other marking, such as topic
marking, on the noun phrase away from the head (118), (119), or to be
marked in addition to the head, e.g. (120), (121). This tendency is
caused by the inherent topicality and referentiality of demonstratives.

There are two ways in which a demonstrative can modify a noun. The first is
anaphoric modification, as illustrated in (118), (119) and (120). In these examples the
referent has been mentioned before. The second way is deictic modification as
illustrated by (121).
8 DEMONSTRATIVES

148
(118) umido uaw kamal sandini.
umido [u =aw kamal] {sandi -ni}
then DST=ACC priest search -FUT
Then, [they] will search the priest.

(119) otokoymo na'dorado uaw rukpek bisi roaymu gumukan thoytokoknowa.
otokoymo [na'] =dora =do [u =aw rupek bisi] {ro} =ay =mu
so.then fish =p =TOP DST=ACC frog poison drink =ADV =SEQ
[gumuk] =an {thoy -thok -ok} =no -wa
all =FC/ID die -ALL -COS =QUOT -FACT
So then, the fish, having drunk that frog poison, all died, it is said.

In the following examples we see how the demonstratives attract the topic and the
accusative enclitics which are repeated in the string of phrasal enclitics. The heads of
the NPs are in order of appearance: jow' mother, di' shit, para goy' sa (reed
CLF:RESIDU one) one culm of reed.

(120) udo jow'gabado uaw di'awba asetca, otokoyan rotaymuna hon'arokno.
[u =do jow'] =gaba =do [u =aw di'] =aw =ba {aset -ca}
DST=TOP mother =DREL =TOP DST=ACC shit =ACC =EMPH throw.away -NEG
[otokoy] =an {rat} =ay =muna {hon' -arok} =no
like.that =FC/ID collect =ADV =SEQ give -PROG =QUOT
That mother does not throw away that shit, having collected [the mustard
leaves, she] is just giving it, it is said.

(121) de na'a re'earocido iaw para goy'saaw kawancoy.
de [na'a] {re'e -aro} =ci =do
well.then 2s go.away -PROG =LOC =TOP
[i =aw para goy' sa] =aw {kaw -an -coy}
PRX =ACC reed CLF:RESIDUE one =ACC shoot -REF -try
Well then, if you are going anyway, try to shoot this one culm of reed.

8 DEMONSTRATIVES

149
A demonstrative can also modify a proper name as the next example illustrates.

(122) ie radiba ato taksoraok ie?
[ie radi] =ba [ato] {tak -sora -ok} [ie]
PRX Pname =EMPH what do -TOTALLY -COS PRX
What the heck is that Radi doing, that one?
8.5 Morophological properties
Demonstratives
have a free form and a bound form,
can take case marking,
can be pluralised,
cannot be counted,
cannot be possessed.
The demonstratives occur in their free form when they are not followed by phrasal
enclitics or predicate suffixes and before the question marker <ma> (Q), e.g. ie ma ie?
(PRX Q PRX) This one or this one?
8.6 Other functions of the demonstratives
The locative-marked distal demonstrative u=ci (DST=LOC), but not the proximal
one, can be used for temporal deixis meaning then or for non-temporal deixis
meaning in that case and are then analysed as having grammaticalised into discourse
connectives, described in Chapter 1. Both demonstratives, but most frequently the
distal one, can function as sentence initial adverbials anaphorically referring to a
proposition or string of propositions. In this function the demonstratives are locative
or genitive marked.
The following example illustrates the use of the distal demonstrative with the
reason postposition gomon reason which governs the genitive. In that example the
demonstrative refers to the whole story that precedes.

8 DEMONSTRATIVES

150
(123) umi gomonsa ie hapawe badri rongdoha'way noanowa, aro raawba mokha
badri mowanowa.
[u =mi gomon] =sa [ie hap] =aw =e [badri rodo ha'way]
DST=GEN reason =DLIM PRX place =ACC =FC Pname
{no -a} =noa aro [ra] =aw =ba [mokha badri]
say -CUST =QUOT and rain =ACC =ADD long.heavy.rain
{mo -wa} =noa
call.a.name -FACT =QUOT
Thats why this place is called Badri Rongdyng Hawai, it is said, and the rain
is also called mykha badri, it is said.

A demonstrative can refer to a third person. When a demonstrative refers to a third
person it can take the highly selective personal pronoun plural suffix <-tom> (ppp).
When the demonstrative refers to a third person, the choice between the proximal and
the distal demonstrative depends on the involvement of this third person in the
conversation. If the referent is more involved, the proximal demonstrative will be
used, whereas if the referent is less involved the distal demonstrative will be used.
A good example of this parameter can be found in TEXT 1, line 16, represented
below as (124). In the video of which that text is a transcription, the speaker directs
his attention to the camera, seeking the attention of whoever will be the future viewer,
and begins to speak about this person. It is obvious that in that instance the person
talked about is greatly involved, hence the use of the proximal demonstrative.

(124) Songken (speaking into the camera)
ie, ie, i =do mamu =an do' -khu -ca.
PRX PRX PRX =TOP nothing =FC/ID IE.be -INCOM -NEG
He, he, he is nothing yet.
8.7 The adverbial demonstrative tky
The adverbial manner demonstrative otokey can refer to objects one like this/that, e.g.
(41), and can be used adverbially doing like this/that. This demonstrative can be
used anaphorically, e.g. (125), cataphorically, as in (126) and deictically. When used
deictically it is usually accompanied by gestures, as in (127).

8 DEMONSTRATIVES

151
(125) atakna jalwa bai'sigadora? ayu! otokoy otokoy do'wacom takrukna san
somay thik kha'wachom.
atak =na {jal -wa} [bay'siga] =dora
do.what =DAT run.away -FACT friend =p
[ayu] [tky] [tky] {do' -wa} =com
interj like.this like.this IE.be -FACT =IRR
[san somay] {thik kha' -wa} =com
day time precisely do -FACT =IRR
Why are you running away, friends? Oh! This [and] this supposedly
happened. [They] supposedly fixed a day [and] a time [to fight with each
other].

(126) otokoymu de'thenge otokoy takokno. mo'sa morot man'ay sa'gabaci wak
rakhina ga'akoknoaro.
otokoymu [de'the] =e [tky] {tak -ok] =no |[mo' sa
so.then 3s =TOP like.this do -COS =QUOT one -CLF:HUMANS
morot {man'} =ay {sa'} =gaba] =ci [wak] {rakhi}| =na
-person in.great.amounts =ADV eat =ATTR =LOC pig look.after =DAT
{ga'ak -ok} =no
be.compeled -COS =QUOT
So then he did like this, it is said. He was forced to look after the pigs at [the
house of] a rich man, it is said.

(127) ge'the dokomaw otokoy [gesture] ca'kuisang tan'thongok..
[ge'the] [dokom ]=aw [tky] [ca'kui] =sa {tan'tho -ok}
3s head =ACC like.this big.knife =INSTR decapitate -COS
He cut the head of with a big knife like this (gesture).

The adverbial demonstrative is not to be confused with the homophonous
demonstrative verb otok- to do like this/that (see section 4.5.1). The non-finite
sequential form of this verb has developed into a discourse connective, viz. otokoymu
~ otokoymo ~ otokoymu ~ otokoymuna indicating the occurrence of a new event in
discourse, e.g. (126). In this connective form the adverbial clausal enclitic <=ay>
(ADV) underwent vowel harmony giving /ey/ and while in most allomorphs the
sequential enclitic <=mu ~ =mo ~ =mu> (SEQ) is still recognisable, another
allomorph, <=muna> has evolved just for this connective. It is usually translated in
this grammar as so then, though the literal translation would be having done like
this/that.
8 DEMONSTRATIVES

152
8.8 Deictic-only demonstratives
Deictic-only demonstratives are a closed word class consisting of two members, viz.
hawe ~ haw remote and non-visual demonstrative (REM) (third deictic degree),
hoyawe ~ hoyaw emphatic remote demonstrative (REMEMPH) (third deictic degree).
The properties of the quasi demonstratives are given here below.
Semantic properties: Deictic-only demonstratives cannot be used as third person
personal pronouns.
Discourse properties: Deictic-only demonstratives can be used deictically but cannot
be used anaphorically,
Clausal properties: Deictic-only demonstratives cannot be the head of a predicate,
but can be an oblique argument (128).

(128) te'do hawci ciakol ruguci jowsawaydonote, magacakdo.
[te'] =do [haw] =ci [ciakol (Garo) rugu] =ci
now =TOP REM =LOC well edge =LOC
{jow -saw -aydo} =no =te [magacak] =do
sleep -SURELY -PROG =QUOT =DCL deer =TOP
Now [he] was fast asleep way over there next to a well, it is said, the deer.

Phrasal properties
Deictic-only demonstratives
cannot modify a noun,
cannot modify a verb,
cannot be modified,
in isolation can be used deictically to refer to objects (129).

(129) uci khambayci nok gana, hoyawe.
[u] =ci [khambay] =ci [nok] {gana} [hyawe]
DST=LOC top =LOC house exist REMEMPH
There on top is a house, that one way over there.

8 DEMONSTRATIVES

153
Morphological properties
Deictic-only demonstratives
can take case marking, except accusative,
cannot be counted,
cannot be possessed.

154
Chapter 9 Interrogatives
_____________________________________________________________________

The sixteen members of the closed word class of interrogatives are listed below in
Table 38.

Table 38 List of interrogatives
section MORPHEMES FORM GLOSS
9.2 opaque ca who?
9.3 opaque ato what?
9.4 what=VIA/LIKE ato=tokoy why? how?
9.5 do.what=DAT ~ opaque atak=na ~atana why? for what purpose?
9.6 what-? ato-may'na why? for what purpose?
9.7 do.what=ADV ~ opaque atakay ~ atokoy how?
9.8 QF unbound form bie ~ bi- which? where?
9.9 opaque biskon, boysok how much, how many?
9.10 QF=EMPH bi=ba
when? / in whatever
place
9.11 QF=VIA bi-tokoy by which way?
9.12 QF=LOC bi=ci where?
9.13
QF=MOB bi=sa to/from where?
QF=MOB=GEN
bi=sa=mi ~
bi=sa=mo
from where?
9.14 QF=GEN bi-mi ~ bi=mo from where?
9.15 QF=ATTR bi=gaba which?

Roughly half of the interrogatives are formed with the interrogative formative
morpheme <bi> (QF), which in its unbound form bie means which or where. The
others, except ca who, are formed with what might be a fossilised prefix <a-> (?),
possibly a remnant of the Proto-Tibeto-Burman prefix *'a- ~ *(')o- ~ 'o- ~ *'a- ~
*'ak- which had quite a number of distinct functions, among which indicator of
stativity or intransitivity which might be relevant here (see Matisoff 2003: 104-107).
The interrogative verb atak- to do what seems to be derived from the verb tak- to
do with this a- prefix.
The interrogative form /biskon/ [bis.ken] how much/many has an affirmative
counterpart /isokon ~ iskon/ this many/much. The morpheme <soken> never appears
anywhere else in the language except in the indefinite proform jesokon however
much/many. Therefore, although isokon ~ iskon this many/much, biskon how
9 INTERROGATIVES

155
much/many and jesokon however much/.many may be historically analysable as
containing a suffix <-sokon> (QUANTITY), these forms are now opaque for Atong
speakers.
The question word atakay ~ atokoy why has developed from the adverbial form
of the verb atak to do what?: atak=ay (do.what=ADV) doing something. The form
atokoy is a more grammaticalised, more opaque form with the vowels reduced to
schwa. About the origin of the element may'na (?) in ato-may'na I can only
speculate that the element may' might be historically related to the Garo question
word may what?(without glottal stop). Of course this is all highly speculative.
9.1 Properties of interrogatives
Interrogatives cannot express any grammatical categories expressed by predicate
heads (see Chapter 22), except for biskon how much/many and bisa to where,
which can take the change of state suffix <-ok> (COS) and thus function as head of a
predicate of a verbless content question interrogative clause. Interrogatives cannot be
modified. All the interrogatives will be treated one by one.
The position of the interrogative in the clause can vary just like the position of
other core and peripheral arguments. However, unlike NPs, interrogatives never
appear in right dislocated position after the predicate, except for one example with the
interrogative biskon how much/many?, presented below in (130).

(130) ma', cua biskonan?
[ma'] {cu -a} [biskon] =an
interj big -DCL how.much =FC/ID
How big [did you say it was]?

The following interrogatives can occur as constituents of predicateless
interrogative clauses. Predicateless interrogative clauses are treated in 26.1.2.

ca who?
ato what?
bie which?,where?
bici where?
atakna ~ atana why?

9 INTERROGATIVES

156
The interrogatives biskon how much/many and bisa to where can express
perfectivity by means of the change of state suffix <-ok> (COS) and can even take
event specifiers. Therefore these interrogatives can always be identified as predicate
head when they appear in verbless interrogative clauses. Examples of interrogatives as
head of a predicate can be found in 26.1.3. More fieldwork is required to find out if
there are restrictions on the types of event specifier that can appear on these
interrogatives.
9.2 ca who
The interrogative ca who has human reference and can be used as an argument,
core (131), (132), or oblique (peripheral), e.g. (132), and as a adnominal modifier with
and without the genitive enclitic, e.g. (1), (134). In (131) we see the interrogative in S
function and in (132) in O function and marked with the accusative enclitic. In
example (132) ca who? functions as peripheral argument.

(131) ca ray'awa?
[ca]
S
{ray'a -wa}
who come -FACT
Who has come?

(132) camu re'eni? aga caaw morot baju man'phanaka?
[ca] =mu {re'e -ni}
who =COM go.away -FUT
[aa] [ca]=aw [morot baju] {man' -pha -naka}
1s who =ACC human friend get -IN.ADDITION -IFT
With whom shall we go? Who will I get as human friend? (Said the lazy
king after the tiger had defied him.)

The following examples illustrate the use of the interrogative ca who? as an
adnominal modifier within an NP. In (1) the interrogative is genitive-marked while in
(134) it is unmarked for case and modifies the noun though juxtaposition.

9 INTERROGATIVES

157

(133) camo git?
[ca =mo git]
who =GEN song
Whose song?
(134) na'a ca sa'?
[na'a] [ca sa']
2s who child
Whose child are you?

Example (1) is the reaction to something that somebody had said about a certain song.
The example is a fragment (see also 20.8.3), i.e. it is not a clause and it does not
mean Whose song is it?.
9.3 ato what
The interrogative ato what has non-human reference. It can replace the interrogated
NP in a clause, and it can be used as an adnominal modifier within an NP. In the next
example the interrogative is E argument and in (136) O.

(135) ue usami? bimu ato mowa?
[ue] [u] =sa =mi
DST DST=MOB =GEN
[bimu]
O
[ato]
E
{mo -wa}
name what call.a.name -FACT
Where is he from? What is [his] name?

(136) otokoimo thomay caybutucie atoaw nukokno ge'thee?
otokoymo {thom} =ay {cay -butu} =ci =e
CONJ lay.in.ambush =ADV look -WHILE =LOC =FC
[ato]
O
=aw {nuk -ok} =no [ge'the] =e
what =ACC see -COS =QUOT 3s =FC
So then, while he was looking while laying in ambush, what did he see?, it is
said.

This interrogative can also be used attributively to nouns as shown in (137), where it
modifies the noun kam work.

9 INTERROGATIVES

158
(137) kamba ato kamaw kha'ay mu'naka ie?
[kam] =ba [ato kam] =aw {kha'} =ay {mu' -naka} [ie]
work =EMPH/ADD what work =ACC do =ADV stay -IFT PRX
And then that work, what work will [he] do when [he] will stays here, this
guy? Lit. What work doing will he stay?
9.4 atotky why, how come
This interrogative questions which event has taken place for a situation to be the way
it is. Therefore atotokoy is translatable as why or how come?, e.g. (138). The
word is transparently made up of the interrogative ato and the perlative/similative
case enclitic <=tokoy> (VIA/LIKE).

(138) atotokoy tay'ni ja'bek thawoksoy? nookno. Ato dowwa ama? nookno.
[atotokoy] [tay'ni] [ja'bek] {thaw -ok] =soy {no -ok} =no
why today curry tasty -COS =MIR say -COS =QUOT
[ato] {dow -wa} [ama] {no -ok} =no
what add -FACT mother say -COS =QUOT
How come, to my surprise, the curry is so very tasty today? [he] said, it is
said. What did you add, mother? [he] said, it is said.
9.5 atakna ~ atana why
This interrogative questions a purpose or reason (139), (140). The allomorph atakna is
transparently made up of the root of the interrogative verb atak to do what? (see
4.5.1vii) and the dative enclitic <=na> (DAT). The more opaque allomorph atana
appears most often in quick speech.

(139) na'a atakna icina rat'awa?
[na'a] [atakna] [i] =ci =na {ray'a -wa}
2s why PRX =LOC =ALL come -FACT
Why have you come all the way here?

(140) atakna korewa, morotma'dorangna?
[atakna] {kore -wa} [morot] =ma' =dora =na
why fear -FACT human =interj =p =DAT
Why do you fear the humans, hey?

The interjection ma' in (140) signals surprise (see also (285) in 16.1.5).
9 INTERROGATIVES

159
9.6 atomayna why
This interrogative can occur anywhere in the clause before the predicate. It questions
a reason.
(141) madam tokwa. atongmay'na? otokoyan tokariwa.
[madam] {tok -wa} [atomay'na] [otokoy] =an {tok -ari -wa}
female.teacher beat -FACT why like.that =FC/ID beat -SIMP -FACT
The teacher beat [me]. Why? [She] beat [me] just like that (i.e. for no
reason).
9.7 atakay ~ atky how
This interrogative can appear anywhere in the clause before the predicate. It questions
a method. The allomorph atakay is morphologically transparent and consists of the
verbal root atak to do what? and the adverbial enclitic <=ay> (ADV). The
allomorph atokoy is less transparent since it has reduced most of the vowels into
schwa. There is no verb *atok. Examples with both allomorphs are given below.

(142) atakay koreca na'a, ado korea!
[atakay] {kore -ca} [na'a] [a] =do {kore -a}
why be.afraid -NEG 2s 1s =TOP be.afraid -CUST
Why are you not afraid? I am afraid.

(143) e alsia raja atokoy keaydok? atokoyan jokaw haldunna man'aydok?
[ie alsia raja] [atokoy] {ke -aydok} [atokoy] =an [jok =aw
PRX lazy.person king how live -PROG how =FC/ID spouse =ACC
{haldun} =na {man' -aydok}
feed =DAT be.able -PROG
How does this lazy king live? How is he able to feed his wives?
9.8 bie ~ bi which, where
This interrogative questions both a place, as in (144), and one item out of a collection.
In the latter function the interrogative can function as modifier of a noun, as in (145),
or as a constituent on its own, as we see in (146). This interrogative has a free and a
bound allomorph. The free form bie is used without phrasal enclitics. When it takes a
phrasal enclitic, the bound form bi is used (147).

9 INTERROGATIVES

160
(144) bie na' jodorae? na' joe bie?
[bie] [na' jo] =dora =e
where 2s younger.brother =p =FC
[na' jo] =e [bie]
2s younger.brother =FC where
Where [are] your younger brothers? As for your younger brothers, where [are
they]

(145) sam manama. bie same?
[sam] {manam -a} [bie sam] =e
medicine stink -CUST which medicine =FC
The medicine stinks. Which medicine?

(146) otokoymu biaw mokcana otokgaraawe?
otokoymu [bi] =aw {mokca} =na {otok} =ga =ra =aw =e
CONJ which =ACC fancy -DAT do.like.that=ATTR =p =ACC =FC
So which one(s) am [I] supposed to fancy, those who do like that?

The context of the next example is as follows. A man is talking to his daughter about
the fish traps he had put up. He says: When I inspected the one upstream, otters had
eaten the fish. When I inspected the one downstream, otters had also eaten the fish.
Then the man utters (147). This example illustrates the bound form of the
interrogative bie ~ bi which?, where?

(147) biaw caykhuna? ana ni'ok nookno.
[bi] =aw {cay -khu -na} [a =na] {ni' -ok} {no -ok} =no
which =ACC look -INCOM -DESI 1s DAT not.exist -COS say -COS =QUOT
Which other one [is there] to look at? I have no more, he said, it is said.
Alternatively: Which other one can I/am I supposed to look at?
9.9 biskn and bysk how much/many
The interrogative biskon how much/many questions a quantity. This can be done as
a phrase, e.g. (130), (149) as an adnominal modifier, e.g. (148), or as a predicate
(150). This interrogative can modify countable as well as uncountable nouns.

9 INTERROGATIVES

161
(148) udo biskon somay do'okte?
[u] =do [biskon somay] {do' -ok} =te
DST=TOP how.much time IE.be -COS =DCL
As for him, how much time has it been?

(149) biskon ra'wa?
[biskon]
O
{ra' -wa}
how.much buy -FACT
For how much did [you] buy [it]?

(150) ie gari biskon?
[ie gari]
S
{biskon}
PRX vehicle how.much
How much is this vehicle? i.e. whats the price?

We know this interrogative, when unmarked, can function as a predicate because it
can take the change of state predicate head suffix <-ok> (COS), as we see in (151).
The headless quantified NP is a right dislocated, antitopical Beneficiary.

(151) biskonok, mo'thamna?
{biskon -ok} [mo' tham] =na
how.much -COS CLF:HUMANS three =DAT
How much is it in total, for three persons?

This interrogative has another form boysok how much/many used after classifiers
and auto-classifiers (see Chapter 12). Example (152) is illustrative. In this example we
see that the interrogative boysok modifies a headless quantified NP with the residual
classifier goy'.

9 INTERROGATIVES

162
(152) na'tome goy'boysok man'phawa ie bolsi? noay so'rukthoka.
[na' -tom] =e [goy' bysk] {man' -pha -wa}
2s -ppp FC CLF:RESIDUAL how.many get -IN.TOTAL -FACT
[ie boylsi] {no} =ay {so' -ruk -thok -a}
PRX year say =ADV ask -RC -ALL -CUST
How many [baskets full of rice] did you get in total this year? [they] all ask
each other. (Lit. [they] sayingly ask each other.)
9.10 biba when, in whatever place
The interrogative biba when, in whatever place has both temporal and spatial
reference. When it is used with spatial reference it has an indefinite meaning and is
used as a pre-nominal modifier, e.g. (153). When it is used to question a temporal
constituent, biba when is used adverbially and its position in the clause is variable
(154).

(153) te'ewe biba so damsacie boba mo'sa ganacomnoro.
[te'ew] =e [biba so dam sa] =ci [boba
now =FC wherever village CLF:VILLAGES one =LOC crazy.person
mo' sa] {gana] =com =no =ro
CLF:HUMANS one exist =IRR =QUOT =EMPH
Now, in a certain village wherever, supposedly was a crazy person, it is said.

(154) ie radi bibaan ray'anaka?
[ie radi] [biba] =an {ray'a -naka}
PRX Pname when =FC/ID come -IFT
When precisely will Radi come?
9.11 bitky by which way?
This interrogative, consisting of the question formative bi (QF) and the
perlative/similative case enclitic <=tokoy> (VIA/LIKE), questions a Pathway. The
following example is illustrative. Although the case marker <=tokoy> (VIA/LIKE)
also marks facsimile adjuncts (peripheral arguments/obliques), the interrogative
bitokoy by which way? is only attested questioning Pathways.

9 INTERROGATIVES

163
(155) bitokoy re'enima? Ie ramtokoyma utokoy?
[bitokoy] {re'e -ni} =ma [ie ram =tokoy =ma] [u] =tokoy
by.which.way go.away -FUT =Q PRX road =VIA =Q DST=VIA
By which way shall we go? By this road or by that one?
9.12 bici where
The interrogative bi=ci (QF=LOC) only questions spatial location. Its place in the
clause is right before the predicate (156). In a verbless clause the position of the
interrogative is variable, e.g. ue bici? (DST where) or bici ue? (where DST) Where is
he?.

(156) ayaw! cabi bicin tanaok?
ayaw [cabi] [bici] =n {tan -a -ok}
interj key where =FC/ID put -AWAY -COS
Oooh! Where did I put my key away (so that I cannot find it any more)?(In
Dutch this could be accurately translated by Waar heb ik mijn sleutel toch
weggelegd?)
9.13 bisa to/from where and bisami from where
These interrogatives question a direction. Their position in the clause is right before
the predicate, e.g. (157).

(157) ha? ato cucu? na'a bisa re'eaydoa?
ha [ato] [cucu] [na'a] [bisa] {re'e -aydoa}
interj:SURPRISE what grandson 2s to/from.where go -PROG
Huh? What is it, Grandson? Where are you going?

The interrogative bi=sa (QF=MOB) to/from where might well be the most
frequently used one in the language since it is used in the common greeting bisa
re'ewa? meaning where have you come from? often just shortened to bisa?
where to /where from?. What direction is implied depends on the verb. The verb
re'e- to go.away, leave in the factitive implies movement away from somewhere
with past time reference and therefore the interpretation of bisa will be from where,
asking for a source. In (157) we see the same verb in the progressive, in which case
the bisa is interpreted as questioning a goal. When no verb is used, it depends
9 INTERROGATIVES

164
entirely on the context or on the choice of the listener how to interpret the question. If
the speaker wants to make it absolutely clear that he is asking for a source, he will use
the interrogative with the ablative/genitive case enclitic <=mi ~ =mo> (GEN/ABL),
e.g. (158).

(158) na' bisamo ray'a -wa
[na'] [bisa] =m {ray'a -wa}
2s to/from.where =ABL come -FACT
Where have you come from?

There are two recorded instances, both in the same story, of bisa to/from where?
with a dative case added onto it. This will emphasise that the speaker questions a Goal
rather than a Source. These are also the two recorded occasions on which this question
word was used with the delimitative enclitic. Examples are given below.

(159) bisanasa na'tyme?
[bi] =sa =na =sa [na' -tom] =e
QF =MOB =DAT =DLIM 2s -ppp =FC
To where exactly [are] you [going]?

(160) aya! phulis bisanasa ray'arumasoy tay'nido?
aya [phulis] [bi] =sa =na =sa
interj police QF =MOB =DAT =DILM
{ray'a -ram -a} =soy [tay'ni] =do
go -SEARCH -CUST =MIR today =TOP
Huh?! Where exactly are the police to my surprise trying to go to today?
9.14 bimi ~ bim (from) where
Another way of questioning a place or a source is by means of the interrogative bimi ~
bimo from where, where, which can be analysed as the bound form of the
interrogative bie ~ bi which, where with the ablative case enclitic <=mi ~ =mo>
(ABL) (homophonous with the genitive). The interpretation of this interrogative
depends on the form of the verb in the same way as described above for bisa
to/from where. This interrogative appears right before the predicate (161).
9 INTERROGATIVES

165
(161) bimi akwa ue?
[bi] =mi {ak -wa} [ue]
where =ABL pluck -FACT DST
Where did you pluck them?

The interrogative bimi ~ mimo from where, where can be used as modifier of an NP
postponed to the interrogative, of which (162) below is illustrative. This example
comes from a story about a man who is selling the ashes of his burnt house. When
people see him selling these ashes on the market they ask (162).

(162) iawe bimo morot ra'naka?
[i] =aw =e [bi] =mo morot] {ra' -naka} =ma
PRX =ACC =FC QF =ABL person get -IFT =Q
A person from where will buy this?
9.15 biga ~ bigaba which
This interrogative functions as modifier to a noun which is the head of an NP. It can
attract the case marking away from the head or the last element in the noun phrase just
as demonstratives are likely to do. Example (163) is illustrative.

(163) bigaaw biskut ra'nima?
[biga =aw biskut] {ra' -ni -ma}
which =ACC biscuit buy -FUT -Q
Which biscuits shall I buy?

The interrogative biga ~ bigaba which consists of two morphemes, viz. the question
formative bi (QF) and the attributive suffix <-gaba ~ -ga> (ATTR) (see 29.12). The
function of the morpheme <gaba ~ ga> is extensively treated in Chapter 29.

166
Chapter 10 Indefinite proforms
_____________________________________________________________________

There are eleven indefinite proforms in Atong. They have different syntactic and
morphological properties and therefore belong to different word classes. All proforms
are listed in Table 39 and will be treated separately below.

Table 39 List of indefinite proforms
section PROFORM LABEL OF
PARTS
GLOSS
10.1 je any, whichever, whatever
je=mi ~ je=mo any=GEN any followed by a time noun in
the locative
10.2 je=sa=ba any=MOB=INDEF to wherever
je=ci=ba any=LOC=INDEF anywhere, wherever
je-sokon any-QUANTITY however much/many
10.3 ca=ba who=INDEF someone
ato=ba what=INDEF something
bi=ci=ba QF=LOC=INDEF somewhere, sometimes
bi=sa=ba QF=MOB=INDEF to somewhere
bi-mi=ba ~
bi=mo=ba
QF=GEN=INDEF from somewhere
10.4 ca=gaba who=ATTR whoever
10.5 daraba opaque anybody
10.6 gumuk=sa all=MOB everywhere
10.1 The indefinite proform je any, whichever, whatever
The indefinite proform je any, whichever, whatever
25
modifies any postposed noun
in any syntactic function, e.g. (164). It has a genitive-marked derived form je=mi ~
je=mo (any=GEN) with a more restricted use. This derived form only appears before
locative-marked nouns indicating a unit of time and the word somay time, e.g. (165).

25
This word may have an Indic origin: cf the Hindi relative and indefinite pronoun /je ~ jo/ the
one who, which; whichever, what ever, whoever.
10 INDEFINITE PROFORMS

167
As we can see in (166) the other attested form modifying locative time nouns is the
focus/identifier-marked form.

(164) ha'cok sogumukdo mokha badri noaria je raawba.
[ha'cok so] =gumuk =do [mokha Badri] {no -ari -a}
Garo village =ALL =TOP long.havy.rain say -SIMP -CUST
[je ra] =aw =ba
any rain =ACC =EMPH
Really all the Garo villages just say mokha Badri to all rain.

(165) jemi sanci dibakhongdaaw matsa kakok.
[je =mi san] =ci [dibakhoda] =aw [matsa] {kak -ok}
any =GEN day =LOC Pname =ACC tiger bite -COS
On a certain day a tiger bit Dibangkongdang .

The only other enclitic that je has been recorded with is the focus identifier enclitic
<=an> (FC/ID), which frequently assimilates its vowel to the indefinite article. The
resultant forms are jeen [jecn] ~ jen [jcn], but the form jean also occurs, e.g. (166).

(166) jean sanci jada mo'sa nukokno.
[je =an san] =ci [jada mo' sa] {nuk -ok} =no
any =FC/ID day =LOC idiot CLF:HUMANS one see -COS =QUOT
On a certain day [he] saw an idiot, it is said.

There are both other proforms and adverbs derived opaquely and transparently from
the morpheme je any, whichever, whatever. The derived adverbs are listed in Table
49 below. The derived proforms are mobilitative, locational and quantificational, and
will be discussed separately below.
10.2 Derivations from je any, whichever, whatever
The indefinite proforms je=sa=ba (any=MOB=INDEF) to wherever replaces a
Direction adjunct, e.g. example (167), which comes from Text 3, line 26.


168
(167) kamalnado jesaba walduk sandukba rokarini, khurutna.
[kamal] =na =do [je] =sa =ba [wal -duk san -duk] =ba
priest =DAT =TOP any =MOB =INDEF night -sorrow day -sorrow =ADD
{rok -ari -ni} {khurut} =na
chase -SIMP -FUT perform.an.incantation =DAT
[People] will search anywhere for a priest, whether it is day or night, to
perform an incantation.

The indefinite locational proforms replaces a Location adjunct, e.g. (168).

(168) jeciba na' mu'ciba a na'na kha'gala.
[je] =ci =ba [na'] {mu' }=ci =ba [a] [na'] =na {kha'gal -a}
any =LOC =INDEF 2s stay =LOC =INDEF 1s 2s =DAT love -CUST
Wherever you are, I love you.

The indefinite quantificational proform jesokon however much/many replaces a
quantity, as in (169). It is not attested with the indefinite enclitic <=ba> (INDEF).
This proform is one of three lexical items that contain the bound morpheme <sokon>
(QUANTITY), the others being i-sokon (PRX-QUANTITY) and biskon
(QF.QUANTITY) how much/many, which is syllabified as [bis.ken].

(169) jesokon na'ci gana conaribo, kamalna.
[je -sokon] [na'] =ci {gana} {con -ari} =bo [kamal] =na
any -QUANTITY 2s =LOC exist offer -SIMP =IMP priest =DAT
However much you have, just offer it to the priest.
10.3 caba, atoba, biciba, bisaba and bimiba
These indefinite proforms are all derived from their respective interrogatives (see
Chapter 9) by means of the indefinite phrasal enclitic <=ba> (INDEF). This means
that Atong has polysemy of indefinites and interrogatives. ca=ba (who=INDEF)
someone has human reference, atoba (what=INDEF) something has non-human
reference, bi=ci=ba (QF=LOC=INDEF) somewhere refers to a Location,
bi=sa=ba (QF=MOB=INDEF) refers to a Direction and bi-mi=ba ~ bi=mo=ba
(QF=GEN
=INDEF) refers to a Source. The proforms caba someone and atoba something

169
can fulfil all argument and adjunct functions and can take case marking. Examples of
all these respective proforms are given here below.

(170) cangba ge'the songmi bayaw badayok.
[ca] =ba [ge'the so =mi bay] =aw {baday -ok}
who =INDEF 3s village =GEN border =ACC cross.a.border -COS
Somebody crossed the border of his village.

(171) na'do tay'nido atoba dowwa. ja'bekan thawokte.
[na'] =do [tay'ni] =do [ato] =ba {dow -wa}
2s =TOP today =TOP what =INDEF add -FACT
[ja'bek] =an] {thaw -ok} =te
curry =FC/ID tasty -COS =DCL
You have added something today. This curry is very tasty, really!

(172) umigomon biciba gisep gisep coti sayietrukarinaka.
[u =mi gomon] [bi] =ci =ba [gisep gisep
26
] [coti]
DST=GEN reason QF =LOC =INDEF middle RED letter
{say -et -ruk -ari -naka}
write -CAUS -RC -SIMP -IFT
Because of that, we will sometimes write each other letters from time to
time/in the mean time.

(173) ge'the bisaba re'eok.
[ge'the] [bi] =sa =ba {re'e -ok}
3s QF =MOB =INDEF go.away -COS
He has gone somewhere.

26
The morpheme gysep ~ gisep ~ gesep means middle or space and is, apart from reduplicated, only
found in the body part noun caksi-gysep space in between fingers and in the expression u=mi
gysep=ci=an (DST=GEN middle=LOC=FC/ID) in the meantime.

170
(174) phorendora bimiba indiami no'dorami raya'aymu []
[phoren] =dora [bi] =mi =ba [india =mi no'] =dora =mi
foreigner =p QF =GEN =INDEF Pname =GEN inside =p =GEN
{ray'a} =ay =mu
come =ADV =SEQ
Foreigners come from the interior [places] of India from somewhere [and ]

The pro-form bisaba to/from somewhere has been recorded as modifier to an NP.
Moreover, this indefinite proform is inflected with the genitive case, marking the
proform as a Source.

(175) raaria, phorensamo, bisabamo morotdara.
{ra -ari -a} [phoren] =sa =mo
get -SIMP -CUST foreign.country =MOB =GEN
[bi] =sa =ba =mo morot] =dara
QF =MOB =INDEF =ABL person =p
[They] will just buy it, from foreign countries, people from somewhere.
10.4 cagaba whoever
The relational/derelational/attributive/adverbial morpheme <gaba ~ ga>
(REL/DREL/ATTR/ADV) has many functions (see 14.2 and Chapter 29, especially
29.12) some of which are more productive than others. However, none of these seem
to fit the lexeme cagaba whoever very well. Since the morpheme <gaba ~ ga> is
not productive as a suffix on indefinite proforms, I consider the lexeme under
discussion to be opaque. However, given the indefinite meaning of this proform, I
suggest that if one had to make a guess as to the segmentation of this word, it would
be ca-ga=ba (who-?=INDEF) with the indefinite enclitic <=ba> (INDEF) as last
element. Examples (176) and (177) below illustrate the use of this indefinite proform.
Example (176) comes from Text 3, line 21.


171
(176) cagaba man'ay sa'a cagaba nokda taka, umi bimogumukawan thalay
moaymusa []
[cagaba] {man'} =ay {sa' -a} [cagaba] [nokda]
whoever in.great.amounts =ADV eat -CUST whoever household
{tak -a} [u =mi bimo] =gumuk =aw =an]
do -CUST DST=GEN name =all =ACC =FC/ID
{thal} =ay {mo} =ay =mu =sa
complete =ADV call.a.name =ADV =SEQ =DLIM
Whoever is rich [lit. eats in great amounts] whoever has a family only after
having called clearly/explicitely upon all their names, []

(177) cagabaaw gawiaw komsomnaka?
[cagaba] =aw [gawi] =aw {kom -sym -naka}
whoever =ACC female =ACC marry -FOLLOW -IFT
Who will [he] remarry? or Who will [he] marry next? or Whatever woman
will [he] marry next? Literally: Whoever woman will [he] marry next?
10.5 daraba anybody
The proform daraba anybody has human reference. Although the form is opaque
for speakers of Atong, we can recognise the indefinite enclitic <=ba> (INDEF). The
first morpheme is homophonous with the nominal plural morpheme <=dara> (p),
and the two might actually be historically related. This proform was found in the
example given below.

(178) a soci daraba atokhu'cuk olna man'ca.
[a so] =ci [daraba] [ato khu'cuk] {ol} =na {man' -ca}
1s village =LOC anybody Atong language speak =DAT get -NEG
In my country [I] get nobody to talk Atong with.
10.6 gumuksa everywhere
The proform gumuk=sa is composed of the lexeme <gumuk> all, everything,
everybody and the mobilitative/locative case enclitic <=sa> (MOB/LOC) making
this a transparent lexeme meaning everywhere. This proform can refer to both a
Location (179) and a Direction (180) depending on the verb. It can take genitive case-
marking to explicitly refer to a Source. It is not attested with dative-marking for
explicit reference to Goals but this is possible. The only other enclitic attested on this

172
proform is the focus/identifier enclitic <=an> (FC/ID). The reason that this proform
refers to both a Direction and a Location is that the second syllable sa can be
interpreted as both the mobilitative case enclitic and as the old bound morpheme
meaning place or side still found in a few other words as well. When this proform
refers to a place, it is always marked with the focus/identifier enclitic <=an> (FC/ID).

(179) gumuksaan gana ukci.
[gumuk] =sa =an {gana} [ukci]
everywhere =place/LOC =FC/ID exist leech
They are everywhere, leeches.

(180) ge'the gumuksa re'eok.
[ge'the] [gumuk] =sa {re'e -ok}
3s all =MOB go.away -COS
He went everywhere.

173

174
Chapter 11 Numerals
_____________________________________________________________________

This chapter gives an overview of the numerals and classifiers in Atong and how they
function morphologically, syntactically and semantically. An Atong speaker
quantifies objects with numerals from four different languages, viz. Atong, Garo,
Hindi and English. Within the Atong language itself there are various ways of
counting, some more popular than others. Some numerals have allomorphs
participating in different paradigms. Moreover, speakers have a choice of using two
types of vigesimal systems and a decimal system for numbers higher than 39. It looks
like the vigesimal systems are on their way out at least in the areas where research has
been conducted so far, i.e. Badri and Siju.
The different types of Atong numerals are discussed in the first section. Borrowed
numerals are treated in section 11.2. The use of English and especially of Hindi
borrowed numerals is restricted in Atong. Section 11.3 answers the question of what
is quantified with which numerals. Classifiers and nouns often occur together. There
are situations in which classifiers can be omitted when numerals are used, viz. in
enumeration and when ordinal numerals precede the noun they modify. There are also
situations in which a classifier needs to be repeated. Section 11.4 treats the position of
the classifiers. The syntactic and morphological properties of Atong numerals are
treated in section 11.5. We will look at ordinal numerals in section 11.6, and finally,
the different functions and grammaticalisations of the number one will be treated in
section 11.7.
When a noun is used in combination with a numeral it is said to be quantified, or
more specifically, the NP is quantified. The term counting is defined as the use of
numerals in a sequence without referencing an entity, i.e. without quantifying an
entity. The terms enumerating and enumeration are defined as sequential
quantification, i.e. three houses, four houses, five houses and three, four, five
houses are acts of enumeration. Table 40 gives a paradigmatic overview of the
different ways of counting in Atong. Loans from English and Hind are listed in Table
43 and Table 44. For an overview of Garo numerals I refer the reader to Burling
(2004: 245-6).
11 NUMERALS

175
We need something to count! said Tononjyw [tontonjew?] and Tontonwa
[tontonwa?] two elderly people, the parents of Tonton, who agreed to count for me in
Atong so that I could record it. So they looked around and found a basket of garlic
cloves. They put them in a heap on the ground and started counting them moving
them from one heap to the other. When the source heap neared depletion the counted
cloves were quickly recycled by the spouse. The analysis in this chapter is based on
the data obtained by this and many other experiences with native speakers counting in
Atong.
11.1 Types of Atong numerals
This section discusses the original Atong numerals and the different paradigms in
which they appear and the use of numerals borrowed from Garo. An overview of
Atong numerals in their respective paradigms is presented in Table 40. Section 11.2
discusses the Hindi and English loans.

Table 40 Counting in Atong.
The analysis and translation of the individual components of the numerals
will be presented in Table 41 and Table 42. The morpheme ro
(CLF:ROUND.THINGS) is functioning as default classifier in this table and
will be commented upon in 11.1.1.
_____________________________________________________________________
UNIT NUMERALS

1 ro sa
2 ro ni
3 ro tham
4 boroy
5 baa
6 korok
7 sene
8 catgok
9 cokhow
10 coygok
11 cit sa
12 ci ni
13 ci tham
14 ci bori
15 ca raa ~ ci baa (calque on Garo)
16 ci dok
17 ci sene ~ ci soni
18 ci cat
19 ci sokhu
20 kholgok ~ kholgrok (Garo loan) ~ khol
11 NUMERALS

176
Paradigm 1 paradigm2 paradigm 3 paradigm 4
21 khole ro sa kholgok sa kholgrok sa
22 khole ro ni kholgok ni kholgrok ni
23 khole ro tham kholgok tham kholgrok tham
24 klole boroy khole ro bori kholgok borey kholgrok borey
25 khole baa khole ro baa kholgok baa kholgrok baa
26 khole korok khole ro korok kholgok korok kholgrok korok
27 khole sene khole ro sene kholgok sene kholgrok sene
28 khole catgok khole ro catgok kholgok catgok kholgrok catgok
29 khole cokhow khole ro cokhow kholgok cokhow kholgrok cokhow
Paradigm 5
30 khole coy khola ci
31 khole cit sa khola ci sa
32 khole ci ni khola ci ni
33 khole ci tham khola ci tham
34 khole ci bori khola ci byri
35 khole ca raa khola ci baa
36 khole ci dok khola ci dok
37 khole ci sene khola ci sene
38 khole ci cat khola ci cat
39 khole ci sokhu khola ci sokow
MULTIPLIED ROUND-NUMBER NUMERALS
Vigesimal 1 Decimal
40 rum' ni sot bori
41 rum' ni ro sa sot bori sa
42 rum' ni ro ni sot bori ni
43 rum' ni ro tham sot bori tham
44 rum' ni boroy sot bori boroy
45 rum' ni baa sot bori baa
46 rum' ni korok sot bori korok
47 rum' ni sene sot bori sene
48 rum' ni catgok sot bori catgok
49 rum' ni cokhow sot bori cokhow
50 rum' ni coygok sot boa
51 rum' ni cit sa sot boa sa
52 rum' ni ci ni sot boa ni
53 rum' ni ci tham sot boa tham
54 rum' ni ci bori sot boa boroy
55 rum' ni ca raa sot boa baa
56 rum' ci dok sot boa korok
57 rum' ci sene sot boa sene
58 rum' ni ci cat sot boa catgok
59 rum' ni ci sokhu sot boa cokhow
11 NUMERALS

177
Vigesimal 1 Decimal
60 rum' tham sotok ~ sotdok ~ sodok
61 rum' tham ro sa sotok sa
62 rum' tham ro ni sotok ni
63 rum' tham ro tham sotok tham
64 rum' tham boroy sotok boroy
65 rum' tham baa sotok baa
66 rum' tham korok sotok korok
67 rum' tham sene sotok sene
68 rum' tham catgok sotok catgok
69 rum' tham cokhow sotok cokhow
70 rum' tham coygok sot sene ~ sot soni
71 rum' tham cit sa sot sene sa
72 rum' tham ci ni sot sene ni
73 rum' tham ci tham sot sene tham
74 rum' tham ci bori sot sene boroy
75 rum' tham ca raa sot sene baa
76 rum' tham ci dok sot sene korok
77 rum' tham ci sene sot sene sene
78 rum' tham ci cat sot sene catgok
79 rum' tham ci sokhu sot sene cokhow
Vigesimal 1 Vigesimal 2 Decimal
80 rum' boroy kholca boroy sot cet
81 rum' boroy ro sa kholca bori ro sa sot cet sa
82 rum' boroy ro ni kholca bori ro ni sot cet ni
83 rum' boroy ro tham kholca bori ro tham sot cet tham
84 rum' boroy ro boroy kholca bori ro boroy sot cet boroy
85 rum' boroy baa kholca bori baa sot cet baa
86 rum' boroy korok kholca bori korok sot cet korok
87 rum' boroy sene kholca bori sene sot cet sene
88 rum' boroy catgok kholca bori catgok sot cet catgok
89 rum'boroy cokhow kholca bori cokhow sot cet cokhow
90 rum' boroy coygok kholca bori coygok sot sokhu
91 rum' boroy cit sa kholca bori cit sa sot sokhu sa
92 rum' boroy ci ni kholca bori ci ni sot sokhu ni
93 rum' boroy ci tham kholca bori ci tham sot sokhu tham
94 rum' boroy ci dok kholca bori ci bori sot sokhu boroy
95 rum' boroy ca raa kholca bori ca raa sot sokhu baa
96 rum' boroy ci dok kholca bori ci dok sot sokhu korok
97 rum' boroy ci sene kholca bori ci sene sot sokhu sene
98 rum' boroy ci cat kholca bori ci cat sot sokhu catgok
99 rum' boroy ci sokhu kholca bori ci sokhu sot sokhu coygok
11 NUMERALS

178
Paradigm 6 Paradigm 7
100 raja sa rum' baa
101 raja sa ro sa rum' baa ro sa
102 raja sa ro ni rum' baa ro ni
103 raja sa ro tham rum' baa ro tham
104 raja sa ro boroy rum' baa ro boroy
105 raja sa ro baa rum' baa ro baa
etc. etc.
115 raja sa ca raa ~ raja sa ci boa (calque on Garo, see Table 41)
etc.
200 raja ni
300 raja tham
etc.
1000 hajal ~ hajar sa (< Hindi (hazr) thousand)
1101 hajal sa raja sa ro sa
etc.
2000 hajal ni
etc.
7895 hajal sene raja catgok kholca boroy ca raa (vigesimal)~
hajal sene raja catgok sot sokhu baa (decimal)

For numbers higher than 99,999 the multiplier lak 100,000 has to be used.

250675 lak kolgok baa raja korok sot soni baa (decimal) ~
lak khole baa raja korok rum' tham baa (vigesimal)
_____________________________________________________________________

All numerals can be preceded by a classifier (see Chapter 12). We distinguish Round-
Number numerals and Unit numerals. These terms are defined on the basis of the
morphosyntactic behaviour and semantics of the numerals and not on the basis of their
everyday use in English. For the purpose of the following formulae Round-Number
numerals are indicated with the symbol R. A multiplied R is written as RX. The Unit
numerals are indicated by U. The following formulae define Unit and Round-Number
numerals distributionally and semantically.

1. A U can occur as a free form.
2. A U cannot be multiplied.
3. An R cannot occur without a following U or RX.

11 NUMERALS

179
The semantic relationship between different numerals can be additive or
multiplicative. Both relationships are obtained by simple juxtaposition and depend on
the three conditions below.

1. If a U is followed by another U they are in an additive relationship, e.g. ca raa
(10 [plus] 5) 15 and kholgok sa (20 [plus] 1) 21.
2. If an R is followed by only a U, R and U are in a multiplicative relationship with
each other, e.g. rum' tham-(TWENTY [times] 3) 60, sot cet (TEN [times] 8) 80,
raja sene (hundred [times] 7) 700.
3. If an RX is followed by another RX, they are in an additive relationship, e.g. hajal
tham raja ni sot cet tham (1000 [times] 3 [plus] 100 [times] 2 [plus] TEN [times] 8
[plus] 3) 3283.

Unit numerals will be treated further in 11.1.1 and Round-Number numerals in
11.1.2. Atong has one multiplier, viz. the Hindi loan lak (< (lakh). A multiplier
adds a numeral value to a number and cannot be preceded by a classifier. Some
speakers treat the archaic numeral rum' TWENTY as a multiplier as shall be
discussed below.
11.1.1 Unit numerals
Unit numerals are all the numerals from 1 to 39.
27
Table 40 gives an overview of the
different paradigms in which the morphemes occurring in Unit numerals occur. The
internal structure of compound Units is morphologically complex and irregular. The
morphemes that can be distinguished are listed in Table 41 with their occurrence
restrictions. The semantic relationship of Unit numerals in compounds is additive.

27
Note that the different morphemes denoting 10 and 20 in Table 41 are not Round-Number
numerals because they do not fit the definition of Round-Number numerals given above, but do fit the
definition of Unit numerals.
11 NUMERALS

180
Table 41 Morphemes participating in the formation of Unit numerals
SET FORM MEANING REMARKS
S
e
t

1

sa 1
Have to be preceded by a classifier, even when counting for the sake
of counting
ni 2
tham 3
S
e
t

2

boroy 4 Can occur as free form
baa 5
Can occur as free form or in a compound with ci 10 which will then
be a calque on Garo ci boa 15. What native speakers identify as the
real Atong form for fifteen is ca-raa (10 [plus] 5) 15.
28

korok 6 Can occur as free form.
sene 7 Can occur as free form or in a compound with ci 10
catgok 8
Can occur as free form. cokhow 9
coygok 10
S
e
t

3

kholgok 20 Is identified by native speakers as the real Atong form. Paradigm 3.
kholgrok 20
Is identified by native speakers as code switching to Garo. However,
this numeral has almost completely replaced the use of kholgok and
can be seen as an allomorph of kholgok. Paradigm 3
S
e
t

4

bori 4 Can only occur in a compound with ci 10 and sot TEN.
raa 5 Can only occur in a compound with ca 10 but not with sot TEN.
29

boa 5 Can only appear in a compound with sot TEN.
dok 6
Can only occur in a compound with ci 10 and sot TEN.
soni 7
cat 8
sokhu 9
S
e
t

5

khole 20
Occurs only in compound Unit numerals. Takes coy as morpheme for
10 to form numerals from 30-39. Paradigm 1
khola 20
Occurs only in compound Unit numerals. Takes ci as morpheme for
10 to form numerals from 30 to 39. Paradigm 2
coy 10 Occurs only in compound Unit numerals after khole 20. Paradigm 1
ci ~ cit ~
ca
10
Cannot occur on its own but have to be followed by a numeral of Set
1 or Set 3. The allomorph cit only occurs before sa one, the
allomorph ca occurs only before raa 5
S
e
t

6

khol 20
Cannot be multiplied, cannot be compounded but can only occur as
free form without classifier. Since khol 20 is not attested as a noun
outside the counting system it is considered a numeral used
exclusively in enumeration

28
Speakers volunteered this information only for this particular number.
29
Atong reflexts both prefixes allofams reconstructed by Benedict (1972:31, see also Matisoff, 2003:
129-30) for the numeral five at the Proto-Tibeto-Burman level, *l-a ~ b-a.
11 NUMERALS

181
When we examine Table 41, we can observe the following: All numerals from Set
2 can occur as free forms. Set 3 numerals can occur as free forms or in additive
relationship with numerals from Set 1 and 2 except coygok 10. The numerals from
Set 4 only occur in compounds with ci ~ cit ~ ca 10. Numerals from Set 5 cannot
occur on their own but have to be compounded to another, lower numeral. Set 6
consists only of the numeral khol 20. Since khol 20 is not attested as a noun
outside the counting system it is considered a numeral used exclusively in sequential
enumeration to indicate that a unit of 20 has been reached as in example (181). Note
that I analyse the word khol+ca (TWENTY+multiplied.by) TWENTY as one
Round-Number numeral morpheme (see the next paragraph).

(181) [] rum' tham ci sokhu, khol. kholca boroy. [] kholca boroy ci sokhu,
khol. raja sa.
[rum' tham ci sokhu] [khol] [kholca broroy]
TWENTY 3 TEN 9 TWENTY TWENTY 4
[kholca broroy ci sokhu] [khol] [raja sa]
TWENTY 4 TEN 9 TWENTY 100 1
[] 79, a Unit of twenty: 80. [] 99, a Unit of twenty: one hundred.

The numerals sa 1, ni 2 and tham 3 never occur without a classifier unless
they are in a compound with ci ~ cit 10 or multiplying a multipliable Round-
Number numeral, e.g. khole ro tham (TWENTY CLF:ROUND.THINGS 3) 23 but
rum' tham (TWENTY 3) 60. In counting for the sake of counting, the classifier for
round objects and money is used obligatorily with the numerals one to three, but not
in 11, 12 and 13. This also applies in numerals above twenty in the vigesimal
paradigms, e.g. khole ro sa (TWENTY CLF:ROUND.THING one) 21, rum' tham
ro tham (TWENTY CLF:ROUND.THINGS three) 63 etc (see Table 40). This
means that the classifier ro (CLF:ROUND.THINGS) is the functionally unmarked or
default classifier (see also Aikhenvald, 2003-b: 335-6). When quantifying objects,
too, the classifier has to be present before the last numeral in every cycle of twenty,
e.g. rum' ni khung sa (TWENTY 2 CLF:FLAT.THINGS 1) 41 flat things, rum' ni
khung ni-(TWENTY 2 CLF:FLAT.THINGS 2) 42 flat things etc. Usually the
classifier is omitted again after 3, e.g. rum' ni khu boroy (TWENTY 2
CLF:FLAT.THINGS 4) 44. The numerals sa 1,ni 2 and tham 3 are
11 NUMERALS

182
monosyllabic and the fact that they attract a classifier might just be to give them a
bisyllabic form.
11.1.2 Round-Number numerals and the use of different paradigms
The Round-Numerals
30
are listed in Table 42. As was mentioned above, a Round-
Number numeral cannot occur unmultiplied.

Table 42 Round-Number numerals
Form Meaning Remarks
sot TEN Used in the decimal paradigm.
kholca TWENTY Used in the Vigesimal 2 paradigm.
rum' TWENTY Used in the Vigesimal 1 paradigm.
raja 100 Preferred to the continuation of the vigesimal
paradigm.
hajal ~ hajar 1000 From Hindi (hazr) thousand.

There is one Hindi loan that is incorporated into the Atong counting system and
combines with Atong classifiers and numerals, viz. hajal ~ hajar 1000 (< Hindi
(hazr) thousand). The word for 100,000 is the Hindi loan lak (< (lakh)
hundred thousand) which is a multiplier since it cannot be used in combination with
another classifier when quantifying nouns.
31

The morpheme sot TEN can only be multiplied by Set 1 and by all of the Set 4
Unit numerals except raa 5 (see Table 41). There is a special morpheme for 5 to
multiply sot TEN, viz. boa 5, that appears nowhere else in the counting system.
When multiplied with dok 6 the result may be one of three forms, viz. a voiceless
fused form /sotok/ [sotok ~ sot:ok], a coordinated form /sotdok/ [sotdok] and a voiced
fused form /sodok/ [sodok ~ sod:ok].

30
I use the term Round-Number numerals in analogy to Matisoffs (1982: 92) Round-number
classifiers.
31
I did not attest the use of the Hindi word (krr) (borrowed into Indian English as crore)
10,000,000 in Atong, which of course does not exclude its use in the language.
11 NUMERALS

183
The morpheme ca times, multiplied by is only attested in combination with
khol TWENTY and nowhere else in the language. However, the combination
kholca TWENTY appears to be transparent to Atong speakers because they
translate the two syllables separately. The compound kholca functions as one
multipliable Round-Number numeral in the language. It is in complementary
distribution with khol, which cannot be multiplied, whereas kholca cannot indicate
only one unit of twenty.
32
Example (181) is illustrative of both the use of khol 20
and kholca TWENTY. In that example the classifier is left out because of the
enumeration of the quantified objects.
Some speakers interpret the multipliable Round-Number numeral rum'
(TWENTY) as a multiplier which cannot be preceded by a classifier. The following
example is illustrative of this opinion.

(182) ni soci morot rum'ni mo'ni gana
[ni so =ci [morot rum' ni mo' ni] {gana}
1pe village =LOC person TWENTY 2 CLF:HUMANS 2 exist
In our village there are 42 people. Literally: In our village exist 42 people.

However, other Atong speakers consider rum' (TWENTY) to be a Round-Number
numeral that can be preceded by a classifier, as we see in example (183). In this
example rum' TWENTY is preceded by the auto-classifier for houses nukhu roof.

(183) uci nukhu raja ni kholgok mu'wano. nukhu raja ni kholgok mu'gabae
thometsongrepha rangkhaymadophae nukhu rum'tham, jekhay ha'coksa
balcido sotok, rokhay toy somsa jalaokno
[u] =ci [nukhu raja ni kholgok] {mu' -wa} =no
DST=LOC roof 100 2 20 stay -FACT =QUOT

32
The fact that kholca TWENTY is used as a Round-Number numeral and not khol TWENTY
might have to do with syllabicity. The ca provides an unstressed syllable in between two stressed
ones. Future investigation into the matter of syllabicity in Atong is required to confirm or reject this
hypothesis.
11 NUMERALS

184
[[nkuhu raja ni kholgok] {mu' =gaba =e}] [th. r.] =e
roof 100 2 20 stay =ATTR =FC Name Name =FC
nukhu rum tham] [jekay] [ha'cik] =sa {bal} =ci =do
roof TWENTY 3 somehow Garo =INSTR speak =LOC =TOP
[sotok] [rokhay toysom] =sa {jal -a -ok} =no
60 Pname river.bank =MOB run.away -AWAY -COS =QUOT
There were 220 roofs (i.e. houses), it is said. As for the 220 roofs which were
there, 60 roofs belonging to Thometsyngrepha [and] Rangkhaimadopha, if
[you] say it in Garo sotok (60), run away to the Rongkhai river bank, it is
said.

The vigesimal paradigm with rum' TWENTY (Vigesimal 1 in Table 40) is archaic
and not even known by all younger speakers in Badri and Siju. Speakers of middle
age, when questioned about this morpheme, are often not sure if its value is 20 or 40
and often argue about this with each other. It might well be that rum' TWENTY was
used exclusively as a multiplier in the past.
Most speakers in Badri and Siju mix Garo and Atong numerals most of the time,
although I have not recorded an Atong speaker using the Garo numeral rica 100 for
Atong raja 100. As for the numbers between 20 and 40, paradigms 1, 2 and 3 (see
Table 40) are still used by middle aged and older Atong speakers in Badri and Siju.
Young people usually use paradigms 4 and 5 with the Garo loan kholgrok 20.
When counting numbers higher than 39, the decimal system of counting now
prevails in Badri and Siju. Young people tend to count in Garo and if they count in
Atong they usually use the decimal system. A lot of younger people do not know the
vigesimal systems any more. Older people who do still know the vigesimal system
seldom use it. The least frequently used is the vigesimal system with the morpheme
ca times, multiplied by, which is called Vigesimal 2 in Table 40. Example (184) is
illustrative of the use of kholca TWENTY. The Round-Number numeral is
preceded by the auto-classifier nukhu roof. Paradigm 7 with the numeral ro'
TWENTY for numerals from 100 up is never used in normal speech any more.
Paradigm 7 with the numeral raja 100 has replaced it.
In example (183) above and (184) below, the speaker, an old man, implies that the
decimal system is actually Garo. Since Atong and Garo are closely related languages
in close contact, it is difficult to say if the decimal system is actually borrowed from
Garo into Atong or not and I will make no attempt to propose any arguments in favour
11 NUMERALS

185
or against the old mans assertions. Note that the speaker uses the vigesimal system
with kholca (TWENTY) in example (184) and the vigesimal system with rum'
(TWENTY) in example (183).

(184) anaktancepa khomabalcepa grot'oksopa sanaga'jophadara cigasa
jalagabae nukhu sotcet ganano. atosa balcido nukhu kholcaboroy
do'anowa.
[a. kh. g. s.] =dara [[ciga] =sa
Name Name Name Name =p Pname =MOB
{jal -a} =gaba] =e [nukhu sot cet] {gana} =no
run.away -AWAY =ATTR =FC roof TEN 8 exist =QUOT
[ato] =sa {bal} =ci =do [nukhu kholca brry]
Atong =INSTR speak =LOC =TOP roof -TWENTY 4
{do' -a} =no -wa}
IE.be -CUST =QUOT -FACT
There where 80 roofs (i.e. houses) belonging to Anaktanchepa
Khymangbalchepa Grytoksongpa [and] Sanagajyngpha, those who had run
away to Chiga, it is said. If said in Atong it is kholchangbyryi (80) roofs, it is
said.

The different systems, vigesimal and decimal are presented in Table 40 in
paradigms but in reality, when people enumerate things, they mix up forms from
different paradigms and often also put in some Garo numerals. The mixing seems to
happen at random and unbound by any rules or principles. We can only remark that it
seems that Atong numerals are becoming obsolescent and the Garo numerals are
taking over, at least in Siju and Badri, where the data for this grammar where
collected. At the moment the language uses the Garo and Atong numerals side by
side. Some Garo numerals are so frequent that they can be considered loans and
therefore I have represented them in the table. The numeral kolgrok 20, invariably
pronounced with a word medial, syllable initial cluster /gr/, is a loan from Garo.
Words borrowed from Garo tend to keep their consonant clusters in Atong whereas
Atong phonology avoids consonant clusters. The Atong numeral ci baa (TEN 5) 15
is a calque on Garo ci boa (TEN 5) 15. The Atong often corrected themselves
when they said ci baa 15, or the Garo form ci boa 15, or kolgrok 20 and then
emphatically said that they meant caraa 15 or kolgok 20. This means that they are
aware of the foreign origin of the numerals they are using.
11 NUMERALS

186
11.2 Borrowed numerals
Atong has borrowed English and Hindi numerals. As we will see in the next section,
these loans are usually used to count English and Hindi loan words referring to
objects associated with modern day life introduced into the Atong society in Hindi or
English.
11.2.1 English loans
Numerals borrowed from English are given in Table 43. The sound changes are
regular. The most salient sound changes are the following: (English > Atong) f > p, v
> b, 0 > t, t
h
> t and > r. Voicing distinction in syllable final stops does not occur in
Atong so the English word final voiced consonants became voiceless in Atong. Some
consonant clusters are preserved, albeit sometimes phonetically modified. Only siks
six can be pronounced with a single final consonant, viz. sik six and the cluster in
sixty is simplified to sikty sixty while sixteen can be pronounced both as sikstin
or siktin. The English numerals borrowed into Atong combine as in English to form
higher numerals. The use of the English loan jero zero is the only way to say zero
since Atong does not have a native term to express this number.
Not only did Atong borrow the morphemes of the English numerals, they
borrowed the counting system, i.e. the order in which the morphemes are combined to
make compound numerals. Thus two hundred fifty two in the Atongs English loan
paradigm is tu handrot pipit tu 2 100 50 2.
Both English and Hindi loans are used without classifiers and the order of the
elements in the noun phrase is NOUN NUMERAL, which is different from the
NOUN CLASSIFIER NUMERAL or CLASSIFIER NUMERAL NOUN make up of
the noun phrase when Atong or Garo numerals are used.

11 NUMERALS

187
Table 43 English numerals borrowed into Atong

0 jero
1 wan
2 tu
3 tri
4 por
5 payp
6 siks ~ sik
7 seben
8 et
9 nayn
10 ten
11 ileben
12 twelp
13 tortin
14 portin
15 piptin
16 sikstin ~ siktin
17 sebentin
18 etin
19 nayntin

20 twenti
30 torti
40 porti
50 pipti
60 sikti
70 sebenti
80 eti [e.ti]
90 naynti
100 wan handrot
1000 wan tawson
1000000 wan milyon
11.2.2 Hindi loans
Numerals borrowed from Hindi are listed in Table 44. Only the numbers one to
twelve are recorded to have been borrowed. This is due to the fact that the use of
numerals borrowed from Hindi is very restricted, as we will see below.
Sound changes occurred when the numerals were borrowed into Atong. The
original nasalisation of the Hindi source language in the numeral (pc) five has
been lost in Atong but consonant clusters are retained in panc 5 and gyara 11. The
retroflex stop in (t) eight has lost its retroflexion. The Hindi aspirated palatal
affricate phoneme (ch) has been replaced by /c/ in Atong. The difference between
the vowels (aw) and (o) in (do) 2 and (naw) 9 and the length
distinctions on other vowels have disappeared in Atong. Finally, the numeral 4 can
be pronounced with or without final /r/ and 10 has two allomorphs dos ~ das
reflecting Hindi (das).

11 NUMERALS

188
Table 44 Numerals borrowed into Atong from Hindi

Atong Hindi transliteration
1 k ek
2 do do
3 tin tn
4 ca ~ car cr
5 panc pc
6 ce ch
7 sat st
8 at t
9 no naw
10 ds ~ das das
11 gyara gyra
12 bara bra
11.3 What is quantified with which numerals?
When telling the time, the hours are quantified in Hindi, because the word baji hour
is a Hindi loan. For example if one asks ato baji? (what hour) Whats the time?, the
answer at eleven oclock will be gyara baji (eleven hour) Eleven oclock. Minutes
and seconds are quantified in English because the words minit ~ minot minute and
sekon second come from English. When it is twelve minutes past eleven, an Atong
speaker says it is gyara baji twelp minit. In the case of time telling, the noun and how
to quantify it were borrowed together. This is also the case for the measure loans
kilomitor kilometre, mitor metre and sentimitor centimetre. But this symmetry is
not always the case. Gears of a car or other vehicle are also quantified in Hindi, even
though the word ger gear itself is a loan from English. When my friends were
teaching me how to drive a motorbike, on different occasions they remarked: te'ew tin
ger now [in] third gear. When they admire a new car or truck they enumerate the
number of gears: ek, do, tin, ca, panc, ce ger [the car has] one, two, three, four, five,
six gears. However, the number of gears a car has in total is stated in English: ie siks
ger gari its a six gear vehicle.
The use of Hindi numerals in Atong is restricted to the quantification of hours and
gears. Since vehicles do not have twelve gears, and hours are counted according to the
12 hour system, twelve is the highest numeral borrowed from Hindi into Atong.
Telephone numbers and numbers on number plates of vehicles are counted with
English loans. This is due to the fact that these items where introduced in English
11 NUMERALS

189
when they entered India and ultimately entered Atong society. This means that the
Atong speakers took over the habit of counting these items in Egnlish from speakers
of Indic languages rather than directly from English speakers.
As was already mentioned above Hindi and English numerals are used without
classifiers and the order of the constituents within the NP is NOUN NUMERAL.
Apart from the items described above, Atong numerals are used to quantify all
other things. Some English and Hindi loans or loans from other Indic languages, such
as Bengali, either came into Atong through Garo and were thus not perceived as
loans, or are assimilated into Atong to such an extent that speakers do not perceive
them as loans any more. These loans are quantified with classifiers and Atong
numerals. Borrowed and incorporated measure nouns, i.e. nouns denoting a receptacle
and its volume (see 12.4), are quantified with Atong numerals but without classifiers,
just as indigenous measure nouns are. Examples of these assimilated loans are listed
in Table 45. In this table I make a distinction between nouns borrowed from English
or Indic languages without specifying which Indic language the noun comes from.
More research is needed to trace the exact origin of all Indic loans and sometimes
similar lexical forms exist in multiple Indic languages, which makes it impossible to
trace the source.
Money and playing cards, despite both the objects and the words being of non-
Atong origin, are enumerated and quantified in Atong. Example (185) is a part of a
recorded speech act of enumerating cards. The classifier is left out when counting
above tham three, as is usual during enumeration. The word for money in Atong is
the Indic loan taka money, rupee. I did not record the word for playing card but
the word for the game is the Indic loan tas card game(cf. Hindi (ts)), which in
Hindi refers both to the game and to the individual cards.

(185) khu sa, khu ni, khu tham, boroy []
khu sa khu ni khu tham boroy
CLF:FLAT.THINGS 1 CLF:FLAT.THINGS 2 CLF:FLAT.THINGS 3 4
One [card], two [cards], three [cards], four [cards] []

11 NUMERALS

190
Table 45 Examples of loans from English (probably through an Indic language) and
Indic languages with their classifiers. This is not an exhaustive list.
ASSIMILATED COMMON NOUNS
Borrowed
noun
Meaning Origin Classifier Meaning
sendel sandal English
sandal
jora CLF:THINGS.OCCURRING
.IN.PAIRS
longpen a pair of
trousers
English long
pants
khung CLF: FLAT THINGS
glas (Can also
be used as
measure noun.)
glass or its
contents
English
glass
goy' CLF:RESIDUE
khap (Can also
be used as
measure noun.)
cup or its
contents
English cup goy',
thay'
CLF:RESIDUE,
CLF:RECIPIENTS
gari vehicle Indic pan CLF:APPARATUS
cola T-shirt, shirt Indic khu CLF:FLAT.THINGS
jama T-shirt, shirt Indic khu CLF:FLAT.THINGS
khiil iron nail Indic co CLF:IRON.NAILS
ASSIMILATED MEASURE NOUNS
inci inch English inch -
pit foot English feet -

When objects are quantified but not enumerated, a classifier is obligatory with any
numeral. The example below is illustrative.

(186) ado morot mo' seneaw wet saci so'otna man'gaba.
[a] =do {|[morot mo' sene] =aw [wet sa] =ci
1s =TOP man CLF:HUMANS 7 =ACC turn 1 =LOC
{so'ot}| =na {man'} =gaba]}
kill =DAT be.able =ATTR
I [am] the one who can kill seven men in one go.
11.4 The position of the classifier
The classifier precedes the numeral. A classifier can be repeated between the last
multipliable Round-Number numeral and Unit numeral. This repetition is obligatory
with the Units sa 1, ni 2 and tham 3 but is not obligatory for the higher Units, as
we can see in the next two examples. In example (187) the classifier for animals, ma,
is repeated before the numeral ni 2 but not before catgok 8. In example (188) the
classifier for trees, pha, is not repeated before the numeral sene 7.
11 NUMERALS

191
(187) morot man'ay sa'gaba biana taw' ma rajasa mani ra'akno, husori ma
rajasa catgok ra'akno.
[morot man'ay sa' =gaba bia] =na
person in.great.amounts eat =ATTR wedding =DAT
[taw' ma raja sa ma ni] {ra' -ak] =no
chicken CLF:ANIMANLS 100 1 CLF:ANIMALS 2 get -COS =QUOT
[husori ma raja sa catgk] {ra' -ak{ =no
rabbit CLF:ANIMALS 100 1 8 get -COS =QUOT
For the wedding of a rich man, [he] got 102 chicken, it is said [and he] got
108 rabbits, it is said.

(188) ie baganci pan pha rum' ni sene gana
[ie bagan] =ci [pan pha rum' ni sene] {gana}
PRX garden =LOC tree CLF:TREES twenty 2 7 exist
There are 47 trees in this garden. Literally: In this garden exist 47 trees.

Classifiers are always used before khole (TWENTY) and kholca (TWENTY), as
demonstrated in the following examples. In example (189) the word nok house is an
auto-classifier, i.e. a noun that can be quantified without intervention of a classifier
(see Chapter 12).

(189) ni soci nok khole nok tham gan.
[ni so] =ci [nok khole nok tham] {gana}
1pe village =LOC house TWENTY house 3 exist
In our village are 43 houses. Literally: In our village exist 43 houses.

(190) morot man'ay sa'gabaci gore ma kholca boroy ma sa gana.
[morot man'ay sa' =gaba] =ci
person in.great.amounts eat =ATTR =LOC
[gore ma kholca broroy ma sa] {gana}
horse CLF:ANIMALS TWENTY 4 CLF:ANIMALS 1 exist
A rich man had/has 81 horses. Literally: At a man [who] ate/eats in great
amounts exist 81 horses.

The same principle of repetition of the classifier applies in very complex numerals
with the incorporated loan numeral hajal ~ hajar 1000, as we can see in the example
below.
11 NUMERALS

192
(191) ni soci morot mo' hajal sa raja baa khole mo' tham gana.
[ni so] =ci
1pe village =LOC
[morot m hajal sa raja baa khole m tham]
person CLF:HUMANS 1000 1 100 5 TWENTY CLF:HUMANS 3
{gana}
exist
In our village there are 1523 people. Literally: In our village exist 1523
people.

Auto-classifiers, too, have to be repeated between the last multipliable Round-
Number numeral and Unit numeral as is illustrated in (189) and with the following
example.

(192) umikonsa san khole-san sa cow rokhuanowa. ha'coksa balcido sal
kholgroksa noay monicom
[u =mi konsa] [san khole san sa] [cow]
DST =GEN after day 20 day 1 rice.beer
{ro -khu -a} =no -wa
drink -INCOM -CUST =QUOT -FACT
[The rain stopped after 14 days]. After that they continued drinking rice beer
for 21 more days, it is said.
11.5 Syntactic and morphological properties of numerals
Numerals are nominal modifiers and are dependent on classifiers for their modifying
function. Classifiers in turn, cannot modify a noun on their own but are not solely
dependent on numerals to exercise their modifying function. The distributive enclitic
<=phek> (DIS), the interrogative morpheme boysok how much/many? and the Type
2 adjective abun other have been observed in place of a numeral after a classifier.
Unlike in Languages such as Thai and Chinese, in Atong demonstratives are not used
with classifiers to modify NPs, which we can see when we compare Chinese and
Atong in (193). In this example we see that Chinese uses a classifier, viz. ge, after
the distal demonstrative ni, whereas in Atong there is no classifier after the distal
demonstrative ue.

11 NUMERALS

193
(193) That girl is pretty
(a) Chinese:
ni ge nhr hn pioling

DIST CLF girl very pretty
(b) Atong:
ue gawi sol -a
DST girl pretty -CUST

The order of the constituents within a quantified NP is either NOUN CLASSIFIER
NUMERAL or CLASSIFIER NUMERAL NOUN. The latter order occurs less often
than the previous but is not uncommon. The meaning of the NP is the same regardless
of the position of the classifier phrase. Possible pragmatic overtones or differences
require further investigation. Only nouns which are auto-classifiers can be modified
directly by a numeral. In the case of auto-classifiers the constituent order is fixed
NOUN
auto classifier
NUMERAL (see 12.3). Examples (194) and (195) constitute a
minimal pair for the position of the classifier and numeral. In both examples the head
of the NP is the first person plural inclusive personal pronoun na'na (1pi). In (194)
the classifier and numeral follow the pronominal head and in (195) they precede the
pronominal head.

(194) otokokodo nana mo' ni sa'ay mu'na man'naka
{otok -ok} =odo [na'na m ni]
do.like.that-COS =TOP 1pi CLF:HUMANS 2
{sa'} =ay {mu'| =na {man' -naka}
eat =ADV sit =DAT be.able -IFT
When [we] will have done like that, us two will certainly be able to sit and
eat.

(195) hay mo' ni na'na kakay sa'na
[hay] [m ni na'na] {kak} =ay {sa' -na}
come.on CLF:HUMANS 2 1pi bite =ADV eat -DESI
Come on! Us two want/intend to bite and eat [you].

Case marking is always encliticised to the last constituent of the NP and has scope
over the whole NP. Case marking can thus enclitisise to a numeral if it is the last
constituent of an NP. Example (196) is illustrative of this phenomenon and contrasts
with (197), where the noun is the last constituent of the NP and thus receives the case
11 NUMERALS

194
marking. Example (196) is also an illustration of the use of the residue classifier goy'.
In this example, para thatch could also have been classified with the classifier for
culms, kun'.

(196) de' na'a re'earocido iaw para goy' saaw kawancoy
de [na'a] {re'e -aro} =ci =do
well 2s go.away -PROG =LOC =TOP
[i =aw para goy' sa] =aw {kaw -an -coy}
PROX =ACC thatch CLF:RESIDUE one =ACC shoot -REF -TRY
Well, if you are going, try to shoot this one [culm of] thatch.

(197) ana mo' tham na' sa'aw poraykhalna watetboto.
[a] =na [mo' tham na' sa ] =aw
1s =DAT CLF:HUMANS 3 2s child =ACC
{poray -khal} =na {watet} =bo =to
study -MORE =DAT send =IMP =IMPEMPH
For my benefit, do send three of your children to study more.

Sometimes, phrasal enclitics occur not attached to the last constituent of the NP but
only encliticised to the constituent over which they have scope, although this seldom
happens in the recorded material. This can be seen in example (198), where the
enclitic <=rara> (AMONGST) is encliticised to the head of the NP and not the numeral
which forms the last constituent. The enclitic has scope only over the noun boba (<
Indic) crazy person. Everywhere in the language where an enclitic does not
enclitisise to the last element in the phrase, it only has scope over constituents that
precede it but not over those that follow it.

(198) te'do ge'thethedo, bobarara mo'ni golpho ka'rukokno.
[te'] =do [ge'thethe] =do [boba =rara mo' ni]
now =TOP 3P =TOP crazy.person =AMONGST CLF:HUMANS 2
{golpho kha' -ruk -ok} =no
story do -RC -COS =QUOT
Now the two crazy persons gossiped amongst each other.

11 NUMERALS

195
In example (199) the scope of the focus/identifier enclitic <=an> stretches only over
the noun. Example (199) contrasts with (200) where the noun is omitted and the focus
identifier appears on the numeral head.

(199) ucie somo morotan mo' tham re'eaymo []
ucie [so =mo morot =an mo' tham] {re'e} =ay =mo
then village =GEN person =FC/ID CLF:HUMANS 3 go.away =ADV =SEQ
Then, the three villagers having gone, [].

In example (200) the NP mo' korokawan is headless. The classifier refers to the
older brothers of a child who, in the story from which this example is taken, is
searching for them.

(200) pholgom cunggaba monokrumokno mo' korokawan.
[pholgom {cu} =gaba] {monok -rum -ok} =no
eagle big =ATTR swallow -ALL -COS =QUOT
[mo' korok] =aw =an
CLF:HUMANS 6 =ACC =FC/ID
The big eagle had swallowed all 6 persons, it is said.

When the quantified noun is understood from the context, it can be elided as we
see in example (200). What remains after the elision of the noun is a headless NP. The
numeral-plus-classifier combination can stand on its own or be preceded by a
demonstrative, as we see in example (201).
Demonstratives invariably occur in NP initial position in Atong. When the NP is
headless, the demonstrative will remain in that position and may thus precede a
classifier-plus-numeral, as we can see in example (201). In this example, the NP ue
mo' sa (DST CLF:HUMANS 1) that (one) person refers to a lazy king. The clause
is of the equation/identity type with a nominal predicate. The head of the NP that
functions as the predicate is ellipsed. What is left in the predicate is an attributive
clause(see Chapter 29) indicated between vertical lines.

11 NUMERALS

196
(201) te'awba ue mog' sa {kam kha'na haratgaba.}
--------attributive clause---------
[te'ew] =ba [ue mo' sa]
S
{|[kam] {kha'} =na {harat}| =gaba]}
now =EMPH DST CLF:HUMANS 1 work do =DAT reluctant =ATTR
Now that one is [someone who is] reluctant to do work.

Auto-classifier noun phrases cannot omit the noun since this would leave the numeral
standing on its own without classifier and that would be ungrammatical. The noun nok
can be used as an auto-classifier, e.g. nok sa (house 1) one house. If nok house
were omitted it would result in ungrammaticality.
Classifier-plus-numeral combinations can function as head of a predicate of an
existential clause. This is shown in example (202), where the whole enumeration
functions as non-verbal predicate head. Classifier-plus-numeral phrases together with
the noun they quantify are not attested together as predicate head. More fieldwork
needs to be carried out to find out if such constructions are possible.

(202) mo' byryi mo' baa mo' korokkhua
{mo' boroy mo' baa mo' korok -khu -a}
CLF:HUMANS 4 CLF:HUMANS 5 CLF:HUMANS 6 -INCOM -CUST
There are still four, five, six persons left.

Quantified NPs can take the plural enclitic <=dara> (p). This enclitic makes the
quantification approximate, e.g. example (203).

(203) imi kilomitor kolgokdoratokoy ni'wa ra.
[i] =mi [kilomitor kolgok] =dra =tokoy {ni' -wa} [ra]
PRX =ABL kilometre 20 =p =VIA/LIKE not.exist -FACT rain
About twenty kilometres from here there is none, rain.

Approximation can also be expressed by enumeration. The following example, from a
story about incantation, is illustrative. In this example we see that the classifier is
present before each numeral. The perlative/similative enclitic <=tokoy> (VIA/LIKE)
is not compulsory in this construction. The context is as follows. There are many
people in the village who claim they know how to perform incantations. In reality
they dont know.
11 NUMERALS

197
(204) mo' sa mo'nitokoy sapa.
[mo' sa mo' ni] =tokoy {sap -a}
CLF:HUMANS 1 CLF:HUMANS 2 =VIA/LIKE know -CUST
One or two actually know.

The plural enclitic can also be used to reinforce the notion of plurality, of which the
next example is illustrative. The context and the use of the delimitative enclitic <=sa>
(DLIM) help to interpret the use of the plural enclitic in this case.

(205) te'ew wenthamdarasa miniksuruni baysigane.
te'ew [wen -tham] =dra =sa {miniksuru -ni} [bay'siga] =ne
now turn -three -p =DLIM flat.haired -FUT friend =TAG
Now three times [and] you will be flat-haired, friend ok. (said the deer to the
fox who was bathing in the river but whose hair kept standing up instead of
getting flat.)
11.6 Ordinal numbers
Atong and Garo numerals both simple and compound, but not borrowed numerals,
can be turned into ordinal numerals with the attributive suffix <-gaba ~ -ga> (ATTR).
The result of this process is translated into English with an ordinal numeral. The
suffix <-gaba ~ -ga> (ATTR) is attached to the last element of a numeral.
Ordinal numerals function as modifiers of nouns. The ordinal numeral can precede
or follow the noun it modifies with no difference in meaning. However, when the
ordinal numeral follows the noun it modifies, it is always accompanied by a classifier
preceding the ordinal numeral, whereas the classifier is absent when the ordinal
numeral precedes the noun it modifies.
Example (206) is an illustration of an ordinal numeral preceding the noun it
modifies. Notice the absence of a classifier. This example contrasts with example
(207), where the ordinal numeral accompanied by the classifier follows the noun it
modifies.

(206) sagaba naw nemkhalbutuci thoyok.
[sa -gaba naw] {nem -khal -butu} =ci {thoy -ok}
1 -ATTR younger.sister good -CP -WHILE =LOC die -COS
When [my] first younger sister was getting better [she] died.

11 NUMERALS

198

(207) unasa boba mo' sagaba te'ew abun boba nukaysigaakno
unasa [boba m sa -gaba] [te'ew] [abun boba]
then crazy.person CLF:HUMANS 1 -ATTR now other crazy.person
{nuk -ay -siga -ak} =no
see -TOWARDS -ALT -COS =QUOT
The first crazy person now saw another crazy person coming towards him, it
is said.

An ordinal numeral can occur in a headless NP just as a numeral, as we have seen in
the previous section. In this function the ordinal number is usually accompanied by a
preceding classifier, which limits the scope of reference of the headless NP. This is
illustrated in the following example.

(208) una mo' sagababa, sa'banthay mo'sagaba, bocomokno.
una [mo' sa -gaba] =ba
then CLF:HUMANS 1 -ATTR =EMPH
[sa' banthay mo' sa -gaba] {bocom -ok} =no
child bachelor CLF:HUMANS 1 -ATTR pull.out -COS =QUOT
Then the first [son] pulled out the other son, it is said.

However, in enumerations the classifier is omitted in headless NPs. Example (209)
comes from a text in which the speaker enumerates all the Atong speaking villages.
The two clauses presented in this example are of the identity/equation type with a
nominal predicate.

(209) sagaba Rodo ha'way. [] sotbyrisagaba bakmara konagothum.
[sa =gaba]
S
{rodo ha'way}
1 -ATTR Pname
[sot -byri sa -gaba]
S
{bakmara konagothum}

TEN 4 1 -ATTR Pname
The first [village is] Rongdyng Hawai. [] The forty first [village is]
Baghmara Konagythum.

As we saw in example (208), the construction [CLF saga(ba)] [CLF saga(ba)]
means one the other. The following example shows that [CLF saga(ba)] on its
11 NUMERALS

199
own in the appropriate context can also mean the other. The context is as follows:
Someone comes home late and his children ask where he has been. He answers that he
was at their uncles house, to which the children reply that he is lying since the uncle
had visited them that day and has not left yet. Then the late person replies (210).

(210) na' awa mo' sagami nokcisate!
[na' awa mo' sa -ga =mi
2s fathers.younger.brother CLF:HUMANS 1 -ATTR =GEN
nok] =ci =sa =te
house =LOC =DLIM =DCL
At your other uncles house I say!
11.7 The numeral sa one: its different functions and grammaticalisations
The morpheme sa has several functions which can be seen as grammaticalisations of
the numeral sa one. We shall now examine the different functions of this morpheme
one by one.

A) Numeral
The numeral may be used to indicate that there is really one item of something, e.g.
example (211).

(211) ue gawici sa' mo' korok gananoro aro de'the pipukci ganakhua mo' sa,
mo.
[ue gawi] =ci [sa' mo' korok] {gana} =no =ro aro
DST woman =LOC child CLF:HUMANS 6 exist =QUOT =EMPH and
[de'the pipuk] =ci {gana -khu -a} [mo' sa]
3s belly =LOC exist -INCOM -CUST CLF:HUMANS 1
The woman has/had six children, it is said, and in her belly there is/was one
more.

11 NUMERALS

200
B) Indefinite article
A noun in Atong is not marked for singular or plural and therefore can be interpreted
as referring to one or more than one entity. So the clause matdam sa'-ak (otter eat-
COS) can mean either the otter had eaten [the fish] or the otters had eaten [the
fish]. The numeral sa one is often used to make the plural reading of a noun
impossible. In such cases the meaning of sa one is more like that of the English
indefinite article a/an, e.g. (212).

(212) ucie ramci pheru ma sa gorowano.
ucie [ram] =ci [pheru ma sa] {goro -wa} =no
then road =LOC FOX CLF:ANIMALS one meet -FACT =QUOT
Then, on the road he met a fox, it is said.

As an indefinite article, the numeral sa one is often used to introduce new referents,
as is the case in example (212). This function of the numeral one has also been
described for Lahu in Matisoff (1982: 87) and it is attested in languages around the
world, (see Heine and Kuteva (2002: 2201) and Givn (1981, 1984: 432-35)

C) Other
The combination CLF one-ATTR means either one..the other if it occurs twice in
a row (213), or just the next one, another one (214). Any classifier can participate in
this construction, as, for example, the auto-classifier (see 12.3) dol group in (213)
and the classifier for humans, mo, in (214).

(213) dol sagaba rajamo jagosisa dol sagaba rajamo jagarasatakay mu'ni.
[dol sa -gaba] [raja =mo jagosi =sa] [dol sa -gaba]
group 1 -ATTR king =GEN right =MOB group 1 -ATTR
[raja =mo jagara] =sa =takay {mu' -ni}
king =GEN left =LOC =LIKE sit -FUT
One group will sit to the kings left hand side and the other group will sit to
the kings right hand side. Literally: the other group will sit kings-left-
sidedly.

11 NUMERALS

201
(214) mo' saba rookno. mo' sa do romanokno, jamancakno. otokoymuna mo'
sagaba rothiriokno. uba jamancano.
[mo' sa] =ba {ro -ok} =no [mo' sa] =do
CLF:HUMANS 1 =EMPH drink -COS =QUOT CLF:HUMANS 1 =TOP
{ro -man -ok} =no {jam -an -ca -k} =no
drink -ALREADY -COS =QUOT finish -REF -NEG -COS =QUOT
otokoymuna [mo' sa -gaba] {ro -thiri -ok} =no
so.then CLF:HUMANS 1 -ATTR drink -AGAIN -COS =QUOT
[u] =ba {jam -an -ca -k} =no
DST=EMPH finish -REF -NEG -COS =QUOT
One [of the brothers] smoked, it is said. One [of them] has already smoked,
[it] is said, [he] did not finish [it], it is said. So then another [brother] smoked,
it is said. That one also did not finish it, it is said.

The next example illustrates the use of the construction CLF one-ATTR meaning
one..the other, but without the classifier. The noun so village is not an auto-
classifier. The classifier normally used for villages is dam (CLF:VILLAGES). The
use of this construction without a classifier is only attested for the noun so village.

(215) so sagabaaw somo mowanowa, so sagabaaw sogadal mowanowa.
[so sa -gaba] =aw [somo] {mo -wa} =no -wa
village one -ATTR =ACC Pname call.a.name -FACT =QUOT -FACT
[so sa -gaba] =aw [sogadal] {mo -wa} =no -wa}
village one -ATTR =ACC Pname call.a.name -FACT =QUOT -FACT
One village was called Songmong, the other village was called Songgadal.

This type of grammaticalisation of the numeral one is also attested for the languages
Bulu and Yagira, (see Heine and Kuteva, 2002: 223).
D) Delimitative
The delimitative enclitic <=sa> (DLIM) is a grammaticalisation of the numeral one
into an enclitic. The delimitative enclitic is found on subordinate clauses, NPs,
adverbs, personal pronouns, discourse connectives and demonstratives, and limits the
reference of the clause or phrase. The meaning of the morpheme can schematically be
represented as only/exclusively/precisely X.
11 NUMERALS

202
The following examples show occurrences of the delimitative on a noun
functioning as predicate head and on a clause with a verbal predicate head
respectively.

(216) na'mi jorado na'mi madamsate!
[na' =mi jora] =do
S
{[na' -mi madam] =sa} =te

2s =GEN lover =TOP 2s =GEN female.teacher =DLIM =DCL
Your lover is no one other than your teacher!

(217) sogumukan ue moma wana waykhurutaysa boli hon'aysa man'ay
sa'thokwano.
[so =gumuk] =an [ue moma wa] =na {way khurut =ay} =sa
village =whole =FC/ID DST elephant tooth =DAT spirit incantate =ADV =DLIM
{boli hon'} =ay =sa
offering give =ADV =DLIM
[man'ay] {sa' -thok -wa} =no
in.great.amounts eat -everyone -FACT =QUOT
The whole village, precisely [because] [they] prayed to the elephant tusk
[and] precisely [because they] gave offerings, [they] all became rich (lit.
[they] all ate in great amounts), it is said.

Example (218) illustrates the use of the delimitative enclitic on a quantified NP.

(218) morot mo'sasa botawano.
[morot mo' sa] =sa {bot -a -wa} =no
person CLF:HUMANS one =DLIM lead -AWAY -FACT =QUOT
Only one man led him away, it is said.

In the next example we see the distal demonstrative with the delimitative enclitic. The
second occurrence of that enclitic is on a headless NP with an attributive clause.
(219) raso man'ay takokno usa, cungcugarasa.
[raso] {man'} =ay {tak -ok} =no [u] =sa
boasting in.great.amounts =ADV do -COS =QUOT DST=DLIM
[cu cu =ga] =ra (Garo) =sa
big RED =ATTR =P =DLIM
Those very [ones] boasted a lot [about themselves], it is said, those eldest
ones.

11 NUMERALS

203
The following example shows an occurrence of the delimitative enclitic on a
discourse connective.

(220) umusa sogumuk thom'aymu ha'ba ha'ron ha'ronaw sowalni.
umu =sa [so] =gumuk {thom'} =ay =mu
then =DLIM village =whole gather =ADV =SEQ
[ha'ba] [ha'ron ha'ron] =aw {sowal -ni}
dry.rice.and.vegetable.field plot RED =ACC divide -FUT
[In the beginning [they] begin with a general meeting.] Only then, after the
whole village has gathered together, will [they] divide the dry rice and
vegetable field plot by plot.

The development of the number sa one to the delimitative enclitic <=sa> (DLIM)
might very well have involved a hypothetical intermediate stage in which the
morpheme was interpreted in certain contexts as meaning only one, unique before it
developed further into the delimitative enclitic, which means only, just.

204
Chapter 12 Classifiers
_____________________________________________________________________

In Chapter 1 we have already discussed the following. Classifiers are only used with
indigenous Atong numerals and with Garo loan numerals. Classifiers are nominal
modifiers in the sense that they limit the number of possible referents of an NP. When
the quantified noun is understood from the context, it can be elided, with the result
being a headless NP. In a headless NP a classifier cannot take over the syntactic
functions of the elided noun in that it cannot take any case or other nominal suffixes.
These affixes will go onto the numeral instead, which is always the last element in the
headless NP.
In 12.2 of this chapter we look at the different types of classifiers and how they
are subcategorised. We will look at the use of classifiers and determine the parameters
that determine the choice of a certain classifier with a certain noun. Subsequently, in
12.3 we will look at a subclass of nouns that does not need a classifier to be
quantified, the so-called auto-classifiers. In 12.4 we will look at the double function
of measure nouns. Section 12.5 treats the origin of the Atong classifiers. All
classifiers are given in Table 47 at the end of this chapter, organised according to
different semantic and formal categories.
12.1 The syntactic and semantic properties of classifiers
As has been mentioned in Chapter 1, the order of the elements in a quantified NP is
either NOUN CLASSIFIER NUMERAL or CLASSIFIER NUMERAL NOUN. The
latter order is not frequently attested but by no means uncommon in the language. The
meaning of the NP is the same regardless of the position of the classifier phrase.
Possible pragmatic overtones or differences require further investigation. Only nouns
which are auto-classifiers can be modified directly by a numeral. In the case of auto-
classifiers the constituent order is fixed as NOUN
auto classifier
NUMERAL (see 12.3).
Examples (194) and (195) in Chapter 1 constitute a minimal pair for the position of
the classifier and numeral.
12 CLASSIFIERS

205
As is discussed in 11.5, the classifier relies on the following numeral to carry
case and other enclitics. As is also mentioned in Chapter 1, numerals are nominal
modifiers and are dependent on classifiers for this modifying function. Classifiers, in
turn, cannot modify a noun on their own but are not solely dependent on numerals to
exercise their modifying function. The distributive phrasal enclitic <=phek> (DIS), the
interrogative morpheme boysok how much/many? and the Type 2 adjective abun
other are attested in place of a numeral after a classifier.
Example (221) below illustrates the most common use of a classifier, i.e. in
combination with a numeral after the head noun in the NP.

(221) ketketa buraci koy' ma sa ganano
[ketketa bura] =ci [koy' ma sa] {gana} =no
Name =LOC dog CLF:ANIMALS 1 exist =QUOT
Ketketa Bura has a dog, it is said.

A classifier can be used to narrow down the scope of reference of an NP. If the
context is sufficiently clear, a classifier can be used without a noun, in which case we
have to do with a headless NP. This phenomenon was described and discussed in
11.5. The following example illustrates the use of a classifier without a noun and
without a numeral. Instead of a numeral the classifier is followed by the distributive
enclitic <=pek> (DIS). This is the only recorded example where the long form of the
first person personal pronoun, aa (1s) (see 17.2), is not in A or S but seemingly in
oblique function.

(222) aca na'tome aa sanci mapek hon'ni nowano.
aca [na' -tom] =e
A
[aa]
OBLIQUE/RECIPIENT
[san] =ci
interj 2s -ppp =FC 1s day =LOC
[ma]
O
=pek {hon' -ni} {no -wa} =no
CLF:ANIMALS =DIS give -FUT say -FACT =QUOT
Right then, you
p
shall give me one of each animal every day, [the lion] said,
it is said.

There is one recorded occurrence of a Type 2 adjective directly after a classifier. This
is the Type 2 adjective abun other, as we can see in example (223). Since there is
12 CLASSIFIERS

206
only one recorded occurrence of this type of construction, chances are that it was a
mistake of speaker or that we are dealing here with a new emerging construction.
More fieldwork needs to be done to find out whether this construction is really
productive in Atong or not.

(223) boba mo'sagabaci mo' abun bobaci te'ewdo bobarara mo'ni golpho
kha'rukokno
[boba mo' sa =gaba] [m abun boba] =ci
crazy.person CLF:HUMANS 1 =ATTR CLF:HUMANS other crazy.person =LOC
[te'ew] =do [boba =rara mo' ni]
now =TOP crazy.person =AMONG CLF:HUMANS 2
{golpho kha' -ruk -ok} =no
The first crazy person to the other crazy person, now, among crazy persons,
the two of them gossiped to each other, it is said.

Example (224) illustrates the occurrence of a classifier with the interrogative
morpheme boysok how much/many?. This postposition can only occur after
classifiers (see also 9.9).

(224) na'tome goy'boysok man'phawa ie bolsie?
[na' -tom] =e [goy bysk] {man' -pha -wa}
2s -ppp =FC CLF:RESIDUE how.many obtain -IN.TOTAL -FACT
[ie bolsi] =e
PRX year =FC
How many did you get this year? Implied: How many baskets were you able
to fill with rice this year? (during the rice harvest).

Classifier-plus-numeral combinations can be the head of a predicate of an existential
clause, as has been remarked in 11.5 and illustrated by example (202), repeated here
as (225). The sentence in this example consists of a coordination of two clauses, each
of which contains a headless quantified NP as a predicate. The second clause consists
of headless quantified NPs which are in an enumerative relationship to each other, the
last one of which is the head of the predicate taking the incompletive and customary
aspect suffixes.
12 CLASSIFIERS

207
(225) mo'thamkhua; mo'boroy mo'banga mo' korokkhua
{mo' tham -khu -a} {[mo' broroy]
CLF:HUMANS -three -INCOM -CUST CLF:HUMANS four
[mo' baa] [mo' korok -khu -a]}
CLF:HUMANS five CLF:HUMANS six -INCOM -CUST
There are still three persons left; there are still four, five, six persons left.
12.2 Categories and types of classifiers and their use
Classifiers are divided into two major classificatory categories, viz. sortal, mensural.
The sortal classifiers are used for animate, inanimate and count nouns. All classifiers
can be divided into two major semantic categories, depending on the animacy of the
nouns they can occur with, viz. Animate and Inanimate. Mensural classifiers are only
attested with inanimate nouns, while Sortal classifiers occur with both animate and
inanimate nouns. The category Animate classifiers is divided into two sub-categories,
viz. Humans and Animals. The category Inanimate is in turn split up into smaller
groups of classifiers, viz. plants, shape and dimension, consistency, function,
mensural, residue, repeater and specific. Mensural classifiers are divided into three
subcategories, viz. those classifying objects by arrangement, those classifying objects
by both arrangement and shape and those classifying objects by quantity, i.e. volume,
length or distance, weight and surface. The categorisation of classifiers is represented
in Table 46.
As for group classifiers, there are none. There are two words that mean group,
viz. dol and jinma. These words are auto-classifiers, i.e. nouns that can be quantified
without classifier (see 12.3). These words have all nominal properties. The words dol
is only attested to refer to groups of people, while jinma can be used for both groups
of people and of animals.
Mensural classifiers, treated in 12.2.3, below, individuate in terms of quantity
(see Lyons 1977:463), while all other categories of classifiers individuate whatever it
refers to in terms of the kind of entity that it is or the way people relate to it (see
Lyons 1977:163) and are thus sortal classifiers. Sortal classifiers will be treated in
12.2.1. Repeaters, treated in 12.2.2, only occur with compounded NP heads and
only very few are attested so far.
Some nouns can take alternative classifiers when they are quantified. The choice
of classifier depends on which property of the quantified noun the speaker wants to
12 CLASSIFIERS

208
focus on or finds most relevant in the context of the utterance, as we shall see in
12.2.4. Not every noun has a fixed classifier. The possibilities of reclassification of
the noun referent are discussed in the same section.

Table 46 The categorisation of Atong classifiers
Sortal Animate humans
animals
Inanimate plants
shape and dimension
consistency
function
repeater
specific
Mensural residue
mensural by arrangement
mensural by arrangement and shape
mensural by quantity volume
length/distance
weight
surface

It is important not to confuse classifiers with auto-classifiers and measure nouns.
The former are a subclass of nouns and the latter are a separate word class. Both will
be treated separately below in 12.3 and 12.4 respectively.
12.2.1 Sortal classifiers
The choice of sortal classifier is determined by the inherent physical and semantic
properties of the quantified object. First of all, objects are categorised according to
animacy into humans, animals and inanimate objects. There is one classifier for all
nouns denoting a human, viz. mo' (CLF:HUMANS) and one for all nouns denoting
animals, viz. ma (CLF:ANIMALS). It is interesting to note that the classifier for
animals is also used for knives and other tools, in which case it can be labelled as
(CLF:TOOLS).
About one score of sortal classifiers denoting different types of inanimate objects
has been recorded so far. The different semantic categories that determine the choice
12 CLASSIFIERS

209
of the classifier are plants, shape and dimension, consistency, function, apparatus/
appliances.
The classifier goy' (CLF:RESIDUE) can be used to classify almost anything in cases
where the speaker does not want to use a more specific classifier. This use of goy' is
illustrated in example (196) in Chapter 1. The classifier goy' (CLF:RESIDUE) is also
frequently used when the object quantified is a loanword. This is the residue classifier
function, e.g. gari goy' tham (vehicle CLF:RESIDUE 3) 3 vehicles. The classifier goy'
(CLF:RESIDUE) is not attested to replace the human or animal classifiers mo'
(CLF:HUMANS) and ma (CLF:ANIMALS), but can replace Inanimate, and, less
frequently, Mensural classifiers. As replacement of a mensural classifier, goy' is only
attested replacing a volume, e.g. (224).
When counting for the sake of counting the default classifier ro
(CLF:ROUND.THINGS) is used, as has been said in 11.1.1. To put it more specifically,
the core semantics of the default classifier ro is round objects, but it can also have an
unspecified referent function i.e. when the speaker does not express a referent (e.g.
counting for the sake of counting), whereas the residue classifier goy' does not have a
core semantic meaning and has both a residue function, to classify nouns that fall
outside the semantic domain of certain other classifiers, and a default function, in
which goy' can be substituted for other classifiers when the speaker wants to
abandon the available precision of a semantically specific classifier in favour of a
semantically neutral [one] (Zubin and Shimojo, 1993: 491).
12.2.2 Repeater classifiers
A repeater classifier can only occur in Atong if this classifier morpheme is the same
form as the last morpheme of a compound which is the quantified head of the NP. Not
just any noun that is the last morpheme of a nominal compound can occur as a
repeater classifier. The number of repeaters seems to be very restricted, since not
many of them have been recorded in Atong.
12 CLASSIFIERS

210
Compounds containing the root khal hole all take khal as a repeater classifier,
e.g. ha'khal khal sa (cave
33
CLF:HOLES 1) one cave and nakhu khal khal ni (nose
hole CLF:HOLES 2) two nose holes, but note that nakhukhal ni two nose holes
without classifier is also possible. Note that the compound na+khal (to.hear+hole)
ear is quantified as a limb, viz. nakhal sam sa (ear CLF:LIMBS 1) one ear and the
lexical compound toy+khal (water+hole) river is quantified with the classifier for
roads and rivers, col, since it is not seen as a type of hole in the language. There are
several other morphemes in the language, both bound and free, that behave similarly
to khal hole in compounds, e.g. the bound morpheme ta ?
34
that only occurs in
the noun kokta type of basket in which the morpheme kok basket is a recurrent
element in many names for baskets. The basket kokta is counted by repeating the last
morpheme, viz. kokta ta sa (type of basket CLF:KOKTA-one) one kokta. Another
repeater is found for the noun wa'su bamboo cylinder used to cook food in
35
which
is counted wa'su su tham three bamboo cylinders used to cook food in. Table 47
contains an exhaustive list of repeater classifiers so far attested in Atong.

33
The morpheme ha' ?, of which the meaning is unknown, has only been attested in the compound
noun ha'khal cave.
34
The morpheme ta only occurs in the noun kokta type of basket and might go back to Proto-
Tibeto-Burman *ta tense/tight (Matisoff 2003: 614) which would mean that it is a tightly woven
basket.
35
The word su as free morpheme has only been recorded in Atong with the meaning remembrance,
thought, mind, brain, intelligence and as the repeater morpheme of the only compound it apparently
occurs in, viz. wa'su bamboo cylinder (used as container). The Atong lexeme wa'su bamboo
cylinder (used as container) is no doubt cognate with the Garo word /wa'si/ (phonological
representation based on Burling 2004). pronounced [wa?si], orthography wasing, with the same
meaning. Atong and Garo are very closely related languages and in both languages wa' means
bamboo and can occur as a free morpheme. In Garo the second element of the compound wa'+si
does not occur as a free morpheme or in any other compound. In the Garo dictionary by Nengminza
(2001: 232) under there entry sing- we read: A numeral prefix [i.e. classifier] used for bamboo cups
as wa[]sing singsa one bamboo cup or a bamboo tube. There is no other entry sing in the
dictionary. It is conceivable, although speculative, that in an older stage of Atong and Garo there was a
word meaning tube, cylinder, receptacle that only survived in the compounds wa'si in Garo and
wa'su in Atong and as their repeater classifiers.
12 CLASSIFIERS

211
12.2.3 Mensural classifiers
The mensural classifiers recorded so far have only been used to quantify quantities of
inanimate objects. There are three types of mensural classifiers in Atong, viz. 1.
mensural by arrangement, 2. mensural by arrangement and shape and 3. mensural by
quantity. The choice of the classifier is determined for

Type 1 only by the arrangement in which a quantity of objects occurs,
Type 2 by the shape of the objects itself and by the arrangement in which the
objects occurs,
Type 3 only by the quantity in which the objects occur.

Type 3 mensural classifiers can be subdivided into those denoting length or distance,
weight, volume and surface.
It is interesting to note that according to Aikhenvald (2003-b: 115) in the majority
of the worlds languages the choice of a mensural classifier is determined by only two
factors, viz. the quantity, or measure, of an entity, and its physical properties. Atong
belongs to this majority of languages. The physical properties that Atong uses to
determine the use of a mensural classifier are both temporary, i.e. arrangement, and
permanent, i.e. shape.
12.2.4 The relationship between noun and classifier
A classifier can help determine the exact denotation of a noun, i.e. can disambiguate
potentially polysemous nouns. Some nouns have alternative choices of classifier
depending on which property of the noun is in focus. If a speaker talks about oranges
as fruits, he will use the classifier for round things, ro, e.g. nara ro ni (orange
CLF:ROUND.THINGS 2) two oranges. But if a speaker indicates oranges in little heaps
on the market, it will be nara com' ni (orange CLF:HEAPS.OF.SMALL.ROUND.FRUITS
.AND.VEGETABLES 2) two heaps of oranges. There is also the possibility to speak
about an orange as a tree, in which case the classifier for trees, pha, is used, e.g.
nara pha sa (orange CLF:TREES 1) one orange tree. An umbrella can be classified
as an apparatus when it is closed and as a flat thing when it is open, e.g. satha pan ni
(umbrella CLF:APPARATUS 2) two umbrellas, satha khu sa (umbrella
CLF:FLAT.THINGS 1) one umbrella. In stories animals are often reclassified as humans
because they act like humans, e.g. (226), where the classifier refers to a toad, a frog
12 CLASSIFIERS

212
and a bird, who together are on their way to beat-up an elephant for destroying their
dwellings.

(226) otokoymo mo'tham re'eokno.
otokoymo [m tham] {re'e -ok} =no
so.then CLF:HUMANS 3 go.away -COS =QUOT
So then the three of them went away.

Nouns with human referents have to be classified as humans and cannot be
classified as something else. Animals can be classified as animals or, in stories, as
humans. Nouns that can be classified according to their arrangement can also be
classified according to their shape and dimension. Nouns that fall into the function
category can also be classified according to their shape and dimension. Nouns that
have specific classifiers can be classified with the residue classifier goy'. Some
apparatus can be classified according to their shape and dimension. Nouns which
require repeater classifiers cannot be reclassified. More research is needed to find out
what the precise dynamics are within the classifier system.
Loan words can be classified by their semantics just like native Atong words.
Certain words are more frequently classified by the residue classifier than others and
this might be due to the degree with which these loans are integrated into Atong.
Moreover, certain semantic classes of borrowed nouns are more frequently classified
according to their semantics than others. The borrowed measure nouns dipot teapot,
gylas glass and khap cup (all English loans), for example, are usually counted with
the residue classifier goy'. Borrowed nouns denoting clothes, like jama shirt, cola
shirt, muja sock (all Indic loans) and lopen long trousers (English loan) are
classified like other clothes, i.e. with the classifier for flat things ku or the classifier
for things that occur in pairs jora.
12.3 Auto-classifiers
Auto-classifiers are a subclass of nouns that can be quantified without intervention of
a classifier. Instead the numeral always directly follows the noun. Auto-classifiers
consist of: 1. Time Nouns, i.e. nouns denoting a unit of time, and 2. some
miscellaneous nouns.
12 CLASSIFIERS

213
The Time Nouns are bolsi year, ja month, nogoltoy week, san day, khantha
hour, minit minute and sekon second. Example (227) shows how the Time Noun
san day is quantified without the intervention of a classifier between it and the
compound numeral.

(227) range san cibori wawano.
[ra] =e [san ci bori] {wa -wa} =no
rain =FC day 10 4 rain -FACT =QUOT
the rain rained fourteen days, it is said.
Examples of other nouns that are auto-classifiers are wen' ~ wet time, turn, nok
house, nukhu ~ nokhu roof, khal hole, mom' fist, jinma group, dol group
and bol' stroke, blow. The allomorph wet time, turn appears only before the
numeral sa one and the allomorph wen' before all other numerals as well as before
sa one. In example (228) we see the noun bol' stroke, blow quantified without the
intervention of a classifier between it and the numeral.

(228) otokoymu bol' sa toketokno.
otokoymo [bl sa] {tok -et -ok} =no
so.then blow 1 beat -CAUS -COS =QUOT
So then [he] gave (Lit. hit) another blow, it is said.

Auto-classifiers can be interrogated with the interrogative morpheme boysok how
much?, how many? just like classifiers. Example (229) illustrates the use of this
interrogative morpheme with a Time Noun, while in (230) we see it with a more
prototypical noun.

(229) sanboysok mu'ni?
[san boysok] {mu' -ni}
day how.many stay -FUT
How many days will [you] stay?

(230) na' soci nokboysok gana?
[na' so] =ci [nok boysok] {gana}
2s village =LOC house how.many exist
How many houses are there in your village?
12 CLASSIFIERS

214
Between the last Round-Number numeral and Unit numeral (see 11.1 for the
definitions), auto-classifiers can be repeated just as other classifiers, as we saw above
in 11.4. Examples (189) and (231) are illustrative. In example (231) the speaker
clarifies the vigesimal number by translating it into Garo, in which counting is
different.

(231) umikonsa san khole san sa cow rokhuanowa. ha'coksa balcido sal
kholgroksa noay monicom
[u =mi konsa] [san khole san sa] [cow] {ro -khu -a}
DST=GEN after day TWENTY day 1 rice.beer drink -INCOM -CUST
=no -wa [ha'cok] =sa {bal} =ci =do [sal kholgrok sa]
=QUOT -FACT Garo =INSTR speak =LOC =TOP day twenty 1
{no} =ay {mo -ni} =com
say =ADV call.a.name -FUT =IRR
[The rain stopped after 14 days]. After that they continued drinking rice beer
for 21 more days, it is said. If [you] say [it] in Garo, [you] would say sal
kholgrok sa.

When they are quantified, the auto-classifiers wet ~ wen' time, turn and bol'
stroke, blow can function as verbal action classifiers, i.e. a type of adverbial phrase
modifying the following predicate, as is illustrated in example (228) with bol' stroke,
blow and in (232) with wet ~ wen' time, turn.

(232) wen' ni ropwacian miniksuru takokno.
[wen' ni]
ACTION CLASSIFIER
{rop -wa} =ci =an
time 2 stay.under.water -FACT =LOC =FC/ID
[miniksuru] {tak -ok} =no
be.flat-haired do -COS =QUOT
When [he] had stayed under water twice, [his fur] was flat-haired, it is said.

The noun nogoltoy week is counted by partial deletion, viz. fore-clipping (233), and
can be made distributive by partial reduplication, i.e. final reduplication of the last
syllable: nogoltoy-toy (week-PARTRED) every week.

12 CLASSIFIERS

215
(233) ro'suci nogoltoysa toyni mu'ni.
[ro'su] =ci [nogoltoy sa] [toy ni] {mu' -ni}
Pname =LOC week one week.PARTIALLY.DELETED.FORM two stay -FUT
[I] will stay one or two weeks in Rongsu
12.4 Measure nouns
Measure nouns are a potentially open grammatical and semantic subclass of nouns
that can function as nouns as well as mensural classifiers. Measure nouns denote
receptacles and their volumes. The most frequently used measure nouns are listed
below.

khap a cup or the volume of whichever glass or cup is used to serve the
substance
golas a glass or the volume of whichever glass is used to serve the substance
paway a bowl to serve curry in or its volume
thali a plate or its volume
botol a bottle or its volume

The words khap cup, golas glass and botol bottle are English loans and the word
thali is a Indic borrowing (cf. Hindi (thl)). All nouns denoting pots, pans,
plates, jugs and baskets can be used to indicate a volume and are thus measure nouns.
The word thothak a drop or its volume is the only auto-classifier that can also be
used as a mensural classifier. e.g. thothak sa one drop (of dew or rain), mokren sam
thothak sa (eye medicine CLF:DROPS 1) one drop of eye medicine moktoy thothak
sa (tear CLF:DROPS 1) one teardrop.
When measure nouns are used as mensural classifiers denoting a volume, they are
preceded by a semantically compatible noun and followed by a numeral, just like
other classifiers. When measure nouns are used as nouns denoting an object, they can
be quantified themselves with the right classifier. The following example shows the
word thali plate being used as a mensural classifier to indicate a number of platefuls.

(234) uci thalibaa s'akno may.
uci [thali baa] {sa' -ak} =no [may]
then CLF:PLATEFUL 5 eat -COS =QUOT rice
Then he ate five plates of rice, it is said.

12 CLASSIFIERS

216
The next examples illustrate the use of the morpheme khap cup as a classifier in
(235) and as a noun in (236).

(235) ca khap boroy hon'bo.
[ca khap boroy] {hon'} =bo
tea CLF:CUPFUL 4 give =IMP
Give four cups of tea.

(236) khap goy' ni bay'ok.
[khap goy' ni] {bay' -ok}
cup CLF:RESIDUE 2 break -COS
[He] has broken two cups.

When the measure nouns are quantified in their function as noun, denoting an object
and not a volume, they are used with their own classifiers. For most measure nouns,
but especially for the borrowed ones, the residue classifier goy' is normally used.
There is also a special classifier for receptacles, viz. thay', that can be used for all
measure nouns, but some measure nouns use other classifiers. The basket kokta type
of basket uses a repeater classifier, e.g. kokta ta ni (type of basket CLF:KOKTAD-
two) two kokta.
12.5 The origin of classifiers in Atong
The origin of the Atong classifiers lies in both nouns and verbs. This makes it a
system of classifiers of mixed origin (see Aikhenvald 2003-b: 352-67). Some of these
nouns and verbs are still attested in the modern language and some might now have
become obsolete. The classifier for cylindrical objects and long sharp or pointy things,
-pho, comes from the still attested noun pho wooden handle for big knives, axes
and spears. Some classifiers correspond to nouns that are only attested with a
fossilised prefix. The classifier for animals, knives and tools, ma, and the one for
spoken things, mo, for example, correspond to the nouns bima body, appearance
and bimo ~ bimu name respectively, in which bi- is a fossilised prefix. It has to be
noted that the vowel variation /i ~ u/ that exists for the noun bimo ~ bimu name
does not occur in the classifier, which has only one shape, i.e. mo. Some classifiers
come from nouns that are only attested as bound morphemes in compounds. The
12 CLASSIFIERS

217
classifier for villages, dam, for instance, corresponds to the bound morpheme dam
place found in the compounds jabol-dam (garbage-place) garbage heap and
ca'wek-dam (chaff-place) place where the chaff is thrown after winnowing the rice.
Loans have also been attested among the classifiers, especially among the
mensural ones. Most loans come from English, e.g. keji kilogram, sentimitor
centimetre, peket (< English: packet) classifier for packets and layn (< English:
line) classifier for a collection of items lined up inside packets or on shelves. One
Indic loan has been detected so far, viz. the classifier for things that occur in pairs (not
body parts), jora. This classifier corresponds to the Hindi noun (jot) pair. In
its use as a noun in Atong, jora means match in love.
Some nouns have grammaticalised into both classifiers and event specifiers, e.g.
the nouns tho' half (the result of a cut across the width) and phak side, half (the
result of a cut along the length or longitudinal cut). The classifiers that are derived
from these nouns are tho' classifier for cylindrical objects and for halves of objects
cut across the width and phak classifier for parts of objects that are the result of a cut
along the length or longitudinal cut. Examples of the use of these classifiers are given
in Table 47 below. The corresponding event specifiers are tho' V in half, and
phak V by the side of something, V side by side, V for a little while, where V stands
for any semantically compatible verb.
Some classifiers correspond to verbs as well as nouns. The classifier for small
round objects etc., ro (without glottal prosody), probably derived from the noun ro'
stone (with glottal prosody) and corresponds to the verb roro- to roll, in which
we see the syllable used for the classifier reduplicated. The classifier for things that
are like a fist, mom', corresponds to the noun mom' fist and to the verb mom' to be
like a fist. The above-mentioned classifier for spoken things, mo, has a
corresponding verb mo to call somebody/something a name and noun bimu ~
bimo name.
For two classifiers no corresponding noun is attested but only a corresponding
verb. These are the classifier for small heaps of round fruits and vegetables, com', and
the mensural classifier for armfuls, khabak, which correspond to the verbs com to
stack, pile up (without glottal prosody) and khabak to embrace respectively.
The attested nouns and verbs corresponding to classifiers are represented in Table
47 below.
12 CLASSIFIERS

218
Table 47 List of classifiers
Organised according to their functional properties, and examples. Where
possible, the corresponding verb, noun or event specifier is given with the
example.
Animate Humans m classifier for humans
Animals ma classifier for animals, knives and tools
bthy ma sa one porcupine, cawky ~
cakuy ma sa one big knife
NOUN: bima body (of human or animal)
Inanimate
Plants pha classifier for trees and flowers, culms and
stalks
samsi pha sa one culm of grass
nara pha sa one orange tree
Shape and
dimension
phek classifier for branches of trees
dala phek sa a smaller but not very small
branch of a tree and not directly derived from
the trunk
ro classifier for small round objects, money, small
stones, seeds, stones in a game (when they have
a value) and fruits, default classifier for
counting for the sake of counting
nara ro sa one orange, tanka ro cygk
ten rupees
NOUN: ro' stone
VERB: roro- to roll Notice that the noun
has a glottal prosody which is absent on the
classifier and the verb.
phak classifier for parts of objects that are the result
of a cut along the length or longitudinal cut
a bucot=aw phak tham kan-ni (1s
mango=ACC CLF:LONGITUDINAL.CUTS three
cut-FUT) I will cut the mango in three pieces.
NOUN: phak side, half/slice/part which is the
result of a longitudinal cut also found in the
compounds dkm phak place where the head
is and ca+phak thigh
EVENT SRECIFIER:
-phak to VERB lengthwise, to VERB by the side
of something, to VERB side by side, to VERB for
a little while, to VERB partly

12 CLASSIFIERS

219
Table 47 continued (a)
Inanimate shape and
dimension
(continued)
tho classifier for cylindrical objects and for parts of
objects cut across the width
batri tho bery four batteries
NOUN: tho half (the result of a cut across the
width)
EVENT SPECIFIER: -tho to VERB in half
(crosswise)
kun classifier for culms
para kun sa one culm of thatch
NOUN: kun stick
khu classifier for flat things, clothes, written things
and pictures, even when the pictures appear on a
computer screen
citi khu ni two letters, lopen khu sa one
pair of trousers
NOUN: khu carapace, the shell of a crab,
tortoise etc.
mm auto-classifier for fists and classifier for things
that are like a fist (cf. Dutch gebald)
mm ni two fists
NOUN: mm fist
VERB: mm- to be like a fist, (in Dutch:
gebald zijn)
dot classifier for long cylindrical things like logs (of
wood), candles and bananas
wa dot sa one culm of bamboo, kendel dot sa
one candle, pan dot sa one log (of wood)
thut ~
thun
classifier for big spherical things, stones, bricks,
rocks, heads, hills, mountains and bars of soap
habri thut sene seven hills, mountains sabun
thut sa one bar of soap
dkm thut sa one head
rothay thut tham three stones/rocks
pho classifier for cylindrical objects and for long
sharp or pointy things
NOUN: pho wooden handle for big knives,
axes and spears
t classifier for long thin things like ropes, chains
and hair
BOUND MORHPEME IN NOUNS: pi-t (?-
string) thread ray-t (reed-string) clothes
line, puk-t (belly-string) small intestine
wa-t (bamboo-string) bamboo string

12 CLASSIFIERS

220
Table 47 continued (b)
Inanimate shape and
dimension
(continued)
khaw classifier for teeth, planks, sheets of corrugated
iron for roofs and flattened bamboos used to
make mats (jow') when they are in a mat
(damdol)
damdl khaw sa one jow' of a damdol, wa
khaw ni two teeth, two tusks, tota khaw
tham tree planks, tin khaw brry four
sheets of corrugated iron
khap classifier for flat materials
tota khap sa one plank tin kahp sa one sheet
of corrugated iron damdl khap sa one
bamboo mat used for the side of a house
kep classifier for small flat things
biskut kep sa one biscuit
ge classifier for long vegetables
rasunok ge sa one spring onion
jora classifier for things that occur in pairs
sendel jora sa one pair of sandals muthay
jora sa one pair of breasts mkren jora sa
one pair of eyes
NOUN: jora match in love (< Indic, cf. Hindi
(jot) pair)
Consistency phel classifier for baked things
barata phel sa one flat bread
biskut phel sa one biscuit
Function phat classifier for cloths
ripan phat sa one cloth
thay classifier for receptacles, e.g. jugs, boxes etc.
boyom thay sa one jug dipot thay sa one
teapot khap thay sa one cup gethe boiom
thay ni bayok. He broke two jugs.
NOUN: thay fruit
pan classifier for apparatus, appliances, mechanical
and electrical things, cars, bikes, bicycles,
mortars and umbrellas
radio pan sa one radio
satha pan sa one umbrella
gari pan sa one car thep pan sa one tape,
tibi pan sa one TV, asam pan tham three
mortars
Inanimate/
Mensural
Residue goy residue classifier e.g. (196), (224) and (236)
Inanimate Mensural by
arrangement
sat classifier for bundles
garu sat tham three bundles of mustard
leaves

12 CLASSIFIERS

221
Table 47 continued (c)
Inanimate Mensural by
arrangement
(continued)
ali classifier for small heaps or piles of
things
nara ali tham three piles of
oranges
cok classifier for bunches or small
heaps
jart cok sa one small heap of
chillies rasunok cok sa one
bundle of spring onions
thay classifier for boxes and other
receptacles
boyom thay sa one jug dipot
thay sa one teapot khap thay sa
one cup
NOUN: thay fruit
thom classifier for things in heaps or
piles jw thom sa a pile of
flattened bamboo used to make
mats
tum classifier for packets
peket classifier for packets
sigret peket sa one packet of
cigarettes
NOUN: peket (< English) packet
thep classifier for heaps and small
packets
phan classifier for food packed in
bundles in ray'cak big leaf used to
pack food
layn classifier for a collection of items
lined up inside packets or on
shelves
NOUN: layn (< English) line
Mensural by
arrangement
and shape
khasot classifier for bundles of things with
stalks
rasun khasot sa one bundle of
onions (that have stalks)
com classifier for small heaps of round
fruits and vegetables the way they
are presented at the market
nara com ni hnbone. (orange
CLF:PILES-two give=IMP-TAG) Give
two little piles of oranges.
VERB: com- to stack, to pile up, to
fuck (NB. no glottal stop in the
verbal root)

12 CLASSIFIERS

222
Table 47 continued (d)
Inanimate Mensural
by
quantity
v
o
l
u
m
e

khabak an armful
VERB: khabak- to embrace
khatom classifier for bagsful
rasunok khatom sa one bagful of spring
onions
cakwak classifier for handfuls ro cakwak citsa
eleven handfuls of stones
litr litre (English loans)
l
e
n
g
t
h
/
d
i
s
t
a
n
c
e

inci inch
pit the length of two fists and two thumbs
when one joins the thumbs at the tip while
making fists (possibly an English loan cf.
feet)
mk the length from the elbow to the top of the
middle finger
maynm length from the elbow to the top of the
fist
khuru length from the top of the thumb to the top
of the middle finger when one puts ones
hand down on the table on these points
bawa length of the widely stretched arms and
hands
capha a foot-length
NOUN: ca leg/foot, pha sole of the
foot; capha foot-sole
sentimitr centimetre (English loans)
mitr metre
kilomitr kilometre
w
e
i
g
h
t

grem gram
keji kilogram
mon weight unit of 40 kg
dora weight unit of 5 kg
s
u
r
f
a
c
e
bikha classifier for surfaces of 80 by 80 pit. (A pit
is the length of two fists and two thumbs
when one joins the thumbs at the tip while
making fists)

12 CLASSIFIERS

223
Table 47 continued (e)
Inanimate SPECIFIC tm classifier for fields
haba tm ni two dry rice and vegetable
fields
m classifier for spoken things, games and of the
word bostu thing (< Indic, cf. Hindi
(vastu) object, thing)
khata m ni two words, golpho m ni
two stories, git m ni two songs bostu
m tham three things
VERB: m- to call somebody/something a
name
NOUN: bim ~ bimu name
col classifier for ways, roads, paths and rivers
tykhal col ni two rivers, ram col tham
three roads, paths, sorok col brry four
roads
khan classifier for log boats
ru khan ni two boats
co classifier for iron nails
khiil co sa one (iron) nail
dam classifier for villages
so dam ni two villages
BOUND NOUN: -dam place found in words
like jabol-dam (garbage-place) garbage heap
and cawek-dam (chaff-place) place where
the chaff is thrown after winnowing the rice
tum classifier for places
hap tum-bysk? (place CLF:PLACES-
how.many) How many places?
tu classifier for things like bridges
dolo tu sa one bridge
sam classifier for limbs: hands, arms, legs, feet, ears
and tires
nakhal sam sa one ear, ca sam sa one
leg/foot, tayr sam ni two tires
khap flat piece of hard material like stone or metal,
classifier for flat pieces of hard material
sorekhap khap ni two pieces of mica
NOUN: flat piece of hard material like stone or
metal
the classifier for pieces of meat
masu+randay the sa (cow+meat
CLF:PIECES.OF.MEAT one) one piece of beef

12 CLASSIFIERS

224
Table 47 continued (f)
Inanimate specific
(continued)
khaw classifier for teeth
wa khaw sa one tooth
ci classifier for bamboo shoots
maywa ~ maywa ci sa one bamboo shoot
Repeater khal classifier for orifices, holes and caves
hakhal khal ni two caves, nakhukhal
khal ni two nose holes (nakhukhal ni two
nose holes is also possible)
NOUN: khal hole
su classifier for hollow cylinders
wasu su tham three bamboo cylinders
BOUND MORPHEME IN NOUN: wasu
bamboo cylinder used to cook food in
(wa bamboo, for su see the footnote 35 in
12.2.2)
ta classifier for kokta type of basket
kokta ta ni two kokta
cak classifier for leaves
pancak cak sa one leaf
NOUN: cak hand pan+cak (tree+hand) leaf

225

226
Chapter 13 Postpositions
_____________________________________________________________________

Postpositions occur after the NP and the case marking enclitic they occur with. Five
postpositions are attested in Atong, viz. daka before, konsa after, gomon
because of, about, dabat (LIMIT) since, until and thol' up to, treated separately
below in this order. Exept for gomon because of, about and dabat (LIMIT), the
other postpositions also have other functions in the grammar. Some aspects of daka
and konsa are treated in Chapter 1. thol' also exits as a verb meaning go very far.
13.1 The postposition daka
As a postposition, daka is only attested followed by the focus identifier/enclitic
<=an> (FC/ID) The complement NP is marked by the dative case enclitic, as we can
see in the example below.

(237) na'tommi nagabaaw nang'tommi pi'aydoabaaw, na'na dakaan
phetaok, na'na dakaan udo re'esawok
[[na' -tom] =mi {na} =gaba] =aw
2s -ppp =GEN need =ATTR =ACC
[na' -tom] =mi {pi' -aydo} =gaba] =aw
2s -ppp =GEN ask -PROG =ATTR =ACC
[[na' =na] daka] =an {phet -a -ok}
2s =DAT before =FC/ID arrive -AWAY -COS
[[na' =na] daka] =an [u] =do {re'e -saw -ok}
2s =DAT before =FC/ID DST=TOP leave -CERTAINLY -COS
[The curse] which you
p
needed, which you
p
were asking for, had arrived
before you [and] it has certainly left before you.

Dative-marked clauses functioning as complement of this postposition are treated in
27.3.
13 POSTPOSITIONS

227
13.2 The postposition knsa
The postposition konsa occurs with the genitive, as is illustrated by example (238)
and (239). Example (658) (in 24.5) illustrates a clause functioning as complement of
this postposition.

(238) te'ewdo pa'a bolsidarangmi konsaan [] ue way alaga sosa jalaokno.
[te'ew] =do [[pa' -a bolsi] =dara =mi knsa]
now =TOP many -CUST year =p =GEN later
[ue way] [alaga so] =sa {jal -a -ok} =no
DST spirit other village =MOB run.away -AWAY -COS =QUOT
Now, many years later, [because the village of Siju has gotten dirty,] that
spirit has run away to another village, it is said.

(239) umi konsa san khole san sa cow rokhuwanowa.
[u =mi knsa] [san khole san sa] [cow]
DST=GEN after day twenty -day one liquor
{ro -khu -wa} =no -wa
drink -INCOM -FACT =QUOT -FACT
After that [the people of Badri] drank liquor for twenty one more days, it is
said.
13.3 The postposition gmn
Examples (240) and (241) below are illustrative of the use of the postposition gomon
reason, about. Clauses of which the predicate carries the factitive suffix
<-wa> (FACT) also take the genitive when they function as complement of this
postposition, as we can see in example (657) in 24.5. The only phrasal enclitic that is
attested to occur after this postposition is the delimitative <=sa> (DLIM).

(240) otokoymo ge'thee alu kobi habijabi otokoy samcakramogomon bagan
takwano.
otokoymo [ge'the] =e [alu kobi habijabi otokoy
so.then 3s =FC potato cabbage all.sorts like.that
samcak =ra (Garo) =mo gmn] [bagan] {tak -wa} =no
vegetable =p =GEN reason garden make -FACT =QUOT
So then, because of potatoes, cabbage and all kinds of vegetables like that, he
made a garden, it is said
13 POSTPOSITIONS

228
(241) uan joramigomon co'sa golpho ka'etwa.
[u =an jora =mi gmn] [co'sa] {golpho kha' -et -wa}
DST=FC/ID love.match =GEN about a.little story do -CAUS -FACT
I have a little bit told a story about that love match.

There are two recorded occurrences of an accusative-marked postposition phrase, both
by the same speaker, represented in examples (242) and (243). A possible analysis is
that the accusative-marking means that this speaker considers gomon to be a noun and
head of the NP. In this case the accusative indicates that the NP is definite and
referential. The accusative-marked argument of the verb bal to say, tell, speak is
always the person or thing talked about, e.g. morotdara na'=aw bal-aydo (person
2s=ACC tell-FUT) People are talking about you. The lexeme gomon only occurs in
genitive constructions after other NPs and never on its own as lexical item.

(242) aa imigomonaw baletni.
[aa] [i -mi gomon] =aw {bal -et -ni}
1s PRX =GEN about =ACC tell -CAUS -FUT
I will tell about this.

(243) aa ie dakami acu ambimigomonaw baletni.
[aa] [ie daka =mi acu ambi =mi gomon] =aw
1s PRX in.the.past =GEN grandfather grandmother =GEN about =ACC
{bal -et -ni}
tell -CAUS -FUT
I will tell about the ancestors of long ago.
13.4 The limitative postposition dabat
The limitative postposition dabat (LIMIT) indicates a temporal limit and marks both
the Source, e.g. (244), and the Goal, e.g. (245). To indicate a spatial Source limit, the
complements of this postposition, e.g. time words, demonstratives and nouns, have to
be genitive-marked.

13 POSTPOSITIONS

229
(244) umi dabatsa iawe dabat mowanoro.
[u -mi dabat] =sa [i] =aw =e [dabat]
DST=GEN LIMIT =DLIM PRX =ACC =FC Pname
{mo -wa} =no =ro
call.a.name -FACT =QUOT =EMPH
Since then this [place] is called Dabat.

(245) tay'nimi dabat na'moan baju takcaka.
[tay'ni =mi dabat] [na'] =mo =an {baju tak -ca -ka}
today =GEN LIMIT 2s =COM =FC/ID friend do -NEG -INCOM
From today onward [I] will not be your friend any more.

Predicates of clauses functioning as complement of this postposition take the factitive
suffix <-wa>, e.g. (246), (247). Since the factitive-marked verb does not take any case
marking, I analyse this as a complementation strategy in which the verb and its
arguments, if any, become the complement of the postposition. This is the same
complementation strategy as the strategy that is used for the verbs like macot- to
finish and jam- to finish, which also take factitive-marked verbal complements, (see
24.3).
Complement clauses for which the predicate head is factitive-marked do not
(usually) take any case marking and therefore rely on another mechanism to
determine whether the construction should be interpreted as a temporal Source or
Goal. This mechanism is polarity. Atong can use the negative suffix <-ca> (NEG) in
non-negative contexts to indicate that an event has not been realised yet. Negated
verbs do not take the factitive suffix, which can be seen in example (246). In example
(246) the negation is put in to signal that the event denoted by the verb has not been
realised yet, and not to express negative polarity. In the corpus collected for this
grammar there are not many instances of this emphatic use of the negative morpheme
<-ca> (NEG), but enough to ensure its function in clearly positive clauses.

13 POSTPOSITIONS

230
(246) otokciba na'a angna aro amo jokna na' khewa dabat a thoyca dabat
aaw mu'ay sa'na hon'bo nookno.
otokciba [na'a] [a] =na aro [a =mo jok] =na]
but 2s 1s =DAT and 1s =GEN spouse =DAT
[[na'] {khe -wa} dabat] [[a] {thoy -ca} dabat]
2s live -FACT LIMIT 1s die -NEG LIMIT
[a] =aw {mu'} =ay {sa'} =na {hon'} =bo {no -ok} =no
1s =ACC stay =ADV eat =DAT give =IMP say -COS =QUOT
However, you keep giving me and my wife to eat as long as you live until I
die, [he] said, it is said. Alternatively in French: jusqu je ne meure.

The use of the negative morpheme <-ca> (NEG) to signal that an event has not yet
been realised is the same as the function of the ne expltif in French, as we can see
in the alternative translation of (246).
The use of the expletive negative to indicate a Goal clause is optional. Example
(247) below shows that the verb sok- to succeed, to hold out, which is the predicate
of the clausal complement of the postposition dabat, is not negative-marked and can
still only be interpreted as the Goal.

(247) otokoymo pherudo ropathiriokno. phaltha sokwa dabatdo toyno'ci ro'ci
poy'aymo wa khu'cephin'ay sakcikaydokno pheruba.
otokoymo [pheru]=do {rop -a -thiri -ok} =no
so.then fox =TOP be.in.the.water -AWAY -AGAIN -COS=QUOT
[[phaltha] {sok -wa} dabat] =do [toy no'] =ci
self hold.out -FACT LIMIT =TOP water inside =LOC
[ro'] =ci {poy'} =ay =mo
stone =LOC hold.on.to =ADV =SEQ
{wa khu'ce -phin'} =ay
tooth bite.teeth.firmly.together -FULLY =ADV
{sak =cik -aydok} =no [pheru]=ba
bear -AS.LONG.AS.YOU.CAN -PROG =QUOT fox =EMPH
So then the fox soaked in the water again, it is said. Until [he] could not hold
out any longer, [he] sat under water as long as [he] could bear [it], holding on
to a stone and biting [his] teeth firmly together, it is said.

There is one recorded instance of a factitive complement clause taking case marking
to indicate that the event is a Goal. This example, presented here below, comes from
13 POSTPOSITIONS

231
an epic story told in an unusually complicated register and translated from Garo. The
verb tak- to do is nominalised with the factitive suffix <-wa> (FACT) and on top of
that we see the locative <=ci> (LOC) and allative <=na> (ALL).

(248) thot tho'thot takwacina dabat sokromaymo khanetsigaaydono.
[[thot tho'thot] {tak -wa} =ci =na dabat]
drop last.drop do -FACT =LOC =ALL LIMIT
{sokrom} =ay =mo {khan -et -siga -aydoa} =no
hold.the.whole.body =ADV =SEQ poor -CAUS -ALT -PROG =QUOT
Until the last drop (was done) he in turn is poring [the liquor into her mouth]
holding her whole body.

As can be noted from the above examples, the postposition dabat (LIMIT) can take
the delimitative enclitic <=sa> (DLIM), as in (244), or the topic enclitic <=do>
(TOP), as in (247). It is not attested with any other enclitics.
13.5 The limitative postposition thl
The postposition thol' up to, until governs the genitive and marks a spatial
boundary, as is illustrated in example (249).

(249) ie ca'masami way khurucido, ue hoysamiaw bagoladesmi thol' koosmi
jaria ha'golsakgumukawan moani.
[ie ca'masa =mi way] {khurut} =ci =do [ue
PROX downstream =GEN spirit summon.a.spirit =LOC =TOP DST
hoy] =sa =mi =aw [bagolades =mi thl] [koos =mi
REM =MOB =GEN =ACC Pname =GEN up.to Pname =GEN
jaria] [ha'golsak] =gumuk =aw =an {moa -ni}
influence world/everything =all =ACC =FC/ID call.upon -FUT
When he summons the downstream spirit, that [priest] will call upon the
influence of all those far away [places] up till Bangladesh [and] the influence
of Kongos, all of them.

This postposition is attested as a verb with the event specifier <-a> attached to it, as
we can see in (250).
13 POSTPOSITIONS

232
(250) ga'thoaymuna thol'aok.
{ga'tho =ay =muna} {thol' -a -ok.}
kick =ADV =SEQ go.very.far -AWAY -COS
Because I kicked [it], [it] went very far.

233

234
Chapter 14 Time words
_____________________________________________________________________

Time words are half way on the cline between nouns and adverbs and share properties
with both. The properties of the time words are summed up in section 14.1. Most time
words are deictic, but daka before, in the past, earlier and konsa later, after can
be deictic or relative depending on the context. The time word daka is found in many
more different syntactic environments than the other time words and is treated
separately in section 14.2. The words daka and konsa can both function as
postpositions. This function is treated in sections 13.1 and 13.2 respectively.
Time words are a closed class, the fifteen members of which are listed here below,
illustrated by examples.

Table 48 List of time words

Deictic
maya the day before yesterday or longer ago
moja yesterday
tay'sa a moment ago, just now, a little while ago
tay'ni today
te'ew now
te'en later but still today
hampoy later today, in the evening
hanep tomorrow
ceknay the day after tomorrow
hambun later but not today, in the far future
teraka last year
taray this year
nayja next year
Deictic/Relative
daka before, in the past, earlier
konsa after, later
14.1 The Properties of time words
i Clausal properties
Time words cannot be the head of a predicate and occur canonically as adjunct to a
clause indicating Temporal Location.
14 TIME WORDS

235
ii Phrasal properties
Time words
can be the head of a temporal phrase indicating Temporal Location,
can modify nouns,
can possess a noun but cannot be possessed.
The following example illustrates the time word moya yesterday modifying a verb
which is the predicate of an attributive clause in a headless arch NP (see Chapter 29),
which functions as the nominal head of the predicate of an identity/equation clause.

(251) ie moya kha'gasega
--------arch NP--------
[ie]
S
{ [ [moya] {kha'} =ga] =sega}
PRX yesterday make =ATTR =ALT
These [are] now the ones made yesterday. (Said about some pictures that
were shown after seeing some other pictures that were made last year)

TEXT 1 line 28, presented here as (252), presents an occurrence of the genitive-
marked time word te'ew now, functioning as Possessor and thus modifying the
following noun gawi girl.

(252) Songken says: otokaria, te'ewrawrawmi gawido.
{otok -ari -a} [te'ew -rawraw =mi gawi] =do
do.like.that-SIMP -IMPF now -CONTINUOUSLY =GEN girl =TOP
Yes. They do like that, the girls from now on. (i.e. todays girls)
iii Morphological properties
Time words
occur with a limited set of case markers corresponding to the limited
types of peripheral argument that time words can be. Note that, as
Temporal Location adjunct, a time word cannot take the locative case
enclitic <=ci> (LOC), the reason being that time words are inherently
temporal locational. Only when a time word functions as a Goal
adjunct can it be locative-plus-dative-marked, e.g. (254).
14 TIME WORDS

236
have been recorded with a limited set of other enclitics and suffixes,
viz. <=do> (TOP) and <=e> (FC), <=an> (FC/ID), <=ba> (EMPH/ADD),
<=sa> (DLIM), which all function as phrasal enclitics, and <-mama>
just, only and <-rawraw> (continuously), which also function as
event specifiers on verbal predicate heads, e.g. (252).
can take the plural morpheme <=dora> (p) meaning approximately
TIME WORD, e.g. (253).
cannot be counted.
cannot occur with classifiers.

The example here below illustrates a pluralised time word. A plural-marked time
word expresses an approximate time reference.

(253) ge'the tay'nidora ray'anikhon
[ge'the] [tayni] =dora {ray'a -ni} =khon
3s today =p come -FUT =SPEC
He might come today or so.

The next example shows the time word te'ew now used as a Goal and hence marked
with the locative and dative enclitics.

(254) ue ha'boriawe te'ewcinakhokho ato khu'cuksa matsa caw'koy asetram
mowano.
[ue ha'bori] =aw =e [teew] =ci =na [khokho]
DST hill =ACC =FC now =LOC =ALL still
[ato khu'cuk] =sa
Atong language =INSTR
[matsa caw'koy asset ram] {mo -wa} =no
tiger big.knife throw.away place call.a.name -FACT =QUOT
[We] call that hill up till now still matsa caw'koy asetram in the Atong
language, it is said.

In the next example the time word tay'sa a little while ago appears as a Facsimile
adjunct marked with the similative enclitic <=tokoy> (LIKE).

14 TIME WORDS

237
(255) uan tay'satokoy kantaraaw korokorok re'eaymo so'etthiriokno.
[u] =an [taysa] =tokoy [kantara] =aw [korok-korok]
DST =FC/ID a.little.wile.ago =LIKE emptiness =ACC quickly
{re'e} =ay =mo {so' -et -thiri -ok} =no
go.away =ADV =SEQ ASK -CAUS -AGAIN -COS =QUOT
Just like a little while earlier he quickly went to the emptiness and asked
again.
iv Semantic properties
Time words express a location in time. The next example illustrates the use of the
deictic/relative time word konsa later, after.

(256) a torucena bay'siga, na'a konsa torubone noaydoano magacakan.
[a] {toru -ce =na} [bay'siga] [na'a] [knsa]
1s take.a.bath -FIRST =DESI friend 2s later/after
{toru} =bo =ne {no -aydoa} =no [magacak] =an
take.a.bath =IMP =TAG say -PROG =QUOT deer =FC/ID
I want to take a bath first, friend. You take a bath later, OK? the deer is
saying, it is said.

When nouns indicating a period of time, i.e. bolsi year, ja month, nogoltoy week,
san day, manap morning and gasam evening, afternoon are preposed to and thus
modifying deictic time words in asyndetic coordination, the resulting new NP has the
properties of a time word, which is the head, and cannot take the locative case marker
any more. Examples of these compounds are tay'ni gasam this
evening/afternoon/later part of the day, moya manap yesterday morning, hambun
nogyltoy a week in the far future, (257).

(257) tay'nigasam ni noksa koy' sa'na re'ebone.
[tay'ni gasam] [ni nok] =sa [koy'] {sa'} =na {re'e} =bo =ne
today evening 1p house =MOB dog eat =DAT go.away =IMP =TAG
Go to our house this evening to eat dog, OK?
14.2 The word daka
The word daka before, in the past, earlier is attested in six syntactic environments,
viz.
14 TIME WORDS

238
1. as time word in topic function,
2. as a genitive-marked Possessor modifying a following NP,
3. with the attributive suffix <-gaba ~ -ba> (ATTR) functioning as NP modifier,
4. with the adverbialising suffix <-gaba ~ -ga> (ADV) functioning as clausal adverb,
5. as underived adverb modifying a predicate,
6. as a postposition governing the dative: see 13.1.

Examples of all of these different occurrences, except as postposition, will be given
below.
14.2.1 As time word
Whether = daka before, in the past, earlier as a time word indicates deictic or
relative time depends on the context. Example (258) illustrates the use of daka
before, earlier, in the past, as a deictic time word, functioning as a pre-clausal topic.

(258) dakado, mamu khem ni'wacido domcorasasa cowgon rowano.
[daka] =do [mamu khem] {ni' -wa} =ci =do
in.the.past =TOP nothing drum not.exist -FACT =LOC =TOP
[domcora] =sa =sa] [cowgon] {ro -wa} =no
snare.instrument =INSTR =DLIM festival.of.the.dead drink -FACT =QUOT
As for the past, when there were no drums, [they] celebrated the festival of
the dead only with the domcora, it is said.

A deictic or relative temporal interpretation is possible for the clause in which daka
occurs in the next example.

(259) a ie khata dakado tocacom, te'ewdo nemay took.
[a] [ie khata] [daka] =do {to -ca} =com
1s PRX word before =TOP know -NEG =IRR
[te'ew} =do {nem} =ay {to -ok}
now =TOP good =ADV know -COS
I did not know this word before/in the past, but this is not the case any more;
as for now [I] know [it] well. (Literally: I have come to know it well)
14 TIME WORDS

239
14.2.2 As a genitive-marked Possessor
The next example illustrates the use of daka as a Possessor, modifying the following
NP caso generation. This ability to occur as Possessor is a property that daka
shares with the other time words.

(260) uan gam moa, dakami casodo.
[u] =an [gam] {mo -a} [daka =mi caso] =do
DST=FC/ID wealth call.a.name -CUST past =GEN generation =TOP
That was called wealth, as for the generation(s) of the past.
14.2.3 With the attributive suffix <-gaba ~ -ga>
The lexeme daka can take the attributive suffix <-gaba ~ -ga> (ATTR). The
resulting form, daka-gaba (before-ATTR) means the first, and can thus function
attributively to nouns, as we can see in (261). The meaning of dakagaba first is
distinct from that of daka=mi (past=GEN) (see 14.2.2 above). The different
functions of the attributive morpheme <gaba ~ ga> (ATTR) are discussed extensively
in Chapter 29.

(261) uci thomaymo ge'thee [] dakagaba bobaan diriceokno.
[u] =ci {thom =ay =mo [ge'the] =e
DST=LOC lay.in.ambush =ADV =SEQ 3s =FC
[daka =gaba boba =an] {diri -ce -ok} =no
before =ATTR crazy.man =FC/ID hold -FIRST -COS =QUOT
[Having lain in ambush,] the first crazy person got hold of [the horses tail]
first, it is said.
14.2.4 With the adverbialising suffix <-gaba ~ -ga>
The suffix <-gaba ~ -ga> (ADV) can turn daka into the clausal adverb dakagaba
first of all, the first time, as we can see in (262) and (263).

(262) dakagabado jineral miticeni.
[daka =gaba] =do {jineral miti -ce -ni}
before =ADV =TOP general -hold.a.meeting -FIRST -FUT
First of all a general meeting will be held.

14 TIME WORDS

240
(263) dakagaba turaci mu'waci mobinaw gorongwa
[daka =gaba] [tura] =ci {mu' -wa} =ci [mobin]=aw {goro -wa}
before =ADV Pname =LOC stay -FACT =LOC Pname =ACC meet -FACT
The first time [I] stayed in Tura, [I] met Mobbin.
14.2.5 As underived adverb
There is one recorded instance of daka being used as an underived adverb, meaning
previously, modifying the following predicate mo to call something/somebody a
name. This instance is shown in (264). The fragment in this example consists of the
village name soma soni khocu Badri which is modified by two attributive clauses
(AC) (see Chapter 29), one pre- and one post-head.

(264) [] daka mogaba soma soni khocu badri nogaba []
----------------------------------------arch NP--------------------------------------
--------AC-------- --AC--
[ | [daka] {mo}| =gaba soma soni khocu badri |{no}| =gaba]
before call.a.name =ATTR Pname say =ATTR
the previously so called Songma Songni Khychu Badri

241

242
Chapter 15 Adverbs
_____________________________________________________________________

The class of adverbs is open since there are productive processes to derive adjectives
from nouns, verbs and adjectives as will be discussed in Chapter 18. There are some
non-derived or opaquely, unproductively derived adverbs which I will call the
primary adverbs, which might form a closed subclass, although it is likely that not all
members have been recorded yet.
Adverbs are morphologically quite invariable, incapable of taking any suffixes or
phrasal enclitics except for some who have been recorded with the delimitative <=sa>
(DLIM) and emphatic <=ba> (ADD/EMPH). Some adverbs can be reduplicated for
intensification (indicated in list below). Although some adverbs seem to be
reduplicated forms, those have a different meaning from what appears to be the
simplex form, i.e. bok suddenly and bokbok quickly. Some adverbs look like
reduplicated forms with vowel alternation, e.g. domdom damdam carelessly,
disorderly. Other adverbs show partial reduplication, e.g. koken naken zigzag.
Adverbs behave differently from adjectives in that they only modify adjectives,
verbs and clauses, but not nouns. The adverb can be separated by other constituents
from the predicate that it modifies. An adverb always modifies something that follows
it, not something that precedes it in the clause. Adverbs cannot function as head of a
predicate, cannot take case marking and cannot express negation, aspect, modality or
any other verbal or nominal category and cannot be nominalised. There are three
adverbs that only modify adjectives (of both types) when they are head of a predicate,
viz. the intensifiers nemen very, boloen very and iskon so much, to this extent.
Table 49 below presents a list of the adverbs and intensifiers that is by no means
exhaustive. Here are some examples of the use of adverbs.
Example (265) illustrates the use of the adverb bok suddenly. We cannot be
certain whether the adverb has scope only over the immediately following predicate
of the adverbial clause, ho to jump, or over both the main and the adverbial
predicate.

15 ADVERBS

243
Table 49 List of adverbs and intensifiers
thaguduk suddenlycaca exactly
phanan ever, never (can be reduplicated)
isokon that much
co'mot ~ co'mot really
ondon simply, in vain, for free
sokatha carelessly, disorderly
kepleplep stretched out flat on your belly
bok suddenly
biciba never
biciba biciba sometimes
bokbok quickly
wetwet quickly
wetancian every time
capcap close together (as in a crowd)
romrom rolling down
jorom jorom quietly
soraksorak exactly, precisely
dokdok for a short while
korok korok swiftly
domdom damdam carelessly, disorderly
koken naken zigzag
alamola somewhat, a little
poltawtaw jerkingly over a rough road (can be reduplicated)
powtawtaw jerkingly over a rough road (can be reduplicated)
sirimonmon at the crack of dawn
biborokhon ~
bibokhoron some day
hawtoy for some time
otokoy like this/that
biciba sometimes
biciba biciba sometimes, seldom
gasam gasam seldom
gisep gisep ~
gysep gysep from time to time
jenethene somehow
jebado anyway, however it may bejekhay for example
jetokoy somehow
khasinsin slowly
phas first (<English: first)
las last (<English: last)
manapmi very early in the morning

Intensifiers
nemen very
boloen very
iskon so much, to this extent
15 ADVERBS

244
(265) matsado uci bok hoay jalaokno.
[matsa] =do [u] =ci bk {ho} =ay {jal -a -ok} =no
tiger =TOP DST=LOC suddenly jump =ADV run.away -AWAY -COS =QUOT
Then the tiger suddenly jumped [and] run away, it is said.

In the next example we see the use of the clausal adverb jekhay for example,
modifying the whole following locative- and topic-marked subordinate clause, of
which the predicate is bal to speak.

(266) jekhay atosa balcido song pidan do'acom.
[jekhay] [ato] =sa {bal} =ci =do [so -pidan]
for.example Atong =INSTR speak =LOC =TOP village new
{do' -a} =com
IE.be -CUST =IRR
For example when [you] speak in Atong [you] supposedly say so pidan.

The adverb phanan can be translated as always when the predicate is not negative,
e.g. (267) and as never when the predicate is negative, e.g. (268). In (267), the
adverb is separated from the predicate or predicates (scope uncertain) it modifies by
the NP moma elephant.

(267) taw'reksorup masa ge'themo thup phanan moma phay'ay sa'rowana
moma mathayaw tapna re'eaydoanowa.
[taw'reksorup ma sa] [ge'the =mo thup]
O
[phanan]
type.of.bird CLF:ANIMALS one 3s =GEN nest always
[moma]
A
{phay'} =ay {sa' -ro -wa} =na
elephant break =ADV eat -usually -FACT =DAT
[moma mathay] =aw {tap =na}
elephant bachelor.elephant =ACC hit =DAT
{re'e -aydoa} =no -wa
go.away -PROG =QUOT -FACT
As for the banana bird, because his nest was always brokenly eaten by an
elephant, [it] went to hit the lonely elephant, it is said.

(268) panaan hon'rocane udo rajado
[panan] {hon' -ro -ca} =ne [u] =do [raja] =do
never give -USUALLY -NEG =TAG DST=TOP king =TOP
15 ADVERBS

245
[He] usually never gives [the drum], that king.

The following three examples illustrate the use of the intensifiers. Both are
interchangeable in all circumstances. In (269) and (270) we see the intensifiers
modifying a Type 2 adjective functioning as predicate head. In (271) the intensifier
modifies a predicate of which the head is a stative verb.

(269) ue waye boloen cuano.
[ue way] =e [boloen] {cu -a} =no
DST spirit =FC very big -CUST =QUOT
That spirit was very big, it is said.

(270) te'ew wensa ropay na'do nemen solnaka noaydoano pherue.
[te'ew] [wen sa] {rop} =ay [na'] =do [nemen] {sol -naka}
now time one soak =ADV 2s =TOP very beautiful -IFT
{no -aydoa} =no [pheru]=e
say -PROG =QUOT fox =FC
Now [you] soak once more [and] you will certainly be very beautiful, the
fox is saying, it is said.

(271) uci mu'butung somayci badri nemen man'ay sa'ano.
[u] =ci {mu' -butu +somay} =ci
DST=LOC live -WHILE +time =LOC
[badri] [nemen] {man'} =ay {sa'} -a =no
During the time [they] lived there, Badri was very rich (Lit. ate in great
amounts), it is said.

The adverbs phas first and las last, borrowed from English via Indic, are the only
adverbs that can be attributivised with the attributive clausal enclitic <=gaba ~ =ga>
(ATTR), as we can see in example (272) below.

(272) aa pasgaba, ge'the lasgaba
[aa] {pas} =gaba [ge'the] {las} =gaba}
1s be.first =ATTR 3s be.last =ATTR
Im the first [sibling], hes the last [sibling].

246
Chapter 16 Discourse connectives
_____________________________________________________________________

Discourse connectives indicate the semantic relationship between stretches of text.
Syntactically a discourse connective comes either at the beginning of a clause or at the
end of one. A discourse connective can belong to either the preceding clause or the
following depending on the prosody, i.e. whether there is a pause before or after the
discourse connective. If there is no pause at all, it is impossible to say to which clause
the discourse connective belongs. Sometimes, too, there is a pause before as well as
after the discourse connective and then it stands alone, syntactically and prosodically.
There are two types of discourse connective depending on their morphological
make up, which correlates with their usage. Both types will be treated separately
below. Both types of discourse connective are grammaticalisations of forms found
elsewhere in the language. The members of the two types of discourse connective are
listed in Table 50 below.
16.1 Type 1 discourse connectives
Discourse connectives of Type 1 are the result of grammaticalisations of various
forms of the verb otok-
36
to do like this/that. There are two groups within Type 1
depending on their morphological make up. Group A consists of the connectives
otokoysa therefore, thats why and otokoymo ~ otokoymu ~ otokoymuna ~ otokoymu
~ otkoymuna so then, having done that/this. These connectives consist of the verbal
root otok- to do like this/that, followed by an altered form of the adverbial enclitic,
viz. <=oy> instead of the normal form <=ay> (ADV), followed by

36
The verb otok- do.like.this/that itself seems to stand in a vowel alternation relationship with the
interrogative verb atak- to do what?. The etymological connection between these two verbs deserves
further study which lies outside the scope of this grammar.
16 DISCOURSE CONNECTIVES

247
Table 50 List of discourse connectives and their historical make up
Type 1
MOPHOLOGICAL MAKE
UP
FORMS GLOSS CONNECTIVE
MEANING
otok-
to.do.like.this/that +
<-oy> + <=ay> (ADV) +
<=mo ~ =mu ~ =mu
~ =muna ~ =muna.>
(SEQ)
otokoymo ~
otokoymu ~
otokoymuna ~
otokoymu ~
otkoymuna
G
r
o
u
p

A

so then, having
done that/this
sequential,
pause filler
otok-
<-oy> < <=ay> (ADV) +
<=mu> (SEQ) + <=an>
(FC/ID)
otokoymuan so then, having
done that/this
sequential,
pause filler
otok-
<oy> < <=ay> (ADV) +
<=sa> (DLIM)
otokoysa therefore,
thats why,
then
reason,
sequential
otok-
<=ci> (LOC) + <=do>
(TOP)
otokcido
G
r
o
u
p

B

in that case condition
otok-
<=ci> (LOC) + <=ba>
(INDEF)
otokciba but contrastive
otok-
ma' well then, ok +
<=ci> (LOC) + <=ba>
(INDEF)
otokma'ciba but contrastive
Type 2
MOPHOLOGICAL MAKE
UP
FORM GLOSS
u (DST), <=ci> (LOC),
<=e> (FC), <=an>
(FC/ID)
uci ~ ucie ~ ucian then sequential
u (DST) + <=ci> (LOC)
+ <=ba> (INDEF)
uciba but then, but contrastive/
sequential,
contrastive
u (DST) + <=na> (DAT) una then, therefore,
because of that
sequential,
reason
u (DST) + <=mi ~
=mo> (GEN) + gomon
(REASON) + <=ci>
(LOC)
umigomonci ~
umogomonci
for that reason,
therefore, thats
why
reason


248
the delimitative enclitic <=sa> (DLIM) or the sequential enclitic <=mo ~ =mu ~
=mu ~ =muna ~ =muna> (SEQ) respectively.
The verb otok- to do like this/like that still exists in Atong (see section 4.5.1) and
its productive adverbial form is otok-ay (do.like.this/that=ADV) doing like this/that.
In Atong as it is today, the form otokoy cannot function as discourse connective but
only as the adverb like this/that.
37

Group B consists of connectives that have a locative enclitic attached directly to
the root of the verb otok- to do like this/that. One exception in this group is the form
otokma'ciba but, which we will discuss below.
Example (273) here below illustrates the use of the verb otok- to do like this/that.
After this example we will look at all the Type 1 discourse connectives separately.

(273) na'ba te'ewsa ana otokaydo, te'en bisa jokay jalna?
[nan'] =ba [te'ew] =sa [a] =na {tk -aydo} [te'en] [bi =sa
2s =EMPH now =DLIM 1s =DAT do.like.this -PROG later QF =MOB
{jok} =ay {jal -na}
escape =ADV run.away -DESI
You are doing like this to me now, where are you intending to run to and
escape later?
16.1.1 The origin of Type 1 discourse connectives
All Type 1 discourse connectives were historically most probably verbal tail-head
linkage devices. As has been said above, they are all grammaticalised forms of the
verb otok- to do like this/that and were once used anaphorically as non-finite verbal
forms referring to the event in the preceding clause. The form otokoymo comes most
likely from otok-ay=mo (do.like.this/that=ADV=SEQ) having done like this/that.
This form will have participated in tail-head linkage of a sequential kind, while the
form otok=ay=sa (do.like.this/that=ADV=DLIM) was most probably involved in

37
Chapter 20 on case marking treats further grammaticalisation of the verb otok- to do like this/that
into the perlative and similative suffixes. The lexeme otokoysa is homophonous with the delimitative-
marked form of the adverb otokoy like this/that, e.g. (279).

249
simultaneous tail-head linkage before it grammaticalised into the modern day reason
discourse connective otokoysa therefore, thats why. The forms in Group B of Table
50 where evidently involved in tail-head linkage of a temporal locative nature.
In the present synchronic state of the language, discourse connectives do not refer
to the event in the previous clause but only indicate a certain abstract relationship
between sentences and paragraphs.
16.1.2 tkym and its allomorphs
The discourse connective otokoymo ~ otokoymu ~ otokoymuna ~ otokoymu ~
otkoymuna ~ otokoymuan so then has a series of allomorphs which are all in free
variation although different forms are more frequently used in certain dialects. In
Badri the allomorph otokoymo and otokoymu are most popular while in Siju the
forms otokoymu, otokoymu and otokoymuna are most frequently used. Only one
instance of the use of the allomorph otokoymuan has been recorded, which was in
Siju.
The connective otokoymo so then and its allomorphs is the most frequently used
connective in Atong. This connective usually signals that the speaker has not finished
talking yet but that more is yet to come in the discourse or narrative. It is also used as
pause filler, giving the speaker time to think what he will say next.
Example (274), from a story told in the Badri dialect, illustrates the use of the
discourse connective so then in the allomorph otokoymo.


250
(274) te'ewe alsia rajano so dam saci. [PAUSE] otokoymo alsia raja so dam saci
noaysa [PAUSE] kam kha'na nobo haratanoaro ue, alsiae. [PAUSE]
otokoymo jokba mo'ni khomanoro. [PAUSE] jok mo' ni khomano. [PAUSE]
otokoymo sa'naba jok payna naano, jowna jok payna naano.
[te'ew] =e {alsia raja} =no [so dam sa] =ci
now =FC lazy.person king =QUOT village -CLF:VILLAGES one =LOC
tkym [alsia raja] [so dam sa] =ci
so.then lazy.person king village CLF:VILLAGES one =LOC
{no} =ay =sa
say =ADV =DLIM
{kam kha'} =na {harat -a} =no =aro [ue alsia] =e
work =do =DAT be.reluctant -CUST =QUOT =EMPH DST -lazy.person =FC
tkym [jok] =ba [mo' ni] {khom -a} =no =ro
so.then spouse =EMPH CLF:HUMANS two marry -CUST =QUOT =EMPH
[jok] =ba [mo' ni] {khom -a} =no
spouse =EMPH CLF:HUMANS two marry -CUST =QUOT
tkym {sa'} =na =ba [jok] {pay} =na {na -a} =no
so.then eat =DAT =ADD spouse carry.by.hand =DAT need -CUST =QUOT
{jow} =na |[jok] {pay} =na {na -a} =no
sleep =DAT spouse carry.by.hand =DAT need -CUST =QUOT
Now, [there is] a lazy king, it is said, in a [certain] village. So then, a lazy
king in a [certain] village, [I]m saying, right. [He] is reluctant to do work, it is
said, that one, the lazy person. So then, [he] is married to two wives, it is said.
[He] is married to two wives, it is said. So then, [his] wives have to carry [the
lazy king] in order to eat, it is said, and [his] wives have to carry [him] in order
to sleep, it is said.

As we can see in example (274) above, the first clause is a presentative clause with
only a nominal predicate head alsia rajano and a postposed locative adjunct so dam
sa ci. The discourse connective following that clause does not refer to any event in
that clause but functions as a pause filler. The second occurrence of this connective is
also as pause filler The second occurrence of the connective is between the third
clause kam kha'na nobo haratanoaro ue, alsiae, which is the end of a sentence, and
the beginning of the next sentence in which more information about the king is given,
i.e. that he is married to two wives. The narrator is not warmed up yet and needs a lot
of time to think about what to say next, so he repeats the previous sentence before he
throws in the discourse connective again and continues with more information about
the living conditions of the king. This example is typical of the use of the discourse
connective otokoymo so then and its allomorphs.

251
The discourse connective otokoymo ~ otokoymu ~ otokoymuna ~ otokoymu ~
otkoymuna so then can be used in tail-head linkage constructions. Tail-head linkage
in Atong is done by repeating the whole or part of the last clause with the predicate in
a sequential form. Example (275) is illustrative of a tail-head linkage construction. In
this example we see two tail-head links, both of which repeat the main verb of the
previous clause in sequential form, i.e. thorokaaymo having jumped in and
ropaymo having stayed in the water.

(275) magacakdo biskutaw toysamci tanaymo caw thorokaokno. thorokaaymo
hawtoy ropokno magacake. bewal ropaymo phetaakno.
[magacak] =do [biskut] =aw [toysam] =ci {tan}=ay =mo
deer =TOP biscuit =ACC river.bank =LOC put =ADV =SEQ
[caw] {thorok -a -ok} =no
TAIL
interj:SPLASH jump.down -AWAY -COS =QUOT
{thorok -a} =ay =m
HEAD

jump.down -AWAY =ADV =SEQ
[hawtoy] {rop -ok} =no
TAIL
[magacak] =e
for.some.time stay.under.water -COS =QUOT deer =FC
[bewal] {rop} =ay =m
HEAD

for.some.time stay.under.water =ADV =SEQ
{phet -a -ak} =no
arrive -TOWARDS -COS =QUOT
The deer, having put the biscuits on the river bank, splash! jumped in, it is
said. Having jumped in, he stayed under water for some time, it is said.
Having stayed under water for some time, he emerged, it is said.

Example (276) illustrates how the discourse connective otokoymo (or one of its
allomorphs) can occur before the head in a tail-head link.


252
(276) otom mo' korokan ha' kamarokno. otokoymuna kamaymuna konsado
jow'gaba noksa ray'aakno.
[otom mo' korok] =an {ha' kam -arok} =no
TAIL
3p CLF:HUMANS six =FC/ID soil work -PROG =QUOT
------------HEAD----------
tkymna {kam} =ay =muna [konsa] =do
so.then work =ADV =SEQ after =TOP
[jow'] =gaba [nok] =sa {ray'a -ak} =no
mother =DREL house =MOB come -COS =QUOT
The six of them worked [weeding] the land, it is said. So then, after working
the land [their] mother came home, it is said.
16.1.3 tkysa
The discourse connective otokoysa can be used in contexts where it has a sequential
meaning and in other contexts where it has to be interpreted as indicating a reason
relationship between sentences or stretches of discourse. The difference in meaning
between otokoysa then as sequential connective and the sequential connective
otokoymo ~ otokoymu ~ otokoymu ~ otokoymuna ~ otokoymuna. is probably very
subtle and they are often used in contexts, which seem completely similar to non-
native speakers of Atong. The connective otokoysa differs from otokoymo and its
allomorphs in that the former does not occur in tail-head linkage constructions and the
latter does.
Although the connective otokoysa could be analysed historically as the verbal root
otok- to do like this/that and the delimitative enclitic <=sa> (DLIM), its meaning is
context dependent and cannot be construed on the basis of the root and the suffix or
enclitics. Therefore I treat this connective as unanalysable in Atong as it is spoken
today.
The following example illustrates the use of the sequential meaning of the
discourse connective otokoysa then. It is the beginning of a description about how to
cook food in a bamboo cylinder.


253
(277) wa'da wa'ru tan'aysa, pay'ak. pay'aak hoyawe somsa rewetsaaw.
otokoysa nokcina do'waci su'bolokna tan'a.
[wa'da wa'ru] {tan'} =ay =sa {pay' -ak}
type.of.bamboo young bamboo cut =ADV =DLIM carry.on.body -COS
{pay' -a -ak} [hoyawe somsa rewet] =sa =aw
carry.on.body -TOWARDS -COS yonder River.name river.bank =MOB =ACC
tkysa [nok] =ci =na {do' -wa} =ci {su' -bolok} =na
then house =LOC =DAT arrive -FACT =LOC pound -INTO.PULP =DAT
{tan' -a}
cut -CUST
Having cut young wa'da, [you] carry it. You have carry it towards [home]
from the river bank of the Symsang way over there. Then, when [you] have
arrived home, [you] cut it so that [you] can pound [the food inside] to pulp.

The next example shows the use of the discourse connective otokoysa in a context in
which it indicates a reason link.

(278) ue toygat rowanasa ue toykhalawe rodo mowano. otokoysa ie hapawe
rodo toykhalci mu'wanasa rodo ha'way noay.
|[ue toygat] {ro -wa}| =na =sa [toykhal] =aw =e [rodo]
DST water.place drink -FACT =DAT =DLIM river =ACC =FC Rname
{mo -wa} =no
tkysa [ie hap] =aw =e |[rodo toykhal] =ci
therefore PRX place =ACC =FC Rname river =LOC
{mu' -wa} =na =sa}|
stay -FACT =DAT =DLIM
[rodo ha'way] {no} =ay
Rname plain say =ADV
Because [the Rongdyng clan] drank from that water place, [they] called the
river Rongdyng, it is said. Therefore, because [they] stayed at the Rongdyng
river, they sayingly [called the village and the area] Rongdyng Plain.

In the following example we see the use of the homophonous adverb otoykoysa like
this/that. This example forms the end of the description about how to cook food in a
bamboo cylinder, of which we have just read the beginning in example (277) above.
An adverb always immediately precedes the predicate it modifies. and thus differs in

254
position from a discourse connective, which is always the first or last element in a
sentence.

(279) atodo otokoy -sa bereay sa'a.
[ato] =do [tky] =sa {bere} =ay {sa' -a}
Atong =TOP like.that =DLIM cook.food.in.bamboo.cylinder =ADV eat -CUST
Like this the Atong eat food cooked in a bamboo cylinder.
16.1.4 tkcido
The discourse connective otokcido in that case seems morphologically the most
transparent of the Type 1 discourse connectives, viz. (do.like.that =LOC=TOP) if do
like that. This discourse connective can be postponed to the predicate of a main
clause and still link that clause or sentence to the preceding stretch of discourse. This
is illustrated by example (280). Other discourse connectives always occur in between
the clauses or sentences they link.
The context of example (280) is as follows. The lazy king wants to convince the
barber to come with him to the jungle, not adding that this is to support the lazy king
in his fight against the tiger. The barber says he is busy, but the lazy king says that he
should come since there will be many wild animals to be seen just like in a zoo. Then
the barber says (280).

(280) ay! cayna nani otokcido.
[ay] |{cay}| =na {na -ni} tkcido
interj watch =DAT need -FUT in.that.case
Ay! [I] will have to watch [that], in that case.

An alternative analysis of otokcido in (280) above is as adverb with anaphoric
reference.
The following example shows the most frequent occurrence of the connective
otokcido before the sentence which it links to the preceding discourse. The context is
as follows. The most powerful god has explained the mission to Bandi. The god has
told Bandi about the dangers of the road which he has to take to the person he is
supposed to meet. This is a lengthy stretch of discourse. The explanation ends with

255
the words If you meet enemies, dont go slowly. Defend yourself. he said, it is
said. Then Bandi answers (281).

(281) otokcido aa a re'ecie aawe tosawnima?
otokcido [aa] [a] {re'e} =ci =e [a] =aw =e
in.that.case 1s 1s go.away =LOC =FC 1s =ACC =FC
{to -saw -ni} =ma
know -CERTAINLY -FUT =Q
In that case, as for me, if I go, will [he] certainly recognise me?

There is a third alternative for the placement of otokcido in that case. It can be
preceded by a pronoun in address term function and/or an interjection, as is the case in
the following example, where otokcido links sentences 1 and 2 (delimited by big
square brackets) but is preceded by an interjection and a pronoun in Vocative
function.

(282) nia jetokoyba takan coyni aro uaw kawna re'earini nido acu nookno.
de acudora, otokcido re'eancoybo.
[[nia] [jetokoyba] {tak -an -coy -ni} aro [u] =aw {kaw}| =na
1pe somehow do -REF -try -FUT and DST=ACC shoot =DAT
{re'e -ari -ni} [ni] =do [acu] {no -ok} =no]
SENTENCE 1

go.away -SIMP -FUT 1pe =TOP grandfather say -COS =QUOT
[de] [acu] =dora tkcido [{re'e -an -coy} =bo]
SENTENCE 2
interj grandson =p in.that.case go.away -REF -try =IMP
We will try to do [it] somehow and we will just try to go in order to shoot
that [eagle], grandpa, [they] said, it is said. Very well, grandchildren, in that
case try to go.
16.1.5 tkciba and tkmaciba
The locative-plus-indefinite combination of enclitics occurs on non-main clause
predicates and indicates the notion whenever event X is the case (see 27.5). It is
therefore not clear where the contrastive sense comes from in the discourse
connective otokciba and otokma'ciba but. Historically they both come from the verb
otok- to do like this/that, and can be analysed as otok=ci=ba (do.like.this/that
=LOC=INDEF) and otok=ma'=ci=ba (do.like.this/that=interj=LOC=INDEF) and

256
that can be translated as whenever do like this/that. When the verbal form otokciba
developed into the discourse connective it became opaque since the meaning cannot
be deduced any more from the sum of the morphemes still visible in the word.
Example (283) here below illustrates the contrastive meaning of this discourse
connective.

(283) ana mamoawan nacawa, otokciba na'a ana aro amo jokna na'
khewa dabat a thoyca dabat aaw mu'ay sa'na hon'bo nookno.
[aa] [mamu] =aw =an {na -ca -wa}
1s nothing =ACC =FC/ID need -NEG -FACT
tkciba [na'a] [a] =na aro [a =mo jok] =na
[a] =aw {mu'} =ay {sa'} =na {hon'} =bo {no -ok} =no
1s =ACC stay =ADV eat =DAT give =IMP say -COS =QUOT
I dont need anything. However, you keep giving me and my wife to eat as
long as you live until I die, he said, it is said.

As mentioned above and discussed in Chapter 27, the combination <=ci=ba>
(LOC=INDEF) on predicate heads indicates an indefinite location in time. Example
(284) here below is illustrative.

(284) je takay pataciba ru botroreaariano.
[je] {tak -ay} {pat -a =ci =ba} [ru]
whatever do =ADV cross -AWAY =LOC =INDEF boat
{bot -rore -a ari -a =no
drive -SPIN -AWAY -SIMP -CUST =QUOT
Whenever [you] cross in whatever way (Lit. doing whatever), the boat will
just spin around and around, it is said.

The form otok=ma'=ci=ba contains the morphemes (do.like.this/that=interj=LOC
=INDEF/ADD) and means but. This discourse connective is interesting because it
contains an interjection in what used to be its string of enclitics. The interjection
signals surprising or unexpected contrasts, as we can see in example (285) here below.

257
The surprising event expressed in the last clause is further signalled by the mirative
clausal enclitic <=toy> (MIR).

(285) gadakciciaymuna thopsetthiriokno. otokma'ciba uba sa'gorayba jumu
kha'thirithirioknotoy.
{gadak -cici} =ay =muna {thop -set -thiri -ok} =no
cut.up -into.pieces =ADV =SEQ throw -DISPOSE.OF -AGAIN -COS =QUOT
tkmaciba [u =ba sa'goray] =ba
but DST=EMPH child =EMPH
[jumu] {kha' -thiri -thiri -ok} =no =toy
collect do -AGAIN -RED -COS =QUOT =MIR
They cut him up into pieces and threw him away again, it is said. But that
child joined together again and again, it is said to our surprise.
16.2 Type 2 Discourse connectives
Type 2 discourse connectives are all forms of the distal demonstrative <u-> (DST).
These discourse connectives indicate a temporally precise relationship between
stretches of text and are often used to indicate the climax of a stretch of text but they
are also used in the same way as the Type 1 discourse connectives to simply indicate
that there is more to follow in a narrative.
These discourse connectives might be formally similar to demonstratives,
functionally they are not for the following reasons. Firstly, demonstratives are
dependents in an NP and are always the first constituent in an NP, discourse
connectives have a fixed position, i.e. in between sentences. Secondly, demonstratives
modify nominal heads, whereas discourse connectives link sentences and stretches of
discourse. Let us look at the use of the Type 2 discourse connectives one by one.
16.2.1 uciba
Although it is clear that uciba but then, but consists of the distal demonstrative
followed by the locative <=ci> (LOC) and the indefinite morpheme <=ba> (INDEF),
the meaning of this lexeme is not predictable from the sum of its morphemes.
Therefore, I treat it as morphologically opaque and gloss it in its entirety instead of
breaking it up into its alleged morphemes. Further study is required to find out what

258
the exact difference is between uciba but (then)and otok(ma')ciba but treated in
16.1.5. The following example illustrates the use of this discourse connective.

(286) aro jagalan kawokno. uciba khirumancano.
aro [jagal] =an {kaw -ok} =no
and everybody =FC/ID shoot -COS =QUOT
uciba {khi -rum -an -ca} =no
but hit.the.mark -ALL -REF -NEG =QUOT
And everyone shot, it is said. But all of them did not hit [the reed culm], it is
said.
16.2.2 umigmnci ~ umgmnci
In the form umigomonci ~ umogomonci thats why, therefore, for that reason we
can clearly discern the distal demonstrative followed by the genitive <=mi ~ =mo>
(GEN), the bound reason postposition gomon (REASON) and the locative enclitic
<=ci> (LOC). Only on the discourse connective and nowhere else in the grammar,
does the locative appear after the bound reason postposition gomon (REASON). This
non-productive occurrence of the locative indicates the non-compositionality of the
lexeme. The following example illustrates the use of this discourse connective. It links
the last clause of the story about a place in the Symsang river called Dabatwari to the
entire story that precedes.

(287) umigomonci iawdo dabatwari mowano.
umigmnci [i] =aw =do [dabatwari] {mo -wa} =no
thats why PRX =ACC =TOP Pname call.a.name -FACT =QUOT
Thats why [we] call this place Dabatwari, it is said.
16.2.3 una
The discourse connective una then, therefore, because of that seems to be
morphologically and semantically transparent, consisting of the distal demonstrative
and the dative case enclitic <=na> (DAT). The meaning therefore can be deduced
from the meaning of the morphemes, as the next example suggests. However, this
analysis means that only dative-marked distal demonstrative phrases but no other
phrase types can be interpreted as a Reason adjunct. Otherwise only dative marked

259
clauses can fulfil this semantic role (see Chapter 27). The discourse connective una is
the result of a grammaticalisation of the dative-marked distal demonstrative.
In the next example we see the discourse connective una therefore which refers
back to the text in the paragraph that precedes the clause in this example. Not that this
discourse connective can take the delimitative enclitic <=sa> (DLIM).

(288) te'cinakhokho unasa badri rongdo ha'way mowano.
[te'] =ci =na [khokho] una =sa [badri rodo ha'way]
now =LOC =DAT still therefore =DLIM Pname
{mo -wa} =no
Until now, still, exactly because of that, [we] call [it] Badri Rongdyng
Hawai, it is said.

A reason to treat una then, therefore, because of that as one unit rather than as a
sequence of the distal demonstrative <-u> (DST) and the dative case <=na> (DAT), is
that the lexeme una also has a temporal sequential interpretation then which cannot
be deduced at all if we analyse it as a distal demonstrative with a dative case enclitic.
The example below illustrates the temporal interpretation of the discourse
connective una then, therefore, because of that.

(289) co'isa rawkhalay nowano. una jom'aymu sinthong'waci aya! noaymu
jalaokno
[co'isa] {raw -khal} =ay {no -wa} =no
a.little long -CP =ADV say -FACT =QUOT
una {jom'} =ay =mu {sinthong' -wa} =ci [aya]
then sneak.up.on =ADV =SEQ cut.in.half -FACT =LOC interj
{no} =ay =mu {jal -a -ok} =no
say =ADV =SEQ run.away -AWAY -COS =QUOT
A bit longer, [he] said, it is said. Then, having sneaked up on [her], when
he cut her in half, she said Ouch! and ran away, it is said.

This leaves us with only three morphologically and semantically transparent discourse
connectives, viz. u=ci (DST=LOC) then and its focused forms u=ci=e
(DST=LOC=FC) then and u=ci-an (DST=LOC=FC/ID) then. The locative case

260
indicates both Spatial and Temporal Location, and the interpretation of the forms
under discussion as temporal demonstratives is pragmatically conditioned.
As example (290) shows, the discourse connective uci ~ ucie ~ ucian then can
function in the same way as the Type 1 discourse connective otokoymo ~ otokoymu ~
otokoymuna ~ otokoymu ~ otkoymuna so then, having done that/this, i.e. to signal
that more is to follow in the story. The example is taken from the beginning of a story
and starts with a presentative clause consisting, as presentative clauses do (see the
chapter on clause types) of a nominal predicate head.

(290) nay'nokholthagaba aro konokholthagabano. ucie konokholthagabado
sansanan dabatwarisa digaray sana re'eroanoro.
[nay'nokhol] =tha =gaba aro [konokhol] =tha =gaba} =no
mother-in-law =OWN =DREL and son-in-law =OWN =DREL =QUOT
ucie [konokhol] =tha =gaba =do [san san] =an [dabatwari] =sa
then son-in-law =OWN =DREL =TOP day RED =FC/ID Pname =MOB
[digaray] {sana} {re'e -wa} =no
fish.trap put.as.trap go.away -FACT =QUOT
A mother-in-law and a son-in-law, it is said. Then, the son-in-law goes to
Dabatwari every day to put fish traps, it is said.

As was mentioned above, Type 2 discourse connectives can be used to indicate
temporally precise relationships between events. The next example illustrates this.
The context is as follows. A small bird, a frog and a toad are punishing an elephant
because he always destroys their houses. After they picked at the elephants eyes, the
elephant becomes blind and because of the toad and the bird pestering him all day, he
has gotten thirsty. As part of the plan, the frog lures the elephant to the edge of a
ravine by quacking at the bottom of it. The elephant will of course not see the cliff,
fall down and die.


261
(291) otokoymo te'do taw'reksorupmo beblokmo baletaydokno. ha parawbo
bay'siga noaymo, uci rupeke hoyawe ro'ka otokno'sa pekpek pekpek
noay parawaydoknowa.
otokoymo [te'] =do [taw'reksorup =mo beblok =mo]
so.then now =TOP banana.bird =COM toad =COM
{bal -et -aydok} =no
speak -CAUS -PROG =QUOT
[ha] {paraw} =bo [bay'siga] {no} =ay =mo
interj make.animal.sound =IMP friend say =ADV =SEQ
uci [rupek] =e [hoyawe] [ro'ka otyk no'] =sa
then frog =FC yonder cliff bottom.of.ravine inside =LOC
[pekpek pekpek] {noay} {paraw -aydok} =no -wa
frog.sound frog.sound say make.animal.sound -PROG -QOT -FACT
So then, now, the banana bird and the toad are speaking, it is said. Hey, call,
friend! they say and then the frog, way over there, at the bottom of the ravine
is calling pekpek! pekpek!, it is said.

Note that in the above example the morpheme <=sa> is glossed as locative.
Historically this morpheme comes from the noun sa meaning place, side and is still
found with that meaning in a few compounds, e.g. saphak ~ samphak side.

262
Chapter 17 Other word classes
_____________________________________________________________________

The word classes treated in this chapter are: the additive conjunction aro and, the
personal pronouns, the generic pronoun, the proclauses, the onomatopoeia and the
interjections. The Prohibitive word <ta> (PROH) is treated in 26.2.3.
17.1 The additive conjunction aro and
The additive conjunction aro and, more is a linking device that links both clauses
and NPs. This word also functions as adjective meaning more. As adjective aro
more, other always precedes the NP it modifies. The word aro is an Indic loan,
borrowed from Assamese or Bengali.
In example (290) above we see how aro and links two nouns in an NP which
functions as nominal predicate head of a presentative clause. In example (292) here
below we see the clause linking function of aro and.

(292) ana bunduk hon'etbo. aro ana curiba hon'etbo.
[a] =na [bunduk] {hon -et} =bo aro [a] =na [curi] =ba
1s =DAT gun give -CAUS =IMP and 1s =DAT knife =ADD
{hon' -et} =bo
give -CAUS =IMP
Give me a gun and give me also a knife.

As mentioned above, aro can also be used adjectivally meaning more, other. The
following example illustrates the use of this adjective meaning more. In example
(427) in 20.2.2 we see its use with the meaning other.

(293) aro ja'bek hon'bo
[aro ja'bek] {hon'} =bo
more curry give =IMP
Give [me] more curry.

Example (294) below illustrates that it is not always possible to tell whether aro
functions as adjective or clause linker. Translation A mirrors the clause linking
interpretation whereas translation B mirrors the adjectival interpretation.
17 OTHER WORD CLASSES

263
(294) mo'sa them! kawoknotoy. khiancano. aro mo'sa kawtheriokno. them!
kawokno. uba khiancano. aro jagalan kawokno. uciba khirumancano.
[mo' sa] [them] {kaw -ok} =no =toy
CLF:HUMANS one bang shoot -COS =QUOT =MIR
{khi -an -ca}- =no
hit.the.mark -REF -NEG =QUOT
aro mo' sa {kaw -theri -ok} =no [them] {kaw -ok} =no
and CLF:HUMANS one shoot -AGAIN -COS =QUOT bang shoot -COS =QUOT
Translation A: One person shot, bang! [he] did not hit [the reed culm], it is
said, to [our] surprise. And another person shot again, it is said.
Bang! he shot, it is said.
Translation B: One person shot, bang! [he] did not hit [the reed culm], it is
said, to [our] surprise. One more person shot again, it is said.
Bang! he shot, it is said.

17.2 Personal pronouns
Personal pronouns are a closed class. Table 51 lists the personal pronouns in Atong.
Personal pronouns are deictic and the third person pronouns can also be used
anaphorically. They constitute the only word class in Atong that expresses number,
viz. singular and plural. In addition the first person has an inclusive versus exclusive
distinction in the plural.

Table 51 Personal pronouns
a ~ aa 1s
na' ~ na'a 2s
ge'the ~ de'the 3s
na'na 1pi
ni ~ nia 1pe
na'-tom (2s-ppp) 2p
ge'thethe 3p
itom 3p
utom ~ otom 3p
phaltha self

Clausal properties
Personal pronouns can function as core or oblique arguments. Personal pronouns can
function as head of a predicate of identity/equation clauses and are attested with the

264
referential suffix <-an> (REF), the negative suffix <-ca> (NEG) and the change of
state suffix under negation <-k> (COS).
Phrasal properties
Personal pronouns
can function as head of an NP,
can modify nouns possessively by juxtaposition with or without
genitive marking, the order is always Personal Pronoun (Possessor)-
Noun (Possessed)
cannot be modified or possessed.
two juxtaposed unmarked personal pronouns can be interpreted as
being in an additive relationship, e.g. (457) 20.5.
Morphological properties
Personal pronouns can take all the cases and other phrasal enclitics but cannot be
pluralised with the plural enclitic <=dora> (p). The second person plural is formed
with the short form na' (2s) and the personal pronoun plural suffix <-tom> (ppp),
which is also used to form the third person plural from the distal demonstrative 
(DST), viz. utom (3p) and can be distinguished in the allomorph with otom (3p), where
the vowel of the distal demonstrative has been reduced to schwa. The third person
ge'the (3s) and the reflexive pronoun phaltha self form their plurals through
partial reduplication, viz. ge'thethe (3p) and phalthatha selves.
The two forms of the first and second person singular, aa ~ a (1s) and na'a ~
na' (2s), are in free variation in A and S function but there are certain differences.
The forms aa (1s) and na'a (2s) only occur as S or A argument in a clause.
Semantically the longer forms aa (1s) and na'a (2s) are more emphatic and they can
serve in situations of contrastive focus and new topic. Morphologically, too, there are
differences between the two forms. First of all, the allomorphs aa (1s) and na'a (2s)
cannot take suffixes. Secondly, the allomorphs aa (1s) and na'a (2s) are unable to
enter into a possessive relation with a following NP by juxtaposition. The long form
of the first person plural exclusive, nia (1pe) is only attested four times in the
recorded material, in stories by two different speakers. In all cases of its appearance,
the pronoun is in A function.
Only the second person singular allomorph na'a is used as an address term at the
end of sentences when the speaker wants to express that the contents of the sentence

265
are of particular importance to the interlocutor. The next examples illustrate the
second person personal pronoun na'a (2s) in address term function. In (295) and
(296) we see two dialogues, at the end of which the speaker adds the pronoun na'a
(2s) to the clause.

(295) Speaker 1: a =do cok -aydoa.
1s =TOP cold -DUR
Im cold!
Speaker 2: atak -wa?
do.what -FACT
What did [you] do?
Speaker 1: te'ew -mama =sa toyru -wa naa.
now -EXCLUSIVELY =DLIM take.a.bath -FACT 2s
[I] only just took a bath, oh you!

(296) thup thokwaci wek nothiriokno. atakwa? nookno. a di'phusa na'a.
[thup] {thok -wa} =ci [wek] {no -thiri -ok} =no
hitting.sound hit -FACT =LOC pigs.cry.sound say -AGAIN -COS =QUOT
{atak -wa} {no -ok} =no [a di'phu] =sa [naa]
do.what -FACT say -COS =QUOT 1s fart =DLIM 2S
When [he] hit [it] Thwack! [the pig] said Squeel! again, it is said. What
did you do/What happened? [he] said, it is said. [its] only my fart, oh
you!

The Atong form na' 2s is the only form used in you-youing
38
. An example of
you-youing is given in (103) below. The pronoun na'a (2s) cannot be used here.

38
As in Dutch jijbakken you-youing, which is the childish activity of passing accusations back and
forth by saying You You.

266
(297) you-youing: Person A and Person B are talking to each other.
Person A: kaltok! Person who never washes!
Person B: na'! You!
Person A: na'! You!
Person B: na'! You!
Person A: na'! You!

The two forms of the third person singular <ge'the ~ de'the> (3s) are in free
variation, the allomorph <ge'the> (3s) occurs much more frequent than <de'the>
(3s). The composition of the first person plural inclusive <na'na> (1pi) is opaque in
the current stage of the language. However, it might have originated from a compound
of which the elements no longer occur as separate morphemes in the language of
today. The demonstratives <ue ~ u-> (DST) and <ie ~ i-> (PRX) can also be used as
third person personal pronouns. As such they can take the highly selective personal
pronoun plural suffix <-tom> (ppp). The resultant forms are utom (3p) and itom (3p)
which are personal pronouns. The other third person plural ge'thethe shows partial
reduplication. Historically the element <-the> might have been a suffix. It might
have been the phrasal enclitic <=tha> (own), of which the vowel was harmonised
with the front vowel in the first syllable /ge ~ de/. The only other enclitic that can be
reduplicated with the meaning plural is <=tha> own as the next example illustrates.
This enclitic probably also forms a fossilised element in the now opaque reflexive
pronoun phaltha self and its plural form palthatha selves.

(298) [] khasin-khasin gumukawan paloci jalgabadoraaw jokthathaaw
jumuphonnaakno.
[khasin -khasin] [gumuk] =aw =an [[palo] =ci
slow -RED all =ACC =FC/ID jungle =LOC
{jal} =gaba] =dora =aw} [jok] =tha =tha =aw
run.away =ATTR -p =ACC spouse =OWN =RED =ACC
{jumu -phon -a -ak} =no
collect =BACK -TOWARDS -COS =QUOT
[They] slowly collected all their husbands back who [had] run into the
jungle.

There is another third person plural pronoun, viz. otom, which is a phonologically
altered form of the distal demonstrative with the personal pronoun plural suffix

267
<-tom> (ppp). As far as I am aware, all the different third person plurals are in free
variation.
17.3 The generic pronoun
Atong has one generic pronoun which has a free form hay'e (GPN) and a bound form
hay' (GPN). The generic pronoun can be used as filler or as replacement for any NP
in a clause when the speaker cannot think of the correct word (Let me see; uh,
whatchamacallit) or does not want to say it, e.g. (299), (300).

(299) te'do ucian pherudo biskutaw payay jalokno, pa'ay pa'ay jalokno
magacakmo, hay'e [pause], bagalmo biskutaw.
[te'] =do [u] =ci =an [pheru]=do [biskut] =aw {pay} =ay
now =TOP DST=LOC =FC/ID fox =TOP biscuit =ACC carry.in.hand =ADV
{jal -ok} =no {pa'} =ay {pa?} =ay {jal -ok} =no
run.away -COS =QUOT many =ADV many =ADV run.away -COS =QUOT
[magacak] =mo [hay'e] [bagal =mo biskut] =aw
deer =GEN FILLER Bengali =GEN biscuit =ACC
Now then the fox ran away carrying the biscuits, he ran [with] a lot, a lot [of]
the deers, uh, the Bengalis biscuits.

(300) Speaker S: ue, usami? bimu ato mowa?
Speaker J: hay'e na'a, joken.
[ue] [bi] =sa =mi [bimu] [ato] {mo -wa}
DST QF =MOB =ABL name what call.a.name -FACT
[hay'e] [na'a] [joken]
FILLER 2s Pname
Speaker S: He, from where [is he]? What [is his] name called?
Speaker J: Let me see, oh you!, Jonken.

There are no clauses recorded in which the generic pronoun co-occurs with a
demonstrative pronoun or any other modifier. The following example illustrates the
use of the determiner. Only the relevant passages of the narrative sample have been
glossed. These passages have been bolded and underlined in the translation.

268
(301) [] otokoymuna hay'aw garu ramgabaaw dethe garu ramgabaaw rotokno.
[] uan phowramu hay'e garutara dowariwate.
[... having packed the whole lunch, having done all this and having done all
this and having [put] the leafy greens outside to dry she left to weed in the dry
rice and vegetable field. [] The mother came back to cook rice. So then she
collected the dried leafy greens. [] Mother, oh! the curry is really tasty!
You dont cook tasty very often, why is it so tasty today, the curry? What did
you add, mother? [the son] said, it is said. [] I added only that rice
powder with the leafy greens, Im telling you!

otokoymuna [hay'] =aw [garu {ram} =gaba =aw
so.then GPN =ACC mustard dry =ATTR =ACC
[[de'the garu] {ram} =gaba] =aw {rot -ok} =no
3s mustard dry =ATTR =ACC collect -COS =QUOT
So then she collected the dried leafy greens.

[u =an phowra] =mu [hay'e garu] =tara
DST=FC/ID rice.powder =COM GPN mustard =EXCLUSIVELY
{dow -ari -wa} =te
add -just -FACT =DCL
I added only that, eh, rice powder with these leafy greens, Im telling you!

There is a homophonous generic pro-verb hay'- (PRO-VERB) which can also be used
as filler for any verbal predicate head and can be translated as do something, do this,
this happens (302).

(302) otokoy phetaaymuna hay'okno []
[otokoy] {phet} -a =ay =muna {hay' -ok} =no
like.that arrive -TOWARD =ADV =SEQ PRO-VERB -COS =QUOT
After arriving this happened, it is said []
17.4 Proclauses
There are six words that form clauses on their own. These are the proclauses listed
below in Table 52. The combination /hm/ stands for a voiceless bilabial nasal [m ].

269
Table 52 List of proclauses

hay Come on!; Lets go!
ha' take it from me!
ho'o yes
om affirmative
'm. hm' thats right This pro clause is clearly bisyllabic, it has a lower pitch on the
first syllable. Pronunciation: glottal stop followed by a long voiced bilabial
nasal that becomes voiceless and ends in a glottal stop.
hm'm no
hayda I dont know.

Proclauses differ from interjections (see 17.6) in various ways. Most proclauses
express polarity, viz. ho'o yes, om affirmative, ?m:hm? thats right and hm'm
no, while interjections do not do this. The pronclause hay Come on! Lets go! is an
adhortative expression and therefore expresses mood, something interjections cannot
do. Some proclauses can be followed by clausal enclitics, which is another property
that interjections do not have. The word hay Come on! Lets go! can take the
imperative emphasiser clausal enclitic <=to> (IMPEMPH) as example (303) below
illustrates. The word ho'o yes can occur with the irrealis clausal enclitic <=com>
(IRR) (see26.8, example (724)), the speculative clausal enclitic <=khon> (SPEC)
(see 26.9, example (740)) and the confirmative clausal enclitic <=mo> (CONF),
e.g.TEXT 2, line23, and example (520) in 21.4.
Proclauses cannot take arguments or any kind of modificatory phrase. Two
proclauses, viz. hay Come on!; Lets go! and ha' take it from me can, however,
have nouns that are associated with them. These nouns are always unmarked for case.
The nouns that can be associated with hay Come on!; Lets go! is the person to
whom the command is directed, i.e. Vocatives, which are not part of the clause, which
can be any person, e.g. (303), (304). The noun associated with ha' take it from me
can only be the object given, e.g. (305). Between the proclauses and their Vocatives,
there will usually be a pause, but not always.

(303) hayto mosa, na'a ana honcakama.
hay =to [mosa] [na'a] [a] =na {hon -ca -ka} =ma
come.on = IMPEMPH friend 2s 1s =DAT give -NEG -IFT =Q
Come on buddy, arent you going to give [some bananas] to me?


270
(304) hay aba nookno bebloke
hay [a] =ba {no -ok} =no [beblok] =e
come.on I =ADD say -COS =QUOT toad =FC
Come on! Me also said the toad, it is said.

(305) ha', cabi.
ha cabi
take.it.from.me key
take it from me, a/the/its a key.

The affirmative and negative proclauses ho'o yes and hm'm no indicate the
attitude of the speaker towards an utterance of his interlocutor. If the speaker agrees
with the utterance he will use ho'o yes and if he disagrees hm'm no (306), (307),
(308). These proclauses can be used in addition to the appropriate form of the
predicate (309) or on their own as a complete answer to the question (310). If both a
proclause and a predicate are expressed, the answer word usually precedes the
predicate.

(306) mamu dowancate nookno. ucie: hm'm ama na'do tay'nido atoba
dowwa.
[mamu] {dow -an -ca} =te {no -ok} =no [u] =ci =e
nothing add -REF -NEG =DECL say -COS =QUOT DST=LOC =FC
hmm [ama] [na'] =do [tay'ni] =do [atong =ba {dow -wa}
no mother 2s =TOP today =TOP what =INDEF add -FACT
[I] did not add anything, I tell you! [she] said, it is said. Then: No, mother,
today you added something.

(307) ram rimolnakakhonay ho'o rimolnakakhon
[ram] {rimol -naka} =khon =ay hoo {rimol -naka} =khon
road slippery -IFT =SPEC =POS yes slippery -IFT =SPEC
The road might certainly be slippery, positively! Yes, [it] might certainly
be.

(308) na' ray'cawa? ho'o.
[na] {ray' -ca -wa} hoo
2s go -NEG -FACT yes
You will not be going? Yes (i.e. I will not be going.)

271
(309) cow ronima? hm'm, rocawa.
[cow] {ro -ni} =ma hmm {ro -ca -wa}
rice.alcohol drink -FUT =Q no drink -NEG -FACT
Shall [we] drink liquor? No, [I] will not drink.

(310) pholgom de'etdapay tanawa? ho'o no'okno.
[pholgom] {de' -et -dap} =ay {tan -a -wa}
eagle shit -CAUS -ON.TOP =ADV put -AWAY -FACT
hoo {no -ok} =no
yes say -COS =QUOT
An eagle dropped shit [on it]? Yes [she] said

In Text 2, line 55 we see the proclause ho'o yes used as answer to the statement in
line 54. The same text provides a good example of the use of the proclause 'm. hm'
thats right in line 47.
There is another affirmative word, viz. om affirmative that is used as the
acknowledgment to statements (311), (312) and imperatives (313).

(311) na'awba man'seganine nookno. om man'niba na'awba.
[na'] =aw =ba {man' -sega -ni} =ne {no -ok} =no
2s =ACC =ADD be.able -ALT -FUT =TAG say -COS =QUOT
m {man' -ni} =ba [na'] =aw =ba
affirmative be.able -FUT =ADD/EMPH 2s =ACC =ADD/EMPH
Ill get you back! he said, it is said. Yes, Ill get you too/indeed!

(312) ici taw'banok nookno. om. raw'bo nookno.
[i] =ci [taw'] {ban -ok} {no -ok} =no
PRX =LOC bird trapped-COS say -COS =QUOT
m {raw' bo} {no -ok} =no
affirmative catch =IMP say -COS=QUOT
Theres a bird trapped here he said, it is said. Ok, catch it, he said it is
said.


272
(313) anaba ata phulni tanbo, nookno. om.
[a] =na =ba [ata phul ni] {tan} =bo
1s =DAT =ADD flour CLF.ROUND.BAKED.THINGS two put =IMP
{no -ok} =no
say -COS =QUOT
m
affirmative
Put aside two round baked things for me he said, it is said. We will.

The words ho'o yes and hm'm no can also be used as affirmative interjections to
react to statements or conceptions, e.g. (314) and (315).

(314) interruption in a story and continuation
speaker A: ucie
then
speaker B: phagoma =ci
shoulder =LOC
speaker A: hoo [phagoma] =ci [sa'] {gat} =ay =mu
yes shoulder =LOC child load =ADV =SEQ
ucie {da -a -ok} =no =ro
then enter -AWAY -COS =QUOT =EMPH
speaker A: Then
speaker B: On [his] shoulder
speaker A: Yes, on [his] shoulder having loaded the children, then, he went
in, it is said.

In the next example a child roaming through the forest sees a deer, and says:

(315) hm'm, iawdo kawcaka.
hmm [i] =aw =do {kaw -ca -ka}
no PRX =ACC =TOP shoot -NEG -IFT
No, I will not shoot this one.

In the above example the child is answering an inner conception, i.e. the question of
whether to shoot that deer or not.

273
The proclause hayda I dont know expresses ignorance on the part of the speaker
and is the answer to a polar question or to a statement, e.g. (316), where a son asks a
question and the mother answers that she doesnt know.

(316) ido di'an thawokona, randaydo atongtokoy thawaronaka, mo ama?!
nookno. hayda
[i] =do [di'] =an {thaw -ok} =ona [randay =do [ato] =tokoy
PRX =TOP shit =FC/ID tasty -COS =DAT meat =TOP what =LIKE
{thaw -aro -naka} [mo] [ama]
tasty -DUR -IFT CONF mother
[hayda]
I.dont.know
Because this, the shit, is so tasty, how tasty must that meat be, arent I right,
mother? I dont know.
17.5 Onomatopoeia
There are many onomatopoeia in Atong covering a large variety of sounds occurring
in their environment. As (318) shows, onomatopoeia can be unmarked O of the verb
tak- to do. Here are some examples.

(317) mmmmm, mmmmm
Both with high falling intonation, mimics the call of an eagle.

(318) otokoymuna tokoreaw man'aymuna ha'cina wuuuuuuuk dom!
takramphinoknotoy pholgom gal'waan.
otokoymuna [tokore] =aw {man'} =ay =muna} [ha'] =ci =na
so.then neck =ACC get =ADV =SEQ ground =LOC =DAT
[wuuuuuuk dom] {tak -ram -phin -ok} =no =toy
shoooossh thud do -INEVITABLY -TOTALLY -COS =QUOT =MIR
[pholgom gal' -wa] =an
giant.eagle fall -FACT =FC/ID
So then, after [the little child] got [the giant eagle] in the neck, it inevitably
did swoooosh, thud! right to the ground the fall of the (giant) eagle.


274
(319) rookno rookno rookno, krrrrr jamoknotoy. wetsacian jamjolay hupokno
sa'goray molgabado.
{ro -ok} =no {ro -ok} =no {ro -ok} =no
drink -COS =QUOT drink -COS =QUOT drink -COS =QUOT
krrrrr {jam -ok} =no =toy [wet sa] =ci =an
vicious.smoking.sound finish -COS =QUOT =MIR time one =LOC =FC/ID
{jam -jol} =ay {hup -ok} =no [sa'goray {mol} =gaba] =do
finish -QUICKLY =ADV inhale -COS =QUOT child small =ATTR =TOP
[He] smoked and smoked and smoked, it is said, krrrrrrrr, finished quickly to
everyones surprise. In one go [he] quickly finishingly inhaled, it is said, the
little child.

Onomatopoeia can modify verbs. Onomatopoeia cannot be the head of a predicate,
cannot be a constituent in a clause, cannot be modified and cannot modify nouns.
17.6 Interjections
There are many interjections in the Atong language expressing a variety of emotions
on the part of the speaker. Interjections are not part of the clause. They cannot be
modified or modify nor can they take any suffixes or enclitics. Table 53 presents just
a few examples of what might very well be a closed class. Those which can be
glossed satisfactorily will be glossed.

Table 53 List of interjections

Anger
hot Hey!
hot sala Damn! / You bastard!
tyi sala Damn! / You idiot!
sala Damn! / Idiot This lexeme can function both as interjection meaning
something like damn! or as a noun meaning idiot. (Indic loan)
Indignation
h yts expresses indignation
Surprise and admiration
atooow pronounced with a long and falsetto /e/ Wooooow!
baaa pronounced in low pitch and with a long [o]. Woooooow! (Indic loan)
baaapre (idem) (Indic loan)
Surprise
hari ~ hare Huh? (Indic loan)
omoy Huh?

275
Table 53 continued
Surprise, astonishment, amazement and grief
From strongest to lightest expression:
ayaw
aya
ayu
All can be translated as Jeez!, Goodness! or Huh?!. The interjections ayaw and
aya can also be used to express grief.
Mutual understanding
ba' OK then
ma' Very well then. This interjection is attested in two instances as part of a
chain of enclitics: a productive chain in (140) in 9.5 and a fossilised chain
in (285) in 16.1.5. In both cases this interjection indicates surprise or
unexpectedness.
de OK then. (According to some Atong speakers this is a Garo loan.)

Attention seeking
hu hu Hello?
oy Oy!
o this interjection is pronounced on a higher pitch than the following word,
usually a proper name. There is no pause between the interjection and the
following word. The proper name that follows is pronounced with falling
intonation, e.g. o samrat! Hey Samrat!
Self location
kow pronounced short and in falsetto Im here!
Acknowledgment
o pronounced long with rising intonation.

276
Chapter 18 Word-class-changing derivation
_____________________________________________________________________
18.1 Types of derivation
Atong shows nine types of word-class changing derivation, two of which are not
productive. Here below is an overview of the types of derivation. Noun incorporation
by means of the support verb constructions is treated in Chapter 22.

NOUN zero derivation VERB (not productive). 18.2
NOUN zero derivation Type 2 ADJECTIVE (not productive).
18.3
ADJECTIVE suffixation VERB. 18.4
NOUN suffixation more Verb-like. 18.5
VERB/ADJECTIVE reduplication ADVERB. 18.6
VERB zero derivation ADVERB. 18.7
NOUN reduplication ADVERB. 18.8
VERB nominalisation PERSON NOUN. 18.9
18.2 Denominal verbs or deverbal nouns, zero derivation
The only aspect a nominal predicate with a prototypical noun as its head can express
is negative change of state. Very few lexical items are attested, however, that can
occur both as contituents in a clause and as verbal predicate heads, and can carry
aspect and modality marking which only occurs on verbal/adjectival predicate heads,
such as non-negative change of state (56), (327), (331), progressive aspect (329),
future modality (331) and customary aspect and the imperative mood (331). All verbal
occurrences of these lexical items are intransitive. Since this phenomenon occurs so
rarely, I have the suspicion that we have to deal here with a closed set of lexical items
that can function both as verbs and as nouns. I cannot say in which function these
words appear most frequently and thus it is impossible to establish whether they are
basically nouns or verbs. Table 54 presents some examples. The list is not exhaustive.
Examples (324)-(327) illustrate these words as head of an NP and as predicate head.

18 WORD-CLASS CHANGING DERIVATION

277
Table 54 Nouns that also occur as verbal predicate heads.
NOUN
example
VERB
valency example
balwa wind (55)
balwa- to blow (as
wind)
S: only
the noun
balwa
wind
(55)
cowgon festival
of the dead
(326)
cowgon- to celebrate
the festival of the dead,
drink for a dead
person
S, A, O
(327),
(328)
golpho story (458)
golpho to talk
extensively
S, A, O (57)
di'phu fart (330) di'phu to fart S (331)
wal night (324) wal to (be) night san day
(322),
(56)
manap morning
manap- to be
morning
zero
gasam evening gasam- to be evening zero (322)
mokha face mokha- to-face
S and
Direction
(320),
(321)
toy' egg toy'- to lay an egg S

(320) a ge'thesa mokhani
[a] [ge'the] =sa {mokha -ni}
1s 3s =MOB face -FUT
I will sit face-to-face with him.

(321) mokharukbo!
{mokha -ruk} =bo
face -RC =IMP
Face each other!

(322) te'ewdo gasamok. ray'na man'ancak. phetana daka walnaka.
[te'ew] =do {gasam -ok} {ray'} =na {man' -an -ca -k}
now =TOP evening -COS go =DAT be.able -REF -NEG -COS
{phet -a} =na [daka] {wal -naka}
arrive -AWAY =DAT before night -IFT
Now it has become evening. We cant go any more. It will certainly be night
before we arrive.

There is a derivation of the body-part noun kon back, i.e. konju- to turn your back
to somebody. Apart from kon back and mokha face no other body-part noun can
function as a verb.

278
(323) tay'ni balwa tha'rakay balwaaok.
[tay'ni] [balwa] {tha'rak} =ay {balwa -a -ok}
today wind strong =ADV wind -AWAY -COS
The wind blew strong today.

(324) walci ha' saw'naka.
[wal] =ci [ha'] {saw' -naka}
night =LOC soil burn -IFT
At night we will burn the soil.

(325) te'ewdo walok.
[te'ew] =do {wal -ok}
now =TOP night -COS
It has become night now.

Notice that the verbs wal- to (be) night, manap- to be morning and gasam to be
evening have a valency of zero, i.e. they cannot take any arguments. This is not
unusual for verbs indicating weather events; however, in Atong these are the only
verbs of such type that have zero valency. The verb wa- to rain is always used with
the noun ra- rain, viz. ra wa-aydok (rain rain-PROG) rain is raining. As far as
other weather events are concerned, they are expressed as follows: soltoy rat-a (iron
water hit IMPF) hail hits, rasan kam-a (sun burn-CUST) the sun burns (alt. its
hot) and balwa gana (wind Exist) wind is(alt. the wind blows) Also recorded but
uncertain is the lexeme balwa wind used as a verb with zero valency, viz. balwa-
aydoa (wind-PROG) [it] is wind-ing in English: the wind is blowing.

(326) dakado mamu khem ni'wacido, domcorasasa cowgon rowano.
[daka] =do [mamu khem] {ni' -wa} =ci =do
before =TOP nothing drum not.exist -FACT =LOC =TOP
[domcora] =sa =sa [cowgon] {ro -wa} =no
snare.instrument =INSTR =DLIM festival.of.the.dead drink -FACT =QUOT
In the past, when there was no drum, [we] celebrated (lit. drunk) the festival
of the dead only with snare instruments, it is said.


279
(327) [] acu ambitokoy dothoyciay takaymo uan me'ma saw'etokno
cowgonokno.
[acu ambi] =tokoy {dothoy -ci} =ay {tak} =ay =mo
grandpa grandma =LIKE kill.ritually -FIRST =ADV do =ADV =SEQ
[u =an me'ma] {saw' -et -ok} =no
DST =FC/ID ghost burn -CAUS -COS =QUOT
{cowgon -ok =no
celebrate.the.festival.of.the.dead -COS =QUOT
Like [their] ancestors (lit. grandfather-grandmother) [in the past] having
ritually killed [a lizard] [they] burned that ghost, it is said [and] celebrated the
festival of the dead, it is said.

(328) sangumuk morot thoygaaw cowgonok.
[san] =gumuk [morot {thoy} =ga] =aw {cowgon -ok}
day =whole person die =ATTR =ACC celebrate.the.festival.of.the.dead -COS
We drank the whole day at the house of the family of the dead person.

(329) konsa golphook golphook golphook. golpho kha'wacie walaaydok.
kosa {golpho -ok} {golpho -ok} {golpho -ok}
later.on talk.extensively -COS talk.extensively -COS talk.extensively -COS
[golpho] {kha' -wa} =ci =e {wal -a -aydok}
story do -FACT =LOC =FC night -AWAY -PROG
Later on they talked and talked and talked extensively. When they talk/talked,
it is/was becoming night. Literally: When they did story it is/was nighting
away.

In the next example we see the lexeme di'phu fart as a possessed noun. The
Possessor is the first person singular a my. The ability to be possessed is an
exclusively nominal characteristic. Example (331) shows three occurrences of the
same lexical item di'phu fart: 1: as the predicate of a nominalised clause, 2: in the
imperative, marked for future modality and 3: marked for change of state.


280
(330) ma' ato korowa? nowacie, aya naqa! a di'phusa nowano
[ma'] [ato] {koro -wa} {no -wa} =ci =e aya [na'a]
interj what make.noise -FACT say -FACT =LOC =FC excl 2s
[a di' -phu] =sa {no -wa} =no
1s shit -blow =SIMP say -FACT =QUOT
When [the turtle] said What? What was/is making that noise?, Aya, oh
you! just my fart [the monkey] said, it is said.

(331) a di'phuna sokaydokay. kha'sinay di'phubo, nookno. hoyts! kha'sinay
di'phuni, khasin khasin, nookno. pho!! di'phuokno. maca thop khi'okno,
jalaokno.
[a] {di' -phu} =na {sok -aydok -ay} {kha'sin} =ay
1s shit -blow =DAT want -PROG -POS soft =ADV
{di' -phu} =bo {no -ok} =no [hoyts] {kha'sin} =ay
shit -blow =IMP say -COS =QUOT interj:indignation soft =ADV
{di' -phu -ni} {kha'sin kha'sin} {no -ok} =no
shit -blow -FUT soft RED say -COS =QUOT
[pho] {di' -phu -ok} =no [maca] thop {khi' -ok} =no
brap! shit -blow -COS =QUOT tiger SOUND.SYMBOL hit -COS =QUOT
{jal -a -ok} =no
run.away -AWAY -COS =QUOT
[I] want to fart really badly! Fart softly!, [he] said, it is said. Huh! [I]
will softly fart, softly, [he] said, it is said. Brap!! [he] farted, it is said. [The
fart] hit the tiger thop, it is said [and the tiger] run away, it is said.

The word di'phu fart consists of the free morpheme di' shit and the bound
morpheme phu which means something like blow and also occurs in lexemes like
ga-phu (be.erect-blow) to swell, haphu (?-blow) to blow, tokhophu (neck-blow)
and thaphu (?-blow) a blister. It is, at least in the current state of research, not clear
whether the lexeme di'phu fart is basically nominal or verbal.
18.3 De-adjectival nouns or Denominal adjectives: zero derivation
There are two cases in which a noun corresponds to a Type 2 adjective with the same
form. One of them is the morpheme alaga, which, as noun, has the meaning someone
else(332), and as Type 2 adjective has the meaning other (333). The other
correspondence is the morpheme bodoy, which means old man as a noun and old
(of persons) as a Type 2 adjective.


281
(332) alagami nok
[alaga =mi nok]
someone.else =GEN house
someone elses house

(333) nok alaga
[nok alaga]
house other
another house

I cannot say in which function these words appear most frequently and thus it is
impossible to establish whether they are basically nouns or Type 2 adjectives.
18.4 De-adjectival verbs
Type 2 adjectives can function as modifiers to nouns and as predicate heads. The
simplicitive suffix <-ari> (SIMP), the event specifier suffix <-a> (AWAY) and
maybe other event specifier suffixes make Type 2 adjectives more verb-like. Firstly,
by expanding the range of aspectual suffixes they can take. Type 2 adjectives marked
with the simplicitive can occur with the customary aspect marker <-a> (CUST), which
otherwise never happens (334). Secondly, on a Type 1 adjective with the event
specifier <-a> (AWAY), the change of state suffix <-ok> (COS) no longer has the
possibility of being interpreted as having an intensifying meaning (see 5.1) and can
only be interpreted as denoting change of state (335). Thirdly, an adjective marked
with an event specifier can only function as a predicate head. More fieldwork is
needed to find out what the exact effects of event specifier suffixation on adjectives
are. De-adjectival verbs are not attested in the recorded stories but appear frequently
in colloquial speech.

(334) gapsanaria
{gapsan -ari -a}
same -SIMP -CUST
[Its/theyre] just the same.


282
(335) otokoymo te'do jan'aokno bagaldo.
otokoymo [te'] =do {jan' -a -ok} =no [bagal] =do
so.then now =TOP far -AWAY -COS =QUOT Bengali =TOP
So now he had gotten far away, it is said, the Bengali.
18.5 Making a noun more verb-like
The simplicitive suffix <-ari> (SIMP), the event specifier <-phin> (V back), and
maybe other semantically compatible event specifiers, make nouns functioning as
predicate heads more verb-like by expanding the range of aspectual categories they
can express. Nominal predicate heads marked with the simplicitive can take the
customary aspect marker <-a> (CUST), which nominal predicate heads otherwise
cannot do. More fieldwork is needed to find out what the exact effects of event
specifier suffixation on nouns are. Nominal predicate heads with the simplicitive and
the customary aspect are not attested in recorded stories but occur frequently in
colloquial speech. Example (336) is an example from colloquial speech in which we
see the noun baju friend functioning as predicate head with the simplicitive event
specifier suffix <-ari> (SIMP) and the customary aspect suffic <-a> (CUST) attached
to it.

(336) ge'themiba bajuaria
{[ge'the] =mi =ba baju -ari -a}
3s =GEN =EMPH friend -SIMP -CUST
[She] is just his friend.

The noun kan' body (of human) can be used as a verb after suffixation of the event
specifier <-phin> (V back), viz. kan'phin- to turn to/on the side (of the body) (337).

(337) ge'the kan'phinay jowaro
[ge'the] {kan' -phin} =ay {jow -aro}
3s side.of.body -RETURN =ADV sleep -PROG
He is sleeping on his side.
18.6 Deverbal and de-adjectival adverbs by reduplication
Verbal and adjectival (both types) roots, possibly enhanced with event specifier
suffixes, can be reduplicated to modify the following predicate head. Example (338)

283
illustrates a reduplicated verbal stem, (339) shows a reduplicated Type 1 adjective
with event specifier suffix, and in (482) and (340) we see a reduplicated Type 2
adjective. The deverbal and de-adjectival adverbs cannot take arguments.

(338) uci amakmo di'sa coret coret hokhotaydoano.
uci [amak =mo di'] =sa {coret coret}
then monkey =GEN shit =DLIM squirt RED
{hokhot -aydoa} =no
come.out -PROG =QUOT
Then the monkeys shit came squirting out. Literally squirtingly came out.

(339) otokoymu te'do jaraw jaraw ro'ci poy'thataymo ropaydokno pherudo.
otokoymu [te'] =do {ja -raw ja-raw} [ro'] =ci
so.then now =TOP long.time -CONTINUOUSLY RED stone =LOC
{poy' -that} =ay =mo
hold.on.to -EXCESSIVELY =ADV =SEQ
{rop -a -thiri -ok} =no [pheru} =do
stay.in.water -AWAY -AGAIN -COS =QUOT fox =TOP
So then, now, [he] held on tightly to a stone and stayed in the water again for
a long time it is said, the fox.

(340) [] khasin khasin gumukawan paloci jalgabadaraaw jokthathaaw
jumuphonnaakno.
{khasin khasin}
slow RED
[gumuk] =aw =an [[palo] =ci {jal} =gaba] =dara =aw
all =ACC =FC/ID jungle =LOC run.away =ATTR =p =ACC
[jok -tha -tha] =aw {jumu -phon -a -ak} =no
spouse -own -RED =ACC collect -again -TOWARDS -COS =QUOT
[the women] slowly collected everybody again, their own husbands, [the
ones who] had run away to the jungle, it is said.

(341) na'nadongda donda hapsan golgolni
[na'na] [doda doda] [hapsan] {golgol -ni}
1pi alone RED together roam -FUT
We both will roam alone in different places.

284
18.7 Deverbal adverbs by zero derivation
Verbal roots and stems, i.e. the root plus stem-forming suffixes (see Table 63), can
function as adverbs modifying an immediately following predicate head. The deverbal
adverb cannot take any arguments. Here below are some illustrative examples of this
phenomenon.
Another possible analysis of this construction is to say that the bare verbal root or
stem is incorporated into the predicate of the verb it modifies. However, it appears to
be possible to separate the two verbs with other elements in colloquial speech. More
fieldwork is needed to test the grammaticality of the construction when elements
intervene between the two verbs.
In (342) the verb ray-thiri-thiri go again functions adverbially to the verb mu'
stay. The verb mu' stay adds the aspectual value of durativity to the clause, which
is reflected in the English translation with keep.

(342) ram tophacaaymo ue otokoy ray'sotwae ray'mangabaaw ray'thirithiri
mu'na naok
[ram] {to -pha -ca} =ay =mo [ue] [otokoy]
road know -IN.ADDITION -NEG =ADV =SEQ DST like.that
[ray' -sot wa] =e
go -directly -FACT =FC
[{ray' -man} =gaba] =aw [ray -thiri -thiri] {mu'} =na [na -ok]
go -ALREADY =ATTR =ACC go -AGAIN -RED stay =DAT need -COS
Because all of [them] did not know the way as well, that, like, shortcut, [they]
had to keep taking [the road] which [they] had already taken again and again.

Adverbialised verbs often occur in support verb constructions with the verbs tak- to
do and kha'- to do (see Chapter 22), as illustrated in the following examples.

(343) magacakmi mon'do toysiwacian miniksuru takjolarianoro
[magacak =mi mon'] =do {toysi -wa} =ci -an
deer =GEN body.hair =TOP wet -FACT =LOC =FC/ID
[miniksuru] {tak -jol ari -a} =no =ro
be.flat-haired do -QUICKLY -SIMP -CUST =QUOT =EMPH
As for the deers body hair, when [it] is wet [it] just quickly gets flat-haired,
it is said.


285
(344) [] phaltha pe'ay tanagabaw ra'phin kha'na, deet kha'na noymu,
bandiaw watetna cayaydokno.
[[phaltha] {pe'} =ay {tan -a} =gaba] =aw [ra -phin] {kha'} =na
self curse =ADV put -AWAY =ATTR =ACC get -back do =DAT
[de -et] {kha'} =na {no} =ay =mu
untie -CAUS do =DAT say =ADV =SEQ
[bandi] =aw {wat -et} =na {canci -aydok} =no
Name =ACC send -CAUS =DAT think -PROG =QUOT
[the supreme god] [I] want to undo [literally: to get back], to untie the
curse which [I] [my]self have put [upon the village], [he] said and, [he] was/is
thinking about sending Bandi, it is said.

(345) otokma'ciba uba sa'gorayba jumu kha'thirithirioknotoy.
otokma'ciba [u] =ba [sa'goray] =ba
but DST=EMPH child =EMPH
[jumu] {kha' -thiri -thiri -ok} =no =tyi}
reassemble do -AGAIN -RED -COS =QUOT =MIR
But that child reassembled once again, it is said to our surprise.
18.8 Denominal adverbs
Adverbs can be derived from nouns by reduplication. The results of these processes
can be classified in terms of their adverbial versus nominal character as illustrated in
Table 55.

Table 55 The properties of denominal adverbs compared to those of adverbs and
nouns
ADVERBS
ADVERBIALISED NOUNS
NOUNS REDUPLICATION
TEMPORAL NOUNS OTHER NOUNS
cannot function as head of a predicate
can function as head
of a predicate
cannot take case
marking
not attested with
case marking
can take case marking
can modify adjectives and verbs
cannot modify verbs
and adjectives

As we can see in Table 55 the adverbialised temporal nouns, i.e. nouns denoting a
period of time, are most adverbial, while all other adverbialised nouns still display a
nominal property, i.e. they can take case-marking. On adverbial temporal nouns, e.g.

286
(346), case marking is never attested, whereas example (347) has case marking on a
non-temporal denominal adverb.

(346) ayaw nawmol san sanba nang'naba kha'galwa jamcaaydok.
ayaw [nawmol] [san san] =ba [na'] =na =ba
excl unmarried.girl day RED =EMPH 2s =DAT =EMPH
{kha'gal -wa} {jam -ca -aydok}
love -FACT stop -NEG -PROG
The girls are not stopping to love you every day.

(347) umusa sogumuk thom'aymu ha'ba ha'ron ha'ronaw sowalni.
umusa [so] =gumuk {thom'} =ay =mu
CONJ village =whole assemble =ADV =SEQ
[ha'ba] [ha'ron ha'ron] =aw {sowal -ni}
dry.rice.and.vegetable.field plot RED =ACC divide-FUT
Then, the whole village having come together, [they] will divide the land plot
by plot.

In example (347) we see an overlap of clausal functions, viz. affected participant and
adverbial clause. The reduplication in (347) has case marking because not only does it
indicate the way in which the action denoted by the verb takes place, the adverbial
function, but the reduplicated noun is also marked by the speaker as the O argument
with the accusative enclitic <=aw> (ACC), because the plots of land are what is seen
as most affected by the verb sowal- to divide and as a topical NP. It is impossible to
determine whether the NP ha'ba dry rice and vegetable field is in S or O function in
the clause, since we can interpret the verb as being used as an S
O
ambitransitive or as
transitive with an ellipsed A, which would be coreferential with so=gumuk
(village=whole) the whole village. O arguments can be unmarked in Atong and case
marking is primarily semantically and pragmatically based
39
.

39
It is possible for a clause in Atong to contain two accusative-marked arguments, i.e. in sentences
with a causativised transitive verb where Causee and O are both accusative-marked and in clauses as

287
18.9 Nominalisation
Nominalisation creates a noun denoting a person from a verb by using only a verbal
root plus nominal derivational enclitic or a root plus an event specifier. Bare verbal
roots are not attested as nominalised person noun derivations. Nouns denoting a
person are attested in predicate head function, as clausal constituent and as term of
address. Both phenomena, the root plus nominal derivational enclitic or a root plus an
event specifier, can be called zero-marked nominalisation because the derivation is
semantic and syntactic but not formal, i.e. there is no specific nominaliser morpheme
involved. As for the verbal roots with phrasal enclitics, one can also say that the
enclitics act as nominalisers.
Most derived person nouns come from transitive verbs, but, when semantically
appropriate in the right context, also from intransitive ones, as we can see in (349).
Derived person nouns can take nominal inflectional and derivational morphology.
Example (348) comes from a contemporary Atong rap song by Samrat Nokrek
Marak. In this example the two derived person nouns mokca sweetheart and kha'gal
love, i.e. person who someone loves, both carry the third person A/S co-referential
possessive derivational enclitic <=tha> (OWN). Both derived person nouns are
interpreted as Patient nominalisations.

(348) mu', mu', ca'methamu mu'. mokcatha kha'galtha je sokgamu mu'.
{mu'} {mu'} [ca'me] =tha =mu [mu']
stay stay sweetheart =OWN =COM stay
[mokca] =tha [kha'gal] =tha [je {sok} =ga] =mu {mu'}
fancy =OWN love =OWN whoever want =ATTR =COM stay
Stay, stay, stay with your sweetheart. With the one [you] fancy, with your
love, with whoever [you] want, stay.

illustrated by the following example. In that example the first NP, sam grass, is semantic patient and
the second, ca' foot/leg an instrument.

samaw ca'aw itokoy tokano.
[sam] =aw [ca'] =aw [i] =tokoy {tok -a} =no
grass =ACC foot =ACC PRX =LIKE hit -CUST =QUOT
[They] trample the grass like this with [their] feet.

288

A verbal root or stem can possibly be an imperative. However, in Text 2, line 32
we find the clause represented here for convenience as (349). And there we see that
the clause contains a topical S argument, which would not be possible if the clause
were an imperative since imperatives in Atong are second person only. Thus, the
verbal form can be analysed as a partial nominalisation indicating a person. Other
examples of this grammatical phenomenon are (350) and (351).
An alternative analysis of the unmarked forms verbal roots or stems in (349) and
(350) is the following: these forms are actually verbs functioning as predicate heads of
declarative clauses but without any predicate marking and hence with an habitual
overtone. This analysis is problematic for (351), since the bare stem is used as address
term, but this could well be explained as a pragmatic interpretation.

(349) ado ni'sora.
[a] =do {[ni' -sora]
NP
}
PREDICATE HEAD
1s =TOP not.exist -TOTALLY
Im a total not-haver. Alternatively: I totally never have [a girlfriend].

Other examples of zero-marked nominalisation are given here below. The context in
which these occur is as follows. The story teller is saying how precious the soil of the
Garo Hills is, but that the Atong and Garo do not know how to make use of it. And
then the foreigners came and they are so great and rich. They have big salaries, they
can read and study and do not bother their mother. How much do they earn in a
month? They always have enough money to meet expenses. This is not so with the
Garo and Atong, because of (350).

(350) na'naacido tiktikca. sa'bobo robobo. man'gabaaw sa'phet rophet.
[na'na] =ci =do {tiktik -ca}
1pi =LOC =TOP be.sufficient -NEG
{[sa' -bobo]} {[ro -bobo]}
eat -MORE.THAN.NECESSARY drink -MORE.THAN.NECESSARY
[{man'}] =gaba =aw {[sa' -phet]} {[ro -phet]}
have =ATTR =ACC eat -TO.ONES.DETRIMENT drink -TO.ONES.DETRIMENT
We do not meet our expenses with out money. (Lit. At us [it is] not
sufficient.) [We are] gluttons and drunks. Alternatively: [We] always eat too

289
much and drink too much. Those who are rich (Literally: those who have)
are self-destructive drunks and gluttons Alternatively: Those who are rich
always eat and drink to their detriment.

Zero-marked nominalisations can be used in an address term function, i.e. something
to call a person. One of the most frequently used is shown in (351):

(351) re'e buta!
{re'e} [but -a]
go.away penetrate -WITHOUT.HOLDING.BACK
Go away, fucker!

Zero-marked nominalisations are not attested with arguments, i.e. as a clause. So it
might well be that they have lost this verbal property. More fieldwork is needed to
find out exactly what the morphological and syntactic properties these zero-marked
nominalisations have. It would be interesting to know if these zero-marked
nominalisers can be modified with demonstratives and occur in possessive
constructions. I have the strong impression that the process of zero-marked
nominalisation is either not fully productive or that only certain, semantically suitable
verbs used in the right context can take part in this process, i.e. zero-marked
nominalisation could be pragmatically constrained. More fieldwork needs to be
carried out to find out what exactly these pragmatic constraints are.

290
Chapter 19 Phrasal enclitics
_____________________________________________________________________

Phrasal enclitics are grammatical words that modify a phrase and occur in a fixed
position at the end of the phrase, irrespective of whether the last constituent of that
phrase is the head or not (see Anderson, 1992). For a phrasal enclitic to occur, its
semantics must be compatible with the semantics of the phrase they enclitisise to.
Table 56 gives an overview of the phrasal enclitics in Atong. Some of the enclitics,
indicated in the table, also function as clausal enclitics. All the enclitics will be
described one by one below, except for the case markers, which are treated in Chapter
20.

Table 56 Overview of NP enclitics
NAME MORPHEME LABEL
possessive <=tha> (OWN)
reciprocal <=maran> (RC)
plural <=dora> (p)
quantifier <=gumuk> all, whole
distributive <=pek> (DIS)
exclusive <=tara> (EXCLUSIVELY)
privative <=noy ~ =ni> (PRIV)
privative <=ri> (LOST)
locational/quantificational-
delimitative
<=rara> (AMONG/ALL.EXCLUSIVELY)
associative <=para> (&co)
alternative <=sega ~ =siga> (ALT)
additive/emphatic <=ba> (ADD/EMPH)
Also clausal
enclitic (see
Table 70)
focus/identifier <=an> (FC/ID)
topic <=do> (TOP)
focus <=e> (FC)
delimitative <=sa>
(DLIM) Also clausal enclitic,
treated in 11.7.
case marking enclitics (see Table 58)
19 PHRASAL ENCLITICS

291
19.1 The possessive enclitic <=tha>
The NP enclitic <=tha> (OWN) has two functions. One function is as a third person
co-referential possessor marker. The enclitic codes intra-clausal co-reference with the
S or A of the clause which is the possessor of the constituent marked by <=tha>
(OWN), e.g. (768) and (353).

(352) otokoymo te'ewdo amakdo nokthacina doawacido na' ni'oknoa.
otokoymo te'ew =do [amak]
S
=do [nok] =tha =ci =na

so.then now =TOP monkey =TOP house -SF.own =LOC =DAT
{do -a -wa} =ci =do [na'] {ni' -ok} =no
enter -AWAY -FACT =LOC =TOP fish NEG.be -COS =QUOT
So then, now, when the monkey reached his own house, there was no more
fish left.

In example (353) the morpheme <=tha> (OWN) is coreferential with the A argument,
viz. amak=ba (monkey-EMPH) monkey of the clause and not with the genitive-
marked NP constituent rupek-mo (frog=GEN).

(353) [] noay takaidoano, amakba, rupekmo bay'sigathagaba budiaw
tosomay takaymo.
{no} =ay {tak =aydoa} =no [amak] =ba
say =ADV do -PROG =QUOT monkey -EMPH
[rupek =mo bay'siga] =tha =gaba [budi]] =aw
frog =GEN friend OWN =DREL -trick =ACC
{to -som} =ay {tak} =ay =mo
know -IMITATE =ADV do =ADV =SEQ
[Just look in my mind later] [he] said, the monkey, having remembered and
imitated the trick of his own friend the frog. Literally: [] sayingly did [he],
the monkey, having think-followed the trick of his own friend the frog.

The possessive is the only enclitic that can be reduplicated; the reduplication indicates
plurality of the S/A referent, as we can see in examples (354).


292
(354) uan baydamdo haw'aman'gaba ha'gun sa'aman'gaba ha'ronthathaaw
khaa.
[u] =an [baydam] =do {haw' -a -man'} =gaba
DST=FC/ID some.people =TOP clear.the.jungle -AWAY -ALREADY =ATTR
ha'gun {sa' -a -man'} =gaba
old.rice.field eat -AWAY -ALREADY =ATTR
[ha'ron] =tha =tha =aw {kha -a}
parcel =OWN =OWN =ACC occupy -CUST
As for those, some people occupy their own parcels: old rice fields which
have already been cleared, which have already been eaten up.

The other function of <=tha> is that of a marker of kinship terms (Chapter 1)
independently of the syntactic function this noun has in the clause.
19.2 The reciprocal enclitic <=maran>
The enclitic <=maran> (RC) indicates reciprocity on NPs and indicates plurality. It
occurs only on NPs referring to kinship relations and other interpersonal relations and
seldom occurs without the possessive enclitic <=tha> (OWN) preceding it. When the
enclitic <=maran> (RC) is added to an NP, the NP refers to a group of two or more
people that are at the same time Possessor and Possessee, which makes the possessive
relationship reciprocal. Examples (355) and (356) are illustrative of the usual co-
occurrence of the reciprocal and the possessive enclitics.

(355) otokoymo bay'sigathamaran toy dukuokno dukuokno dukuokno.
otokoymo [bay'siga] =tha =maran [toy] {duku -ok} =no
so.then friend OWN =RC water dam.up -COS =QUOT
{duku -ok} =no {duku -ok} =no
dam.up -COS =QUOT dam.up -COS =QUOT
So then, the mutual friends dammed up the water and dammed [it] up and
dammed [it] up.


293
(356) otokoymo ge'the thedo bodoythamaran koksikoda dukuaydoknoa.
otokoymo [ge'thethe] =do [bodoy] =tha =maran koksikoda
so.then 3p =TOP old.person =OWN =RC disorderly
{duku -aydok} =noa
dam.up -PROG =QUOT
So then, they, the old couple, dammed [the water] up disorderly.

When, in the two examples above, the reciprocal enclitic would be removed, the
meaning would change: bay'siga=tha (friend=OWN) [somebodys] friend(s) and
bodoy=tha (old.person=OWN) [somebodys] old person(s).
There are very few recorded instances of the enclitic <=maran> (RC) being used
without <=tha> (OWN), e.g. (357) and (358).

(357) [] nosalimu kasalimusa ja'nawmaran na'ce jow'na reoaymusa [].
[nosali =mu [kasali] =mu =sa [ja'naw] =maran
Name =COM Name =COM =DLIM sister =RC
[na'ce] {joy'} =na {re'e} =ay =mu =sa
shrimp catch =DAT go.away =ADV =SEQ =DLIM
The mutual sisters [sisters to each other] Kasai and Kasali went to
catch shrimps

(358) [] gambirimu gamsilimu ja'nawmarane senthiokno.
[gambiri] =mu [gamsili] =mu [ja'naw] =maran =e
type.of.tree =COM type.of.tree =COM sister =RC =FC
{senthi -ok} =no
lament -COS =QUOT
[When are you going to dress us in clothes?] lamented the gambiri and
gamsili tree to the mutual sisters.

The titles of TEXT 1 and 2, viz. sadu=tha=maran mo' tham (the.relationship.
between.men.whose.wives.are.sisters=OWN=RC CLS:HUMANS three) The three
brothers-in-law whose wives are sisters, prove that NPs with a reciprocal enclitic can
also refer to a reciprocal possessive relationship between more than two persons.

294
19.3 The plural enclitic <=dara ~ =dra>
The morpheme <=dara ~ =dora> (p), whose allomorphs are in free variation,
indicates plurality on NPs with countable nouns and quantification on NPs with
uncountable nouns as their head. This enclitic also occurs on demonstratives. The
distal demonstrative functioning as personal pronoun has a personal pronoun plural
form utom ~ otom (3p). NPs in Atong do not have to be marked for plural to indicate
plurality of the referents. The plural enclitic is thus used on NPs with countable
nouns:
to indicate plurality in a context in which the plurality of the nominal referent
would not otherwise be evident, e.g. (359), (360).
to emphasise the notion of plurality of the referents, e.g. (361), (366),
to indicate plurality or multiple occurrence when it appears on time words (362).

The following example is the opening sentence of a story. Plural marking on the NP is
used here to disambiguate the fact that there was more that one animal, since no
previous context is available for disambiguation.

(359) sagaba matborodora : na'nado raja ni'khua.
[sa -gaba] [matboro] =dara [na'na =do] [raja] {ni' -khu -a}
one -ATTR animal =p 1pi =TOP king NEG.be -INCOM -CUST
Firstly the animals [said]: We
i
dont have a king yet.

The plural marking in (360) is used by the speaker to indicate that he is taking about
all the foreign people, i.e. white people, instead of just about the one in the audience,
i.e. the author.

(360) mayawdo pa'ay sa'ca phorenmi morotdorado.
[may -aw] =do {pa'} =ay {sa' -ca}
rice =ACC =TOP much =ADV eat -NEG
[phoren =mi morot] =dara =do
foreign =GEN person =p =TOP
[They] dont eat a lot of rice, foreign people.


295
Example (361) illustrates the use of the plural enclitic to emphasise the notion of
plurality of the referents.

(361) otokoymo na'dorado uaw rukpek bisi roaymu gumukan thoytokoknoa.
otokoymo [na'] =dra =do [u =aw rupek bisi] {ro} =ay =mu
so.then fish =p =TOP DST =ACC frog poison drink =ADV =SEQ
[gumuk] =an {thoy -ok} =noa
all =FC/ID die -COS =QUOT
So then, the fish, having drunk that frogs poison, all died, it is said.

That the plural enclitic <=dara ~ =dora> (p) can indicate plurality or multiple
occurrence in a very abstract way becomes clear when it appears on abstract nouns
like in example (362) here below, where somay, an Indic loan, is an abstract noun
meaning time. The example describes a recurring event, a custom, which is also
indicated by the customary aspect morpheme on the prediacte head. The use of the
word somay time in the plural is reminiscent of the plural use of the English word
time in sentences as in those times one went to school at seven or during times of
crisis.

(362) umudo marsja somaydaraci saw'a.
umu =do [mars ja =ci somay] =dra =ci {saw' -a}
CONJ =TOP March month =LOC time =p =LOC burn -CUST
So then, whenever March comes [you] burn [the jungle]. Literally: So then,
in the times of month of March...

Example (363) here below shows that when the event is not recurring there is no
plural marking on somay time in the same construction as above in (362).


296
(363) dakami somaydo ni pi'sa molbutucido nokma nodagdo man'ay sa'gasa
gam pa'gasa nokma moa.
[daka =mi somay] =do [[pi'sa (Garo)] {mol -butu} =ci =do
before =GEN time =TOP child small -WHEN =LOC =TOP
[[nokma] {no} =ga] =do [{man'} =ay {sa'} =ga] =sa
big.shot say =ATTR =TOP in.great.amounts =ADV eat =ATTR =DLIM
[gam {pa'} =ga =sa [nokma] {mo -a}
wealth big =ATTR =DLIM big.shot call.a.name -CUST
In the past, when [I] was a small child, as for a so called nokma, only
[someone] who was rich (lit. eats in great amounts) [and] whose wealth was
great was called a nokma.

The plural morpheme on NPs referring to uncountable or mass nouns can indicate a
great quantity of some substance, e.g. (364). In that example we see the NP with the
uncountable noun may rice as its head.

(364) a ie maydoraaw sa'cawa.
[a] [ie may] =dara =aw {sa' -ca -wa}
1s PRX rice =p =ACC eat -NEG -FACT
Im not going to eat all that rice.

The enclitic <=dara ~ =dora> (p) can occur on quantified NPs with a numeral
higher than one:
to reinforce the notion of plurality,
to indicate that the number is approximate (see 11.5).

Context is an important factor for the interpretation of the plural marker on a
quantified NP. The plural marker is not attested on NPs containing the numeral one.
Atong has no numeral for zero.
Example (365) illustrates that it is not necessary to mark a noun for plural when the
context already indicates this. The words sa' child and sa'goray child, in bold
face, are not marked for plural throughout the text, although it is clear from the start
that the referents are plural. This example contrasts with example (366) below where
the plural is used to emphasise the plurality of the referents of the same word sa'
child.


297
(365) te'ewba gorialci sa' mo' sene ganano. [] aya bay'siga, na'cido sa'
pa'ate! Aa mo'tham na' sa'aw poraykalna watetbo, noaymo pherue
pi'okno. []. ucido pheru balokno: na' sa'do boloen leka porayna
man'a. [] na' sa' ataknaay nokci tanna? noaymo. [] te'ewdo
pherudo kakay sa'manokno sa'goraydo.
[te'ew] =ba [gorial] =ci =e [sa mo' sene] {gana =no
now =EMPH crocodile =LOC =FC child CLF.HUMANS seven exist =QUOT
aya bay'siga [na'] =ci =do [sa] {pa' -a} =te
interj friend 2s =LOC =TOP child many -CUST =DCL
[aa] [mo' tham na' sa] =aw]
1s CLF.HUMANS three 2s child =ACC
{poray -khal} =na} {watet} =bo {no} =ay =mo
teach -CP =DAT send =IMP say =ADV =SEQ
[pheru]=e {pi' =ok} =no [u] =ci =do [pheru]{bal -ok} =no
fox =FC ask -COS =QUOT DST=LOC =TOP fox say -COS =QUOT
[na' sa] =do [boloen] [leka] {poray} =na {man' -a} [na' sa]
2s child =TOP very book read =DAT be.able -CUST 2s child
[atakna] =ay [nok] =ci {tan} =na {no} =ay =mo [te'ew] =do
why =POS house =LOC keep =DAT say =ADV =SEQ now =TOP
[pheru]=do {kak} =ay {sa' -man -ok} =no [sagray] =do
fox =TOP bite =ADV eat ALREADY -COS QUOT child =TOP
Now the crocodile had seven children, it is said. Jee, friend, you have a lot of
children! Send me three of your children to teach, sayingly asked the fox.
Then the fox said, it is said: Your children can real books very well. Why
would you like to keep your [other] children at home? [he] said. [and he
asked to teach the other four as well]. Now the fox had already devoured (lit.
bitingly eaten) [them], it is said, the children.

(366) otokoymu sa'daraba pa'anoa.
otokoymu [sa'] =dara=ba {pa' -a} =noa
so.then child =p =EMPH many -CUST =QUOT
So then, there where many children, it is said.

The morpheme dara occurs as a modifier preceding a noun in example (385). In that
example the gloss of dara is every. The morpheme dara also occurs as a free
noun, as in example (178), where it is glossed anybody.
19.4 The quantifier enclitic <=gumuk>
As can be seen in example (361), the enclitic gumuk can occur as a noun on its own
meaning all, everything, everybody. The same morpheme can be cliticised to other

298
NPs and demonstratives with the meaning all, whole. This is demonstrated in
examples (367) and (368). The enclitic can attach to NPs with both countable and
uncountable nouns. In the example below the enclitic occurs on countable nouns. An
example of <=gumuk> on an uncountable noun is [jagi khe-wa]=gumuk (life live-
FACT=whole) [your] whole life.

(367) sogumukan ue moma wana waykhurutaysa boli hon'aysa man'ay
sa'thokwano.
[so] =gumuk =an [ue moma wa] =na
village =whole =FC/ID DST elephant tooth =DAT
{way khurut} =ay =sa {boli hon'} =ay =sa
spirit incantate =ADV =DLIM offering give =ADV =DLIM
{man' =ay} {sa' -thok -wa} =no
in.great.amounts =ADV eat -ALL -FACT =QUOT
The whole village, [because] [they] prayed to the elephant tusk [and because]
[they] gave offerings, [they] all became rich, it is said.

(368) otom mo' nie sangumuk golgolarono
[otom mo' ni] =e [san] =gumuk {golol -aro} =no
3p CLF:HUMANS two =FC day =whole roam -PROG =QUOT
The two of them are roaming the whole day, it is said.
19.5 The distributive enclitic <=pek>
The distributive enclitic <=pek> (DIS) can be reduplicated to reinforce a notion of
plurality of the referent to which it is enclitisised. The enclitic occurs on nouns (369)
and numeral classifiers (370) and is only attested with reference to countable nouns.


299
(369) palema burubaabaa haw'waan pupek phiano.
[palema buru baa baa] {haw' -wa} =an
type.of.plant bush five RED clear -FACT =FC/ID
[pu] =pek {phi -a} =no
rice.stock.house =DIS be.full -CUST =QUOT
[They] cut five bushes of Barebinia xariegata each [and] one rice stock house
each was filled, it is said.

(370) aca na'tome aa sanci mapek hon'ni []
aca [na' -tom] =e [aa] [san] =ci [ma] =pek] {hon' -ni}
interj 2s -ppp =FC 1s day =LOC CLF.ANIMALS =DIS give -FUT
Right then, you
p
shall give me one of each animal [every] day, [the lion
said, it is said].
19.6 The exclusive enclitic <=tara>
The morpheme <=tara> (EXCLUSIVELY) can be understood as indicating
exclusively the referent. The following examples illustrate this.

(371) morot mo' sa ganano. uba jow'taraanokno. wa' ni'okno.
[morot mo' sa] {gana} =no
person CLF:HUMANS one exist =QUOT
[u] =ba {[jow'] =tara =an -ok} =no
DST=EMBH mother =EXCLUSIVELY =FC/ID -COS =QUOT
[wa'] {ni' -ok} =no
father not.exist -COS =QUOT
There is one person, it is said. She has become a single mother, it is said.
There is no more father, it is said.

(372) umigomon na'nae ie ha'golsakci phalthaawtara cunukna naarica,
gumukan hapsan ra'na naa.
[u =mi gomon] [na'na] =e [ie ha'golsak] =ci
DST=GEN reason 1pi =FC PRX world =LOC
[phaltha] =aw =tara {cu} {nuk} =na {na -ari -ca}
self =ACC -only big see =DAT must -SIMP -NEG
[gumuk] =an [hapsan] {ra'} =na {na -a}
everybody =FC/ID same take =DAT must -CUST
Therefore we must not just consider only ourselves as big in this world, [we]
have to consider everybody [as being] the same.

300
19.7 The privative enclitics <=ny ~ =ni> and <=ri>
The enclitic <=noy ~ =ni> (PRIV) is most probably an Indic loan, cf the Hindi
negative morpheme /neh
/. The allomorphs of the privative enclitic <=noy ~ =ni>

(PRIV) are in free variation, the allomorph <=noy> is predominant in the Badri area
whereas <=ni> is most often used in Siju. The difference between the two morphemes
is that <=ri> (LOST) means that something was lost and therefore not present
whereas <=noy ~ =ni> (PRIV) simply indicates that something is not present (374).
There are only two occurrences of <=ri> (LOST) in the recorded corpus; they are
represented here below in (373).

(373) ha'golsakci aa jokri mu'waba, uanarinaka, sa'ri parawaba, [].
[ha'golsak] =ci [aa] [jok] =ri {mu' -wa} =ba
world =LOC 1s wife =LOST stay -FACT =EMPH
{[u] =an -ari -naka} [sa'] =ri {para -wa} =ba
DST =FC/ID -SIMP -IFT child =LOST wander.around -FACT =EMPH
I lived in the world having lost a wife indeed, it will just be the same, [I]
wandered around having lost [my] children indeed, [it will be the same. What
benefit is there in being married to you?].

(374) cininoyca takbo
[cini] =ny [ca] {tak} =bo
sugar =PRIV tea make =IMP
Make tea without sugar.
19.8 The enclitic <=rara>
The morpheme <=rara> among, all/exclusively has two meanings, viz. a locational
one and an quantificational/delimitative one. The locational sense is illustrated in
(375), (376) and (381) below, where seems to indicate category of origin to which
something belongs.


301
(375) te'do ge'thethedo bobarara mo'ni golpho ka'rukokno.
[te'ew] =do [ge'thethe] =do [boba] =rara [mo' -ni]
now =TOP 3p =TOP crazy.person =AMONG CLF.HUMANS -two
{golpho ka' -ruk -ok} =no
Now then, the crazy men amongst each other, two of them, talked [lit. did
story] to each other, it is said.

(376) alsia rajado morotraraanno.
[alsia raja] =do {[morot] -rara =an} =no
lazy king =TOP human -AMONG =FC/ID =QUOT
The lazy king is only a human, it is said. Alternatively: the lazy king is from
among the humans.

The quantificational/delimitative function of the enclitic <=rara> is illustrated in the
following example.

(377) utome morote gawigababa biphagababa bobirara bobararanowa.
[u -tom] =e [morot] =e [gawigaba] =ba [biphagaba] =ba
DST-ppp =FC person =FC wife =ADD husband =ADD
{[bobi] =rara
crazy.woman =ALL/EXCLUSIVELY
[boba] =rara} =no -wa
crazy.man =ALL/EXCLUSIVELY =QUOT -FACT
They, these people, the wives and husbands, [are] all crazy women and crazy
men, it is said.
19.9 The associative enclitic <=para>
The associative only occurs on personal names (378), kinship terms (379) and on
nouns denoting persons or animals, but only when the animals act like persons in
stories (380). The associative indicates that the referent belongs to a group of people
associated with a named person (see Moravcsik, 2003). In two instances an
associative-marked noun has been lexicalised, viz. ama=para (mother-&co) and
jow'=para (mother-&co) both mean mothers household/mothers house. The
associative can be enclitisiseded to an NP in any function in the clause.


302
(378) moya jenetaparamo wakaw sa'wa.
[moya] [[jenet] =para =mo wak] =aw {sa' -wa}
yesterday Name =&co =GEN pig =ACC eat -FACT
Yesterday we ate the pig of Janet and her company.

(379) ah! na' dadaprardo usa pholgom cugaaw kawna re'ewanote.
ah [na' dada] =para [u] =sa [pholgom {cu} =ga] =aw
interj 2s older.brother =&co DST=MOB eagle big =ATTR =ACC
{kaw} =na {re'e -wa} =no =te
shoot =DAT go.away -FACT =QUOT =DCL
Oh! The company (group) of your elder brothers went there (in that direction)
to shoot the big eagle, it is said.

(380) otokoymu rupekba taw'reksorupparaba beblokparaba moma mathayaw
thokni tay'nido.
otokoymu [rupek] =ba [taw'reksorup] =para =ba

so.then frog =ADD banana.bird -&co =ADD
[beblok] =para =ba [moma mathay] =aw
toad -&co =ADD elephant bachelor.elephant =ACC
{thok -ni} [tay'ni] =do
hit -FUT today =TOP
So then the frog and the banana bird and his company and the toad and his
company are going to beat up the elephant today.
19.10 The alternative enclitic <=sega ~ =siga>
The two allomorphs of the alternative enclitic <=sega ~ =siga> (ALT) are in free
variation. However, the allomorph <=siga> predominates in the Badri area while the
allomorph <=sega> predominates in Siju. Apart from alternative marker on nouns,
this morpheme also occurs with the same meaning as event specifier on verbal and
adjectival predicate heads (see Table 68). The meaning of <=sega ~ =siga> (ALT) is
X in turn, the other, next.
The noun marked with the alternative often stands in a possessor-possessed
relationship with another noun which can be marked with the genitive case as in (381)
or be unmarked for case as in (382).


303
(381) morotmi morotsigaaw jomi josigaaw bay'sakrara kakrukok. sosami
sosigacina nawrukok tan'rukok.
[morot] =mi [morot] =siga =aw [jo] =mi [jong] =siga =aw
man =GEN man =ALT =ACC brother =GEN brother =ALT =ACC
[bay'sak] =rara {kak -ruk -ok}
friend =AMONG bite -RC -COS
[so sa] =mi [so] =siga =ci =na]
village one =GEN village =ALT =LOC =DAT
{naw -ruk -ok} {tan -ruk -ok}
scold -RC -COS slay -RC -COS
Fellow men and brothers fought with each other among friends. From one
village to the next [people] scolded each other and slew each other. Literally:
man [and] man in turn, brother [and] brother in turn.

(382) ucie sunibalsansegaci phetokno
ucie [sunibal san] -sega =ci {phet -wa} =no
then Saturday day -ALT =LOC arrive -FACT =QUOT
Then he arrived the day after Saturday, it is said.

The following example shows the only recorded occurrence of the alternative enclitic
on a postpositional phrase with the postposition konsa (see 13.2) This example also
illustrates the use of the alternative enclitic as an event specifier, i.e. predicate head
suffix, in this case on the predicate ray'a to come.

(383) umo konsasiga te'ew cancicopay dolsamo te'edo ray'asiganaka.
[u =mo konsa] =siga te'ew {cancicop} =ay [dol] =sa =mo
DST=GEN after =ALT now suppose =ADV group =MOB =GEN
[te'e] =do {ray'a -siga -naka}
now =TOP come -ALT -IFT
In turn after that, supposing that from the group now will come another
[person] in turn.
19.11 The additive/emphatic enclitic <=ba>
The additive/emphatic enclitic <=ba> (ADD/EMPH) indicates either addition or
emphasis, depending on the context. This enclitic functions on both phrases and
clauses. The difference between addition and enumeration is that addition highlights
the whole of the added constituents whereas in enumeration the whole is irrelevant. It

304
also marks speakers in speech report constructions. We will address all types of
occurrences separately below. The appearance of <=ba> (EMPH) in complex
predicates is treated in 22.6.1. This enclitic is homophonous with the indefinite
clausal enclitic <=ba> (INDEF), which occurs on locative-marked clauses (see
Chapter 27) and indefinite proforms (see Chapter 1)
19.11.1 Addition
The additive/emphatic enclitic is frequently found on two or more NPs in any
syntactic function in an emumeration. This can be NPs in the same clause as in (380)
or in different clauses as in (384). It is not obligatory to mark enumerated NPs with
<=ba> (EMPH) as we can see in (385). Juxtaposition of NPs or clauses can also
signal enumeration. Example (385) follows on from (384) in the original text.

(384) ucie bothoyba re'eok, macokba re'eok, magacakba re'eok, se'elba
re'eok, moma re'eok. sa'ak sanci maphek.
ucie [bothoy] =ba {re'e -ok} [macok] =ba {re'e -ok}
then porcupine =ADD go.away -COS barking.deer =ADD go.away -COS
[magacak] =ba {re'e -ok} [se'el] =ba {re'e -ok}
small.deer =ADD go.away -COS jackal =ADD go.away -COS
[moma] =ba {re'e -ok} {sa' -ak} [san] =ci ma =phek
elephant =ADD go.away -COS eat -COS day =LOC CLF.ANIMALS =DIS
Then the porcupine went, the barking deer went, the small deer went, the
jackal went, the elephant went. [The lion] ate one animal a day.

(385) maca re'eok, amak re'eok hu'raw re'eok, da'ra matan re'eok, jamok.
[maca] {re'e -ok} [amak] {re'e -ok} [hu'raw] {re'e -ok}
deer go.away -COS monkey go.away -COS gibbon go.away -COS
[dara mat] =an {re'e -ok} {jam -ok}
every animal =FC/ID go.away -COS complete -COS
The tiger went, the monkey went, the gibbon went, every animal went,
completed [i.e. all of them].
19.11.2 Emphasis
Examples of the enclitic <=ba> used as marker of emphasis are (28), where it is used
on a personal pronoun; (494), where it appears on a prototypical noun; and (38),

305
where we can see it used on a demonstrative. The following example illustrates the
use of the enclitic <=ba> (ADD/EMPH) in its function as emphasiser on a clause.

(386) ami joraaw cayna man'nima? man'niba.
[a =mi jora] =aw {cay} =na {man' -ni} =ma
1s =GEN lover =ACC see =DAT be.able -FUT =Q
{man' =ni} =ba
be.able -FUT =EMPH
Can I see my lover? You can indeed.
19.11.3 Marker of speaker
NPs referring to speakers are often marked by the enclitic <=ba> (EMPH), as is
illustrated in the following example, where the toad, beblok, is the speaker. Speakers
can also be marked with the focus/identifier enclitic <=an> (FC/ID), the topic enclitic
<=do> (TOP) or the focus enclitic <=e> (FC). More fieldwork is needed to find out if
there are any conditions that determine the appearance of different enclitics on NPs
denoting speakers.

(387) bisa ray'na bay'siga? noay so'aydoano beblokba.
[bi] =sa {ray'} =na [bay'siga]
QF =MOB go =DAT friend
{no}=ay {so' -aydoa} =no [beblok] =ba
say =ADV ask -PROG =QUOT toad =EMPH
Where are you intended to go, friend? sayingly asked the frog.
19.12 The focus/identifier enclitic <=an>
The focus/identifier enclitic can be used to uniquely identify a phrase or clause. The
border between unique identification and focus is difficult to define. It might be better
to say that the two functions go hand in hand and that one is more salient according to
the context and the type of phrase to which <=an> (FC/ID) is enclitisised. If, for
example, <=an> (FC/ID) is enclitisesed to a question word, it is for its focus function,
whereas on NPs headed by animate nouns, the enclitic is likely to be used for its
unique identification function. In (388) we see the focus/identifier enclitisesd to a
personal pronoun phrase. The focus/identifier enclitic always marks factitive-marked

306
clauses when they function as Comparee in an event comparison, of which example
(653) is illustrative.
In the context out of which the following example has been taken, the wild
animals are all running away in fear for the seemingly brave and strong lazy king,
when they meet a fox. The fox says that the wild animals do not need to be afraid of
the human and then continues saying:

(388) hay, a gana. aan raja aan balthumni.
hay [a] {gana} [a] =an {raja}
come.on 1s exist 1s =FC/ID king
[a] =an {bal -thum -ni}
1s =FC/ID speak -on.behalf.of -FUT
Come on! I am here. I am the king. I shall speak on your behalf.

Examples (389) and (390) are illustrative of the focus identifier enclitic on an
indefinite pronoun and a question word phrase respectively.

(389) thoymanaydonaka. atobaan takaydonaka.
{thoy -man -aydo -naka} [ato] =ba =an {tak -aydo -naka}
die -already -PROG -IFT what =INDEF =FC/ID do -PROG -IFT
He will almost certainly already be dying. [Someone] will almost certainly be
doing something [bad to him]. (thought the father whose son had run away).

(390) atokoyan jokaw halduna man'aydok?
atokoy =an [jok] =aw {haldun} =na {man' -aydok}
how =FC/ID spouse =ACC feed =DAT be.able -PROG
How can [he] feed his wives?

In (391) we see the focus identifier enclitic on a headless quantified NP.

(391) utom mo'korokan ha'kamarokno.
[u -tom] [mo' korok] =an [ha'] {kam -arok} =no
DST-ppp clf.humans six =FC/ID soil to.work -PROG =QUOT
They, the six of them, worked in the field.


307
In verbless clauses of the type PRONOUN-NOUN or DEMONSTRATIVE- NOUN the
focus/identifier enclitic marks a potentially modifying constituent. In a PRONOUN-
NOUN construction the pronoun is always understood as the possessor of the
following noun, unless the pronoun is marked by the focus/identifier enclitic <=an>
(FC/ID) as in (388). The phrase a raja (1s king) would always be understood as my
king. Alternatively the focus/identifier mopheme can also mark the predicate head in
a verbless clause of the type PRONOUN-NOUN provided that the pronoun phrase is
marked by the emphatic enclitic <=ba> (EMPH) or the topic enclitic <=do> (TOP).
A demonstrative usually modifies the noun it precedes, e.g. ie daba (PRX
coconut) this coconut. When the demonstrative phrase takes the enclitic <=an>
(FC/ID) it stops being a modifier and becomes a focused/identified NP on its own or a
predicate head. In verbless clauses (393) as well as in clauses with a verbal or
adjectival predicate like (394) or even when the predicate is nominalised like in (395),
when the demonstrative is not a modifier, or when it is a predicate head, marking it
with <=an> (FC/ID) is obligatory. This enclitic is not the marker of Copula Subject
per se in a verbless identity clause, as it can also occur on the predicate head.

(392) daba ian
[daba] {i} =an
coconut PRX =FC/ID
A coconut is this.

(393) ian badrimi oltho.
[i] =an {badri =mi oltho}
PRX =FC/ID Pname =GEN meaning
This is the meaning of Badri.

Keeping in line with the AOV/SV order of the language, the last constituent of the
clause is the predicate. The order of the elements in the clause can be permutated
according to topicality. It is the first constituent that is talked about and specified by
the second. This is according to the topic first principle.

(394) uawan gam moa.
[u] =aw =an [gam] [mo -a}
dst =ACC =FC/ID wealth call.a.name -CUST
That is called wealth.'


308
(395) ian balgabaawba jametarinaka.
[i] =an {bal} =gaba] =aw =ba {jam -et -ari -naka}
PRX =FC/ID speak =ATTR =ACC -EMPH complete -CAUS -SIMP -IFT
I will now just make this which is told (story) finish.

The focus/identifier enclitic also occurs in copula clauses, as we can see in (726),
repeated here as (396), and in (397)a), and in clauses which contain a single overt
constituent, i.e. a referent which also functions as predicate, e.g. (397)b).

(396) ue bihape cigacak te'ew kol india kolani hapan do'wacomnoa.
[ue bihap (<Garo) ]
CS
=e [cigacak]
CS
te'ew

DST place =FC Pname now
[kol india kolani hap]
CC
=an

{do' -wa} =com =noa
Pname place =FC/ID IE.be -FACT =IRR =QUOT
The place Chigachak is now supposedly the Coal India Colony place.

(397) a) ana daygabaan gana.
b) sihoan
a) [[a] =na {day} =gaba]
CS
=an {gana}
1s =DAT be.bigger=ATTR =FC/ID exist
b) {siho} =an
lion =FC/ID
a) There is one bigger than me.
b) [It]s the lion.

There is one example where the focus/identifier enclitic <=an> (FC/ID) marks the
Focus in a sentence which has an ellipsed A argument. It can thus be said that this
sentence contains a focused extraposed topic.

(398) bajolwamian khu'cuk olna sapjolariano.
[{ba -jol -wa} =mi]
FOCUS
=an [[khu'cuk] {ol} =na
be.born -QUICKLY -FACT =GEN =FC/ID language speak =DAT
{sap -jol -ari -a} =no
know.a.skill -QUICKLY -SIMP -CUST =QUOT
As far as [his] quick birth is concerned, [he] was just able to speak language
quickly, it is said.


309
NPs can be postposed to the predicate as afterthoughts or anti-topics. In narratives
postposed NPs almost always carry a focus, topic or focus/identifier enclitic, e.g.
(399). In colloquial speech, however, postposed NPs are almost always unmarked for
these categories.

(399) ca torucenaka? noaydoano magacakan. a torucena bay'siga. na'a
konsa torubone noaydoano pheruan.
[ca] {toru -naka} {no -aydoa} =no [magacak] =an
who take.bath -IFT say -PROG =QUOT deer =FC/ID
[a] {toru -ce} =na bay'siga [na'a] konsa {toru} =bo =ne
1s take.bath -FIRST =DAT friend 2s later take.bath =IMP =TAG
{no -aydoa} =no [pheru]=an
say -PROG =QUOT fox =FC/ID
Who will take bath first? [he] said, the deer. I will take bath first, friend.
You take bath later, ok. [he] said, the fox.

The focus/identifier anclitic <=an> (FC/ID) also occurs on relator nouns and
demonstratives used as temporal adverbs and conjunctions. In these cases it is attested
as co-occurring with the topic enclitic <=do> (TOP), e.g. (400). Example (401) shows
the focus/identifier morpheme on a question-word phrase.

(400) otokoy takaymu konsadoan khusi do'aymu ray'aphinwana []
otokoy {tak} =ay =mu [konsa] =do =an
like.that do =ADV =SEQ later =TOP =FC/ID
{khusi do'} =ay =mu {ray'a -phin -wa} =na
happy IE.be =ADV =SEQ come -back -FACT =DAT
Having done like that, being very happy because [he] had returned []

310
(401) biskonan jaci sa'aydoa? takaawba tiktika.
[biskon] =an [ja] =ci {sa' -aydoa}
how.much =FC/ID month =LOC eat -PROG
[taka] =aw =ba {tiktik -a}
money =ACC =EMPH save -CUST
How much are they eating in a month? [They] save the money.

Climaxes in event sequences are also marked by the use of the focus/identity
morpheme <=an>(FC/ID). In these cases it occurs on the discourse connective uci
then, as is illustrated in examples (402) and (403). Observing the use of the
discourse connectives uci then and uci=an (then=FC/ID) then, it seems that, for
many young speakers, there is no difference in meaning between the two forms.
Example (402) comes from a story about the deer and the fox who stole biscuits
from a banggal non-tribal Indian or Bangladeshi. Before eating the biscuits the deer
and the fox want to take a bath in a river. First they argue who will take a bath first. It
is decided that it will be the deer. The deer quickly finishes his bathing and then its
the foxs turn. The fox takes a long time to bathe and while he is in the water the deer
quickly eats as many of the biscuits as he can. When the fox comes out of the water,
the deer says that hes not yet clean and makes the fox bathe again and again. Every
time the fox goes into the water, the deer eats more of the biscuit loot and then the
storyteller says (402).

(402) ucian magacakdo biskutan man'ay man'nay man'ay sa'aydokno.
uci =an [magacak =do] [biskut =an] {man'} =ay
then =FC/ID deer =TOP biscuit =FC/ID in.great.amounts =ADV
{man'} =ay {man'} =ay {sa' -aydok} =no
in.great.amounts =ADV in.great.amounts =ADV eat -PROG =QUOT
Then the deer ate the biscuits in great great great amounts, it is said.

Example (403) comes from a story about a hanging root (type of vine) which changes
into an old lady at night. If you are unlucky enough to fall asleep under one of those
roots, the old lady will pester you with scratching her long arm all night. If you dont
do it, she will threaten to devour you. In the story a woman has been harassed by the
old hanging root lady for some nights in a row while her husband was away to the

311
market. In the end the husband comes back, takes the place of his wife under the
hanging root, dressed up in womens clothes, and cuts the arm of the old lady with a
sword. Then the storyteller continues with (403), the last sentence of the story.

(403) ucian manapci caywacido karadolsa thoy' cokcokay mu'aydono.
[uci] =an [manap] =ci {cay] =ci =do
then =FC/ID morning =LOC look =LOC =TOP
[karadol] =sa [thoy'] {cokcok] =ay {mu' -aydo} =no
hanging.root =DLIM blood drip =ADV stay -PROG =QUOT
Then, when [they] looked in the morning, the hanging root was hanging (lit.
staying) there dripping blood.
19.13 The topic enclitic <=do>
The topic enclitic <=do> (TOP) occurs on both phrases and clauses. It also occurs
in complex predicates with identical verbs (see 22.6.1). When it occurs on clauses
immediately after the change of state predicate head suffix <-ok ~ -ak ~ -k> (COS),
the topic enclitic will appear as the allomorph <=odo> (TOP). I analyse this enclitic
as a topic-marker for the following reasons:
it occurs on left-most clausal constituents which are topical, in the sense that the
they designate the referent that proposition is about (see Lambrecht 1994: 127-
128).
it occurs on locative-marked clauses and change of state-marked predicates with
conditional interpretation (see 27.5 and example (614) in 23.13 respectively, see
Haiman, 1978),
it occurs on right dislocated antitopic constituents, (see below),
In addition to these functions, the topic enclitic is found on time words in pre-clausal
position, functioning as discourse connectives, or in tail-head linkage constructions.
Time words in this environment are almost always topic-marked, e.g. (404).

(404) konsado jow'cana naokno.
konsa =do {jow' -ca} =na {na -ok} =no
later =TOP sleep -SUDDENLY =DAT need -COS =QUOT
Later, [he] suddenly needed to sleep, it is said.

312
In example (276) we see how the relative time words konsa after participates in a
tail-head linkage construction and is topic-marked.
It is not unusual for another a topic-marked constituent to occur immediately after
the topic-marked time word (405). Since I consider the time word to be in discourse
connective function, and therefore pre-clausal, the topic-marked constituent is still the
left-most constituent in the clause.
As was mentioned above, the topic enclitic also marks right dislocated
constituents as antitopic (see Lambrecht 1994: 117-205), of which examples (680)
and (405) are illustrative as well as Text 2 line 39, presented below as (406).

(405) konsado matsado morot son man'aimo ray'wilokno alsiado.
konsa =do [matsa]
TOPIC
=do [morot son] {man'} =ay =mo
later =TOP tiger =TOP human scent get =ADV =SEQ
{ray' -wil -ok} =no [alsia]
ANTITOPIC
=do
go -AROUND -COS =QUOT lazy.person =TOP
Later, the tiger, having smelled the scent of a human, walked around [him
j
], it
is said, [around] the lazy person
j
that is.

(406) dokom saa ado
[dokom]
S
{sa -a} [a]
ANTITOPIC
=do
head hurt -IMPF 1s =TOP
[My] head hurts, as far as im concerned.

Topic-marked clausal constituents can be indistinguishable from subjects when they
are in clause initial position and no other constituent is present that could be
interpreted as the subject, as we can see in examples (532), (585), (603). Non-subject
phrases can also be topic-marked, e.g. (577) and (360). Whereas the antitopics in
examples (405) and (406) are not coreferential with the subject of the clause, in (407)
the antitopic is coreferential with the implied subject of the clause, i.e. the deer.

(407) otokoymo jalathiriokno magacakdo.
otokoymo {jal -a -thiri -ok} =no [magacak]
ANTITOPIC
=do
so.then run.away -AWAY -AGAIN -COS =TOP deer =TOP
So then [he
j
] ran away again, the deer
j
.

313
Most clauses on which the topic enclitic occurs are subordinate. There are very few
cases of topic-marked main clauses and their predicates are all marked with the
factitive suffix <-wa> (FACT) (see Chapter 1) e.g. (666).
19.14 The focus enclitic <=e>
I have to stress that the label focus of the enclitic <=e> is preliminary, since it
occurs in several different pragmatic environments, which I will describe one by one.
Most salient is the use of this enclitic when new referents are introduced in the text or
discourse.
The enclitic <=e> marks new discourse topics conveyed by NPs that are not part
of the argument structure of the clause they precede, e.g. (408) which is the opening
sentence of a narrative about the history of the meaning of the place Badri Rongdyng
Hawai. The subject of the clause is an implicit third person plural. The NP Badri
Rongdyng Hawai is the pre clausal focus-marked NP referring to the new discourse
topic: the story will be about the meaning of Badri Rongdyng hawai.

(408) badri rodo ha'waymi oltoe, dakado cigacakcisa mu'wano.
[badri rodo ha'way =mi oltho] =e
Pname =GEN meaning =FC
daka =do [cigacak] =ci =sa {mu' -wa} =no
in.the.past =TOP Pname =LOC =DLIM stay -FACT =QUOT
As for the meaning of Badri Rongdyng Hawai, in the past, [they] lived in
Chigachak, it is said.

The focus enclitic marks NPs in topic position, i.e. clause initial, which are also
subjects, when these NPs refer to newly introduced referents. These referents do not
need to be topical. The following example, taken from Text 1 (line 20) is illustrative.
The speaker who produced this example, presented below as (409), was speaking
about something totally different, when he suddenly turns to Nongken and starts
speaking about the break up of Silat and Manchi, marking Silat with the enclitic <=e>
(FC). When we read the following lines in the text, we notice that the topic of the
conversation is clearly not Silat but Manchi.


314
(409) silate atakna manchiaw wat-ok?
[Silat] =e [atakna] [Manchi] =aw {wat -ok}
Pname =FC why Pname =ACC send.away -COS
Why did Silat send Mansi away? (i.e. Why did Silat break up with
Mansi?)

The focus enclitic is often encountered marking NPs in situations where antitopical
referents alternate, in other words the <=e> (FC) marks contrasting antitopics, e.g.
(410). The context of this example is as follows. The deer is pretending to be lame to
lure the Bengali away from the place where he has put down his load of biscuits. The
Bengali sees the deer walking slowly and chases it. But when the Bengali almost
catches up with it, the deer quickly runs away. What kind of deer is this? Now it
walks again as if it is lame.

(410) una roktheriokno bagale. tharapna guduk takwachiba tarakay jalariano
magachake. otokoymo bakrawraw bakrawraw jan'aoknoa. te'do uchian
pherudo biskutaw payay jalokno []. otokoymo te'do jan'aokno bagaldo,
man'anchaknoa. ian ja'naka nowachi te'do magachakdo [].
una {rok -theri -ok} =no [bagal] =e
then chase -AGAIN -COS =QUOT Bengali =FC
{tharap} =na {guduk tak -wa} =ci =ba
catch.up =DAT almost do -FACT =LOC =INDEF
{tarak} =ay {jal -ari -a} =no [magacak] =e
quick =ADV run.away -SIMP -CUST =QUOT deer =FC
Then [he] chased [it] again, it is said, the Bengali. Whenever he almost
caught up [with it], [it] ran away quickly, it is said, the deer.

Another example of the focus enclitic indicating a new topic constituent can be found
in (411). The elephant, moma, has already been introduced in the first sentence, but
not as a topic. Because the topic alternates in the second sentence, the new topic is
marked by the enclitic <=e> (FC).


315
(411) taw'reksorup masa ge'themo thup phanan moma phay'ay sa'rowana,
moma mathayaw tapna re'eaydoanoa. te'edo ue moma'ay rekci thupay
thupay mu'gabaaw phanan phay'ay phay'ay sa'roano.
[taw'reksorup ma sa] [ge'the thup] [phanan] [moma]
banana.bird CLF:ANIMALS one 3s nest always elephant
break =ADV eat -USUALLY -FACT =DAT
[moma mathay] =aw {tap} =na {re'e -aydoa} =no -wa
elephant bachelor.elephant =ACC hit =DAT go.away -PROG =QUOT -FACT
[te'e] =do [ue moma] =e [rek] =ci {thup} =ay {thup} =ay
now =TOP DST elephant =FC banana.tree =LOC to.nest =ADV to.nest =ADV
{mu'} =gaba] =aw {sa' -ro -a} =no
stay =ATTR =ACC eat -USUALLY -CUST =QUOT
An elephant always breaks and eats (lit. breakingly eats) a banana birds
nest (lit. his nest). Now as for the elephant, [he] always breaks the [thing]
that stays nestingly in the banana tree.

The focus enclitic also occurs on case-marked phrases, as in the following example,
where it occurs on an accusative-marked demonstrative phrase. In this example the
focus enclitic does not indicate the change of a referent or the introduction of a new
one. In this example the enclitic indicates that the phrase is in focus in the clause. The
preceding context translates as follows: Now in a village supposedly lives Bil, it is
said. Then, there is also a wife, it is said. Now, [Bil] is reluctant to do work and so
spends his time sleeping and sitting, it is said.

(412) maymamaaw sa'naan uawe gawigabasa khu'tipetna naphinano.
[may -mama] =aw {sa'} =na =an
rice -CONTINUOUSLY =ACC eat =DAT =FC/ID
[u] =aw =e [gawigaba] =sa {khu'tip -et} =na
DST=ACC =FC wife =DLIM shut.mouth -CAUS =DAT
{na -phin -a} =no
need -AGAIN -CUST =QUOT
Because he eats rice continuously, that one [i.e. Bils mouth], [his] wife has to
mouth-close again [all the time], it is said. (Because he is too lazy to do that
himself.)

The topic enclitic <=do> (TOP) and the focus enclitic <=e> (FC) cannot occur
together on the same NP, but both enclitics can co-occur in the same clause on
different NPs.

316
Chapter 20 Case Marking
_____________________________________________________________________

Transitivity plays a role in Atong. There are transitive-intransitive verb pairs, (see
Table 25 in 4.6). Intransitive verbs appear in constructions where O arguments are
not usually conceivable and transitive verbs are those that appear can appear in
constructions where O arguments are conceived or implied. There is a transitive
derivation of intransitive verbs and adjectives of Type 1 with the causative predicate
suffix <-et> (CAUS). A and S are never marked for case and O arguments can
optionally be marked with the accusative phrasal enclitic <=aw ~ =taw> (ACC). This
enclitic does not only mark Patients, but has a range of other functions as well (see
20.8), and therefore, not all accusative-marked NPs are O arguments. Whether an
unmarked NP can be interpreted as Agent, Actor or Undergoer (i.e. entity most
effected by the event denoted by the verb) depends on the context.
In this chapter I will talk about the syntactic, semantic and pragmatic factors that
condition case marking. Case assignment on NPs depends on the following factors:
the semantics of the verb
the semantics of the NPs
the semantics of the case marker
pragmatic factors
Atong distinguishes the following types of arguments: the core arguments A, S, E
(the third argument of an extended transitive verb), CS (copula subject), CC (copula
complement), O and adjuncts or peripheral arguments. No NP is obligatorily
expressed in Atong when it is retrievable from the context. Only the E argument of
the verb mo- to call something/someone a name, the name of a named entity,
cannot be ellipsed. The order of the NPs in the clause depends on semantic and
pragmatic conditions. When NPs are expressed in a clause, their case marking
depends on semantics and pragmatics and not on their syntactic role. Transitive
subjects (A arguments) and intransitive subjects (S arguments) are never marked for
case. Patients, Recipients, Beneficiaries, Directions and Possessors do not need to be
case-marked when their semantic role is clear from the context. NPs indicating a
Location in space and time, Source, Pathway, Goal or Standard of comparison are
20 CASE MARKING

317
obligatorily case-marked for these semantic functions. Table 57 gives an overview of
marked and unmarked arguments.

Table 57 Marked and unmarked syntactic and semantic argument types
ALWAYS
UNMARKED
MAY BE UNMARKED OBLIGATORILY MARKED
ARGUMENT ARGUMENT CASE MARKER ARGUMENT CASE MARKER
S O <=aw ~ =taw> (ACC) Location in
space and time
(Location in
time can only
be fulfilled by
locative
clauses)
<=ci> (LOC)
A Patient <=aw ~ =taw> (ACC) Source <=mi ~ =mo>
(ABL) Beneficiary <=na ~ -ona> (DAT)
Possessor <=mi ~ =mo> (GEN) Pathway <=tokoy> (VIA)
Causer Causee <=aw ~ -taw> (ACC) Goal <=ci=na>
(LOC=ALL)
Experiencer as the
argument of an
experiencer verb
Direction, if
expressed by an
inherently
locational noun
<=sa> (DIR) Instrument <=sa> (INSTR)
Comitative
adjunct
<=mu ~ =mu ~
mo> (COM)
Reason (can
only be fulfilled
by a reason
clause)
<=na> (DAT)
CS Recipient <=na ~ =ona> (DAT)
CC
Result
Direction, if
expressed by a
noun which is
not inherently
locational
<=sa> (DIR)
E = Name
Comparee
Duration
Agent
Actor Facsimile <=tokoy> (LIKE)
Dative marked O arguments
of verbs of emotion and interaction
<=na> (DAT)

Standard of
comparison in
comparative,
non-equative
constructions
<=na> (DAT)
Standard of
comparison
equative
constructions
<=mi ~ =mo>
(GEN)
Reason (can
only be fulfilled
by dative-
marked clause)
<=na ~ =ona>
(DAT)
Purpose (can
only be fulfilled
by dative-
marked clause)
<=na ~ =ona>
(DAT)
20 CASE MARKING

318
The semantic and syntactic role of any expressed unmarked NP is determined by
the
context
semantics of the verb
semantics of the NP
position of the NP in the animacy hierarchy
co-occurrence with other NPs
The language does not mark the syntactic role but only the semantic role of NPs.
What is marked as S, A and O are those NPs that are most likely to be interpreted as
such.
More fieldwork needs to be undertaken to find out whether A and S control
reflexivisation. Control over reflexivisation is not a universal property of these core
arguments. Though most of the time the controller and the reflexive appear in the
same clause, and it may therefore appear to always control reflexivity, it takes some
special contexts to see whether this is indeed the case. In Chinese, for example,
reflexives are pragmatically or semantically controlled (see LaPolla, 1993). The sole
property of O is that it can be accusative-marked when referential and definite. A
clause can occur with only one or with more case-marked NPs, e.g. (413) and (414).
In these examples the case-marked NP or NPs indicate primarily semantic role,
syntactic function being less important because all of the NPs are peripheral. A clause
can also occur with a mix of marked and unmarked NPs none of which is S or A and
the semantic and syntactic function of the unmarked NPs is determined by the factors
described above, e.g. (415).
In example (413) we see a clause with only one Location adjunct. This adjunct is
locative-marked to show its semantic function.

(413) sanci re'eni
[san]
LOCATION
=ci {re'e -ni}
day =LOC go.away -FUT
[We] will go during the day.

20 CASE MARKING

319
Inherently locational nouns can be mobilitative-marked, as in (414), but do not have
to be. When used in the right context, their meaning is enough to assign the semantic
role of Direction to them, as we see in (415).

(414) sanci nogolsa re'eni
[san]
LOCATION
=ci [nogol]
DIRECTION
=sa {re'e -ni}
day =LOC market =MOB go.away -FUT
[We] will go to the market during the day.

(415) sanci nogol re'eni
[san]
LOCATION
=ci [nogol]
DIRECTION
{re'e -ni}
day =LOC market go.away -FUT
[We] will go to the market during the day.

Since we can only determine the syntactic status of an NP after having determined its
semantic role, I argue that case marking in Atong determines primarily semantic roles
and that syntactic role indication is secondary. The accusative case, however,
indicates primarily a syntactic role but also has pragmatic functions treated below.
Case-marking on NPs helps to determine:
the semantic role of an NP in a clause,
the relationship between nouns within an NP,
the syntactic role of an NP in a clause.
The animacy hierarchy and identifier marking can help, too, to determine who does
what to whom. Because no NPs are obligatorily expressed in a clause, it is not always
possible to determine with certainty which NPs are obligatorily conceptualised, i.e.
are core arguments of a verb and which are not.
The identification of CS and CC in copula clauses is easy because of the limited
freedom in constituent order and the limited number of NPs possibly expressed in the
clause. In identity/equation copula clauses the constituent order is either CS-CC-
COPULA or CC-COPULA+ADDitional NP marked for topic, focus,
focus/identification or emphasis. CC and CS are always unmarked.
Case marking in Atong is done with phrasal enclitics. The accusative is frequently
found marking only one constituent, in a NP which consists of more than one
nominal. In these cases the case marker still has scope over the whole NP. All case
20 CASE MARKING

320
markers listed in Table 58 and their functions will be treated one by one in this
chapter. We will discuss the conditions for occurrence of a case marker as well as the
non-occurrence and repetition of it. In 20.10 we will see how Atong makes use of
multiple case marking. Finally, in 20.11 we summarise the motivations for repeated
case marking,

Table 58 The Atong case markers and the types of NPs they can mark.
20.1 Zero marking
A, S, E, CS, CC and Result (417) NPs are never marked for case. O arguments may
be left unmarked and of the obliques Recipients, Beneficiaries, Possessors and
Directioon adjuncts may be left unmarked. Whether or not these adjuncts receive case
marking depends on three factors, viz.
CASE MARKERS MEANING
SYNTACTIC
FUNCTION
SEMANTIC ROLE
none unmarked
A,S,CS,CC,O,
oblique
Agent, Patient,
Goal, Source,
Possessor,
Beneficiary,
Recipient,
Comparee
<=sa>
MOB/LOC/INSTR
mobilitative/locative/
instrumental
oblique
Goal, Source,
Location,
Instrument
<=ci>
LOC
locative oblique Location
<=mo ~ =mi>
GEN/ABL
genitive/ablative A, oblique
Agent, Possession,
Source
<=mu ~ =mu ~
=mo>
&CO
comitative A, S, oblique Comitative
<=na>
DAT/ALL
dative/allative O, oblique
Recipient,
Beneficiary, Goal,
Standard of
comparison,
Emotor, Purpose
(only clauses),
Reason (only
clauses)
<=aw ~ =taw>
ACC
accusative O
Patient, Material,
Road
<=tokoy>
VIA/LIKE
perlative/similative oblique Pathway/Facsimile
20 CASE MARKING

321
Type of NP
Context
Pragmatic factors, which differ per case

Inherently locational nouns can be Direction adjuncts without being marked for
case, as has been mentioned above. The higher a noun is in the animacy hierarchy the
more chance it has to be an unmarked Possessor. Animate nouns can be unmarked
Recipients or Beneficiaries.
In the next example the verb hon'- to give has four NPs, viz. two core arguments
A and O, and two adjuncts/peripheral arguments, viz. Recipient and Temporal
Location. Three of these NPs are unmarked for case. The context is as follows. The
animals have gathered to appoint a king. Nobody wants to be the king because they all
say that there is a greater animal than them who is more apt to be the king. When the
animals ask the lion, he says that he will only accept when the animals give him what
he asks in (416). In that clause all NPs belong to a different level on the animacy
hierarchy: aa (1s) > na'tom (2p) > ma classifier for animals. Furthermore there
is no ambiguity as to who gives what to whom because everybody is supposed to
know that lions eat animals.

(416) aca na'tome aa sanci mapek hon'ni nowano.
aca [na' -tom]
A
=e [aa]
RECIPIENT
[san] =ci
interj 2s -ppp =FC 1s day =LOC
[ma]
O
=pek {hon' -ni} {no -wa} =no
CLF:ANIMALS =DIS give -FUT say -FACT =QUOT
Right then, you
p
shall give me one of each animal every day, [the lion] said,
it is said.

The next example shows a Result adjunct that is unmarked for case.

(417) a bu'cotaw phaktham kan'ni
[a] [bu'cot]
PATIENT
=aw [phak tham]
RESULT
{kan' -ni}
1s mango =ACC longitudinal half three cut -FUT
I will cut the mango in three pieces.

20 CASE MARKING

322
Let us now explore the different case markers, their functions, semantics and the
conditions under which these markers are applied to an NP.
20.2 The mobilitative/locative/instrumental case marker <=sa>
Depending on the context, the mobilitative/locative/instrumental case enclitic <=sa>
(MOB/INSTR) marks adjuncts that are semantically Directions, e.g. (428), Locations
and Instruments, e.g. (429). Although they are expressed with the same morpheme,
there are functional differences between the mobilitative and the instrumental case.
NPs expressing an instrument are obligatorily case-marked whereas mobilitative
marking can be omitted on NPs headed by an inherently locational nouns. As
instrumental marker <=sa> (INSTR) is not attested in combination with other case
markers, whereas case stacking is possible on a Direction adjunct (see section 20.10).
The different functions of the enclitic <=sa> (MOB/LOC/INSTR) will be
exemplified one by one.
20.2.1 Mobilitative interpretation
When the enclitic <=sa> (MOB) marks a Direction adjunct it indicates movement.
The movement can be from a source (418), to a destination (418), (419), or in a
certain direction (423). The mobilitative enclitic, however, does not indicate the
direction of the movement. The direction of the movement is most often made clear
by the context and by the form of the verb of movement. It is possible to stack the
genitive/ablative case <=mo ~ =mi> (GEN/ABL) (with ablative interpretation in
these cases) onto the mobilitative to unambiguously express movement away from
somewhere as in (420). This method is rarely practised in colloquial speech.
Mobilitative marking can be omitted on inherently locational nouns (421) but can be
added when needed for emphasis (418) or contrast as in (422).
In example (418) the mobilitative designates the noun tura Tura either as Source
or as Goal depending on the context.

(418) turasa re'ewa.
[tura] =sa {re'e -wa}
Pname =MOB go.away -FACT
[I] left from Tura ~ [I] come from Tura. ~ [I] went to Tura (but now Im
back). [Somebody] went to Tura (and is still gone). etc.
20 CASE MARKING

323
In (419) here below, the fact that the noun so village marked by the mobilitative is
a Goal is disambiguated by the future form of the predicate head.

(419) sosa re'eni
[so] =sa {re'e -ni}
village =MOB go.away -FUT
[I] will go to the village.

Example (420) here below illustrates the morpheme <=mi ~ =mo> (GEN/ABL) as
marker of a Source adjunct, i.e. in ablative function, stacked onto a mobilitative case
in the underlined NP phoren=sa=mi. The nouns in the NP phoren=mi morot
(foreign.country=GEN person) can be interpreted as being in a Possessor-Possessed
relationship and thus the morpheme <=mi> is glossed as genitive.

(420) phorenmi morot ray'adoa, phorensami ray'aydoa.
[phoren =mi morot] {ray' -aydoa} [phoren]
SOURCE
=sa =mi
foreign.country =GEN person come -PROG foreign.country =MOB =ABL
{ray'ay -doa}
come -PROG
Persons belonging to foreign countries are coming, [they] are coming from
foreign countries.

In example (421) the Direction adjunct is an unmarked inherently locational noun,
jadi. This example contrasts with (422), where the mobilitative is used. The inherently
locational nouns jadi Jadi in (422) is presumably mobilitative-marked because the
speaker needs to distinguish between to and from. The source in (422) is
indicated only with the genitive.

(421) sunibal sanci jadi re'ewano biphagabae.
[sunibal san] =ci [jadi] {re'e -wa} =no [biphagaba =e
Sunday day =LOC Pname go.away -FACT=QUOT husband =TOP
On Sunday the husband went to Jadi.

20 CASE MARKING

324
(422) te'ewba roo ha'waymo morot mo' sa [] dasawaymo ca'aw
re'eaydonano jadinogolsa.
[te'ew] =ba [[rodo ha'way] =mo [morot mo' sa]]
S

now =EMPH Pname =GEN person CLF:HUMANS one
[dasaway]
SOURCE
=mo [ca'] =aw {re'e -aydoa}
Pname =ABL foot =ACC go.away -PROG
[jadi nogol] =sa
Pname market =MOB
Now then, a person from/belonging to Rongdyng Hawai went from
Dangsawai on foot to Jadi market.

The following example comes from a story in which the context makes clear that the
persons mentioned were first living in a place called Songmong. At a certain moment,
described in example (463), the people start running away from there. Thus the
mobilitative-marked place Durama can only be interpreted as the direction in which
these people go and the place Gandyrung as the limit.

(423) aro cesonmatra renadolkhajosikjebal asegosane durama'sa
gandorucina jaltawokno.
aro [cesonmatra renadolkha josikjebal asegosan] =e
and Name Name Name Name =FC
[durama']
DIRECTION
=sa [gandoru]
LIMIT
=ci =na
Pname =MOB Pname =LOC =DAT
{jal -taw -ok} =no
run.away -UPWARD -COS =QUOT
And Chengsynmatrang, Rynangdylkhang, Jongsikjengbal [and]
Asenggongsan run away upward in the direction of Durama up till Gandrung,
it is said.

The mobilitative can be used to indicate a Direction, both concrete, as in (424), and
abstract as we can see in example (425).

(424) Speaker: cabi hon'bo (Giver makes a mistake and gives the key to the wrong
person) Speaker, irritated: asa hon'bo!
[cabi] {hon'} =bo [a]
DIRECTION
=sa {hon'} =bo
key give =IMP 1s =MOB give =IMP
Give the key. Give the key to me! Literally: Give the key in my direction!

20 CASE MARKING

325
(425) otokoymu ucie nay'nokholthagabado konokholthagasa nukaymuna
cayancaknoay. baratokno.
otokoymo [u] =ci =e [nay'nokhol] =tha =gaba =do
so.then DST=LOC =FC mother-in-law =OWN =DREL =TOP
[konokhol -tha =ga] =sa {nuk} =ay =muna
son-in-law -OWN =DREL =MOB see =ADV =SEQ
{cay -an -ca -k} =no =ay {barat -ok} =no
look -REF -NEG -COS =QUOT =POS be.ashamed -COS =QUOT
So then, at that moment, his own mother-in-law having thrown a look in the
direction of her own son-in-law, she stopped looking, it is said. She was
ashamed, it is said.
20.2.2 Locative interpretation
I think that, historically, this morpheme comes from the noun sa meaning place,
side and is still found with that meaning in a few compounds, e.g. saphak ~
samphak side. Thus it is not surprising that the enclitic <=sa> (MOB/LOC/INSTR)
can sometimes still be used as a locative, i.e. marking Location adjuncts, as is
illustrated in examples (426) and (427). Line 23 from Text 3 is represented below as
(426). In this example no movement of the frog is implied; the animal is just sitting at
the bottom of the ravine and is making a noise.

(426) uci rupeke hoyawe ro'ka otokno'sa pekpek pekpek noay
parawaydoknowa.
uci [rupek] =e [hoyawe ro'ka otok no'] =sa
then frog =FC yonder cliff bottom.of.ravine inside =LOC
[pekpek pekpek] {noay} {paraw -aydok} =no -wa
frog.sound frog.sound say make.animal.sound -PROG =QUOT -FACT
Then the frog way over there at the bottom of the ravine is calling pekpek!
pekpek!, is said.
20 CASE MARKING

326
(427) otokoymudo uan, ge'the, sagaba: ado nemkhalanca nocido aro kamalsa
thama caithiria.
otokoymudo [u] =an [ge'the] [sa =gaba]
so.then DST=FC/ID 3s be.ill =ATTR
[a] =do {nem -khal -an -ca} {no} =ci =do
1s =TOP good -CP -REF -NEG say =LOC =TOP
[aro kamal] =sa {thama cay -thiri -a}
other priest =LOC divination look -AGAIN -CUST
So then, as for that sick [person], if [he] says: I am not better, [they] will
practice divination again at the place of another priest.
20.2.3 Instrumental interpretation
Examples (428) and (429) here below are illustrative of the use of the morpheme
<=sa> (INSTR) as instrumental case marker.

(428) a rong'sa depowaw ratwa
[a] [ro'] =sa [dopow] =aw {rat -wa}
1s stone =INSTR snake =ACC hit -FACT
I hit the snake with a stone

(429) a ie biskutaw takasa ra'wa
[a] [ie biskut] =aw [taka] =sa {ra' -wa}
1s PRX biscuit =ACC money =INSTR get/buy-FACT
I bought the biscuits with money.

The instrumental is also used with verbs of speaking to mark the language that one
speaks (430), (431) or that one speaks a certain language using the word khu'cuk
language (432).

(430) [] ha'cok khu'cuksa mokha badri mooy mowano.
[ha'cok khu'cuk] =sa [mokha badri]
Garo language =INSTR rain heavy.incessant.rain
{no} =oy {mo -wa} =no
say =ADV call.a.name -FACT =QUOT
[Because they held a drinking competition with the rain in that place], [they]
sayingly call [it] mokha badri in Garo, it is said.

20 CASE MARKING

327
(431) atosa balcido so pidan do'acom.
[ato] {bal} =ci =do [so pidan] {do' -a} =com
Atong speak =LOC =TOP village new IE.be -CUST =IRR
If you would say it in Atong, it would be so pidan.

(432) khu'cuk abunsa balcido ble payri noay balna man'nicom.
[khu'cuk abun] =sa {bal} =ci =do [ble payri]
language other =INSTR say =LOC =TOP blank firing
{no} =ay {bal} =na {man' -ni} =com
say =ADV say =DAT be.able -FUT =IRR
If you would say it in another language, you could sayingly say ble payri
(< English blank firing).
20.3 The locative case marker <=ci> (LOC)
The locative case <=ci> (LOC) marks oblique arguments (adjuncts/peripheral
arguments). Both temporal (433) and spatial (434) locations are marked by the
locative case.

(433) sanci re'eni
[san] =ci {re'e -ni}
day =LOC go.away -FUT
[We] will go during the day.

(434) de'the pipukci ganakhua mo' sa.
[de'the pipuk] =ci {gana -khu -a} [mo' sa]
3s belly =LOC exist -INCOM -CUST CLF:HUMANS one
In her belly she had one [child] more.

The locative case can be omitted from NPs with generic time reference, whereas NPs
with specific time reference need to be locative-marked. The following two examples
form a contrastive pair. In (435) the locative <=ci> (LOC) on somay time is omitted,
whereas in (436) it is present. The time reference in (435) is generic whereas in (436)
it is precise and topical.

20 CASE MARKING

328
(435) rodomi oltoe daka somay jaksoram maca nok phandaymi macamu rodo
maharimu takrukwanoa.
[rodo =mi oltho] =e [daka somay] [jaksonram maca
Pname =GEN meaning =FC in.the.past time Pname tiger
nok phanday =mi maca] =mu [rodo mahari] =mu
house bachelor =GEN tiger =COM Pname clan =COM
[takruk -wa] =noa
fight -FACT =QUOT
As for the meaning of Rongdyng, in ancient times, the tigers of the tigers
bachelor house of Jaksonram fought with the Rongdyng clan, it is said.

In example (436) here below we see the concomitant action clause uci mu'butu
when [they] stay there. The clause functions as a temporal adjunct in the main
clause. The concomitant action suffix is compounded with the noun somay time.
This phenomenon is treated extensively in 27.6.

(436) uci mu'butu somayci badri nemen man'ay sa'ano.
|[u] =ci {mu' -butu +somay}| =ci
DST =LOC stay -WHILE +time =LOC
[badri] nemen {man'} =ay {sa' -a} =no
At the time of [their] stay there, Badri was very rich. Literally: Badri ate in
great amounts.

Inherently locational nouns also need to be locative-marked when they function as
locative adjuncts in a clause (437). An inherently locational noun can function as S
argument and is then unmarked as in example (438). Since the predicate mu'- stay
of the attributive clause is not a verb of movement, interpretation of the place name
Siju as a movement oblique without mobilitative marking is impossible.

(437) dakado cigacakcisa mu'wano.
[daka] =do [cigacak] =ci =sa {mu' -wa} =no
before =TOP Pname =LOC =DLIM stay -FACT =QUOT
In the past [they] just lived in Chigachak.

20 CASE MARKING

329
(438) te'ew sijow so mu'aydogaba bihapci
[te'ew] [sijow so] {mu' -aydo} =gaba] [bihap] =ci
now Pname village stay -PROG =ATTR place =LOC
at the place where the village of Siju is now situated

There are some verbs that take their third argument in the locative case, e.g. caduk-
to bump (intransitive)(439) and thot- to hit (intransitive) (440).

(439) dokom bolbanci cadukwa
[dokom] [bolban] =ci {caduk -wa}
head crossbeam =LOC hit -FACT
[I] bumped [my] head against the wooden beam above the door.

(440) cak ro'ci thotwa
[cak] [ro'] =ci {thot -wa}
foot stone =LOC hit -FACT
[I] hit [my] foot on a stone
20.4 The genitive/ablative/nominaliser case marker <=mi ~ =m>
The two allomorphs of the genitive /ablative/nominaliser morpheme <=mi ~ =mo>
(GEN/ABL/NR) are in free variation although the allomorph <=mo> is used mostly
in the Badri area and <=mi> more in Siju. The genitive has a number of functions in
Atong. 1) It indicates the relationship between nouns within an NP. 2) The genitive
enclitic marks adjuncts which indicate a Source. This is the ablative function of the
morpheme. 3) The genitive marks the standard of comparison in equative
constructions. 4) The genitive appears as a nominalising enclitic on clauses of which
the predicate carries the factitive suffix <-wa> (FACT) (see also 24.3.1). 5) The
genitive marks nominal obliques which are the complements of the postpositions
<gomon> (REASON), <konsa> later, after and <dabat> (LIMIT). Example (420)
shows types 1 and 3. Examples of 5 can be found in Chapter 1. We will now explore
1, 2, 3 and 4 separately below.

20 CASE MARKING

330
20.4.1 Indication of the relationship between nouns within an NP
Within an NP the genitive marks the possessor in a Possessor-Possessed relationship
between nouns, e.g. (441).

(441) baydam rosa thoykhalmi ha'waycina jalaok.
[baydam] [rosa thoykhal =mi ha'way] =ci =na
some.people RIVERname river =GEN plain =LOC =ALL
{jal -a -ok}
run.away -AWAY -COS
Some ran away to the plains of the river Rongsa.

The first and second person pronouns have a long and a short form (see 17.2), viz.
aa ~ a (1s), na'a ~ na' (2s) and ,very rarely encountered, nia (1p) the long form of
the usual form ni (1p).When the short form of personal pronoun is followed by a
noun it will almost always be interpreted as a Possessor-Possessed relationship and
the personal pronoun need not to be marked, e.g. (434), (442). An example of a
context in which this interpretation is unlikely is found in (428), where the stone is
non-referential and not possessed. The Possessor-Possessed interpretation is also not
likely when the noun following the personal pronoun is compounded to the verb and
thus not an argument. The Possessor-Possessed interpretation is not possible with the
long forms of the personal pronouns, which are inherently A/S (or, in the case of na'a
(2s), used as address term), e.g. (443).

(442) phanan ni nokaw thokroa.
[phanan] [ni nok] =aw {thok -ro -a}
always 1p house =ACC destroy -USUALLY -CUST
[The elephant] always destroys our house.

In the next example (443) the pronoun aa (1s) is a separate NP. If it were in the form
without final /a/, it would be interpreted to be the possessor of sona gold in the NP
a sona (1s gold) my gold. Note that it is perfectly normal for the allomorphs a
(1s) and na (2s) to function as A/S argument as is illustrated in (429).

20 CASE MARKING

331
(443) [...] aa sonamo bi'camcom ician ha'mamo piriaymo nok rophiwaci
co'cetnagaba gana.
[aa] [sona =mo bicam] [i] =ci =an [ha'ma] =mo
1s gold =GEN piece PRX =LOC =FC/ID soil =COM
{pirin} =ay =mo [[nok] {co'cetna} =gaba] {gana}
to.mix =ADV =SEQ house to.glitter =ATTR EXIST
after I mix golden pieces with soil now, when [I] plaster my house [I] will
have a shiny one
20.4.2 Marker of a Source
A source can be marked with both the mobilitative and genitive/ablative case, as in
(420), or solely with the genitive/ablative case, as in (422) above and (444), (445) and
(446) below. Case marking of a Source is obligatory.

(444) nokmi hokotangbo!
[nok] =mi {hokot -a} =bo
house =ABL to.exit -AWAY =IMP
Exit away from the house! Alternatively: Go outside! (Said when both the
speaker and hearer where inside.)

(445) imi ang wal' nuketca.
[i] =mi [wal'] {nuk -et -ca}
PRX =ABL fire see -CAUS -NEG
From here I do not see the fire at all.

(446) na, aa otokoy colie colisemcaaydok noaymo te'ewba jokmo jalaokno.
na [aa] [otokoy] {coli =e coli -sem -ca -aydok}
interj 1s like.that succeed =FC succeed -CERTAINLY -NEG -PROG
{no} =ay =mo [te'ew] =ba [jok] =mo {jal -aydok} =no
say =ADV =SEQ now =EMPH spouse =ABL run.away -PROG =QUOT
Nah! like this, as far as succeeding is concerned, Im not succeeding, [the
lazy king] said and now he is running away from his wives.
20 CASE MARKING

332
20.4.3 Marking of the standard of comparison in equative clauses
The genitive marks the standard in equative clauses as illustrated here below.

(447) a na'mi hapsan cua
COMPAREE STANDARD MARK INDEX PARAMETER
[a] [na'] =mi hapsan {cu -a}
1s 2s =GEN same big -CUST
I am as big as you.
20.4.4 Nominalisation
I define nominalisation as a derivational process of which the resultant forms can
function as head of an NP. The morpheme <=mi ~ =mo> occurs as a nominalising
enclitic on clauses of which the predicate is marked by the factitive suffix <-wa>
(FACT). When it occurs as a nominaliser, the morpheme <=mi ~ =mo> will be
labeled (NR). The result of clausal nominalisation by means of the
genitive/nominaliser is an action/state or object nominalisation, depending on the
context. It is important to note that factitive-marked predicates are only attested as
heads of NPs in very few cases, which I consider to be lexicalisations of factitive-
marked predicates (see 24.3.1). Genitive-marked clauses can always function as head
of a predicate. Deriving NPs from clauses of which the predicate is factitive-marked is
a fully productive process.
Examples (448) and (449) illustrate action/object nominalisations. In (450) the
nominalisation can only be interpreted as an object nominalisation.

(448) ami balwami ician jametwa.
[|a =mi bal -wa| =mi] [i] =ci =an {jam -et -wa}
1s =GEN talk-FACT =NR PRX =LOC =FC/ID end -CAUS -FACT
I will end my talking/talk here.

20 CASE MARKING

333
(449) morot thoywami somayci atongtakay na'nangae taknaka?
[|morot thoy -wa| =mi somay =ci]
person die -FACT =GEN time =LOC
[ato] =takay [na'naa] =e {tak -naka}
what =LIKE 1pi =FC do -IFT
At the time of a persons death/dying, how will we do [it]?

(450) pheru nuksegaakno sa'wamiaw.
[pheru]{nuk -sega -ak} =no [|sa' -wa| =mi] =aw
fox see -ALT -COS =QUOT eat -FACT =NR =ACC
This time the fox saw [it], it is said, the food.

In the next example we see a state nominalisation.

(451) [] ni atodo dakado mamo thoromaw ni'wami somaycido wayaw mania.
[ni ato] =do [daka] =do
2p Atong =TOP in.the.past =TOP
[[|mamo thorom =aw ni' -wa| =mi ]somay] =ci =do
nothing religion =ACC not.exist -FACT =NR time =LOC =TOP
way =aw mani -a
spirit =ACC worship -CUST
[] we the Atongs, in the past, in a time when religion did not exist, [we]
worshiped the spirits.
20.4.5 Repeated genitive case marking
Genitive case marking is repeated when nouns are in an additive relationship within
an NP as illustrated in the next example. In that example acu ambi (grandfather
grandmother) ancestors is a fixed collocation.

20 CASE MARKING

334
(452) [] sansa dotamanca thariaysa kamalna rokaysa way khuruta, ni acu
ambimi nimi pi'sacido.
[san sa] [dothamanca] {thari} =ay =sa [kamal] =na
day one special prepare=ADV =DLIM priest =DAT
{rok} =ay =sa [way] {khurut -a}
search =ADV =DLIM spirit summon.a.spirit -CUST
[[ni acu ambi] =mi [ni'] =mi pi'sa]] =ci =do
1pe grandpa grandma =GEN 1pe =GEN childhood =LOC =TOP
[and] one special day [they] prepare [stuff], search a priest and perform the
incantation to summon a spirit, in the childhood of our forefathers and of us.
20.5 The comitative case marker <=mu ~ =mu ~ =m>
The allomorphs of the comitative morpheme <=mu ~ =mu ~ =mo> (COM) are in
free variation, although <=mo> is predominantly used in Badri and <=mu ~ =mu>
are predominantly used in Siju. The comitative marks a relationship between nouns in
different NPs which are in a comitative or additive relationship. When the nouns in a
comitative or additive relationship at phrase level and at inter-NP level are marked for
comitative case, all of them are always marked, not just one of them. When two nouns
are juxtaposed, they need not be marked for case at all to receive an additive
interpretation (see 6.6). A noun in a comitative relationship can be ellipsed when
retrievable from the context. Comitative marked NPs can be in A, S and oblique
function in the clause. Comitative marked nouns in O function are not attested. The
comitative functions only at NP level in a clause. In example (453) here below we see
comitative-marked nouns which function as A arguments in the clause.

(453) te'ewba magacakmo lukwakmoe khen' ra'aknote.
[te'ew] =ba [magacak]
A
=m [lukwak]
A
=m =e

now =EMPH deer =COM toad =COM =FC
[khen']
O
{ra' -ak} =no =te
crab get -COS =QUOT =DCL
Now the deer and the toad got (river) crabs, it is said, Im telling you.

The following example illustrates how the comitative case marks nouns belonging to
different oblique NPs in a clause. The example also illustrates double case marking
with the dative and the comitative. The dative enclitic <=na> (DAT) marks the NPs
20 CASE MARKING

335
jahas ship and bagaji fortune-teller as Recipients and the nominalised dative-
marked clause as a Purpose adjunct. The comitative indicates the relation between the
NPs. The dative marking is repeated because the NPs are enumerated.

(454) jahasnamu, ca ronamu, bagajinamu rajani taka jamok.
-Purpose clause--
[jahas] =na =mu | [ca] {ro} | =na =mu
ship =DAT =COM tea drink =DAT =COM
[bagaji] =na =mu [raja ni taka] {jam -ok}
fortune-teller =DAT =com hundred two money/rupee finish -COS
On the ship (for the benefit of the ship), on drinking tea, on the fortune-teller
the two hundred rupees were all spent [lit. were finished].

The comitative marks A and S arguments. Example (455) shows how the comitative
can appear only once in an NP and clause. In this example one of the constituents of
the A argument of the reciprocal verb ol-ruk- (speak-RC) speak to each other is
ellipsed as it is clear from the context. The two nouns in additive relation in the stated
A NP are not in a comitative relation to each other, but are together in a comitative
relation to the ellipsed A argument. Hence only one of the constituents of the NP is
comitative-marked with the marker having scope over the whole NP.

(455) mama manithagamuba olrukancakno.
[mama mani]
A
=tha =ga =mu =ba

elder.brother mother-in-law =own =DREL =COM =EMPH
[He and] his elder brother [and] his mother-in-law did not speak to each other
any more.

In some cases, where the constituents of a comitative NP are doing something
amongst themselves, the alternative is not used on both nouns as in the next example
in which the comitative and the alternative mark the constituents of an S argument.
The two nouns belong to the same NP, which is focused with the focus marker <=e>
(FC) attached to the second noun in the clause.

20 CASE MARKING

336
(456) morotmo morotsega'e takrukaydono.
[[morot] =mo [morot] -siga] =e
S
{takruk -aydo =no
person =COM person -ALT =FC fight -PROG =QUOT
One person and another person are fighting, it is said.

The above examples all show comitatives used in coordinative constructions.
Examples (435), (443) and (748) are good exmples of basic comitative-marked
participants in non-coordinative constructions.
As was mentioned above, an additive relation between nouns does not have to be
marked with the comitative morpheme <=mu> (COM), but can be obtained by simple
juxtaposition of two or more unmarked nouns. This is also the case with two
juxtaposed unmarked personal pronouns. Example (457) here below illustrates an
unmarked additive relation between two personal pronouns within the same NP. Note
that Atong has a separate pronoun for the first person plural inclusive, viz. na'na.
Note also that the combination pronoun-plus-noun is interpreted as a possessive
relationship with the pronoun as the Possessor, e.g. a gawi (1s girl) my girl.

(457) re'eni na' a.
{re'e -ni} [na' a]
go.away -FUT 2s 1s
[We] will go, you [and] I.
20.6 The dative/allative case marker <=na>
The morpheme <=na> will be labelled as allative (ALL) when it marks Goal adjuncts,
and will be labelled dative (DAT) in all other cases. The dative case marks the
following NPs:

(A) Adjuncts,
(B) Clauses in adjunct function to any other verb (see Chapter 27).
(C) Complements of the postposition daka before, earlier, in the past (see
27.3),

The morpheme <=na> (DAT/ALL) marks NPs which are either Beneficiaries (458),
Recipients (459), Goals, Experiencer (460) or Standard of comparison in comparative
20 CASE MARKING

337
clauses (461), (462). Goal marking involves the locative case and is treated under
multiple case marking in section 20.10. Example (416) above illustrates an unmarked
Beneficiary/Recipient.

(458) hanep aa na'na golpho balni.
hanep [aa]
A
[na']
BENEFICIARY
=na [golpho]
O
{bal -ni}
tomorrow 1s 2s =DAT story tell -FUT
Tomorrow I will tell a story for you.

(459) na'a ana hon'cakma?
[na'a]
A
[a]
RECIPIENT
=na {hon' -ca -k} =ma
2s 1s =DAT give -NEG -COS =Q
You will not give [some] to me any more?

(460) ana ato khu'cuk nemen raka.
[a]
EXPERIENCER
=na] [ato khu'cuk] {nemen rak -a}
1s =DAT Atong language very difficult -CUST
To me the Atong language is very difficult. Alternatively: I find the Atong
language very difficult.

(461) abun sona dayay ie so hansekhala.
STANDARD ------------MARK------------ COMPAREE PARAMETER INDEX
[abun so] =na {day} =ay [ie so] {hanse -khal -a}
next village =DAT be.bigger=ADV PRX village beautiful -CP -CUST
The next village is much more beautiful than this village.

(462) a na'na cukhala.
COMPAREE STANDARD MARK PARAMETER INDEX
[a] [na'] =na {cu -khal -a}
1s 2s =DAT big -CP -CUST
I am bigger than you.
20.7 Repeated dative case marking
Dative case marking is found to be repeated on enumerated nouns in a coordinated
structure which are all NPs of the same verb, as in (463), where we see a sequence of
three nouns marked by the dative. The dative-marked nouns all belong to different
NPs which are coordinated.

20 CASE MARKING

338
(463) ucisa macana makbulna momana paycaaymu boldobolda jalna
ha'baceok.
ucisa [maca] =na [makbul] =na [moma] =na {pay -ca} =ay =mo
then tiger =DAT bear =DAT elephant =DAT bear -NEG =ADV =SEQ
[boldobolda] {jal} =na {ha'bace -ok}
all.over.the.place run.away =DAT begin -COS
Then, not bearing the tigers, the bears and the elephants any more, [they]
stared to run away all over the place.
20.8 The accusative case <=aw ~ =taw>
The allomorph <=taw> (ACC) is used when the NP it attaches to ends in /t/. The
allomorph <=aw> (ACC) is used everywhere else. The accusative marks a variety of
NPs, viz.
Patients (syntactic O arguments), treated in 20.8.1,
Materials of which some artefact is made, treated in 20.8.2,
the word ram road as Pathway of the verbs ray'- to go, ray'a to come and
re'e to go away,
the word ca' foot/leg is used as an instrument, e.g. in the expression ca'=aw
re'e- (foot=ACC go.away), e.g. (422) and (492),
Moreover, the morpheme <=aw ~ =taw> (DREF) can be used purely to mark
referentiality, individuation or definiteness on NPs that are fragments, as we will see
in 20.8.3. In this case the label (DREF) definite & referential is used. Fragment
NPs are not predicative, as in presentative clauses (see 26.4) and are not part of a
clause as argument, adjunct, nor are they topics. Finally, the morpheme <=aw ~
=taw> (DREF) is found on clause initial NPs that cannot be interpreted as Patient or
O argument. This use will be treated in 20.8.4.

(464) otokoy ray'sotwae ray'mangabaaw ray'thirithiri mu'na naok.
otokoy {ray' -sot -wa} =e [{ray' -man} =gaba] =aw
like.that -go -DIRECTLY -FACT =FC go -ALREADY =ATTR =ACC
[ray'-thiri -thiri] {mu'} =na {na -ok}
go -AGAIN -RED stay =DAT need -COS
Like that [they] went directly [and they] had to keep taking (lit. going) [the
road] which they had already gone again and again.
20 CASE MARKING

339
20.8.1 The marking of O arguments
An O argument does not have to be case-marked. There are certain pragmatic
conditions under which an O argument receives accusative marking. There conditions
are summed up in Table 59.

Table 59 Pragmatic conditions for accusative case-marking of O
A) position of the referent on the animacy hierarchy relative to any other NP
or NPs in the clause or in the context
B) ambiguity about the semantic status or affectedness of the argument
C) referentiality

A) RELATIVE POSITION OF THE REFERENT IN THE ANIMACY HIERARCHY. The higher
an O argument is in the animacy hierarchy, the more likely it is to be accusative
marked. This pragmatic criterion is connected to the ambiguity criterion treated below
under B. The clause in the next example preceded the clause in (416) in the story of
the Lion and the Fox. Here below in (465) we find the only argument siho lion
marked with the accusative. If there were no case marking the lion, being high in the
animacy hierarchy, could be misinterpreted as the agent, as the one who did the
appointing. Apart from that the lion is the new topic of the story. So here two criteria
together are responsible for the accusative marking of the O, viz. position of the
referent on the animacy hierarchy and topicality.

(465) te'do sihoaw sookno.
[te'] =do [siho]
PATIENT
=aw {so -ok} =no
now =TOP lion =ACC appoint-COS =QUOT
Now [they] appointed the lion.

B) AMBIGUITY ABOUT THE SEMANTIC STATUS OR AFFECTEDNESS OF THE
ARGUMENT. An argument can be accusative marked to disambiguate its semantic
status. Ambiguity is likely to occur if arguments of the same clause are on a par to
each other in the animacy hierarchy, as in example (466). If the name of the person
had not been accusative-marked, there could be confusion about who saw whom.

20 CASE MARKING

340
(466) na' ranustaw nukama nukanca?
[na'] [ranus] =taw {nuk -a} =ma {nuk -an -ca}
2s Name =ACC see -CUST =Q see -REF -NEG
Have you seen Ranus or not?

In the next example the speaker marks the NP wiliamnagal Williamnagar (a village
name) with the accusative, because the verb me- to call a name is used transitively
and its O argument, the named entity, is a semantic Patient. Wiliamnagal
Williamnagar and Symsanggre (the former name of the same village) are both low
on the animcy hierarchy. Therefore case marking is necessary in his case because
there could be confusion about which is the affected participant, the O, if both were
left unmarked since the order of the NPs in Atong is free. Since a place name is
always uniquely referential and definite, affectedness of the participant is the
dominating factor for the case-marking here. In Atong the named entity is seen as the
affected argument and thus wiliamnagal Williamnagar is accusative-marked and the
other argument is unmarked for case. It is not possible for both NPs to be accusative-
marked in this construction.

(467) wiliamnagalaw somsagre noay moa.
[wiliamnagal]
O
=aw [somsagre]
E
{no} =ay {mo -a}
Pname =ACC Pname say =ADV call.a.name -CUST
Williamnagar used to be called Symsanggre. Alternatively: [People]
sayingly used to call Williamnagar Symsanggre.

In the next example we see two animate NPs of the verb mo- to call someone or
something a name. Again the named entity is accusative-marked and the name, the E
argument, is unmarked for case.

(468) amo amaw goje em sama mogwa.
[a =mo ama]
O
=aw [goje em sama]
E
{mo -wa}
1s =GEN mother =ACC Pname Sname1 Sname2 call.a.name -FACT
My mothers name is Goje M. Sangma.

The normal way to make people to know what your name is in Atong is the following
expression (469), where the verb mo- to call a name is an extended intransitive and
20 CASE MARKING

341
neither of the arguments S or E are marked for clausal case. In this case there is no
ambiguity about the syntactic status of the NPs.

(469) ami bimu todan em sama mowa.
[a =mi bimu]
S
[todan em sama]
E
{mo -wa}
1s =GEN name Pname Sname1 Sname2 call.a.name -FACT
My name is called Todan M Sangma.

If there is no ambiguity, i.e. when the situation of who did what to whom is clear from
the context, the affected participant can be left unmarked for case, as in the next
example, where the O argument jok wife is unmarked despite the high position in the
animacy hierarchy. The context is as follows. In a country supposedly lived/lives a
lazy king. He has two wives who have to carry him on their hands when he eats and
sleeps. He is so lazy that he does not work, has no money and cannot sustain his two
wives any more. The people are speaking disdainfully about him and so he feels
ashamed and decides to run away. The ellipsed A in (470) is the lazy king.

(470) te'ewe jok asteay jalawacie soreaokno.
[te'ew] =e [jok]
O
{asset} =ay {jala -wa} =ci =e
now =FC wife throw.away =ADV run.away -FACT =LOC =FC
{sorea -ok} =no
wander.around -COS =QUOT
Now when [he] had divorced [lit. threw away/disposed of] his wives, and
ran away, [he] wandered around, it is said.

In the next example we see a mix of animate and non-animate affected participantss.
When the context is sufficiently clear, the affected participants are unmarked for case.
When the speaker thinks there is reason to doubt about the affected status of a
participant, it is accusatively marked.

20 CASE MARKING

342
(471) taw'reksorup masa ge'themo thup phanan moma phay'ay sa'rowana,
moma mathayaw tapna re'eaydoanoa. te'edo ue momaay rekci thupay
thupay mu'gabaaw phanan phay'ay phay'ay sa'roano. otokoymo hawtoy
re'ewacian beblok masa'aw gorookno [] sakhapci a thupay mu'waci
phanan a thupawba pay'ay pay'ay sa'roa.
[taw'reksorup ma sa] [gethe thup] [phanan] [moma]
banana.bird CLF:ANIMALS one 3s nest always elephant
break =ADV eat -USUALLY -FACT =DAT
[moma mathay] =aw {tap} =na {re'e -aydoa} =no -wa
elephant bachelor.elephant =ACC hit =DAT go.away -PROG =QUOT -FACT
[te'e] =do [ue moma] =ay [rek] =ci {thup} =ay {thup} =ay
now =TOP DST elephant =FC banana.tree =LOC to.nest =ADV to.nest =ADV
{mu} =gaba] =aw {sa' -ro -a} =no
stay =ATTR =ACC eat -USUALLY -CUST =QUOT
otokoymo [hawtoy] {re'e -wa} =ci =an
so.then REM go.away -FACT =LOC =FC/ID
[beblok ma sa] =aw {goro -ok} =no
toad CLF:ANIMALS one =ACC meet -COS =QUOT
[sala] [sakhap] =ci [a] {thup} =ay {mu' -wa} =ci
idiot type.of.tree =LOC 1s to.nest =ADV stay -FACT =LOC
[phanan] [a thup] =aw =ba {pay'} =ay {pay'} =ay
always 1s nest =ACC =EMPH break =ADV break =ADV
{sa' -ro -a}
eat -USUALLY -CUST
An elephant always breaks and eats (lit. breakingly eats) a banana birds
nest (lit. his nest). Now as for the elephant, [he] always breaks the [thing]
that stays nestingly in the banana tree. So when [the banana bird] goes way
over there he meets a toad. [Where are you going, friend?, asks the toad. I
am going to beat-up the elephant today, he says. Come on! I will also go.]
The idiot, when [I] nestingly stay in the sakhap tree [the elephant] always
breaks and eats my nest.

The next example shows an O argument without case marking but with focus
marking. The situation, who does what to whom, is clear from the context.
20 CASE MARKING

343
(472) [] gambirimu gamsilimu ja'nawmarane senthiokno.
[gambiri] =mu [gamsili] =mu [ja'naw] =maran =e
type.of.tree =COM type.of.tree =COM sister =RC =FC
{senthi -ok} =no
lament -COS =QUOT
[When are you going to dress us in clothes?] lamented the gambiri and
gamsili tree to the mutual sisters (those who were sisters to each other).

C) Referentiality. In the next example the biscuits are referential. They are not just
biscuits, they are the biscuits which the fox and the deer are trying to steel from the
Bangladeshi as is already clear from what preceded in the story. The biscuits being
significantly lower in the animacy hierarchy than the Bangladeshi, do not necessarily
need to be accusative-marked to understand who does what to whom in this sentence.

(473) otokoymo bagale biskutaw tanaymo []
otokoymo [bagal] =e [biskut] =aw {tan} =ay =mo
so.then Bangladeshi =FC biscuit =ACC put =ADV =SEQ
So then, the Bangladeshi, having put [down] the biscuits[: Hey, that deer is
lame, he said and chased it].

In an NP consisting of a demonstrative and a noun it may occur that the demonstrative
receives the accusative case marking and not the following noun. The case marker
still has scope over the whole NP. Maybe this preferential marking of the
demonstrative instead of the noun has to do with the fact that demonstratives are
inherently more referential than nouns.

(474) umido uaw kamal sandini. uaw wa'phekgumuk, wa' pangumuk
tharitholoaymusa []
umido [u =aw kamal] {sandi -ni}
[u =aw wa'phek] =gumuk [wa' pan] =gumuk
DST=ACC small.type.of.bamboo =all bamboo firewood =all
{thari -tholo} =ay =mu =sa
prepare-NICELY =ADV =SEQ =DLIM
Then [they] will look for that priest, all that small bamboo [and] bamboo for
the fire (alternatively bamboo and firewood) are nicely prepared
20 CASE MARKING

344
20.8.2 Marking of material of which something is made
The enclitic <=aw ~ =taw> (ACC) is found marking the Material of which some
artefact is made. The NP referring to the artefact itself can be unmarked for case when
it is a Patient, as in (475). In the first clause of example (475) we see that Material
ha'boku sand is marked with the accusative, whereas the Patient NP morot person
is clearly a Patient and unmarked for case. More fieldwork is needed to find out if the
accusative-marked Material NP functions syntactically as a modifier in a larger,
complex NP with the noun denoting the artefact.
The last clause of example (475) contains one of the few nouns that can also fully
function as a verb, i.e. cowgon the festival of the dead and cowgon- to celebrate the
festival of the dead.

(475) ha'bokuaw morot takaymuba kakelekaw so'otaymu cigorosa
domcorasa dakami acu ambitokoy dothoyciay takaymu uan me'ma
sa'wetokno, cowgonokno.
[habku] =aw [morot]
O
{tak} =ay =mu =ba [kakolek] =aw
sand =ACC person make =ADV =SEQ =EMPH lizard =ACC
{so'ot} =ay =mu [cigoro] =sa [domcora] =sa
kill =ADV =SEQ snare.instrument =INSTR snare.instrument =INSTR
[daka =mi acu ambi] =tokoy {dothoy =ci =ay}
before =GEN grandpa grandma =LIKE kill -FIRST =ADV
{tak} =ay =mu [u] =an [me'ma] {saw' -et -ok} =no
do =ADV =SEQ DST=FC/ID ghost burn -CAUS -COS =QUOT
{cowgon -ok} =no
celebrate.the.festival.of.the.dead -COS =QUOT
Having made a person out of sand and having killed a lizard, with [their]
string instruments, like [their] ancestors from long ago [they] killed and
burned the ghost, it is said [and] celebrated the festival of the dead.
40

40
After a person dies the spirit continues to live in the house for a year. Sometimes a small house is
built in front of the deceaseds house for its spirit to live in. The burning of the ghost is a ceremony
which is performed one year after the death of a person as described above or, when appropriate, by
burning the small house of the spirit. The spirit of the dead person then leaves the house and travels to
the land of the dead through the way of the forest fire. The land of the dead is Balpakram, now a
national park on the border with Bangladesh and the Khasi Hills.
20 CASE MARKING

345
In example (476) we see that both the Material NP, pan wood, and the artefact NP,
in this case modified by an Attributive clause (see Chapter 29), are accusative-
marked.

(476) [] bondokaw payay, panaw jap kha'gabaaw kawtawna thomokno.
[bondok] =aw {pay} =ay [pan] =aw [jap]
gun =ACC carry.in.hand =ADV wood =ACC defence.wall
{kha} =gaba =aw {kaw -taw} =na {thom -ok} =no
make =ATTR =ACC shoot -UPWARD =DAT lie.in.ambush -COS =QUOT
carrying guns [using] the defence wall made of wood [they] laid in ambush
to shoot upward, it is said.
20.8.3 Purely referential/individuating/definiteness usage of the morpheme
<=aw ~ =taw>
The morpheme <=aw ~ =taw> can be used purely as a marker of referentiality,
individuation and definiteness on NPs that are neither predicates, as in presentative
clauses (see26.4), nor part of a clause, i.e. fragments or free constituents (see Sadock
and Zwicky 1985: 187). The best label for the morpheme in these cases is (DREF)
definite & referential. Definiteness & referential marking in this case is used for
emphasis. Examples (477) and (478) are illustrative. The context from which example
(478) is taken is as follows. A cunning man called Thengton [t
h
c?ton] has been
trapped in a big basket by the people of his village, who want to drown him in the
river. On the way to the river, Tengtons carriers take a break to eat. While they are
not paying attention to the basket, a Nepali arrives and asks:

20 CASE MARKING

346
(477) ha! na'a ca koksechi dagabae? nowano. ucie: acom, nowano.
atakgaba morot? nochie a, aawdo!
ha [na'a [ca koksep
41
] =ci {da} =gaba =e {no} -wa =no
interj 2s who cage.basket =LOC enter=ATTR =FC say -FACT =QUOT
uchie |{a}|
PRESENTATIVE CLAUSE
=com {no} -wa =no
then 1s =IRR say -FAT =QUOT
[atak =gaba morot] {no} =ci =e
do.what =ATTR person say -FACT =FC
[a]
FRAGMENT
[a]
FRAGMENT
=aw =do
1s 1s =DREF =TOP
Hey! Who are you, [the one who] entered into the cage-basket?, [the
Nepali] said, it is said. Then: It is supposedly me! [Thengton] said, it is
said. When [the Nepali} said: What kind of person? [literally: a person who
does what?], [Thengton said:] Me, me!

In the above example we see that the definite- & referential-marked fragment
(underlined) in the last line is the emphatic answer to the question asked in the line
above. The context illustrates the reason for the emphasis: the Nepali had already
asked who Thengton was before, and when he asked a second time, Thengton
answered emphatically to make sure the Nepali got the message. The definite- &
referential-marked fragment NP has no other reason to be marked as such other than
to emphatically indicate its referentiality or individuation, since it does not comply
with any of the other criteria to be accusative-marked, viz. being a Patient or O
argument, a Material etc. as listed in 20.8. The fragment is not only definite- &
referential-marked but also topic-marked for extra emphasis, although this topic-
marking is not obligatory on accusative-marked fragments, as we can see in the next
example.
Example (478), below, illustrates another instance of a definite-& referential-
marked fragment. The speaker starts telling a story, and breaks off the second
sentence which ends in a rising intonation indicating that the sentence was not

41
A koksep is the term for a variety of big, loosely woven bamboo baskets in which animals, usually
chicken and pigs, are kept while they are being transported to and sold at the market.
20 CASE MARKING

347
finished yet in order to state his name for the recording. He first announces that he
wants to do this by the definite- & referential-marked fragment a=mi bimu=aw
(1s=GEN name=DREF) this name of mine. This fragment has a falling intonation
and is therefore a statement on its own; it does not belong to the next clause. The
fragment is referential, definite and introduces a new topic. Contrary to example (477)
above, the fragment is not topic-marked, despite the fact that a new topic is
introduced. The morpheme <=aw ~ =taw> (ACC/DREF) marks only referential NPs
which can be topical in addition (see 20.8).

(478) aa imi gomonaw baletni. ie ooo dakami acu ambi. ooo ami bimoaw.
ami bimo genda ar marak.
aa i =mi gomon =aw bal -et -ni
1s PRX =GEN ABOUT =ACC tell -CAUS -FUT
ie ooo daka =mi acu ambi
PRX interj:HESITATION in.the.past =GEN grandfather grandmother
ooo a =mi bimo =aw
FRAGMENT
interj:HESITATION 1s =GEN name =ACC
a =mi bimo genda ar marak
1s =GEN name Pname SURNAME
I will tell about this. These, uh, ancestors of the past Uh, this name of mine.
My name is Genda R Marak.

The marking of fragments with the morpheme <=aw ~ =taw> (ACC/DREF) is very
common in Atong and speakers of all ages do it.
20.8.4 The morpheme <=aw ~ =taw> on clause initial topical S arguments
Two instances have been discovered in the recorded corpus of language material
where a clause initial NP that cannot possibly be interpreted as Patient or O argument
is marked with the accusative enclitic <=aw> (ACC). These instances, represented
below as examples (479) and (480), occur in texts of two different older speakers. The
speaker who produced (480) was very old. Example (479) comes from TEXT 2 (line
24) and the wider context of (480) can be found in (350).

20 CASE MARKING

348
(479) khurutna sapgaba morotawsa [pause] sogumukciba soci pa'ramaria.
[khurut =na sap =gaba morot]
TOPIC/S
=aw =sa

perform.an.incantation =DAT know.a.skill =ATTR person =ACC =DLIM
[so] =gumuk =ci =ba [so] =ci
village =whole =LOC =emph village =LOC
pa' -ram -ari -a
be.many -FORTUITOUSLY -SIMP -CUST
As for precisely those people [who] know how to perform incantations, in all
villages [and] in this village, [there are] many [of them] for no good reason.

(480) man'gabaaw sa'phet rophet.
[man' =gaba]
TOPIC/S
=aw {sa' -phet} {po -phet}
have =ATTR =ACC eat -TO.ONES.DETRIMENT drink -TO.ONES.DETRIMENT
The ones who are rich (lit. who have) [are] eaters to their [own] detriment
[and] are drunkards to their [own] detriment.

Although not marked by the topic enclitic <=do> (TOP) (see 19.13) or the focus
enclitic <=e> (FC) (see 19.14), the accusative-marked NPs in both examples above
introduce new topics. These topics are referential and definite and therefore marked
by the accusative enclitic <=aw> (ACC); in (479) the accusative-marked NP is also
delimitative-marked for a more precise reference. In (479) the new topic is separated
from the rest of the clause by a pause, but there is no such pause in (480).
Both accusative marked NPs function as S argument in their respective sentences.
In (479) the NP khurutna sapgaba morotawsa precisely those people [who] know
how to perform incantations is the S of the predicate pa'ramaria are many for no
good reason. In (480) the NP man'gabaaw the ones who have is a headless
attributive clause (see Chapter 29) functioning as the S argument of the predicates
sa'phet eat to ones detriment and rophet drink to ones detriment. In 18.9 is
argued that these are actually nominal predicates (see also 22.5).
In the Linguistic Survey of India (Grierson, 1902: 86-88), reverend E.G. Philips
uses the gloss the every time the accusative/definite morpheme occurs. He correctly
recognised that this morpheme only occurs on referential and definite NPs. Maybe a
hundred years ago the sole purpose of the morpheme <=aw ~ =taw> was one like the
definite article in English, marking NPs as referential and definite. In the language of
today this morpheme has developed a second function, i.e. that of accusative marker
20 CASE MARKING

349
that helps to distinguish O arguments from other NPs, but only when the O is definite
and referential.
Since S and A are always unmarked for case and O can only be marked with the
morpheme <=aw ~ =taw> (ACC) when it is referential and definite and since the
morpheme <=aw ~ =taw> (ACC/DREF) does not exclusively mark O arguments, the
syntactic and semantic function of unmarked NPs has to be assessed pragmatically.
20.8.5 Repeated accusative case marking
Accusative marking is found to be repeated for emphasis as in (481) and (482). The
accusative-marked NPs in (482) are in apposition to each other and have the same
semantic and syntactic roles. That the accusative-marked NPs are in apposition and
not part of one complex NP is apperent by the pauses in between the accusative
marked clausal constituents. In example (482) the pauses are marked in seconds.

(481) udo jow'gabado uaw, di'awba, asetca.
[u] =do [jow'] =gaba =do [u] =aw [di'] =aw =ba {asset -ca}
DST=TOP mother =DREL =TOP DST=ACC shit =ACC =EMPH throw.away -NEG
That mother does not throw that [stuff], the shit, away. [She simply collected
the vegetables and cooked them]

The next example illustrates the co-occurrence of accusative case marking and
focus/identification-marking on the noun gumuk all, everything. The reason for the
accusative marking here is ambiguity about the semantic status of gumuk all,
everything. If unmarked it could be interpreted as referring to the women who
collected their husbands and not to the husbands.

20 CASE MARKING

350
(482) khasin khasin (2.8) gumukawan (0.4) paloci jalgabadaraaw (0.7)
jokthathaaw (0.4) jumuphonnaakno.
{khasin khasin}
slow RED
[gumuk] =aw =an [palo] =ci {jal} =gaba] =dara =aw
all =ACC =FC/ID jungle =LOC run.away =ATTR =p =ACC
[jok] =tha =tha] =aw {jumu -phon -a -ak} =no
spouse -OWN -RED =ACC collect -AGAIN -TOWARDS -COS =QUOT
[the women] slowly collected everybody again, their own husbands, [the
ones who] had run away to the jungle, it is said.

When a noun is modified by a demonstrative, the demonstrative and the noun often
both take the case-marker, e.g. (481) and (496).
20.8.6 More than one accusative marked NP in a clause
In (492) we can see accusative marking on two different NPs with different syntactic
and semantic functions. In that example sam grass is the actual O argument, whereas
ca' foot is an adjunct in the semantic function of instrument. The noun ca' foot is
always accusative-marked when used as an instrument while other body parts and
other instruments are marked with the instrumental case <=sa> (INSTR), e.g.
cak=sa kha'gaba (hand=INSTR make=ATTR) made by hand.
Another clause type where we can find two accusative-marked NPs is the
causative clause. Causative clauses with two accusative-marked NPs have been
obtained by elicitation as they did not occur in any of the recorded materials (483).
Causativisation is treated in the chapter on valency changing derivation. In the next
example the animacy hierarchy makes it clear that the third person singular (ge'the)
is the Causee argument acting in turn on the girl (gawi) which is lower on the animacy
hierarchy. The real O argument (483) is gawi girl because, as we can see in (484),
the Causee can be instrumental-marked if lower on the animacy hierarchy than the O.
Common nouns are lower in the animacy hierarchy than pronouns.

(483) a ge'theaw gawiaw so'otetwa.
[a]
A
[ge'the] =aw [gawi] =aw {so'ot -et -wa}
1s 3s =ACC girl =ACC kill -CAUS -FACT
I made him kill the girl.
20 CASE MARKING

351
When you want to express the reverse, i.e. I made the girl kill him, the girl has to be
instrumental-marked because it is lower in the animacy hierarchy than the third
person ge'the (3s) (484).

(484) a ge'theaw gawisa so'otetwa.
[a] [ge'the] =aw [gawi] =sa {so'ot -et -wa}
1s 3s ACC girl =INSTR kill -CAUS -FACT
I made the girl kill him
20.9 The homophonous markers <=tky> (VIA) for the perlative and
<=tky> (LIKE) for the similative
Although the markers are homophonous, they are presumably used in two historically
related but synchronically different grammatical processes. We can hypothesise that
similative is a derivational morpheme, viz. an adverbialiser. A similative-marked NP
indicates how the action denoted by the verb takes place. The perlative is not
adverbialising because semantically the NP it occurs on refers to a pathway, and so it
does not indicate how the action denoted by the verb is done, but via which place or
path an action takes place. The reading of the specific function of the morpheme
depends on the word class it enclitisises to, the semantics of the verb and the context.
The analysis seems problematic since we cannot know if the Atong speakers
conceptualise the two possible functions of the morpheme <=tokoy> (VIA/LIKE) as
one cognitive operation or not. If for the Atong speakers marking an NP or adverbial
clause with the marker under discussion is one cognitive operation, then I would
assume it is an adverbialising one. This would mean that Atong speakers
conceptualise doing something in the way their forefathers did (475) in the same
way as going in a Baghmara-like fashion (485).

(485) na'na bagmaratokoy re'eni.
[na'na] [bagmara]
PATHWAY
=tky {re'e -ni}
1pi Pname =VIA go.away -FUT
We will go through/via Baghmara.

A purely adverbialising function is also what one would expect considering the origin
and grammaticalisation path of the morpheme <=tokoy> (VIA/LIKE) as explained
20 CASE MARKING

352
below, i.e. coming from a grammaticalised adverbial clause with the predicate otok-
do like this/that.
Since it is impossible to know the conceptualising processes that go on in the
mind of an Atong speaker, one could argue that splitting the morpheme into two
homophonous enclitics with different grammatical functions is arbitrary or based on
the grammar of the language in which this grammar is written, e.g. English.
There is one argument in favour of a cognitive split. There are separate
interrogatives in Atong that ask for the pathway, viz. bi=tokoy (QF=VIA) via which
way? and for the way some action came about or is done, viz. ato=tokoy
(what?=LIKE) doing what?/why?/how? (486).
An argument against the split is the fact that the marker <=tokoy> (VIA/LIKE)
appears on the predicate heads of non-verbal clauses like (494). Adverbs cannot be
predicate heads of non-verbal clauses.
Thus I describe the perlative/similative enclitic as a case enclitic that marks
constituents as Pathway or Facsimile adjuncts (peripheral arguments), depending on
contextual interpretation.

(486) atotokoy tay'nido thawoksoyi ja'bek?
ato =tokoy [tay'ni] =do {thaw -ok} =soy [ja'bek]
what =LIKE today =TOP tasty -COS =MIR curry
Why is [it] so tasty today, the curry?

The following examples illustrate the use of the similative enclitic on a prototypical
noun (487), and on a classifier-numeral phrase (488).

(487) phulistokoy nukramphinokno bunduk paygana.
[phulis] =tokoy {nuk -ram -phin -ok} =no
police =LIKE see -INADVERTENTLY -OBVIOUSLY -COS =QUOT
[[bunduk] {pay} =ga] =na
gun carry.in.hands =ATTR =DAT
[They] inadvertently obviously looked like the police, it is said, because of
the carrying of guns.
20 CASE MARKING

353
(488) udo pa'aydo sapdamca. mo'sa mo'nitokoy sapa.
[u] =do {pa'} =ay =do {sap -dam -ca}
DST =TOP many =ADV =TOP know.a.skill -TRUELY -NEG
[mo' sa mo' ni] =tokoy {sap -a}
CLF:HUMANS one CLF.HUMANS two =LIKE know.a.skill -CUST
Not many [people] truly know that skill. [They only] one-or-two-ingly know
the skill.(i.e. only one or two of them know the skill).

It seems that morpheme <=tokoy> (VIA/LIKE) is a grammaticalised form of the
non-finite form of verb otok=ay (do.like this/that=ADV) with reduced vowel quality of
the non-finite enclitic <=ay> (ADV) and loss of the non-productive prefix <o-> (?)
which makes it phonologically less heavy and more apt to be an enclitic.
Grammaticalised differently, but from the same source is the adverb otokoy like that,
e.g. (489), which in turn is more grammaticalised than the homophonous discourse
connective otokoy so then which still has a verbal property in that it can take the
sequential marker <=mo ~ =mu ~ =muna> (SEQ).

(489) otokoy taw'phinbo.
otokoy {taw' -phin} =bo
like.that ascend -BACK =IMP
Go back up like that/through that way.

The perlative is the most grammaticalised form since it is no longer adverbialiser but
marks adjuncts indicating a pathway. Tracing this path of grammaticalisation helps us
to explain the adverbialising character of the similative function of the case marker
<=tokoy> (LIKE). The whole grammaticalisation path of this case marker is
summarised in Table 60.

Table 60 The grammaticalisation path of the case marker
<=tokoy> (VIA/LIKE)
verb non-finite form

discourse
conjunctive
adverb
otok- to do
like
this/that
otok=ay
(do.like.this/that=ADV)
doing thusly
otokoy
so then
otokoy
like
this/that
similative perlative
<=tokoy>
(LIKE)
<=tokoy>
(VIA)
20 CASE MARKING

354
Here below is an example of the perlative on demonstratives.

(490) (Speaker A) itokoy re'enima itokoy? (Speaker B) utokoy.
[i =tokoy] {re'e -ni} =ma [i] =tokoy [u] =tokoy]
PRX =VIA go.away -FUT =Q PRX =VIA DST=VIA
(Speaker A) Will we be going through here or through here? (Speaker B)
Through there.

Relativised clauses can also take the perlative enclitic as is illustrated in the following
example.

(491) samna jekay rokgabatokoy roka, kamalnado.
[sam] =na [jekay {rok} =gaba] =tokoy {roka} [kamal] =na =do
medicine =DAT some search =ATTR =VIA search priest =DAT =TOP
Medicine is sought through some searcher, for the priest.

The following example shows the similative on a demonstrative. The utterance was
accompanied by gestures of the speaker.

(492) samaw ca'aw itokoy tokano.
[sam] =aw [ca'] =aw [i] =tokoy {tok -a} =no
grass =ACC foot =ACC PRX =LIKE hit -CUST =QUOT
[They] trample the grass like this with [their] feet. Alternatively: [They] hit
the grass like this on foot.

In the example here below the similative appears on a time word, i.e. tay'sa just a
little while ago. Time words are noun-like but do not have all the properties of a
prototypical noun.

(493) uan tay'satokoy kantaraaw korokorok re'eaymo so'etthiriokno.
[u] =an [tay'sa] =tokoy [kantara] =aw [korokorok]
DST=FC/ID a.little.wile.ago =LIKE emptiness =ACC quickly
{re'e} =ay =mo {so' -et -thiri -ok} =no
go.away =ADV =SEQ ask -CAUS -AGAIN -COS =QUOT
Like just a little while earlier he quickly went to the emptiness and asked
again.
20 CASE MARKING

355
Example (494) illustrates the similative on a demonstrative which functions as the
predicate head of a verbless clause.

(494) phalthaci wa'curek sa'na man'arino. na'naba utokoy.
[[phaltha =ci wa'curek] {sa'} =na {man' -ari} =no
self =LOC capacity eat =DAT be.able -SIMP =QUOT
[na'na] =ba {[u] =tky}
1pi =ADD DST =LIKE
[Earthworms] can only eat their own capacity. We are also like that.
20.10 Multiple case marking
This section sums up all the examples with multiple case marking from this chapter
and presents some new examples as well. Dixon (2002: 148-9), in his summary of
double case in Australian languages, presents three types of double occurrence of case
morphemes on an NP:

(a) marking of phrasal function (genitive, comitative, privative) plus marking
of clausal function (ergative, accusative, dative, instrumental, etc.);
(b) local marking plus marking of clausal function;
(c) marking of clausal function plus marking of clausal function.

Atong has interesting variations on (a) and (b); there is no construction akin to
Dixons type (c). We can clearly see that Atong uses case marking to indicate
semantic roles which I will indicate where appropriate.
20.10.1 Local/Direction marking + marking of clausal function
This type of double case-marking is akin to Dixons type be (b) mentioned above.
i Location and O
In (495) and (496) we see a use of the marker <=sa> as a locative.
42
Baljo=sa
(Pname=MOB) at Baljong is a local modifier and functions as afterthought in post-

42
This marker grammaticalised from the noun *sa side, place now only found as a bound morpheme
in some nouns. Compounding with this bound morpheme is not productive.
20 CASE MARKING

356
verbal position. It is marked with accusative marker <=aw> (ACC) because the NP is
referential and definite.

(495) daka ha' saw' wacom, baljosaaw.
[daka] [ha'] {saw' -wa} =com [baljo] =sa =aw
before land burn -FACT =IRR Pname =MOB =ACC
In the past [people] supposedly burned the land, [the land] at Baljong. (i.e.
they practised slash-and-burn agriculture).

(496) otokoymuna ha' haw'ay sa'ayronotoy, usaaw raponramsaaw ha'
haw'ay sa'ayrono.
otokoymuna {ha' haw'} =ay {sa' ayro} =no =oy
so.then soil clear.the.jungle =ADV eat -PROG =QUOT =MIR
[u] =sa =aw [aponram] =aw {ha' haw'} =ay
-DST =MOB =ACC Pname =ACC soil clear.the.jungle =ADV
{sa' -ayro} =no
eat -PROG =QUOT
So then, [she] is living off a rice field, it is said, [she] is living off that rice
field at Rangpynram, it is said.
ii Marking a Location as a Goal
Besides marking Beneficiary, Recipient and Experiencer (see section 20.6, this
chapter) the morpheme <=na> (DAT/ALL) can also mark a Goal, in which case the
morpheme is labelled as the allative case marker. NPs referring to Goals have to be
obligatorily allative-marked. Only locations can be Goals and thus a Goal also needs
to be obligatorily locative-marked. The locative <=ci> (LOC) marks the NP as a
location. The allative marker has a clausal function. Thus we find double marking for
semantic role on NPs functioning as Goals. The locative case is marked closest to the
root or stem, followed by the allative marker, i.e. <=ci=na> (LOC=ALL) as is
illustrated here below in (497) and above in (423) and (441).

20 CASE MARKING

357
(497) otokoymo ketketa bura re'eokno re'eokno, jow'paracina phetaokno.
otokoymo [ketketa bura] {re'e -ok} =no {re'e -ok} =no
so.then Name go.away -COS =QUOT go.away -COS =QUOT
[jow'] =para =ci =na {phet -a -ok} =no
mother =&co LOC =ALL reach -AWAY -COS =QUOT
So then Ketketa Bura went and went and reached the house of his mothers
family.

Locative-allative marking is found repeated on two nouns within the same NP which
are in an additive relation to each other in (498). Both nouns are modified by the
reduplicated type 2 adjective dotha different. The reduplication of the adjective
indicates plural in this case. In (499) the locative-allative marking is found repeated
on two nouns belonging to different NPs in coordination.

(498) otokoysa dothadotha socina bihapcina jalthokna ga'akok.
otokoymo [dotha dotha so] =ci =na [bihap] =ci =na
so.then different RED village =LOC =ALL place =LOC =ALL
{jal -thok} =na {ga'ak -ok}
run.away -TOGETHER =DAT be.compelled -COS
So then [they] were compelled to all run away to different villages [and] to
[different] places.

(499) may botwamodo pucina soncina khayrata.
[may bot] =wamo =do [pu] =ci =na
rice pull =NR =TOP rice.stock.house =LOC =ALL
[so] =ci =na {khay -rat -a}
village =LOC =DAT carry.on.body -DOWNWARD -CUST
The rice harvest is carried down on the body to the rice stock house, to the
village.
20.10.2 Local marking + local marking: Direction and Source
In (500) the implication of the mobilitative plus ablative marking is movement away
from something. Therefore, semantically, the combination of mobilitative + ablative
marks the NP for movement and source.

20 CASE MARKING

358
(500) phorenmi morot ray''adoa, phorensami ray'aydoa.
[phoren =mi morot] {ray' -aydoa} [phoren] =sa -mi
foreign.country =GEN person come -PROG foreign.country =MOB =ABL
{ray' -aydoa}
come -PROG
Persons belonging to foreign countries are coming, [they] are coming from
foreign countries.
20.10.3 Local marking + local marking + clausal function: Direction, Source
and O
The following example illustrates how an NP, in this case a demonstrative, is marked
by three cases. It is marked as Direction by the mobilitative, as a Source by the
ablative. The motivation for the accusative marking is referentiality and definiteness,
and to mark the constituent as a Patient. The NP under discussion is underlined.

(501) ie ca'masagmi way khurutcido ue hoysamiaw bagoladesmi thol'koosmi
jaria ha'gorsakgumukawan moani.
[ie ca'ma sa =mi way] {khurut} =ci =do [ue
PRX downstream =MOB =GEN spirit summon =LOC =TOP DST
hoy] =sa =mi =aw [bagolades =mi thol'] [koos =mi
far.away =MOB =ABL =ACC Pname =GEN up.to Pname =GEN
jaria] [ha'gorsak] =gumuk =aw =an {moa -ni}
influence the.world/everything =everything =ACC =FC/ID call.upon -FUT
When he summons the downstream spirit, that [priest] will call upon the
influence of all those far away [places] up till Bangladesh [and] the area of
Kongos, all of them.
20.10.4 Marking of clausal function first and then of phrasal function
The following example illustrates double case marking with the dative and the
comitative. The example is the same as (454) above and also illustrates how the
comitative case marks nouns belonging to different peripheral NPs in a clause. The
dative marker <=na> (DAT) has a clausal function, marking the NPs as Recipients
and the nominalised clause ca ro (tea drink) as a Purpose adjunct, and the comitative
indicates the relation between the NPs, which is a phrasal function. The dative
marking is repeated because the NPs are enumerated and therefore coordinated.
20 CASE MARKING

359
(502) jahasnamu, ca ronamu, bagajinamu rajani taka jamok.
[jahas]
RECIPIENT
=na =mu |[ca] {ro}|
PURPOSE
=na =mu
ship =DAT =COM tea drink =DAT =COM
[bagaji]
RECIPIENT
=na =mu [raja ni taka] {jam -ok}
fortune-teller =DAT =COM hundred two money/rupee finish -COS
On the ship (for the benefit of the ship), on drinking tea, on the fortune-teller
the two hundred rupees were all spent [lit. were finished].
20.10.5 Stem-forming genitive governed by <gmn> reason, about
The postposition <gomon> reason, about occurs with the genitive case. Thus the
genitive functions as a stem-forming suffix as treated in section 20.4 B. Examples can
be found in 13.3. This postposition only marks adjuncts.
In the following example the proximal demonstrative carries the genitive
governed by the postposition <gomon> reason, about. The whole NP receives
accusative marking because it is referential.

(503) aa imigomonaw baletni.
[aa] [i =mi gomon] =aw {bal -et -ni}
1s PRX =GEN about =ACC tell -CAUS -FUT
I will tell about this.
20.11 Repeated case marking summary
Repeated case marking means that more than one noun in a sequence of nouns is
marked with the same case. Double case is always repeated as a whole. The
motivations for repeated case marking differ from case to case. The motivations and
the cases for which the motivation holds are listed in Table 61.

Table 61 What cases are found repeated and why.
MOTIVATION CASE EXAMPLE
Nouns in additive relationship within an NP
GENITIVE (452)
LOCATIVE-dative (498)
Enumeration/ Coordination
DATIVE (463)
DATIVE=COMITATIVE (454)
LOCATIVE-dative (499)
Emphasis / Referentiality ACCUSATIVE (481), (482)

20 CASE MARKING

360
Cases for which repetition is not attested: mobilitative/instrumental/locative <=sa>
(MOB/INSTR/LOC), locative <=ci> (LOC), perlative/similative <=tokoy>
(VIA/LIKE). All NPs in a comitative relationship are marked with the comitative
case. A single comitative-marked NP in a clause can occur when the other comitative
argument is ellipsed, e.g. (455), when the action is reciprocal e.g. (456), or when the
comitative is used to express a simple comitative relation on an NP rather than
comitative coordination., e.g. (435), (443) and (748).

361

362
Chapter 21 Transitivity
_____________________________________________________________________

Transitivity in Atong is a property of a construction that corresponds to a clause and
therefore a discourse phenomenon that involves more than the argument frame of a
verb, as has been described for Iatmul (Ndu family, East Sepik, Papua New Guinea,
see Jendraschek, 2008). A clause is transitive when it either contains an NP that
functions as O argument, or when an O argument must be inferred from the context. A
clause is intransitive when there is neither an overt NP in the clause that can function
as O argument, nor an implicit O argument that is recoverable from the context. As
we will see below, it is not always possible to know whether the speaker
conceptualises an O when it is not explicitly stated in the clause.
Transitivity is also related to valency, i.e. the property of the verb to take
complements (core arguments). I adopt Jendrascheks definition of transitive verb: a
verb that can be used in transitive constructions. We will see that transitive verbs in
Atong can also appear in intransitive constructions, whereas an intransitive verb
cannot appear in a transitive construction. Thus we can define intransitive verbs as
verbs that cannot be used in transitive constructions.
Transitivity plays a role in Atong, given the existence of a small number of
transitive and intransitive verbal lexical pairs (see Table 25 in 4.6) and the existence
of the causative predicate head suffix <-et> (CAUS). Other than the occurrence in
transitivity pairs of some verbs, there are no morphological criteria to distinguish
transitivity classes. Since, as was mentioned above, no NP needs to be expressed
obligatorily in any clause, we can say that there is no obvious relationship between the
transitivity of the clause and the valency of the verb when we look at the
morphosyntactic structure of the language. The number of overt dependents that are
expressed in a clause and their case marking is a matter of semantics and pragmatics
rather than syntax.
The following observations can be made for Atong syntax:
1. There is no formal distinction between core syntactic roles such as case marking
or word order.
2. Complements need not be overtly stated when retrievable from the context.
21 TRANSITIVITY

363
3. S=A ambitransitivity is impossible to diagnose.
4. S=O ambitransitivity is possible when no A can possibly be retrieved from the
context.
5. NPs can coalesce with the verb, i.e. be incorporated into the predicate, and thus
lose their argument status. As a result the predicate becomes intransitive (see
22.7.1).
6. There are pivots in coreference for a small number of clause combinations.

Point 1 will be discussed in section 21.1. Points 2 and 3 are treated in section 21.2,
followed by a discussion on point 4, S=O ambitransitivity, in section 21.3 which also
involves observations concerning point 5. Finally, point 6 will be treated in section
21.4.
21.1 No formal distinction between core syntactic roles
The core arguments intransitive subject (S), transitive subject (A), (CS) and copula
complement (CC) cannot be distinguished on the basis of case marking. S and A, CS
and CC are always unmarked, and transitive object (O) can optionally be accusative-
marked, but only when the NP is referential and definite. However, the
accusative/definite-&-referential morpheme <=aw ~ =taw> (ACC/DREF) does not
only mark O arguments, as is discussed in 20.8, but also Materials of which artefacts
are made, e.g. the word ca' foot/leg when used as an instrument. A clause with more
than one accusative marked NP is not exceptional, e.g. (504). In this example we see
that both the Patient, sam medicinal plant, and the Instrument, ca' leg/foot, are
marked with the morpheme <=aw> (ACC). Neither the syntactic nor the semantic
role of these NPs can be assessed on the basis of the case-marking, although the case
marking does narrow down the number of possible interpretations.

(504) [] samaw ca'aw itokoy [gestures] tokano.
[sam]
PATIENT
=aw [ca']
INSTRUMENT
=aw [i] =tokoy {tok -a} =no
medicinal.plant =ACC foot/leg =ACC PRX =LIKE beat -CUST =QUOT
[ our ancestors] beat the medicinal plants with [their] feet like this
(gestures), it is said.

21 TRANSITIVITY

364
Moreover, NPs with other semantic roles can be unmarked for case as well. Semantic
roles that can optionally be left unmarked for case are given in Table 58. Goals, for
example, can be left unmarked when the noun is inherently locational. The Name NP
in a clause with the verb mo- to call someone/something a name is always
unmarked for case, while the named entity can optionally be accusative-marked. In
the next example, we see a clause where both the named entity and the Name are
unmarked for case. Thus, context has to make clear what is named what. Examples
(467) and (468) in 20.8.1 shows a case where the named entity is accusative-marked.

(505) ami bimu Samrat mowa.
[a =mi bimu]
NAMED ENTITY
[Bairik]
NAME
{mo -wa}
1s =GEN name Bairik call.a.name -FACT
[One] calls my name Bairik. Alternatively: My name is called Bairik.

As for word order, most often the Topic is the first NP of the clause. The topic most
often corresponds to the Actor or Agent of the clause, or, when seen from a syntactic
perspective, with the A or S argument. When not the A argument but the O is the
topic, the O can be accusative-marked when referential, e.g. (506). Non-referential O
arguments cannot be accusative-marked (see 20.8).
The referents of the two core argument NPs expressed in the next example,
Dilbangkhongdang, the name of a man, and matsa tiger, are both high on the
animacy hierarchy and are both referential. The O argument, Dilbangkhongdang, is
the Topic and therefore preposed to the A argument, matsa tiger. To indicate its O
status, the inherently referential NP Dilbangkhongdang is accusative-marked.

(506) jemi sanci dibakhongdaaw matsa kakok.
[je =mi san] =ci [dibakhoda]
TOPIC/O
=aw [matsa] {kak -ok}
any =GEN day =LOC Pname =ACC tiger bite -COS
On a certain day a tiger bit Dibangkongdang .

In the context of example (507), all the animals have gathered to elect a king. They
elect several animals, who all decline. Finally they elect the lion, who accepts. We can
see that the accusative-marked NP siho lion is the only NP in the clause. The
referent is high in the animacy hierarchy and could therefore easily be interpreted as
21 TRANSITIVITY

365
the Agent. However, the fact that siho lion is accusative-marked allows the hearer
to infer that an A argument is implied, i.e. the animals, retrievable from the context,
are the ones who elect.

(507) te'do songhoaw sookno.
te' =do [siho]
TOPIC/O
=aw {so -ok} =no
now =TOP lion =ACC elect -COS =QUOT
Now, [the animals] elected the lion, it is said.

In addition to this, Atong lacks cross reference of arguments on the predicate and
the predicate does not agree in any way in person or number with any NP in a clause.
The syntactic function of every unmarked NP and every NP marked by the morpheme
<=aw ~ =taw> (ACC/DREF) must therefore be assessed pragmatically. All NPs that
are marked as being S, A or O in this grammar are marked based on the inference of
the syntactico-semantic function of the NPs in the context of the example.
21.2 Optionality of complements and S=A ambitransitivity
Adverbials are optional in Atong, i.e. they need not be overtly stated when they are
clear from the context. It is therefore not possible to diagnose complements with a
deletion test. English differs from Atong in this respect, since omission of NPs in
English can lead to ungrammatical clauses. In Atong omission of a Location NP with
the verb tan- to put, for instance, results in a perfectly grammatical clause, as we can
see in (508).

(508) otokoymo bagale biskutaw tanaymo []
otokoymo bagal =e biskut =aw tan =ay =mo
so.then Bengali =FC biscuits =ACC put =ADV =SEQ
So then, the Bengali put the biscuits down [and said ]

In the same way, all NPs in Atong can be omitted when retrievable from the context.
In line 1 of TEXT 1 the subjects of the two transitive clauses are omitted since it is
clear that the speaker is speaking about himself. This example is represented below as
(509).

21 TRANSITIVITY

366
(509) aaah jowna sokarokte. coy'a.
aaah jow =na sok -arok =te. coi' -a.
interj sleep =DAT want -PROG =DCL tired -IMPF
Oh! [I] want to sleep. [I]m tired.

Since all verbal complements are optional, it is impossible to distinguish between the
S=A ambitransitive use of a verb (i.e. the intransitive use of a transitive verb) and the
transitive use with omitted O. So (510) could well be an ellipsed version of (511).

(510) rasan khama.
rasan kham -a
sun burn -CUST
The sun burns (Possible interpretations: Its very hot or the sun burns X.)

(511) rasan aaw khama
rasan a =aw kham -a
sun 1s =ACC burn -CUST
The sun burns me.

Jendraschek (2008: 2) states about Iatmul, that [i]n intransitive use, any inferred
patient [] would be non-referential. An omitted O in contrast would have been
introduced in the context and therefore correspond to a definite NP []. In Atong it is
possible to walk into a kitchen, see someone eating and say a=ba sa'-ni (1s
=ADD eat-FUT) I will also eat, but we cannot be sure that the speaker is not
implying an ellipsed O, namely the same food as that of the person who is already
eating. Therefore, we cannot be sure if the construction is transitive or not. The
uncertainty holds true for the clause with the transitive verb in the next example,
where, although it has not been stated anywhere during the conversation, we can
assume that the one that is being hit is the student. Since this referent can be retrieved
from the extra-linguistic context, the clause can be interpreted as transitive with an
omitted O argument.

21 TRANSITIVITY

367
(512) teacher: bi =sa re'e -wa na'a?
QF =MOB go.away -FACT 2s
Where did you go to?
student: u =sa nalsasa re'e -wa.
DST =MOB the.other.side.of.the.water go.away -FACT
[I] went there, to the other side of the [sea].
teacher: hm'm, bal =bo atak =na re'e -wa.
no tell =IMP do.what =DAT go.away -FACT
bal -ca =ci =do tok -ni.
tell -NEG =LOC =TOP beat -FUT
No! Tell [me], why did you go? If you dont tell, Ill beat [you].

The point is that there is usually something in the real world context that is understood
by the speaker and the hearer as an implied O in those cases where a transitive verb is
used without overt O. We cannot look into the speakers mind to see whether this
possible O is implied or not, i.e. whether we have a case of S=A ambitransitivity or
not.
Atong often inserts a prototypical noun in cases where English uses an
ambitransitive verb without O. This prototypical noun coalesces with the verb, i.e. it
gets incorporated into the predicate and loses its argument status altogether. As a
result of the incorporation, the predicate becomes intransitive (see 22.7.1). In Text 2
line 65, represented here as (513), Atong incorporates the prototypical noun may
rice into the predicate, whereas the non-literal English translation does not have an
O. We can see that the NP may rice is not an O argument, because the quantifying
event specifier thok ALL functions on an S/O basis (see 25.5), and in this case it
quantifies the S of this intransitive clause.

(513) may sa'hokokma na'tome?
{may sa' -thok -ok} =ma [na -tom] =e
rice eat -ALL -CoS =Q 2s -ppp =FC
Have you all eaten? Literally: Have you all rice-eaten?
21.3 S=O ambitransitivity
In the same way that it is almost impossible to know whether a speaker implies an O
in a seemingly S=A ambitransitive clause, it is in some cases equally impossible to
diagnose whether a speaker implies an A or not when no A NP is overtly stated, i.e. in
21 TRANSITIVITY

368
what could be S=O ambitransitive clauses. When there is no A in the linguistic
context, we can only assume none is implied, but our assumption might be wrong, as
when, for example, someone asks if a certain fruit is eaten or not, and when the
interlocutor answers that it is not, and there is no particular A mentioned in the whole
conversation, e.g. (514). The interpretation could be one of S=O ambitransitivity or of
an implied, non referential A.

(514) Speaker 1: [cicot]
S=O?
{sa' -a} =ma?
dud.jackfruit eat -CUST =Q
Do you/does one/do people eat dud jackfruit? Alternatively: Can
you/one/people eat dud jackfruit? Alternatively: Is dud jackfruit eaten?
Speaker 2: [cicot]
S=O?
{sa' -ca}
dud.jackfruit eat -NEG
You/one/people dont eat dud jackfruit. Alternatively: You/one/people cant
eat dud jackfruit. Alternatively: Dud jackfruit is not eaten. Alternatively:
Dud jackfruit is not edible.

The NP cicot in (514) is non-referential. In example (516) we see a referential NP that
appears in a clause that may be interpreted as being S=O ambitransitive. However, the
potential A, i.e. the monkey children, has already been introduced in the preceding
part of the clause. The verb nuk- to see is transitive, as we can see in (515).

(515) aa na'aw nukjorooaria.
[aa]
A
[na']
O
=aw {nuk -joro -ari -a}
1s 2s =ACC see -DAILY -SIMP -CUST
I just see you every day.

21 TRANSITIVITY

369
(516) walmokrakwaci te'edo sa'dore: awamo nakhaldo thawsu thawsu nukwa
noaydoanoa amak sa'ay.
walmokrak -wa =ci =e te'e =do sa' =dora =e
saty.awake.all.night -FACT =LOC =FC now =TOP child =p =FC
[awa =m nakhal]
S=O?
=do [thaw -su thaw -su] {nuk -wa}
father =GEN ear =TOP tasty -REALLY tasty -REALLY look.like -FACT
no -aydoa =no amak sa' =ay
say -PROG =QUOT monkey child =FC
When they stayed up all night, now, the children: Fathers ears sure look
tasty (Alternative, more literal translation with implied A argument:) [We]
see fathers ears tastily, said the monkey children, it is said. [after the mother
had killed the father monkey accidentally, and then the mother and the
children ate the father monkey.]

In the context of example (517), a mother has just told her newly born prodigious
baby, who can already talk, that his brothers are away to hunt a giant eagle. Just after
the example, the baby tells his mother that he will go in search of his brothers. The NP
gandurian umbilical cord, in this example, is not very likely to be interpreted as an
A argument, since it is very low on the animacy hierarchy. A possible interpretation
of this clause is the S=O ambitransitive one, since the mother, although mentioned not
long before in the context, does not have to be conceptualised as the implicit A of the
clause. But we have no proof that the speaker does conceptualise an A argument, and
there we have two choices, the mother, who was said to be all alone in her house at
the time of the birth, or the child, since it is expressed as an antitopic and because it is
prodigious and could therefore cut its own umbilical cord.

(517) gandurian cotkhucano ue sa'goraye.
[gandurian]
S=O?
{cot -khu -ca} [ue sa'goray]
ANTITOPIC
=e
umbilical.cord tear -INCOM -NEG DST child =FC
The umbilical cord was not yet torn, as far as that child is concerned.
Alternatively: [The mother] had not yet torn the umbilical cord
Alternatively: [The child] had not yet torn the umbilical cord

The first clause in the next example has the best chance of being intended as a real
S=O ambitransitive, albeit with an omitted S, since there is no particular A available
anywhere in the context. A group of brothers are walking through the jungle carrying
guns. They accidentally meet an old woman who thinks that they are the police.
21 TRANSITIVITY

370
(518) phulistokoy nukramphinokno bunduk paygana.
[phulis] =tokoy {nuk -ram -phin -ok} =no
police =LIKE look.like -UNINTENTIONALLY -TOTALLY -COS =QUOT
[bunduk |{pay}| =ga] =na
gun carry.in.hand =ATTR =DAT
[They] unintentionally looked totally like the police because of the guns
carried.

We can conclude that although Atong has no morphosyntactic means to distinguish
transitive from ambitransitive clauses, it is possible to interpret certain clauses, in
contexts where no A argument can be recovered, as being S=O ambitransitive.
21.4 Pivots
There seem to be some co-reference restrictions on subjects (S/A) in certain clause
combinations in Atong, examples of which are presented in a later section of this
grammar to which I will only make reference in this section. Because of these
restrictions, Atong can be said to have pivots (see Dixon 1994) which constrain clause
combinations and possibly the occurrence of more than one S/A for the two combined
clauses, but this topic needs further fieldwork research. The subject (S/A) of dative-
marked complement clauses (see 27.2.1) of purposive clauses (see 27.2.3) and of
adverbial clauses (see 28.1) is always co-referential with the subject of the matrix
clause. When the subjects are not co-referential, other syntactic constructions must be
used.
In all other clause combinations, there are no co-reference restrictions of any kind,
e.g. (519). In this example the O argument of the sequential clause (see 28.2), viz.
cep=gaba (imprison=ATTR) the prisoner, is coreferential with the S argument of
the main clause, viz. ge'the he. In a language like English, one would expect the
referent of ge'the he to be the implied A of the sequential clause. In Atong there
are no such grammatically determined expectations about the reference of the third
person singular, which is pragmatically determined.

21 TRANSITIVITY

371
(519) otokoymo cepgaba deaymo ge'the hokotokno.
otokoymo [cep =gaba]
O
{de} =ay =mo
so.then imprison =ATTR untie =ADV =SEQ
[gethe]
S
{ho'khot -ok} =no
3s come.out -COS =QUOT
So then, having untied the prisoner
j
[Lit. the imprisoned], he
j
came out, it is
said.

Another example of a lack of co-reference restrictions is (520), where we see a reason
clause (see 27.1.1), syntactically subordinate, where the implied O of the transitive
verb tan'- to cut (up), to slay is coreferential with the S of the verb in the matrix
clause predicate thoy- to die.

(520) tan'manokona thoyok udo, mo
{tan' -man -ok} =ona {thoy -ok} [u] =do mo
cut -already -COS =DAT die -COS DST=TOP CONF
Because [they] had cut [him
j
] up, [he
j
] died, that one, isnt it.

372
Chapter 22 The Predicate
_____________________________________________________________________
22.1 Defining the predicate and the predicate head
The predicate is the part of the clause which contains the predicate head and anything
semantically tightly linked or prototypically associated to that head. Words from
several different word classes can function as predicate head, viz. verbs, Type 2
adjectives, demonstratives, personal pronouns, numeral-plus-classifier compounds,
the interrogatives bisa to/from where? and biskon how much/many? and nouns.
Table 62 below summarises the properties of the different types of predicate heads.
As we can see, there are no clear boundaries between the different types; their
properties overlap. We will refer to predicates with a noun as head as nominal
predicates, to those with verbs as head as verbal predicates, etc. First, in 22.2 we
will look at the morphological structure of the predicate head. Verbal predicates are
treated in 22.3, Type 2 adjectival ones in 22.4, and verbal ones in 22.5.
There are two types of predicate, viz. simple and complex. Only verbs and Type 2
adjectives can be the head of a complex predicate. In complex predicates, the head is
always the right-most constituent. It is only the head that can take predicate suffixes,
although in complex predicates the constituent which is not the head can carry
enclitics. Complex predicates without incorporated nouns will be treated in 22.6, and
those with incorporated nouns 22.7.
22 THE PREDICATE

373
Table 62 Properties of different types of predicate depending on the head
Type of head
VERB TYPE 2 ADJECTIVES NOUN
Syntactic criteria
can occur as the predicate of a non-finite clause
can take all argument types can only take S arguments that are Attributants
can be the predicate of an
imperative clause
occurs only in identity/equation clauses
cannot be a predicate of an imperative clause
can be modified by adverbial clauses and adverbs
cannot be modified by adverbial
clauses and adverbs
Morphological criteria
can be the head of a complex predicate
cannot be the head of a complex
predicate
can take all event specifiers
difficult to use with event
specifiers
can take the causative suffix <et> (CAUS) cannot be causativised
cannot occur as bare stem as
predicate of a main clause,
except in imperative clauses
can also be used without any marking, i.e. as bare root
does not have to express but can
take most of the same categories as
a verbal predicate head but is not
attested with all of them, can occur
without predicate suffixes in main
clauses
can express fewer categories
than verbal and adjectival
predicates, can occur without
predicate suffixes in main
clauses
22.2 The morphological structure of the predicate head
A predicate head consists of a root, followed optionally by one or more stem-forming
suffixes, followed optionally by one or more inflectional suffixes indicating negation,
aspect, modality and polarity. An overview of all predicate suffixes is given in Table
63. As we can see in that table, the suffixes are ordered in echelons and columns.
There are four columns in Echelon 1 and three in Echelon 2. The suffixes in Echelon
1 appear on both main and subordinate clause predicates. Some suffixes in Echelon 2
only appear on subordinate clause predicates, some only on main clause predicates
and some on both. The suffixes that appear in the same column are not attested to
occur simultaneously. The only exceptions are the event specifiers. Although all event
specifiers are ranked in column 2, more than one event specifier can occur on a single
predicate head. The function and meaning of all the individual predicate head suffixes
is treated in Chapter 23. Event specifiers are treated in Chapter 25.
Suffixes from Echelon 1 and 2 are never attested to occur in the reverse order.
Within Echelon 2, suffixes from column 5, 6 and 7 do not occur in other orders. The
columns in Echelon 1, however, do not represent absolutely fixed positions, but rather
a strong tendency for morphemes to appear in a certain order. Variations in the order
are attested. These variations might signify variations in scope of the suffixes, but
22 THE PREDICATE

374
there is also a chance that some variations are in free variation. More fieldwork is
needed to find out more about this positioning variability. A frequently attested

Table 63 Predicate head suffixes in their respective slots.
Morphemes that are attested on nominal predicate heads and on clauses
with nominal predicate heads are in bold face.

Echelon 1 Stem-forming suffixes
column 1
Causative <-et> (CAUS)
column 2
Event specifiers
There are a great number of event specifiers in Atong, possibly hundreds,
not all of which have been recorded yet. Those which have been
discovered fall into twelve categories, viz. manner, manner/direction,
aspect, extent, direction/extent, direction, epistemic, deontic, determinacy,
location, conative and quantification. Event specifier suffixes are treated in
Chapter 25.
column 3
Reciprocal <-ruk> (RC)
Comparative <-khal> (CP)
Excessive <-duga> (XS)
column 4
Simplicitive <-ari> (SIMP)
Incompletive aspect <-khu> (INCOM)

Echelon 2 Inflectional suffixes
column 5
Main clause predicate suffixes
Customary aspect <-a> (CUST)
Desiderative <-na> (DESI)
Future modality <-ni> (FUT)
Imperious future modality <-naka> (IFT) (in non-negative predicates)
Referential <-an> (REF)
column 6
Main and non-main clause predicate suffix
Negative <-ca> (NEG)
column 7
Main clause predicate suffix
Imperious future modality <-ka> (IFT) (after the negative suffix)
Main and non-main clause predicate suffixes
Factitive <-wa> (FACT)
Concomitant action <-butu>(WHILE)
Change of state <-ok ~-ak ~ -k> (COS)
Progressive/durative aspect <-aydoa ~ -aydo ~ -aydok ~ -aroa ~ -aro
~ -arok> (PROG)
22 THE PREDICATE

375
example of the variable position of two Echelon 1 suffixes is given bolow. In
(521) the progressive suffix precedes the incompletive suffix and in b) the situation is
reverse. Both clauses in have the same meaning.

(521) Variation in the position of the progressive and incompletive aspect suffixes:
a) sa' -aydo -khu -a
eat -PROG -INCOM -CUST
b) sa' -khu -aydo -a
eat -CUST -PROG -CUST
[I] am still eating.

The two examples below show that Echelon 1 suffixes can be combined to indicate
two contrastive views of the same event on one predicate. In the English translations
we need to express with two clauses what Atong does with one. In example (522)
(from TEXT 2, line 59) we see how the meanings of incompletive and negative
suffixes contrast with the meaning of the future suffix.

(522) hayda ray'akhucaaronikhon.
hayda ray'a -khu -ca -aro -ni =khon.
I.dont.know come -INCOM -NEG -PROG -FUT =SPEC
I dont know. He has not come yet but he might still be coming.

The next example illustrates how two contrastive directions of the movment expressed
by a motion verb are expressed on one predicate. The suffix <-a> indicates
movement away from the deictic centre, while the suffix <-theri> indicates the
opposite.

(523) londonsa jalatheriaymu
london =sa jal -a -theri =ay =mu
London =MOB run.away -AWAY -BACK =ADV =SEQ
having run away to London and back again

The stem-forming suffixes in Echelon 1 can occur in combination with all of the
Echelon 2 suffixes. Suffixes from Echelon 1 can occur simultaneously, if they are
semantically compatible. Co-occurrences of Echelon 2 suffixes are heavily restricted.
22 THE PREDICATE

376
Suffixes from column 5 and 7 never co-occur. The imperative, customary aspect and
both future modality suffixes in column 5, cannot co-occur with the negative suffix
<-ca> (NEG). The only column 5 suffix that is attested to cooccur with the negative
suffix is the referential suffix <-an> (REF). The imperious future suffix has two
allomorphs, viz. <-naka> (IFT), which occurs in non-negated predicates, and the
allomorph <-ka> (IFT), which occurs after the negative suffix <-ca> (NEG). All
column 7 suffixes can occur under negation.
22.3 The verbal predicate
The verbal predicate can take all types of arguments and can occur in all clause types
depending on the verbs valency and its subtype (for subtypes of verbs see Table 21).
For example Type 1 adjectives (stative verbs denoting a quality) are not usually
attested as heads of imperative clauses, unless causativised. This is of course due to
the semantics of these verbs, which are not usually combined with the semantics of
the imperative category. An Atong speaker does not often want to order someone to
have a certain physical property, age, dimension etc.
43
The only Type 1 adjective
found as head of an imperative predicate is tharak- fast in (27), repeated here as
(524).

(524) tarakboto na'a!
{tarak} =bo =to [na'a]
Be fast, oh you!

There are two types of verbal predicate, viz. simple and complex. The simple
predicate consists only of a head possibly followed by predicate suffixes. Complex
predicates consist of the head and a semantically tightly linked or prototypically
associated preceding lexical item, i.e. a verbal root with a limited set of possible

43
For a list of different semantic categories expressed by Type 1 and 2 adjectives, see Table 26 in 5.1.
22 THE PREDICATE

377
enclitics or a bare nominal root. Only verbs, Type 1 and Type 2 adjectives can occur
in complex predicates. Complex predicates are treated in the next section.
Except in imperative clauses, the verbal predicate of a main clause cannot occur
without an Echelon 2 predicate suffix expressing either negation, aspect or modality,
or a combination of these. Verbal predicates of non-main clauses can occur without
predicate suffixes, but in these cases the clause will have a clausal enclitic attached to
it, marking the clause as subordinate (see Chapter 27 and Chapter 28).
22.4 The type 2 adjectival predicate
Type 2 adjectival predicates share properties with both verbs and nouns (see Table 20
in Chapter 3). A Type 2 adjectival predicate cannot occur as the head of an imperative
clause, unless causativised with the suffix <-et> (CAUS). A Type 2 adjectival
predicate can occur without any suffixal marking (525), as can nominal predicates.
Type 2 adjectival predicates are intransitive and can only have an S argument that is
semantically an Attributant (see Van Valin and LaPolla 1997:115).

(525) ie ram thombolo
[ie ram]
Attributant/S
{thombolo}
PROX road have.holes.in.it
This road is damaged.
22.5 The nominal predicate
Nominal predicates can express negation, and aspect, but not modality. Aspect
marking on nominal predicates is reminiscent of what Nordlinger and Sadler (2004)
describe as independent nominal TAM in that the aspectual and modality suffixes
specify information intrinsic to the nominal itself (2004:778). The marking in
Atong is different from Nordlinger and Sadlers description in that they mostly
discuss TAM on NPs in NP functions, i.e. as arguments or obliques, and not on nouns
functioning as predicates. The same clausal enclitics that occur on clauses with verbal
predicates also occur on clauses with nominal predicates. Nouns cannot occur as
predicates of imperative clauses. Table 63 shows the suffixes that are attested on
nominal predicates in bold.
22 THE PREDICATE

378
It is important to note that nominal predicates occur in main as well as in
subordinate clauses. Main and subordinate clauses will be discussed separately.
Nouns functioning as predicate head can still carry NP enclitics, except case-marking,
as we will see in the examples below.
22.5.1 Main clause nominal predicates
Nominal predicates of main clauses occur only in identity/equation clauses (this label
includes identity, equation and attributive clauses which are all formally
indistinguishable) and can only take S arguments that are semantically Attributants.
When a clause consists of just two nouns, which are both devoid of suffixes, it is
impossible to determine on formal grounds which one is the predicate, given that in
main clauses any constituent can be right dislocated, i.e. occur after the predicate, for
backgrounding. The predicate has to be determined on semantic and pragmatic
grounds. In pragmatically unmarked clauses, the predicate is the final constituent.
Example (526) shows an identity/equation clause with two nouns, neither of which
carries any predicate suffixes. Given the situation in which the utterance occurs, viz.
in a kitchen where someone is cooking, and the semantics of both nouns, the hearer
can deduce that the curry, ja'bek, has to be the thing talked about, and therefore the
Attributant, and that pumpkin, gomonda, is the predicate. The enclitic <=thara>
exclusively is an NP enclitic.

(526) jabekan gomondathara
[jabek] =an
Attributant/S
{[gomonda] =thara}

curry =FC/ID pumpkin =EXCLUSIVELY
The curry is only pumpkin.

The focus/identifier enclitic <=an> (FC/ID) does not help in the determination of the
Attributant, since it can occur both on NPs and clauses, depending on its scope. In
(526) the enclitic has phrasal scope and in (527) clausal scope.

(527) jabek gomondatharaan
|[jabek]
Attributant/S
{gomonda -thara}| =an
curry pumpkin -EXCLUSIVELY =FC/ID
The curry is only pumpkin.

22 THE PREDICATE

379
Only when nominal predicates carry predicate suffixes are they formally recognisable
as predicates. The example below shows two main clauses in apposition. The NP san
tham (day three) three days is functioning as main clause predicate, and because it
carries the change of state suffix <-ok> (COS) it can only be interpreted as a
predicate.

(528) santhamok karen ni'wa.
{san tham -ok} [karen] {ni' -wa}
day three -COS electricity not.exist -FACT
It has been three days [and] there is no electricity.

Main clause clausal enclitics also occur on clauses with nominal predicate heads.
Example (529) illustrates the use of the speculative aspect enclitic <=khon> (SPEC)
and the declarative enclitic <=te> (DCL) on a main clause with a nominal predicate.
The use of the irrealis on a main clause with nominal predicate is illustrated by
example (530). Clausal enclitics are treated in Chapter 26.

(529) na'tome bobamorotkhonte.
|[na' -tom]
S
{boba morot}| =khon =te
2s -ppp crazy.person person =SPEC =DCL
Maybe you all are crazy persons, Ill say!
22.5.2 Subordinate clause nominal predicates
Nominal predicates are attested in the following clauses: the reason clause (see
27.1.1), e.g. (745), the concomitant action type locative-marked clause (see 27.6),
e.g. (783) below and (530) below, and the sequential clause (see 28.2), e.g. (808). In
all these cases the predicate is followed by at least one clausal enclitic and is therefore
formally identifiable as predicate. In the following example we see the nominal
predicate bay'sega-tha-maran (friend=OWN=RC) followed by the irrealis clausal
enclitic <=com> (IRR).

(530) ge'thegthe balrukbutuci bay'segathamarancom.
[ge'thethe] {bal -ruk -butu} =ci {[baysega] =tha =maran} =com
3p speak -RC -while =LOC friend =OWN =RC =IRR
When they were speaking to each other, they were no longer friends.
22 THE PREDICATE

380
22.5.3 Not only nouns
Nominal predicates do not have to be only nouns. Other non-verbal word classes
can fulfil this function as well, viz. demonstratives, personal pronouns, numeral plus
classifier compounds and even two interrogatives, viz. bisa to/from where? and
biskon how much/many?. The interrogatives, apart from being able to take event
specifiers, can only take the change of state suffix <-ok> (COS) (for examples see
Chapter 9). Demonstratives are attested to take negation, irrealis and speculative
modality suffixes (examples can be found in Chapter 1). Personal pronouns are
attested to express only negative, change of state, speculative modality, and irrealis
(see 17.2 for examples).
22.6 Complex predicates
Complex predicates can be headed by either a verb or a Type 2 adjective. There are
four types of complex predicates in Atong, divided into two categories:
A. Complex predicates without nouns (22.6).
1. Complex predicates with two identical verbs or Type 2 adjectives (22.6.1).
2. The Type 2 adjective-plus-support-verb predicate (22.6.2).
B. Complex predicates with incorporated nouns (22.7)
3. The predicate with a prototypically associated noun (22.7.1).
4. The nouns-plus-support-verb predicate (22.7.2).
22.6.1 Complex predicates with identical verbs or Type 2 adjectives
The first type of complex predicate consists of two identical verbs or Type 2
adjectives and is emphatic or bestows a greater intensity on the event denoted by the
predicate. The first verb or Type 2 adjective is marked as Topic, Focus or emphatic
with one of the following enclitics: topic <=do> (TOP), focus <=e> (FC),
focus/identifier <=an> (FC/ID) or emphatic <=ba> (EMPH). These enclitics are
attached to the first verb or Type 2 adjective, which cannot take any other affixes or
enclitics and can thus not be negated separately from the predicate head. The second
verb, the predicate head, is marked for other categories. Nothing can intervene
between the two verbs or Type 2 adjectives. The order of the verbs/Type 2 adjectives
is fixed. The two verbs/Type 2 adjectives share transitivity and share all their
22 THE PREDICATE

381
arguments. All types of verbs and Type 2 adjectives can occur in this type of complex
predicate. Complex predicates of this type are only attested in independent clauses. In
examples we see complex predicates in which the first constituents are marked with
the emphatic enclitics <=ba> (EMPH) and the topic enclitic <=do> (TOP)
respectively.

(531) ucian gadakba-gadakjolokno
ucian {gadak =ba gadak -jol -ok} =no
then cut.in.pieces =EMPH cut.in.pieces -QUICKLY -COS =QUOT
Then, as for cutting, [he] cut [the big eagle] quickly in pieces, it is said.

(532) ado sa'do sa'ak.
[a] =do {sa' =do sa' -ak}
1s =TOP eat =TOP eat -COS
As for me, as far as eating is concerned, [I] have eaten.

Further fieldwork needs to be conducted to find out how many nominal properties the
first verb of a complex predicate with identical verbs has before we can say that this
verb is nominalised or not. Example (533) illustrates a complex Type 2 adjectival
predicate in which the first constituent is marked with the focus enclitic <=e> (FC).

(533) na' abu bodoye bodoyok.
[na' abu] {bodoy =e bodoy -ok}
2s grandmother old =FC old -COS
Your grandmother is really old.

One example has been recorded of a complex predicate in which the first verb carries
the focus/identifier enclitic <=an> (FC/ID), viz. (534). The reason why in this
construction it is the focus/identifier enclitic <=an> (FC/ID and not the referential
suffix <-an> (REF) is that the referential suffix <-an> (REF) never occurs without the
negative suffix <-ca> (NEG). This complex predicate construction is clearly a means
of emphasising or reinforcing the action denoted by the verbs in the predicate.

22 THE PREDICATE

382
(534) lukwake ro'aw jamcano. uci magacakdo jaman jamsoracano.
[lukwak] =e [ro'] =aw {jam -ca} =no [u] =ci
toad =FC stone =ACC complete -NEG =QUOT DST=LOC
[magacak] =do {jam -an jam -sora -ca} =no
deer =TOP complete =FC/ID complete -COMPLETELY -NEG =QUOT
The toad could not complete [move] the stone, it is said. Then, the deer could
not complete [move] [the stone] at all, it is said.

As the following example demonstrates, the locative/existential verb gana to exist
has a truncated root ga- which is only found in this complex predicate.

(535) gado ganacom, gawi thogiok.
{ga} =do gana} =com [gawi] {thogi -ok}
exist =OP exist =IRR girl cheat -COS
I did have [a girlfriend] but not any more, the girl cheated on me.
22.6.2 Type 2-adjective-plus-support-verb compounds
Bare roots of Type 2 adjectives can be compounded with the support verb tak- to do.
The two elements are tightly knit together, the bare root form of the verb modifying
the light verb predicate head. The verb in the bare root form provides the lexical
content for the predicate while the head carries the grammatical information
concerning negation, modality, aspect etc. The support verb compound can add a
sense of liveliness or irony to the clause. In the following examples we see the Type 2
adjectives cega upright, miniksuru flat-haired and cu'ret stuck with the
support verb tak- to do.

(536) co'sa ropaymo jarawacian mon'an cega takariano.
co'sa {rop} =ay =mo {jaraw -wa} =ci =an
a.little.bit be.in.the.water =ADV =SEQ be.a.long.time -FACT =LOC =FC/ID
[mo'] =an] {cega tak -ari -a} =no
body.hair =FC/ID be.upright do -SIMP -CUST =QUOT
Having been under water a bit for some time, [the foxs] hair was still
upright.

22 THE PREDICATE

383
(537) otokoymo ha! wenni ropwacian miniksuru takokno solokno magacakmi
non'do.
otokoymo [ha] [wen' -ni] {rop -wa} =ci =an
so.then interj turn/time -two be.in.the.water -FACT =LOC =FC/ID
{miniksuru tak -ok} =no
be.flat-haired do -COS =QUOT
{sol -ok} =no [magacak -mi mon'] =do
pretty -COS =QUOT deer =GEN body.hair =TOP
So then, ha! when he was in the water the second time, [it] was flat-haired
and very pretty, the deers fur.

(538) ri'do cu'ret takawano ca'masa na'pitsa.
[ri'] =do {curet tak -a -wa} =no [ca'ma] =sa [na'pit] =sa
penis =TOP stuck do -AWAY -FACT =QUOT down =MOB barber =MOB
[The foxs] penis was stuck in a downward direction towards the barber.

The verb guduk- means to wobble, to move in an unstable way and can be used in a
compound with the verb tak- to do. The result is a lexicalised compound, viz. guduk
tak- to almost VERB. This predicate requires a verbal complement marked with the
dative marker <=na> (DAT), e.g. (539) and (540).

(539) te'do tharapna guduk takwaci ha'tokoy jalwano magacake
[te'e] =do {tharap} =na {guduk tak -wa} =ci [ha'] =tokoy
now =TOP catch.up =DAT almost do -FACT =LOC ground =VIA
{jal -wa} =no [magacak] =e
run.away -FACT =QUOT deer =FC
Now when [he] almost caught up with [the deer], [it] run away, the deer.

(540) okhana guduk takaydo
{okha} =na {guduk tak -aydo}
be.full =DAT almost do -PROG
I am almost full.

The following example shows a complex predicate with two Type 2 adjectives before
the head.
22 THE PREDICATE

384
(541) ca'e dabakuntokoy omo ja'jot takarioknotoy.
[ca'] =e [daba kun] =tokoy
leg =FC coconut stick =LIKE
{ompo ja'jot tak -ari -ok} =no =toy
bulging narrow do-SIMP -COS =QUOT =MIR
[His} legs were just like a coconut stick bulging and narrow, it is said to [our]
surprise.
22.7 Complex predicates with incorporated nouns
This section describes the phenomenon whereby a noun is incorporated into a
predicate with a verbal head. We are not dealing here with noun incorporation that
refers to a type of compounding in which a verb and a noun combine to form a new
verb (see Aikhenvald 2007: 11), because in this construction, morphemes can
intervene between the noun and the verb, as we will see below. What gets
incorporated into the predicate is the free form of the noun, sometimes even with an
emphatic enclitic as in (546). The noun and the verb are simply juxtaposed and form a
tight bond. The relationship between the noun and the verb is somewhere in the
middle on the cline between an argument/adjunct-predicate relationship on the one
hand, and incorporation through compounding on the other. The noun loses its
syntactic status as argument/adjunct of the clause (see Mithun 1984: 849). The
incorporated noun and the verb, i.e. the predicate head, can form one phonological
word but do not have to. Atong distinguishes two types of predicates with
incorporated nouns, i.e. nouns incorporated into support verb predicates and
predicates with a prototypically associated noun which is I call prototypically
associated noun incorporation. They are treated separately below.
22.7.1 The predicate with a prototypically associated noun
There are transitive and intransitive verbs which can be, but do not have to be,
accompanied by a prototypically associated noun which is not referential and does not
belong to the argument structure (i.e. is not an argument or adjunct) but is
incorporated into the predicate. The arguments for the non-argument status of this
type of noun are:

22 THE PREDICATE

385
they are non-referential,
they have a fixed position before the predicate head, whereas
arguments can appear in post predicate position,
they cannot receive case marking,
they cannot be modified,
they are not affected by the action denoted by the predicate head,
their meaning is closely semantically related to or specifies the event
denoted by the verb that is the head of the predicate. Both the noun and
the verb denote a unitary activity in which the components lose their
individual salience,
they cannot be referred to by quantifying event specifiers.

Hence I consider the prototypically associated noun to be part of the predicate, e.g.
(544), (545) and (546). Most of these arguments are also mentioned in Mithun (1984).
The loss of argument/adjunct status is most clearly seen in predicates where the
semantic relationship between the noun and the verb is that of Location, e.g. toy hu-
to water-swim where toy water is not an adjunct because it is not locative-marked.
Location adjuncts (peripheral arguments) are otherwise obligatorily locative-marked,
e.g. toy=ci hu- (water=LOC swim) to swim in the water.
As mentioned before, the prototypically associated noun is not obligatory. All
verbs with which they occur can also be used without these nouns. The incorporation
of the prototypically associated noun is a pragmatic mechanism, making the predicate
more explicit. I call this phenomenon prototypically associated noun incorporation.
The relationship between the prototypically associated noun and the predicate is
somewhere in between an argument/adunct-predicate relationship and incorporation
through compounding. Although the noun is not case-marked, it does occupy an
argument/adjunct-like position in front of the verb. However, since the noun cannot
be case-marked, the relationship between the noun and the verb is semantic rather
than syntactic. The semantic relations that are so far attested between the verb and the
incorporated noun are Patient, viz. ha' haw- (land/earth/soil clear) to clear the land
and may sa' (rice eat) to eat, Instrument, viz. nakhal na- (ear hear) to hear and
mokren nuk- (eye see) to see, Location, viz. ha' kon (earth/land/soil collect) to
collect the remaining cinders after burning the jungle and ha' kham (earth/land/soil
22 THE PREDICATE

386
burn) to burn the land (although what is actually burnt is the jungle) and Pathway,
viz. ram ray'- (road go) to go. Only the compounds nakhal na- (ear hear) to hear
and mokren nuk- (eye see) to see are still transitive, e.g. (548). The other noun-verb
constructions cannot take an O argument any more. One could say that the nouns
nakhal ear and mokren eye are more incorporated than the other nouns.
A good example of incorporated Patient can be found in the Text 2 line 65,
represented here as (542).

(542) may sa'hokokma na'tome?
{may sa' -thok -ok} =ma [na -tom] =e
rice eat -ALL -CoS =Q 2s -ppp =FC
Have you all eaten? Literally: Have you all rice-eaten?

In this example the word may rice is incorporated in the predicate. The quantifying
event specifier -thok all refers on an S/O basis. If may rice functioned as O
argument, the suffix thok all would have to refer to it. However, in this clause the
suffix thok refers to na'tom you.plural, which is in S function in this clause, and
not to may rice because may rice is not an O argument but part of the predicate.

Table 64 Prototypically associated nouns with their verbs.
SEMANTIC
RELATIONSHIP
BETWEEN NOUN
AND VERB
COMPOUND FORM GLOSS OF PARTS TRANSLATION
Theme jagi ke- life live to live
Theme jagi hoy- life die to die
Instrument mokren nuk- eye see to see
Instrument nakhal na- ear hear to hear
Location toy hu- water swim to swim
Location ha' khon- earth/soil/land collect to collect the
remaining cinders
after burning the
jungle
Location ha' kham- earth/soil/land burn to burn the land
(actually what is
burnt is the
jungle)
Pathway ram ray- road go to go
Patient ha' haw- earth/soil/land clear to clear the land
Patient may sa'- rice eat to eat
22 THE PREDICATE

387
The following example shows an incorporated Pathway, ram road.

(543) utokoymu ie ha'boritokoy ramra'ano.
utokoymu [ie ha'bori] =tokoy {ram ray' -a] =no
CONJ PRX hill =VIA road go -CUST =QUOT
So then, they went via this mountain, it is said Lit. they road-went over this
mountain.

Prototypically associated nouns can have a detransitivising effect on transitive verbs
when they get incorporated, e.g. (544). In that example the transitive verb nuk- to
see is detransitivised by the associated noun mokren eye. The semantic relation of
the noun to the verb is that of Instrument.

(544) na' walci mokrennukama?
[na'] [wal] =ci {mkren nuk -a} =ma
2s night =LOC eye see -CUST =Q
Do you see at night? Lit. Do you eye-see at night?
Since mokren eye in the above example is low on the animacy hierarchy, it is not
likely to be interpreted as A. This would then mean that wal night was in a
Possessor-Possessed relationship with the personal pronoun na' and then we would
get a ridiculous meaning like *[Does person X] see eyes in your night? Moreover,
the interpretation as unmarked instrument for mokren eye is not possible since
instruments cannot be unmarked for case.
The next example illustrates the incorporation of the noun toy water which is
semantically a Location in relation to the verb hu- to swim.

(545) na' toyhuna sapama?
[na'] {toy hu} =na {sap -a} =ma
2s water swim =DAT know.a.skill -PUR =Q
Do you know how to swim? Lit. Do you know to water-swim?

Since toy water in the example above is not an inherently locational noun, it cannot
be interpreted as an unmarked location.
In (546) we see the incorporated noun jagi life which is a Theme in relation to
the verb thoy- to die.
22 THE PREDICATE

388
(546) aa jagiba thoymanok.
[aa] {jagi =ba thoy -man -ok}
1s life =EMPH die -ALREADY -COS
I have already died. Lit. I life-died.

Since the verb thoy- to die is intransitive and aa I is the S argument, jagi life
cannot be interpreted as O. Moreover, the emphatic pronoun aa (1s) cannot be in a
Possessor-Possessed relation to a following noun since emphatic pronouns cannot be
possessors. Thus the noun jagi life, despite the emphatic enclitic <=ba> (EMPH),
is not an argument/adjunct and therefore has to be incorporated into the predicate.
It is always possible for the prohibitive free morpheme <ta> (PROH) to come in
between the prototypically associated noun and the predicate head, as can also be seen
in examples (547) with the verb hu to swim and its prototypically associated noun
toy water and in (548) with the verb na to hear and its prototypically associated
noun nakhal ear, which is still transitive, as we can see by the accusative-marked O
argument.

(547) toy ta hu.
{toy ta hu}
water PROH swim
Dont swim. Literally: Dont swim water.

(548) noksemi malgabaaw nakhal ta na!
[[nokse] =mi {bal} =gaba] =aw {nakhal ta na}
Name =GEN say =ATTR =ACC ear PROH hear
Dont you listen to the things that Nikseng says!

Since prototypically associated nouns are unmarked they could be mistaken for S,
A or unmarked, non-referential O.
44
Prototypically associated nouns in complex
predicates differ from non-referential O arguments in a clause. The referent of a

44
Which arguments can occur unmarked and under what conditions is treated in Chapter 20.
22 THE PREDICATE

389
prototypically associated noun is not in any way affected by the action denoted by the
predicate head, i.e. it cannot be interpreted as an unmarked O argument. Neither are
they likely to be interpretable as A or S argument in the context in which they appear.
Rather, the complex predicate as a whole just denotes one event. The prototypically
associated noun denotes something that is circumstantially involved in the event
denoted by the predicate head like an Instrument in (544), a Location in (545) or
something that can be associated semantically with the predicate head, like the Theme
in (546), to make the predicate as it were more explicit.
The construction may sa- (rice eat) is used to indicate the action of eating while it
does not actually have to be rice that is eaten. The context will determine whether or
not the referent of the noun may rice is interpreted as affected by the action denoted
by the verb sa'- to eat and whether or not may rice can be interpreted as O
argument. It is impossible to draw the line between prototypically associated nouns,
which are not part of the argument structure, and non-referential O arguments, which
are. Therefore it might be a better analysis to consider a scale with on the one hand
case-marked referential Os and on the other hand prototypically associated nouns. In
(549) for example, may rice is non referential and might or might not be seen as a
prototypically associated noun.

(549) may sa'akma?
[may]
O
{sa' -ak} =ma
rice eat -COS =Q
Have you eaten (rice)?

An argument in favour of may rice in (549) being an O argument is that it is affected
by the event denoted by the verb. Moreover, may rice can be accusative-marked and
be questioned as O argument. If we were to mark the noun may rice with an
accusative, it becomes referential, and the sentence would still be felicitous, viz. may
=aw sa'-ak=ma? (rice=ACC eat-COS=Q) have you eaten the rice? If, on the other
hand, we would put a case marker on a prototypically associated noun, and thus make
it referential and cast it out of the predicate into the argument structure, the clause
would have a very strange meaning, as in the following examples, which are
22 THE PREDICATE

390
manipulations of (544), (545) and (546) respectively. This test shows that these
compounds are lexicalised.

(550) ? na' walci mokrenaw nukama?
[na'] [wal] =ci [mokren] =aw {nuk -a} =ma
2s night =LOC eye =ACC see -CUST =Q
Do you see (your/the) eyes at night?

(551) * na' ty-iaw hu=na sap-a-ma?
[na'] [toy] =aw {hu} =na {sap -a} =ma
2s water =ACC swim =DAT know.a.skill -CUST =Q
Do you know how to swim the water?

(552) * aa jagiawba thoymanok.
[aa] [jagi] =aw =ba {thoy -man -ok}
1s life =ACC =EMPH die -ALREADY-COS
I died my life.
22.7.2 The noun-plus-support-verb predicate
Nouns can be incorporated into a complex predicate with the support verbs tak- to
do, kha'- ~ kha- to do, make, and ra'- to take, get. There are four arguments that
support incorporation instead of co-ordination of predicates. 1) The incorporated noun
is not referential. 2) The incorporated noun is not part of the argument structure
because the referent of the noun is not affected by the event denoted by the verb. 3) In
the support verb construction nothing can come between this noun and the inflected
form of the support verb. 4) The incorporated noun cannot be modified. The two
elements are tightly knit together. The noun provides the lexical content for the
predicate while the head carries the grammatical information concerning negation,
modality, aspect etc. More fieldwork is needed to find out what the criteria are for the
use of either tak- to do or kha'- ~ kha- to do, make as support verb in the complex
predicate. More fieldwork is also needed to find out whether the prohibitive word
<ta> (PROH) can come between the noun and the support verb.
22 THE PREDICATE

391
i The support verbs kha- ~ kha- to do, make and tak- to do
Any noun seems to be incorporable in a support verb predicate. This type of
incorporation is used in several cases:
where the language has no verbal equivalent for the NOUN-VERB compound, as
in (553) with the noun and bay'siga friend and in (554) with the noun degu
extortion inside the predicate;
to create the notion of pretending to do something, as in (555), where the noun
khora lame person is part of the predicate;
to express the notion be like something as in (556), where the noun nawa
confused person, idiot is part of the predicate.
Sometimes loans have to be incorporated in order to use them as we can see in (557).

(553) te'ewe amakmo rupekmo bay'siga kha'wano.
[te'ew] =e [amak] =mo [rupek] =mo {baysiga kha' -wa} =no
now =FC monkey =COM FROG =COM friend do -FACT =QUOT
Now the monkey and the frog are friends, it is said.

(554) a degu takni na'a, na'a payay jalbone noaydoano magacakan.
[a] {degu tak -ni} [na'a] [na'a] {pay} =ay
1s extortion do -FUT 2s 2s carry.in.hand =ADV
{jal} =bo =ne {no -aydoa} =no [magacak] =an
run.away =IMP =TAG say -PROG =QUOT deer =FC/ID
I will do extortion, oh you! you carry [the biscuits and] run away, ok? [he]
is saying, it is said, the deer

(555) a khora taknane, na'a payay jalbone bay'sigane noaydoano.
[a] {khora tak -na} =ne [na'a] {pay} =ay
1s lame.person do -DESI =TAG 2s carry.in.hand =ADV
{jal} =bo =ne [bay'siga] =ne {no -aydoa} =no
run.away =IMP =TAG friend =TAG say -PROG =QUOT
I will pretend to be lame, you carry [the biscuits and] run away, ok, friend,
right, he is saying, it is said.

(556) jagalan otokoy nawa takthokaroknoro.
[jagal] =an [otokoy] {nawa tak -thok -ok} =no =ro}
everybody =FC/ID like.that confused.person do -ALL -COS =QUOT =EMPH
So everybody was like really confused.
22 THE PREDICATE

392
In the next example the Indic loan thik agreement is incorporated into the support
verb complex predicate. In this example the word osokon is a phonologically reduced
form of isokon this much. The loan thik agreement also occurs in complex
predicates with the identity/equation copula do'- ~ do- (IE.be) of which example
(562) is illustrative.

(557) konsado beanbebe osokon san osokon somay thik ka'akno.
[konsa] =do [beanbebe][osokon san] [osokon somay]
after =TOP truly this.much day this.much time
{thik ka' -ak} =no
agreement make -COS=QUOT
Then they truly agreed on a certain day and a certain time for the event.

There are cases in which the support verbs tak- to do and kha'- ~ kha- to do, make
are juxtaposed to bare verbal roots or verbal stems. It is possible for verbal roots or
stems to modify immediately following predicates and therefore it is possible to
consider verbal roots or stems followed by the support verbs to be normal (zero-
derived) adverb plus verb constructions. It appears that other elements can intervene
between the two verbs in colloquial speech, although more fieldwork is needed to test
the grammaticality of the construction with intervening elements. Examples can be
found in 18.7.
ii The support verb ra- to take, get
There is a third support verb that allows incorporation of nouns into the predicate, viz.
ra'- to take, get. Contrary to the other two support verbs treated above, ra'- to take,
get only allows incorporation of a small number of specific nouns. The nouns
incorporated into predicates with ra'- to take, get which have been discovered up to
now are listed here in Table 65, followed by some examples.

22 THE PREDICATE

393
Table 65 Elements incorporated into predicates with the support verb ra'- to take
GLOSS
INCORPORATED
ELEMENT WORD CLASS COMPOUND TRANSITIVITY GLOSS
snore hogol noun hogol ra'- intransitive to snore
thought su noun su ra'- transitive to remember
word
khatha ~ khata
~ katha ~ kata
noun khata ra'- transitive to obey
respect man noun man ra'- transitive to respect

(558) na' ge'thegaw man ra'na naa
[na'] [ge'the] =aw {man ra'} =na {na -a}
2s 3s =ACC respect take =DAT need -CUST
You have to respect him.

(559) jowcewaci na' hogol ra'wa
{jow -ce -wa} =ci [na'] {hogol ra' -wa}
sleep -first -FACT =LOC 2s snore take -FACT
When you were asleep first, you snored

(560) nokgumuk khusidoay isolaw sug ra'ay, je kristen dogabado isolaw phi'ay
sa'cea.
[nok] =gumuk {khusi do =ay} [isol] =aw {su ra'} =ay
house =whole happy IE.be =ADV God =ACC thought get =ADV
[[je kristan] {do} =gaba] =do
any Christian IE.be =ATTR =TOP
[isol] =aw {phi'} =ay {sa' -ce -a}
God =ACC beg =ADV eat -FIRST -CUST
The whole house is happy [and] remembers God (Alt.: praises God) [and
then] whoever is a Christian, prays to God first [and] starts eating.
iii The copula as support verb
There are certain predicative constructions, all of which contain loanwords, which
require the presence of the identity/equation copula do'- ~ do- (IE.be) as support
verb to be able to occur in a clause. The word khusi happy in (561) is an Indic loan
as is thik agreement in (562) below. The word phel ~ pel [p
h
e:l ~ pe:l] fail comes
from the English word fail (563). All these loans are part of the predicate of the
clauses in which they occur. There is also an expression thik kha- (agreement make)
which means to agree upon a place and time for the event which is illustrated in
example (557) above.

22 THE PREDICATE

394
(561) ge'thethe khusi do'thamakaymo gore di'mayci phaltha cak diriga
sawalaymo watokno.
[ge'thethe] {khusi do' -thamak} =ay =mo
3p happy IE.be -EXCESSIVELY =ADV =SEQ
[gore di'may] =ci [[phaltha cak] {diri} =ga] {sawal =ay =mo
horse tail =LOC self hand hold =ATTR forget =ADV =SEQ
{wat -ok} =no
depart -COS =QUOT
They were so excessively happy that they forgot their own hands which were
holding the horses tail and they let go, it is said.

(562) isamo moaytanamangaba bimu aro usamy ray'agaba morot
cancicopay thik do'okodo []
[[i] =sa =mo {mo} =ay {tan -a -man} =gaba bimu]
PRX =IRIN =ABL call.a.name =ADV put -AWAY -already =ATTR name
aro [[u]=sa =mo {ray'a} =gaba morot]
and DST-IRIN =GEN come =ATTR person
{cancicop} =ay {thik do' -ok} =odo
suppose =ADV agreement IE.be -COS =TOP
Then now suppose that if the name that [someone] gave from this side and the
person from that side are in agreement, then [in] this game ajot the king will
certainly say ajot to him. (In the childrens game called ajot).

(563) ge'the lekha nemay poreanca, otokotymu poreka pel do'ok.
[ge'the] [lekha] {nem} =ay {pore -an -ca}
3s book good =ADV study/read -REF -NEG
otokoymu [poreka]
O
{pl do' -ok}
so.then exam fail IE.be -COS
He did not study the book well, so then he failed the exam.

Integrating or incorporating loans into the language by combining them with native
verbs is not uncommon in other languages as well. Haig (2001: 213) describes how
Iranian languages make extensive use of combinations of often borrowed nominal
elements plus a semantically bleached native support verb to extend their verb
lexicon. The same has been described for Japanese and Korean (see Sohn 1999: 254-
5 for Korean).

395

396
Chapter 23 Predicate head suffixes
_____________________________________________________________________

In this chapter the predicate head suffixes will be treated separately in the order in
which they appear in Table 63, except for the event specifiers, which are treated in
Chapter 25, the factitive suffix <-wa> (FACT), which is treated in Chapter 1, and the
concomitant action suffix <-butu> (WHILE) , which is treated in 27.6.1.
23.1 The causative suffix <-et>
The causative is signalled by the morpheme <-et> (CAUS). It occurs on both
transitive and intransitive verbs. The causative is not attested on type 2 adjectives or
nouns functioning as predicate. On intransitive verbs and adjectives the causative
increases the transitivity and the valency, i.e. it makes it possible for the verb or
adjective to have an A argument, semantically a Causer. On transitive verbs the
causative can have a valency increasing effect, or a different effect, which will be
discussed below.
The next example illustrates the occurrence of the causative on the Type 1
adjective tu- to be hot. Type 1 adjectives are a subtype of intransitive verb. In its
causative form tu-et (hot-CAUS) the meaning is straightforward to make hot, to
heat up.

(564) may tuetnima?
[may]
O
{tu -et -ni} =ma
rice hot -CAUS -FUT =Q
Shall [I] make the rice hot?
23.2 The causative on transitive verbs
On transitive verbs the causative can have a valency increasing effect, i.e. it adds a
Causer to the clause. However, the causative does not have to add any participants.
The causative on transitive verbs certainly does not always add the meaning of
causation. There are many instances of the use of a causative without any apparent
difference in the meaning of the verb to which it is attached. More fieldwork needs to
23 PREDICATE HEAD SUFFIXES

397
be conducted to find out what subtle meaning the causative adds to transitive verbs in
these cases.
The following example was uttered by a boy who wanted me to make a girl write
a letter to him. We can assume that the causative in this clause has a valency
increasing function. Another example with the verb say- to write where the
morpheme may or may not have a valency increasing function is (566).

(565) na'a citi sayetbo!
[na'a]
A
[citi]
O
{say -et} =bo
2s letter write -CAUS =IMP
You, make [her] write a letter!

In the following example, which was dictated to me by the speaker, the transitive verb
bal- to speak, say, tell is used with the causative for an apparent reason, i.e. the
speaker cant write, so he makes someone else write the letter for him. This means
that, given this context, one could analyse Miksrang as the Causer and the Causee
being ellipsed in the clause. However, it might also be that Miksrang does not want
Barbara to know that he cant write and so he uses the causative to stress his own
involvement in the action, or something else. Again, we can only conclude that more
fieldwork needs to be conducted to find this out.

(566) barbara, na'na miksra salam baletwa
[barbara] [na'] =na [miksra]
CAUSER
[salam]
O
{bal -et -wa}
Name 2s =DAT Name greeting say -CAUS -FACT
Barbara, Miksrang makes [someone] say greetings to you. (Via a letter
written by someone else because Miksrang himself is illiterate.)

The next example is the opening sentence of a letter. Here the author of the letter
might well have used the causative to express his intense involvement in the event.


398
(567) a na'na pa'ca khatadoraaw sayay baletna.
[a] [na'] =na [{pa' -ca} [khata] =dora =aw
1s 2s =DAT many -NEG word =p =ACC
{say} =ay {bal -et -na}
write =ADV say -CAUS -DESI
I am intending to writingly say some words to you.

In the next example the causative might well be attached to the transitive verb dow-
to add to express that the event was difficult or involved effort. More fieldwork is
needed to find out if the causative suffix can be left out in this example and if there
would then be a difference in meaning.

(568) phikhe phikhe romay na'amdo pipuk pheciba thoycagabakona pawayci
dowetwacie biphagaba man'nagabaaw gawigaba kumiri romacie na'lame
gudukokno.
phikhen phikhe {rom} =ay [na'lam] =do
raw RED cook =ADV fish =TOP
[pipuk] {phe} =ci =ba {[{thoy -ca} =gaba] -ak} =ona
stomach to.gut =LOC =EMPH die -NEG =ATTR -COS =DAT
[pawa] =ci {dow -et -wa] =ci =e
pan =LOC add -CAUS -FACT =LOC =FC
[[biphagaba] {man' -a} =gaba] =aw
husband obtain -TOWARDS -REF =ACC
[gawigaba] [kumiri] {rom -wa} =ci =e
wife Name cook -FACT =LOC =FC
[na'lam] =e {guduk -ok} =no
type.of.fish =FC wiggle -COS =QUOT
Cooking [it] raw, the fish, really when they [had] gut[ted] the belly [it] was/is
not yet dead, when [they] put [it] into the pan [with difficulty], that which the
husband obtained, when the wife Kumiri cooked [it] the fish wiggle[d] about.

The next example the morpheme <-et> (CAUS) seems to indicate telicity.


399
(569) ucie sosokol thow'gaba nukayok, ciakol. otokoy cayetwacie phalthaaw
nukok.
ucie [sosokol thow' =gaba] [nuk -ay -ok] [ciakol]
then deep.hole.in.the.ground deep =ATTR see -POS -COS well
otokoy {cay -et -wa} =ci =e [phaltha] =aw {nuk -ok}
like.that look -CAUS -FACT =LOC =FC self =ACC -see -COS
Then [he] saw a deep hole in the ground. When [he] looked like that [he] saw
himself.

What do other languages tell us about the semantic effects of causative marking on
verbs? In Manambu (Ndu family) the causative prefix kay- derives straightforward
causatives from intransitive verbs (see Aikhenvald, 2008). However, when used with
transitive verbs, it does not add any participants: instead, it marks manipulative effort,
forceful action or multiplicity and extent of the object (Aikhenvald, forthcoming b).
In Garo (Burling, 2004: 143-144) the causative apparently also has two functions
which resemble those of the Atong causative, which is not surprising given that the
two languages are closely related. In Garo intransitive, transitive and ditransitive
verbs can be causativised with the meaning make/cause to V. Burling remarks that
in some cases, a causative added to a transitive verb seems simply to emphasize the
transitive nature of the verb or indicates that the action denoted by the verb was done
with more force or that the O argument of the verb was fully affected.
45

23.3 The reciprocal suffix <-ruk>
The reciprocal morpheme <-ruk> (RC) is a stem-forming morpheme which signals
that the event denoted by the verb is reciprocal. It occurs only on transitive verbal
predicate heads, e.g. (571) and (572). It can occur on Type 1 adjectives functioning as

45
David A. Peterson (personal communication) says it is common for morphemes that have a causative
effect on intransitive verb to have an applicative effect with transitive verbs. Peterson treats
causative/applicative isomorphism this in his book on applicative constructions (2007: 64-66). The
causative morpheme could also have a reference-tracking function (LaPolla, personal communication)
in discourse. I need to do extensively more research than is possible within the time limits for this PhD
to find out what the exact interpretational possibilities are for predicates with what I now call the
causative suffix <-et> (CAUS).

400
predicate head when they are causativised and thus transitivised, e.g. sak-et-ruk- (red-
CAUS-RC) to make each other red but also when they are not causativised. In that
case the adjective expresses a quality that mutually holds for two things as in example
(570) here below. Type 2 adjecives are not attested with the reciprocal.

(570) na'nae so jan'rukok.
[na'na] =e so {jan' -ruk -ok}
1pi =FC village far -RC -COS
Our villages are very far apart from each other.

(571) jalaaymo kosado jan'gabami otokoy olrukokno.
{jal -a} =ay =mo [kosa] =do [{jan'} =gaba] =mi
run.away -away =ADV =SEQ after =TOP far =ATTR =GEN
otokoy {ol -ruk -ok} =no
like.this speak -RC -COS =QUOT
Having run away, later, from far [they] spoke to each other like this, it is
said. [The tiger and the lazy king after the lazy king had bitten the tiger in the
thigh and the tiger had run off].

(572) sosami sosigacina nawrukok tan'rukok. tanrukciba patok ni'wa.
[so sa] =mi [so] =siga =ci =na {naw -ruk -ok}
village one =GEN village =ALT =LOC -ALL scold -RC -COS
{tan' -ruk -ok} {tan' -ruk} =ci =ba [patok] {ni' -wa}
slay -RC -COS slay -RC =LOC =EMPH prison NEG.be -FACT
From one village to another [they] scold each other, slay each other. When
[they] slay each other there is no prison.

There is one example in the recorded corpus of the reciprocal preceding an event
specifer, represented here as (573). As we can see in Table 63, event specifiers usually
precede the reciprocal morpheme. This example shows that the order of the suffixes is
not fixed and probably depends on their scope.


401
(573) na'tome goy'-boysok man'phawa ie bolsie? noay sy'rukthoka.
[na' -tom] =e [goy' boysok] {man' -pha -wa}
2s -ppp =FC CLF:RESIDUE how.many obtain -IN.TOTAL -FACT
[ie bolsi] =e {no} =ay {so' -ruk -thok -a}
PRX year =FC say =ADV ask -RC -ALL -CUST
How many did you get this year? everybody asks. Lit. sayingly all ask
each other. Implied: How many baskets were you able to fill with rice this
year? (during the harvest).
23.4 The comparative/superlative suffix <-khal>
The morpheme <-khal> (CP/SUP) occurs on both Type 1 adjectives as well as other
verbs to form comparatives, superlatives and predicates with the meaning V more, V-
er, where V stands for any verb. Whether a predicate with the suffix <-khal> has to
be interpreted as a comparative or a superlative dependes on the context. The suffix is
labelled (SUP) when the context prompts a superlative interpretation and (CP) in all
other contexts. The suffix <-khal> (CP) functions as Index in comparisons, whether
the Parameter is qualitative, i.e. a Type 1 adjective, as we can see in (653), (654),
(750) (752), or any other verb, as in (574).

(574) ge'the ana dayay sa'khala.
STANDARD COMPAREE ----------MARK----------- PARAMETER INDEX
[ge'the] [a] =na {day} =ay {sa' -khal -a}
3s 1s =DAT be.bigger=ADV eat -CP -CUST
He eats more than me.

It has to be noted that the Comparee and the Mark are not compulsory for the clause
to be comparative, e.g. a=an cu-khal-a (1s=FC/ID big-CP-CUST) Im bigger.
The next two examples, from the story about a lazy king called Bil, illustrate how
the interpretation of the suffix <-khal> (CP/SUP) as comparative or superlative
depends on the context. In (575) <-khal> indicates a superlative, whereas the three
occurrences of this suffix in (576) all have to be interpreted as comparatives. In the
clause leading up to (575), king Bil says to the king of another country: I dont need
anything.


402
(575) na'a ana na'mo gore jalna rakkhalgabaaw hon'etaribo.
na'a a =na na' =mo gore jal =na rak -khal =gaba =aw
2s 1s =DAT 2s =GEN horse run =DAT strong -SUP =ATTR =ACC
hon' -et -ari =bo
give -CAUS -SIMP =IMP
You just give me your fastest running horse.

(576) gore jalna rakbebeokno. kha'sinkhalay jalkhalna noaymo ga'dukdukciba
rakkhalay rakkhalay jalariokno
gore jal =na rak -bebe -ok =no
horse run =DAT strong -TRULY -COS =QUOT
kha'sin -khal =ay jal =na no =ay =mo ga'dukduk =ci =ba
slow -CP =ADV run =DAT say =ADV =SEQ prod.with.leg =LOC =INDEF
rak -khal =ay rak -khal =ay jal -ari -ok =no
strong -CP =ADV stong -CP =ADV run -SIMP -COS =QUOT
The horse ran really quickly, it is said. Having told [it] to run slower,
whenever [he] prodded [it ] with his legs, [it] just ran faster and faster, it is
said.

The comparative suffix does not just function in comparative constructions, as the
example below illustrates.

(577) [] nemkhalciba nemkhalcaciba ue morotnado dokoksa caysakni.
{nem -khal} =ci =ba {nem -khal -ca} =ci =ba
good -CP =LOC =EMPH good -CP -NEG =LOC =EMPH
[ue morot] =na =do [dokdok] =sa {cay -sak -ni}
DST person =DAT =TOP a.little.while =DLIM look -APPROPRIATELY -FUT
[After having called upon all those who are rich and their family members,
after having invoked the spirits and after having eaten and drunk,] whether
[the sick person] is better or not better, [one] will maybe wait a little while [to
see how the persons condition will develop].

23.5 The excessive suffix <-duga>
The excessive suffix <-duga> (XS) indicates that the event denoted by the predicate is
carried out to an excessive extent or, in the case of a predicate indicating a quality, is
too much. This suffix is attested on verbal and Type 2 adjectival predicates.

403
(578) umynggymynci thogigaba morotaw agrai bebe ra'dugana baino.
umogomonci [{thogi} =gaba morot] =aw [agray]
therefore lie =ATTR person =ACC EXCESSIVELY
{bebe ra' -duga} =na =bay =no
truly get -XS =DAT =PROH =QUOT
Therefore [you] shouldnt too much believe persons who are lying, it is said.

(579) tibimi korogaba rakdugabutuci caba ni'etok.
[tibi =mi koro =gaba] {rak -duga -butu} =ci
television =GEN make.sound =ATTR strong -XS -when =LOC
[ca] =ba {ni' -et -ok}
who =INDEF not.exist -CAUS -COS
When the sound of the TV was too strong, someone turned [it] off.

(580) ie ram thmblongduga'a
[ie ram] {thombolo -duga -a}
PRX road have.holes.in.it (of roads/bridges) -XS -CUST
This road is too damaged.

The meaning of a Type 1 adjectives can still be reinforced by the change of state
morpheme (see 5.1) when the excessive suffix is attached, e.g. mol-duga-ak (small-
XS-COS) much too small, alternatively: has become too small.
23.6 The simplicitive aspect suffix <-ari>
The simplicitive expresses that the event denoted by the verb is executed with ease,
without any trouble or hindrance or without discussion or further ado, and is signalled
by the morpheme <-ari> (SIMP). The Atong simplicitive differs from the English
lexeme just in that <-ari> (SIMP) has no delimitative meaning. Example (581) is
illustrative of the use of this morpheme.

(581) a otokoy balaymo tanarinaka.
[a] [otokoy] {bal} =ay =mo {tan -ari -naka}
1s like.this speak =ADV =SEQ PUT -SIMP -IFT
I, having spoken like this, will just/simply stop now [without further ado].

404
23.7 The incompletive aspect suffix <-khu>
The incompletive aspect suffix <-khu> (INCOM) indicates that the action denoted by
the verb has not yet reached the point of completion or is still the case. It can also be
used to form the polite imperative (see 26.2).
When used together with the progressive/durative aspect suffix, both the order
incompletive-progressive/durative, e.g. (582), and progressive/durative-incompletive,
e.g.(583), are attested. The speakers judgement is that the former order, which has
been recorded more often (see Table 63) is better than the latter. Example (202)
illustrates the use of the incompletive on a nominal predicate head.

(582) sa'khuaydoa ge'thedo.
{sa' -khu -aydoa} [ge'the] =do
eat -INCOM -PROG 3s =TOP
Hes still eating.

(583) a pipuk babaaydokhua
[a pipuk] {baba -aydo -khu -a}
1s belly/stomach empty -DUR -INCOM -CUST
My stomach is still empty.

Under negation, the incompletive signals that something is not yet the case, e.g. (584)
and (585).

(584) sa'khuca.
sa' -khu -ca
eat -INCOM -NEG
[I] have not yet eaten.

(585) ado sawamigomon te'ewrawrawdo re'ena man'khucaaydoa.
[a] =do {sa -wa} =mi gomon te'ew -rawraw =do
1s =TOP be.ill -FACT =GEN cause now -continuously =OP
{re'e} =na {man' -khu -ca -aydoa}
go.away =DAT be.able -INCOM -NEG -PROG
As for me, [I] can not yet go to school because of [my] illness.

405
23.8 The customary aspect suffix <-a>
The customary aspect suffix <-a> (CUST) is used to express regularly recurring,
unchangeable or ongoing events. The customary aspect cannot be expressed under
negation. The following example presents prototypical instances of the use of the
customary aspect suffix. The example is taken from a story about the slash-and-burn-
cultivation as practices by the Atong people.

(586) ha'khonmanwamosa maysi khita. umo abodora cala, dacadora cala.
otokoymosa calmanwa macotwamosa may kay'cea.
[ha'] {khon -man -wa} =mo =sa [maysi] {khit -a}
soil collect/clear -ALREADY -FACT =SEQ =DLIM millet sow -CUST
umo [abo] =dora {cal -a}
then corn =p sow.by.making.a.hole.in.the.ground -CUST
[daca] =dora {cal -a}
type.of.shrub =p sow.by.making.a.hole.in.the.ground -CUST
otokoymo =sa {cal -man -wa}
so.then =DLIM sow.by.making.a.hole.in.the.ground -ALREADY -FACT
{macot -wa} =mo =sa [may] {kay' -ce -a}
finish -FACT =SEQ =DLIM rice plant -INCEPT -CUST
Having already cleared the charred detritus off the ground, millet is sown.
thereafter corn is sown, daca is sown. Only then, having already finished
sowing, [you] begin to plant the rice.

The next example is the conclusion to a story in which the storyteller makes a
statement, i.e. he says that something is always like that.

(587) ni jatdo otoka
[ni jat] =do {otok -a}
1p tribe =TOP do.like.this/that -CUST
Our tribe does like that. Alternatively: Thus are the ways of our tribe.
23.9 The desiderative suffix <-na>
The desiderative suffix <-na> (DESI) can occur on the predicate head of independent
clauses and indicates a desire or wish, e.g. (588), (589), or an implied impossibility,
e.g. in Text 1 line 36 and example (590). The desiderative suffix can occur on verbal
predicate heads of content questions as we can see in (588) and declarative clauses as

406
illustrated in (589). Desiderative-marked predicate heads of independent clauses of
any type cannot take any modality or polarity suffixes, i.e. inflectional suffixes, but
the stem-forming incompletive aspect is attested, e.g. in Text 1 line 36 and example
(590). (Types of predicate suffixes are treated in Chapter 23.) Dependent desiderative
clauses can express change of state. Although the desiderative suffix
<-na> (DESI) is a homophone of the allomorph <=na> of the dative enclitic <=na ~
=ona> (DAT), the two morphemes are clearly functionally distinct since they occur in
different clause types. It cannot be excluded that the two morphemes are
etymologically related. Dative-marked clauses are treated in Chapter 27.

(588) bisa ray'na bay'siga?
[bi] =sa {ray' -na} [bay'siga]
QF =MOB go -DESI friend
Where do you intend to go, friend?

(589) hay bay'siga, biskut sa'khawna
[hay] [bay'siga] [biskut] {sa' -khaw -na}
come.on friend biscuit eat -SNEAKILY -DESI
Come on, friend, I intend/want to steal the biscuits.

In the context of the next example a man is talking to his daughter about the fish traps
he had put up.

(590) ca'masaba cayok! khambaysaawba cayok. ca'masami cayciba matdam
sa'ak, khambaysami cayciba matdam sa'ak. biaw caykhuna? ana ni'ok
nookno.
[ca'masa] =ba {cay -ok} [khambaysa] =aw =ba {cay -ok}
downside =ADD look -COS upstream =ACC =ADD look -COS
[ca'masa] =mi {cay} =ci =ba [matdam] {sa' -ak}
downside =GEN look =LOC =ADD otter eat -COS
[khambaysa] =mi {cay} =ci =ba} [matdam] {sa' -ak}
upstream =GEN look =LOC =ADD otter eat -COS
[bi] =aw {cay -khu -na}
which =ACC look -INCOM -DESI
[a] =na {ni' -ok} {no -ok} =no
1s =DAT not.exist -COS say -COS =QUOT

407
I looked downstream. I also looked upstream. Whenever I looked at the one
downstream, the otters had eaten it. Whenever I looked at the one upstream,
the otters had eaten it. Which other one [is there] to look at? I have no more,
he said, it is said. Alternatively: Which other one can I/am I supposed to
look at?

The following example comes from the story of the prodigal son told by Kempai A
Sangma. Apparently, the son of the wealthy king has returned to his father after years
of wandering. The king sits in his palace and hears the rumours. He then thinks (591)
to himself:

(591) bici ray'aphin'khuna? aa iskon gamaw hanthietok. iskon jan'gaba sosa
de'the jalaok. bici ucie acina ray'akhuna?
[bi] =ci {raya -phin -khu -na} [aa] [iskon gam] =aw
QF =LOC come -RETURN -INCOM -DESI 1s so.much wealth =ACC
{hanti -et -ok}
divide -CAUS -COS
[[iskon] {jan'} =gaba so] =sa [de'the] {jal -a -ok}
so.much far =ATTR country =MOB 3s run.away -AWAY -COS
[bi] =ci [ucie] [a] =ci =na {ray'a -khu -na}
QF =LOC t hen 1S =LOC =ALL come -CP -DESI
What should [he] return for? I bestowed so much wealth on him. He ran away
to such a far country. What then should he return to me for?
23.10 The Future modalities
There are two morphemes indicating that an event takes place in the future, viz the
future sufix <-ni> (FUT) and the imperious future suffix <-naka ~ -ka> (IMF). The
suffix <-ni> (FUT) indicates the uncertain future, and the suffix <-naka> (IFT) a more
certain future. The differences between these two future modalities will be discussed
in greater detail below.
The future has characteristics of both tense and modality; it contains the epistemic
element of probability and the tense-related element future temporal location of an
event. How are we now going to decide whether the future in Atong is a modality, or
a tense? Bhat (1999: 175-177) argues that the notion of future would be temporal or
modal depending upon the prominence that the language attaches to the categories of
tense, aspect and mood respectively. Atong can be said to be an aspect and modality

408
prominent language since both categories play an equally important role in the
language.
There are two arguments in favour of calling the future categories modalities in
Atong. Firstly, the future <-ni> (FUT) exists alongside the imperious future <-naka ~
-ka> (IMF). Together they form an epistemic modal system in which the former
indicates less certainty about the occurrence of the event and the latter indicates more
certainty. The second argument in favour of considering the future and the imperious
future as modal categories comes from the Atong speakers themselves. One of my
Atong friends and consultants pointed out to me that with <-ni> (FUT) you are 50%
sure and with <-naka ~ -ka> (IMF) you are 95% sure that the event you are talking
about will indeed happen.
The morphemes under discussion do certainly not indicate prospective aspect. It is
not possible for these future modality morphemes to be used in situations such as
Yesterday, just as I was about to cook the soup. The next sections will treat in detail
the uses of the two future categories.
23.10.1 The imperious future suffix <-naka ~ -ka>
The imperious future modality is signalled by the morpheme <-naka ~ -ka> (IMF).
The allomorph <-ka> (IMF) is used when immediately following the negative
morpheme <-ca> (NEG), whereas the allomorph <-naka> is used in all other
environments. The imperious future in non-negative polarity expresses that an event
will almost certainly take place in the future or that an event is about to take place,
depending on the context. In negative polarity it expresses that an event will almost
certainly not take place or that an event will not go on any more from now on,
depending in the context.
The imperious future forms an epistemic pair with the future category (see next
section) and is also used as indicator of what will happen or happens usually in certain
circumstances. Example (600) in the next section illustrates the difference in degree of
certainty between the two future categories. The fact that the future cannot co-occur
with other modal categories, such as the irrealis, confirms its status as a modality. The
following examples illustrate some uses of the imperious future.


409
(592) ama, may sa'naka nido.
ama {may sa' -naka} [ni] =do
mother rice eat-IMF 1se =TOP
Mother, well eat rice now.

(593) ra nemcie ataknakasoy?
[ra] {nem} =ci =e {atak -naka} =soy
rain good =LOC =FC do.that -IFT =MIR
Now that the rain has stopped, what the heck shall we do?

(594) ca ronaymo, may saaymo, raynaka.
[ca] {ro} =ay =mo [may] {sa'} =ay =mo {ray' -naka}
tea drink =ADV =SEQ rice eat =ADV =SEQ go -IFT
After drinking tea and eating rice, we will go.

One of the most common occurrences of the imperious future is in the following
expression whose functional equivalent in English would come close to good bye. It
is impolite to leave somebody without saying (595).

(595) ray'naka.
{ray' -naka}
go -IFT
[I/We]ll go now. (And then the speaker or the speaker and his company
usually leave instantly after saying this.)

On adjectives of Type 1 the imperious future imparts the meaning of enough to the
predicate, i.e. som-naka (sweet-IFT) sweet enough, cu-ca-ka (big-NEG-IFT) not
big enough. The imperious future is not attested on Type 2 adjectives. Nominal
predicate heads cannot take modality suffixes.
Under negation the meaning of the imperious future does not change. When, for
example, someone asks you: pipuk sa'-ni=ma? (intestines eat-FUT=Q) Will you
eat intestines?, and you want to answer in the negative, you can say: sa'-ca-ka.
(eat-NEG-IFT) I will certainly now not eat [it].

410
23.10.2 The future suffix <-ni>
Future modality is signalled by the morpheme <-ni> (FUT), which indicates that an
event might take place in the future and that there is a rather big chance that it will
not. The future is also used to point out what event will occur or occurs usually in a
certain circumstance. The future modality forms an epistemic pair with the imperious
future treated in the previous section. A noticeable fact about the future modality is
that it cannot be negated. This means that the opposition between the two futures is
neutralised under negation as the imperious future can be negated.
The use of the future suffix <-ni> (FUT) as indicator of a possible future event is
illustrated below.

(596) o came, ami na'na kha'galgabaaw na'mi khathoci daetna
man'phanima?
[o came] [a] =mi [na'] =na {kha'gal} =gaba] =aw
interj sweetheart 1s =GEN 2s =DAT love =ATTR =ACC
[na' =mi kha'tho] =ci {da -et} =na {man' -pha -ni} =ma
2s =GEN heart =LOC enter-CAUS =DAT be.able -IN.ADDITION -FUT =Q
O sweetheart! will you be able to insert also into your heart me who loves
you?

(597) morot so'otgabaaw gobormen so'otsigani, sakhawgabaaw jurimana kamna
nani.
[[morot] {so'ot} =gaba] =aw [gobormen] {so'ot -siga -ni}
person kill =ATTR =ACC government kill -ALT -FUT
{sa' -khaw} =gaba] =aw [jurimana] {kam} =na {na -ni}
eat -secretively =ATTR =ACC penalty remove =DAT must -FUT
Those who kill people, the government will kill in turn, those who steal will
have to pay a fine.


411
(598) rodo ha'wayci mu'gaba mosora maharidoraba caksan macadu
hon'phani nooy khu'rasakokno.
[[rodo ha'way] =ci {mu'} =gaba} mosora mahari] =dora =ba
Pname =LOC live =ATTR Sname1 family =p =EMPH
[caksan] [macadu] {hon' -pha -ni} {no} =oy
bracelet tiger/man give -IN.ADDITION -FUT say =ADV
{khu'rasak -ok} =no
promise -COS =QUOT
Those very Mosera families who lived in Rode Ha?way promised to also
give a bracelet to the macadu (creatures who are human during the day but
tigers at night), it is said. Literally: Those very Mosera families who lived
in Rode Ha?way, to also give a bracelet to the macadu, [they] sayingly
promised, it is said.

The future can be used as indicator of what will happen or happens usually in certain
circumstances. It is logical that the epistemic (less certain) future category is used in
these cases because the speaker does not know, when creating a condition, if that
condition will be met or not as in (599). And in case of a certain circumstance in
which something will usually be done, the speaker does not know if this circumstance
will actually occur or not. Only in those cases that the speaker is certain that an event
must occur, will he use the imperious future, as in (600) c) and g).

(599) balcacido tokni.
{bal -ca} =ci =do {tok -ni}
tell -NEG =LOC =TOP hit -FUT
If [you] dont tell, Ill hit [you].

The next example comes from TEXT 3, lines 8 to 14, about incantation of spirits. The
storyteller explains the process.


412
(600) a) umido uaw kamal sandini.[]
b) otokoymu kamal khurutni. []
c) na'a way cugabaaw nukok noay cancibo, ma'su ra'naka, purun ra'naka,
taw' ra'na nani, wak ra'na nani, unado.
d) umi cowba som'na nani, ue kmalna.
e) otokoymu kamalna cow som'ay hon'aymu aro unaba may jabek ra'ana
nani, kamalnaba. []
f) otokoymu takaba kharayci coyni. []
g) kamal uan kamaldo doda hokhotanaka.
h) khurutgabami niamaw otokoy balnine.
a) umido [kamal] {sandi -ni}
then priest search -FUT
Then [they] will maybe search for a priest.
b) otokoymu [kamal] {khurut -ni}
so.then priest perform.incantation -FUT
So then the priest will maybe perform an incantation.
c) [na'a] [way {cu} =gaba] =aw} {nuk -ok} {no} =ay
2s spirit big =ATTR =ACC see -PF say =ADV
{canci} =bo [ma'su] {ra' -naka} [purun] {ra' -naka}
imagine =IMP cow get -IFT goat -get -IFT
[taw'] {ra'} =na {na -ni} [wak] {ra'} =na {na -ni}
chicken -get =DAT need -FUT pig get =DAT need -FUT
[u] =na =do
DST=DAT =TOP
[Sayingly] imagine that you see a big spirit, [you] will certainly get a cow,
[you] will certainly get a goat, maybe [you] will need to get a chicken, maybe
[you] will need to get a pig, for him.
d) umi [cow] =ba {som'} =na {na -ni} [ue kamal] =na
then liquor =ADD prepare=DAT need -FUT DST priest =DAT
Then [you] will need to prepare some liquor for that priest.
e) otokoymu [kamal] =na [cow] {som} =ay {ho'} =ay =mu aro
so.then priest =DAT liquor prepare=ADV give =ADV =SEQ and
[u] =na =ba [may] [jabek] {ra' -a} =na {na -ni}
DST=DAT =EMPH rice curry get -AWAY =DAT need -FUT
[kamal] =na =ba
priest =DAT =EMPH
So then, after having given the prepared liquor to the priest, [you] will maybe
also have to give rice and curry, to the priest that is.

413
f) otokoymu [taka] =ba [kharay] =ci {coy -ni}
so.then money =ADD big.pan =LOC offer -FUT
So then, [you] will offer money in a big pan.
g) [kamal] [u =an kamal] =do [doda] {hokhot -a -naka}
priest DST=FC/ID priest =TOP alone come.out -AWAY -IFT
The priest, that priest, will come out alone.
h) [{khurut} =gaba =mi niam] =aw [otokoy] {bal -ni} =ne
perform.incantation =ATTR =GEN rules =ACC like.that tell -FUT =TAG
I will now tell the rules of performing an incantation, ok.
23.11 The referential suffix <-an>
The referential marker <-an> (REF) occurs on negated predicate heads and indicates
that the S or O argument refers to a specific referent. The absence of this morpheme
indicates that the S or O is non-referential, i.e. does not refer to a specific referent. On
negated predicate heads before the morpheme <-ca> (NEG) is the only place the
referential morpheme can occur. The following examples illustrate the use of the
referential marker <-an>-(REF). Proper names, place names and personal pronouns
are always referential and therefore <-an> (REF) occurs in (601), (602) and (603). In
(602) we see the homophonous focus/identifier NP enclitic <=an> (FC) on the S
argument, i.e. the place name cokpot Chokpot.

(601) ranustaw nukama nukanca?
[ranus]
O
=taw {nuk -a} =ma {nuk -an -ca}
Name =ACC see -CUST =Q see -REF -NEG
Have [you] seen Ranus or not?

(602) cokpotan ga'suanca.
[cokpot]
S
=an {ga'su -an -ca}
Pname =FC/ID splendid -REF -NEG
Chokpot is not splendid.

(603) angdo nemkhalanca
[a]
S
=do {nem -khal -an -ca}
1s =TOP good -CP -REF -NEG
As for me, [I] am not better.


414
In the following example the main predicate indicates that the falling rocks, which are
in O function, are referential.

(604) rong'galetgaaw nukrumancak
[ro' {gal -et} =ga]
O
=aw {nuk -rum -an -ca -k}
rock fall -CAUS =ATTR =ACC see -all -REF -NEG -COS
[You] cannot see the rocks [I] all caused to fall any more. Alternatively:
The rocks that were all caused to fall down cannot be seen any more. (said
Samrat when he threw some stones into a vast and deep ravine.)

Even when the S or O are ellipsed the negative predicate can still take the referential
suffix to indicate that the implied S or O refers to a specific referent, e.g. (605). In this
example the money, taka money, is mentioned in the first sentence, but in the third
sentence it is ellipsed. Still the negative predicate in that sentence contains the
referential marker. In this case the referential marker has an anaphoric function.

(605) aya! taka do'tawcakthay. ado rocoygoksan raariwa. hoys! do'tawancak.
aya [taka] {do' -taw -an -ca -k}
interj.surprise money be.enough -UPWARDS -REF -NEG -COS
[a] =do [ro coygok] =sa =an {ra -ari -wa}
1s =TOP CLF:ROUND.THINGS ten =DLIM =FC/ID take -SIMP -FACT
hoys {do' -taw -an -ca -k}
interj.disapproval be.enough -UPWARDS -REF -NEG -COS
Huh?! I dont have enough money any more. I took only ten rupees. Damn! it
is not enough any more.

Apart from the plain negative, <-an> (REF) can only be expressed on the negative
change of state <-ca-k> (NEG-COS) and does not co-occur with any other predicate
head suffixes (see Table 63). The referential marker does not co-occur with the
negated incompletive <-khu-ca> (INCOM-NEG) nor with the negated simpicitive
<-ari-ca> (SIMP-NEG) nor in any negative subordinate clause.
I have one elicited example of a negated Type 2 adjective with a referential S but
without the referential marker (606). More fieldwork needs to be conducted to find
out if it is really impossible to have <-an> (REF) on a Type 2 adjective with
referential S argument.


415
(606) ie ram thomboloca
[ie ram]
S
{thombolo -ca}
PRX road to.have.holes.in.it -NEG
This road is not damaged.
23.12 The negative suffix <-ca>
The negative suffix <-ca> (NEG) indicates that something is not the case. As an
extended function it is also used to indicate that an event has not yet been realised, as
we shall see further below. The negative suffix is a predicate head marker and occurs
on nominal, both types of adjectival and verbal predicate heads. Demonstratives,
interrogatives and other possible non-verbal predicate heads (see Chapter 22) are not
attested under negation. Both main and dependent clause predicates can be negated.
The verb ni to not exist cannot be negated. There are no recorded examples of
negated purpose clauses (see 27.2.3). Furthermore, customary aspect and future
modality cannot be expressed under negation:
In Atong the grammatical system relating to the predicate is dependent on
polarity. The fact that fewer categories relating to the predicate can be expressed
under negation than in positive clauses is in accordance with observations made in
Aikhenvald and Dixon (1998). It is relevant to quote a short passage from page 63 of
that article as it describes exactly what happens in Atong.

The type of dependency is simple. Since positive is always the unmarked
term, another type of grammatical system, if it depends on polarity, will have
more choices available in the positive than in the negative.

The position of the negative morpheme is variable as is mentioned below in this
section. However, in the overwhelming majority of the recorded occurrences its
position relative to the other morphemes is as represented in Table 63. The scope of
the negation is always the predicate.
The following example illustrates the use of the negative suffix on a non-
referential nominal predicate. The conversation is not about a particular house but any
house built by a Khasi.


416
(607) Speaker A: noksamci simen, tota, tin pirinay hama.
Speaker B: nokhuci?
Speaker A: nokhuca! noksamci.
[nok sam] =ci [simen tota tin]
house side =LOC cement plank corrugated iron
{pirin} =ay {ham} -a
mix =ADV build -CUST
[nokhu] =ci
roof =LOC
{nokhu} -ca [nok sam] =ci
roof -NEG house side =LOC
A: For the wall of a house (lit. at the side of a house) [the Khasis] mix
cement, planks and metal plates.
B: For the roof?
A: Not the roof! For [lit. at] the wall of the house.

The following example illustrates the occurrence of a negated verbal predicate.
Example (606) shows a negated Type 2 adjective.

(608) aa ketketa bura nogabaawan tokhuca.
[aa] [ketketba bura {no} =gaba] =aw =an {to -khu -ca}
1s Name say =ATTR =ACC =FC/ID know -INCOM -NEG
I dont yet know this so-called Ketketa Bura.

There are very few recorded instances of stem-forming suffixes that come after the
Echelon 1 suffix <-ca> (NEG) viz. (609) and (610). In those examples the stem-
forming suffixes are event specifiers. This adds to the evidence that the position of
suffixes is to a great degree variable depending on the sope the speaker wants the
suffix to have (see also Chapter 22).


417
(609) otokoymu agalaw ram ray'aymu thoycaphingadorado agal rugumi
wayphinano.
otokoymu [agal] =aw ram {ray'} =ay =mu
so.then forest.fire =ACC road go =ADV =SEQ
{thoy -ca -phin} =ga] =dora =do] [agal rugu] =mi
die -NEG -FULLY =ATTR =p =TOP forest.fire edge =GEN
{way -phin -a} =no
return -RETURN -CUST =QUOT
So then, having gone the road of the forest fire, those who are not dead
altogether return [via] the edge of the forest fire.

(610) te'ewrawraw morot nemcabatsoragaba.
te'ew -rawraw [morot] {nem -ca -bat -sra} =gaba
now -CONTINUOUSLY person good -NEG -VERY -COMPLETELY =ATTR
Still now people are extremely bad.

As was said above, another function of the negative morpheme is to indicate that
an event has not yet taken place. Examples of the negation of a not yet realised event
are given above in (248) in 13.4 and below in (611). The latter example comes from
a story in which a cunning man called Thengthon [the?ton] has just impoverished
his village people by getting them to burn their houses with the promise that they
would be able to sell the ashes and cinders at the market and get rich. However, after
the villagers have discovered that nobody wants to buy their ashes, they become very
angry with Thengthon and utter (611). The fact that the event is bound to take place
is marked by the imperious future suffix <-ka> (IFT).

(611) ramci hampoy na'nado watcaka ge'theawdo.
[ram] =ci [hampoy] [na'na] =do {wat -ca -ka}
road =LOC this.evening 1pi =TOP get.rid.of -NEG -IFT
[ge'the] =aw =do
r3s =ACC =TOP
This evening we will certainly get rid of him on the road.
23.13 The change of state suffix <-ok ~ -ak ~ -k>
The change of state aspect is signalled by the morpheme <-ok ~ -ak ~ -k> (COS). The
allomorph <-ak> (COS) occurs after a root or stem ending in /a/ or /a?/. This means

418
that for roots and stems ending in /a/, there are three possible pronunciations,which all
occur in the recorded material. First possibility: the two vowels can collapse into one.
Second, a glottal stop can occur on the morpheme boundary. Third: the two vowels
are both pronounced, the result of which sounds like one long vewel. The allomorph
<-k> (COS) occurs always after the negative morpheme <-ca> (NEG). There is not
one recorded instance of a form *<-ca-ak> (NEG-COS) with one of the
pronunciations [tca:k] or [tca?ak] which are phonetically possible. For an explanation
of what happens when two of the same vowels are juxtaposed across morpheme
boundaries, see 2.10.
This suffix is attested on verbs, including stative verbs and Type 1 adjectives (i.e.
stative verbs expressing a quality), nouns and interrogatives functioning as predicate
head; it is not attested on Type 2 adjectives. An alternative might be to label this
suffix as perfective aspect, however, the fact that it occurs on Type 1 adjectives and
nouns, makes the label change of state more suitable, even though the semantics of
this suffix are compatible with activity verbs.
The semantic interpretations of the change of state suffix are different according
to the type of predicate head it suffixes to. These interpretations will be treated one by
one below. First we will see the effects of this suffix on verbal predicates, excluding
Type 1 adjectives, then on Type 1 adjectives, then on nouns and on other types of
predicate head. Examples of this suffix on interrogatives can be found in Chapter 9.
Negated predicates are treated separately after that.
23.13.1 On verbal predicates
When it occurs on verbal predicate heads the change of state suffix indicates that a
change has taken place in the situation that supposedly held for the S or A argument
before the event denoted by the predicate took place. In text example (612) we see the
change of state contrast with the factitive. The contrasitve factitive- and change of
state-marked predicates are underlined. In line 2 the speaker indicates that the fish
trap has been set up, where before it had not yet been set up, hence the change of state
suffix on the predicate sa-ak (put.as.trap-COS). This change of state contrasts with the
already existing situation in line 3, where nothing changes and thus the factitive suffix
is used on the predicatte sa-wa (put.as.trap-FACT). In line 4 the speaker puts forth the
fact that there are no fish in the river, and therefore marks the predicate with the

419
factitive suffix, viz. ni'-wa (not.exist-FACT). This statement contrasts with that in
line 8 where the speaker indicates a change of state in the existence of fish,
presupposing that there had been fish before the otter ate them. hence the change of
state suffix on the predicate ni'-ok (not.exist-COS).

(612) konsado manapmi sirimonmon re'eaymona dabat warisa, digaray
saakno. sangumuk dingaray sawado, ni'wanoro, na'ba. otokoymuna
kambaysami digarayaw na' cayokno. uciba matdam sa'akno otokoymuna
kambaysami digaraymi cayciba, uciba na'an ni'okno. uciba matdam
sa'akno uawba.
a. [konsa] =do [manapmi] [sirimonon] {re'e} =ay =mona
after =TOP very.early.in.the.morning at.the.crack.of.dawn go.away =ADV =SEQ
b. [dabat wari] =sa [digaray] {sa -ak} =no
Pname deep.place.in.river =MOB fish.trap put.as.trap -COS =QUOT
c. [san] =gumuk [digaray] {sa -wa} =do
day =whole fish.trap put.as.trap -FACT =TOP
d. {ni' -wa} =no =ro [na'] =ba
not.exist -FACT =QUOT =EMPH fish =EMPH
e. otokoymuna [ca'ma] =sa =mi digaray] =aw [na'] {cay -ok} =no
so.then downstream =MOB =GEN fish.trap =ACC fish look -COS =QUOT
f. [u] =ci =ba [matdam] {sa' -ak} =no
DST=LOC =EMPH otter eat -COS =QUOT
g. otokoymuna [kambay =sa =mi digaray =mi] {cay} =ci =ba
so.then upstream =MOB =GEN fish.trap =GEN look =LOC =EMPH
h. [u] =ci =ba [na'] =an {ni' -ok} =no
DST=LOC =EMPH fish =FC/ID not.exist -COS =QUOT
i. [u] =ci =ba [matdam] {sa' -ak} =no [u] =aw =ba]
DST=LOC =EMPH otter eat -COS -QOUT DST=ACC =EMPH
Later, in the morning, at the crack of dawn, having gone to Dabatwari, [he]
set fish traps, it is said. The whole day he set fish traps, it is said, [but] there
was none, fish that is. So then, [he] looked for fish [in] the fish trap(s) from
downstream, it is said. There an otter had eaten [them], it is said. So then he
indeed looked at [those] from the upstream funnel(s). There (also) there was
no more fish, it is said. There (also) an otter had eaten them, it is said, those
[fish].

A series of at least two, usually three or four, repeated predicate heads marked by the
change of state aspect indicate an event of long duration (613). As an important side
effect, the speaker can repeat the predicate head, marked for all categories, as many
times as he wants and in the meantime think about what he will say next.

420
(613) jalaok ue banthay mo'sado. te'do sagal toysamci poreok poreok porepok
poreok. konsado cugaba kam man'ok.
{jal -a -ok} [ue banthay mo' sa] =do [te'] =do
run.away -away -COS DST bachelor CLF.HUMANS one =TOP now =TOP
[sagal toysam] =ci {pore -ok} {pore -ok} {pore -ok} {pore -ok}
sea water.side =LOC study -COS study -COS study -COS study -COS
[He] run away that young lad. Now, at the seaside he studied [and] studied
[and] studied [and] studied. [After that he got a great job.]

The change of state suffix is also attested on predicates of subordinate clauses, viz.
dative-marked reason clauses (see Chapter 27) and on topic-marked clauses, e.g.
(614), where it refers to a possible future event.

(614) hap pidan rama man'okodo jotnaka?
|[hap pidan] {ram} =na {man' -ok}| =odo {jot -naka}
place new search =DAT obtain -COS =TOP move -IFT
When/if [you] have found a new place youll immediately move?

23.13.2 On Type 1 adjectival predicates
On Type 1 adjectival predicate heads the change of state aspect can have two possible
interpretations according to the context. The first possible interpretation is a
reinforcement of the property denoted by the adjective, as we can see in the
translation one of my consultants made of the Atong sentences in (615). The second
possible interpretation is change of state, as is illustrated in (67) and (771).

(615) ama! ah! tanido jabek thawokte. na'a panando thaway romca. atotokoy
taynido thawoksoy ja'bek?
[ama] ah [ta'ni] =do [ja'bek] {thaw -ok} =te
mother Interj today =TOP curry tasty -COS =DCL
[na'a] [panan] =do {thaw} =ay {rom -ca}
2s always =TOP tasty =ADV cook -NEG
[ato] =tokoy [tay'ni] =do {thaw -ok} =soy ja'bek
what -through today =TOP tasty -COS =MIR curry
Mother! Oh! today the curry is very tasty! You dont cook tasty all the time.
Why is [it] today to my surprise so very tasty, the curry?

421
23.13.3 On nominal predicate heads
Nouns can only function as predicate heads in identity/equation clauses. On nominal
predicate heads the change of state effect of the suffix <-ok ~ -ak ~ -k> (COS) is very
clear, as we can see in (683) repeated here as (616).

(616) ido the'thonte. me'maokma jowmaok?
[i] =do {the'thon} =e {me'ma -ok} =ma {jowma -ok}
PRX =TOP Name =FC ghost -COS =Q dream -COS
This is Thengthon, Im telling you. Has [he] become a ghost or a dream?
23.13.4 On other types of predicates
On numeral plus classifier constructions the change of state aspect implies a totality,
as is illustrated in the answer of example (617).

(617) na'ci roboysok te'ewe? roboroyok.
[na'] =ci [ro boisok] [te'ew] =e {ro boroy -ok}
2s =LOC CLF.MONEY how.many now =FC CLF.MONEY four -COS
How much money do you have now? Four rupees in total.

There is one recorded instance of an adverb functioning as predicate head marked by
the change of state. This is represented in the following example.

(618) konsado amakdo dodaanok.
[konsa] =do [amak] =do {[doda] =an -ok}
after =TOP monkey =TOP alone =FC/ID -COS
After [that], the monkey was alone.
23.13.5 On negated predicates
When occurring on negated predicates, the change of state suffix indicates that
something is not the case any more, e.g. (619) with a nominal predicate and with a
verbal predicate in (620). Interrogatives functioning as predicate head cannot be
negated.


422
(619) ge'the mastelancak
[ge'the] {mastel -an -ca -k}
3s male.teacher -REF -NEG -COS
He is not a teacher any more.

(620) te'ewdo matsa caw'koy asetram mongancak. ue ha'boriawe seng'sotay matsa
caw'koy mongsigaariok
[te'ew] =do [matsa caw'koy aset ram] {mo -an -ca -k}
now =TOP tiger big.knife throw.away place call.a.name -REF -NEG -COS
[ue ha'bori] =aw =e {se'sot} =ay
DST hill =ACC =FC abbreviate =ADV
[matsa caw'koy] {mo -siga -ari -ok}
Pname call.a.name -ALT -SIMP -COS
Now [we] dont call [it] Matsa Chawkyi Asetram
46
any more. This hill has
come to be abbreviatedly called Matsa Chawkyi
23.14 The progressive/durative aspect suffix
The interpretation of the progressive/durative suffix depends on the semantics of the
predicate.The progressive/durative aspect indicates that an event is ongoing
(progressive interpretation) or that a state is continuing (durative interpretation). All
its allomorphs, viz. <-aydoa ~ -aydo ~ -aydok ~ -aroa ~ -aro ~ -arok>
(PROG/DUR) are in free variation. When a root ends in /-i/ the sequence /ay/ and the
vowel /a/ of the progressive/durative morpheme assimilate to /e/ creating the
allomorphs <-edoa ~ -edo ~ -edok ~ -eroa ~ -ero ~ -erok> (PROG/DUR). The
progressive/durative aspect suffix is only attested on verbal (including Type 1
adjectival) predicates. Example (621) illustrates the use of this suffix on a Type 1
adjective, in which case we get a durative interpretation, because Type 1 adjectives
are stative verbs that denote a quality (see Chapter 5). The example is taken from a
story about the fox and the deer. The fox has just found the deer after a wild chase.

46
In this name the word matsa is short for matsadu. A matsadu is a creature which is human during
the day and changes into a tiger at night. The name of the place in the example is related to an event in
the history of the Atong and translates roughly as the place where the matsadu threw away their
weapons.

423
The deer says it is guarding the kings royal fan. The fox says that he is hot and wants
to use the fan. The deer says that the king does not give it to anyone, not even to him.

(621) aan tuaydoa, anaan hon'anca raja.
[a] =an {tu} -aydoa [a] =na =an {hon' -ca} [raja]
1s =FC/ID hot -DUR 1s =DAT =FC/ID give -NEG king
Im hot, the king doesnt even give [it] to me.

A series of at least two, usually three or four, repeated predicate heads marked by the
progressive/durative aspect, and often also the quotative clausal enclitic, indicate an
event of long duration, just as the same construction with the change of state suffix
described in the previous section. As an important side effect, the speaker can repeat
the predicate head, marked for all categories, as many times as he wants and in the
meantime think about what he will say next. The folowing example is illustrative. In
this example we encounter the progressive interpretation of the progressive/durative
suffix, since the verb re'e- to go away is an activity verb.

(622) nia kawarini noaymu re'earokno re'earokno re'earokno re'earokno.
[nia] {kaw -ari -ni} {no} =ay =mu
1pe shoot -SIMP -FUT say =ADV =SEQ
{re'e -arok} =no {re'e -arok} =no {re'e -arok} =no
go.away -PROG =QUOT go.away -PROG =QUOT go.away -PROG =QUOT
{re'e -arok} =no
go.away -PROG =QUOT
Having said: We will just shoot [it], they are going, it is said, [and] going, it
is said, [and] going, it is said, [and] going, it is said.

One example has been recorded where the suffix under discussion occurs on a Type 1
adjective (a stative verb denoting a quality, in this case cu to be big) but has to be
interpreted as the progressive and not the durative, viz. TEXT 1, line 22. This
example is presented below as (623). Because of the event specifier suffix <-a>
(WITHOUT.HOLDING.BACK), the predicate cu-a (be.big-WITHOUT.
HOLDING.BACK) get really big is dynamic. This dynamic predicate takes the
suffix <-arok> (PROG/DUR) which now has to be interpreted as the progressive, i.e.

424
the action of getting big is ongoing. This is why my Atong consultant insisted on
translating the predicate in the example as are getting really big.

(623) muthayba cuaarok, te'ewe.
[mu'thay] =ba {cu -a -arok} [te'ew] =e.
bosom =EMPH big -WITHOUT.HOLDING.BACK -PROG now =FC
[Her] breasts are getting really big, though, now.

The progressive/durative aspect is also attested on predicates of subordinated locative-
marked clauses, of which example (121) is an illustration.

425

426
Chapter 24 The factitive suffix
_____________________________________________________________________

The factitive suffix <-wa> (FACT)
47
occurs in five different syntactic environments,
viz.

1 on main clause predicates, treated in 24.1,
2 on predicates of subordinate clauses which are governed by a matrix clause
predicate, i.e. complement clauses in S or O function, discussed in 24.3,
3 on subordinate clause predicates in adjunct function, modifying a matrix
clause, see 24.4,
4 on lexicalised object nominalisations, also treated in.24.4,
5 on complement clauses of the limitative postposition dabat (LIMIT), see
24.5.

The function of the factitive on main clause predicates is reification, i.e. presenting
the event denoted by the verb as a fact. It does not appear to have this function in all
types of subordinate clauses, but it is very clear that the suffix does have this function
on Temporal Location adjuncts. If a noun functions as predicate of a Reason clause, it
has to be factitive-marked to appear in this function. As we shall see below, this is not
in conflict with the function this suffix has on verbal predicates. A summary of the
functions of the factitive suffix is given in 24.6 followed by a note on a possible
diachronic development of the suffix in 24.7.
24.1 Factitive-marked main clause predicates
The factitive suffix <-wa> (FACT) on main clause predicates is a marker of
reification, i.e. it presents the event denoted by the verb as a fact, as something that is
the case, which is an epistemic modality function. I adopt Lyonss definition of
epistemic modality, which is:

47
I have taken the term factitive from Roland Rutgers (1998: 231).
24 THE FACTITIVE SUFFIX

427
Any utterance in which the speaker explicitly qualifies his
commitment to the truth of the proposition expressed by the sentence
he utters, whether this qualification is made explicit in the verbal
component [] or in the prosodic or paralinguistic component, is an
epistemically modal or modalized, utterance. (1977: 797)

The factitive modality is one of the possibilities an Atong speaker has to modify the
information denoted by the predicate. The other modality suffixes, treated in Chapter
23, are presented in Table 66 together with clausal enclitics indicating modality (see
26.826.9). The factitive can occur in clauses which are marked by the speculative or
the irrealis clausal enclitics.

Table 66 Suffixes and clausal enclitics indicating a modality
TYPE OF MODALITY MORPHEME LABEL
factitive -wa FACT
imperious future (more certain) -naka ~ -ka IFT
future (less certain) -ni FUT
irrealis =com IRR
speculative =khon SPEC

The factitive suffix is not attested on Type 2 adjectives and nominal predicates of
main clauses, but does occur on nominal predicates of co-subordinate clauses, as we
will see below.
Polarity has an influence on the interpretation of the time reference of a factitive-
marked predicate, except when the predicate is a Type 1 adjective (see 24.2). When
the factitive suffix occurs on a negated predicate the event can be interpreted as
having future or present time reference but never as having past time reference. On a
non-negated predicate the factitive suffix usually induces past time reference
interpretation of the verb, but can also have a present time reference interpretation
depending on the verb and the context. Note that the factitive suffix is by no means a
tense marker. Let us look at negated verbs first.
The following two examples present a contrasting pair of negated main clause
predicates. The main predicate man'-ca (be.able-NEG) cannot in (624) is not
factitive-marked and has an habitual interpretation, whereas the main predicate in
(625), man'-ca-wa (be.able-NEG-FACT) will not be able is factitive marked and is
interpreted as having future time reference. The context of example (624) is as

428
follows. The fox has chased after the deer and finds it lying next to a well. The deer
says it is guarding the kings well. The fox is thirsty and wants to drink water from the
well but the deer says that the king does not allow anyone to drink water from the well
and then says (624).

(624) aan rona man'ca iaw
[a] =an {ro} =na {man' -ca} [i] =aw
1s =FC/ID drink =DAT be.able -NEG PRX =ACC
Even I cannot drink this [water].

Example (625) comes from a story in which a group of brothers go into the jungle to
hunt a giant eagle. They meet an old man who says that they are only fit to shoot the
eagle if they can smoke his tobacco. But the tobacco is so strong that none of the
brothers can smoke it. One of the brothers exclaims (625).

(625) nido kawna man'cawa udo
[ni] =do {kaw} =na {man' -ca -wa} [u] =do
1pe =TOP shoot =DAT be.able -NEG -FACT DST=TOP
We will not be able to shoot that [eagle].

When a non-negated main verb is marked by the factitive it can be interpreted as
an event that happened in the past. This is not surprising if one were to state that a
reified event is something that must already have occurred in order to be a fact. When
people meet on the road, they usually ask where you have come from and do this with
a factitive-marked predicate. The reply will also be factitive-marked. Example (626)
is illustrative of such a conversation on the road.

(626) bisa re'ewa na'a? turasa re'ewa.
[bi] =sa {re'e -wa} [na'a]
QF =MOB go.away -FACT 2s
[tura] =sa {re'e -wa}
Pname =MOB go.away -FACT
Literally: From where have [you] left, oh you?! [I] have left from Tura.
Alternatively: Where do you come from? I come from Tura.

429
Not all occurrences of non-negated factitive-marked main clause predicates have to be
interpreted as having past time reference. The usual way to say what you are called in
Atong is with the predicate marked by the factitive suffix, as we see in (627). The
verb in the predicate, mo- to call someone/something a name, is factitive-marked
and has no past time interpretation. The predicate just states a fact.

(627) ami bimu Samrat mowa.
[a =mi bimu] [Samrat] {mo -wa}
1s =GEN name Samrat call.a.name -FACT
[One] calls my name Samrat. Alternatively: My name is called Samrat.

In the next example a factitive-marked predicate is contrasted with a customary
aspect-marked predicate. The predicate in (628) refers to an event which may take
place in general and its object is non-referential, whereas in (629) we see that the
factitive makes the situation more concrete, that the event is interpreted as having past
time reference, and that, in this case, the object can be interpreted as referential and
can hence take the accusative suffix <=aw> (ACC).

(628) may sa'ama?
[may] {sa' -a} =ma
rice eat -CUST =Q
Do [you] eat rice (in general)?

(629) may sa'wama?
[may] =aw {sa' -wa} =ma
rice =ACC eat -FACT =Q
Did [you] eat the rice (which was provided for you)?

Example (635) below comes from a story about a fox and a deer. This sentence can be
translated into English with a past time interpretation, using the past tense, or with a
present time interpretation, using the present tense. In Atong there is no tense, so that
nothing makes you have to choose between one or the other temporal interpretation. A
story, like the one about the deer and the fox, which does not explicitly refer to a
certain time by using time words or other lexical means, can always be translated in

430
English using either the past tense or present tense. My consultants use past and
present tense in English indiscriminately when translating their language in cases
when there is no explicit temporal reference. This can indicate that they dont know,
or are not aware of the difference between the past and present tense in English, or
that they do not care, since the story makes sense no matter how you translate it.
The factitive also occurs in combination with the irrealis clausal enclitic <=com>
(IRR) or the speculative modality enclitic <=khon> (SPEC). The irrealis and
speculative enclitics occur exclusively on main clause predicates. The irrealis also co-
occurs with the customary aspect suffix <-a> (CUST). The customary aspect and
factitive suffixes are mutually exclusive. The temporal reference interpretation of
factitive and irrealis- and factitive and speculative-marked predicates depends on
polarity for negated verbs and on the meaning of the verb and the context for non-
negated verbs, as described above. In example (630) the interpretation of the factitive
and irrealis-marked predicate is of past time reference and in (631) of present time
reference, but there is nothing in the verbal forms that indicates this reading.

(630) raw'na bakwacom otokciba man'anca
|{raw'} =na| {bak -wa} =com otokciba {man' -an -ca}
catch =DAT try -FACT =IRR but be.able -REF -NEG
He attempted to catch it in vain, but he could not.

(631) ana daygaba ni'wacom.
[[a] =na {day} =gaba] {ni' -wa} =com
1s =DAT be.bigger=ATTR not.exist -FACT=IRR
There is supposedly no one greater than me. Alternatively: There cannot be
anyone greater than me (i.e. it is unimaginable).

The only occurrence of a factitive and speculative-marked predicate is presented in
example (632).

(632) na'tomdo nukcawakhonay.
[na' -tom] =do {nuk -ca -wa} =khon =ay
2pe -ppp =TOP see -NEG -FACT =SPEC =POS
You
p
might not see (the eagle) at all!

431
A factitive-marked main clause predicate can take the topic enclitic <=do> (TOP)
and the focus/identifier enclitic <=an> (FC/ID) as is illustrated with the next
examples. In (633) we see the intransitive verb koro to (make a) sound with the
factitive suffix and the topic enclitic. The context from which this example is taken is
as follows. The fox wants to play on the kings drum but the deer, who is guarding it,
says that it is not possible. The kings drum makes a very special sound, says the deer.

(633) ha'no' tara cinina imo korowado rajami dama noaydoano.
[ha' no' tara ci ni] =na [i] =mo {koro -wa} =do
earth inside layer TEN two =DAT PRX =ABL sound -FACT =TOP
[raja =mi dama] {no -aydoa} =no
king =GEN drum say -PROG =QUOT
The kings drum sounds from here to the twelve layers of the earths inside,
[the deer] is saying, it is said.

In example (634) the first clause indicates the topical event, of which has been spoken
in the two preceding clauses as well, and the second clause a comment which is also
the conclusion of the paragraph. The two co-ordinated main clauses are in a
contrastive relationship, due to the semantics of both clauses.

(634) sangumuk digarayi sawado, ni'wanoro na'ba
[sani] =gumuk [digaray] {sa -wa} =do
day =whole fish.trap put.as.trap -FACT =TOP
{ni' -wa} =no =ro [na'] =ba
not.exist -FACT =QUOT =EMPH fish =EMPH
[He] put up fish traps the whole day, [but] there was no fish.

The following example comes from the story about the deer and the fox. The fox and
the deer are friends. The deer has stolen a pack of biscuits from a Bengali. The two
friends went to a beautiful place on the riverside to take a bath and eat the biscuits
together. But the deer deceives the fox by eating all the biscuits alone while the fox is
bathing. The example contains a complex predicate consisting of a repetition of the
same verb (see 22.6). The factitive-marked main verb is focused by the
focus/identifier enclitic <=an> (FC/ID).

432
(635) pheru ropwa ropwaci magacak sa'waan.
[pheru]{rop -wa rop -wa} =ci [magacak] {sa' -wa} =an
fox bathe -FACT bathe -FACT =LOC deer eat -FACT =FC/ID
When the fox was/is bathing and bathing, the deer ate/eats.
24.2 The factitive on Type 1 adjectives
Type 1 adjectives are a subclass of verb and indicate qualities. The main characteristic
that sets Type 1 adjectives apart from other verbs is the reinforcing interpretation that
the change of state suffix <-ok ~ -ak> (COS)

can have on the meaning denoted by
these adjectives when they are used predicatively.
48

The factitive suffix is not widely attested on Type 1 adjectives. In the instances
that this suffix is attested, it always occurs in contexts where one cannot have any
other interpretation than that it denotes emphatic reinforcement of the quality denoted
by the Type 1 adjective. One such instance is presented in example (636) from a story
about a child and a giant eagle. A very small prodigious child, actually a newly born
baby, is walking though the jungle in search of his brothers when he meets an old
woman. He talks to the old woman to get her attention. So the old woman tries to look
for a person but sees nobody. This is because the child is very small, as is stated by
the narrator in (636). The factitive suffix is stressed by a pronunciation in falsetto
voice and a very long vowel /a/ as can be seen in the IPA transcription of the
predicate. This intensified pronunciation of the factitive suffix intensifies its emphatic
reinforcement function on the Type 1 adjective.

(636) sa'goray molwa [molwa] hagolsakno isa ganano.
[sa'goray] {mol -wa} [hagolsak] =no [i] =sa {gana} =no
child small -FACT interj =QUOT PRX =LOC exist =QUOT
A very small child indeed, my goodness, it is said, is here, it is said.

48
It should be noted that the change of state interpretation of the suffix <-ok ~ -ak> (COS) is also
possible on Type 1 adjectives, depending on the context.

433
In the next example the predicative Type 1 adjective is functioning as a main
clause predicate. This example comes from a song about a beautiful girl. The singer,
Wilseng S Marak, wants to emphasise how beautiful she is and uses the factitive on
the predicative Type 1 adjective and in addition to that he adds the emphatic positive
clausal enclitic <=ay> (POS) for extra emphasis.

(637) [] na' bima solwa'ay.
[na' bima] {sol -wa} =ay.
2s appearance/body beautiful -FACT =POS
your appearance/body is very beautiful indeed
24.3 Factitive-marked complement clauses
This section treats factitive-marked complement clauses. One could argue that the
factitive morpheme has the function of clausal nominaliser on subordinate clauses,
whereas it is a modality marker on main clauses. However, as we will see in section
24.4.3, these two functions are not very clearly separable. Moreover, factitive-marked
clauses have almost no nominal properties, except the possibility to be case-marked,
although the number of cases attested on factitive-marked clauses is very limited, as
we shall see below.
Factitive-marked complement clauses only occur in S and O function, just like
dative-marked complement clauses treated in Chapter 27. Complement clauses in S
and O function are treated in separate sections. The first section also treats some of
the characteristics of factitive lexicalisations.
24.3.1 Factitive-marked object complement clauses and nominalisation
The phasal verbs macot- to complete/finish an activity, jam- to complete, finish ,
which are Primary-B verbs, and the Secondary verb da'- to enter (into a state)
49

49
The verb da'- to enter is intransitive. The thing that one enters into is marked with the locative
<=ci> (LOC) or mobilitative <=sa> (MOB) e.g. toy no'=sa da'-a-ok=no (water inside=MOB
enter=AWAY-COS=QUOT) [he] entered into to water, it is said and u=ci-an so=ci da'-ok

434
take factitive-marked complement clauses as O argument as we can see in, (638),
(639) and (640)respectively.

(638) may sa'wa jamkhuca.
[[may] {sa' -wa}]
O
{jam -khu -ca}
rice eat -FACT finish -INCOM -NEG
[I] have not finished eating rice yet.

(639) ucian aa na'aw nukjorowacian na'na kha'galwa da'ok.
[ucian] [na']
O
=aw {nuk -joro -wa} =ci =an
then 2s =ACC see -DAILY -FACT =LOC =FC/ID
[[na'] =na {kha'gal -wa}]
O
{da' -ok}
2s =DAT to.love -FACT enter -COS
Then, when [I] saw you every day, [I] started loving you. Alternatively: [I]
entered into the state of loving you.

No factitive-marked object complement clause has been recorded which has a
different implied subject (S/A) from that of the matrix clause. It has to be investigated
through future fieldwork whether or not there is a pivot constraint that prohibits this
clause type from having a different subject from the matrix clause, like dative-marked
complement clauses (see 27.2.1) and purpose clauses (see 27.2.3).
In example (640) we see a factitive-marked complement clause functioning as O
argument of the matrix verb macot- to finish. Even though no overt A arguments are
expressed in either the complement or the matrix clause, the implied arguments in
both clauses have to be co-referential. In this example the implied argument of the
subordinate verb, cal- to sow by making a hole in the ground with a stick and putting
a seed into it, and the verb of the matrix clause, macot- to finish, is abo corn.
The matrix clause with the predicate macot- to finish is subordinate to the main

(DST=LOC=FC/ID village=LOC enter-COS) then [they] entered the village. This verb takes factitive
complement clauses only when the meaning is extended to enter into a state.

435
clause with the predicate kay- to plant, i.e. not embedded, not governed, but
functioning as a modifier to the main clause (see 28.2).

(640) calmanwa macotwamusa may kay'cea.
|[cal -man -wa]
O
{macot -wa}| =mu =sa|
sow -ALREADY -FACT finish -FACT =SEQ =DLIM
[may] {kay' -ce -a}
rice plant -FIRST -CUST
Only after sowing is finished is the rice planted first.

Factitive object complement clauses with arguments are not attested with accusative
marking on the clause, which is a good indication that the clause is not nominalised,
because the accusative is a phrasal and not a clausal enclitic. Accusative marking is
attested only when the factitive-marked derivations are lexicalised and therefore have
more nominal properties. The following example, in which we find the lexicalised
object nominalisation sa'-wa (eat-FACT) food, is illustrative.

(641) otokoymuna pankambayci powaymuna otokoy sa'khucano sa'waawdo.
otokoymuna [pan kambay] =ci {pow =ay =muna [otokoy]
so.then tree top =LOC fly =ADV =SEQ like.this
{sa' -khu -ca} =no [sa' -wa] =aw =do
eat -INCOM -NEG =QUOT eat -FACT =ACC =TOP
So then, having flown up in the treetop like this, [the crow] had not yet eaten
[it], it is said, the food.

As was mentioned in 24.3, the factitive suffix <-wa> (FACT) sometimes competes
with the construction in which a clause of which the predicate is carrying the factitive
suffix is followed by the genitive/nominaliser enclitic <=mi ~ =mo> (GEN/NR) to
derive lexical nominalisations from verbs, i.e. nouns denoting physical objects. In the
same text from which example (641) is taken, where the word meaning food is
derived from the verb sa'- to eat by means of the factitive suffix <-wa> (FACT), we
also find the meaning food derived from a clause with the same verb by means of
the factitive-plus-nominaliser-enclitic construction <-wa=mi> (FACT=NR). The latter
derivation is given in (642).

436
[pheru] {nuk -sega -ak} =no [|sa' -wa| =mi =aw
The fox saw [it] this time, the food.

Since <=mi ~ =mo> (NR) is also used to derive action/state nominalisations from
verbs, the word sa'-wa=mi (eat-FACT=NR) can also refer to the act of eating, as is
illustrated in the next example. In that example the storyteller describes what the main
character of the story sees when he looks at his potato garden in the morning after a
horse has eaten the plants.
An alternative analysis of this example is to treat sa'-wa (eat-FACT) as a
factitive-marked lexicalisation and the morpheme <=mo> to be the genitive, since
the postposition gomon reason requires the preceding NP to take the genitive case
marker (see 13.3). Cases where clauses are nominalised by the enclitic <=mi ~
=mo> (NR) are followed by the genitive, so as to have two of the same morphemes
in a row, has not been recorded.

(643) una aludaraaw rodomabutu sokaw sa'wamogomon te'ew manap
caywacido gumukan cokrumokno.
una [alu] =dara =aw [[{rodom -a -butu} [sok] =aw
then potato =p =ACC sprout -TOWARDS -WHILE sprout =ACC
[sa' -wa =mo gomon]
eat -FACT =NR/GEN reason
|[te'ew] [manap] {cay -wa} =ci =do
now morning look -FACT =LOC =TOP
[gumukan] {cok -rum -ok} =no
all rip.off -ALL -COS =QUOT
Then, the potatoes, because of the eating of the sprouts while [they] where
sprouting, when [he] now looked in the morning, were all torn off, it is said.
Alternative free translation: When he looked in the morning at the potatoes,
he saw that they were all ripped off because the young leaves had been eaten
while they were sprouting, it is said.

The derivations sa'wami ~ sa'wamo and sa'wa meaning food are used to refer to
food in a very general way. More fieldwork is needed to find out which nominal
properties factitive derivations have. The fact that factitive-marked verbs can still be

437
modified by adverbial clauses, as we can see in example (648), means that they are
not fully nominal, since NPs cannot be modified by adverbial clauses. The derivations
with the clausal enclitic <=mi ~ =mo> (NR) do have all nominal properties. It seems
that in the case of examples (641) and (642) both derivations have exactly the same
meaning even though they are derived with different morphemes. In every-day speech
the different derivations sa'wami ~ sa'wamo and sa'wa are used side by side, in
what appears to be the same contexts, to denote food.
Lexicalised factitive nominalisations occur with one other phrasal enclitic, viz.
<=gumuk> all, whole. Example (644) illustrates this phenomenon. From the
material gathered in the corpus, it seems that factitive lexicalisations have a limited
ability to take nominal derivational morphology but further fieldwork is required to
investigate their precise morphological properties.

(644) botwagumuk khayrataysa macota
[bot -wa] =gumuk {khay -rat} =ay =sa
pull -FACT =whole carry.on.body -DOWNWARD =ADV =DLIM
{macot -a}
finish -CUST
[They] finish by carrying the whole harvest down (on their body).

When a factitive-marked clause functions as the complement clause of a verb of
emotion and interaction (see 4.5.1ii), it has to take the dative enclitic <=na> (DAT),
as we can see in examples (645) and (646). As can be seen in example (645), which is
elicited, speakers find it acceptable to mark the predicate head of the complement
clause for modality with an inflectional suffix, in this case the uncertain future
modality suffix <-ni> (FUT). Modality marking with inflectional suffixes on factitive
complement clauses has not been recorded in narratives or spontaneous conversation.

(645) muma aaw ga'phonekniwana koreya.
|[muma] [a] =aw {ga'phonek -ni -wa} =na| {koroy -a}
elephant 1s =ACC stamp.to.death -FUT -FACT =DAT be.afraid -CUST
[I]m afraid an elephant will stamp me to death.

438
Factitive- and dative-marked object complement clauses are formally, but not
semantically or syntactically, similar to Reason clauses (described in 27.1.1). The
complement clause is functioning as a core argument within the matrix clause and is
thus governed by the predicate, while the Reason clause is an adjunct and is thus a
modifying clause. Both factitive-marked clause types take the dative case enclitic
<=na> (DAT) and both clause types are dependent. The semantic and syntactic
difference between the Reason adjunct and the factitive-marked complement clause is
caused by the semantics of the matrix clause predicate. The following two examples
are illustrations of a factitive-and-dative marked complement clause and a Reason
adjunct respectively. The valency of verbs of emotion and interaction, like kore- to
fear in (646), is discussed in 4.5.1ii. The delimitative in (647) is optional and can
also occur on adjunct clauses such as (646).

(646) morot toywana korea.
|[morot] {toy -wa}] =na {kore-a}
person die -FACT =DAT fear -CUST
I fear the persons death. Alternatively: Im afraid the person will die.

(647) ue toygat rowanasa ue toykhalawe rodo mowano.
|[ue toygat] {ro -wa}|
REASON
=na =sa

DST water.place drink -FACT =DAT =DLIM
[ue teykha]
O
=aw =e [rodo]
E
{mo -wa} =no
DST river =ACC =FC Rname call.a.name -FACT =QUOT
Because [they] drunk from that water place, [they] named that river
Rongdyng, it is said.

Multiple embedding is possible with factitive-marked complement clauses just as with
other types of nominalisation. The following example shows a factitive-marked clause
embedded in a headless complex NP (i.e. an NP of which the head, which is modified
by an attributive clause, is ellipsed) which is in turn embedded in a main clause. This
example also illustrates that factitive-marked clauses, like verbs, can be modified by
adverbial clauses. The adverbial clause rom-ay (cook=ADV) modifies sa'-wa (eat-
FACT).

439
(648) gasamci romay sa'wa jamcagabaaw nawepara kumiriparae garan baw'ay
tanoknokhon.
---------------------------------arch NP--------------------------------
--------------------attributive clause---------------------
--------complement clause---------
[gasam] =ci [ | |{rom} =ay {sa -wa}|
O
{jam -ca}| =gaba]
O
=aw

evening =LOC cook =ADV eat -FACT finish -NEG =ATTR =ACC
[nawe =para kumiri]
A
=para =e {garan baw} =ay
Pname =&co Pname =&co =FC jerky make.jerky =ADV
{tan -ok} =no =khon
put -COS =QUOT =SPEC
It might be the case that in the evening Naweng and Kumiri (in each others
company), put [the food which they] had not finished eating cooked [above the
fire] to make jerky, it is said. Literally, Naweng and Kumiri jerky-makingly
put [the food above the fire].

Elicitation has shown that factitive-marked complement clauses cannot be pluralised,
i.e. cannot take the plural enclitic <=dara> (p). This has to do with the fact that this
clause type is not a prototypical nominalisation in the sense that it does not have all
the nominal properties. More fieldwork is needed to find out if lexicalised factitive
nominalisations can be pluralised although they do not occur in the recorded data. It
might well be that even lexicalised factitive nominalisations do not have the full array
of nominal properties and that the action/state nominaliser enclitic <=wami ~
=wamo> (NR) is upcoming as a derivational morpheme of object nouns from verbs
in addition to its function as action state nominaliser.
24.3.2 Factitive-marked subject complement clauses
The factitive is used to mark the predicate of complement clauses in S function in
matrix clauses with a Type 1 adjective as predicate, as is illustrated in examples (649)
and (650) below. This is in contrast to the small group of verbs, treated in the
paragraph above that can take complement clauses in O function.

(649) morot dokomci tokwa ga'ca.
|[morot] [dokom] =ci {tok -wa}|
S
{ga' -ca}

person head =LOC hit -FACT good -NEG
Hitting a person on the head is not good.


440
(650) joksa ray'soracaka. sala! jalporawaan nemnaka aa, sala!
[jok] =sa {ray' -sora -ca -ka} sala
spouse =MOB go -COMPLETELY -NEG -IFT interj
|{jal -pora -wa}|
S
=an {nem -naka} [aa] [sala]
run.away -WITHOUT.DESTINATION -FACT =FC good -IFT 1s interj
[I] will not go back to my wives at all, damn! Running away without
destination will certainly be good that damn!

One secondary verb has been recorded with both dative and factitive complement
clauses. This is the verb ga'a- to be compelled. In Text 2 line 58 we find an
example with a factitive complement clause and an example with a dative
complement clause is shown in (758).
24.3.3 The syntactic status of factitive-marked complement clauses
Factitive-marked complement clauses, with the exclusion of dative-marked
complements of verbs of emotion and interaction, are formally the same as
independent or main clause factitive-marked predicates, as we will see in the
examples below. Moreover, case marking and argument structure in this type of
complement clause are the same as in main clauses. Since factitive-marked
complement clauses can occur as independent clauses forming a sentence on their
own, they are not dependent on a matrix clause for their occurrence. However, since
the complement clause does function as core argument, and is thus governed by the
predicate in the matrix clause, it is subordinate. It is worth quoting Christian Lehmann
(2007).

If the term asyndesis
50
is applied to verbal expressions (verbal sentences and
clauses), it presupposes their finiteness. This is because a non-finite verb form
signals its syntactic dependence morphologically. As a consequence,

50
The term asyndesis means that there is a lack of morphological marking. In the case of the type of
clause type under discussion there is no morphological marking that signals its subordination.

441
subordination can be diagnosed under conditions of asyndesis only for
intrinsic interpropositional relations, as in E22.

E22. I thought you were younger.

Here the valency of the verb in the first clause forces the analysis of the
second clause as a subordinate one.

In Atong, it is the semantics of the sentence that forces the analysis of the first clause
as being its complement and therefore subordinate.
Let us look at a set of examples that demonstrate the independent status of the
factitive-marked complement clause. In example (651) we see a factitive-marked
complement clause is S function in a matrix clause. In (652) we see that this same
clause can stand alone as a sentence.

(651) niba otokoy takwa ga'nima?
|[ni]
S
=ba [otokoy] {tak -wa}|
S
{ga' -ni} =ma
1pe =ADD like.that do -FACT be.good -FUT =Q
Will our doing like that be good?

(652) niba otokoy takwa.
[ni]
S
=ba [otokoy] {tak -wa}
1pe =ADD like.that do -FACT
We also did like that.
24.4 Factitive-marked clauses with dative and locative case-marking
Factitive-marked clauses with dative and locative case marking cannot occur as a
clause on their own and are thus dependent on a matrix clause for their appearance.
Factitive-marked adjunct clauses occur with various semantic roles, viz. Standard of
Comparison (dative-marked), Temporal Location (locative-marked) and Facsimile
(perlative/similative-marked). The various types of adjuncts will be treated one by one
in this order.

442
24.4.1 Factitive-marked Standard of comparison and Comparee clauses
The Standard of comparison and the Comparee in an event comparison are factitive-
marked clauses. As is discussed in Chapter 27, Standard of comparison clauses take
the dative enclitic <=na> (DAT), of which example (751), repeated below for
convenience as (653), is illustrative. There are no co-reference restrictions between
the arguments in the Standard of comparison clause and the main clause. This is
illustrated in Chapter 27 by comparing examples (751) and (752). The dative case
functions as mark of the Standard of comparison clause just as it does on Standard of
comparison NPs.
The Comparee NP is also a factitive-marked clause and is always marked by the
focus/identifier enclitic <=an> (FC/ID). The syntactic function of the Comparee is S
and thus it is a complement clause of the verb nem- be good which functions as the
Parameter. Since factitive-marked main clauses can also occur with the
focus/identifier enclitic <=an> (FC/ID), of which example (635) is illustrative, this
complement clause, like the others treated above, represents a case of asyndetic
subordination (see 24.3.3).

(653) umi gomonci aa na'aw khomana dayaydo asetwaan nemkhalnaka.
STANDARD ----------MARK----------
umigomonci [aa]
A
[na'] =aw {khom -a
51
} =na {day} =ay =do
therefore 1s 2s =ACC marry -FACT =DAT be.bigger=ADV =TOP
------------COMPAREE------------ PARAMETER INDEX
{asset -wa}
S
=an {nem -khal -naka}
throw.away/dispose -FACT =FC/ID good -CP -IFT
Therefore [it] will certainly be much better to throw [you] away than to be
married to you.

The following example illustrates the use of the dative enclitic on the NP khasi
khu'cuk Khasi language functioning as Standard of comparison. This example also

51
The form of the factitive suffix in this example is <-a> due to the phonological rule that the /w/
elides when the verbal root or stem ends in /m/ or /p/ (see Chapter 2).

443
illustrates that nominal Comparee NPs do not need to take the focus/identifier enclitic
<=an> (FC/ID) as do clausal Comparees.

(654) khasi khu'cukna dayay atong khu'cuk rakkhala
STANDARD ----------MARK---------- COMPAREE PARAMETER INDEX
[khasi khu'cuk] =na {day} =ay [ato khu'cuk] {rak -khal -a}
Khasi language =DAT be.bigger=ADV Atong language hard -CP -CUST
The Atong language is more difficult than the Khasi language.
24.4.2 Factitive-marked adjunct clauses with the dative case
Factitive-marked clauses can be dative-marked to fulfil an adjunct function in a
clause. The following semantic roles of dative-marked adjunct clauses can be
distinguished: Standard of comparison and Reason. Reason clauses are described in
27.1. It is interesting to note that nouns can also function as predicate head of a
reason clause. The function of the factitive on nominal predicate heads is to mark
them as predicate.
24.4.3 Factitive-marked adjunct clauses with the locative case
Factitive-marked clauses can be locative-marked to fulfil the adjunct function of
Temporal Location. This adjunct clause type is discussed in 27.5. In that section,
factitive-marked temporal adjunct clauses are contrasted with temporal adjuncts
without the factitive suffix, but with the locative encliticised directly to the root of the
verb. It is shown that the function of the factitive suffix is the same as that on main
clause predicates, i.e. reification, presenting the event denoted by the verb as a fact.
24.4.4 Factitive-marked adjunct clauses with the similative case
Factitive-marked clauses can take the perlative/similative enclitic <-tokoyi ~ -takay ~
-takoy> (LIKE) (see 20.9), illustrated in (655) here below. In all recorded instances
the factitive-marked clauses with this suffix were Facsimile NPss and not Pathways.
More fieldwork needs to be done to find out if it is possible for factitive-marked
clauses to occur as Pathway. There are no co-reference restrictions between the
arguments of the Facsimile clause and those of the main clause.

444
(655) ian maja na'balwatokoy de'the gam jamok
[i] =an [maja] [na'] {bal -wa} =tky
PRX =FC/ID in.the.past 2s say -FACT =LIKE
[de'the gam] {jam -ok}
3s wealth finish -COS
This [is] like you said in the past, his wealth has finished.
24.5 Factitive-marked complement clause of postposition
The limitative postpositions dabat (LIMIT), gomon reason, about and konsa after
require verbal complements to be marked with the factitive suffix <-wa> (FACT).
Example (656) is illustrative of a complement clause of the postposition dabat
(LIMIT). The postpositions gomon reason, about and konsa after govern the
genitive, hence the occurrence of the enclitic <=mo> (GEN) on the complement
clauses in (657) and (658).

(656) otokoymo jaraw jaraw ge'the sokwa dabatdo sakcikaydoano pheruba.
otokoymo [jaraw jaraw] [ge'the] [{sok -wa} dabat] =do
so.then LONG RED 3s not.hold.out -FACT LIMIT =TOP
{sak -cik -aydoa} =no [pheru]=ba
hold.out -AS.LONG.AS.YOU.CAN -PROG =QUOT fox =EMPH
So then, for a long time, until he did not hold out any longer, [he] was
holding out as long as he could, it is said, the fox.

(657) rongdo toykhal ha'wayci mu'wamigomonsa rodo ha'way noay mowano.
[[rodo toykhal ha'way] =ci {mu' -wa} =mi gmn]
RIVER.name river plain =LOC stay -FACT =GEN reason
[rodo ha'way] {no} =ay {mo -wa} =no
Pname say =ADV call.a.name -FACT =QUOT
Because they live in the plains of the river Rongdyng, [they] sayingly call
[the village] Rongdyng Hawai, it is said.


445
(658) nok rophiwamo konsado te'ew ge'the nokawan alaga morotdorado
tecopcopay nukariokno.
[[nok] {rophi -wa} =m knsa] =do
house plaster -FACT =GEN after =TOP
[te'ew] [ge'the nok] =aw =an
now 3s house =ACC =FC/ID
[alaga morot] =dora =do {te -copcop} =ay {nuk -ari -ok} =no
other person =p =TOP shine -ALL.OVER =ADV see -SIMP -COS =QUOT
After the plastering of the house, now other people found his house just shiny
all over, it is said. (Because he had plastered it with a mix of cow dung and
gold flakes.)

Postpositions are treated in detail in Chapter 1.
24.6 Summary of properties of factitive-marked clauses
The factitive suffix clearly functions as a modality suffix on main clause
predicates and can be seen as a clausal nominaliser on subordinate clauses. However,
factitive-marked clauses have no nominal properties, except for the fact that they can
be case-marked, and the number of attested cases on subordinate clauses is very
limited, viz. dative (Standard of comparison, Reason), locative (Temporal Location)
and perlative/similative (Facsimile; Pathway not attested).
The difference between the modality function and the nominalising function of
<-wa> (FACT) is clearest when we compare main clauses and complement clauses.
Complement clauses refer to actions or states, depending on the semantics of the verb,
while main clauses do not. By contrasting factitive-marked with non-factitive-marked
Temporal Location adjunct clauses, we see that the role of the factitive is the same as
in main clauses, viz. reification, presenting the event denoted by the verb as a fact.
This means that, although the factitive was probably once a nominaliser in an earlier
stage of the language, it has developed in the direction of a modality marker, although
it still preserves a hint of its old nominalising function.
There are some lexicalised object nominalisations with the factitive suffix, but it
appears that these do not possess all nominal properties. More fieldwork is needed to
find out what the differences are between factitive lexicalisations and object
nominalisations with the enclitic <=wami ~ =wamo> (NR).

446
Factitive-marked Reason clauses can have a noun as predicate head. If the
factitive suffix were not there, they could not be interpreted as Reason adjuncts since
nouns cannot fulfil this function, only clauses can. Dative marked nouns can be
Patient, Recipient, Beneficiary, Experiencer, Goal, and Standard of comparison. The
factitive on nouns thus indicates that the noun functions as a predicate. This is not
converse to the function of the factitive on subordinate clauses with verbal predicates,
since it is not the factitive but the case marker that indicates that the clause is used as
adjunct. Factitive-marked predicates without case can occur as main clause on their
own, but a predicate head that is case-marked cannot. Table 67 presents a schematic
summary of the different functions of the factitive suffix <-wa> (FACT) on different
types of predicates with and without case marking.

Table 67 The functions of the factitive suffix <-wa> (FACT)
on different types of predicates with and without case marking

on verbal predicates
on nominal
predicates NO CASE MARKER
WITH CASE
MARKER
CLAUSE
TYPE
independent/main
clauses
governed subordinate
clauses in S or O
function, i.e.
complement clauses
or complement of
postposition
modifying
subordinate
clauses in
adjunct function
to a matrix
clause
modifying
subordinate
clause in
adjunct
function to a
matrix clause
FUNCTION
reification
modality
clausal
nominaliser/reification
modality
modality marker,
nominaliser
marker of
predication
SEMANTIC
ROLE AS
ADJUNCT

Reason, Standard
of comparison,
Temporal
Location,
Facsimile
Reason
24.7 Diachronic note
No internal reconstruction of Atong has thus far been attempted let alone historical
comparison with other languages of the Bodo-Koch group Therefore, any diachronic
statement about the language is highly speculative. However, it is important to note
that, although it may be probable that the morpheme <-wa> (FACT) historically
derives from the Proto Tibeto-Burman nominaliser *pa, it does not serve as a
nominaliser in main clauses in the current stage of Atong but rather as a modality

447
suffix of reification, i.e. indicating that the event denoted by the verb is a fact. This
modality function has spread to locative adjunct clauses as we can see in 27.5. The
nominalising function of the factitive suffix is still preserved in other semantic types
of subordinate clauses and lexicalised object nominalisations. It is conceivable that a
main clause predicate marked with <-wa> (FACT) could once have been considered a
stand-alone nominalisation in the sense of Matisoff (1972: 246) and Noonan (1997:
380-1), but this stage has clearly come to an end.

448
Chapter 25 Event specifiers
_____________________________________________________________________

Atong has a wide variety of event specifiers. Due to limitations of space, they cannot
all be treated in this grammar. Therefore, this chapter simply provides an overview of
the event specifiers recorded to date. Some examples of the use of event specifiers
will be given as illustration.
25.1 The function of event specifiers
Event specifiers are predicate head suffixes that give information about how the state
or event depicted by the predicate comes about. These suffixes may also simply
reinforce the meaning of the predicate. A verbal predicate head can take more than
one event specifier suffix. Type 2 adjectival and nominal predicate heads with more
than one event specifier are not attested. The event specifiers seem to appear in a
semantically motivated order, the suffixes that come later having scope over those
preceding. Event specifiers are Echelon 1 column 2 suffixes (see Table 63) and can
co-occur with all other types of suffixes in the suffixal string of the predicate head, as
long as the result of the combination is semantically felicitous. As has been mentioned
in Chapter 22, the order of the suffixes within Echelon 1 is not fixed but varies. There
is, however, a strong tendency for the suffixes to appear in the order in which they are
depicted in Table 63.
The combinatory possibilities of event specifiers with predicate heads depend on
the semantics of the head and the suffix. Some event specifiers have a very specific
meaning and a semantically felicitous combination can only be made with a select
group of verbs. Other event specifiers can be used on almost any verb and even with
Type 2 adjectives and nouns, because their meaning is less specific.
As has been illustrated in Chapter 18, certain event specifiers participate in the
process of word class changing derivation, bestowing verbal properties on the non-
verbal predicate head they attach to. More fieldwork is needed to find out exactly
which event specifiers participate in this grammatical process and exactly what verbal
properties they can and cannot transfer to a non-verbal predicate head.
25 EVENT SPECIFIERS

449
25.2 Origin and meaning differentiation
Some event specifiers correspond to verbs, nouns and adverbs which are found in the
language as separate lexical items. This suggests that event specifiers are the result of
grammaticalisation through compounding and reanalysis (see Hopper and Traugott
1993: 32-62). Event specifiers have lost part of the meaning that they had as lexical
items and came to have a more abstract meaning. The event specifier taw V
upwards, for instance, is homophonous to the verbal root taw to go up, ascend, and
the event specifier phak V in half lengthwise, V lengthwise, V and go through
lengthwise, V by the side of something, V side by side has a nominal cognate form
phak half (the result of a longitudinal cut). In example (659) we see the morpheme
taw function as the verb to go up, ascend, contrasting with its semantic counterpart
wol- to descend. In (660) we see the same form functioning as event specifier.

(659) ray'sotwae tawaaymu ue gripnok
52
hamgabatokoysa wolathiriokno.
[ray' -sot -wa] =e {taw -a} =ay =mu
go -DIRECTLY -FACT =FC ascend -AWAY =ADV =SEQ
[ue grip nok ham =gaba] =tokoy =sa
DST Name house build =ATTR =VIA =DLIM
{wol -a -thiri -ok} =no
descend -AWAY -AGAIN -COS =QUOT
Having gone up the shortcut, [he] descended again via the G.R.E.E.F. house,
which is built [there]

52
Grip is the Atong pronunciation of the abbreviation G.R.E.E.F. (some Atong speakers say), which
was probably the name of a British coal mining company operating in the Garo Hills. Although the
company does not exist any more, the remnants of the mining business can still be found between Jadi
and Badri.
25 EVENT SPECIFIERS

450
(660) gore [] kambaysa powtawaokno.
[gore] [kambay] =sa {pow -taw -a -ok} =no
horse above =MOB fly -UPWARD -AWAY -COS =QUOT
the horse, [having seen the idiot, having run away], flew away upward to the
above [i.e. into the sky].

Some event specifiers can be interpreted differently, depending on the verb they are
suffixed too, and on the context. The next example illustrates the use of the event
specifier -a V away on the verb bot- to lead, and can be compared to example
(765) in Chapter 27 where the same event specifier is used with the meaning V
without holding back on the verb khu'mo to conspire.

(661) [] ge'the ma'su maboroyaw botaaymo [].
[ge'the] [ma'su ma boroy] =aw {bet -a} =ay =me
3s cow CLF:ANIMALS four =ACC lead -AWAY =ADV =SEQ
he lead the four cows away [and ate them, it is said].
25.3 Categories
I am convinced that there are many more event specifiers in Atong than those
recorded during the fieldwork for this grammar. The ones so far collected can be
divided into twelve categories according to their meanings. These categories are
manner, manner/direction, aspect, extent, direction/extent, direction, epistemic,
deontic, determinacy, location, conative and quantification.
25.4 Striking phonetic feature
One of the striking features of event specifiers is the versatility in the pronunciation of
the glottalised consonant coda, if they have one. Different speakers pronounce the
same event specifier in different ways, i.e. with or without glottalised consonant coda.
The same speaker may use the same morpheme in the same context with a glottalised
and plain consonant coda. The pronunciation does not seem to have anything to do
with a preceding morpheme containing a glottalised consonant or glottal stop coda as
in Garo (see Burling 2004: 35-38).
25 EVENT SPECIFIERS

451
25.5 Overview and some comments
All the event specifiers recorded to date are given by category and in alphabetical
order in Table 68. In the column Meaning in this table, the capital V stands for any
Atong verb, or to be more precisely, predicate,
53
although nouns and Type 2
adjectives are almost never used with event specifiers.
The suffix <-ton> lead in V-ing, V as the leader can have a transitivising effect
on intransitive verbs. In example (662) we see how the transitive intransitive verb jal-
to run away becomes transitive and can therefore take the accusative-marked O
argument sipay soldiers.

(662) tay'nido do'ancak na'nado noaymo sipaydoraaw jaltonokno.
[tay'ni] =do {do' -an -ca -k} [na'na] =do {no} =ay =mo
today =TOP sufficient -REF -NEG -COS 1pi =top say =ADV =SEQ
[sipay] =dora =aw {jal -ton -ok} =no
soldier =P =ACC run.away -AS.THE.LEADER -COS =QUOT
Today [things] are not right any more, as far as we are concerned, [he] said
and lead the soldiers away running.

The quantifying event specifiers (see Table 68 (c)) work on an S/O basis. In example
(663) it is the S argument phulis police that is quantified on the predicate by the
event specifier rum all, while in (664) it is the O argument mo' korok
(CLF:HUMANS four) four people that is quantified by the same event specifier.

(663) phulis bisagsa ray'arumwasoy tay'nido?
[phulis]
S
[bi] =sa {ray' -rum -wa} =soy [tay'ni] =do
police QF =MOB go -ALL -FACT =MIR today =TOP
Where are the police all going today?

53
This way of representing the meaning of event specifiers is inspired by Okell and Allott 2001.
25 EVENT SPECIFIERS

452
(664) pholgom cugaba monokrumokno mo' korokawan.
[pholgom cu =gaba]
A
{monok -rum -ok} =no
eagle big =ATTR swallow -ALL -COS =QUOT
[m korok]
O
=aw =an
CLF:HUMANS four =ACC =FC/ID
The big eagle had swallowed them all, the four [of them].

As was said above, sometimes an event specifier can be used to simply reinforce the
meaning of the predicate, to add emphasis. Especially the suffix -sora V completely,
V wholly, V till the end, V very much is frequently used in colloquial speech for this
purpose. Good examples of this event specifier as emphasiser can be found in Text 2
lines 33, 35, 38 and 57. Line 35 is presented here as example (665).

(665) hay, wal'botawba hon'etsora
[hay] [wal'bot] =aw =ba {hon -et -sora}
come.on match =ACC =EMPH give -CAUS -TOTALLY
Come on, give the matches too already.

25 EVENT SPECIFIERS

453
Table 68 Event specifiers
listed according to category with corresponding lexical items where these
were found.
EVENT
SPECIFIER
MEANING
CORESPONDING
LEXICAL ITEM
MEANING
MANNER
-ca V suddenly
-cap
V along with
someone/something

-cep V alone
-cop
V wastefully, V
unsuccessfully

-cici V with force, V into pieces
-cikcak V in a swarm
-damdam V in different places dam (noun) place
-dap V and add, V on top dap- (verb) to cover, to be on top
-gak V accidentally
-jokjok V up and down
-jol V quickly
-joljol V very quickly
-khaw V secretly, V and steal
-khelek V for fun khele- (verb) to play
-khep V firmly
-na V in a beautiful or nice way
-nap V with all your heart
-para ~
-pora
V without destination,
without goal, aimlessly
prara- ~ pora-
(verb)
to journey, to travel,
to wander, go astray
-pol V rapidly
-porak V and cut
-phak
V in half lengthwise, V
lengthwise, V and go through
lengthwise, V by the side of
something, V side by side
phak (noun,
classifier)
NOUN: half (the
result of a
longitudinal cut);
CLASSIFIER for
halves of objects cut
lengthwise
-phet V detrimentally
-phetphet V repeatedly
-phin' ~
-phon'
V back, over-V, obviously V,
V fully

-ramram V normally, V naturally ramram (adverb) usual
-sak V appropriately sak- (verb) to fit (into)
-saw
V and wait, V expectantly, V
for sure

-sega ~ -siga
V in turn, (alternative) (also
phrasal enclitic, see Chapter
19)

-sek V and steal
-som
V and follow; imitate
someones V-ing

-seme V reluctantly

25 EVENT SPECIFIERS

454
Table 68 continued (a): manner, manner/direction
EVENT
SPECIFIER
MEANING
CORESPONDING
LEXICAL ITEM
MEANING
Manner continued
-somsom V continuously
-set ~ -sot
to V so as to dispose of
something

-sot V directly
-susa V competitively
-ton V as the leader, lead in V-ing
-tota V all over the place
-thol V and avoid, V ahead
-tholo V nicely
-tho ~ -thi only V
-thotho
only V and nothing else,
continuously V (more intense
than -tho)

-thirithiri ~
-therirtheri
V again and again
-tho'
V in half, V crosswise, V and
go through crosswise
tho' (noun,
classifier)
NOUN: half (the
result of a crosscut);
CLASSIFIER for
halves of objects cut
crosswise
-thum
V on behalf of someone else,
V for the benefit of someone
else/something

-wenwen V in circles
-wowa V in a confused way
-wil ~ -wilwil V around
MANNER/DIRECTION
-rore V while spinning around
-thiri ~ -theri
V again, back,
reversely/backward

25 EVENT SPECIFIERS

455
Table 68 continued (b): aspect, extent
EVENT
SPECIFIER
MEANING
CORESPONDING
LEXICAL ITEM
MEANING
ASPECT
-ce V first
ha'ba-ce
(?-first)
54
(verb)
to begin
-dokdok about to V
-gat
V upon to, to start V-ing
(inceptive)

-man ~ -man' already V-ed
-mu' keep V-ing mu'- (verb) to sit, stay
-rawraw continue to V
EXTENT
-an still V-ing
-baray V always
-bat V even more, V most
-bolok V into pulp
-bi very
-bobo
V more than necessary, V in
abundance (pejorative), V
scandalously much
bobo (noun) liar
-cik ~ -cok V as long as you can
-gacak V until it is red hot
-joro V daily, V all the time
-lala very
-mama
V simply, as best you can,
barely, just

-porot over-V
-ro usually, always
-ruru
V more and more, V around,
V all over the place

-sora
V completely, V wholly, V
till the end, V very much

-te still too V
-tete still much too V
-tham barely V
-thamak barely V
-that V excessively
-thethe still too V (more intense
than -te)

54
I consider the two parts of the verb ha'ba-ce to begin to be bound morphemes. While the second
part is readily identifiable as the morpheme corresponding to the event specifier -ce V first the first
part ha'ba- does not occur as a separate verb in the language and has not been attested in any other
lexical item. There is a noun ha'ba dry rice and vegetable field which I do not consider to have
anything to do with the bound morpheme in question.
25 EVENT SPECIFIERS

456
Table 68 continued (c): direction/extent, location, epistemic, deontic, determinacy,
conative, quantification
EVENT
SPECIFIER
MEANING
CORESPONDING
LEXICAL ITEM
MEANING
DIRECTION/EXTENT
-a V away, V affluently, V
without holding back

DIRECTION
-a ~ ay V towards
-pat V across pat- (verb) to cross, go across
-rat V downward rat- (verb) to go down, descend
-soso V to/on the ground
-taw V upward taw-(verb) to go up, ascend
-wil V around
-wilwil V around and around
LOCATION
-dap V on top of something dap- (verb) to be on top, to press,
keep together by
force, pinch together,
to pinch, to crush, to
stack
-tota V all over the place
EPISTEMIC
-asol really V, verily V, actually V asol (adverb) really
-bebe truly V verily V bebe (adverb) truly
-cay ~ -coy try to V cay- (verb) to look (at)
-dam truly V
DEONTIC
-tat compulsory V
DETERMINACY
-comot ~
-comot
V determinedly, V certainly,
V definitely

-thel surely V
CONATIVE
-cay ~ -coy try to V, V and see cay- (verb) to see
-ram V inadvertently, V
unintentionally, V
fortuitously, V because of the
situation
ram- (verb) to search
QUANTIFICATION S/O
-gorop V together, V with a whole
group

-korom V in a group
-pha V also, V in addition, V along
with, V together, V in total

-rum V all, all V
-thok V together, everybody V, all
V

457

458
Chapter 26 Clause Types
_____________________________________________________________________

This chapter treats the general properties of dependent and independent clauses, and
subsequently focuses on independent clauses. Dependent clauses are also called
subordinate clauses and the independent ones can also be termed main clauses.
The major clause types in Atong are presented in Table 69 below. The predicates of
all clause types can be headed by a verb except in presentative clauses and
predicateless interrogative clauses. There are fewer clause types in which Type 2
adjectives are attested as predicate head and even less of those in which nominals can
head the predicate.
Main clauses can, but do not have to carry a clausal enclitic, while all subordinate
clauses are signalled by a clausal enclitic. All main and subordinate clause clausal
enclitics are summed up in Table 70. The main clause types will be treated one by one
below. Subordinate clauses are treated in separate chapters as can be seen in Table 69.
The last two sections treat the functions of the irrealis and the speculative enclitics.

Table 69 Clause types in Atong
Their syntactic status and the type of predicate head that they can occur with are ordered by section.
The abbreviations used in this table are as follows. V verb, ADJ2 Type 2 adjective, N nominal.
Section
SYNTACTIC
STATUS
CLAUSE TYPE
ATTESTED
PREDICATE
HEADS
26.1
I
N
D
E
P
E
N
D
E
N
T
/
M
A
I
N

Interrogative clauses
V, ADJ2, N
26.1.1 content questions
26.1.2 predicateless interrogative clauses none
26.1.4
marked polar questions
V, ADJ2, N
unmarked polar questions
26.2
Imperative clauses
V different levels of politeness, prohibitives
and optative
26.3
Declarative clauses
V, ADJ2, N marked declarative clauses
unmarked declarative clauses
26.4 presentative clauses N
26.5 copula clauses V (copula)
Chapter 27
D
E
P
E
N
D
E
N
T
/

S
U
B
O
R
D
I
N
A
T
E
Dative-marked V, N
Locative-marked clauses V, ADJ2, N
Chapter 28
Sequential clauses
V, ADJ2, N
Adverbial clauses
Chapter 29
Attributive clauses V
26 CLAUSE TYPES

459
Table 70 Clausal enclitics
The ones in bold are attested on clauses with nominal predicates
Main clause clausal enclitics
column 1
Irrealis modality <=cm> (IRR) Also particle, see 26.8.
Speculative modality <=khon> (SPEC) Also particle, see 26.9.
Imperative mood <=bo> (IMP)
Prohibitive mood <=bay> (PROH)
column 2
Declarative <=te> (DCL)
Mirative <=thay ~ =ty ~ =sy> (MIR)
Emphatic positive <=ay> (POS)
Confirmative tag <=mo> (CONF) obviously,naturally, Also used as
particle.
Affirmation seeking tag <=ne> (TAG) Also used as particle
Emphatic/Additive <=ba> (EMPH/ADD) Also phrasal enclitic (see
Chapter 19)
Interrogative <=ma> (Q) Also particle.
Emphatic <=aro ~ =ro> (EMPH)
Imperative emphasiser <=to ~ =ta> (IMPEMPH)
Main clause movable clausal enclitic: Evidentiality
Quotative <=no> (QUOT)
Subordinate clause clausal enclitics
column 1
Attributive <=gaba ~ =ga> (ATTR)
Locative <=ci> (LOC) (see 24.4)
Dative <=na ~ =ona> (DAT) (see Chapter 27)
Adverbial <=ay ~ =e> (ADV)
column 2
Sequential <=m ~ =mu ~ =muna ~ =mu> (SEQ)
column 3
Focus <=e> (FC) Also phrasal enclitic (see Chapter 19)
Main and subordinate clause clausal enclitic
column 3
Focus/identifier <=an> (FC/ID) Also phrasal enclitic (see Chapter 19)
Delimitative <=sa> (DLIM) Also phrasal enclitic (see 11.7)
Indefinite <=ba> (INDEF)
Topic <=do ~ =odo> (TOP) Also phrasal enclitic (see
Chapter 19)
26 CLAUSE TYPES

460
General properties of independent and dependent clauses
Both independent and dependent clauses can have verbal and non-verbal predicates.
Presentative clauses, as can be seen above in Table 69, only contain nominal
predicates, and in predicateless interrogative clauses no predicate can be identified.
The majority of clauses in the recorded material are predicate final. However, any
constituent, i.e. argument or adjunct, of a main clause can be right dislocated, i.e.
appear after the predicate for backgrounding. Right dislocation is impossible in
subordinate clauses.
All clause types can be juxtaposed to clauses of the same type and there are no
restrictions on the focusability or topic marking of constituents in any clause type.
However, constituents of dependent clauses tend to be less frequently marked for
topic or focus than those of independent clauses. When we look at topic or focus
marking of whole clauses, we find that only dependent clauses can be focused and
that both dependent and independent clauses can be topic-marked. However, only
independent clauses with a predicate head carrying the factitive suffix <-wa> (FACT)
can be topic-marked, whereas dependent clauses of all types can be topic-marked.
The following example illustrates a topic-marked main clause with right
dislocated subject (S) raja=mi dama (king=GEN drum) the kings drum.

(666) ha'no' tara cinina imo korowado rajami dama noaydoano.
[ha' no' tara ci ni] =na [i] =mo {koro -wa} =do
earth inside layer TEN two =DAT PRX =ABL sound -FACT =TOP
[raja =mi dama]
S
king =GEN drum
The kings drum sounds from here to the twelve layers of the earths inside,
[the deer] is saying, it is said.

Case marking of constituents (described in Chapter 20) is the same in all types of
clauses.
Table 71 below summarises the general properties of independent and dependent
clauses.

26 CLAUSE TYPES

461
Table 71 General properties of independent and dependent clauses
PROPERTY INDEPENDENT CLAUSES DEPENDENT CLAUSES
Type of predicate head verbal and non-verbal
Constituent order
Relatively free, predicate final in
unmarked scenario. Any constituent
can be right dislocated, i.e. postposed
to the predicate.
Strictly predicate final
Marking of predicate
head
More marked. Less marked.
Can be negated
Referentiality Can be expressed. Cannot be expressed.
Evidentiality with
quotative suffix
Can be expressed. Cannot be expressed.
Declarative Can be expressed. Cannot be expressed.
Imperative Can be expressed. Cannot be expressed.
Interrogative Can be expressed. Cannot be expressed.
Juxtaposition with
clauses of the same type
Yes.
Focus/Topic marking of
constituents
Often Rare
Focus/Topic marking of
clause
Rare Yes
Grammatical relations Case-marking of NPs is the same for all types
26.1 Interrogative clauses
There are three types of interrogative clauses, viz. marked polar questions and content
questions and predicateless content questions. In this section I will also treat the
alternative question strategy, which is a sentence which consists of minimally two
clauses. We will see how unmarked polar questions are just a function of declarative
clauses. The predicate head of an interrogative clause can be inflected for all
independent clause categories (see Table 63 and Table 69) except declarative <=te>
(DCL) and mirative <=thay ~ =toy ~ =soy ~ =si> (MIR). The constituent order in
interrogative clauses is not fixed, just like in other clause types.
26.1.1 Content questions
Content question clauses contain an interrogative or question word. Interrogatives are
treated in Chapter 9. There is a tendency for interrogatives to appear in clause initial
position, but this is not obligatory. Examples of interrogative clauses can be found at
the beginning of Text 2. As we can see in lines 7, 10 and 25 of Text 1, content
question clauses can also consist of just a question word, in which case they are
26 CLAUSE TYPES

462
elliptical. Examples (667) and (668) are illustrations of content questions with more
than one constituent.

(667) ato balanaka?
[ato] {bal -a -naka}
what tell -WITHOUT.HOLDING.BACK -IFT
What more shall [I] tell?

(668) ca ana daygaba raja?
[ca] {|[a] =na {day}| =gaba raja]}
who 1s =DAT be.bigger=ATTR king
Who is a bigger king than me?

Example (669) is a desiderative content question. Desiderative content questions
question a possibility. The context of the next example is as follows. A man explains
to his wife that he has inspected both the fish traps upstream and downstream in the
river. Then he says (669).

(669) biaw caykhuna? ana ni'ok
[bi] =aw {cay -khu -na} [a] =na {ni' -ok}
which =ACC look.at -INCOM -DESI 1s =DAT not.exist -COS
Which other one can I look at? I have no more.

An example of a desiderative content question can be found in the text1, line 36.
26.1.2 Predicateless focus content question clauses
In a predicateless focus content question clause the noun which is the focus of the
interrogation is always marked by the focus enclitic <=e> (FC). In this clause type it
is impossible to determine which constituent is the predicate head since none of the
constituents can take any predicate suffixes as can nouns when they are the head of a
predicate of a non-interrogative clause (see Table 63). It has also been impossible to
determine any subject properties for the constituents of these clauses. All
interrogatives can occur as constituent of predicateless content question clauses as in
(670) (671) and (672) below.
26 CLAUSE TYPES

463
(670) bie na' jodorae? na' joe bie? nookno.
|[bie] [na' jo] =dora =e|
CLAUSE

where 2s younger.brother =p =FC
|[na' jo] =e [bie]|
CLAUSE
{no -ok} =no
2s younger.brother =FC where say -COS =QUOT
Where are your younger brothers? Where are your younger brothers? [she]
said, it is said.

(671) na'mi jorae ca?
[na'] =mi jora =e [ca]
2s =GEN love.match =FC who
Whos your love match?

(672) bisanasa na'tome?
[bi] =sa =na =sa [na' -tom] =e
QF =MOB =DAT =DLIM 2s -ppp =FC
To where exactly [are] you [going]?

One could hypothesise, by analogy with the interrogatives treated below in 26.1.3,
that the interrogative is the predicate head.
26.1.3 Clauses with interrogatives as predicate head
The interrogatives biskon how much/many and bisa to where can express
perfectivity by means of the change of state suffix <-ok> (COS), e.g. (673) (674) and
can even take event specifiers as is illustrated in (675). Therefore these interrogatives
can always be identified as predicate head when they appear change of state-marked
in verbless interrogative clauses. More fieldwork is required to find out if there are
restrictions on the types of event specifier that can appear in these interrogatives.

(673) ama, dadaparae bisaok?
ama [dada] =para =e {bi =sa -ok}
mother older.brother =&co =FC QF =MOB -COS
Mother, where did [my] elder brothers go off to?

26 CLAUSE TYPES

464
Example (151), here repeated as (674), illustrates a change of state-marked
interrogative functioning as predicate head with a right-dislocated dative adjunct.

(674) biskonok, mo'thamna?
{biskon -ok} [mo' tham] =na
how.much -COS CLF:HUMANS three =DAT
How much is it in total, for three persons?

In example (675) we see the interrogative bi=sa (QF=MOB) to/from where?
functioning as predicate head and carrying an event specifier suffix, viz. the quantifier
<-rum> ALL and the change of state suffix <-ok> (COS).

(675) bisarumok na'tom sendele?
{bi =sa -rum -ok} [na -tom sendel] =e
QF =MOB -ALL -COS 2s ppp sandal =FC
Where have all your sandals gone?
26.1.4 Marked and unmarked polar questions
Marked polar questions have the same structure and intonation as declarative clauses
except that the question enclitic <=ma> (Q) appears on the clause, which is an
epistemic modal indicating the speakers uncertainty about the event. This enclitic
receives stress and that means that it is usually pronounced at a higher pitch than the
rest of the clause. The use of the question enclitic <=ma> excludes all other enclitics
from the clause.
Examples (676) and (678) below are examples of marked polar questions. In (678)
the polar question is an embedded direct speech complement of the verb no to say.

(676) na' soci may sa'ama?
[na' so] =ci [may] {sa' -a} =ma
2s village =LOC rice eat -CUST =Q
Is rice eaten in your village (alternatively: country)?
26 CLAUSE TYPES

465
(677) o came, ami na'na kha'galgabaaw na'mi khathoci daetna
man'phanima?
[o came] [[a] =mi [na'] =na {kha'gal} =gaba] =aw [na' =mi
interj sweetheart 1s =GEN 2s =DAT love =ATTR =ACC 2s =GEN
kha'tho] =ci {da -et} =na {man' -pha -ni} =ma
heart =LOC enter -CAUS =DAT be.able -IN.ADDITION -FUT =Q
O sweetheart! will you be able to insert also my love for you into your heart?

(678) ucie nepale otokcido a re'esiganima na'mo phal? nowano.
ucie [nepal] =e {otok} =ci =do
then Nepali =FC do.like.that=LOC =TOP
[a] {re'e -siga -ni} =ma [na' =mo phal] {no -wa} =no
2s go.away -ALT -FUT =Q 2s =GEN place say -FACT =QUOT
Then the Nepali: In that case, shall I go instead of you? sayingly said, it is
said.

Nominals and adjectives can also be the head of a predicate of a marked polar
question as examples (679) and (680) respectively illustrate. In (680) the first person
personal pronoun, a, is right dislocated as antitopic (see Lambrecht 1994: 202).

(679) cikorakca, aa sa'gorayma?
{cikorak -ca} [aa] {sa'goray} =ma
joke -NEG 1s child =Q
[I] dont joke. Am I a child?

(680) mokha pisakma ado?
[mokha] {pisak} =ma [a] =do
face red =Q 1s =TOP
Is my face red? Literally: Is the face read, as far as Im concerned?

Unmarked polar questions are formally indistinguishable from declarative clauses and
can therefore be said to be just one of the pragmatic functions of a declarative clause.
Intonation does not help to distinguish declarative clauses indicating a statement from
unmarked polar questions. Marked polar questions can have the same falling
intonation as declarative clauses, but are formally distinguishable from declarative
clauses because of the clausal enclitic <=ma> (Q). Usually unmarked polar questions
26 CLAUSE TYPES

466
have a falling intonation, just like declarative clauses indicating a statement. When a
speaker is surprised or otherwise emotional, the unmarked clause will have a rising or
fairly level intonation with maybe a slight dip at the end and can still be interpreted
either as a polar question, as a statement, or even an exclamation, e.g. (718).
26.1.5 Alternative question sentences
The alternative question sentence is a strategy in which two clauses are combined.
The predicate head of the first clause is marked with the enclitic <=ma> (Q) and the
second is not. The predicate head of the second clause cannot take any clausal
enclitics. The alternative question has falling intonation stretched out over the entire
sentence.
The following two examples show alternative questions with verbal predicates. As
example (682) illustrates, both predicates do not have to consist of the same verb.

(681) ranustaw nukama nukanca?
[ranus] =taw {nuk -a} =ma {nuk -an -ca}
Name =ACC see -CUST =Q see -REF -NEG
Have you seen Ranus or not?

(682) ra nemceama na'na cow jamcea
[ra] {nem -ce -a} =ma [na'na cow] {jam -ce -a}
rain good -FIRST -CUST =Q 1pi liquor finish -FIRST -CUST
Will the rain get better [i.e. stop] first or will we run out of liquor first? [the
Badri people said and they started a drinking competition with the rain.]

The following examples show alternative questions with non-verbal predicates. In
(683) the predicates are nominal, whereas in (684) they consist of demonstratives.
Since the question morpheme <=ma> (Q) is a clausal enclitic and not a suffix, the
demonstratives appear in their unbound form, viz. ie (PRX) instead of their bound
form (PRX).

(683) ido the'thonte. me'maokma jowmaok?
[i] =do {the'thon} =e {me'ma -ok} =ma {jowma -ok}
PRX =TOP Name =DCL ghost -COS =Q dream -COS
This is Thengthon, Im telling you. Has [he] become a ghost or a dream?
26 CLAUSE TYPES

467
(684) iema ie?
{ie} =ma {ie}
PRX =Q PRX
This one or this one?
26.2 Imperative clauses
The imperative is the only clause type in which a verbal predicate head can appear as
a bare root or stem without any predicate head suffixes or clausal enclitics, and it is
the only main clause type in which a verbal predicate can occur without predicate
head suffixes. The bare stem is the root plus stem-forming suffixes (predicate head
suffixes are listed in Table 63). An imperative can optionally be signalled by the
imperative clausal enclitic <=bo> (IMP), depending on the level of politeness, as we
will see below. The structure of the predicate head of an imperative clause is shown is
Table 72 and will be explained below. Imperatives are second person only. Atong also
has an optative, treated below in 26.2.4, which is only attested with third person.

Table 72 The structure of the fully inflected imperative predicate head
r
o
o
t
Column 1 Column 2 Column 3 Column 4
causative <-et>
(CAUS)
event specifier
imperative
<=bo>
(IMP)
other clausal
enclitics
(only after
<=bo> (IMP))
simplicitive <-khu>
(IMP)
incompletive <-
ari> (SIMP)

Not all stem-forming suffixes can occur on imperatives; only the derivational
causative <-et> (CAUS), event specifiers, the simplicitive <-ari> (SIMP) and
incompletive aspect suffix <-khu> (INCOM) can. Atong imperatives cannot take the
negative suffix <-ca> (NEG). The only aspectual suffix which can possibly occur on
an imperative is the incompletive aspect stem-forming suffix <-khu> (INCOM).
Imperative clause predicates cannot take any inflectional suffixes (see Table 63).
Event specifiers, the simplicitive and the incompletive suffixes are all mutually
exclusive on predicates of imperative clauses. When an event specifier is used, the
simplicitive or incompletive cannot be added to the stem, which is a restriction not
found in other clause types.
The incompletive suffix <-khu> (INCOM) has two meanings in imperative
clauses, viz. as marker of incompletive aspect and as politeness marker. Politeness
26 CLAUSE TYPES

468
levels will be treated below. In all other clause types the suffix <-khu> is not
polysemous and only indicates incompletive aspect. In imperative clauses the suffix
<-khu> (INCOM) can be followed by the imperative clausal enclitic <=bo> (IMP). In
those cases the context has to provide clues as to whether the incompletive suffix has
to be understood as politeness marker or as marker of incompletive aspect. The
imperative enclitic <=bo> (IMP) can optionally be followed by one of the following
other clausal enclitics, whereas the incompletive suffix <-khu> (INCOM) cannot.

a. affirmation seeking tag <=ne> (TAG)
b. emphatic positive <=ay> (POS)
c. emphatic imperative <=to ~ =ta> (IMPEMPH)
26.2.1 Politeness
Atong imperatives display three levels of politeness. These levels are marked by the
following forms from rude to most polite: the bare imperative, the imperative with the
imperative clausal enclitic <=bo> (IMP), and with the incompletive suffix <-khu>
(INCOM). They will be treated one by one below.
i The bare imperative
The bare imperative is a clause of which the predicate head consists only of the root
or the stem, (i.e. the root plus stem-forming suffixes). The bare imperative is the
impolite form of imperative. It can be used to children and close friends or family
when addressing younger kin or in emotional situations. Examples (685), (686) and
(687) are illustrations of bare imperative clauses. Example (687) comes from Text 2,
line 33. Line 35 in the same text is another example of the bare imperative.

26 CLAUSE TYPES

469
(685) re'e.
{re'e}
go.away
Go away!

(686) sa'ari.
sa' -ari
eat -SIMP
Just eat.

(687) hay sigoret honetsora na'a uaw.
[hay] [sigoret] {hon' -et -sora} [na'a] [u =aw]
come.on cigarette give -CAUS -TOTALLY 2s DST=ACC
Come on, give the cigarettes, oh you, those!

The event specifier -sora (TOTALLY) in (687) is used to give an emphatic effect to the
command.
ii The imperative with <=bo>
The imperative signalled by the imperative mood clausal enclitic <=bo> (IMP) is the
mid-level politeness form. It can be used to anybody, in any situation and can even be
very cordial or compelling depending on the tone of voice of the speaker. Examples of
this type of imperative can be found in abundance in Text 2, lines 4, 11, 13, 18, 20, 30
and 45. In order to emphasise the imperative, one of the discourse level suffixes
treated above can be added after the imperative enclitic, e.g. in Text 1 line 41 and in
Text 2, lines 4, 40, 42 and 43. The next example is a dialogue between a mother and
her sons. This example comes from a story in which six sons are leaving their
mothers house to shoot a giant eagle.

26 CLAUSE TYPES

470
(688) otokoymuna ama ni ue pholgomaw kawna re'enane nookno. ucie, om,
nemay re'eboay. bunduk ra'abo nookno.
[otokoymuna] [ama] [ni] [ue polgom] =aw {kaw} =na
so then mother 1pe DST eagle =ACC shoot =DAT
{re'e -na} =ne} {no -ok} =no ucie
go.away -DESI =TAG say -COS =QUOT then
[om] {nem} =ay {re'e} =bo =ay
yes good =ADV go.away =IMP =POS
[bunduk] {ra' -a} =bo {no -ok} =no
gun take -AWAY =IMP say -COS =QUOT
So then: Mother, we want to go to shoot that eagle, [they] said, it is said.
Then: Yes. Do go carefully. Take the guns. [she] said, it is said.
iii The imperative with <-khu>
An imperative ending in the incompletive suffix <-khu> (INCOM) signals a request
and is the most polite form of imperative in Atong. Examples (689) and (690) are
illustrative.

(689) baba, ana taka ratjani hon'khu.
[baba] [a] =na [taka ratja ni] {hon' -khu}
father 1s =DAT money hundred two give -INCOM
Father, please give me two hundred rupees.

The following example comes from the story of the hanging root that changes into an
old woman at night. When you happen to sleep under a hanging root, the old woman
will ask you to scratch her arm the whole night or else she will rip you open with her
long nails and eat you.

(690) ha' ambi a cakaw khenetkhu nookno.
[ha'] [ambi] [a cak] =aw {khen -et -khu} {no -ok} =no
interj grandchild 1s arm/hand =ACC scratch -CAUS -INCOM say -COS =QUOT
Hey grandchild, please scratch my arm. [she] said, it is said.

As mentioned above, the incompletive suffix <-khu> (INCOM) can be followed by
the imperative mood enclitic <=bo> (IMP). In that case the incompletive can be
understood either as politeness marker or as marker of the incompletive aspect.
26 CLAUSE TYPES

471
Example (691) below was uttered during a meal and is clearly a command to eat
more. So there the incompletive suffix functions as marker of the incompletive aspect.

(691) may sa'khubo
[may] {sa' -khu} =bo
rice eat -INCOM =IMP
Eat more rice.

Example (692) below comes from the story about a cunning man called Thengthon
[the?thon]. Thengthon is much hated in his village for being such a bad person. The
villagers burn down his house because of his bad behaviour. Thengthon collects the
ashes and cinders and goes to the market to sell them. The village people think hes
mad. Who will buy ashes and cinders? they say. On his way back home from the
market, Thengthon steals a large sum of money somewhere. Later he sits in front of
his house counting the coins which attracts the attention of the other villagers. When
Thengthon tells them that he sold all his ashes and cinders for this large sum of
money, the other people wonder if they cannot become rich too when they burn down
their houses and sell the remains on the market. Thengthon says they certainly can.
The result is that the other village people have found out that they have been conned,
they capture Thengthon and want to torture him. Then Thengthon utters (692),
which can be interpreted as a request or a polite request.
If we consider that it is the first time in the story that Thengthon asks for
forgiveness, it is a request and the incompletive signals politeness. But because
Thengthon has been bad in the past, which was why his house was burned down in
the first place, it might have occurred in the past that he has asked the other villagers
for forgiveness and so (692) can be interpreted as a request to forgive him again, in
which case the incompletive suffix signals incompletive aspect.

(692) na'tom angaw wetsado khema kha'khubo.
[na' -tom] [a] =aw [wet sa] =do {khema kha' -khu} =bo
2s -ppp 1s =ACC time one =TOP forgiveness do -INCOM =IMP
Please forgive me one more time. or Please forgive me this once.
26 CLAUSE TYPES

472
26.2.2 The prohibitive with <=bay>
The construction most frequently used to form prohibitives in Atong is formed with
the prohibitive clausal enclitic <=bay> (PROH) attached to a dative-marked clause.
The prohibitive clausal enclitic <=bay> (PROH) is homophonous with the intransitive
verb bay- to break.

It is very likely that the prohibitive under discussion
grammaticalised from a secondary verb construction (see Chapter 27) with the verb
bay- to break as main clause predicate and the action that is prohibited in the form of
a dative complement clause.
55
I analyse <=bay> (PROH) as prohibitive enclitic for
the following reasons:

No other constituent can intervene between the dative-marked verb and the
prohibitive morpheme and the position of the dative-marked verb is fixed
before this morpheme. If bay was considered a main clause predicate and the
dative-marked verb its complement, it would have been possible for
constituents to intervene between dative complement clause and the predicate
of the matrix clause (see Chapter 27).
The grammatical unity between the dative-marked verb and the prohibitive
morpheme is reinforced by the linking element /m/.
A prohibitive clause cannot take the imperative mood enclitic <=bo> (IMP).

The structure of a prohibitive is given in Table 73 below. We can see that the
inflectional possibilities of the prohibitive are more limited than those of the
imperative shown in Table 72 above. The change of state suffix <-ok> (COS) can be
attached to a prohibitive for emphasis. The result is a stronger prohibition. The change
of state suffix can have the same emphatic function on Type 1 adjectival predicate

55
The same construction has been described for Achik (Garo) by Burling (2004:127), except that there
seems to be no linking phoneme /m/ in Achik. The element na, which immediately precedes be in
Achik, is homophonous with the dative case marker, like in Atong, and be in the Achik prohibitive
(negative imperative) is also homophonous with the verb be- to break. Although the
grammaticalisation of the verb to stop into a marker of the prohibitive is well attested in languages
around the world (Heine and Kuteva 2002: 283-284), there are no other cases known to me of this
happening to the verb to break except for Atong and Achik.
26 CLAUSE TYPES

473
heads, as is discussed in 5.1. This means that the meaning of the change of state
suffix <-ok> (COS) varies depending on the clause type and the word class.

Table 73 The structure of the prohibitive with <=bay>
r
o
o
t

stem-forming
suffixes
clausal
enclitic
linking
element
clausal
enclitic
inflectional
suffix
clausal
suffix
<-et>
(CAUS)
<-khu>
(INCOM)
<=na>
(DAT)
/m/
<=bay>
(PROH)
<-ok>
(COS)
<=ne>
(TAG)
optional optional optional

Here below are some illustrative examples of the prohibitive with <=bay> (PROH)
We clearly see that the linking element /m/ is optional. Examples (693) and (694)
below come from the same paragraph in the same story. The prohibitive verbs in
(693) have the linking /m/, whereas the verb in (694) does not.

(693) otoknambay baba. otokoy takkhunambay.
{otok} =na -m- =bay [baba]
do.like.this =DAT -LINK- =PROH son
[otokoy] {tak -khu} =na -m- =bay}
like.this do -INCOM =DAT -LINK- =PROH
Dont do like this, son. Dont do like this any more.

(694) na'ba mokhasaba otokoy taknabay nooknoaro.
[na'] =ba [mokha] =sa =ba [otokoy] {tak} =na =bay
2s =EMPH face =MOB =EMPH like.this do =DAT =PROH
{no -ari -ok} =no
say -SIMP -COS =QUOT
Dont you do like this [any more] in the future, they just said, it is said.

The following example exhibits change of state-marked prohibitive forms indicating a
stronger prohibition. This constrution is comparable to the English V already
imperative, e.g. Stop already!, and to the Dutch and German imperative with the past
participle, e.g. Dutch: Afgelopen nu! [ofelope(n) ny] Stop now! (where afgelopen
is the past participle of the verb aflopen [oflope(n)] to finish).
26 CLAUSE TYPES

474
(695) te'ewdo morot so'otnam bayok sa'khawnambayok darinambayok.
[te'ew] =do [morot] {so'ot} =na -m- =bay -ok
now =TOP person kill =DAT -LINK- =PROH -COS
{sa' -khaw} =na -m- =bay -ok
eat -SECRETIVELY =DAT -LINK- =PROH -COS
{dari =na} -m- =bay -ok
behave.badly =DAT -LINK- =PROH -COS
Now dont kill people, dont steal, dont behave badly.

In the above example, the combination sa'-khaw (eat-SECRETIVELY) means to
steal in this case, and not to eat secretively.
Predicates of prohibitives can be modified by adverbial clauses and adverbs, as is
illustrated in examples (696) and (697) here below.

(696) ramci dolo khagaba gana. kha'sinay re'enabayne.
[ram} =ci [dolo {kha'} =gaba] {gana}
road =LOC bridge make =ATTR exist
{kha'sin} =ay {re'e} =na =bay =ne
slow =ADV go.away =DAT =PROH =TAG
There is a bridge on the road. Dont go slowly, ok.

(697) alaga morotna domdom damdam hon'na bay
[alaga morot] =na [domdamdamdam] {hon'} =na =bay
other person =DAT just.like.that give =DAT =PROH
Dont give it to someone else just like that.
26.2.3 The prohibitive with <ta>
The use of the prohibitive construction with ta (PROH) is much rarer than the
construction with the clausal enclitic <=bay> (PROH) (see above). Predicates in
prohibitiveconstructions cannot express aspect or modality. The word ta (PROH) is a
free morpheme with only one restriction: it has to precede the predicate. Thus the next
sentence has three variants with the same meaning; the only difference between them
is the position of the morpheme <ta> (PROH).

26 CLAUSE TYPES

475
(698) a: ta ie nok doda ham.
b: ie nok ta doda ham.
c: ie nok doda ta ham.

[ta] [ie nok] [doda] {ham}
PROH PRX house alone build
Dont build this house alone

To emphasise the imperative force of the utterance the imperative emphasiser clausal
enclitic <=to ~ =ta> (IMPEMPH) (allomorphs in free variation) can be cliticised to
the clause, e.g. (699). The result is a stronger prohibition. This enclitic is usually used
when the speaker is annoyed or impatient.

(699) na'a ta dokoroto!
[na'a] [ta] {dokoro} =to
2s PROH make.noise =IMPEMPH
You, dont make noise!

The prohibitive morpheme ta (PROH) can even be combined with the imperative
enclitic <=bo> (IMP) as the next example illustrates. In this example the prohibitive
morpheme has scope over both predicate heads.

(700) ta rophilrambo cikarakbo!
ta {rophil -ram} =bo {cikarak} =bo
PROH joke -try =IMP joke =IMP
Dont try to joke [or] be funny!

The predicate head of a prohibitive clause cannot take any inflectional suffixes, i.e.
cannot express modality, aspect and polarity by means of the suffix
<-ca> (NEG). (For the types of predicate suffixes see Table 63.) The predicate head
can, however, take stem-forming suffixes and the prohibitive clauses can take clausal
enclitics as is illustrated by the next example.

(701) oy ta so'dugasi, ado cokaydok.
[oy] [ta] {so' -duga} =si [a] =do {cok -aidok}
interj PROH ask -XS =MIR 1s =TOP cold -DUR
Oy! Dont ask too much, Im cold.
26 CLAUSE TYPES

476
26.2.4 The optative
A root plus the comparative suffix <-khal> (CP) signals the optative in Atong. In
other constructions the comparative suffix has a comparative function comparable to
the suffix er in English, e.g. nem-khal-a (good-CP-CUST) [it is/was] better. The
optative is only attested with a third person, both singular and plural, e.g. (702), and is
not attested in negated form.

(702) ge'the re'ekhal!
[ge'the] {re'e -khal}
3s go.away -CP
May he go away! or in French: Quil sen aille!
26.2.5 The hortative strategy
Atong has the proclause hay Come on!, Lets go! which can be used to imply a
hortative for first or second person singular or plural just by putting another clause
after it. This second clause will usually be a desiderative clause, as in (703), or
declarative clause with future-marked predicate, e.g. (704) or an imperative clause, of
which example (705) is illustrative. A simple fragment may suffice too, as in (706).

(703) dada, dada! hay na'nado, ama garu ramgabaci pholgom de'etay
tanagabaaw, pholgomaw kawna ray'na nookno.
[dada] [dada] [hay] [na'na] =do
elder.brother elder.brother come.on 1pi =TOP
[ama garu {ram} =gaba] =ci
mother mustard dry =ATTR =LOC
[pholgom {de'et} =ay {tan -a} =gaba] =aw [pholgom] =aw
eagle shit =ADV put -AWAY =ATTR =ACC eagle =ACC
{kaw} =na {ray' -na} {no -ok} =no
shoot =DAT go -DESI say -COS =QUOT
Older brother, older brother! Come on, us, [we] want to go to shoot the
eagle, the eagle which shat in mothers dry mustard [leaves], [he] said, it is
said.

The context for the next example is as follows. An animal says: The animals dont
have a king yet. In the whole world, in foreign countries, in India, in the Garo Hills,
there is no king, after which comes (704).
26 CLAUSE TYPES

477
(704) hay raja sonaka nowanowa.
[hay] [raja] {so -naka} {no -wa} =no -wa
come.on king elect -IFT say -FACT =QUOT -FACT
Come on, [we] will elect a king now, they said it is said.

In the following example a frog gets pinched by a river crab. His friend the deer urges
the frog to hit the river crab by uttering (705).

(705) hay, hay iaw tokboto tokboto tokboto!
[hay] [hay] [i] =aw {tok} =bo =to {tok} =bo =to
come.on come.on PRX =ACC beat =IMP =IMPEMPH beat =IMP =IMPEMPH
{tok}=bo =to
beat =IMP =IMPEMPH
Come on, come on, beat it, beat it, beat it!

The following example comes from a story in which a banana bird is on his way to
beat up an elephant for always destroying his nest and eating it. The bird meets a toad
and explains where he is going to. Then the toad says that the elephant also always
destroys his nest and joins the bird saying (706). This construction cannot be negated.

(706) hay aba
[hay] [a] =ba
come.on 1s =ADD
Come on, me too. Alternatively: Let me go too.
26.3 Declarative clauses and identity/equation clauses
The declarative clause is a type of independent clause that is used to express
statements and facts. The predicate head of a declarative clause can be marked for all
independent clause categories (see Table 63). Declarative clauses can take the
declarative clausal enclitic <=te> (DCL) or the mirative <=thay ~ =toy ~ =soy ~ =si>
(MIR) whereas interrogatives cannot.
Lines 19 the second clause, 28, 30 and 31 in Text 1 are illustrative of declarative
clauses. The first clause in line 19 of this text is a presentative clause which will be
treated in the next paragraph. The following example contains three declarative
clauses, two with a nominal predicate head and one with a verbal predicate head. A
26 CLAUSE TYPES

478
declarative clause with a noun as predicate head is interpreted either as a (verbless)
identity/equation/attributive clause. There is no formal distinction in Atong between
these three clause types; the term identity/equation clause is used in this grammar to
cover the three different interpretations. Example (707) comes from the story about
the giant eagle in which six sons are on their way to shoot it when they come to a
clearing in the jungle with a huge tree in it. In the tree the giant eagle is perching and
one of the brothers says (707).

(707) iannokhon, amami garu ramgabaci de'etgaba iankhonte ie. hay na'nado
kawnaka.
{i =an} =khon [[ama =mi garu {ram} =gaba] {de'et} =gaba}]
PRX =FC/ID =SPEC mother =GEN mustard dry =ATTR shit =ATTR
{i =an} =khon =te [ie] [hay] [na'na] =do {kaw -naka}
PRX =FC/ID =SPEC =DCL PRX come.on 1pi =TOP shoot -IFT
That might [be it], that might [be] the one which shat in mothers dry mustard
[leaves]. Come on, we will certainly shoot it.
Members of other non-verbal word classes, like demonstratives (see Chapter 1 and
Chapter 22), can also be predicate head of an identity/equation clause, e.g. (708).

(708) daba ian
[daba] {i =an}
coconut PRX =FC/ID
This [is] a coconut.
26.4 The presentative clause
A presentative clause is a declarative clause that contains only a nominal predicate
head. The properties of nominal predicate heads are treated in Chapter 22. Its function
is to present background information for what follows, as the first clause of line 19 in
26 CLAUSE TYPES

479
Text 1 and the first two clauses of example (709) below, which is the beginning of a
story.
56

(709) morot mo'sanoromo. jow'taraannokmo, wa' ni'ok. ue gawicie sa'
mo'korok gananoro aro de'the pipukci ganakhua mo'sa, mo.
{morot mo' sa} =no =ro =mo
person CLF:HUMANS one =QUOT =EMPH =CONF
{[jow' -tara] =an -ok} =mo [wa'] {ni' -ok}
mother -EXCLUSIVELY =FC/ID -COS =CONF father not.exist -COS
[ue gawi] =ci =e [sa' mo' korok] {gana} =no =ro
DST woman =LOC =FC child CLF:HUMANS six exist =QUOT =EMPH
aro [de'the pipuk] =ci {gana -khu -a} [mo' sa] =mo
and 3s belly =LOC exist INCOM -CUST CLF:HUMANS one =CONF
One person, OK. Only a mother [is left], OK. There is no father any more.
The woman has six children, it is said, and in her belly there is one more, OK.
26.5 Copula clauses
The identity/equation copula in Atong is do'-~ do- (IE.be). The allomorphs are in
free variation. A copula clause can be identified as such if it contains the
identity/equation copula don'- ~ do- (IE.be) and at least one core argument, viz. the
copula complement. Both copula subject (CS) and copula complement (CC) are
obligatorily unmarked for case. This obligatory unmarkedness of the two core
arguments is one feature which distinguishes copula clauses from transitive clauses in
which O can be accusative-marked under certain pragmatic conditions, (see 20.8).
The equation/identity copula is mostly used in emphatic (710) or contrastive (711)
circumstances or to express the sense of become (712). In non-emphatic and non-
contrastive circumstances the copula can be dropped in most cases, and then we have
a (verbless) identity/equation clause, treated above as declarative clause subtype.

56
It was interesting to note that some Atong speakers did not approve of presentative clauses in the
written language. When one of my friends read the book of compiled stories, he wanted to add a verb
to the first clause in example (709) so that it would read morot mo' sa gana=no=ro=mo (person
CLF:HUMANS one exist=QUOT=EMPH=CONF) [There] is one person, OK.

26 CLAUSE TYPES

480
Cases in which the copula cannot be dropped are constructions in which loanwords
are involved, as we will see below.

(710) sala, na'a moltete. ang iskon madam kam kha'phin'ok. nawmol dophinok.
[sala] [na'a] {mol -te -te} [a]
CS
{iskon madam
bastard 2s small -still.too -RED 1s so.much female.teacher
kam kha' -phin' -ok} [nawmol]
CC
{do -phin' -ok}

work make -FULLY -COS marriageable.girl IE.be -FULLY -COS
Bastard! you are still too small. I am already working as a teacher. I am
already a marriageable girl.

(711) ue soma sogni khocu badrido ato khu'cuk do'ca, ha'cok khu'cuk
do'sigaak.
[ue soma sogni khocu badri]
CS
=do [ato khu'cuk]
CC
{do' -ca}
DST Pname =TOP Atong language IE.be -NEG
[ha'cok khu'cuk]
CC
{do' -siga -ak}
Garo language IE.be -ALT -COS
That soma sogni khocu badri is not Atong, it is Garo.

(712) moma wa ni'wamian man'ay sa'cak, kagal do'ok..
[[moma wa] {ni' -wa} =mi =an {man'ay} {sa' -ca -k}
elephant tooth not.exist -FACT =GEN =FC/ID in.great.amounts eat -NEG -COS
[kagal]
CC
{do' ok}
poor.person IE.be -COS
The non-existence of the elephant tusk being the case, they did not have any
riches any more. They became poor persons.

In example (713) Ketketa Bura, who used to be a very skinny man, returns from the
house of his parents where he ate so much that he got fat. On his way he meets a fox
who recognises him and says that he is going to devour Ketketa Bura. But Ketketa
Bura simply lies and says I am not Ketketa Bura. Ketketa Bura is thin and I am fat as
you can well see.

(713) aa ketketa bura do'ca.
[aa]
CS
[ketketa bura]
CC
{do' -ca}

1s Name IE.be -NEG
I am not Ketketa Bura.
26 CLAUSE TYPES

481
As mentioned above, the identity/equation copula is mostly used in emphatic or
contrastive situations and does not need to be expressed under normal conditions
because nominal predicate heads can also express negation, aspect and modality and
can also take dependent clause marking. The copula is, however, obligatory when it is
used as support verb in complex predicates with certain loan words as treated in
22.7.2iii.
26.6 Quotative clauses
Quotative clauses are marked by the quotative enclitic <=no> (QUOT). This enclitic
is a marker of evidentiality that indicates that the speaker has the information from
hearsay. The quotative enclitic is historically derived from the verb no to say. Older
speakers often pronounce the enclitic with two syllables, [noa ~ noo], or an off-glide
[a ] or [e ], [noa ~ =noe ], but younger speakers do not. Older speakers indicated that
this second syllable or off glide is in fact the factitive suffix <-wa> (FACT), which is
a predicate head suffix (see Chapter 23). This means that in the language of older
speakers the grammaticalisation process from verb to clausal enclitic has not yet been
fully completed, since they can still attach a factitive-like element to the quotative.
Younger speakers who helped me to transcribe texts never wanted to write anything
else but /no/ for the quotative, even when I insisted that I heard something else in the
recording. For the younger speakers the process of grammaticalisation of the verb into
an enclitic is complete, and all traces of verbness have vanished. Examples of
quotative clauses can be found throughout this grammar.
57

57
I use the term quotative to label this hearsay evidential morpheme so as to keep in line with the
terminology used in other grammars of the gourp of languages that Atong belongs to, viz. the grammar
of Garo by Burling (2004) and the grammar of Rabha by Joseph (2007).
26 CLAUSE TYPES

482
26.7 Reactions to independent clauses involving proclauses
Atong has all three systems treated in Sadock and Zwicky (1985: 189-91) that involve
proclauses ho'o yes and hm'm no and om affirmative which are reactions to
independent clauses, viz.

the agree/disagree system,
the yes/no system,
the echo system.

Systems 1 to 3 can be used to react to all types of independent clauses except that the
echo system does not apply to imperatives. For imperative clauses the proclause om
affirmative must be used as positive reaction or agreement proclause. The echo
system is freely interchangeable with either the agree/disagree or the yes/no system.
The proclauses and echo answers can be combined, the most frequently used one is
PRO-CLAUSE followed by PREDICATE. All reactions will be treated separately
below.
26.7.1 The agree/disagree system
Negative questions and statements are answered according to the agree/disagree
system in which the affirmative proclause ho'o yes means that the presupposition
of his interlocutor was correct, e.g. line 26 and 27 in Text 1. The negative proclause
hm'm no means that the speaker disagrees with the interlocutors presupposition,
e.g. (714) and (715).

(714) may sa'khuca? hm'm.
{may sa' -khu -ca} [hm'm]
rice eat -INCOM -NEG no
Havent [you] eaten yet? Yes [I have].

26 CLAUSE TYPES

483
(715) ato dowa ama? nookno. ho? ni'wate baba, aa mamuawan dowancate.
uan phowramu hay'e garutara dowariwate. mamu dowancate nookno. ucie:
hm'm ama na'do tay'nido atoba dowa. ja'bekan thawokte.
[ato] {dow -wa} [ama] {no -ok} =no [ho'] {ni' -wa} =te
what add -FACT mother say -COS =QUOT interj not.exist -FACT =DCL
[baba] [aa] [mamu] =aw =an {dow -an -ca} =te [u] =an
son 1s nothing =ACC =FC/ID add -REF -NEG -DCL DST=FC/ID
[phowra] =mu [hay'e garu] =tara {dow -ari -wa} =te
rice.powder =COM GPN mustard =EXCLUSIVELY add -SIMP -FACT =DCL
[mamu] {dow -an -ca} =te {no -ok} =no [ucie] [hmm] [ama]
nothing add -REF -NEG -DCL say -COS -QUPOT then no mother
[na'] =do [tay'ni] =do [ato] =ba {dow -wa} [ja'bek] =an
2s =TOP today =TOP what =INDEF add -FACT curry =FC/ID
{thaw -ok} =te
tasty -COS =DCL
What did you add, mother?, [he] said, it is said. Huh?! Nothing, son, I did
not add anything, Im telling you. I only added that rice powder and these, um,
mustard [leaves], Im telling you. I added nothing, really, [she] said, it is said.
Then: No, mother, you did add something today. The curry is very tasty,
really.

Prohibitives require om affirmative as agreement reaction and hm'm no as
disagreement reaction proclause, e.g. (716). Other examples of the use of the
proclause om affirmative can be found in 17.4.

(716) re'enabay! hm'm / om
{re'e} =na =bay
go.away =DAT =PROH
Answer A: [hm'm]
no
Answer B: [om]
affirmative
Dont go! Answer A: Yes, [I will go]. Answer B: No, [I wont go].
26.7.2 The yes/no system
Positive questions and statements can be answered with the yes/no system. In this
case the positive proclauses ho'o and 'm. hm' thats right mean affirmation of the
interlocutors presupposition, e.g. (717) below. The negative hm'm no contradicts
26 CLAUSE TYPES

484
the interlocutors presupposition, of which example (718) and (719) below are
illustrative and as can be seen line 68 of Text 2, which is the negative answer to the
positive question in line 65.
58

(717) tot! di'phuram'ama? nookno. ho'o manamaydomo noatakokno,
phalthaba, amakba.
[tot] {di'phu -ram -a} =ma {no -ok} =no
interj fart -INADVERTENTLY -FACT =Q say -COS =QUOT
[ho'o] {manam -aydo} =mo
yes stink -PROG =CONF
{no -wa} {tak -ok} =no [phaltha] =ba [amak] =ba
say -FACT do -COS =QUOT self =EMPH monkey =EMPH
Hey! Did you just accidentally fart?, he said, it is said. Yes, it stinks,
doesnt it, [he] said, it is said, [he] himself, the monkey.

Text 2 provides ample examples of yes/no-system answers. In lines 9 and 10 we
find an example of a marked polar question followed by a proclause answer. Line 22
spoken by Jongken in the same text is interpreted as a polar question by Songken,
who answers with a proclause. Note that the intonation does not mark line 22 as a
polar question. The same holds for the polar question in line 30. In line 46 and 47 we
find a presupposition in the form of a declarative clause answered by the affirmative
proclause 'mhm' thats right. Line 51 presents a proclausal answer to the rhetorical
content question in line 50.

58
Note that in this question {may sa'-thok-ok}=ma [na'-tom]=e? (rice eat-ALL-COS=Q 2s-ppp=FC)
Have all of you eaten rice? the word may rice is incorporated in the predicate since the quantifying
S/O event specifier suffix -thok all refers to na'tom you.plural, in S function, and not to may rice.
For more detail see Chapter 22.
26 CLAUSE TYPES

485
(718) pholgom de'etdapetanawa?! nookno. ho'o nookno.
[pholgom] {de'et -dap -et} {tan -a -wa} {no -ok} =no
eagle shit -ON.TOP -CAUS put -AWAY -FACT say -COS =QUOT
[ho'o] {no -ok} =no
yes say -COS =QUOT
An eagle shat on [it]?! he said, it is said. Yes [she] said, it is said.

(719) hay co'sa sa'na na'a bay'siga. hm'm man'ca noariano.
[hay] [co'sa] {sa' -na} [bay'siga]
come.on a.little eat -DESI friend
[hm'm] {man' -ca} {no -ari -a} =no
no be.able -NEG say -SIMP -CUST =QUOT
Come on, [I] want to eat a bit, friend. No, [you] cant, [he] just said, it is
said.

Imperative clauses require the proclause om affirmative as positive answer and
hm'm no as negative answer, e.g. (720) here below. Other examples of the use of the
proclause om affirmative can be found in 17.4.

(720) uci nepaldo: na'a a ma'su ma rajasaaw tonasegabone nookno. om
noaimo the'thonba tonaokno.
uci [nepal] =do [na'a] [a ma'su ma raja sa] =aw
then Nepali =TOP 2s 1s cow CLF:ANIMALS hundred one =ACC
{ton -a -sega} =bo =ne {no -ok} =no
lead -AWAY -ALT =IMP =TAG s ay -COS=QUOT
[om] {no }=ay =mo [the'thon} =ba {ton -a -ok} =no
affirmative say =ADV =SEQ Name =EMPH lead -AWAY -COS =QUOT
Then the Nepali: You lead my hundred cows away [instead of me], ok, [he]
said, it is said. Having said: Yes [I will], Thengthon led them away, it is
said.
26.7.3 The echo system
The echo system functions in answers to polar questions where the predicate of the
interrogative clause is repeated in the desired positive or negative form, e.g. (721) and
(722).

26 CLAUSE TYPES

486
(721) may sa'akma? sa'ak. / sa'khuca
{may sa' -ak} =ma
rice eat -COS =Q
Answer A: {sa' -ak}
eat -COS
Answer B: {sa' -khu -ca}
eat -INCOM -NEG
Have you eaten? Literally: have you rice-eaten? Answer A: Yes
Answer B: Not yet.

(722) kha'rekma ie? kha'rekanca, saman.
{kha'rek} =ma [ie] {kha'rek -an -ca} {sam} =an
yardlong.bean =Q PRX yardlong.bean -REF -NEG weed =FC/ID
Is this a string bean [plant]? [They are] not beans, [They are] weeds.

If the predicate of the polar question was marked with <=ma> (Q), the predicate of the
answer will frequently be marked by the emphatic marker <=ba> (EMPH), as in the
greeting and answer expressed in (723). This emphatic marking is not obligatory. The
enclitic <=ba> (EMPH) can also occur on predicates when the polar question was not
marked by <=ma> (Q).

(723) nemaydoama? nemaydoaba.
{nem -aydoa}=ma {nem -aydoa} =ba
good -DUR =Q good -DUR =EMPH
Are you well? Well indeed.

The echo clause can be an emphatic complex predicate as is the case in line 67 of,
Text 1, which is the reaction to the marked polar question in line 65.
We will now turn our attention to two clausal enclitics in Atong, viz. the irrealis
and the speculative. These enclitics do not indicate different clause types, but only
occur on independent declarative clauses.
26.8 The irrealis enclitic <=cm>
The irrealis only occurs on independent declarative clauses and is signalled by the
morpheme <=com> (IRR) and indicates that something is supposedly the case
(epistemic modality connotation) or that a certain event was going on but was
26 CLAUSE TYPES

487
stopped, which is the irresultative interpretation. Depending on the context the use
of the morpheme <=com> (IRR) can also have a frustrative implication, or it can
mean that an event could have been otherwise, i.e. the implicative interpretation. All
these interpretations will be treated one by one.
The use of the irrealis is independent of that of the predicate head suffixes
indicating aspect and modality. The irrealis also appears as a particle after the
proclause ho'o yes (see 17.4), as we can see in line 55 of TEXT 1, which is
represented here with its context as (724). The comma indicates the pause that usually
precedes the irrealis in this construction. The speculative clausal enclitic <=khon>
(SPEC) (see 26.9) and the confirmative tag enclitic <=mo> (CONF), e.g. TEXT 2,
line 23 and (520), are also attested on this proclause.

(724) Joken says:
{banthay} =ci =ba [taka poysa] [na'] {na -arok} =ona
bachelor =LOC =INDEF money money 2s need -PROG =DAT
when you are a bachelor, because you need money'
Nongken says:
ho'o, com.
yes IRR
Yes, supposedly
26.8.1 Supposition interpretation
In examples (725), (726) and (727) the irrealis functions as a modality marker,
indicating that something is supposedly the case. The first two examples are the
opening sentences of stories, creating a fictional character of whom an adventure will
be told, the third example is a simple supposition. When a clause is both irrealis- and
quotative-marked, the irrealis usually comes after the quotative enclitic, as we see in
(714). Less usual is the position of the irrealis before the quotative, as we see in (715).
When the clause is marked for emphasis with the emphatic enclitic <=aro ~ =ro>
(EMPH) in addition to the irrealis and quotative enclitics, the order of the enclitics is
always =IRREALIS=QUOTATIVE=EMPHATIC, e.g. (716). The different positions
of the quotative have no effect on scope of the enclitic, or on the meaning of the
clause.

26 CLAUSE TYPES

488
Irrealis after quotative:
(725) so damsaci morot mo'sa man'ay sa'gaba gananocom.
[so dam sa] =ci [morot mo' sa
village CLF:VILLAGES one =LOC person CLF.HUMANS one
man' =ay sa' =gaba] {gana} =no =cm
in.great.amounts =ADV eat =ATTR exist =QUOT =IRR
In a village is/was supposedly a man who ate in great amounts (a rich man), it
is said.

Irrealis before quotative:
(726) ue bihape cigacak te'ew kol india kolani hapan do'wacomnoa.
[ue bihap (<Garo) ] =e [cigacak] te'ew

DST place =FC Pname now
[kol india kolani hap]

=an

{do' -wa} =cm =noa
Pname place =FC/ID I E.be -FACT =IRR =QUOT
The place Chigachak is now supposed to be the Coal India Colony place.

Irrealis preceeds quotative followed by emphatic:
(727) Te'ewe biba so damsacie boba mosa ganacomnoro.
[te'ew] =e [biba so dam sa] =ci =e
now =FC in.whatever.place village CLF:VILLAGES one =LOC =FC
[boba mo' sa] {gana} =cm =no =ro
crazy.man CLF.HUMANS one exist =IRR =QUOT =EMPH
Now, in a village wherever supposedly is/was a lazy king, it is said.
26.8.2 Irresultative interpretation
In example (728) we see the irrealis on a clause of which the predicate carries the
change of state suffix. A certain event had come about during a game of cards, but the
event did not continue. The discontinuity of the event is signalled by the irrealis.

(728) a tay'sa rajasa lapokcom. thoyok.
[a] [tay'sa] [raja sa] {lap -ok} =cm {thoy -ok}
1s just.now hundred one make.profit -COS =IRR die -COS
I just made one hundred [rupees] profit. I died (i.e. lost the game).
Alternatively: I just won a hundred rupees but I lost at the end.

The sentence in (729) was uttered when a young woman came into the room to get a
small stool which had been there earlier, which she had seen. At the moment of
26 CLAUSE TYPES

489
utterance the small stool had been removed. Just like in the example above, the
irrealis is used to signal the discontinuation of an event, in this case the existence of
the small stool.

(729) mura tay'sa ganacom, te'ew ni'wa.
[mura] tay'sa {gana} =cm te'ew {ni' -wa}
small.stool just now exist =IRR now not.exist -FACT
There was supposed to be a small stool here [but] now its gone.

Discontinuation of a negated event can also occur. This means that something was
once not the case, but is the case now. The next example is illustrative.

(730) a ie khata dakado togcacom. te'ewdo nemayan took
[a] [ie khata] daka =do {to -ca} =cm
1s PRX word before =TOP know -NEG =IRR
[te'ew] =do {nem} =ay =an {to -ok}
now =TOP good =ADV =FC/ID know -COS
I did not know this word before. Now I know it well.
26.8.3 Frustrative interpretation
I adopt the definition of frustrative as used in Aikhenvald (2003: 280) for the
frustrative enclitic -tha in Tariana. This definition states that the frustrative is used
when an action has failed already or is bound to fail; or that the success of an
attempted action is not yet certain. The frustrative interpretation of the irrealis
morpheme <=com> (IRR) is illustrated in the next examples. In the context of (731), a
small child is walking through the forest in search of his brothers, but he cannot find
them anywhere. When he meets an old woman, the child says:

(731) abu, ado dadaparaaw sandiedoacom.
abu [a] =do [dada] =para =aw {sandi -edoa} =cm
grandma 1s =TOP older.brother =&co =ACC search -PROG =IRR
Grandma, I am searching in vain for my elder brothers.
While in the above example the event is ongoing but frustrated, because it is in vain,
the event in the following example has failed already.
26 CLAUSE TYPES

490
(732) na'awdo boro tay'sa sandiwacom.
[na'] =aw =do [boro] [tay'sa] {sandi -wa} =com
2s =ACC =TOP Name just.now search -FACT =IRR
Boro just searched you in vain.
26.8.4 Implicative interpreation
The morpheme <=com> (IRR) also signals another kind of irrealis, viz. that
something could be or could have been otherwise. This is the implicative
interpretation, illustrated by examples (733) - (735).

(733) biskut dotha ramacom
[biscut dotha] {ram -wa} =cm
biscuit different search -FACT =IRR
You could have searched for other biscuits. (said the mother to her daughter
who came in with some biscuits which were apparently not the ones the
mother wanted.)

In the next example both occurrences of the irrealis can be interpreted as signaling an
implication.

(734) a na'aw boloen nukna sokaydoacom. a phaltha re'enado
sokaydokcom, otokciba aa sawamigomon re'ena man'caaydoa.
[a] [na'] =aw [boloen]{nuk}=a {sok -aydoa} =cm
1s 2s =ACC very see =DAT want -PROG =RR
[a phaltha] {re'e} =na =do {sok -aydok} =cm
1s self go.away =DAT =TOP want -PROG =IRR
otokciba [aa] [sa -wa =mi =gomon] {re'e} =na
but 1s be.ill -fact =GEN =REASON go.away =DAT
{man' -ca -aydoa}
be.able -NEG -PROG
I am really wanting to see you, (but something prevents me from attaining
this). I would have liked to come myself but because of me being ill I am not
able to go.

26 CLAUSE TYPES

491
(735) ha'coksa balcido sal kolgroksa noay monicom.
[ha'cok] =sa {bal} =ci =do [sal kolgrok sa]
Garo =INSTR say =LOC =TOP day twenty one
{no} =ay {mo -ni} =cm
say =ADV call.a.name -FUT =IRR
When you say it in Garo, it would be called sal kolgroksa (twenty one
days).

The next example is a comment from a speaker whom I had just told that if Atong
people would go to my country (Holland) they would not become white like me.
My interlocutor did not agree and said:

(736) boknicomdo.
{bok -ni} =com =do
white -FUT =IRR =TOP
[We] might become white anyway.

I am not sure whether to interpret this use of the irrealis as an implicative or a
supposition.
26.9 The speculative enclitic <=khon>
The speculative enclitic <=khon> (SPEC) can only be encliticised to independent
declarative clauses, and expresses speculative modality, which indicates that the
speaker is uncertain and speculating about the coming about of an event which he did
not witness. The occurrence of the speculative modality clausal enclitic is independent
of the occurrence of any other aspect or modality predicate head suffixes. When used
in combination with the quotative enclitic, the speculative enclitic always follows the
quotative, e.g. (737). Clauses where the speculative enclitic precedes the quotative are
not attested. In colloquial speech the speculative marker is most often followed by the
declarative enclitic<=te> (DCL) whereas this combination is rare in narrative texts.
The position of the speculative enclitic in relation to the quotative is purely
morphosyntactically conditioned and does not change the scope of the enclitics nor
the meaning of the clause.
26 CLAUSE TYPES

492
The following example illustrates the use of the speculative enclitic in a narrative
about the past. The speaker tells us that a family with some children wanted to swim
across a dangerous river. Why they wanted to swim across is speculated upon in
(737).

(737) ucie te'ewdo ru ni'wanokhon.
ucie [te'ew] =do [ru] {ni' -wa} =no =khon
then now =TOP boat not.exist -FACT =QUOT =SPEC
Then, well, there might not have been any boats, it is said.

Example (60), repeated here as (738), is a speculation about a future event, as can be
seen by the appearance of the future modality predicate head suffix <-ni> (FUT) in
the second, underlined clause.

(738) rasan raboromaydoa, waynikhon.
[rasan] {raborom -aydoa} {wa -ay -ni} =khon
sun be.shrouded.in.clouds -DUR rain -TOWARDS -FUT =SPEC
The sun is blocked by the clouds, it might rain.

Examples of the speculative on clauses with predicates marked by other aspect and
modality suffixes can be found in line 22 of TEXT 2, with the change of state suffix,
line 46 of TEXT 2, with the incompletive and the negative, and in example (307) in
17.4, with the imperious future. Example (116) in 8.3 illustrates the occurrence of
the speculative enclitic on a clause with a demonstrative as predicate head. The
following example shows the speculative enclitic on a clause with a nominal predicate
head. While walking through the jungle, one of my friends heard a sound and said:

(739) mumakhonte
{muma} =khon =te
elephant =SPEC =DCL
It might be an elephant.

Like the irrealis <=com> (IRR), see 26.8, and the confirmative tag enclitic <=mo>
(CONF), e.g. TEXT 2, line 23 and (520), the speculative enclitic occurs as a particle
26 CLAUSE TYPES

493
after the proclause ho'o (see 17.4), as the next example illustrates. The comma
indicates the pause which usually precedes the occurrence of the particles in this
construction.

(740) Speaker 1: na ama nogolsa re'ewama?
Speaker 2: ho'o, khon.
[na ama] [nogol] =sa {re'e -wa} =ma
2s mother market =MOB go.away -FACT =Q
ho'o, khon
yes SPEC
Has your mother gone to the market? Yes, possibly.

494
Chapter 27 Dative- and locative-marked
clauses
_____________________________________________________________________

The dependent status of dative- and locative-marked clauses is signalled by case
marker morphemes.
59
These clauses can fulfil a number of semantic roles in a matrix
clause. Dative marked clauses range in syntactic status from adjuncts to complements
whereas locative-marked clauses only occur as adjuncts, modifying a matrix clause.
There are two kinds of predicates, according to formal criteria, on which the
dative and locative cases occur, viz. inflected and non-inflected. The variety of
different predicate inflections is bigger for dative clauses than for locative clauses and
all of these are main clause inflections, i.e. without the case marker, the predicate
could stand as a sentence on its own. The dative enclitic can occur after the change of
state suffix, the progressive/durative aspect or the factitive suffix, as we will see in
27.1, while the locative enclitic can only occur after the factitive suffix, as will be
discussed in 27.5. The factitive suffix on locative-marked clauses has the same effect
as the factitive on main clauses, i.e. as marker of factitive modality, as is discussed in
Chapter 1.
Dative marking on non-inflected clauses is discussed in 27.2. The dative case
enclitic also marks complement clauses of the temporal postposition daka before,
as we will see in 27.3. Section 27.4 summarises the most important properties of
dative-marked clauses. Everything about locative-marked clauses will be discussed in
27.5. Finally, 27.6 is a description of the different functions of the concomitant
action suffix <-butu> (WHILE).

59
Atong is not the only language that uses case marking as a clause linkage device. A list of languages
from around the world that also exhibit this phenomenon is given in Aikhenvald (forthcoming 2009).
27 DATIVE- AND LOCATIVE-MARKED CLAUSES

495
27.1 Dative marking on inflected predicates
Predicate heads inflected with either the factitive suffix <-wa> (FACT), the change of
state suffix <-ok ~ -ak ~ -k> (COS) or the progressive/durative suffix <-aydoa ~ -
aydo ~ -aydok ~ -aroa ~ -aro ~ -arok> (PROG) can take the dative enclitic <=na>
(DAT) to fulfil an adjunct role in a matrix clause. The factitive, change of state and
progressive/durative are all main clause suffixes, i.e. without the dative case enclitic,
predicates with these suffixes can stand as sentences on their own; the dative enclitic
is the marker of subordination.
Change of state-plus-dative- and progressive/durative-plus-dative-marked
predicate heads can only be interpreted as Reason adjuncts. Factitive-plus-dative-
marked predicates can be either Reason adjunct, Standard of comparison or O
argument, depending on the main clause predicate and the context. Factitive-plus-
dative marked complement clauses in O function are treated in Chapter 1.
There are no co-reference restrictions between the arguments of an inflected
dative clause and a main clause. Case marking of NPs in dative-marked subordinate
clauses is the same as in main clauses, i.e. A and S unmarked for case and O can be
optionally accusative-marked when the NP is referential and definite (see 20.8).
Argument structure in dative-marked and main clauses is also the same. Reason
clauses and standard of comparison clauses are treated separately below.
27.1.1 Reason clauses
A Reason clause is a dative-marked subordinate clause that is syntactically an adjunct
to a matrix clause. The allomorph <=ona> occurs after the change of state suffix
<-ok ~ -ak ~ -k> (COS) and after the allomorphs <-aydok ~ -arok> (PROG) of the
progressive/ durative aspect suffix.
60
The allomorph <=na> occurs elsewhere.

60
The appearance of the homophonous desiderative suffix <-na> (DESI) on independent clauses
indicating an implied possibility, a wish or intention is treated in 23.9. See also example (762) in this
chapter.

496
Examples (741) and (742) illustrate Reason clauses whose verbal predicate head is
inflected with the factitive suffix <-wa> (FACT) and the change of state suffix
<-ok> (COS) respectively. In example (741) we see that the subordinate clause
headed by the transitive predicate sa'rowana has a fully fledged argument structure.
The NPs inside the subordinate clause are marked in the same way as in a main
clause, viz. the O argument ge'themo thup his nest is marked as accusative by the
enclitic <=aw> (ACC) and the A argument moma elephant is unmarked for case.

(741) taw'reksorup masa ge'themo thupaw phanan moma phay'ay
sa'rowana, moma mathayaw thapna re'eaydoanowa.
[taw'reksorup ma sa] |[ge'the =mo thup]
O
=aw [phanan]
babana.bird CLF:ANIMALS one 3s =GEN nest =ACC always
[moma]
A
{phay'} =ay {sa' -ro -wa}| =na [moma
elephant break =ADV eat -USUALLY -FACT =DAT elephant
mathay] =aw {thap} =na {re'e -aydoa} =no -wa
bachelor.elephant =ACC beat.up =DAT go.away -PROG =QUOT -FACT
A banana bird, because an elephant always breaks/broke and eats/ate his nest
[lit. breakingly eats], is/was going to beat up the bachelor elephant, it is
said.

In example (742) the S argument of the intransitive subordinate predicate thawokona
is unmarked for case just like in a main clause.

(742) ido di'an thawokona, randaydo atongtokoy thawarongnaka, mo ama?!
|[i] =do [di']
S
=an {thaw -ok}| =ona [randay] =do [ato] =tokoy
PRX =TOP shit =FC/ID tasty -COS =DAT meat =TOP how =LIKE
{thaw -aro -naka} [mo] [ama]
tasty -DUR -IFT CONF mother
Because this, the shit, is so tasty, how tasty will the meat certainly be, wont
it, mother?!

497
Nominal predicate heads of Reason clauses can also either be change of state-
marked, e.g. (743), (744)
61
or factitive-marked, e.g. (745). Factitive-marked nominal
predicate heads are rare in Atong. To express the factitive category instead of the
nominal, the copula can be used as predicate, as we can see in example (746).
The context of the next example is as follows. The lazy king, having become a
tramp, is at the market and wants to buy a bullock cart because he is too lazy to walk.
He tries to find the cheapest cart so he asks several salesmen. Then he returns to the
first salesman and asks again how much the cart costs and the salesman replies (743).
The dative-marked nominal predicate which is the focus in this example consists of
the round number numeral hajar 1000 and the unit numeral sa one followed by the
suffix <-ak> (COS) and the dative morpheme. The head noun and classifier are elided
since they are understood from the context, i.e. the predicate is a headless NP.

(743) cewamo damdo hajarsaakona, hajarsa.
|[ce -wa =mo dam]
S
=do {hajar sa -ak}| =ona {hajar sa}
begin -FACT =GEN price =TOP thousand one -COS =DAT thousand one
Because the first price was one thousand, (it is still) one thousand.

Whereas the change of state-marked nominal predicate head of the Reason clause in
(743) consists of a numeral, the following example shows a prototypical noun as
change of state-marked head of a Reason clause.

(744) balphakram ha'borigumukokona, ray'sotna man'ca.
|[balphakram] {[ha'bori] =gumuk -ok}| =ona
Pname hill =all -COS =DAT
|{ray' -sot}| =na {man' -ca}
go -DIRECTLY =DAT be.able -NEG
Because Balphakram is all hills, you cannot go [there] directly.

61
Other examples of change of state-marked Reason clauses are (766) and line 32 and 33 of the Text 1
and line 50 of the text Text 2.

498
The following example illustrates the use of the factitive and dative enclitics on a
nominal predicate head.

(745) sa'goraywana komcawa.
|{sa'goray -wa}| =na {kom -ca -wa}
child -FACT =DAT marry -NEG -FACT
Because shes a child, I will not marry her.

The factitive marks the noun sa'goray child as predicate head. Without the factitive
on sa'goray child this noun cannot be interpreted as Reason adjunct since only
clauses can fulfil this semantic role. The sentence sa'goray=na kom-ca-wa
(child=DAT marry-NEG-FACT) would mean I will not marry for (the benefit of) the
child). The dative is interpreted as the marker of a Beneficiary. For a more detailed
discussion, see 24.6.
More fieldwork is needed to find out what the difference in meaning is between a
factitive-marked and a change of state-marked nominal predicate head of a Reason
clause.
Example (746) below illustrates the use of the copula verb do'- taking the
factitive suffix <-wa> (FACT) instead of the noun in copula complement function.
The copula is used for emphasis in this example, instead of suffixing the factitive
directly to the nominal predicate head as in (745).

(746) a pi'saci amapara babapara khagal do'wana []
|[a pi'sa] =ci [ama]
CS
=para [baba]
CS
=para
1s childhood =LOC mother =&co father =&co
[khagal]
CC
{do' -wa}| =na
poor.person EXIST -FACT =DAT
In my childhood, because my mother [and] father and those associated with
them were poor people []

The only clausal enclitic that is attested on a factitive-marked Reason clause is the
delimitative enclitic <=sa> (DLIM). The example below is illustrative.


499
(747) sa'khawca na'a. ni'wa na'a, go'wanasa balwa sakay mu'aro.
{sa' -khaw -ca} [na'a] {ni' -wa} [na'a]
eat -SECRETLY -NEG 2s not.exist -FACT 2s
|{go' -wa}| =na =sa [balwa] {sak =ay} {mu' -aro}
be.willing -FACT =DAT =DLIM wind enjoy =ADV sit -PROG
[I] did not steal, oh you! [Theres] nothing, oh you! Only because [I] want to
[I]m sitting [here] enjoying the wind.

A Reason clause can be negated, as the next example illustrates.

(748) ra sokcawanasa te'ewcinakhokho ramu cow rosusawanasa mokha badri
bimu mewanowa.
|[ra] {sok -ca -wa}| =na =sa [te'ew] =ci =na [khokho]
rain succeed -NEG -FACT =DAT =DLIM now =LOC =ALL still
|[ra] =mu {cow ro -susa -wa}| =na =sa
rain =COM liquor drink =COMPETITIVELY -FACT =DAT =DLIM
[mokha badri] [bimu] {mo -wa} =no -wa
Pname name call.a.name -FACT =QUOT -FACT
Because the rain did not succeed, still up till now, precisely because [they]
competitively drank with the rain, the name is called Mykha Badri, it is said.

In Text 2, line 3, presented as example (749) below, we see a Reason clause as an
afterthought postposed to the predicate of the main clause and separated from the
main clause by a pause.

(749) o, golgolaro otoken, haratwanasa.
[o] {golgol -aro} [otoken], |{harat -wa}| =na =sa
interj roam -PROG just.like.this lazy -FACT =PUR =DLIM
Oh, [I]m just roaming like this, just because [I]m lazy.
27.1.2 The standard of comparison clause
Events, expressed by verbs, can be compared in Atong just as objects, expressed by
nouns, can be compared. To compare events, the dative case is attached to the
factitive-marked predicate of the clause which functions as Standard of comparison
(see Dixon 2006-c). There are no co-reference restrictions between the arguments of
the standard of comparison clause and the main clause.

500
The dative case functions as mark of the Standard of comparison clause just as it
does on Standard of comparison NPs. The next example shows the dative enclitic
marking the NP abun so other village as Standard of comparison (see also 20.6
(A)), whereas in (751) and (752) the dative marks a whole clause as such.

(750) abun sonna dayay ie son hansenkhala.
STANDARD ------------MARK---------- COMPAREE PARAMETER INDEX
[abun so] =na {day} =ay [ie so] {hanse -khal -a}
other village =DAT be.bigger=ADV PRX village beautiful -CP -CUST
This village is more beautiful than the other village.

In (653), here repeated as (751), the arguments of the Standard of comparison clause
are co-referential with those in the Comparee clause. In (752) the arguments in both
clauses are different.

(751) umigomonci aa na'aw khomana dayaydo asetwaan nemkhalnaka.
STANDARD ------------MARK----------
[umigomonci] [aa]
A
|[na'] =aw {khom -a
62
}| =na {day} =ay| =do
therefore 1s 2s =ACC marry -FACT =DAT be.bigger=ADV =TOP
COMPAREE PARAMETER INDEX
{asset -wa} =an {nem -khal -naka}
throw.away/dispose -FACT =FC/ID good -CP -IFT
Therefore it will certainly be much better to throw you away than to be
married to you. Literally: Therefore, than being married to you, throwing
you away is biggerly better.

It is not possible to determine whether the A argument, aa (1s), in (751) belongs to
the dative-marked clause or to the main clause. Therefore alternative bracketing for
the dative clause in (751) would be: |[aa]
A
[na'=aw] {khom-wa=na}|.

62
The form of the factitive suffix in this example is <-a> due to the phonological rule that the /w/
elides when the verbal root or stem ends in /m/ or /p/ (see Chapter 2).

501
(752) khaw kan'wana dayaydo na'nae ici cayay mu'waan ga'sukhalnaka.
STANDARD ------------MARK----------- COMPAREE
|[khaw] {kan' -wa}| =na {day} =ay =do [[na'na] =e
hair cut -FACT =DAT be.bigger=ADV =TOP 1pi =FC
PARAMETER INDEX
[i] =ci {cay} =ay {mu' -wa}] =an {ga'su -khal -naka}
PRX =LOC look =ADV sit/stay -FACT =FC splendid -CP -IFT
It is better that we sit here and watch than that you cut hair. Literally: Than
cutting hair, us watchingly sit here will certainly be more splendid.
27.2 Dative marking on verbal roots or stems
As we saw in the previous sections, predicate heads inflected with the factitive or the
change of state marker can be dative-marked to mark Reason adjuncts or Standard of
comparison adjuncts. The dative-marked constructions treated in this paragraph are
those in which the dative case encliticises directly to the verbal root or stem.
The semantic and syntactic function that the dative clause fulfils in the matrix
clause depends on the type of verb in the predicate of the matrix clause.
Dative complement clauses of Primary-B and Secondary verbs (see 4.5 for an
explanation of these terms) are always in O function. These are thus non-
canonically-marked O arguments just like the O arguments of the verbs of
emotion and interaction. Syntactic status: embedded (subordinate).
Dative complement clauses of any intransitive verb can function as S argument.
Syntactic status: embedded (subordinate).
Dative clauses in a subordinate linkage relationship with other verbs, both
transitive and intransitive, can function as adjuncts indicating a Purpose.

The subject (S/A) of the dative-marked clause has to be co-referential with the subject
in the main clause. It is not always possible to tell if this NP belongs to the matrix or
to the dative-marked clause.
27.2.1 Dative-marked complement clauses
All arguments of Primary-B verbs (see 4.5) can be personal pronouns or NPs;
however, alternatively the O argument can also be a dative clause. Secondary verbs
can only take dative clauses as O argument. A list of Primary-B and Secondary verbs
can be found in Table 23 in 4.5.2.

502
Dixon (2006-b: 1) defines a complement clause as a clause that functions as a core
argument in a matrix clause. When a dative-marked clause appears embedded in a
matrix clause headed by a Primary-B or Secondary verb predicate, it is always in O
function and hence a complement clause. Complement clauses can be questioned with
the question ato what? followed by a form of the main clause verb, e.g. what do
you want? or what are you thinking?. Dative complement clauses are the only non-
canonically-marked O arguments.
Example (753) illustrates a dative complement clause of the Primary-B verb hon'-
to give and (754) of the Secondary verb sok- to want.

(753) otokciba na'a angna aro amo jokna na' khewa dabat a thoyca dabat
aaw mu'ay sa'na hon'bo nookno.
otokciba [na'a] [a] =na aro [a =mo jok] =na
|[a] =aw
63
{mu'} =ay {sa'}| =na {hon'} =bo {no -ok} =no
1s =ACC stay =ADV eat =DAT give =IMP say -COS=QUOT
However, you keep giving me and my wife to eat as long as you live until I
die, he said, it is said.
64

The subject (S/A) of the dative-marked O complement clause is always co-
referential with that of the matrix clause and, if overt, it is only stated once, usually
before the complement clause. This might be a syntactic restriction or just be cause it
is the only logical inference that can be made in the contexts in which these clauses
usually occur, but more fieldwork is needed to find this out. There are no recorded

63
The accusative marking might be a mistake of the speaker. I have no idea why else it would appear in
this example, because the phrase is clearly a Beneficiary and one would expect a dative case marker.
64
Example (753) contains an instance of the expletive negative in the predicate thoy-ca (die-NEG).
The expletive negative reinforces the notion that the event denoted by the verb is unrealised. This
phemomenon is treated in more detail in 13.4. A more accurate translation of the clause in question
can be obtained in French, viz. [[a]{thoyca}dabat] jusqu ce que je ne meure, which involves the
subjunctive.

503
examples of the subject being stated twice, once in the subordinate clause and once in
the main clause. The next example, containing a Secondary main verb, illustrates that
it is not always possible to determine to which clause an NP belongs. The NP
phaltha self in example (754) is in S function in the dative-marked clause and in A
function in the matrix clause and it is impossible to determine to which clause it
belongs. The alternative analysis of the first line of the example is presented below the
translation.

(754) a phalthaan re'enado sokaydokcom.
a phaltha =an
65
|{re'e}| =na =do {sok -aydok} =com
1s self =FC/ID go.away =DAT =TOP want -PROG =IRR
I would like to go myself, [but because of my illness I cannot go.]
Alternatively: I myself would like to go[, but]

Alternative bracketing including the NP phaltha self in the dative-marked clause:

|[a phaltha =an] {re'e =na =do| {sok -aydok} =com
1s self =FC/ID go.away =DAT =TOP want -PROG =IRR

There are no arguments in favour or against either analysis.
Dative complement clauses can be used as an indirect speech report strategy. The
speech report is cast as a dative complement of the verb indicating speech, e.g. (755)
or thought, e.g. canci- to think in (756). Direct speech report constructions can stand
alone or can be embedded in a higher clause when they are cast as a complement of
the direct speech report verb no- to say, e.g. (62) above. Example (755) is a
continuation of the same story that example (767) below comes from. The horse was
really quick, it is said. then (755) happened.

65
There are two possible ways to bracket this part of the example, viz. [a] [phaltha-an] (1s
self=FC/ID), corresponding to the first translation, or, [a phaltha-an], corresponding to the second
translation. Both ways of bracketing are equally felicitous considering the context from which this
example is taken.

504
(755) kha'sinkhalay jalkhalna noaymo ga'dukdukciba, rakkhalay rakkhalay
jalariokno.
|{kha'sin -khal} =ay {jal -khal} =na| {no} =ay =mo
slow -CP =ADV run.away -CP =DAT say =ADV =SEQ
{ga'dukduk} =ci =ba
prod.with.legs =LOC =INDEF
{rak -khal} =ay {rak -khal} =ay {jal -ari -ok} =no
strong -CP =ADV strong -CP =ADV run.away -SIMP -COS =QUOT
Having told [the horse] to run slower, whenever he prods [it] with his legs, [it]
just runs faster and faster, it is said.

(756) bandiaw watetna canciaydokno.
|[bandi] =aw {watet}| =na {canci -aydok} =no
Name =ACC send =DAT think -PROG =QUOT
He is/was thinking to send Bandi. Or: He was thinking about sending
Bandi.

When Primary-B verbs are used with a noun as O argument, as in (757), the noun
cannot be marked by the dative case as in the next example.

(757) taka nacawa
[taka]
O
{na -ca -wa}
money need -NEG -FACT
[I] dont need money.

One secondary verb has been recorded with both dative and factitive complement
clauses. This is the verb ga'a- to be compelled. In Text 2 line 58 we find an
example with a factitive complement clause and here below is an example with a
dative complement clause.

(758) otokoysa dothadotha socina hapcina jalthokna ga'akok.
[otokoysa] |[dotha dotha so] =ci =na [hap] =ci =na
thats.why different RED village =LOC =ALL place =LOC =ALL
{jal -thok}|
O
=na {ga'ak -ok}
run.away -ALL =DAT compel -COS
Precisely like that/Thats why/So [they] were all forced to run away to
different villages [and] places.

505
Contrary to other Primary-B and Secondary verbs, phasal verbs cannot take dative
adjuncts but only factitive-marked complement clauses (see 24.3.1). In the following
example we see the verb jam- to complete, finish with a dative-marked adjunct. In
this example the subject (S/A) of the main and the dative-marked subordinate clause
are co-referential, as is always the case in Purpose constructions.

(759) otokoymo raja sonae matkaketdo jamok.
otokoymo |[raja] {so}| =na =e [mat -kaket] =do {jam -ok}
so.then king appoint=DAT =FC animal -ALL =TOP complete -COS
So then, all the animals were gathered in order to appoint a king. Literally:
all the animals were complete in order to appoint a king.
27.2.2 Dative-marked subject complement clauses
Dative-marked clauses can also occur in S function with intransitive main verbs as
long as the verb of the main predicate and the verb of the dative-marked predicate are
semantically compatible. Examples (760) and (761) are illustrative.

(760) mamu taka ni'wa aro sa'na ronaba ni'wa
[mamu taka]
S
{ni' -wa} aro |{sa'}| =na |{ro}|
S
=na =ba

nothing money not.exist -FACT and eat =DAT drink =DAT =ADD/EMPH
{ni' -wa}
not.exist -FACT
[I] dont have any money and nothing to eat [and/or] to drink. Literally:
Money does not exist and eat [and/or] drink does not exist.

(761) sigoret ronaan thawa
|[sigoret] {ro}|
S
=na =an {thaw -a}
cigarettes smoke =DAT =FC/ID tasty -CUST
Smoking cigarettes is tasty.

Dative-marked clauses cannot occur in transitive subject (A) function since they will
always be the O argument of Primary B and Secondary verbs and Purpose adjunct of
other transitive verbs.

506
27.2.3 Purpose adjunct clauses
Purpose adjunct clauses function as adjuncts to a matrix clause and give additional
information about the event depicted by the main clause predicate. The specific
properties of this clause type are given here below. There are no recorded occurrences
of negated purpose adjuncts.

Purpose adjuncts cannot appear in a clause with a Primary-B or Secondary verb as
predicate head for the reason that a dative-marked clause will be a complement
clause in O function. Apart from this caveat, any verb can have a Purpose adjunct.
Purpose adjuncts provide information to the question as to why the event
expressed by the main verb is taking place.

Example (762) shows a Purpose clause with an O argument. The verb in the main
clause is intransitive. The S of the main clause is co-referential with the implied A of
the dative adjunct clause.

(762) ama ni ue pholgomaw kawna re'enane.
[ama] [ni]
S
|[ue pholgom]=aw {kaw}|=na {re'e -na} =ne
mother 1pe DST eagle =ACC shoot =DAT go.away -DESI =TAG
Mother, we want/intend to go to shoot that eagle, OK.

As said above, the subject (S/A) of the uninflected dative-marked clause is always co-
referential with the subject of the matrix clause. Just like with the dative-marked O
complement clauses treated above, the subject which is common to both clauses, if
overt, is only stated once, i.e. before the purpose clause. More fieldwork needs to be
conducted to find out whether this is a syntactic restriction or not. When the subjects
are not coreferential, other constructions must be used, for instance an imperative or
an attributive clause, e.g. (763). In this example there is just one subject, i.e.
Cherogi, who is the A argument of the verbs nuk see and cay look, watch and
bandi ray'agaba Bandi who comes is the O argument. Bandi is modified by the
Attributive clause and is therefore not a clausal constituent on its own (see Chapter
29).


507
(763) bandi ray'agabaaw nukaymo caysawaymu balaydoano cerogiba []
[bandi ray'a =gaba] =aw {nuk} =ay =mu
Pname come =ATTR =ACC see =ADV =SEQ
{cay -saw} =ay =mu {bal -aydoa} =no [cerogi] =ba
watch -EXPECTANTLY =ADV =SEQ speak -PROG =QUOT Name =EMPH
[He (Cheronggi)] saw Bandy who comes, having watched expectantly, [and]
is saying, it is said, [this] Cheronggi: [why are you arriving so tired?]

In addition to this co-reference restriction, other arguments can also be shared by the
main and the co-subordinate clause. In the following example the noun wak pig is
both the O argument of the verb raw'- to catch and of the verb khat to slaughter. In
this example the implied A argument, i.e. first person plural, of the Purpose adjunct
clause is the same as the A of the main clause. This example is the answer to the
question What are you doing?.

(764) biana wak khatna raw'aydoa
[bia] =na [wak]
O
{khat} =na {raw' -aydoa}
wedding =DAT pig slaughter =DAT catch -PROG
[We] are catching a pig to slaughter for the wedding.

It is not possible to prove whether or not the NP wak pig is a constituent of the
Purpose adjunct clause or not. Because of this, the boundaries of the Purpose clause in
(764) are not indicated by vertical lines.
In example (765) here below the accusative-marked personal pronoun
(underlined) can be considered either as constituent of the Purpose adjunct clause or
of the main clause. Both the underlined NP and the purposive adjunct clause can be
omitted from the clause without making the clause ungrammatical or changing the
meaning. In fact, all NPs can always be omitted if they are recoverable from the
context.


508
(765) ge'theaw watkhuna so'otthelarinaka noay khu'moaokno.
[ge'the] =aw =do |{wat -khu}| =na {so'ot -thel -ari -naka}
3e =ACC =TOP banish -INCOM =DAT kill -SURELY -SIMP -IFT
{no} =ay {khu'mo -a -ok} =no
say =ADV conspire -WITHOUT.HOLDING.BACK -COS =QUOT
In order to get rid of him once more [we] will surely kill him, [they] freely
conspired, it is said.

The alternative boundaries of the Purpose clause indicated by columns would be
|[ge'the] =aw =do {wat -khu}| =na.
Type 1 adjectives, which are a subtype of intransitive verb, also take Purpose
adjuncts, as we can see in examples (766) and (767). In (766) it is the Type 1
adjective ja- be quick that takes a dative adjunct. As is mentioned in the chapter on
word classes, the change of state suffix on Type 1 adjectives can intensify the
meaning of the adjective, which is what happens in this example.

(766) somi nokgumukan wa'raraakona wal' khamna jaok.
|[so =mi nok] =gumuk =an {[wa'] =rara -ak}| =ona
village =GEN house =all =FC/ID bamboo =EXCLUSIVELY -COS =DAT
|{wal' kham}| =na {ja -ok}
fire burn =DAT be.quick -COS
Because all the houses of the village are made only of bamboo, [they] are
very quick to catch fire. Literally: [they] are very quick to fire-burn.

The following example comes from a story about a weak and cowardly king called
Bil, who has to fight the enemy of his ally-king. Bil asks his ally for the fastest horse.
But not knowing how to fight, Bil thinks (767) to himself.

(767) gore jalna rakcido joljol jalaarinaka noay canciokno.
[gore] |{jal}| =na {rak} =ci =do [joljol]
horse run.away =DAT strong =LOC =TOP quickly
{jal -a -ari -naka} {no} =ay {canci -ok} =no
run.away -AWAY -SIMP -IFT say =ADV think -COS =QUOT
If the horse is strong to run, [I/it] will just quickly run away, [he] sayingly
thought. Alternatively: If the horse can really run fast

509
27.3 Dative-marked clauses as complement of postposition
The postposition daka before, in the past, earlier occurs with the dative case (see
13.1). If the complement of the postposition is a noun, it will be dative-marked, e.g.
a=na daka (1s=DAT before) before me (in time). Similarly, if the complement of
the postposition is a verbal clause, the predicate head will also receive dative marking,
as we can see in examples (768) and (769) here below. There are no co-reference
restrictions between the subject (S/A) of the complement of the postposition clause
and that of the matrix clause or the main clause.

(768) ca hon'na daka balokno.
[[ca] {hon'} =na daka] {bal -ok} =no
tea give =DAT before say -COS =QUOT
Before [she] gave the tea [she] spoke, it is said.

(769) unasa ru cawna dakaan otokoy ru dokomaw ga'toaymuna kha dawa!,
kha dawa! noaymusa ru cawaymu patroanoro.
[unasa] [|{ru caw}| =na daka] [otokoy] [ru dokom] =aw
then boat stream =DAT before like.this boat head =ACC
{ga'ton} =ay =muna [kha dawa kha dawa]
stamp =ADV =SEQ interj Sname1 interj Sname1
{no} =ay =mu =sa
say =ADV =SEQ =DLIM
[ru] {caw} =ay =mu {pat -a -a} =no =ro
boat stream =ADV =SEQ cross -AWAY -CUST =QUOT =EMPH
Thats exactly why, before going by boat, after having stamped on the head of
the boat like this [and] only after saying Kha Dawa! Kha Dawa! [one] goes
by boat and usually crosses, it is said.
27.4 Summary of dative-marked clauses
The dative enclitic appears on predicates inflected with the change of state suffix, the
progressive/ durative aspect suffix or the factitive modality suffix as well as on
uninflected verbal roots. In all cases the dative case enclitic indicates the dependency
of the clause and limits the interpretations of the semantic role of the clause. Only
Reason clauses can have a noun functioning as predicate. In those cases, the factitive
suffix indicates the predicative role of the noun. Table 74 summarises the effect of the
dative case enclitic on-inflected and non inflected clauses.

510
Table 74 The effects of the dative case enclitic on clauses

on inflected predicates
on verbal root
CHANGE OF STATE- OR
PROGRESSIVE/DURATIVE-
MARKED PREDICATE PLUS
DATIVE ENCLITIC
FACTITIVE-
MARKED
PREDICATE PLUS
DATIVE ENCLITIC
SYNTACTIC
ROLE
subordinate: adjunct modifying
a matrix clause
subordinate: as
adjunct
modifying a
matrix clause or
governed as O
subordinate:
governed as S, O
or complement of
postposition or as
adjunct
SEMANTIC
ROLE AS
ADJUNCT
Reason
Reason, Standard
of comparison
Purpose
27.5 Locative-marked clauses
Locative clauses are marked by the locative case enclitic <=ci> (LOC) and fulfil the
adjunct function of Temporal Location in a matrix clause. This clause type indicates
punctual temporal location, except when the predicate carries the concomitant action
suffix <-butu> (WHILE), treated in 27.6. The predicate of a Location clause can
either be a bare root or stem or an already inflected verb with a factitive suffix, <-wa>
(FACT) as in (770), or after the concomitant action suffix <-butu> (WHILE), which
will be treated in the next section. The function of the factitive suffix on predicates of
locative-marked clauses will be discussed below. When the predicate does not carry
the factitive suffix, the locative enclitic has to be accompanied by either the topic
enclitic <=do> (TOP) as in (772), or the indefinite enclitic <=ba> (INDEF),
illustrated in (776), the delimitative enclitic <=sa> shown in (778), the
focus/identifier enclitic <=an> (FC/ID) (771) or the focus enclitic <=e> (FC)
illustrated in (779). This means that the temporal locative meaning of the independent
predicate head, which is characterising the clause type, has to be further specified as
being factual, topical, both, or indefinite, delimited, or focused. Factitive-
plus=LOCative-marked predicates can also be topic-marked, e.g. (775). The arguments
of a locative-marked clause do not have to be co-referential with those of the matrix
clause.
Examples (770) and (771) below illustrate the occurrence of the locative enclitic
on a verbal predicate head marked by the factitive. In this example the event denoted
by the locative-marked predicate has already occurred, i.e. Ketketa Bura, the main

511
character of the story from which this example has been taken, is already fat when the
other events described in the sentence occur. In (771) the event described by the
locative-marked predicate is pragmatically interpreted as a condition for a general
truth expressed by the locative clause and the main clauses combined.

(770) una jow'parae may hon'aymo mel'waci juta lopen muja tupi dokopaymo
bondok pansa hon'aymo watetokno.
una [jow'] =para =e [may] {hon'} =ay =mo
then mother =&co =FC rice give =ADV =SEQ
|{mel' -wa}| =ci
fat -FACT =LOC
[juta] [lopen] [muja] [tupi] {dokop} =ay =mo
shoe trousers sock hat dress.someone =ADV =SEQ
[bondok pan sa] {hon'} =ay =mo {watet -ok} =no
gun CLF:APPARATUS one give =ADV =SEQ send.away -COS =QUOT
Then, [Ketketa Buras] mother, having given [him] rice, when [he] was fat,
after [she] had dressed [him] in shoes, long trousers, socks [and] a hat, having
given [him] a gun, [she] sent him away, it is said.

A more literal translation of the locative clause mel'-wa=ci (fat-FACT=LOC) in the
above example would be when it was the case that he was fat. Although the factitive
can be seen as a nominaliser in dependent clauses, the abstract noun fatness would
be derived by means of the construction in which a factitive-marked verb takes the
genitive/nominaliser enclitic <=mi ~ =mo> (GEN/NR), viz. mel'wami ~ mel'wamo
(fat-FACT=NR) fatness, which could then take cases and function as a real noun
(see Chapter 24.1 24.6).

(771) magacakmi mon'do toysiwacian miniksuru takjolarianoro.
[magacak =mi mon'] =do {toysi -wa} =ci =an
deer =GEN body.hair =TOP wet -FACT =LOC =FC/ID
[miniksuru] {tak -jol -ari -a} =no =ro}
flat-haired do -QUICKLY -SIMP -CUST =QUOT =EMPH
As far as the deers body hair goes, when it is the case that it is wet, it quickly
becomes flat-haired.

The question now arises what exactly the function of the factitive is in examples (770)
and (771). As is discussed in 24.1, the factitive is an epistemic modality suffix that is
used by speakers to present the event denoted by the verb as a fact. Since factitive-
marked predicates in dependent clauses can take case marking, the factitive can be

512
considered to be a nominaliser, or, the fact that the factitive licences case-marking on
predicates can be seen as a remnant of its role as a nominaliser in an earlier stage of
the language. Nominalising the event, or reifying it, makes the occurrence or non-
occurrence of the event expressed by a factitive-marked verb factual. Example (770),
with nominalised predicate, contrasts with (772), where the event denoted by the
predicate of the locative-marked clause is not nominalised or reified by the factitive
and therefore interpreted as hypothetical in that it did not occur. But independently
from the presence or absence of the factitive, a locative-marked clause functions as a
Temporal location adjunct indicating temporal simultaneity.

(772) aro a ma'su pa'khucido tan'aymo kereaw phalkhunicom.
aro [a ma'su] {pa' -khu} =ci =do {tan'} =ay =mo
and 1s cow many -INCOM =LOC =TOP slaughter =ADV =SEQ
[kere] =aw {phal -khu -ni} =com
bone =ACC sell -INCOM -FUT =IRR
And if I [had had] more cows, having slaughtered [them] I could have sold
more bones. Literally: And at my cows greater quantity.

In the above example there are three clauses, represented here below as 1, 2 and 3.
1. aro a ma'su pa'-khu=ci=do at my cows greater quantity
2. tan'-ay=mo having slaughtered
3. kere=aw phal-khu-ni=com would have sold more bones
The conditional relationship between clauses 1 and 3 is pragmatically inferred. As we
see below in example (775), locative-marked clauses can also be topic-marked and the
semantic relationship between the two clauses in that example is temporal and not
conditional. The only difference between nominalised and non-nominalised locative
clauses is the degree of certainty of the taking place of the event denoted by the
predicate head of the locative clause. Events of nominalised locative predicate heads
are more certain to occur or to have occurred than those of non-nominalised ones.
Because non-nominalised locative-marked events are unlikely to occur or to have
occurred, they are interpreted as conditional, as in (772) above, but what is marked on
the predicate head of the locative clause is just Temporal Location and topicality. Of
course topics are easily interpreted as conditionals as is discussed at length in Haiman

513
(1978). Below, in example (777) we shall see that locative clauses without topic
marker can also be interpreted as conditionals.
The two following examples are a minimal pair. In example (773) the subordinate
predicate is factitive- marked and in example (774) it is not. In example (773) the
subordinate predicate is not topic-marked whereas in (774) it is. However, factitive-
marked clauses like the one in (773) can be topic-marked, and when they are, they
will not be interpreted as conditionals, as we can see in example (775).

(773) turasa re'ewaci ana topi ra'bone.
[tura] =sa {re'e -wa} =ci [a] =na [topi] {ra'}=bo =ne
Pname =MOB go.away -FACT =LOC 1s =DAT hat get =IMP =TAG
When you go to Tura, buy me a hat. or When your going to Tura is a fact,
buy me a hat.

(774) turasa re'ecido ana topi ra'bone.
[tura] =sa {re'e} =ci =do [a] =na [topi] {ra'} =bo =ne
Pname =MOB go.away =LOC =TOP 1s =DAT hat get =IMP =TAG
If you go to Tura, buy me a hat.

In both examples the event of going is expressed by the verb re'e- to go away. In
(773) the verb is marked by the factitive and in (774) the factitive is absent. Although
the events in both examples have not yet taken place, and are thus hypothetical, the
event of going in (773) is reified, i.e. presented as a fact, and, in the speakers opinion,
is much more likely to occur than the event in (774). This is another way of saying
that the event in (773), marked by the factitive, is presented by the speaker as much
more factual or realistic than the event in (774) where the factitive is absent. In (773),
according to the speaker, the person will certainly go to Tura sooner or later, we only
do not yet know when. In (774) the speaker does not expect the person to go to Tura
at all.
Example (775) below illustrates how a nominalised locative predicate head can
also be topicalised. As we can see in example (772) above, when the factitive suffix
<-wa> (FACT) is left out and only the locative and topic suffixes are used on the
predicate head, the event is interpreted as hypothetical because the event was not

514
realised, and the semantic relation between the two clauses is then interpreted as
conditional.
Example (775) comes from a recipe for beri food cooked in a bamboo cylinder.
The topic enclitic on the subordinate predicate indicates a new topical event.

(775) otokoymu berewa monwacido su'boloka.
otokoymu [bere -wa]
S
{mon -wa} =ci =do
so.then cook.food.in.bamboo.cylinder -FACT ripe -FACT =LOC =TOP
{su'bolok -a}
mash -CUST
So then, when the food cooked in the bamboo cylinder is ready, [we] mash
it.

Example (776) here below illustrates the effect of the indefinite enclitic <=ba>
(INDEF) on a clause with locative-marked predicate heads. The effect of the
indefinite enclitic is that the Temporal Location is indefinite. When the factitive suffix
<-wa> (FACT) is not present on the locative predicate, the event is more hypothetical,
whereas the presence of the factitive makes the event actual, more real or factual, as
we can see in (777) below.

(776) sosami sosigacina nawrukok tan'rukok. tan'rukciba patok ni'wa.
[so sa] =mi [so -siga] =ci =na {naw -ruk -ok}
village one =ABL village -ALT =LOC =ALL scold -RC -COS
{tan' -ruk -ok} {tan' -ruk} =ci =ba [patok] {ni' -wa}
slay -RC -COS slay -RC =LOC =INDEF prison not.exist -FACT
From one village to another [they] scold each other, slay each other.
Whenever [they] slay each other there is no prison.

515
In the next example the storyteller is giving an account of something that is actually
happening and therefore uses the factitive on the predicate of the locative clause. The
indefinite enclitic <=ba> (INDEF) supplies the meaning of indefinite time.
66

(777) tarapna guduk takwaciba tarakay jalariano magacake.
|{tarap}| =na [guduk]
67
{tak -wa} =ci =ba
catch.up =DAT wobble do -FACT =LOC =INDEF
{jal -ari -a} =no [magacak] =e
run.away -SIMP -CUST =QUOT deer =FC
Whenever [the Bengali] almost caught up, [he] just run away, it is said, the
deer.

As we can see in the following example, the delimitative-marked locative clause can
be interpreted as either temporal or conditional. The absence of a topic enclitic on the
locative clause proves that the topic enclitic <=do> (TOP) is not the only morpheme
responsible for the interpretation of non factitive-marked locative clauses as
conditionals.
The delimitative defines the boundaries of the temporal interpretation of the
locative clause.

(778) nemsakca takcisa, mo'sa kamal takthirini.
[nem -sak -ca] {tak} =ci =sa [mo' sa kamal]
good -APPROPRIATELY -NEG do =LOC =DLIM CLF:HUMANS one priest
{tak -thiri -ni}
do -AGAIN -FUT
Only at those times when/if it is not done appropriately well, another priest
will do it again.

66
The indefinite enclitic <=ba> (INDEF) also occurs on indefinite proforms. This suffix is
homophonous with the emphatic/additive enclitic <=ba> (EMPH/ADD) that occurs on arguments and
adjuncts, both phrasal and clausal (see Chapter).
67
As was mentioned in 22.6.2 ,this is the root of the verb guduk- to wobble, to move unstably which
functions adverbially in a light verb construction with the verb tak- to do. The construction guduk tak-
means to almost V.

516
The next two examples illustrate that even when the factitive suffix <-wa> (FACT) is
not present on the predicate of the locative clause, the clause does not have to be
interpreted as conditional and is in fact still temporal. Example (779) comes from a
story about the history of the Badri area. Example (780) was a spontaneous utterance
of wonder by one of my friends.

(779) ramu cow rosusaay coycie, rae san ci bri wawano.
[ra] =mu {cow ro -susa} =ay} {coy} =ci =e
rain =COM liquor drink -COMPETITIVELY =ADV try =LOC =FC
[san ci bri] {wa -wa} =no
day TEN four rain -FACT =QUOT
When [the people of Badri] tried to drink competitively with the rain, the rain
fell for fourteen days, it is said.

(780) ra nemcie ataknakasoy?
[ra] {nem} =ci =e {atak -naka} =soy
rain good =LOC =FC do.what -IFT =MIR
Now that the rain has stopped (Lit. is good), what the hell shall [we] do?

Locative clauses can be negated. The following example is illustrative. Remarkably,
no negated factitive-marked locative clauses are attested in Atong. More fieldwork is
necessary to find out if it is possible to negate such clauses or not.

(781) balcacido tokni.
{bal -ca} =ci =do {tok -ni}

tell -NEG =LOC =TOP hit -FUT
If [you] dont tell it, [I]ll hit [you].
27.6 The concomitant action suffix
The aspectual suffix <-butu> (WHILE) is found on the predicates of subordinate
clauses that function as Temporal Location adjunct in the matrix clause. This
morpheme indicates concomitant action, i.e. that the event in the subordinate clause
takes place simultaneously with the event in the matrix clause, or that the event in the
subordinate clause is already ongoing when the event in the main clause occurs.

517
Predicates carrying the concomitant action suffix occur in two different syntactic
constructions, which will be treated one by one below. The constructions are:

1. on predicates of Temporal Location adjunct clauses,
2. on the predicates of Temporal attributive clauses
68

These constructions will be treated separately below. Predicates carrying the
concomitant action suffix can be negated but are not attested with any other aspect or
modality suffixes.
27.6.1 Temporal Location adjunct clauses
The subordinate clause carries the locative case enclitic <=ci> (LOC). There are no
coreference restrictions between the arguments of the Temporal Location clause and
the matrix clause. Verbs, nouns and Type 2 adjectives have all been recorded as
predicates of concomitant action type Temporal Location clauses.
The following example presents a Type 2 adjective, viz. thombolo have holes as
the head of the predicate of the subordinate clause. Type 2 adjectives can function as
modifiers as well as predicates and are treated in 5.2.

(782) ie ram thombolobutuci gari galatok.
|[ie ram]
S
{thombolo -butu}| =ci [gari] {galat -ok}
PRX road have.holes -WHILE =LOC car fall -COS
While this road was damaged, cars fell.

Example (783) illustrates the concomitant action suffix on a subordinate nominal
predicate. The locative-marked clause functions as Temporal Location.

68
Out of a total of 23 predicates with the concomitant action suffix, 18 occurred in Temporal Location
adjunct clauses, i.e. 78%, and 5 occurred in temporal attributive clauses, i.e. 22%

518
(783) ge'the sa'goraybutuciba sansan palosa na' punna re'ewa
|[ge'the] {sa'goray -butu}| =ci =ba
3s child -WHILE =LOC =INDEF
[san san] [palo] =sa {na' pun} =na {re'e -wa}
day RED jungle =MOB fish catch.fish =DAT go.away -FACT
When he was a child, he went to the jungle every day to catch fish.

The next examples show verbal predicates carrying the concomitant action suffix. In
(784) there is no coreference between the implied subject of the subordinate clause
and the stated subject of the main clause. The implied S of the first clause (the
Temporal Location adjunct) is the lazy king, who is sad because he does not have any
friends, while the S of the main clause is na'pit barber.

(784) phepci sontibutuci te'ewe na'pit mo' sa ray'phaknoro.
|[phep] =ci {sonthi -butu}| =ci
banyan.tree =LOC suffer -WHILE =LOC
[te'ew] =e [na'pit mo' sa]
S
{ray' -pha -k} =no =ro
now =FC barber CLF:HUMANS one come -IN.ADDITION -COS=QUOT =EMPH
While he was suffering in the banyan tree, a barber came by, it is said, really.

(785) otokoymo thomay caybutucie atoaw nukokno ge'thee?
otokoymo |{thom} =ay {cay -butu}| =ci =e
so.then lay.in.ambush =ADV watch -WHILE =LOC =FC
[ato] =aw {nuk -ok} =no [ge'the] =e
what =ACC see -COS =QUOT 3s =FC
So then, while he
i
was lying in ambush and watching, what did he
i
see?

The subordinate predicate in (786) is the Type 1 adjective (a subclass of intransitive
verb), nem to be good with the event specifier suffix <-khal> (CP).

(786) sagaba naw nemkhalbutuci thoyok
{sa} =gaba [naw]
S
{nem -khal -butu} =ci {thoy -ok}
be.ill -ATTR younger.sister good -CP -WHILE =LOC die -COS
When my younger sister was getting better, she died.

519
The absence of locative enclitic on the predicate sa'-but (eat-WHILE) in (787) can
be explained when we consider the two predicates sa'-butu (eat-WHILE) and ro-
butu (drink-WHILE) to be part of a complex predicate with the locative enclitic
attached to it.

(787) a babaci amaci maha maha sa'butu robutucian, randay sa'na jamca.
[a baba] =ci [ama] =ci [maha maha]
1s father =LOC mother =LOC great RED
{sa -butu ro -butu} =ci =an
eat -WHILE drink -WHILE =LOC =FC/ ID
[randay] {sa'} =na {jam -ca}
meat eat =DAT finish -NEG
While I ate and drank in great amounts at my father and mother [s place],
[we] didnt finish eating meat.

The next example contains an illustration of a negated Temporal Location clause.

(788) ge'theng jowcabutucii korogaba nawa
[ge'the] {jow -ca -butu} =ci [koro =gaba] {na -wa}
3s sleep -NEG -WHILE =LOC make.sound =ATTR hear -FACT
While he was not [yet] sleeping, he heard a sound

There are two recordings of a Temporal Location clause of which the noun somay
time occurs compounded on the predicate after the concomitant action suffix
<-butu> (WHILE). The noun somay time does not add any semantic content to the
predicate, but simply emphasises the ongoing or temporally stretched-out character of
the event expressed by the predicate. The concomitant action suffix together with the
noun somay time form a double marker of subordination. It is conceivable that this
noun is in the process of grammaticalising as a linker. The two recorded examples
with this construction are almost identical and therefore just one of them is
represented here. The noun somay time is an Indic loan related to Hindi
(samay) time.

520
(789) uci mu'butu somayci badri nemen man'ay sa'ano
[u =ci {mu' -butu +somay} =ci
DST =LOC stay -WHILE +time =LOC
[badri] [nemen] {man'} =ay {sa' -a =no
During the time [they] lived there, Badri was very rich (ate in great amounts),
it is said.

A very similar phenomenon is described for Tamil (see Lehmann, 1989: 341 cited in
Heine and Kuteva, 2002: 299), where the noun pootu time can function as a
temporal clause marker. Heine and Kuteva write: This is an instance of a process
whereby a noun, on account of some salient semantic property, gives rise to a
grammatical marker highlighting that property []. The same can be said about the
use of somay time in Atong.
27.6.2 Temporal attributive clauses
Attributive clauses are treated in Chapter 29, which the reader is advised to consult
first for theoretical details. In summary, the attributive clause functions as modifier to
a noun which is the head of a complex NP called the arch NP
69
. The arch NP
functions as a constituent in a matrix clause and can be case-marked with enclitics.
The attributive clause can occur on either side of the head of the arch NP without
difference in meaning. There can be a semantic relationship between the head of the
arch NP and the predicate of the attributive clause, which has to be inferred
semantically and/or pragmatically. There are no grammatical constraints that force
any semantic interpretation of an arch NP.
In (790) we see that the temporal attributive clause modifies the head sok sprout
and that the whole arch NP is accusative-marked for its O function in the matrix
clause of which the predicate is sa'-wa (eat-FACT). The inferred semantic

69
The term arch NP refers specifically to an NP of which the head is modified by a clause. I invented
this term to make it easy to refer to such NPs without using a lot of words, like NP of which the head
is modified by a clause.

521
relationship of the head of the arch NP, the noun sok sprout, to the predicate of the
attributive clause is that of Actor.

(790) una aludaraaw rodomabutu sokaw sa'wamo gomon te'ew manap
caywacido gumukan cokarumokno
una [alu] =dora =aw
therefore potato =p =ACC
---------------------matrix clause--------------------
--------------arch NP---------------
---------AC---------
[|{rodoma -butu}| sok
ACTOR
]
O
=aw {sa' -wa} =mo [gomon]
sprout -WHILE sprout =ACC eat -FACT =GEN reason
[te'ew] [manap] {cay -wa} =ci =do
now morning look -FACT =LOC =TOP
[gumuk] =an {coka -rum -ok -no}
all =FC/ID tear -ALL -CoS -QUOT
Therefore, when he looked in the morning, the potatoes were all torn, because
all the sprouts were eaten while they were sprouting, it is said.

In the following example we see a headless arch NP that carries prototypical nominal
morphology, i.e. the plural enclitic <-dora> (p), The accusative case enclitic <-aw>
(ACC) marks the arch NP for its semantic role as O argument in the matrix clause.

(791) sala burbok sa'goray na'a ni jowsukbutudora atakna halaka'wa?
sala [burbok sa'goray] [na'a]
interj idiot child 2s
-------------------arch NP--------------------
-----------------AC------------------
[[ni] |{jow -suk -butu}|]
O
=dra =aw
1pe sleep -COMFORTABLY -WHILE =p =ACC
[atakna] {hala ka' -wa}
why disturb do -FACT
Damn you idiot child! Why did you disturb us, [the ones who] were
comfortably asleep?

In (792), from the same story as (790), is certainly a headless arch NP in which the
implied head is gore horse. The predicate of the temporal attributive clause is sa'-

522
butu (eat-WHILE) and the predicate of the matrix clause in which the arch NP
functions as O argument is nuk-ok-no (see-COS-QUOT).

(792) te'ewe beanbebe rarasami gore masa ray'aaymo ge'the alubagan
habijabi samcakaw sa'butuaw nukokno.
---------------------------arch NP-------------------
[te'ew] =e beanbebe [|[rara] =sa =mi [gore ma sa]
now =TOP truly sky =MOB =ABL horse CLF:ANIMALS one
{ray'a} =ay =mo
come =ADV =SEQ

----------------------------------------arch N--------------------------------------------
---------------------------------------------AC-------------------------------------------
[ge'the alu bagan bari habijabi samcak] =aw {sa' -butu}|]
O
=aw
3s potato garden garden all.sorts vegetable =ACC eat -WHILE =ACC
{nuk -ok} =no

see -CoS =QUOT
Now [he] truly saw a horse having come from the sky while [the horse was]
eating all kinds of vegetables of his garden. Alternatively: He truly saw a
horse having come from the sky, [which horse] at that time was eating all
kinds of vegetables of his garden.

Hale (1976) uses the label adjoined relative clause to refer to a clause type that can
have both adverbial and relative functions, similar to the clause with a predicate head
marked by <-butu> (WHILE) in Atong. The classic illustration of the adjoined
relative clause comes from Warlpiri. (Hale, 1976: 78, example 1), here represented as
(793).

(793) Ngajulu-rlu rna yankirri pantu-rnu, [kuja-lpa ngapa nga-rnu].
I-ERG AUX emu spear-PAST COMP-AUX water drink-PAST
I speared the emu which was/while it was drinking water

523

524
Chapter 28 Adverbial and sequential clauses
_____________________________________________________________________

Adverbial and sequential clauses are subordinate clauses marked with special clausal
enclitics that have no function elsewhere in the grammar, viz. the adverbial enclitic
<=ay ~ =e> (ADV) and the sequential enclitic <=mo ~ =mu ~ =mu ~ =muna ~
=muna> (SEQ). Adverbial clauses function as manner adverbs modifying the
following clause and are treated in 28.1. Sequential clauses, treated in 28.2, give
background information and can be used to imply a cause and effect relationship
between two clauses.
28.1 Adverbial clauses
Clauses with the clausal enclitic <=ay ~ =e> (ADV) function as adverbial adjunct
clauses in a matrix clause. The allophone <=e> of the adverbial enclitic appears after
a stem ending in /i/, and the allomorph <=ay> occurs elsewhere. Verbs and Type 2
adjectives, and nouns (see example (808)), are attested to function as predicate of this
clause type. Adverbial predicates cannot take aspect or modality suffixes but can take
the negative suffix <-ca> (NEG). The subject (S/A) of the adverbial predicate is
always co-referential with the subject of the predicate it modifies. More fieldwork
needs to be done to find out if this is a syntactic restriction or not. Argument structure
in adverbial clauses is the same as in main clauses and so is NP marking, i.e. S and A
unmarked for case and O can be accusative-marked (see Chapter 20).
Since adverbial clauses cannot occur as sentences on their own, they are
dependent on the matrix clause for their appearance. Adverbial clauses function as
manner adverbs, indicating how the event denoted by the modified predicate comes
about, examples (794), (795) and (796) are illustrative.

(794) pheru panci cagakay thoyokno
[pheru][pan] =ci {cagak} =ay {thoy -ok} =no
fox tree =LOC hit =ADV die -COS =QUOT
The fox hit a tree and died, it is said.

28 ADVERBIAL AND SEQUENTIAL CLAUSES

525
(795) lekha kirinay sayok.
[lekha] {kirin} =ay {say -ok}
paper to.tear =ADV write -COS
[I] wrote [on] the paper and tore it. Literally: [I] tearingly wrote [on] the
paper.

(796) wak romay sa'a nido
[wak] {rom} =ay {sa' -a} [ni] =do
pig cook =ADV eat -CUST 1p =TOP
We eat pig cooked.

An adverbial predicate modifies the following predicate even when this following
predicate is itself subordinate. In (794)-(796), the following predicates are all main
clause predicates. looking back at example (779) we can see how an adverbial clause
modifies a Locative-marked clause, which is subordinate. In example (648) in 24.3.1
we see the adverbial clause rom-ay (cook=ADV) modifying the subordinate predicate
sa'-wa (eat-FACT). In (797) below we see a subordinate clause functioning as
Facsimile adjunct modified by the adverbial clause takruk-ay (fight=ADV).

(797) [] takrukay ray'asemgabatokoy nuksawphinokno.
{takruk =ay} [{ray'a -sem} =gaba] =tokoy
fight =ADV come -CERTAINLY =ATTR =LIKE
{nuk -saw -phin -ok} =no
look.like -CERTAINLY -COMPLETELY -COS =QUOT
[ he] certainly looks like someone who certainly came fightingly [or in
fighting manner or as if he had been fighting
70
], it is said.

Example (798) shows how an adverbial clause is used in a context in which a cause
and effect relationship between the adverbial and main clause can be inferred

70
The adverbial clause can be translated into English in different ways. In Atong this adverbial clause
simply functions as a manner adverb, indicating in which way the event denoted by the following
predicate comes about.

526
pragmatically. The adverbial clauses indicate the cause and the main clause the effect.
What is important in this example is that the adverbial construction is used to convey
the message that the praying and the offering to the elephant tusks happened
simultaneously to becoming very rich. In this respect the adverbial clause differs from
the sequential clause that is also used in contexts in which cause and effect can be
inferred pragmatically. The sequential clause indicates that the events expressed in the
sequential and main clause happened in succession.

(798) soggumukan ue momawana way khurutaysa boli hon'aysa man'ay
sa'thokwano.
[so] =gumuk =an [ue moma wa] =na
village =whole =FC/ID DST elephant tooth =DAT
{way khurut}
cause
=ay =sa

spirit perform.an.incantation =ADV =DLIM
{boli hon'}
cause
=ay =sa
offering give =ADV =DLIM
{man'} =ay {sa' -thok -wa}
effect
=no
in.great.amounts =ADV eat -ALL -FACT =QUOT
Precisely because the whole village prayed and offered to the elephant tusks,
they all became very rich, it is said. Alternatively: They became very rich, it
is said, the whole village [by/whilst] offering and praying to the elephant
tusk.

Adverbial predicates can be negated, which is illustrated in example (799) here below.

(799) phowra sow'aymuna garu susetcaay dowetoknoay.
[phowra] {sew'} =ay =mona [garu] {suset -ca} =ay
rice.powder pound =ADV =SEQ mustard.leaves wash -NEG =ADV
{dow -et -ok} =no =ay
add -CAUS -COS =QUOT =POS
Having pounded rice powder, [she] added the mustard leaves without
washing [them], really!

Type 2 adjectives can also function as predicate of a non-finite clause. The following
example illustrates a Type 2 adjective as predicate head of a simultaneous clause. The

527
example comments on a girl with dark skin whose beauty was renowned among
unmarried boys throughout the region.

(800) pinakay sola
{pinak} =ay {sol -a}
black =ADV beautiful -CUST
[Shes] black and beautiful.
28.2 Sequential clauses
A sequential clause cannot occur as a sentence on its own and is therefore dependent
on a main clause for its occurrence. As a sequential clause is not governed by the
main clause predicate, i.e. it does not function as core argument, but is dependent on
the main clause for its occurrence and is hence grammatically a modifier and thus
subordinate.
A sequential clause is marked with both the adverbial enclitic <=ay ~ =e>
(ADV)
71
and the sequential enclitic <=mo ~ =mu ~ =mu ~ =muna ~ =muna>
(SEQ). The allomorphs of the sequential morpheme are in free variation. In the Badri
dialect there is a strong preference for the allomorph <=mo> whereas in the dialect
of Siju the other allophones are preferred.
There is only one story in the recorded corpus in which the sequential enclitic is
not preceded by the adverbial enclitic <=ay ~ =e> (ADV) but by the factitive suffix
<-wa> (FACT). Example (640) is illustrative. Why the speaker uses the factitive in
sequential clauses has to be investigated in future fieldwork.
Sequential clauses usually provide background information and therefore usually
occur at the beginning of a sentence. However the position of a sequential clause is
quite variable. Example (801) shows a sequential clause before the main clause
predicate and in example (802) the sequential clause comes after the main clause
predicate. When a sequential clause follows a finite predicate it is considered to be an

71
As was mentioned in the previous section, the allophone <-e> of the adverbial enclitic appears after a
stem ending in /i/, the allomorph <-ay> occurs elsewhere

528
afterthought. Example (803) illustrates that sequential clauses can also come in
between a subordinate clause and a main clause.

(801) ca roaymo, may sa'aymo, ray'naka.
|[ca] {ro}| =ay =mo |[may] {sa'}| =ay =mo {ray' -naka}
tea drink =ADV =SEQ rice eat =ADV =SEQ go -IFT
Having drunk tea, having eaten rice, well go.

(802) otokoyimuna phaltha digaray sagaawdo caythirina noaymu re'eokno,
awanaymu.
otokoymuna |[[phaltha digaray] {sa} =ga] =aw =do
so.then self fish.trap put.as.trap =ATTR =ACC =TOP
{cay -thiri} =na {no =ay}| =mu {re'e -ok} =no
look -AGAIN =DAT say =ADV =SEQ go.away -COS =QUOT
|{awan}| =ay =mu
forget =ADV =SEQ
So then, having looked again at the fish trap that he had put up himself, as
said before, he went away, it is said, having forgotten [it]. (i.e. having
forgotten that he was not wearing any underwear under the gamusa cloth tied
around the waist with a knot he was wearing and which he had wound around
his head against the sun.)

The arguments of a co-subordinate sequential clause, even when implied, do not have
to be co-referential with those of the main clause, as we can see in example (807)
below and (803) below. In example (803), the implied A argument in clause 2, i.e. the
person who shot Arong Nokma in the face, is not the same as the person who dies. In
fact, it is Arong Nokma who dies, who is the S argument of the first clause and of the
third clause.

(803) aro nokma caykhawwaci aro nokmami mukhaaw khiemu thoyokno.
clause 1 [aro nokma]
S
{cay -khaw -wa} =ci =e
Name headman look -SURREPTITIOUSLY -FACT =LOC =FC
clause 2 [aro nokma =mi mokha]
O
=aw {khi} =e =mu

Name headman =GEN face =ACC hit.the.mark =ADV =SEQ
clause 3 {thoy -ok} =no
die -COS =QUOT
When headman Arong
i
looked surreptitiously, having hit Arongs face, [he
i
]
died.
Sequential clauses are a clause chaining device. Speakers easily produce clause chains
involving four or more sequential clauses of which example (804) is illustrative. As

529
we can see in this example, the adverbial phrase manapmi=an (very.early.
in.the.morning=FC/ID) very early in the morning has scope over the whole sentence
and each of the following sequential clauses has scope over all subsequent clausal
constituents of the sentence. The order of the sequential clauses is iconic.

(804) manapmian may ja'bek romaymuna, may ja'bek monmanaymuna romay
sa'aymuna, maysangumuk pon'aymuna, hay'aw garu balagaci ramay
tanaokno.
clause 1 [manapmi] =an |[may ja'bek]
O
{rom}| =ay =muna
very.early.in.the.morning =FC/ID rice curry cook =ADV =SEQ
clause 2 |[may ja'bek]
S
{mon -man}| =ay =muna}
rice curry ripe -already =ADV =SEQ
clause 3 |{rom}=ay {sa' =ay}| =muna
cook =ADV eat =ADV =SEQ
clause 4 |[may san]
O
=gumuk {pon'}| =ay =muna
rice day =all pack =ADV =SEQ
clause 5&6 [hay' =aw garu]
O
[balaga] =ci
GPN =ACC mustard.leaves outside =LOC
{ram} =ay {tan -a -ok} =no
dry =ADV put -AWAY -COS =QUOT
Very early in the morning, having cooked rice, the rice having become ready,
having eaten it cooked, having packed all the rice for lunch, [she] put that, eh,
mustard leaves outside to dry, it is said.

The sequential nature of the relationship between a sequential and main clause can be
reinforced by the relative time postposition konsa later, after as we see in example
(805).

(805) otokoymuna kamaymuna konsado jow'gaba noksa ray'aakno.
otokoymuna {kam} =ay =muna [knsa] =do [jow' =gaba]
so.then work.on.the.land=ADV =SEQ after =TOP mother =DREL
[nok] =sa {ray'a -ak} =no
house =MOB come -COS =QUOT
So then, after working in the rice field, the mother came home, it is said.
Note that the postposition konsa later, after governs the genitive case and that
sequential clauses do not take the genitive enclitic because sequential clauses are not
nominalisations.

530
Sequential clauses can be negated, which is illustrated in the following example.
The fox has jumped into a deep well because he wanted to drink water. But (806):

(806) ga'khatna man'caaymo thoyokno.
||{ga'khat}| =na {man' -ca}| =ay =mo {thoy -ok} =no
climb =DAT be.able -NEG =ADV =SEQ die -COS =QUOT
Not having been able to climb out, [he] died, it is said.

Sometimes a cause and effect relationship between the sequential and the main clause
can be pragmatically inferred. Examples (807) and (808) are illustrative and contrast
with examples (801) and (802) above in which the non-finite sequential clauses
simply indicate that the events in the stretch of narrative happened in sequence. When
a cause-effect relationship can be inferred, as in the examples below, the sequential
clause will indicate the Cause and the main clause the effect. This semantic role of the
sequential clause is inferred pragmatically from the context since the clause is not
marked for its role as Reason, but rather only as sequential.

(807) balwa rakaymo wa' bay'ok.
[balwa] {rak} =ay =mo [wa' bay' -ok]
wind strong =ADV =SEQ bamboo break -COS
The wind having been hard, the bamboo has broken. Alternatively: Because
of the hard wind the bamboo has broken.

Example (808) illustrates the use of a sequential nominal predicate head. The word
calak cunning is a Type 2 adjective modifying the head noun morot person.


531
(808) heg'thondo se'ano, calak morotaymo udo te'ewcinaan kegaydoano.
[thethon] =do {se' -a} =no {calak morot} =ay =mo
Pname =TOP intelligent -CUST =QUOT cunning person =ADV =SEQ
[u] =do [te'ew] =ci =na =an {ke -aydoa} =no
DST=TOP now =LOC =ALL =FC/ID be.alive -DUR =QUOT
Thengthon is intelligent, it is said, [because] he is a cunning man, he is now
still alive, it is said. Alternatively: having been a cunning man, he is now still
alive.

In everyday speech the Atong seem to use the sequential clause construction much
more frequently than the factitive-plus-dative-marked reason clause construction to
express cause and effect. This might have to do with the truth value of the Reason
clause. Since it is factitive-marked (see Chapter 1), the reason clause has a strong
truth value. A speaker will only use it if he is certain or wants to imply that he is
certain that what he says is factual, i.e. that the causal event was factual. To avoid
taking responsibility for letting people think that a causal event was factual, a speaker
uses the sequential clause construction. He can then express a temporal connection
between two events and the hearer can decide whether or not to take it as an implied
cause and effect relationship.
We find Reason clauses particularly often when speakers are talking about
themselves and in stories about history in which reason clauses are used to present
historical facts. These are all situations in which the speaker can easily express causal
events with certainty. In the case of first person, because you usually know what you
yourself did, and in the case of historical fact, there is really no discussion possible.
So a correlation between person and the construction used to express causality is
certainly expected. More examples need to be found through future fieldwork.
It might be that the use of the sequential clause construction when talking about
events other than in the first person has attained the status of politeness. It would be
interesting to find out, through more fieldwork, if people find it rude to speak about
things that happened to others with reason clauses and therefore prefer to use
sequential clauses.
In the next example the speaker uses a sequential clause to talk about something
that happened to a third person.


532
(809) ie morot toy huna sapcaaymu toy caw'wa.
[ie morot] {toy hu} =na {sap -ca} =ay =mu
PRX person water swim =DAT know.a.skill -NEG =ADV =SEQ
{toy caw' -wa}
water stream -FACT
This person, not having known how to swim, drowned. Alternatively:
Because this person did not know how to swim, he drowned.

An example with a Reason clause used in a first person situation can be found in Text
2 line 3, which is also represented in example (749).
Example (810) illustrates the use of a Reason clause to tell an undisputable
historical fact in the story about the history of the Badri area.

(810) ue toygat rowanasa ue toykhalawe rodo mowano.
[ue toygat] {ro -wa} =na =sa [ue toykhal] =aw =e
DST water.place drink -FACT =DAT =DLIM DST river =ACC =FC
[rodo] {mo -wa} =no
Pname call.a.name -FACT =QUOT
Because [the Rongdyng clan] had drunk at that water place, that river was
called Rongdyng, it is said.

It is quite probable that the sequential enclitic <=mo ~ =mu ~ =mu ~ =muna ~
=muna> (SEQ) derives historically from the allomorph <=mo> of the
genitive/ablative phrasal enclitic <=mo ~ =mi> (GEN/ABL). Some allomorphs seem
to have fused with the dative case enclitic <=na> (DAT). Today, as has been
mentioned above, these allomorphs are in free variation, although there is a strong
preference for certain allomorphs in certain dialects.

533

534
Chapter 29 Attributive clauses
29.1 Terminological preliminaries
When I set out to describe attributive clauses in Atong, I found the traditional
theoretical literature on relative clauses (Keenan and Comrie 1977, Comrie 1981,
Keenan 1985) not very helpful to describe the phenomenon in the language. Keenan
(1985), for example, tells the reader that a restrictive relative clause (RC) is an NP
(see page 141) and that such an NP consists of an optional determiner, an omitable
common noun also called the domain noun and a restrictive clause (S
rel
) that
modifies the domain noun (see page 142). The domain noun is said to be the head
of the RC (see page 145). Keenan gives no arguments why the noun should be called
common noun. Furthermore, Keenan indicates that he will only treat RCs of which
the domain noun or head occurs outside the S
rel
and then calls these constructions
external RCS (external relative clauses).
72

This terminology is confusing, because the so called head is not external to the
RC but to the S
rel
. The head is internal to the RC, because that is how Keenan
defines an RC (see page142). For me the head of a clause is the predicate. A
determiner cannot be seen as a separate constituent in a clause since it is part of an
NP. And it is confusing to have the same name or abbreviation ((restrictive) relative
clause (RC), restrictive clause (S
rel
)) for the NP in which the modified noun occurs
and the clause that modifies it. Regretfully, Keenans terminology has had many
reverberations in later literature and so the confusion has been perpetuated.
The solution to this confusion, when describing Atong at least, is to have separate,
clear, typologically transparent terms for the separate elements that we find in relative
or attributive clause constructions. For Atong we can define these elements as
follows. An attributive clause construction involves two clauses, a matrix clause and

72
In section 5 of his article, Keenan (1985: 168 ff.) treats non-restrictive relative clauses as other
relative-like constructions. In Atong there is no grammatical distinction between non-restrictive and
restrictive attributive clauses.
29 ATTRIBUTIVE CLAUSES

535
an attributive clause. The attributive clause predicate is marked with the clausal
enclitic <=gaba ~ =ga> (ATTR), of which the allomorphs are in free variation.
73
An
attributive clause modifies a noun. Together, the attributive clause and the modified
noun form the so called arch NP of which the modified noun is the head.
74
The arch
NP as a whole functions as a constituent in the matrix clause. Aside from the
attributive clause an arch NP can also contain other modifiers, as we will see below.
As will be argued in section 29.2, the head of the arch NP is not a constituent of the
attributive clause. The evidence for this is the inability of the noun to be marked for
case. A semantic relationship between the head of the arch NP and the predicate of the
attributive clause can be inferred, but is not compulsory in any way. As we will see in
section 29.3, there are also arch NPs where no semantic relationship can be inferred
between the head and the predicate of the attributive clause, e.g. (831). This example
is also evidence for the lack of a gap in Atong attributive clauses.
We can see in example (811) how this terminology explains the syntactic situation
in Atong. The predicate of the attributive clause is cu to be big. The head of the
arch NP is pholgom eagle. The head is modified by the attributive clause. Atong
marks case with phrasal enclitics. In this example we see that the accusative enclitic
indicates the O function of the arch NP in the matrix clause. The semantic relation of
the head of the arch NP to the predicate of the attributive clause is that of Attributant.
The semantic role of Attributant indicates the relationship between a Type 1 adjective
(a stative verb denoting a quality, see Chapter 5), and its S argument.

73
Older speakers prefer to use the allomorph <=gaba> while younger speakers prefer <=ga>.
74
Andrews (2007: 206) calls an NP whose reference is being restricted by a relative clause the NP
mat
,
because this NP occurs in the matrix clause. NP
mat
and arch NP are thus synonyms, since an arch NP
also occurs in the matric clause. I prefer to use the term arch NP to avoid any misreading of NP
mat
as
matrix noun phrase. An NP containing a noun-modifying clause can occur embedded as a modifier in
a higher NP (as we will see below, e.g. (813)), which could in that case be the matrix NP, and we
would still have terminological confusion.

536
(811) ucie pholgom cunggabaaw nukokno
--------------------------matrix clause-------------------------
---------------arch NP---------------
---AC---
ucie [phlgm
Atbtnt
|{cung}| =gaba]
O
=aw {nukok} =no
then eagle big =ATTR =ACC see =QUOT
Then [he] saw the big eagle, it is said.

It is important to note that, except in more complicated cases, which will be treated in
sections 29.6 and 29.7 below, the attributive clause can precede or follow the head of
the arch NP without any conceivable difference in meaning.
75
In (812) we see how the
head of the arch NP, kam work is modified by a preceding attributive clause with the
predicate cu to be big. The arch NP is not accusative-marked for its O function in
the matrix clause because it is not referential.

(812) konsado cugaba kam man'ok, sagaltoysamci.
------------------------------------matrix clause------------------------------------
--------------arch NP-------------
--AC--
konsa =do [|{cu}| =gaba kam
Attributant
]
O
{man' -ok} [sagal toysam] =ci
later =TOP big =ATTR work get -CoS sea waterside =LOC
Later [he] got a big job at the seaside.

The arch NP as a whole can be embedded as a modifier within another NP, as we can
see in (813), where the noun dada elder brother is the head.

75
There might be subtle pragmatic differences depending on whether the attributive clause follows or
precedes the head, although thorough investigations in the field and of the recorded data have revealed
no differences at all.

537
(813) sa?geray melgabami dadadara
------------arch NP-----------
-AC-
[[sa'goray |mol| =gaba] =mi dada]
MATRIX NP
=darang
child be.small =ATTR =GEN elder.brother =p
the small childs elder brothers

The clausal enclitic <=gaba ~ =ga> (ATTR) is not a nominaliser. I define
nominalisation as a derivational process of which the outcome, whether it be a clause
or a single morpheme, can function as the head of an NP, which is the most salient
property of nouns. The attributive enclitic in Atong turns clauses (with or without
NPs) into nominal modifiers. Modifying a noun is not an exclusively nominal
property, but is also a property of demonstratives (see Chapter 1), some interrogatives
(see Chapter 9), personal pronouns (see 17.2), Type 2 adjectives (see Chapter 5) and
some indefinite proforms (see Chapter 1). Therefore I think that it is infelicitous to
call clauses with the attributive enclitic <=gaba ~ =ga> (ATTR) nominalisations.
Only members of the word class of verbs are attested as head of the predicate of
an attributive clause. Attributive clause predicates seem to behave exactly like main
clause predicates in the possibilities of inflection that they can express and the
arguments they can take. Being able to take arguments, core or oblique, is one of the
most important verbal properties. Although not all aspectual and modality suffixes are
attested on predicates of attributive clauses, when elicited, speakers find attributivised
verbs marked for any type of aspect and modality acceptable and even natural.
Example (818) is an illustration of a progressive-marked attributivised predicate.
Example (814) below exhibits a factitive-marked attributive clause predicate. The
biggest difference between main and attributive clause predicates is the fact that
nouns and Type 2 adjectives cannot function as predicates of attributive clauses, while
they can function as main clause predicates.
76

76
Type 2 adjectives can modify an NP in post head position, can function as predicate of
identity/equation clauses, like nouns, but cannot take the customary aspect suffix.

538
In the example below we see a headless, genitive-marked arch NP which functions
as a Possessor in a larger NP of which the noun bimu name is the head. The
predicate of the attributive clause is factitive-marked.

(814) badrido cigacakci mu'wagabami bimu do'acom.
--------------------arch NP------------------
-----------------AC-------------
[badri] =do [ [ | [cigacak] =ci {mu' -wa}| =gaba] =mi bimu]
Pname =TOP Pname =LOC stay -FACT =ATTR =GEN name
{do' -a} =com
IE.be -CUST =IRR
As for Badri, [it] is supposedly the name from [the people] that were living in
Chigachak. i.e. Badri was supposedly the name that the people who lived in
Chigachak gave to the village.

In some languages the predicate of a noun-modifying clause is non finite, e.g. the
subject relative clause in Kham (Tibeto-Burman, Bodic Branch, Nepal, see
Watters2002: 2001), Hayu (Tibeto-Burman, Nepal, see Michailovsky, 1988: 185 ff)
and Turkish (see Comrie 2006: 147 and 150), in the sense that these predicates
include no marking for person or number agreement, while cross reference does occur
on main clause predicates. Atong predicates never show cross-reference with any of
their NPs in any clause type, so the distinction finite versus non-finite is not relevant
for the language if one defines finiteness in terms of showing cross-reference. If,
however, one defines a finite verb as any verb whose form is such that it can stand in
a simple declarative sentence (Matthews, 1977: 129), predicates of attributive
clauses in Atong are finite. Although most attested forms lack inflectional predicate
head suffixes (see Table 63), and verbal forms without inflectional predicate head
suffixes usually occur in imperative clauses, they also arguably appear in declarative
clauses of which the predicate has an habitual overtone (see 18.9).
In section 29.3 it will be argued that there is no need to posit a gap in attributive
clauses in Atong. The terms pre- and post-head attributive clause will be explained in
section 29.4. Arch NPs with post-head attributive clauses are the focus of section
29.5. Because the head of the arch NP is not a clausal constituent, and can thus not
function as argument or adjunct (peripheral argument) in the attributive clause, there

539
is no such thing as an internally headed attributive clause. Besides this, section 29.6
treats the way in which genitive-marked NPs should be analysed. Section 29.7 treats
constraints in the variation of the position of the attributive clause within the arch NP.
We will see that these constraints are determined by the transitivity of the attributive
clause predicate, the number of nouns expressed in the arch NP and the animacy of
the nouns. Which semantic relationships can obtain between the head of the arch NP
and the predicate of the attributive clause will be treated in section 29.8. That arch
NPs can function as head of a predicate of a verbless equation/identity clause, like any
other noun or NP, is illustrated in section 29.9. Headless arch NPs, in which the noun
modified by the attributive clause is ellipsed, are treated in section 29.10. Headless
arch NPs can become lexicalised as we will see in section 29.11. Lexicalisations are
nominalisations of the participant and abstract type. Abstract nominalisations provide
more evidence against the presence of a gap in Atong attributive clauses.
In his 1998 (a) article, Comrie proposes the term attributive clause construction
for Asian languages that present a single grammatical construction that covers
European relative clauses, fact-S constructions (or noun-complement construction, see
for example Matsumoto (1997)), e.g. (815) from Korean (Comrie 1998 a: 52 example
(4), my bracketing and labelling) and other possibilities/interpretations, such as the
sound of knocking at the door (Comrie 1998 a: 54-55). Matsumoto (1997) has already
discussed such noun-modifying constructions in Japanese.

(815) Korean:
--------------------------attributive clause----------------------------
[[ku namca -ka ku yeca -eyke cayk -ul cwu -n] sasil]
NP
the man -NOM the woman -to book -ACC give -PRS.PRT fact
the fact that the man gave a book to the woman

Although attributive clauses in Atong do not cover the interpretations mentioned
above, other than relative clauses and other possibilities/interpretations (Comrie
1998 a: 54-55), as we can see in example (831), it would still be appropriate, in my
view, to term attributive clauses in Atong as such, because of the evidence that the
morpheme <gaba ~ ga> (ATTR) has an attributivising function on other word classes
apart from verbs, i.e. numerals, the attributive time postposition daka before, in the

540
past and the bound interrogative formative morpheme <bi-> (QF) as we shall discuss
in section 29.12. Finally, section 29.13 explicitly points out the fact that attributive
clauses in Atong are not part of a nominalisation ~ relativisation ~ genitivisation
syncretism.
29.2 No common argument
The term common argument refers to the syntactic relationship which holds
simultaneously between the head of the arch NP on the one hand and the predicates of
the matrix clause and the relative clause on the other hand. For Atong it would be a
mistake to call the head of the arch NP the common argument, suggesting that it
simultaneously partakes in the argument structure of the attributive as well as the
matrix clause. This assumption is mistaken, because it is not the head of the arch NP
that functions as argument in the matrix clause, but the arch NP as a whole. In the
English clause I eat a big apple, the noun apple is not the O argument of the clause
but the NP [a big apple] is. The head of an NP can not be seen as a separate clausal
constituent from the modifier of that head. Semantically the head of an NP denotes,
whereas an NP as a whole refers (see Lyons1977: 174 ff. for a discussion on
denotation and reference).The referent is structurally represented by the whole NP
constituent (if represented at all). Hence the head of the modified NP cannot be a
constituent of the matrix clause.
If one says that the head of the arch NP is a common argument (e.g.
Aikhenvald, 2008: 469 ff and Dixon, 2004 b: 525), this would entail that this noun is
simultaneously governed by the predicate of the attributive clause and controlling the
same predicate as a modifier within the NP. In reality the attributive clause modifies
the head noun of the arch NP. When we make a diagram of the arch NP in (811),
according to the theory that prescribes the common argument, we would get (816),
constituent analysis, or (817), dependency analysis. One has to bear in mind that cu
to be big is a verb in Atong. These diagrams tell us that Atong people are actually
thinking the big eagle is big, which is then later transformed into the big eagle
after the deletion of one of the two occurrences of eagle. The head of the arch NP is
governed by the relative clause predicate and at the same modified by it. This would
be a very strange representation of the facts attested in the language, a representation
to which I do not adhere.

541
(816) -----------arch NP----------- Consituent analysis
--AC--
[pholgom |{cu}| =gaba]
eagle big =ATTR
a/the big eagle

Before deletion After deletion In reality
NP NP NP

VP VP VP

NP VP VP
pholgom pholgom cu pholgom cu=gaba pholgom cu=gaba
eagle eagle big eagle big=ATTR eagle big=ATTR

(817) Dependency analysis of the arch NP in (811).
Before deletion After deletion In reality
eagle eagle eagle

big big big

eagle (S argument)

In Atong, the fact that the head of the arch NP is not a constituent of any clause
can be seen by the fact that it cannot be case-marked. Case marking in Atong
functions roughly as follows. S and A are always unmarked, O can be accusative-
marked when the NP is referential, adjuncts (or peripheral arguments) have to be
case-marked according to their semantic function. Case-marking of NPs in attributive
clauses is the same as in main clauses.
However, although the head of the arch NP is not a clausal constituent, in Atong,
and other languages with attributive (or relative) clauses, a semantic relationship
between the head of the arch NP and the predicate of the attributive clause can be
inferred, but no semantic relationship is grammatically required, as we shall see in the

542
next section. In Atong the factors that limit the interpretation of the possible semantic
relationship are the semantics and argument restrictions of the attributive clause
predicate, the position, semantics and case marking of other NPs within the arch NP,
the semantics of the head itself and the context. We shall discuss these factors in
detail below. Now consider the following example.

(818) a mu'aydogaba mura ga'anca.
----------------------------matrix clause-------------------------------
---------------------arch NP---------------------
----------AC-----------
[|a
S
{mu' -aydo}| =gaba mura
Location
]
S
{ga' -an -ca}
1s sit -PROG =ATTR stool good -REF -NEG
The stool on which I sit is not good.

Example (818) consists of two clauses, viz. the attributive clause, between vertical
lines, of which the intransitive verb mu' to sit is the predicate, and the matrix clause,
of which ga' to be good is the predicate. The noun mura stool is the head of the
arch NP which functions as S argument in the matrix clause. Since S arguments are
always unmarked for case, the arch NP has no case enclitic following it. The semantic
relationship that obtains between the head of the arch NP and the predicate of the
attributive clause has to be pragmatically inferred. The semantics of the predicate and
of the head prompt the hearer to understand mura stool as the thing sat on, not the
thing doing the sitting. In this example we also see that the argument of the attributive
clause, the first person singular personal pronoun a, appears in the normal argument
position before the predicate. Personal pronouns cannot function as the head of an
arch NP in Atong. Hence identification of the head of the arch NP is easy: it is the
right-most constituent of the arch NP.
As was said above, adjuncts have to be marked for case according to their
semantic role. Thus when mura stool would be a constituent of a clause, for instance
a main clause, it would have to be locative-marked. The main clause version of the
attributive clause involving the word mura stool, could be either (819) or (820),
depending on the speakers desire to make sure that the S argument is not confused
with the possessor of the stool. In Atong an unmarked personal pronoun followed by a
noun can always be interpreted as Possessor-Possessed. In (819) only one

543
interpretation is possible, viz. a I is the S argument. In (820) it is not possible to
distinguish S argument from Possessor.

(819) muraci a mu'aydoa.
[mura] =ci [a]
S
{mu' -aydoa}
stool =LOC 1s sit -PROG
I am sitting on a stool.
(820) a muraci mu'aydoa
a mura =ci {mu' -aydoa}
1s stool =LOC sit -PROG
I am sitting on a stool. or [X]
am/is/are sitting on my stool.

In (821) we again see an arch NP with a head that has the semantic role of
Location, but, whereas in (818) above, the clause precedes the head, below, the
attributive clause follows the head. We see that the noun nok house is the head of an
arch NP with an intransitive attributive clause of which the predicate is mu' to stay,
to sit. The head is not marked for case because it is neither a constituent of the
attributive clause nor of the matrix clause. The arch NP as a whole functions as S
argument in the matrix clause, which is an adverbial-marked subordinate clause. The
head should be interpreted as a Location given its semantics in combination with the
semantics of the predicate of the attributive clause. The arch NP is unmarked for case.
The first person personal pronoun a, which is the S argument of the attributive
clause, is in the same position as in the example above, i.e. in argument position,
immediately preceding the predicate. Identification of the head of the arch NP is again
easy, since it is now the left-most phrasal constituent of the arch NP, immediately
preceding the modifying clause.

(821) nok ang mu'gaba gurumok.
--------------------matrix clause----------------------
---------------arch NP---------------
-------AC------
[nok
Location
|[a]
S
{mu'}| =gaba]
S
{gurum -ok}
house 1s stay =ATTR collapse -CoS
The house in which I lived has collapsed.

In (822) the most likely interpretation of the semantic role of the noun bostu thing
with respect to the transitive predicate of the attributive clause, is that of Patient, i.e.
the thing eaten and not the thing doing the eating.

544
(822) phanan sa'rocagaba jilami bostudoraaw raay honaymo []
-----------------------------------arch NP---------------------------------
-------------------AC------------------
[|[phanan] {sa' -ro -ca}| =gaba jila =mi bostu] =dora =aw
always eat -USUALLY -NEG =ATTR district =GEN thing =p =ACC
{ra} =ay {hon'} =ay =mo
bring =ADV give =ADV =SEQ
Having brought and given things from the district which are usually never
eaten

Note that, in the above example, the head of the arch NP is not only modified by the
attributive clause, but also by a genitive-marked Possessor. Identification of the head
of the arch NP is easy: it is the right-most phrasal constituent in the arch NP.
Nouns can be modified by more than one attributive clause. In example (823) the
head of the arch noun, wa tooth, is modified by an attributive clause on either side.
In addition the head noun is modified by another noun in apposition, viz. muma
elephant (see 6.6 for possible interpretations of nouns in juxtaposition). The
semantic relationship of the head of the arch NP to the predicate of AC1 is Patient,
and to that of AC2 Attributant. The whole arch NP functions as O argument in the
matrix clause.

(823) [] nokphanday do'khakhuci khacapay tangaba momawa dora boroy
do'gabaaw ra'ay jalaokno.
--------------------------------------AC 1------------------------------------
[|[nokphanday do'khakhu] =ci {kha -cap -ay} {tan}| =gaba
bachelors.house king.post =LOC tie -ALONG.WITH -ADV put =ATTR
moma wa
Patient/Attributant
elephant tooth
--------------AC 2----------------
|[dora boroy] {do'}| =gaba]
O
=aw
weight.of.5kg four IE.be =ATTR =ACC
{ra'} =ay {jal -a -ok} =no
take =ADV run.away -AWAY -CoS =QUOT
[they] ran away taking the four dora (20 kg) weighing elephant tusk which
was tied to the king post of the bachelors house, it is said.
a
r
c
h

N
P


545
In example (824), of which we have seen a shortened version in (822), all attributive
clauses are grouped to the left of the noun they modify. The head of the arch NP has a
certain semantic relationship to the predicates of the attributive clauses AC1 and AC2
and another to the predicate of AC3. To AC1 and AC2 the head is Attributant. The
relationship between the head of the arch NP and AC3 is Patient.

(824) thawgaba somgaba phanan sa'rocagaba jilami bostudoraaw raay
hon'aymu []
--AC1-- --AC2-- -----------------AC 3-----------------
[|{thaw}| =gaba |{som}| =gaba |[phanan] {sa' -ro -ca}| =gaba
tasty =ATTR sweet =ATTR always eat -USUALLY -NEG =ATTR
jila =mi bostu
Attributant/Attributant/Patient
]
O
=dora =aw
district =GEN thing =p =ACC
{ra} =ay} {hon'} =ay =mo
bring =ADV give =ADV =SEQ
Having brought and given tasty, sweet things from the district which are
usually never eaten

In examples like (825) below it is not the noun ram road that functions as
topicalised S in the matrix clause, but the NP dajosa re'egaba ram the road that
goes to Dajong as a whole. The semantic relation of the head of the arch NP to the
predicate of the attributive clause is that of Actor.

(825) dajosa ray'gaba ramdo tuka.
------------------------------matrix clause-------------------------------
--------------------arch NP--------------------
------------AC------------
[|[dajo] =sa {ray'}| =gaba ram
Actor
]
S
=do {tuk -a}
Pname =MOB go =ATTR road =TOP overgrown -CUST
The road which goes to Dajong is overgrown.

To recapitulate, the arch NP contains an attributive clause and a modified noun,
which is the head of the arch NP. This analysis is also discussed by Lehmann (1984).
Although he did not invent the term arch NP, Lehmann describes the
relative/attributive clause construction as being a complex NP consisting of a
modifying relative clause, the Relativsatz, and the head, i.e. the noun modified by
a
r
c
h

N
P


546
the relative clause, which Lehmann calls the Bezugsnominal. Given the great
relevance of his work to the analysis proposed in this article, Lehmann is worth
quoting extensively:

[] nennen wir das Nominal, das von Bezugsnominal und RS [Relativsatz]
konsituiert wird, das hhere Nominal. Dieses, d.h. jedes Nominal, das einen
RS als unmittelbare Konstituente hat, heit Relativkonstruktion [] Es
handelt sich hier um ein endozentrische Konstruktion; das Bezugsnomen is
ihr Nukleus, der RS ihr Satellit. Ein Satellit ist, semantisch gesprochen, ein
Modifikator. Ein Satellit, der ein Nominal modifiziert und also ein
(komplexes) Nominal mitkonstituiert, ist ein Attribut. Daher sind RSe []
Attribute und heien auch Attributstze. (1984: 44)
77

The head of the arch NP does not function as constituent of either the attributive cause
or the matrix clause;
78
therefore, it cannot be case-marked. The arch NP as a whole
functions as a constituent of the matrix clause, and can be case-marked for its
syntactic function in the matrix clause when appropriate (see above). Any semantic
relationship between the predicate of the attributive clause and the head of the arch
NP needs to be inferred pragmatically. There are no grammatical constraints that force
any semantic interpretation, nor indeed the presence of a semantic relationship. The
arch NP functions in all respects as a prototypical noun and there are no restrictions
on inflection or syntactic functions that it can have in the matrix clause.

77
English Translation: [] we call the NP that consists of the head and the relative clause, a higher
NP. This, i.e. every NP that has a relative clause as constituent, is called a relative construction. The
construction is endocentric; the noun modified by the relative clause is the nucleus, the relative clause
is its satellite. A satellite is, in semantic terms, a modifier. A satellite that modifies a nucleus and
therefore participates in the construction of a higher (complex) NP is an attribute. Therefore relative
clauses are attributes, and are also called attributive clauses.
78
Here my analysis also concurs with Lehman (1984: 45) who says: Eine Subkonstituente eines
[Nominalsyntagma]s hier der Nukleus des [Relativsatz]es in Gestalt des Bezugsnomens kann keine
eigene syntaktische Funktion im Matrixsatz haben. English translation: A subconstituent of an NP
in this case the nucleus of the relative clause in the form of the modified head cannot have a syntactic
function of its own in the matrix clause.

547
In the next example, the head of the arch NP, so country, is in an Attributant
(see Van Valin and LaPolla 1997:115) function relation to the predicate of the
attributive clause. The arch NP functions as a directional adjunct in the matrix clause
and is therefore marked with the Mobilitative case enclitic <-sa> (MOB).

(826) iskon jan'gaba sosa jalaok.
---------------------------------------matrix clause------------------------------------
----------------arch NP--------------
--------AC-------
[|[iskon] {jan'}| =gaba so
Atbtnt
]
Direction adjunct
=sa {jal -a -ok}
so.much far =ATTR country =MOB run.away -AWAY -CoS
[He] has run away to such a far country.

The next examples confirm the nominal character of arch NPs. In (827) we see the
phrasal enclitic <=rara> EXCLUSIVELY, whereas in (828) the arch NP is inflected
with the phrasal enclitic <=dara> (p). Both enclitics occur exclusively on NPs. The
arch NPs in both examples are headless. Headless arch NPs are treated in more detail
in section 29.10. The headless arch NP in (827) functions as the predicate of an
identity/equation clause. This is another nominal property: all nouns can function as
the predicate of an identity/equation clause. In the example below the S argument is
ellipsed.

(827) kara khorogabararasano
-------------arch NP------------
---------AC-------
{ [ |[kara] {khoro}| =gaba] =rara =sa} =no
vein narrow =ATTR =exclusively =DLIM =QUOT
[They are] exclusively narrow veined [men], it is said.

(828) paloci jalgabadoraaw
--------------------arch NP------------------
----------------AC--------------
[|[palo] =ci {jal -a}| =gaba] =dora =aw
jungle =LOC run.away -away =ATTR =p =ACC
the ones that run away to the jungle

548
29.3 No gapping and no obligatory semantic relationship
About the examples of arch NPs in (825) and (826) above one could of course say that
the modified nouns, ram road and so country, have been extracted from the
attributive clause and that there is a gap in the attributive clause that corresponds to
the modified nouns, which has been put after the predicate. A gap represents a
syntactically obligatory omission. In other words, a gap assures a syntactic
representation, a zero, of the head of the arch NP inside a relative clause of languages
in which main clauses need all core arguments to be present in order to be
grammatically correct. However, this would make the analysis unnecessarily
complicated for Atong. I adopt Matsumoto (1997) and Comries (1998 a and b) point
of view that certain languages do not need a gap in the attributive clause to account
for any missing NP. Like Korean and Japanese, Atong makes extensive use of zero
anaphora, i.e. NPs can be left out of any clause if they are retrievable from the
context. This means that main clauses without NPs are perfectly grammatical in
Atong and no gap needs to occur in the attributive clause. Moreover, it is difficult to
determine which arguments should be conceptualised as core arguments of any
potentially multivalent verb, since all arguments can be omitted and then transitive or
intransitive interpretation depends on the context in which a clause occurs. When
there is no case marking on an NP, its semantic relationship with the predicate can
only be inferred. Even with possibly omitted A or O arguments there is no constraint,
syntactic or semantic, that forces us to posit a gap in the attributive clause in Atong.
Semantically we could imagine numerous participants, but none of these are
grammaticalised in Atong, to the point where there are syntactic constraints on their
occurrence in a clause. In English, for example, the S (intransitive subject) argument
has to be expressed in intransitive and A (transitive subject) and O (transitive object)
in transitive clauses and the A, O and Location all must be expressed in a clause with
the verb to put. In Atong we could multiply the gaps ad infinitum for lack of syntactic
proof of what should be conceived of as a core argument or obligatorily conceived
NP, which therefore should be present in a clause. In reality, the appearance of NPs in
a clause is pragmatically conditioned. Any construction has to be interpreted in a
context, which is when the relationship between the head of the arch NP and the
attributive clause becomes clear. The utterance in (829) (person kill=ATTR) can be
interpreted in three ways:

549
(829) morot so'ot =gaba
person kill =ATTR

a) The verb so'ot to kill can be interpreted as intransitive: a person who kills,
where the stated NP is the Agent, or the S argument of the intransitive verb.
b) The verb can be interpreted as transitive: a person who kills persons, in
which, according to my consultants, the stated NP is most likely to be
interpreted as the Patient and the Agent is implied.
c) The stated noun can be interpreted as Patient and no Agent is implied: a killed
person.
The construction in (829) can only be interpreted, i.e. gets a meaning, in a context. In
the context of (830), the different interpretations are listed in order from more to less
felicitous; interpretation c) is not felicitous in this context, whereas if we would
change the verb to, for example, hit, it could be felicitous.

(830) morot so'ot =gaba =aw gobormen so'ot -siga -ni
person kill =ATTR =ACC government kill -ALT -FUT
a) A person who kills, the government will kill in turn.
b) [A person] who kills persons, the government will kill in turn.
c) *A killed person, the government will kill in turn

It was already mentioned above that a semantic relationship between the head of the
arch NP and the predicate of the attributive clause is a matter of inference. Sometimes
it is impossible to say what the main clause equivalent of an arch NP could be.
What could have been gapped? In (831) we see an arch NP where different semantic
relationships between the predicate of the attributive clause and the head can be
inferred.


550
(831) kam pa'gaba morot
------------------arch NP-----------------
-------AC-------
[|kam]
S
{pa'}| =gaba morot]
work much/many =ATTR person
a person whose work is much, alternatively: a person who has a lot of work

The predicate of the attributive clause, pa' to be much/many is intransitive and it is
difficult to imagine any other participants than those already stated. However, two
different main clause equivalents (they are of course totally different clauses with
different meanings from the attributive clause) of the attributive clause in (831) can be
conceived. One where morot person is the Possessor of the kam work, as we can
see in (832), and one where the person is a Location, as in (833).

(832) morotmo kam pa'a
morot
POSSESSOR
=mo kam pa -a
person =GEN work be.much -CUST
The persons work is much.

(833) morotci kam pa'a
morot
LOCATION
=ci kam pa' -a
person =LOC work be.much -CUST
At the person the work is much..

It would be unnecessary to analyse a gap in any attributive clause in Atong because
there is no syntactic or semantic ground for it. There is nothing inside the attributive
clause that is co-referential with the head of the arch NP.
The fact that Atong attributive clauses cover the other interpretations or
possibilities type, although they do not cover the fact-S or noun complementation
type in Comries (1998 a) typology of attributive clauses, is a strong argument in
favour of calling them attributive clauses and not relative clauses.
29.4 Pre- and post-head attributive clauses
Attributive clauses that precede the head of the arch NP, as in (818), will be called
pre-head attributive clauses, whereas those that follow the head of the arch NP, as in

551
(821), will be termed post-head attributive clauses. In arch NPs with pre-head
attributive clauses the head is the right-most phrasal constituent in the NP, always
immediately following the predicate of the attributive clause. In arch NPs containing
post-head attributive clauses, the head is the phrasal constituent immediately
preceding the attributive clause. Personal pronouns that are Possessors (818) and
demonstratives (834) are always the first constituents in any arch NP, regardless of
whether the attributive clause precedes or follows the head. In the following example
we see how the distal demonstrative precedes the attributive clause while the head of
the arch NP comes after the attributive clause.

(834) ue gorogaba acu
-----------arch NP-----------
--AC--
[ue |goro| =gaba acu]
DST meet =ATTR grandfather
the old man (lit. grandfather) [whom they] met

Thus, identification of the head is easy given its position before or after the modifier.
The syntactic principles are the same in arch NPs containing pre-and post-head
attributive clauses, only the order of the constituents is different, hence it is possible
for a head to be modified by attributive clauses on either side of it, as we have seen in
(823) above. Let us now look closer at arch NPs containing post-head attributive
clauses.
29.5 Arch NPs with post-head attributive clauses
The interpretation of arch NPs containing pre-head attributive clauses is very
straightforward, since the NP immediately following the attributive clause can only be
interpreted as the head, e.g. (811). Arch NPs containing post-head attributive clauses
with a transitive predicate can be ambiguous, because they can sometimes be
interpreted as headless, as we will se below. In between those two extremes lies the
arch NP with a post-head intransitive attributive clause, such as (821), repeated here
as (835).


552
(835) nok ang mu'gaba gurumok.
--------------------matrix clause----------------------
-----------------arch NP---------------
-------AC------
[nok
Location
|[a]
S
{mu'}| =gaba]
S
{gurum -ok}
house 1s stay =ATTR collapse -CoS
The house in which I lived has collapsed.

In the example above, the verb mu' to stay is intransitive and its only argument, the
first person personal pronoun a, cannot be the head of the arch NP because personal
pronouns cannot be the head of an arch NP.
Arch NPs containing transitive post-head attributive clauses, then, can be analysed
in the same way. It is important to remember that the head of the arch NP cannot be
case-marked because it is not a constituent of the attributive clause or of the matrix
clause. When we look at example (836) we see that the head of the arch NP, bandi
stands in a (semantic) Agent relationship to the predicate of the transitive attributive
clause.

(836) bandi payanggabaaw mokren wa'thok sophin'ay grogra caysomaydoano.
---------------------arch NP-------------------
-----------AC----------
[bandi
Agent
|{pay -a}| =gaba] =aw
Pname carry.by.hand -AWAY =ATTR =ACC
[mokren wa'thok so -phin']
adverbial expression
=ay

eye bamboo.stick raise -COMPLETELY =ADV
[grogra]
A
{cay -som -aydoa} =no
Name look.at -follow -PROG =QUOT
Grynggrang
79
is watching the carrying Bandi (alternatively: Bandi who is
carrying) with eyes raised on bamboo sticks (i.e. attentively), it is said.

79
The name grogra comes from Garo and is used in Atong without breaking the clusters up with the
phoneme /e/.

553
The alternative analysis of (836) is to see Bandi as Possessor of an arch NP with an
ellipsed head, i.e. a headless arch NP (treated below). The translation would be:
Grynggrang is watching [the thing which] Bandi is carrying, but this
interpretation is not intended in this context.
80
It is identification of the head of the
arch NP that will play an important role to determine the superiority of the analysis
described above, to other possible analyses.
Identification of the head in an arch NP with a post-head attributive clause is easy
since it is always the noun that immediately precedes the attributive clause, as was
mentioned above. When there are no other non-clausal modifiers, the head is the left-
most constituent in the arch NP. When the head is modified by non-clausal modifiers
such as demonstrative, a personal pronoun or another noun, this modifier will always
come before the head. When two nouns stand in apposition they can, but do not have
to, be interpreted as the first modifying the second (see also 6.6). When this occurs
in an arch NP, pragmatics will have to make clear whether it has to do with a head
modified by the preceding noun or whether the first noun is the head and the second
noun a constituent of the attributive clause. In (837) we see how two nouns occur in
apposition. The most likely interpretation is the one where morot person, human is
the head of the arch NP and not a modifier of the following noun wa' bamboo.

(837) morot wa' sa'gaba
-----------------arch NP-------------------
------AC------
[morot |[wa'] sa'| =gaba]
person/human bamboo eat =ATTR
a person who eats bamboo, i.e. a tough person.
Unlikely interpretation: *the human bamboo which is eaten, bracketing:
[morot wa |sa?| =gaba]
person/human bamboo eat =ATTR

80
The attributive clause in example (836) is non-restrictive, as there arent lots of Bandis around in the
context to be distinguished. This example, then, shows that non-restrictive and restrictive attributive
clauses have the same structure.

554
In the next example it is clear from the context in the story from which the example is
taken that the first noun, ama mother, is a modifer, viz. a Possessor, of the second
noun, garu mustard and not the head of the arch NP, which is garu mustard.

(838) ama garu ramgabaci de'etgaba iankhonte ie.
----------------arch NP---------------
-AC-
[ama garu | ram | =gaba] =ci
mother mustard cook =ATTR =LOC
[|de'et| =gaba] {i =an} =khon =te [ie]
shit =ATTR PRX =FC/ID =SPEC =DCL PRX
This might be [the eagle that] shat in mothers cooked mustard.
Wrong interpretation in context of the story: This might be [the eagle that]
shat on the mother who cooked the mustard.
29.6 Genitive-marked A argument or Possessor? / No internal head
Consider the following example of an arch NP with a post head attributive clause. It is
important to note that the predicate of the attributive clause, hon'- to give, is
transitive. The noun mola tobacco is the head of the arch NP.

(839) ie, ami mola dolay hon'gabaaw nang'tym iaw ryngna man'cido []
[ie] [a =mi mola {dol} =ay {hon'} =gaba]
O
=aw
PRX 1s =GEN tobacco roll.up =ADV give =ATTR =ACC
[na' -tom] [i] =aw {ro} =na {man'} =ci =do
2s -ppp PRX =ACC drink =DAT be.able =LOC =TOP
If you
p
can smoke this, my tobacco that [I] give [you
p
] rolled up, []

The constituent order in the example above is Possessor Head Attributive clause.
It is important to note that the constituent order Head Possessor Attributive clause
does not occur in corpus of recorded material. The head of the arch NP is mola
tobacco This constituent is modified to the left by a personal pronoun, which is a
genitive-marked possessor, a=mi (1s=GEN) my. The nucleus is modified to the
right by the attributive clause. As we argued above, the head of the arch NP cannot be
case-marked, since it is not syntactically a constituent, hence the lack of accusative

555
marking although the semantic relationship with the predicate of the attributive clause
is that of Patient.
Since the predicate of the attributive clause is transitive, an A argument is implied,
which is the same as the genitive-marked constituent, i.e. first person singular. When
the attributive clause is pre-head, the Possessor always preceeds the attributive clause,
as we can see in (840).

(840) babami hantigaba gam jom
----------------arch NP-----------
---AC---
[baba =mi [ |{hanti}| =gaba gam jom] ] =aw
father =GEN divide =ATTR wealth riches =ACC
fathers divided wealth [and] riches

I have no recorded examples of transitive attributive clauses in which the A
argument is different from the genitive-marked constituent in the arch NP. This leads
us to suspect that the genitive-marked constituent is the A argument of the attributive
clause. This is unlikely for three reasons:
Firstly, we have already established that the impossibility to receive case marking
means that the head is not a clausal constituent. If the genitive-marked noun were a
clausal constituent, this would lead to the syntactically strange situation in which the
head of the attributive clause as non-clausal constituent appears in the middle of a
clause.
The second reason is that genitive-marked nouns also occur in arch NPs with
intransitive attributive clauses, in which case it is very easy to see that they are
possessors and not S arguments. The following example is illustrative. In this
example, the horse, gore, is the one that runs and not the genitive-marked second
person singular na'=mo.


556
(841) na'a anna na'mo gore jalna rakkhalgabaaw hon'etaribo.
[na'a] [a] =na
2s 1s =DAT
-----------------------------arch NP-----------------------------
---------------AC------------
[na =m gore |{jal} =na {rak -khal}| =gaba] =aw
2s =GEN horse run =DAT strong -SUP =ATTR =ACC
{hon' -et -ari} =bo
give -CAUS -SIMP =IMP
You just give me your horse that is strongest in running. Alternatively: You
just give me your horse that runs fastest.

Thirdly, as we can see in (844), an actor in an arch NP that is not the head, in this
case the Agent phulis police, will not necessarily have genitive marking. Therefore,
those constituents that do have genitive marking should be true genitives, and not
actors.
Having taken these arguments into account, we can conclude that the genitive-
marked noun in (839) is a Possessor and not an A argument.
81
The implied A
argument of the attributive clause is co-referential with the possessor. More fieldwork
needs to be done to find out if it is possible to have attributive clauses in which the A
argument is not co-referential with the Possessor of the head of the arch NP.

Recalling the observation that heads of arch NPs cannot be case-marked, when we
look again at (836) we can understand why bandi can only be interpreted as the head
of the arch NP, the phrasal constituent modified by the attributive clause: because it is
not genitive-marked. If bandi were genitive-marked, it could not be the head of the
arch NP and we would have a situation equal to that in (842) below, where the
genitive case prevents dokhi from being interpreted as the A of the predicate of he
arch NP and thus must be an A argument in the attributive clause of a headless arch

81
This use of genitive case marking is different from Shigatse (Haller 2000: 114) where all modified
nouns preceding the predicate of the relative clause are usually genitive-marked. Unfortunately Haller
does not explain what conditions cause the genitive-marking of nouns that precede relative clauses
predicates.

557
NP, since there is no other constituent within the arch NP to be interpreted as the
head. As was said above, genitive-marked nouns modifying the head of an arch NP
can only precede the attributive clause. In (842) we see how the genitive-marked noun
dokhi modifies the ellipsed head of the arch NP.

(842) dokhimi balgabatokoy kha'sin-kadomay re'eca.
--------------arch NP-------------
--AC--
[dokhi =mi |{bal}| =gaba] =tokoy
Name =GEN speak =ATTR =LIKE
{kha'sin kadom} =ay {re'e -ca}
slow slow =ADV go.away -NEG
Like [the words] Dykhi had spoken, [Bandi] does not go slowly.

When there would be another animate candidate for head in the arch NP, i.e. if both
potential arguments of a transitive attributive clause predicate are expressed and both
are animate nouns, we get a situation as in (843). This example shows us an arch NP
where the Agent (semantic relation to the predicate of the attributive clause) phulis
police is the head. The Patient, mobbin, who is the O argument of the attributive
clause, is thus accusative-marked. In other words, genitive case-marking
disambiguates, because a case-marked noun cannot be the head of an arch NP.

(843) phulis mobbin tokgaaw a ka'petaydo.
------------------------arch NP----------------------
---------------AC-------------
[phulis
Agent
|[mobbin]
Patient
=aw {tok}| =ga]
O
=aw
police Name =ACC beat =ATTR =ACC
[a] {ka'pet -aydo}
1s be.angry.with -PROG
I am angry with the police who beat Mobbin.

In the example above it would be difficult to perceive the police as modifier of
Mobbin, and the most felicitous interpretation is thus for the police to be the head and
Mobbin to be the argument of the predicate of the attributive clause.

558
In (843) both nouns in the arch NP are animate and the Agent is the head of the
arch NP. If the Patient is the head, as in (844), it cannot be accusative-marked,
because of the impossibility of heads to be case-marked, but the Agent can also not be
genitive-marked, or it would be a Possessor modifying the head. In example (844),
mobbin, the Patient of the predicate of the attributive clause, is the head of the arch
NP and is therefore not marked for case. The head is the left-most phrasal constituent.
The Agent and transitive subject (A) of the attributive clause, phulis police, is
unmarked for case, just as A arguments in all other clauses, and is positioned after the
head, directly in front of the predicate. These are the expected case-markings and
positions for the head of the arch NP and the NP constituents of any clause.

(844) mobbin phulis tokgaba ha'ci mu'aro.
----------------------arch NP---------------------
-----------AC---------
[mobbin
Patient
|[phulis]
Agent
{tok}| =gaba] [ha'] =ci {mu' -aro}
Name police beat =ATTR ground =LOC sit -PROG
Mobbin who has been beaten by the police, is sitting on the ground.

The above evidence shows again that genitive-marked constituents of an arch NP
are Possessors and not arguments of the attributive clause. This means that an
attributive clause modifies the head of an arch NP in exactly the same way whether
the attributive clause is preposed or postposed to the head. In other words, there are no
so called internally headed attributive clauses in Atong in which a genitive-marked
Agent as argument of the predicate of the attributive clause precedes the head of the
arch NP, thus creating the odd situation in which this head would be inside the
attributive clause.
29.7 Variation constraints in the position of the attributive clause
In cases like (844) above with two animate nouns within the arch NP with a transitive
attributive clause, where the Patient is the head and the Agent is expressed in the
attributive clause, the attributive clause will always be post-head. When there is only
one noun expressed in the arch NP and this is the head, the attributive clause can
precede or follow the head. The same is true for arch NPs in which there is also a
pronoun apart from the head, because pronouns cannot be the head of an arch NP and

559
will thus always be a constituent of the attributive clause. When a pronoun precedes
the head, it will always be interpreted as a Possessor. In Atong a personal pronoun or
other noun need not be genitive-marked to function as possessor; simple juxtaposition
is enough.
82
Example (845) is illustrative (see also (851) and (853) below).

(845) a garu ramay tanagaci
-------------------------------arch NP-------------------------------
------------------AC----------------
[a
Possessor
garu
Patient
|{ram} =ay {tan -a}| =ga] =ci
1s mustard dry.in.the.sun =ADV put -AWAY =ATTR =LOC
in my mustard (leaves) which [I] left (literally: put away) in the sun to dry

The constituent order in the arch NP of (818) can be changed to (846).

(846) mura a mu'aydogaba ga'anca.
--------------------------matrix clause----------------------------
--------------------arch NP---------------------
----------AC----------
[mura
Location
|a
S
{mu' -aydo}| =gaba]
S
{ga' -an -ca}
stool 1s sit -PROG =ATTR good -REF -NEG
The stool on which I sit is not good.

Example (811) can have (847) as variant.

(847) ucie cunggabaaw pholgom nukokno
-------------------arch NP----------------
---AC---
ucie [|{cung}| =gaba pholgom
Attributant
]
O
=aw {nuk -ok} =no
then big =ATTR eagle =ACC see -COS =QUOT
Then [he] saw the big eagle, it is said.

82
In these cases, the left noun is always the possessor and the right noun the Possessed. Juxtapositions
can also form compounds (see 6.6).

560
Finally, example (836) can be transformed into (848).

(848) payanggaba bandiaw [].
---------------------arch NP---------------------
-----------AC---------
[|{pay -a}| =gaba bandi
Agent
] =aw
carry.by.hand -AWAY =ATTR Pname =ACC
[Grynggrang
83
is watching] the carrying Bandi [with eyes raised on bamboo
sticks (i.e. attentively), it is said.]

Thorough investigations in the field have made me conclude that there is no
difference in meaning between arch NPs of which the head is modified by a preceding
or following attributive clause.
It appears that there is a strong tendency for both transitive and intransitive
attributive clauses that contain adjuncts (peripheral arguments) to be pre-head, e.g.
(849). In this example the attributive clause contains a Location adjunct. The head is
the noun sotmay fly.

(849) phaltha khu'cukci dumgaba sotmay
----------------------------------arch NP-------------------------------
-----------------------AC--------------------------
[|[phaltha khu'cuk]
Location
=ci {dum}| =gaba sotmay]
self mouth =LOC swarm =ATTR fly
the flies that swarmed in his own mouth

The following example shows one of only two examples in the corpus of an arch NP
with post-head attributive clause that contains an adjunct.

83
The name grogra comes from Garo and is used in Atong without breaking the clusters up with the
phoneme /e/.

561
(850) bolsi sene abek akankhambayci tangaba
----------------------------------------arch NP---------------------------------------------
------------------------AC------------------
[bolsi sene abek
84
|[akan khambay] =ci {tan}| =gaba]
year seven drinking.spoon rack.above.fire top =LOC put =ATTR
a seven year [old] drinking spoon which had been put on top of the rack
above the fire
29.8 Attested attributivisations
An attributivisation is the semantic relationship that obtains between a noun
modified by an attributive clause and the predicate of the attributive clause. So far we
have seen Location (818), (821), Patient (822), (823), (824), (839), (840), (844), (845)
(850), Actor,(825), Agent (836), (841), (843), (849), and Attributant (811), (812),
(824), (826) attributivisations. Together with Attribute (823), Target (851), Instrument
(852) and Beneficiary (853) attributivisations, these are the ones attested in Atong.
Example (831) shows a type of arch NP where there is no semantic relationship
between the head and the predicate of the attributive clause.

84
It has to be mentioned that (850) is only one of two recorded appearances of a modifying
construction like this, viz. bolsi sene abek (year seven drinking.spoon) a seven year old drinking
spoon. The noun bolsi year is an autoclassifier (see 12.3) quantified by the numeral sene seven.
The quantified noun functions as a modifier to the following noun. The text from which this example is
taken is an epic story told in a register that is far from colloquial and very difficult to understand, even
for native speakers. This particular speech register could well be the reason that this type of
modification of a noun shows up here. The other recorded example of this construction comes from the
same story, viz. bolsi sene=mi cow (year seven=GEN liquor) seven year old liquor, but with a small
difference: the modifier is genitive-marked. The colloquial way of saying that something is seven years
old requires an attributive clause with the identity/equation copula do'- ~ do- (IE.be): cow bolsi sene
do'=gaba (liquor year seven IE.be=ATTR) liquor which is seven years old.

562
Target attributivisation:
(851) a jagimi kha'galgaba baju
--------------arch NP--------------
-----AC-----
[a jagi
Possessor
=mi |{kha'gal}| =gaba baju
Target
]
1s life =GEN love =ATTR friend
the friend whom I love my [whole] life Litterally: my lifes beloved friend.

Instrument attributivisation:
(852) rodo maharimu takruknagaba bostuaw tansetay jalphinaoknowa.
------------------------------------arch NP---------------------------------
---------------------AC
85
----------------------
[|[rodo mahari] =mu {takruk}=na| =gaba bostu
Instrument
]
O
=aw
Name family =COM fight =DAT =ATTR thing =ACC
{tan -set -ay} {jal -phin -a -ok} =no -wa
put -DISPOSE.OF -ADV run.away -back -AWAY -COS =QUOT -FACT
[They] ran away back, leaving behind the things for fighting with the
Rongdyng family, it is said.

Beneficiary attributivisation:
(853) a taka hon'gba morot uci gana.
---------------------arch NP-----------------------
---AC--
[[a taka] |{hon'}| =gaba morot
Beneficiary
]
S
[u] =ci {gana}
1s money give =ATTR person DST=LOC exist
The person to whom [I] gave my money is there.

The meaning of the above example has to be determined by the context. Without
context three interpretations are possible, including the one given above. The other
two interpretations are: The person who gave me money is there and The person
who gave my money (to someone else) is there.

85
The dative enclitic <=na> (DAT) marks the clause as purposive (see Chapter 27). Thus the purposive
clause rodo mahari takruk=na in order to fight with the Rongdyng family is attributivised as a
whole.

563
Given the miscellaneous collection of possible relationship that are attested in
Atong between the head of the arch NP and the predicate of the attributive clause, it is
my impression, that, like in the Japanese and Korean attributive clauses, any
relationship is possible, i.e. that there are no grammatical restrictions on what can be
attributivised upon, if the context is specific enough to interpret it correctly. More
fieldwork is needed to find out if this impression is right.

The verb no- to say signals direct speech reports. The verbatim repeated words
are the Quote of a speech report construction. Although Quotes are not attested as
heads of arch NPs, they can be a constituent of an attributive clause with no say as
predicate, as in (854). In this example the arch NP is headless and the implied head is
the speaker who says the quote. The whole arch NP functions as Beneficiary in the
matrix clause.

(854) aya! awado na' man'wate nogabana ma tham tanaaydoanowa.
---------------------------------arch NP---------------------------
---------------------------AC-----------------------------
-----------------------Quote--------------------
[|[aya {awa =do na' man' -wa} =te] {no}| =gaba]
Beneficiary
=na
interj uncle =TOP fish get -FACT =DCL say =ATTR =DAT
[ma tham] {tan -a -aydoa} =no -wa
CLF:ANIMALs three put -AWAY -PROG =QUOT -FACT
For [the one who] said Hey! Uncle got fish! [he] left three [fishes] behind,
it is said.
29.9 Arch NPs as predicates of verbless clauses
Verbless clauses are of the identity/equation type. Arch NPs, i.e. head noun plus
modifying attributive clause, can function as predicates of verbless clauses, just like
any NP or prototypical noun. In (855) we see a prototypical noun, raja king, as head
of a predicate. Other examples of nominal predicates are given in 22.5.

564
(855) aan raja
[a] =an {raja}
1s =FC/ID king
I am the king.

Just like other NPs, an arch NP does not need to be marked for aspect, modality,
mood or polarity, as we see in (856) and (857). In those examples the arch NPs are
headless, as are the majority of arch NPs that were recorded as predicates. The S or
Topic of the clause, of which the arch NP is the predicate, is always coreferential with
the implied head of the arch NP. An arch NP can also function as S or Topic of a
verbless clause, as is shown in (861).

(856) ado usami paragaba.
----------------------predicate-----------------
-------------------arch NP-----------------
---------------AC-------------
[a] =do { [ | [u]=sa =mi {para}| =gaba] }
1s =TOP DST=MOB =ABL travel =ATTR
I am [someone who] has travelled from there. Alternatively: As for me, [I
am someone who] has travelled from here.

(857) na'tom thol'ramay balgabae.
------------------predicate----------------
-----------------arch NP--------------
-------------AC-----------
[na -tom] { [ | {thol'am -ay} {bal}| =gaba] } =e
2s -ppp lie -ADV speak =ATTR =FC
You
p
are liars. Literally: You [are persons] who speak lyingly.

The following example shows a nominal predicate that is change-of-state-marked.
The arch NP is headless. The context is as follows. The wild animals meet the fox on
the road while they are running away in fear of the lazy king. The fox asks the
animals why they are afraid of the lazy king. The animals say that the lazy king is
very brave, since he sat all alone in the banyan tree and jumped out to fight with them.


565
(858) thoroksoraok una, nido jalgabak.
--------predicate--------
-----arch NP-----
--AC--
{thorok -sora -ok} una [ni] =do { [ |{jal}| =gaba] -k}
jump.out -TOTALLY -CoS then 1pe =TOP run.away =ATTR -CoS
[He] jumped out, then we became [the ones] that ran away.

Example (859) is an illustration of a headed arch NP functioning as predicate. The
head of the arch NP is raja king. The S argument of the verbless identity/equation
clause is the interrogative ca who.

(859) ca ana daygaba raja?
------------------------predicate----------------------
-----------------------arch NP---------------------
------------AC---------
[ca]
S
{ [ | [a] =na {day}| =gaba raja
Attributant
] }
who 1s =DAT be.bigger=ATTR king?
Who is a greater king than me?
29.10 Headless arch NPs
When the head of the arch NP is not expressed, the arch NP is headless. The NP that
the attributive clause modifies is only implied. The ellipsed NP and its intended
relation to the predicate of the attributive clause must be deduced from the context
and the restrictional properties on semantic roles of the predicate of the attributive
clause. In most of the recorded cases, the ellipsed NP, head of the arch NP, is in an
Agent role relationship to the predicate of the attributive clause. Other semantic roles
can of course also be implied, of which some are shown in the examples below, where
the attributive clauses are underlined.


566
Implying a Location NP
(860) phanan rupek mu'gabaacido toy gana.
-----------arch NP-----------
[phanan] [|[rupek] {mu'}| =gaba] =ci =do [toy] {gana}
always frog stay =ATTR =LOC =TOP water exist
At [the place where] a frog stays, is always water.
Incorrect interpretation: *At the sitting frog is always water.

Implying an Agent NP with transitive attributive clause predicate
(861) iaw balgabae derus ar marak.
-------------arch NP------------
[|[i] =aw {bal}| =gaba] =e {derus ar marak}
PRX =ACC tell =ATTR =FC Name Sname1 Sname2
The one who tells this [is] Derus R Marak.

Implying an Object NP
(862) ian balgabaaw jametarinaka.
------arch NP-----
[i] =an [|{bal}| =gaba] =aw {jam -et -ari -naka}
PRX =FC/ID speak =ATTR =ACC finish -CAUS -SIMP -IFT
This, [the story] which [I] told, [I] will now just make it come to an end.

Implying an Attributant NP
(863) jalaaymo konsado jan'gabamo otokoy olrukokno:
------arch NP------
{jal -a} =ay =mo} [konsa] =do [|{jan'}| =gaba] =mo
run.away -AWAY =ADV =SEQ after =TOP far =ATTR =GEN
[otokoy] {ol -ruk -ok} =no
like.this speak -RC -COS =QUOT
After having run away, from [a place which is] far, [they] spoke to each other
like this.

567
Headless arch NPs are the most frequently attested form in the collected fieldwork
data, followed closely by post-head attributive clauses, while both of these types
greatly outnumber pre-head attributive clauses in the language.
86

29.11 Lexicalisations
There are constructions that look like headless arch NPs, but it is impossible to find
any implied ellipsed NP. Moreover, these headless arch NPs have a fixed,
unpredictable meaning and the verb does not function as a predicate any more, i.e. it
cannot take any arguments or verbal modifiers. These constructions are
nominalisations. Lexicalised nominalisations with the attributive enclitic <=gaba ~
=ga> (ATTR) can be participant or abstract nominalisations, body parts, objects,
artefacts, places and persons. Most of the recorded lexicalisations are participant
nominalisations. The meaning of kha'gal=gaba (love=ATTR), in example (864), is the
abstract notion love and is therefore an abstract nominalisation. One could say that
the attributivised verb is the head of the NP.

(864) o came, ami na'na kha'galgabaau na'mi khatoci da'etna
man'phanima?
------------------------ NP----------------------
---AC---
[o] [came] [a =mi [na'] =na |khagal| =gaba] =aw
interj sweetheart 1s =GEN 2s =DAT love =ATTR =ACC
|[na' =mi kha'tho] =ci {da -et}| =na
2s =GEN heart =LOC enter -CAUS =DAT
{man' -pha -ni} =ma
be.able -IN.ADDITION -FUT =Q
O sweetheart, will you be able to insert also my love for you into your heart?

86
Out of the 317 arch NPs counted for this thesis 146 are headless (46%), 120 are post-head (38%), and
51 are pre-head (16%). I did not count elicited and incomprehensible examples, but only the ones
that come from spontaneous speech.

568
Lexicalised headless attributive clause predicates never take any aspect or modality
suffixes. Examples of lexicalised verbs with the attributive enclitic <=gaba ~ =ga>
(ATTR) are listed in Table 75.

Table 75 Examples of lexicalised attributivised verbs
lexical item gloss of parts meaning
mokca=gaba fancy=ATTR sweetheart, someone you fancy
khom=gaba marry=ATTR spouse
cicu=gaba blister=ATTR a blister
nangthay=gaba swell=ATTR abscess
bas ne=tak=gaba bus rest=do=ATTR bus stop
cal=gaba support=ATTR a support
rin=gaba keep.as.domestic.animal=ATTR fishery
koro=gaba make.noise=ATTR sound
okgona=gaba pregnant=ATTR pregnancy
ha'bace=gaba begin=ATTR beginning
tak-sak=gaba do-appropriately=ATTR help

Abstract nominalisations provide more evidence for the lack of gap in attributive
clauses in Atong. As we have seen in (831), if there is a head of the arch NP there
need not be a semantic relation between it and the predicate of the attributive clause.
In the case of abstract nominalisations, we see that a head does not even have to be
implied and that the nominalised verb has no arguments.
29.12 The morpheme <-gaba ~ -ga> as attributive suffix
As was already said in the introduction, attributive clauses in Atong do not cover fact-
S clauses like in Japanese and Korean, and other possibilities/interpretations type
clauses, such as the smell of meat cooking (Comrie 1998 a: 57). The reason to call
noun modifying clauses in Atong attributive clauses and not relative clauses is
that the same morpheme <gaba ~ ga> is used as attributiviser on other word classes.
The morpheme <-gaba ~ -ga> as attributive suffix occurs on i) numerals, ii)

569
interrogatives and iii) the time word daka before, in the past.
87
The reason that the
morpheme <-gaba ~ -ga> is analysed as a suffix in this section and not as enclitic, is
that the scope of the morpheme is not the NP but the lexical item itself.
i Numerals
Suffixed to numerals, the morpheme <-gaba ~ -ga> (ATTR) derives ordinal numbers
e.g. boroy four boroy-ga (four-ATTR) fourth (see also 11.6). Ordinal numbers
are nominal modifiers. This means that the suffix <-gaba ~ -ga> (ATTR) has an
attributivising function, transforming a numeral into an entity that can modify nouns.
The following example is illustrative.

(865) gonigaba so badri maydugotom.
[goni -gaba so] {badri maydugotom}
two -ATTR village Pname
The second village is Badri Maidugytym
ii The bound interrogative formative
The suffix <-gaba ~ -ga> (ATTR) can be combined with the bound interrogative
formative morpheme <bi-> (QF) to form the interrogative bigaba ~ biga which?
(see also 9.15), e.g. (866). The suffix has an attributivising function in this case, just
as with the numerals.

(866) bigaaw biskut ra'nima?
[bi -ga =aw biskut]
88
{ra' -ni} =ma
QF -ATTR =ACC biscuit buy -FUT =Q
Which biscuits shall I buy?

87
The same morpheme also occurs as adverbialiser suffix on the relative time postposition daka
before, in the past (see 14.2.4), as derelationaliser suffix on kinship terms (see7.1) and
unproductively on certain other nouns as relational suffix (see 7.1), e.g. the word nokgaba landlord
can be analysed diachronically as nok-gaba (house-RELATIONAL).
88
The accusative case enclitic in Atong is the only case marker that does not need to occur on the last
element of the NP. It also often happens that more than one NP constituent is accusative-marked. The
prosody of example (866) is that of a normal clause without any pause or hesitation.

570
iii The time word daka
The time word daka before, in the past (see 14.2 for more details about this
lexeme), but not kosa after, is frequently attested with the suffix <-gaba ~ -ga->
(ATTR). The lexeme daka-gaba (before-ATTR) means the first, and can thus
function attributively to nouns, as we can see in example (867).

(867) dakagaba boba alu kay'aymo []
[daka -gaba boba] [alu] {kay' =ay =mo}
before -ATTR crazy.person potatoes plant =ADV =SEQ
The first crazy person, having planted his potatoes []
29.13 The nominalisation ~ relativisation ~ genitivisation syncretism
Many Tibeto-Burman languages use the same morpheme or morphemes to mark
nominalised clauses, relative clauses and genitive. Although this seems to be a very
common pattern in Tibeto-Burman languages, it does not hold for Atong, where these
functions are marked differently, and where attributive clauses are not
nominalisations, as we shall see below. Let me give a brief overview of what has been
written in the Tibeto-Burman literature about the nominalisation ~ relativisation ~
genitivisation syncretism.
Matisoff (1972) describes the three functions of the Lahu morpheme ve, viz.
relativisation, nominalisation and genitive. Noonan (1997) describes the impressive
range of different functions of the morpheme <-wa> in Chantyal, among which are
nominalisation and relativisation. (Bickel (1999) introduced the term Standard Sino-
Tibetan Type Nominalisation (SSTN) inspired by the work of Matisoff, (1972),
Genetti (1992) and others. Bickel comments on their work by saying that: It is well-
known that in many if not most Sino-Tibetan languages relative clause and
attribute/genitive markers are identical with nominalisation devices and that sentences
bearing such markers can also function as independent utterances [] (1999: 271).
In most Bodic languages, however, relative clauses are nominalised clauses marked
by the genitive (Noonan 1997, DeLancey 1999). DeLancey (1999: 233) proposes a
basic pattern of TB relativisation: the use of a nominalised clause to modify a noun.
Since the clause is syntactically a nominal, it is typically marked by the genitive when
it is subordinate to another nominal.

571
As was said above, nominalisation, genitive and relativisation, here called
attributivisation, are marked differently in Atong. Attributive clauses in Atong are
marked with the clausal enclitic <=gaba ~ =ga> (ATTR), which is different from the
genitive/ablative/nominaliser enclitic <=mi ~ =mo> (GEN/ABL/NR) (see 20.4) and
example (868) below. The genitive can also function as action/state/object
nominaliser, labelled (NR), on clauses of which the predicate head is marked by the
factitive suffix <-wa> (FACT) (see 20.4 and 24.3.1), e.g. (868) and (869) below. In
these examples the verbal roots bal speak, say, tell and sa' eat, marked by the
factitive suffix <-wa> (FACT) and the genitive enclitic <=mi> (GEN); these
nominalised claues function as heads of NPs. In (868) we see that this head can be
modified, in this case by a Possessor a=mi (1s=GEN) my, and in (869) we see a
nominalised clause with accusative case-marking <=aw> (ACC).

(868) ami balwami ician jametwa.
[|a =mi bal -wa| =mi] [i] =ci =an {jam -et -wa}
1s =GEN talk-FACT =NR PRX =LOC =FC/ID end -CAUS -FACT
I will end my talking here.

[pheru]{nuk -sega -ak} =no [|sa' -wa| =mi] =aw
This time the fox saw [it], it is said, the food.

An attributive clause in Atong is not a nominalisation since it cannot function as the
head of an NP, except when it is lexicalised (as was treated in 29.11). Attributive
clauses are modifiers within an NP. This chapter provided ample examples in which
this fact was demonstrated.
29.14 Conclusion
As I found the terminology used in the canonical literature about relative clauses
confusing and unsatisfactory to describe attributive clauses in Atong, I set out to find
a more appropriate and transparent way in which this grammatical phenomenon can
be analysed. This led me to coin the term arch NP to designate the grammatical unit

572
which comprises the attributive clause and the noun it modifies, i.e. the head. The
arch NP as a whole functions as constituent in the matrix clause.
There is no such thing as a common argument. The head of the arch NP is
always unmarked for case, just because it is neither a constituent of the attributive
clause nor of the matrix clause. For the same reason, there is also no such thing as an
internally headed attributive clause in Atong, where the head of the arch NP occurs
within the attributive clause surrounded by its constituents.
It is not necessary to posit a gap in the attributive clause, because Atong allows
zero anaphora or ellipsis of NPs, when they are retrievable from the context. I see a
gap as a zero representation of a syntactically required NP in the attributive clause. It
is not possible for the head of the arch NP to be at the same time governed or
controlled by the predicate of the attributive clause and to be modified by it. There are
no grammatical constraints that force the inference of NPs when these are not
expressed in the clause.
A semantic relation between the predicate of the attributive clause and the head of
the arch NP can, but does not have to be inferred.
Attributive clauses in Atong do not cover interpretations of what Matsumoto
(1997) calls noun-complement constructions, and Comrie (1998 a and b) terms
fact-S constructions, e.g. (815). Despite these two shortcomings, it would still be
appropriate, in my view, to consider attributive clauses in Atong as such. Firstly,
because Atong attributive clauses are used to cover constructions with other
possibilities/ interpretations (Comrie 1998 a: 54- 55). Secondly, because of the
evidence that the attributive morpheme <gaba ~ ga> (ATTR) has an attributivising
function on other word classes apart from verbs, i.e. numerals, the attributive time
postposition daka before, in the past and the bound interrogative formative
morpheme <bi> (QF).
.

573

574
Appendix 1 Texts
_____________________________________________________________________

This section presents five texts of different genres. The first two texts represent the
spontaneous speech of two and three unmarried men respectively. The third text is an
informative text, telling us about the summoning of spirits. The text that follows is an
incantation. The last text is a fictional story about a lazy king. All texts are
represented almost exactly as they were recorded. The only alternations that are made
are the removal of hesitations, false entries and unnecessary repetitions. These
alternations are, however, very few in number. Apart from the incantation, all texts
are glossed and translated. The incantation, being untranslatable according to my
friends, is not recognisable as either Atong or Garo and might be some language that
is only used by priests in incantations.

TEXT 1 Saduthamaran mo'tham
The three brothers-in-law whose wives are sisters
part 1

Saduthamaran part 1 and part 2 are the transcriptions of two short films taken by
one of my friends, Samrat N Marak, in Siju in the summer of 2007 with my Sony
digital photo camera. The films are improvised plays but treat serious every day
matters in the lives of the main characters Songken, Jongken and Nongken two of
which feature in Part one, i.e. Songken and Nongken. The speakers age lies
somewhere between 16 and 23 years old. Because the plays are improvised, the
language is spontaneous and colloquial and therefore the films give an excellent
impression of the every day conversational speech of unmarried men from Siju.
The word sadu is a reciprocal kinship term indicating the relation of men whose
wives are sisters (see Chapter 1). The title saduthamaran mo. tham is
morphologically analysed as follows: sadu-OWN-RC CLF:HUMAN three three brothers-
in-law whose wives are sisters.

TEXT 1 Saduthamaran mo' tham Part 1

575
1. Songken
aaah jow =na sok -arok =te. coi' -a.
interj sleep =DAT want -PROG =DCL tired -IMPF
Oh! [I] want to sleep. [I]m tired.

2. Nongken
ayaw!
interj
Jeez!

3. Songken
ie radi jadi tak -a. biba =an ray'a -naka?
PRXC Name crazy.woman do -IMPF when =FC/ID come -IFT
This Radi is acting like a fool. When will [she] be coming?

4. johan de'et =na re'e -wa?
Name shit =DAT go.away -FACT
Johan went away to shit?

5. Nongken
m'm. cabi =ba. ho'o.
yes key =ADD yes.
Yes. And the key (Implies And he went about the key). Yes.

6. Songken
ayaw! cabi bi =ci =n tan -a -ok? na'na =e, ie mm
interj key QF =LOC =FC/ID put -AWAY -COS 1pi =FC PRX interj
Oh! Where did [X] put the key? We, this um

7. a =do ie na'na ato =aw, hay' =aw, golpho thari =ga =aw
1s =TOP PRX 1pi what =ACC GPN =ACC story prepare =ATTR =ACC

8. lapstori, a =mi gomon tari =na sok -aroa =com. asol =an!
love.story 1s =GEN about prepare=DAT want -PROG =IRR really =FC/ID
As for me, this our, what whats it?... prepared story, the love story, I want to
prepare [it] about me. Really!

9. ga'su =ay =sa tari -ni =com. thari -ok =odo
splendid =ADV =DLIM prepare-FUT =IRR prepare-COS =TOP
I should prepare it really splendidly. When Ill have prepared it

10. Nongken
ho'o
yes
Yes

11. Songken
ca?
who
Who?

576

12. Nongken
hayda.
I.dont.know
I dont know.

13. Nongken and Songken
hu!
interj
Hey!

14. Songken
hu! ca?
interj who
Hey! Who?

15. Nongken
hayda.
I.dont.know
I dont know.

16. Songken
aya!
interj
What a pity!

-Hindi singing-

17. Songken (speaking into the camera)
ie, ie, i =do mamu =an do' -khu -ca.
PRX PRX PRX =TOP nothing =FC/ID IE.be -INCOM -NEG
He, he, he is nothing yet.

18. ie burbok =taka =do i =do.
PRX idiot =LIKE =TOP PRX =TOP
Hes like an idiot, [yes] he.

19. i =soi, nawa =taka =soi.
PRX =MIR fool =LIKE =MIR
[It is] him to my amusement, [he] to my amusement is like a fool!

(Speaking to Nongken)


577
20. silat =e atakna manchi =aw wat -ok?
89

Pname =FC why Pname =ACC send.away -COS
Why did Silat send Manchi away? (i.e. Why did Silat break up with Manchi?)

21. kam =ni duk =ni. nem -pha =com ue.
wealth -WITHOUT grief -WITHOUT good -IN.ADDITION =IRR DST
Good for nothing, no grief. [She] was supposedly no good either, she.

22. mu'thay =ba cu -a -arok, te'ew =e.
bosom =EMPH big -WITHOUT.HOLDING.BACK -PROG now =FC
[Her] breasts are getting really big, though, now.

23. te' =do wat =na na -ca =com!
now =TOP send.away =DAT need -NEG =IRR
You should not have sent her away!

24. atakna wat =na na'a? [pause] ma?
why send.away =DAT 2s Q
Why send her away, oh, you? Well?

25. Nongken
gawi =an thik -an -ca.
girl =FC/ID right -REF -NEG
That girl was not decent.

26. Songken
o thik -an -ca?
interj:ACKNOWLEDGEMENT right -REF -NEG
Oh, she was not decent?

27. Nongken
ho'o.
yes
Yes

28. Songken
otok -ari -a, te'ew -rawraw =mi gawi =do.
do.like.that-SIMP -IMPF now -CONTINUOUSLY =GEN girl =TOP
Yes. They do like that, the girls from now on. (i.e. todays girls)

89
Silat and Manchi are not the real names of the persons involved. Their names have been altered for
reasons of privacy.

578
29. Nongken
hm'm.
interj
Yeah.

30. Songken
u =na -n som -sak =na na -a -ro.
DST=DAT =FC/ID follow -APPROPIATELY =DAT need -IMPF =EMPH
You have to be careful with them, really. Literally: You have to follow them
appropriately.

31. Nongken
na' =mi =ba thik -an -ca -khon =te.
2s =GEN =ADD right -REF -NEG -SPEC =DCL
Yours might not be decent either, I tell you.

32. Songken
m'm. kam ni' -wa na'a, nuk -a =no =na
yes wealth NEG.be -FACT 2s see -IMPF =QUOT =DAT

mokca -arok -ona.
fancy -PROG -DESI
Yes. They are worth nothing, because [whoever] they see, it is said, theyll want
to fancy.

33. nuk -wa =no -wa =na
see -FACT =QUOT -FACT =DAT

mokca -damdam -ari -a =te,
fancy -ONE.AFTER.THE.OTHER -SIMP -IMPF =DCL

te'ew -rawraw gawi =e.
now -CONTINUOUSLY girl =FC
It is said that because [whoever] they see, theyll fancy one after the other, Im
telling you, the girls of nowadays.

34. cancip -ay, te'ew na' =aw mokca -ni,
suppose =ADV now 2s =ACC fancy -FUT

u =mi =do alaga =aw mokca -naka.
DST=GEN =TOP other =ACC fancy -IFT
Suppose, now [a girl] will fancy you, after that shell certainly fancy someone
else.

35. otok -ram' -ari -a.
do.like.that-INADVERTENTLY -SIMP -IMPF
It just inadvertently happens like that. Alternatively: It just happens like that
and there is nothing we/they can do about it.


579
36. otokoymu bi =aw mokca =na otok =ga =ra (Garo) =aw =e?
CONJ which =ACC fancy =DAT do.like.that=ATTR p =ACC =FC
So which ones am [I] supposed to fancy, those who do like that?

36.a ga'su -ca ma' ca. kam =ni'.
cool -NEG interj NEG wealth =PRIV
Its really not cool. Worthless

37. Nongken
ri'gol!
penis (this is the form of the word when it is used as swearword)
Dick!

38. Songken
ca ri' ? na' ri'gol ma a ri'gol? aaah! toi sala!
who penis 2s penis Q 1s penis interj interj idiot
Whos a dick? Are you a dick or am I a dick? Argh! Jeez, idiot!

39. Nongken
johan!
Name
Johan!

40. Songken
ca?
who
Who?

41. Nongken
tarak =bo =to na'a!
Hurry up, oh you!

42. Songken
di'su =na =ba sasok sasok tak -arok.
piss =DAT =EMPH feel.an.urge RED do -PROG
[He]s feeling an urge to piss.

-Hindi singing-

580
TEXT 2 Saduthamaran mo'tham
The three brothers-in-law whose wives are sisters
part 2

1. Jongken
o morot =dora! ato tak -aro?
interj person =p what do -PROG
Hey people! What are [you] doing?

2. Nongken
atakna ray'a -wa?
why come -FACT
Why have you come?

3. Jongken
o, golgol -aro otoken, harat -wa =na =sa.
interj roam -PROG just.like.this lazy -FACT =DAT =DLIM
Oh, [I]m just roaming like this, just because Im lazy.

4. Nongken
mu' =bo. ca cini =ba co'oysa ra =bo =to.
sit =IMP tea sugar =EMPH a.little give =IMP =MPEMPH
Sit. Give some tea and sugar, come on.

5. Songken
ca ray'a -wa?
who come -FACT
Who has come?

6. Nongken
ue, u =sa =mi. bimu ato mo -wa?
DST DST=MOB =GEN name what call.a.name -FACT
He, from over there. Whats [your] name?

7. Jongken
hay'e na'a, joken.
whatchamatsallit 2s Name
Um, oh, you! Jongken.

581
8. Nongken
joken.
Pname
Jongken.

9. Songken
ooo. ca =ma?
interj:ACKNOWLEDGEMENT tea =Q
Oooh. Tea?

10. Nongken
ho'o.
yes
Yes

11. Songken
oo. ha', ro =bo.
inter:jACKNOWLEDGEMENT take.this drink =IMP
Ok. Take this, drink.

12. Jongken
ayaw na -ca -wa =com =te.
interj need -NEG -FACT =IRR =DCL
Jeez, that should not have been necessary, really.

13. Nongken
ro =bo!
drink =imp
Drink!

14. Songken
biskut =an i =tara =an -ok =ay bay'.
biscuit =FC/ID PRX =EXCLUSIVELY =FC/ID -COS =POS friend
There are only these biscuits [left], friend. Lit. The biscuits have become only
these really, friend.

15. Jongken
om.
AFFIRMATIVE
Ok.
16. Songken
atakna na'a ray'a -wa =com?
why 2s come -FACT =IRR
Why might you have come?

17. Jongken
ni' -wa otokoy =an golgol -ari -wa, a =do te'ew -mama
NEG.be -FACT like.that =FC/ID roam -SIMP -FACT 1s =TOP now -ONLY


582
may sa' -wa =do sa' -wa =com ca =ba.
rice eat -FACT =TOP eat -FACT =IRR tea =EMPH
Nohting, [I] just roamed around like that, I have just now eaten rice and tea.

18. Nongken & Songken
o, sa' =bo!
inter:jACKNOWLEDGEMENT eat =IMP
Oh, eat!

19. Jongken
harat -wa.
lazy -FACT
[Im] lazy.

20. Songken
kha'sin sa' =bo. ie na'na hay' =ci dorego =e bewal
slowly eat =IMP PRX 1pi GPN =LOC Pname =FC maybe

jan' -a =ma?
far -IMPF =Q
Eat slowly. This Our, this, eh, Dorenggo, is it far, by chance?

21. Jongken
hayda =e a =ba taw =do taw -a -khu -ca.
I.dont.know =FC 1s =EMPH go.up =TOP go.up -AWAY -INCOM -NEG
I dont know, even I have not yet gone up [there].

22. otokciba morot =dora bal -wa jan' -ok =khon =te.
but people =p say -FACT far -COS =SPEC =DCL
But people say that it might be very far for sure.

23. Songken
ho'o mo. ayaw.
yes CONF interj
Yes, of course. Jeez!

24. Nongken
a =ba nuk =na sok -aro =com
1s =EMPH see =DAT want -PROG =IRR
Even I would like to see it.

25. Songken
na'a mokca kha'gal -wa ni' -wa =ma gana?
2s fancy love -FACT NEG.be -FACT =Q exist
Do you have [someone] you fancy [and] love or not?

26. Jongken
ayaw te'ew -dabat =do ni' -way.
interj now -until =TOP NEG.be -FACT.POS
Jeez, untill now [I] havent.

583
27. Songken
ni' -wa.
NEG.be -FACT
[I] havent.

28. Jongken
na'a.
2s
[What about] you?

29. Songken
ga =do gana =com, gawi thogi -ok.
exist =TOP exist =IRR girl betray -COS
There certainly was, but not any more, the girl betrayed [me].

30. Jongken
gawi thogi -ok?
girl betray -COS
The girl betrayed [you]?

31. Songken
ho'o
yes
Yes.

32. Nongken
a =do ni' -sora.
1s =TOP NEG.be -TOTALLY
As for me, [I] totally dont have [one]. Literally: As for me, [Im] a total not-
haver.

33. Songken
hay sigoret hon' -et -sora na'a u =aw.
come.on cigarette give -CAUS -TOTALLY 2s DST=ACC
Come on, give cigarettes!, oh you, those.

34. Nongken
ooo! oh! awan -arok a =do.
interj interj forget -PROG 1s =TOP
Oh! Ah! I forgot. Lit. Im forgetting.

35. Songken
ha wal' +bot hon' -et -sora =bo =to.
hey fire +drive give -CAUS -TOTALLY =IMP =IMPEMPH
Hey, give matches too!

36. Nongken
ha! ro =bo. ayaw!
take.this drink =IMP interj
Take this! Smoke. Jeez!

584
37. Jongken
a -ba daka canci -a tak -wa =com
1s =EMPH in.the.past think =ADV do -FACT =IRR

gawi otokoy tak -wa =nan aa.
gilr like.that do -FACT =DAT.FC/ID 1s
I also used to think like that because girls dolike that, me.

38. Songken
ie radi =ba ato tak -sora -ok ie?
PRX Pname =EMPH what do -TOTALLY -COS PRX
What the hell did this Radi do?

39. Nongken
dokom sa -a a =do.
head hurt -IMPF 1s =TOP
My head hurts.

40. Songken
ie jok =ba sa' =ba otok -rum -ari -a na'a.
PRX spouse =EMPH child =EMPH do.like.this -ALL -SIMP -IMPF 2s
The women and children are all doing like like this, oh you!

atakna ie? ha! ro -sora =bo =to.
why PRX take.this drink -TOTALLY =IMP =IMPEMPH
Why this? Take this! Smoke it all, go on.

41. Jongken
a =do ca ro -aro.
1s =TOP tea drink -PROG
Im drinking tea.

42. Songken
(to Nongken) robo. (to Jongken) ca ro -ce =bo =to.
drink tea drink -FIRST =IMP =IMPEMPH
(to Nongken) Smoke. (to Jongken) Drink tea first, go on.

43. Nongken
(to Jongken) ro =bo =to.
drink =IMP =IMPEMPH
Drink!

44. Songken
na'a, sadu =e.
2s man.who.married.my.wifes.sister =FC
Oh you, sadu!


585
45. Nongken (about Songkens sigaret)
tho' sa hon' -ari =bo sadu.
longitudinal.half one give -SIMP -=IMP man.who.married.my.wifes.sister
Just give me half.

46. Songken
o do' -ari -ni =ba. a =do na'na jok
interj:ACKNOWLEDGMENT IE.be -SIMP -FUT =EMPH 1s =TOP 1pi spouse
Sure, thatll be all right. As for me, our wives

na' jok =ba ray'a -khu -ca =khon =te.
2s spouse =EMPH come -INCOM -NEG =SPEC =DCL
Your wife has maybe not come yet?

47. Nongken
'm. hm'
thats.right
Nope.

48. Songken
'mmm.
interj:PENSIVE
Mmm.

49. Jongken
a =do golgol -ro -ca -wa =na =sa te'ew nok =ci
1s =TOP roam -USUALLY -NEG -FACT =DAT =DLIM now house =LOC

ra wa =ay =mu kam kha' =na harat -ok,
rain rain =ADV =SEQ work do =DAT reluctant -COS

man'dok -ok kam kha' =na.
difficult -COS work do =DAT
As for me, precisely because [I] usually roam [around] now [that] it is raining,
Ive become reluctant to work at home, its very difficult to work. Alternatively
It has become difficult to work.

50. Songken
ho'o u =do na' =do banthay morot -ok =ona
yes DST=TOP 2s =TOP bachelor person -COS =DAT
golgol -a =ci =ba atakna ie?
roam -WITHOUT.HOLDING.BACK =LOC =INDEF why PRX
Thats right. As for that [doing work], because you are a bachelor person, if you
roam around as much as you like, why this [doing work]?

51. Nongken
ho'o
yes
Thats right.


586
52. Jongken
ni' -wa. harat -ay =sa.
NEG.be -FACT lazy =ADV =DLIM
Nothing, just lazy.

53. Songken
'hmmm. u.! dokom sa -phin -a a =do.
interj PUFFING.SOUND head hurt -FULLY -IMPF 1s =TOP
Sigh. Oofff! My head totally hurts.

nokda =mi gomon otokoy cancie =ci =do
people.that.live.in.one.house =GEN reason like.that think =LOC =TOP
Because of the people that live in your house, it you think like that

54. Jongken
'hm! kam =ba ato kam =aw kha' =ay mu' -naka ie?,
interj work =EMPH what work =ACC do =ADV stay -IFT PRX
GRUNT! And then that work, what work will he be doing while he stays here?

54.a banthay =ci =ba taka poysa na' na -arok =ona
bachelor =LOC =INDEF money money 2s need -PROG =DAT
when you are a bachelor, because you need money'

55. Nongken
ho'o, com.
yes IRR
Yes, supposedly

56. Songken
te'ew ni me'apha tak -gaba =ra (Garo) =ba alamala
now 1p married.man do =ATTR =p =EMPH a.little.bit

i =ba, i =mi ha'selsa khom =na na -wa na' a =ba.
PRX =ADD PRX =GEN for.no reason marry =DAT must -FACT 1pi =EMPH
Now that we are like married man, this because of this we had to marry for no
reason, you and me.

57. Nongken
sa' mo' sa ba' =ay =mu, man'dok -arok.
child CLFHUMAN one be.born =ADV =SEQ difficult -DUR
After one child has been born, it is difficult.

Songken
man'dok -asol -a ho'o. ie jenkon =para =ba
diffult -truely -IMPF yes PRX Name =&co =EMPH

ray'a -sora -ca -k.
come -TOTALLY -NEG -COS
Difficult indeed, yes. This Jenkon and those associated with him never come any
more.

587
58. Nongken
wel'a wel'a cugalgal -wa =sa ga' -naka =com.
quickly RED grow.up -FACT =DLIM be.compelled -IFT =IRR
[He] will almost certainly be compelled to grow up quickly.

jekon =para =ba ray'a -khu -ca =khon?
Name =&co =EMPH come -INCOM -NEG =SPEC
Jenkon and company might not have come yet.

59. Jongken
hayda ray'a -khu -ca -aro -ni =khon.
I.dont.know come -INCOM -NEG -PROG -FUT =SPEC
I dont know. He has not come yet but he might still be coming.

60. Songken
sayno =mi naw =aw a =do khom -thiri -ni,
Name =GEN younger.sister =ACC 1s =TOP marry -AGAIN -FUT
I will again marry Seinos younger sister,

a jok khom =ami khambay =ci =an.
1s spouse marry =NOM top =LOC =FC/ID
?

khom =ay =mu a =do i =sa bot -ai -ni -khon.
marry =ADV =SEQ 1s =TOP PRX =MOB lead -TOWARDS -FUT SPEC
After marrying [her], I will maybe bring [her] here.

61. Jongken
'hm
interj
Ok.

62. Songken (giving Nongken his cigarette)
ha' sadu.
take this man.whose wife is your wifes sister
Take this, sadu.

63. Nongken (taking the cigarette)
o.
interj:ACKNOWLEDGEMENT
Ok.

64. Songken
ayaw!
interj (maybe of boredom, maybe because the speaker is worried, reason not sure)
Oh


588
65. Jongken
may sa' -thok -ok =ma na -tom =e?
rice eat -ALL -COS =Q 2s -ppp =FC
Have you all eaten?

66. Songken
o sa' -ak =ba. sa' -ak =com =ay, na'a sa' -ak?
interj eat -COS =EMPH eat -COS =IRR =pos 2s eat -COS
Yea, [I] have indeed. [I] supposedly have indeed, have you eaten?

67. Jongken
a =do sa' =do sa' -ak.
1s =TOP eat =TOP eat -COS
I sure have eaten.
68. Songken
hm'm.
no
No.

69. Jongken
may sa' =nan ja'bek wal'bek thaw -ca -wa =nan.
rice eat =DAT.FC/ID curry burnt.curry tasty -NEG -FACT =DAT.FC/ID
As for eating rice, the curry was burnt, it was not tasty.

70. Songken
ue pipuk =an okha kham' =ay mu' -aroa a =do.
DST belly =FC/ID full burn =ADV stay -PROG 1s =TOP
That stomach is full [and/but] keeps burning, as far as Im concerned.

71. Nongken
ue gumuk =an. okhi -an -ca.
DST all =FC/ID hungry -REF -NEG
All of that. [I]m not hungry.

72. Jongken
ato =aw sa' -wa no na'?
what =ACC eat -FACT say 2s
Tell, oh you! what did you eat?
73. Songken
ja'bek =an ga'su -an -ca na'a ni =ba.
curry =FC/ID splendid -REF -NEG 2s 1p =ADD
Our curry wasnt nice either, oh you.

589
TEXT 3 Way khuruta
To perform an incantation

Told by Genda R Marak,
in the village of Siju in 2006

Having been a priest himself, before he turned to the Christian religion a few years
before I did my first fieldwork on Atong, Mr. Genda R Marak tells us about the art of
spirit incantation. Spirits are summoned by priests to cure the sick. As we will read, it
is by no means a cheap practice.

1. ni sosorek =do, ni ato =do daka =do
1pe heathen =TOP 1pe Atong =TOP in.the.past =TOP

mamo thorom =aw ni' wa =mi somay =ci =do
nothing religion =ACC not.exist -FACT =GEN time =LOC =TOP

way =aw mani -a.
spirit =ACC worship -CUST
We heathens, we the Atong, in the past, in times when there was no religion, [we]
worshipped spirits.

2. way =aw mani -wa =mi oltho =do thama cay -a,
spirit =ACC worship -FACT =GEN meaning =TOP devination look -CUST

kamal =ci thama cay -a.
priest =LOC devination look -CUST
The meaning of spirit worship is divination, devination is practised at the
priests.

3. thama cay =ay =mu ato way do' -ok.
devination look =ADV =SEQ what spirit IE.be -COS
Having practised devination, [we can see] which spirit has appeared.

4. cancicop =ay way cu=gaba do' -ok.
suppose =ADV spirit big =ATTR IE.be -COS
Suppose a big spirit has appeared. Literally: Supposingly a big spirit has
appeared.

5. way cu =gaba do' =ci =do purun ra' -a.
spirit big =ATTR IE.be =LOC =TOP goat get -CUST
If it is a big spirit, we buy [lit. get] a goat.

TEXT 3 Waykhuruta

590
6. purun ra ='ay =mu ma'su ra' -a.
goat get =ADV =SEQ cow get -CUST
Having bought [lit. gotten] a goat, we buy a cow.

7. ma'su na -a, wak na -a, taw' na -a, cow na -a
cow need -CUST pig need -CUST chicken need -CUST liquor need -CUST

way khurut =na.
spirit perform.an.incantation =DAT
[You] need a cow, [you] need a pig, [you] need a chicken, [you] need liquor in
order to perform an incantation.

8. umido u =aw kamal sandi -ni.
Then [you] will search for the priest.

u =aw wa'phek =gumuk, wa' pan =gumuk
DST =ACC type.of.small.bamboo =all bamboo wood =all

thari -tholo =ay =mu =sa,
prepare-NICELY =ADV =SEQ =DLIM

san sa dothamanca thari =ay =sa,
day one especially prepare=ADV =SEQ

kamal =na rok =ay =sa, way khurut -a,
priest =DAT chase =ADV =DLIM spirit perform.an.incantation -CUST

ni acu ambi =mi ni =mi pi'sa =ci =do.
1pe grandfather grandmother =GEN 1pe =GEN childhood =LOC =TOP
Only after having prepared all that small bamboo, only after having nicely
prepared the bamboo and the fire wood, only on an especially prepared day, [you]
call for [lit. chase] a priest [and] perform the spirit ncantation, in the [time] of
our ancestors [lit. grandfather and grandmother] in our childhood.

9. otokoymu kamal khurut -ni.
so.then priest perform.an.incantation -FUT
So then, the priest performs the incantation.

10. na'a way cu =gaba =aw nuk -ok no =ay canci =bo,
2s spirit big =ATTR =ACC see -COS say =ADV suppose =IMP
ma'su ra' -naka, purun ra' -naka, taw' ra' =na na -ni,
cow get -IFT goat get -IFT chicken get =DAT need -FUT

wak ra' =na na -ni, u =na =do.
pig get =DAT need -FUT DST=DAT =TOP
Suppose you have seen a big spirit, [you] will get a cow, [you] will get a goat,
[you] will need to get a chicken, [you] will need to get a pig, for him [that is].

TEXT 3 Waykhuruta

591
11. umi cow =ba som' =na na -ni, ue kamal =na.
then liquor =ADD soak
90
=DAT need -FUT DST priest =DAT
Then [you] will also need to take out some liquor for the priest.

otokoymu kamal =na cow som' =ay hon' =ay =mu aro
so.then priest =DAT liquor soak =ADV give =ADV =SEQ and

u =na =ba may jabek ra' -a =na na -ni,
DST=DAT =ADD rice curry get -AWAY =DAT need -FUT
kamal =na =ba.
priest =DAT =EMPH
So then, having taken out and given the liquor ot the priest, [you] will also have
to give him rice and curry, to the priest.

12. otokoymu taka =ba kharay =ci coy -ni.
so.then money =add big.pan =LOC offer -FUT
So then [you] will also offer money in a big pan.

13. kolgok =e coygok =e je sokon gana
twenty =monetary.unit ten =monetary.unit whatever many exist

pal =ci =do sot boa =ba con -ari -a, kamal =na =do.
sell =LOC =TOP hundred five =ADD offer -SIMP -CUST priest =DAT =EMPH
[They] will just offer twenty [or] thirty [rupees?] whatever [you] have, if [you]
sell [it], they just offer fifty, to the priest.

14. kamal =do u =an kamal =do doda hokhot -a -naka.
priest =TOP DST=FC/ID priest =TOP alone come.out -AWAY -IFT
The priest, that priest, will come out alone.

khurut =gaba =mi niam =aw otokoy bal -ni =ne.
perfom.an.incantation =ATTR =GEN rules =ACC like.this tell -FUT =TAG

khurut =gab =mi niam =do:
perfom.an.incantation =ATTR =GEN rules =TOP
[I] will tell the rules for the incantation like this, ok, as for the rules for the
incantation:

90
The verb som'- here refers to the way in which liquor is gotten out of the pot it is prepared in, called
gora. You slowly dip a hollow spoon, called abek with repeated small up-down movements deeper
and deeper into the filter, janti, which stands in the middle of the gora. When the spoon in full, you
take it out carefully and drink out of it or empty it in a glass.
TEXT 3 Waykhuruta

592
15. ca'masa =mi way do' =ci =do,
downstream =GEN spirit IE.be =LOC =TOP

kambaysa =mi way =aw =do man' -pat -ca -wa.
upstream =GEN spirit =ACC =TOP be.able -CROSS -NEG -FACT
When the spirit of downstream [along the Symsang river] appears, [he] will not
have any influence on the spirit of upstream.

16. u -tom =do sima gana, way =ba. na' =mi eria thokthok =tokoy.
DST-ppp =TOP limit exist spirit =ADD 1pe =GEN area exactly =LIKE
As for them, spirits too have limits, exactly like our areas.

17. kambaysa =mi way nuk =ci =do
upstream =GEN spirit see =LOC =TOP

dotha khurut -a.
different perform.an.incantation -CUST

ca'masa =mi way nuk =ci -do
downstream =GEN spirit see =LOC =TOP

dotha khurut -a.
different perform.an.incantation -CUST
If [you] see the upstream spirit, a different incantation if performed, if [you] see
the downstream spirit, a different incantation is performed.

18. mani mo -wa =an, hap =aw =an
worship call.a.name -FACT =FC/ID place =ACC =FC/ID

dotha dotha moa -a, thokthok moa -a.
different RED call.upon -CUST according.to.the.division call.upon -CUST
As for what [we] call the worshipping, different places are called upon, [they] are
called upon accodring to the division.

19. ie ca'masa =mi way khurut =ci =do
PRX downstream =GEN spirit perform an incantation =LOC =TOP

ue hoy =sa =mi =aw baglades =mi thol'
DST REM =MOB =GEN =ACC Bangladesh =GEN up.to

koos =mi jaria ha'gorsak =gumuk =aw =an moa -ni.
Kongos =GEN influence the.lot =all =ACC =FC/ID call.upon -FUT
When [he] summons the downstream spirit, that [priest] will call upon the
influence of all those far away [places] up till Bangladesh [and] the influence of
Kongos, of them all.

TEXT 3 Waykhuruta

593
20. baghmara takmara gumuk =aw =an moa -ni u =sa [long, fasetto voice]
Baghmara RED all =ACC =FC/ID call.upon -FUT DST=MOB

roara toara gumuk =aw =an moa -ni.
Rongara RED all =ACC =FC/ID call.upon -FUT

jaksoram taksoram ha'gorsak =aw =an moa -ni.
Jaksonram RED all =ACC =FC/ID call.upon -FUT
[He] will call upon Baghmara Takmara all of them way over there Rongara
Tongara [he] will call upon [them]all, Jaksongram Taksongram [he] will call upon
all of them.

21. cagaba man' =ay sa' -a cagaba nokda tak -a,
whoever in.great.amounts =ADV eat -CUST whoever family do -CUST

u =mi bimo =gumuk =aw =an thal =ay
DST=GEN name =all =ACC =FC/ID clearly/explicitely =ADV

moa =ay =mu =sa,
call.upon =ADV =SEQ =DLIM

way khurut =ay =mu, sa' =ay ro =ay =mu,
spirit perform.an.incantation =ADV =SEQ EAT =ADV drink =ADV =SEQ

nem -khal =ci =ba nem -khal -ca =ci =ba
good -CP =LOC =INDEF good -CP -NEG =LOC =INDEF

ue morot =na =do dokdok =sa cay -sak -ni.
DST person =DAT =TOP for.a.short.while =DLIM look -APPROPRIATELY -FUT
Whoever is rich [lit. eats in great amounts] whoever has a family only after
having called clearly/explicitely upon all their names, having performed the
incantation, having eaten and drunk, whether or not [the patient] has improved,
[they] will wait for a short while for that person. Alternatively: [they] will wait
for a little while to see if the patient has improved or not.

22. ue nem -sak -ca, nem -sak -ca tak =ci =sa,
DST good -APPROPRIATLY good -APPROPRIATELY -NEG do =LOC =DLIM

mo' sa kamal tak -thiri -ni.
CLF:HUMANS one priest do AGAIN -FUT
He is not appropriately well, only when [the patient] is not appopriately well,
another priest will do [it] again.

23. otokoymudo u =an ge'the sa =gaba:
so.then DST=FC/ID 3s be.ill =ATTR

a =do nem -khal -an -ca no =ci =do
1s =TOP good -CP -REF -NEG say =LOC =TOP

TEXT 3 Waykhuruta

594
aro kamal =sa thama cay -thiri -a.
and priest =LOC devination look -again -cust
So then, as for that sick [person], if [he] says: I am not better, [they] will
practice diviantion again at the place of another priest.

24. khurut =na sap =gaba morot =aw =sa
perform.an.incantation =DAT know.a.skill =ATTR person =ACC =DLIM

so =gumuk =ci =ba so =ci pa' -ram -ari -a.
village =whole =LOC =EMPH village =LOC be.many -FORTUITOUSLY -SIMP -CUST
As for precesely those people [who] know how to perform incantations, in all
villages [and] in this village, [there are] many [of them] for no good reason.

25. u =ba pa' =ay =do sap -dam -ca.
DST=EMPH many =ADV =TOP know.a.skill -TRULY -NEG
Many of them dont truly know [how to do it].

mo' sa mo' ni tokoy sap -a.
CLF:HUMANS one CLF:HUMANS two =LIKE know.a.skill -CUST
[They only] one-or-two-ingly know the skill.i.e. only one or two of them know
the skill.

26. kamal =na =do jesaba wal duk san duk =ba
priest =DAT =TOP wherever night sorrow day sorrow =ADD

rok -ari -ni, khurut =na;
chase -SIMP -FUT perform.an.incantation =DAT
[People] will search anywhere for a priest, whether it is day or night, to perform
an incantation;

27. sam =na je tak =ay rok =gaba =tokoy rok -a,
medicine =DAT whatever do =ADV chase =ATTRT =like chase -CUST

kamal =na =do.
priest =DAT =TOP
just like a medicine searches searches for medicine in whatever way, [they
search] for a priest.
91

91
This passage can be interpreted as follows: A skilled priest is searched with great zeal just like
someone searches for medicine if a person is sick and will die.
TEXT 3 Waykhuruta

595
28. otokoymu khurut =ay =mu cancicop =ay nem =ok.
so.then perform.an.incantation =ADV =SEQ suppose =ADV GOOD -COS
So then, having performed the incantation, [the patient] has supposingly
improved.

29. ayaw na' khurut =ay =sa no =ay =mu
interj:SURPRISE 2s perform.an.incantation =ADV =DLIM say =ADV =SEQ

kamal =aw =ba mothel -a
priest =ACC =ADD thank -CUST

aro thama cay -thum =gaba =aw =ba
and devination look -ON.BEHALF.OF.SOMEONE.ELSE =ATTR =ACC =ADD

mothel -a.
thank -CUST
Having said: Wow! Because of your incantation [I am better], [he] thanks the
priest and [he] thanks [the person who] has practiced the devination on his behalf.

30. i =an a =mi bal =gaba =aw =ba jam -et -ari -naka.
PRX =FC/ID 1s =GEN speak =ATTR =ACC =EMPH finish -TRANS -SIMP -IFT
As for this, [I] will now just end my story.

596
TEXT 4 Ca'masami way
The downstream spirit

An incantation by Genda R Marak
in the village of Siju in 2006
This text presents part of the incantation used to summon the downstream spirit, i.e.
the spirit who dwells in the area downstream of the Symsang river. My friends did not
want to translate this text, saying that it contained too many difficult words, that it was
not real language, or that the language was too complicated and ultimately, that it was
Hachyk (Garo). To me the incantation appears mostly incomprehensible I recognise
some words that exist in Atong and many words that look like Atong words but are
just a bit different. The language is not recognisable as Garo either. It might be some
language that is only used by priests in incantations. The text is presented here below
to preserve some of the pre-Christian culture for future generations.
The stressed vowels in each line are underlined. Stress is obtained by means of
intensity and a lower pitch. The phenomenon of prolonged consonants to mark stress
(as described in 2.9) does not occur in this incantation. Most lines consist of two
times an equal amount of syllables, although some lines have additional syllables
before or after the first rhyme. Most lines are around twelve syllables long, but there
is a lot of variation. The rhymes occurs with the last syllable or syllables of each half
of a line, e.g. line 1 kusuma ginggpha ja'sonmal congdpha, where the
bigger, bold syllables rhyme. The incantation presents a variety of different meters
usually consisting of two or four of the same type of foot, but there are meters that
consist of different types of foot. We find the dactylic meter, e.g. ku.su.ma
gi.go.pha, the trochaic meter, e.g. line 8 ian nag'na (i=an na' =na
PRX=FC/ID 2S=DAT this [is] for you), the iambic meter, e.g. line 14 ka.mal kha.li
go.ra khan.di.
Apart from these we find meters that consist of prolonged iambic feet and of
prolonged trochaic feet. A prolonged iambic foot cosists of four syllables, with the
accent falling on the second syllable. Line 13 is an example of such a meter: gokabara
bansogetok. A prolonged trochaic foot consists of four syllables, with the accent
falling on the first syllable. Line 12 is an example of a meter with prolonged trochaic

597
feet: cawdonaymu cawgataymu. The same words, or words that are almost the
same, can be stressed in different ways to create different meters, compare line 42 to
line 67:

42 hon'cuaydok hon'botaydok (hon'cu-PROG hon'bot-PROG)
67 hon'cuokay hon'botokay (hon'cu-COS-POS hon'bot-COS-POS)

Both lines contain predicates with the stem hon'cu and hon'bot of which I have no
translation, and it is not in the Garo school dictionary (Nengminza 2001). Although
the predicates have different suffixal morphology, the stems have the same number of
syllables in each line. However both stems are stressed differently in the two lines,
viz. on the second syllable in 42 and on the first in 67.
For a better understanding of the origin, context and meaning of incantations such
as the one presented here, I will need to learn more about the culture that preceded the
advent of Christianity. I hope that future fieldwork will provide me with this
opportunity.

1. oooy! aaah!
2. kusuma gigopha ja'sonmal congdongpha
3. ha'jo ba'gabaw udi me'gaba,
4. seek ha'sona pukil cigana,
5. na'na,
6. khajananaka na'na miluwanaka,
7. aaah!
8. ian nag'na,
9. na'na ian,
10. aaah!
11. agaraaw matgacamaw,
12. cawdonaymu cawgataymu,
13. gokabara bansogetok,
14. kamal khali gora khandi,
15. na'na raqatwataymu ra'tawaymu,
16. cakophagyna beraruguna,
17. na'na hon'cuaydok hon'botaydok,
18. aaah!
19. na' ruraaysa kokhalaysa,
20. jakromaysa jakdaaysa,
21. aaah!
22. ku'karekwa ku'kabolwa,
23. ca'aw napjawa, roaw thojawa,
24. aaah!
25. dawqde kha'sidokbo bolnindokbo,
26. aaah!

598
27. dawqde jajumaciba jakhariciba,
28. na'ba jakromku'nam jakdaku'nam,
29. ruraku'nam kondaku'nam,
30. jaksepku'nam jaknolku'nam,
31. na'na ian a'na matcu salna do'bok,
32. hon'cuaydo hon'botaydoa,
33. na' karaw rokwana nawa somana,
34. na'na hon'cuaydokay, hon'botaydokay,
35. jumasaba karisaba,
36. na'ba ja'khu rokkhunambay bido sokunambay,
37. na'na kaja ni'ok meluwa'ak,
38. caw'pa gona beraru gona,
39. ceksi balbo wa'ma thapdo,
40. sotho ro'ay sobakjakaymu,
41. na'na ian,
42. hon'cuaydok hon'botaydok,
43. aaah!
44. ka'sindokbo bolnindokbo,
45. ca'aw naphakal'na roo thopha'khal'na,
46. bugi koksothe'wacina, wa'ma se'kisawacina,
47. jama rakphakhal'na jagi thaphakhalna,
48. ha'watphabo dedatphabo,
49. aaah!
50. ian na'na te'ewdo,
51. na'na,
52. growa'thariaw khi'mi sawariaw,
53. ra'suaymu ra'botaymu,
54. na'na ian khajan ni'aydok beri doaydok,
55. hon'cuaydok hon'botaydok,
56. te'ewdo na'ba ka'sindokphaboay bolnindokphaboay,
57. aaah!
58. jajumaciba jakhariciba,
59. rurakhunam kon'daku'dam,
60. jaksepku'nam jaknolku'nam,
61. da'na khajam ni'ok na'na miluwa'ak,
62. na'na gorowa' thariaw khi'mi sawariaw,
63. ra'suaymu ra'gataymu,
64. goka boyra bangsogotokaw,
65. aaah!
66. saw'donaymu saw'gataymu,
67. hon'cuokay hon'botokay,
68. ie na'na!

599
TEXT 5 Alsia Raja
The lazy king

Told by Tonton M Sangma
in the village of Badri Maidugytym in 2005

A typical fictional narrative about a lazy king, who is also somwhat stupid. The king
runs away from home, and gets into all sorts of unfortunate situations. Ultimately he
ends up in the jungle where he has an appointment to fight with a tiger. Due to a series
of incredible coinsidences, the king escapes unharmed and victorious.

1. te'ew =e alsia raja =no, so dam sa =ci.
now =FC lazy.person king =QUOT village CLF:VILLAGES one =LOC
Now, a lazy king, it is said, in a village.

2. otokoymo alsia raja so dam sa =ci
so then lazy.person king village CLF:VILLAGES one =LOC

no =ay =sa.
say =ADV =DLIM
So then. I just said: A lazy king in a village.

3. kam kha' =na toen harata =no =aro ue, alsia =e.
work do =DAT very reluctant =QUOT =EMPH DST lazy.person =FC
He is very reluctant to do work, it is said, the lazy one.

4. otokoymo jok =ba mo' ni khom -a =no =ro.
so.then spouse =EMPH CLF.HUMANS two marry -FACT =QUOT =EMPH
So then, he is married to two wives, it is said.

5. jok mo' ni khom -a =no.
spouse CLF:HUMANS two marry -FACT =QUOT
[He] is married to two wives, it is said.

6. otokoymo sa' =na =ba jok pai =na na -a
=no,
so.then eat =DAT =ADD spouse carry.by.hand =DAT need -CUST =QUOT
So then, to eat, [his] wives have to carry [him] on [their] hands, it is said,

7. jow =na =ba jok pay =na na -a =no.
sleep =DAT =ADD spouse carry.by.hand =DAT need -CUST =QUOT
and to sleep, [his] wives have to carry [him] on [their] hands, it is said.

TEXT 5 Alsia raja

600
8. konsa morot =dora conok -phin -aydok =no =e.
later person =p look.down.upon -COMPLETELY -PROG =QUOT =FC
Later, the people are looking very much down on him, it is said.

9. ie alsia raja atokoy khe -aydok?
PRX lazy.person king how live -PROG
How does this lazy king live?

10. atokoy =an jok =aw haldun =na man' -aydok?,
how =FC/ID spouse =ACC feed =DAT be.able -PROG

no =ay =mu morot cancip -aydok =no =ro, morot =dora.
say =ADV =SEQ person think -PROG =QUOT =EMPH person -p

How can [he] feed his wives? they said and thought,it is said, the people.

11. otokoymo konsa conuk =te cunok =te
so.then later look.down.upon =DCL look.down.upon =DCL
So then, [they] are looking down, looking down [upon the lazy king]

12. cunuk =gaba =aw na =ay =mo alsia raja =e:
look.down.upon =ATTR =ACC hear =ADV =SEQ lazy.person king =FC
Having heard [those that] look down [upon him], the lazy king [says]:

13. na aa otokoy coli =e coli -sem -ca -aydok,
interj 1s like.this succeed =FC succeed -CERTAINLY -NEG -PROG

14. no =ay =mo te'ew =ba jok =mo jal -aydok =no.
say =ADV =SEQ now =EMPH spouse =abl run.away -PROG =QUOT
Well, like this I am certainly not succeeding. (Literally, as far as
succeeding is concerned, I am certainly not succeeding.) [he] said and now he
has ran away from his wives, it is said.

15. coli =e coli -ca -aydok =no
succeed =FC succeed -NEG -PROG =QUOT
As far as succeeding is concerned, [he] is certainly not succeeding, it is said.

16. jok -mo jal -a -ok =no.
spouse -ABL run.away -AWAY -COS =QUOT
[He] has ran away from his wives, it is said.

17. te'ew =e jal -a -wa =ci =e
now =FC run.away -AWAY -FACT =LOC =FC
Now when [he] runs away

TEXT 5 Alsia raja

601
18. jok =aw aset =ay jal -a -wa =ci =e de
spouse =ACC dispose.of =ADV run.away -AWAY -FACT =LOC =FC interj

sore -a -ok =no =ro.
travel -AWAY -COS =QUOT =EMPH
He disposed of his wives (i.e. divorsed) and when he ran away, well, he
travelled away, it is said.

19. taka =noy mamu taka ni' -wa aro
money =PRIV nothing money not.exis -FACT and

sa' =na ro =na =ba =noy.
eat =DAT drink =DAT =ADD =PRIV
Without money, no money at all (lit. no money exists) and without food and
drink.

20. te'ew ro ni sa'khaw -a -wa =no =khon,
now CLF:ROUND.THINGS two steal -AWAY -FACT =QUOT =SPEC

gawigaba =me =aw.
wife =GEN =ACC
Now, he might have stolen two rupees, it is said, from his wife.

21. otokoymo nogol =ci =na phet -ok =no.
so.then market =LOC =DAT arrive -COS =QUOT
So then he arrived at a market, it is said

22. ma'su gari dam =aw so' -ok =no ue.
cow vehicle price =ACC ask -COS =QUOT DST
He asks about bullock carts, it is said.

23. ie ma'su gari dam =e biskon?
PRX cow vehicle price =FC how.much
How much is this bullock cart?

24. hoy sala, ie alsia raja =e taka =e
interj interj PRX lazy.person king =FC money =FC

ni' =e ni' -wa.
not.exist =FC not.exist -FACT
Oh Damn! that lazy king has absolutely no money!

25. mamu =tokoy =an do' -an -ca.
nothing =LIKE =FC/ID IE.be -REF -NEG
He as absolutely nothing!

26. na' =na =to =do ro baa phal -ari -ni.
2s =DAT =ONLY =TOP CLF:ROUND.THINGS five sell -SIMP -FUT
Just for you I will sell [it] for five rupees.

TEXT 5 Alsia raja

602
27. aya dam no'm -a =te ma'.
interj price soft -CUST =DCL interj
Jeez! the pice is cheep, goodie!

28. sala aa ama =para =mi sa'khaw =ay =mo
interj 1s mother =&co =GEN steal =ADV =SEQ

ray' -naka san abun =ci, canci -aydok =no.
go -IFT day other =LOC think -PROG =QUOT
Damn! having stolen money from my mother and her company, I can will go
away the next day, he is thinking, it is said.

29. otokoymo canci =ay =mo te'e =do gari nuk -ok =no =aro.
so.then think =ADV =SEQ now =TOP vehicle see -COS =QUOT =EMPH
So then, having thought like this, he saw a vehicle, it is said.

30. ie gari biskon?
How much is this vehicle?

31. nehi (Indic)na'a taka ni' =e ni' -wa.
no 2s money not.exist =FC not.exist-FACT
No! You have absolutely no money! Literally: your money, as far as not
existig is concerned, does not exist.

32. na' =na =do ro coygok phal -ari -ni,
2s =DAT =TOP CLF:ROUND.THINGS ten sell -SIMP -FUT

canci -ok =no.
think -COS =QUOT
I will just sell [it] to you for ten rupees, he thought, it is said.

33. tokoymo: sala aa ama =para =mo taka sa'khaw =ay =mo
so.then interj 1s mother =&co =GEN money steal =ADV =SEQ

ra' =na na -ni gari aro ma'su gari =aw.
get =DAT need -FUT vehicle and cow vehicle =ACC
So then: Damn! Having stolen money from my mother and her company
(i.e. his mothers house ot household), I will need to get/buy a vehicle and a
bullock cart.

34. te' =do ro baa sa'khaw =ay
now =TOP CLF:ROUND.THINGS five steal =ADV

ray'a -k jow' =para =mo =aw.
come -COS spouse =&co =ABL =ACC
Now, he went and stole five rupees from his mothers house.

TEXT 5 Alsia raja

603
35. sa'khaw =ay ray' -wa =ci =e te'ew =e
steal =ADV go -FACT =LOC =FC now =FC

so' -thiri -ok =no.
ask -AGAIN -COS =QUOT
Having gone and stolen, now, he asked again, it is said.

36. ie gari bisken?
How much is this vehicle?

37. how na'a ra' =na =e so' =e se' -ca =gaba tak =na =e
interj 2s get =DAT =FC ask =FC ask -NEG =ATTR do =DAT =FC

so =e so' -ca =gaba.
ask =FC ask -NEG =ATTR
Hey! you are only pretending to be someone to asks to buy it, someone who
pretends to ask.

38. ah hazar sa.
interj thousand one
Ah, one thousand.

39. che -wa =mo dam =do hazar sa -ak =ona, hazar sa.
begin -FACT =GEN price =TOP thousand one -COS =DAT thousand one
Because the price was one thousand to begin with: one thousand.

40. aya do' -taw -an -ca -k =te a taka =do.
interj be.enough UPWARDS -REF -NEG -COS =DCL 1s money =TOP
Jeez! It is not enough any more, my money!

41. sala ma'su gari =do aa man' -ni
interj cow vehicle =TOP 1s get -FUT

dam komi -khal =gaba =aw =do, canci -aydoa =no.
price cheap -CP =ATTR =ACC =TOP think -PROG =QUOT
Damn! As for a bullock cart, I will get [one which is] cheaper, he is
thinking, is is said.

42. so' -ok =no: ie ma'su gari biskon?
ask -COS =QUOT PRX cow vehicle how.much
He asked again: How much is this bullock cart?

43. ra' =na =e tak =e tak -ca =gaba
get =DAT =FC do =FC do -NEG =ATTR

maja =an so' =e so' -man' -ok na'a domdam
yesterday =FC/ID ask =FC ask -ALREADY -COS 2s fortuitously

TEXT 5 Alsia raja

604
bal -ari =gaba =e, raja sa, no -ok =no.
speak -SIMP =ATTR =FC hundred one say -COS =QUOT
[You] are only pretending to buy it, yesterday you already asked, you are
someone who just talks foruitously, one hundred, he said, it is said.

44. aya taka do' -taw -an -ca -k =te.
interj money be.enough -UPWARDS -REF -NEG -COS =DCL
Jeez! my money is not enough!

45. a =do ro coygok =sa =an ra' -ari -wa.
1s =TOP CLF:ROUND THINGS ten =DLIM =FC/ID get -SIMP -FACT
I only brought ten rupees.

46. hoys do' -taw -an -cha -k.
interj be.enough -UPWARD -REF -NEG -COS
Damn! [it] is not enough any more.

47. jok =sa =ba ray' -ca -ka aa.
spouse =MOB =EMPH go -NEG -IFT 1s
I will certainly not go to my wives.

48. a =aw =e ama =para =e naw =ba naw -naka
1s =acc =FC mother =&co =FC scold =EMPH scold -IFT

tok =ba tok -naka.
beat =EMPH beat -IFT
My mother and her company will scold me [and] beat me hard.

49. jok =sa =ba ray' -sora -ca -ka.
spouse =MOB =EMPH go -totally -NEG -IFT
[I] will certainly also not go to my wives.

50. sala jal -pora -wa =an nem -naka aa, sala.
interj run.away -AIMLESSLY -FACT =FC/ID good -IFT 1s interj
Damn, it is better to run away without destination, as far as Im concerned,
damn!

51. conthay ro ni ra' -a -ok =no,
meolon clf.round.things two get -AWAY -COS =QUOT

hay' =mo ro baa ro baa =mo.
GPN =GEN clf.round.things five clf:round.things five =GEN
He brought two melons [with him], it is said. from those, eh, five from
those five rupees.

52. cinthay ro ni ra' =ay =mo
melon CLF:ROUND.THINGS two get =ADV =SEQ

TEXT 5 Alsia raja

605
sore -a =te sore -a.
travel -AWAY =DCL travel -AWAY
Having gotten two melons, he travelled away, travelled away.

53. konsa =do jow' -ca =na na -ok =no.
later =TOP sleep -SUDDENLY =DAT need -COS =QUOT
Later he suddenly had to sleep, it is said.

54. ucie jow -ca =na na -wa =ci =e
then sleep -suddenly =DAT need -FACT =LOC =FC

matsa goro -tat -ok =no =te,
tiger meet -COMPULSARY -COS =QUOT =DCL

maykap =ci jow -wa =ci.
hay =LOC sleep -FACT =LOC
Then, when [he] suddenly needed to sleep, [he] could not help but meet a
tiger, it is said, Im telling you, when he was sleeping in the hay.

55. konsa =do matsa =do morot son man' =ay =mo
later =TOP tiger =TOP person smell obtain =NF =SEQ

ray' -wil -ok =no alsia =do.
go -AROUND -COS =QUOT lazy.pesron =TOP
Later, having cought the smell of a human, [the tiger] walked around the lazy
king.

56. ray' -wilwil -ok =no.
go -AROUND -COS =QUOT
[He] went round and round, it is said.

57. matsa' =e atak =na? canci -aydoa =no ue,
tiger =FC do.waht =DESI think -PROG =QUOT DST

kore -aydok =no kore -wa =ba.
be.afraid -PROG =QUOT be.afraid -FACT =EMPH
What is the tiger planning to do? he thought, [he] is afraid, it is said, [he] is
afraid indeed.

58. otokoy te'ew jow =ci =e konsa =e:
so.then now sleep =LOC =FC after =FC

aa thoy =e thoy -man -ok.
1s die =FC die -ALREADY -COS
So then, now, later, when he is asleep: I am certainly dead.

59. ician: aa jagi =ba thoy -man -ok.
then 1s life =EMPH die -ALREADY -COS
Then: I already life-died.

TEXT 5 Alsia raja

606
60. nok =sa ray' =na =ba pha' -phin -ca -aydok.
house =MOB go =DAT =EMPH dare -TOTALLY -NEG -PROG
I totally dont dare to go home.

61. jebado aa takruk -sora -ari -naka.
somehow 1s fight -TILL.THE.END -SIMP -IFT
Somehow I will just fight to the end.

62. mat[sa] ca'phu =aw. matsa =mi ca'phu =aw
tiger thigh =ACC tiger =GEN thigh =ACC

wa' -jol -ok =no =aro, khabak =ay =mo.
bite -QUICKLY -COS =QUOT =EMPH hold.tightly =ADV =SEQ
[On the] ti[gers] On the thigh; he quickly bites the tiger on the thigh, it is
said, having grasped him tightly.

63. wa' -wa =ci =e konsa =e: ay i =do alsia =e
bite -FACT =LOC =FC after =FC interj PRX =TOP lazy.person =FC

kak -a =te.
bite-CUST =DCL
Having bitten, later on: Auch! that lazy person bites!

64. sok -ca -ka =te =ma a =do no =ay =myng matsa =do,
succeed -NEG -IFT =DCL =Q 1s =TOP say =ADV =SEQ tiger =TOP

jenethene jok =ay =mo jal -a -ok =no =ro.
barely escape =ADV =SEQ run.away -AWAY -COS =QUOT =EMPH
Having said: I will not succeed, what?! the tiger, [he] barely escaped and
run away, it is said.

65. jal -a =ay =mo konsa =do jan' =gaba =mo
run.away -AWAY =ADV =SEQ after =TOP far =ATTR =GEN/ABL

otokoy ol -ruk -ok =no =aro:
like.this speak -RC -COS =QUOT =EMPH
After having ran away, [they] spoke to eachother like this, it is said:

66. aa =e sok =e ok -ca -k.
1s =FC succeed =FC succeed -NEG -COS
I lost. Litterally: As far as succeeding is concerned, I dont succeed any
more.

67. jal -a -an -ca =noo.
run.away -AWAY -REF -NEG =QUOT
He did not ran away, it is said.

TEXT 5 Alsia raja

607
68. aa =e sok =e sok -ca -k
1s =FC succeed =FC succeed -NEG -COS

'mhm, sok =e sok -ca -k.
no succeed =FC succeed -NEG -COS
I lost, yes, I lost.
Litterally: As far as succeeding is concerned, I dont succeed any more, no,
as far as succeeding is concerned, I dont succeed any more.

69. a =ba baju phi -e -naka,
1s =ADD friend invite -TOWARDS -IFT

na' =ba baju phi -e =bo,
2s =ADD friend invite -TOWARDS =IMP

jat morot no =gaba, a =ba matburu no =gaba
spieces human say =ATTR 1s =ADD wild.animals
92
say =ATTR

khaket =aw phi -a -ni, no -ok =no =ro.
all =ACC invite -TOWARDS -FUT say -COS =QUOT =EMPH
I will invite friends over here, you also invite friends over here, [of] the so
called humans species, and I will invite all the wild animals over here, [he]
said, it is said.

70. konsa =do bean bebe iskon san iskon somay
later =TOP truly this.much day this.much time

thik kha' -ak =no =aro.
exactly do -COS =QUOT =EMPH
Later, truly, they fixed a time and a day, it is said.

92
The word matburu ~ matboru wild animal is a close cognate to or a loan from Garo and is a
compound of mat animal and buru jungle and means thus literally jungle animal. In Garo the
word mat means both animal and squirrel. There is another cognate, which speakers identify as
real Atong, which is matpalo wild animal, which is also a compound, with the same literal
translation of jungle animal. The word mat means only animal in Atong and not squirrel, as in
Garo (the Atong word for squirrel being karat). The second element of the compound, palo, means
jungle. The morpheme buru ~ boru is not used outside this compound. When talking about the
jungle, Atong speakers use the cognate lexeme palo jungle.
TEXT 5 Alsia raja

608
71. phep japa =ci mu' -saw -khal =na
banyan.tree foot.of.tree =LOC sit -EXPECTANTLY -CP =DAT

bal -ok =no.
speak -COS =QUOT
[They] said to wait at the foot of a banyan tree, it is said.

72. konsa =do matsa =ba baju phi =aydok =no:
ater =top tiger =EMPH friend invite =PROG =QUOT
Later the tiger indeed is inviting [his] friends, it is said:

73. jat matboru no =ga =do ne'khat =ba do'
species wild.animals say =ATTR =TOP bee =add ?

je ha'golsak =ba muma =ba matsa =ba
any all =ADD alaphant =ADD tiger =ADD

jekhay amak tak =ga =aw gumuk =an.
for.example monkey do =ATTR =ACC all =FC/ID
as far as the species of so called wild animals is concerned, bees and all kinds
[of animals] and elephants and tigers [and] for example different types of
monkeys, all [of them].

74. i =sa =do morot =do na'pit oh mayba =do
PRX =MOB =TOP human =TOP barber uh interj =TOP

alsia raja =do moro =tara =an =no.
lazy.person king =TOP human =EXCLUSIVE =FC/ID =QUOT
Here [however], as far as the species of humans is concerned, the barber uh,
whats it? the lazy king [is] the only human, it is said.

75. aa ca =aw morot baju man' -pha -naka?
1s who =ACC human friend obtain -IN.ADDITION -IFT
Who else will I get as a human friend?

76. a =do a doda =an, canci -aydoa =no =ro.
1s =TOP 1s alone =FC/ID think -PROG =QUOT =EMPH
As for me, I am alone, [he] is thinking, it is said.

77. phep =ci sonti =butu =ci, te'ew =e
banyan.tree =loc lament =WHILE =LOC now =FC

na'pit mo' sa ray' -pha -k =no =ro.
barber CLF:HUMANS one come -IN.ADDITION -COS =QUOT =EMPH
While [he] is lamenting in the bunyan tree, a barber came along, it is said.

78. alsia raja =do thol' =ok =no.
lazy.person king =TOP lie -COS =QUOT
The lazy king lied, it is said.

TEXT 5 Alsia raja

609
79. napit, na'a bi =sa?
barber 2s QF =MOB
Barber, where [are you going]?

80. khaw khan' =na ray' -aydoa a =do.
hair cut =dat go -prog 1s =top
Im going to cut hair.

81. aya khaw kha' -wa =na day =ay =do na'na =e
interj hair cut -FACT =DAT be.bigger=ADV =top 1pi =FC

i =ci cay =ay mu' -wa =an ga'su -khal -naka.
PRX =LOC watch =ADV sit -FACT =FC/ID splendid -CP -IFT
Hey! It is better that we sit here and watch than that you cut hair. Literally:
Than cutting hair, us watchingly sit here will certainly be more splendid.

82. matboru no =ga khakhet ray' -a -ni =no tay'ni
=do.
wild.animals say =ATTR all come -TOWARDS -FUT =QUOT today =TOP
All so called wild animals will come today, it is said.

83. u =aw na'na =e cay =ay mu' -naka, no -ok =no
=ro.
DST =ACC 1pi =FC watch =ADV sit -IFT say -COS =QUOT =EMPH
We will sit and watch them, [he] said, it is said.

84. cirokhana =takay cay =ay mu' -naka.
zoo =like watch =adv sit -ift
[We] will sit and watch them like a zoo.

85. ray' -a =bo, thol' -ok =no.
come -TOWARDS =IMP lie -COS =QUOT
Come!, [he] lied, it is said.

86. ay cay =na na -ni otokcido.
interj watch =DAT need -FUT in.that.case
Oh! In that case, [we] will have to watch.

87. te'ew =do phep kambay =ci du =ay
now =TOP banyan.tree top =LOC climb =ADV

cay -saw -arok =no =te, alsia raja =para
look -EXPECTANTLY -PROG =QUOT =DCL lazy.person king =&co

na'pit =para =do.
barber =&co =TOP
Now climbing in the top of the banyan tree, the Lazy king and his company
[and] the barber and his company are expectanly looking out fot the animals, it
is said.

TEXT 5 Alsia raja

610
88. konsa =do u =sa matboru =dora =ba ray' -adok =no
later =TOP DST=MOB wild.animals =p =ADD come -PROG =QUOT

jamjol =ay =an.
complete =ADV =FC/ID
Later, the wild animals are also coming, all of them.

89. thep! gaw! no -arok =no =an.
[SOUND] [SOUND] say -PROG =QUOT =FC/ID
[They] go thep! gaw! [the noise the animals produce while they are moving
en masse], it is said.

90. ha'golsak =an ray' -arok =no =aro.
all =FC/ID come -PROG =QUOT =EMPH
All [of them] are coming, it is said.

91. pheru, pheru =ba ray' -adok =no =ro.
fox fox =ADD come -PROG =QUOT =EMPH
The fox, the fox is also coming, it is said.

92. pheru =an raja =no.
fox =FC/ID king =QUOT
The fox [is] the king, it is said.

93. konsa =do ray' -wa =ci =e na'pit =do muma matsa =na
later =TOP come -FACT =LOC =fc barber =TOP elephant tiger =DAT

neka -rawraw -wa =na kore -tho =ay =mo
close -CONTINUOUSLY -FACT =DAT fear -SO.MUCH =ADV =SEQ

phep =mo gal' -sora -ok =no na'pit =do.
banyan.tree =ABL fall.down -COMPLETELY -COS =QUOT barber =TOP
Later, the barber was so afraid of the tigers [and] elephants because [they]
were coming continuously closer, [that] he fell down all the way out of the
banyan tree, it is said, the barber.

94. ucie muma matsa =gumuk =an
then elephant tiger =all =fc/id
Then all the elephants and tigers

95. oh mayba a tay'sa sawal -wa.
interj maybe 1s a.little.while.ago forget -fact
Oh, maybe I forgot something just then.

96. mayba =do pheru =do ni' -khu -a =aro.
maybe =TOP fox =TOP not.exist -INCOM -CUST =EMPH
Maybe the fox was not there yet.

TEXT 5 Alsia raja

611
96.a muma matsa ha'golsak matboru khakhet =do
elephant tiger all wild.animals all =TOP

alsia raja thorok -ok.
lazy.person king jump.down -COS
The elephants, the tigers, all the wild animals, the lazy king jumped down.

97. jal =bo! no =ay =mo jal -theri -ok =no
run.away -IMP say =ADV =SEQ run.away -AGAIN -COS =QUOT

na'pit gal' =gaba =na.
barber fall.down =ATTR =DAT
Run away! [someone]
93
said, and they run away again, it is said.

98. konsa =do thik ue na'pit =e phep
later =TOP exactly DST barber =FC banyan.tree

ca' kok no' =sa galat -wa =e,
foot basket inside =MOB fall.down -FACT =FC

da? -jol -a -wa =no =ro.
enter -QUICKLY -AWAY -FACT =QUOT -EMPH
Later, precisely, that barber fell into a hollow between the roots of the tree
and quickly disappeared in it, it is said.

99. otokoymo matboru jal =gaba kaket =aw
so.then wild.animals run.away =ATTR all =ACC

pheru goro -ok =no =aro.
fox meet -COS =QUOT =EMPH
So then the fox met all the animals [that] are running away, it is said.

100. pheru goro -wa =ci =e:.
fox meet -FACT =LOC =FC
When the fox meets [them]:

101. atakna jal -wa bay'siga =dora?
why run.away -FACT friend =p
Why do you run away, friends?

93
Here it is not clear, even to my consultants, who says Run away!. At first it was believed to be the
lazy king, telling the barber to run away. Later however, the context seems to suggest that it is one of
the animals that says it. This is a good case of contextual ambiguity due to lack of referents expressed
in the clause.
TEXT 5 Alsia raja

612
102. ayu, otokoy otokoy do' -wa =com takruk =na san somay
interj like.that like.that IE.be -FACT =IRR fight =DAT day time

thik kha' -wa =com;
exactly do -FACT =IRR
Oh! This and this supposedly happened, [they] supposedly fixed a day and
time to fight;

103. te'ew =e ni =an kore -phin' -a.
now =FC 1pe =FC/ID fear -BACK -CUST
Now it is us who are afraid [of him]. Alternatively: now we we are afraid
back.

104. alsia raja =do kha'do -ari -a.
lazy.person king =TOP be.courageous -SIMP -CUST
The lazy king is just courageous.

105. thorok -sora -ok u =na, ni =do jal =gaba -ak.
jump.down -totally -COS DST=DAT 1pe =top run.away =ATTR -COS
He totally jumped out [of the banyan tree] and so we became the ones who
ran away.

106. atakna kore -wa, morot =ma' =dora =na?
why fear -FACT human =interj =P =DAT
Why are you afraid of the humans?

107. hay a gana. a =an raja.
come.on 1s exist 1s =FC/ID king
Come on! I am here, I am the king,

108. a an bal -thum -ni,
1s =FC/ID speak -ON.BEHALF.OF.SOMEONE.ELSSE -FUT

no -ok =no =ro pheru =e.
say -COS =QUOT =EMPH fox =FC
I will talk on your behalf, [he] said, it is said, the fox.

109. ray' -ak =no.
go -COS =QUOT
[The fox] went.

110. phep =ci otokoy mu' -aydo =no =aro.
banyan.tree =LOC like.this sit -PROG =QUOT =EMPH
[He] is sitting down in the bunyan tree like this, it is said.

111. mu' -wa =ci =e ri' =do cu'ret tak -a -wa =no
sit -FACT =LOC =FC penis =TOP stuck do -AWAY -FACT =QUOT

ca'ma =sa na'pit =sa.
lower.side =MOB barber =MOB
TEXT 5 Alsia raja

613
When [he] sits, his penis got totally stuck, it is said, downward, in the
direction of the barber.

112. a kak =bo no -wa =ci =e kak =bo.
1s bite=IMP say -FACT =LOC =FC bite=IMP
When I say Bite!, bite!

113. a kak =bo no -wa =ci =e kak =bo, no =ay,
1s bite=IMP say -FACT =LOC =FC bite =IMP say =ADV

cin tak -ayro =no =ro.
sign do -PROG =QUOT =EMPH
Saying: When I say Bite!, bite!, [he] is making a sign, it is said.

114. ri' =do cu'ret gal -a -aydok =no ca'ma =sa.
penis =TOP stuck fall.down -AWAY -RPOG =QUOT lower.side =MOB
His penis is falling down and is stuck, it is said, downwards.

115. a tambo no -wa =ci =e tambo.
1s wait.IMP say -FACT =LOC =FC wait.IMP
Whan I say Wait!, wait!

116. ucie jamjol =ay jal
94
-a -ok =no =ro.
then complete =ADV ran.away -AWAY -COS =QUOT =EMPH
Then [they] all ran away, it is said.

117. hay, no -wa =ci =an, ri' =do ri' =an
come.on say -FACT =LOC =FC/ID penis =TOP penis =FC/ID

na'pit =do ri' cu'ret tak =gaba =aw, hay' =mo
barber =TOP penis stuck do =ATTR =ACC whatchamacallit =GEN

pheru =mi ri' =aw kan' cot -et -ok =no =aro.
fox =GEN penis =ACC cut tear -CAUS -COS =QUOT =EMPH
When [he] said Come on! the penis, this penis, as for the barber, the
stuck penis, the whats it, the foxs penis, [he] cuts and tears it, it is said.

118. aya! a ri' =do kan' cot -sora -ok na'pit.
interj 1s penis =TOP cut tear -TOTALLY -COS barber
Ouch! the barber has cut and torn my penis!

94
The speaker used the verb gal- to fall down here, but this was replaced by the verb jal- to ran
away by my friend who transcribed this text for me, because otherwise the sentence would not make
sense in the context.
TEXT 5 Alsia raja

614
119. jal =bo!, no =ay =mo paraw -ok =no.
run.away =IMP say =ADV =SEQ yell -COS =QUOT
Run away! [he] said and shouted, it is said.

120. matboru =ra (Garo plural) =do jal -ok =no.
wild.animals =p =TOP run.away -COS =QUOT
The wild animals ran away, it is said.

121. otokoymo alsia raja =an jam -gop -ari -ok =no =aro.
so.then lazy.person king =FC/ID win -? -SIMP -COS =QUOT =EMPH
So then the lazy king had just won, it is said.

122. i =an golpho jam -ok =ay a =do.
PRX =FC/ID story finish -COS =POS 1s =TOP
This story is finished, as far as im concerned.

123. Other speaker in background:
jam -ok i =ci =an.
end -COS PRX =LOC =FC/ID
[It] ends here.

124. Tonton:
ho'o.
yes
Yes

125. thol' -ok =ma bebe -ok =ma?
lie -COS =Q true -COS =Q
Did [I] lie ot tell the truth? Alternatively: Did [I] lie or was [I] true?

126. aa na -wa =tha =tokoy =sa bal -ay -wa.
1s hear -FACT =OWN =LIKE =DLIM tell -TOWARDS -FACT
I told it [to you] like I myself heard it.

615

616
Appendix 2 Atong-English Dictionary
_____________________________________________________________________

This dictionary contains well over 3000 entries and contains all the lexical items that I
collected during my fieldwork (see 1.9) as well as grammatical morphemes. Since it
is my intention to publish the dictionary in India, to make it accessible to the Atong
speaking community, it is written in the orthography I developed for the language,
which is explained in 1.5. Table 2 from that section, showing the relationship
between the phonemes and the orthography, is repeated here as Table 76 for
convenience.

written in the orthography.
Phonemes Graphemes Phonemes Graphemes Phonemes Graphemes
p
h
ph m m i i
t
h
th n n e e
k
h
kh ng a a
p p r r o o
t t l l u u
k k s s y
b b c ch ii
d d j j ee
g g h h oo
w w y i aa
glottalisation

or

Apart from an indication of the word class or the type of grammatical morpheme
(according to the list of abbreviations given below), the semantic field of each noun,
too, is indicated in the dictionary. The semantic fields and their abbreviations are
listed below. Most of the verbs, and some of the other morphemes, are accompanied
by stretches of text that illustrate their use. Examples of the use of the grammatical
morphemes can be found in the grammar.
ATONG-ENGLISH DICTIONARY

617
The order of the symbols in the Atong alphabet is given below. Every symbol has a
name and an example in which the letter occurs, which makes it easyer for speakers to
remember the alphabet. The examples are, for the convenience of the readers of this
thesis, accompanied by a phonemic transcription and an English gloss.

The Atong Alphabet
symbols name example phonemically gloss
raka mym sa /mom' sa/ one fist
A a a Atong /ato/ Atong
B b ba baju /baju/ friend
C c cha chak /cak/ hand
D d da dam /dam/ price
E e e era /era/ type of fish
G g ga gawi /gawi/ girl
H h ha haba /ha'ba/ dry rice and vegetable field
I i i ichi /ici/ here
J j ja jabek /ja'bek/ curry
K k ka kuchuk /khu'cuk/ mouth
L l la laha /laha/ resin
M m ma mai /may/ rice
N n na net /net/ type of basket
O o o ongang /o'a/ type of frog
P p pa panchung /pancu/ jackfruit
R r ra rai /ray/ reed
S s sa symgong /somgo/ type of plant
T t ta tyi /toy/ water
U u u uching ~ ukching /u'ci ~ ukci/ leech
W w wa wak /wak/ pig
Y y y ymbyng /ombo/ bamboo fluite

618
Abbreviations for word classes
adj1 adjective class 1
adj2 adjective class 2
adv adverb
bound bound morpheme
clf classifier
conj conjunction
cop copula verb
dem demonstrative
disccon discourse connective
encl.cl clausal enclitic
encl.phr phrasal enclitic
encl.phr.cl phrasal and clausal
enclitic
evsp event specifier suffix
interj interjection
interr interrogative
khjyks khatha jyksai,
coordinate compound
that consists of two
synonyms
n noun
num numeral
onom onomatopoeia
postp postposition
ppron personal pronoun
procl proclause
prof proform
prtcl particle
sfx suffix
tw time word
v transitive, intransitive or
ambitransitive verb
V verb, used in the definition of
event specifiers, e.g.
ang V away etc. Instead of
the V a semantically appropriate
verb can be inserted, e.g. byt- to
drive etc. The result will then
be bytang- to drive away.
vB Primary-B verb (transitive verb
that can take a dative-marked O
complement clause)
vdat verb which takes its Theme,
Patient or Target NP in the
dative case
v verb that cannot take any
argument, i.e. with zero valency
vphase phasal verb
vS1 intransitive verb which can only
take one specific S (intransitive
subject) argument
vsec Secondary verb (can take dative-
marked verbal complements in
O function)
The tilde ~ indicates a variant spelling and/or pronunciation of a word
Abbreviations for semantic domains of nouns
ABSTR abstract noun
ACT human activities,
results of or
circumstances related
to such activities
ANIM animals
ART artefacts including
materials used in their
production
BODY body parts of humans
and animals
CORP diseases or substances
produced by the body
FOOD food items, ingredients
used for food
GEO geographic, geological
or natural phenomena
KIN kinship terms
MSRE measure terms
PERS persons, designations for people
or groups of people
PLACE places, landmarks and points
used for orientation
PLANT plants and parts of plants
QUAL qualities
QUANT quantity
SHAPE geometrical and other forms and
shapes
SUBST substances (not those used for
food or for the production of
artefacts)
TIME time expressions

619

mhm procl Thats right.
-a sfx customary/imperfective aspect
suffix
-a ~ -ai evsp V towards the speaker
Nangdo hambun isang
raiaphinwachi, jykthangaw
bytaibone. Later, when you come
back here, bring your wife, OK.
Angdo hanep nangsang
reengni. Raiabo. I will
go/come to your place tomorrow.
Come.
abek n ART long hollow drinking
spoon made of a dried fruit with a
hole at the top and a hole in the
side of the bulge at the bottom used
to scoop rice beer (chyw) out of the
filter (janti) which stands in the
middle of a large earthen pot (gora)
filled with fermented rice (sithi).
The spoon is held by its long slim
part and slowly lowered into the
liquor in the filter. The liquor seeps
into the spoon though the hole in
the bulge. Then the spoon is
carefully pulled out again, and the
liquor is drunk by emptying the
spoon in the mouth through the
opening in the top of the slim end
without touching it with the lips.
abi n KIN elder sister. Mothers-in-law
can call each other abi.
abong n PLANT corn
abu n KIN grandmother. Is also used
to address an unrelated elderly
woman.
abun adj2 next, following,
neighbouring, other, someone
elses Bil nogabamynggymyn
abundyrangmyngaw naakno.He
heard about this so called Bil from
other people.
achepchep n ANIM type of cricket
achi- v to be born Ang Dajongchi
achiwa. I was born in Dajong.
achu n KIN grandfather. Can also be
used by a grandson when
addressing his grandfather. Is also
used to address an unrelated elderly
man. achu ambi grandparents,
ancestors. This word is also used to
talk about or address an elephant
when you are in the jungle.
ade n KIN stepmother
aganggi n ANIM type of grasshopper
agos n TIME August
agre ~ agrai adv too much
aguk n ANIM grasshopper
-ai encl.cl adverbial clausal enclitic
-ai sfx emphatic positive suffix Nang
reenganchate. Angdo reengokai.
You will not go. I will go!
-ai ~ -a evsp V towards the speaker
-ai ~ -e encl.phr.cl focus enclitic,
occurs on NPs and on predicate
heads of locative clauses.
aiaiai interj interjection to call a pig
aiai interj interjection of surprise
aiaw interj interjection of excitement
aiding n ACT hopscotch (a childrens
game)
-aidonga ~ -aidong ~ -aidok ~ -
aronga ~ -arong ~ -arok sfx
progressive/durative aspect suffix
ain n ACT custom, law, tradition ain
niam laws, customs, traditions
aina n ART mirror Ainachi chaiwachi
phalthangau nuka. When I look in
the mirror I see myself.
aiy interj interjection of inquiry and
surprise: what are you doing?!
aja n KIN elder sister
ajip n ART fan
ajot n ART childrens game played
with two groups of unlimited size.
Between the two groups sits a
person called the king. Children
from both groups have to whisper
the name of a child from the other
group into the kings ear, first a
child from one group, then a child
from the other group. If two
children whisper the same name,
the king will call Ajot! and the
person whose name has been
whispered is out. The group that is
depleted first loses.

620
ak- v to pluck (leaves, fruit etc. not
feathers), to pick (flowers)
-ak ~ -ok ~ -k sfx change of state
suffix
akai n KIN aunt: mothers elder sister. I
also used to address an unrelated
married woman older than the
speaker.
akan n ART wooden rack above the
cooking fire
akyrudygyl n PLANT pumpkin
alabok n ANIM white crane bird
alaga n PERS/ABSTR other
(person/thing), somebody else,
different Alaga morotna dymdym
damdam hynna bai. Dont just
give it to someone else.
ali clf classifier for small heaps or
piles of things narang ali tham
three piles of oranges
alsia ~ halsia n PERS lazy person
althu- adj1 easy
alu n PLANT potato
alukotar n ART helicopter
alupren n ART aeroplane
ama n KIN mother. Can also be used
to talk about or address an aunt.
Can also be used by a mother to
address her doughter.
aman n ART pestle, heavy wooden
pole used for flattening rice in an
assam by pounding aman goi ni
two pestles
ambisuthyk n ANIM type of gold-
coloured metallic beetle that flips
itself back on its feet when it lies
on its back
ambret bambret n ACT childrens
game
ambyrai n PLANT type of tree
-an encl.phr.cl focus/identifier
enclitic, occurs on NPs and clauses
-an evsp still V-ing
-an sfx referential suffix
anai n KIN aunt: fathers sister
anaros n PLANT pineapple
ang ppron I, me, first person singular
-ang evsp V away, V without holding
back, V affluently
anyng n KIN 1. aunt: fathers sister, 2.
sister-in-law: husbands wife
aphap n PLANT yeast used to make
sithi fermented rice
apun n ART fishing hook
apunkara n ART fishing line
apunphong n ART fishing rod
aragong n PERS a person who is too
big for his age
-ari sfx simplicitive suffix, just,
simply
aro conj/disccon and, furthermore,
moreover
-aro ~ -ro encl.phr.cl 1. emphatic
enclitic 2. enclitic signalling that
the speaker will say more
asalchong n ANIM type of black
hairy caterpillar that lives on
jackfruit trees
asam n ART mortar, big heavy
hollowed log in which rice is
flattened by pounding it with an
aman asam pan byryi four mortars
aset- ~ asyt- v to throw away, to
dispose of, jyk aset- to divorce
Alsia rajae jyk asetai jalangwachie
songreangokno. The lazy king
divorced his wives and travelled
away, it is said.
aski ~ askhui ~ askui n GEO star
asok n ART type of woven bamboo
basket to keep live pigs in to sell at
the market, type of fish trap
asol adv really
-asol evsp really
asu n PLANT thorn
asyi ~ asi n KIN aunt: mothers
younger sister
asynthalak n ANIM type of fish
asyt- ~ aset- v to throw away, to
dispose of, jyk aset- to divorce
Alsia rajae jyk asetai jalangwachie
songreangokno. The lazy king
divorced his wives and travelled
away, it is said.
ata n FOOD flour
atak n PLANT type of tree
atak- v to do what (interrogative verb)
Aiaw! Angdo chykaidonga.
Atakwa? Teewmangmangsa

621
tyiruwa naa. Oh! Im cold! What
have you done? I just took a bath,
man! Kynsangdo atakoknowa
jamjolai gopcha amakawe. What
happened later? / What did they do
later? They didnt burry him at all, the
monkey. Na?a atakna jumuaidonga
ie hathaphyraawe? nowano. Why
are you collecting these ashes? they
said, it is said. Atakgaba raja naa
angna gore lapchagabaaw watetwa
nookno. What kind of king are you
that you send me a good-for-nothing
horse?!
atakai interr how?
atakna interr why?
atha n MSRE half
athom n BODY stomach
atong interr what?, Atong maina?
Why?
Atong n ABSTR/PERS the Atong
language, Atong person
atongba prof something Gethengna
atongawba hynbo. Buy something
for him Seino naa teew nang
songsang reengchido angna
atongbaaw rabone. Seino, if you
now go to your country, buy me
something, OK.
atongtykyi How?, How come?,
Why?
atyw interj interjection of surprise
-aw ~ -taw encl.phr accusative
enclitic
awa n KIN father
awan n ART winnowing basket
Awanchi mai chawa. Rice is
winnowed in a winnowing basket.
awan- v to forget
awang n KIN 1. uncle: fathers younger
brother, 2. what children call their
stepfather
awyi n KIN grandmother
-ba encl.phr additive/emphatic
enclitic: and, also, too; indeed
Ytykchido nangna randai angna
kerengne nowachie. Hy! Angba
randai sani nowano. In that
case you will get the meat and I
will get the bones, he said and
then: Hey! I also want to eat
meat, he said, it is said. Naru
bimyngaw poknaka bydyi?
noatakaidonga amak gawian.
Teen ang sungaw chaiaribo
noai takaidongano amakba
rukpekmyng baisigathanggaba
budiaw tyngsymai takaimyng.
Where will we uproot the naru,
old man? the wife of the monkey
is asking. Just look into my mind
later, the monkey is saying who
wants to imitate the idea of his
friend the frog.
-ba encl.cl clausal enclitic:
additive/emphatic suffix on main
clauses: also, too; indeed
Nangtym ang nokaw sawwaba
nemariok. Anga nangtymaw
mythelbiok aro ang nok
chungkhuchido, ina daiai
mannichym anga tangka nookno.
Ytykchido ningba phalthang nokaw
sawaimyng hathapyra phalchie
mannima? Manniba,
nangtymba nowano. You burnt
my house indeed and thats just
fine. Im very grateful to you all
and when my house was even
bigger, I could have gotten more
money for it, he said, it is said. In
that case, when we burn our own
houses and sell the ashes, can we
also get money? Yes, you, too,
will indeed get money, he said, it
is said.
-ba encl.cl indefinite enclitic on
dependent locative clauses:
whenever Tharapna guduk
takwachiba tarakai jalariano
magachake. Whenever he almost
caught up with it, it quickly ran
away, the deer.
ba- ~ ba- v to be born Sa
baak/baak. The child is born.
ba- v to carry a child Dada jojongaw
baaidonga. The older brother is
carrying his younger brother.
Nawgabaaw janawgaba okmachi
baaidok. The elder sister is

622
carrying her younger sister on the
front of her body.
baba n KIN father, daddy, dad. Can
also be used by father or mother
when they address their son. Can
also be used to talk about or
address an uncle.
babaji n PERS fortune teller
babelsi ~ babylsi n PLACE kitchen
babu n PERS child or baby (used to
call a small child or baby)
Babyra n PERS supreme god
bada n MSRE a bunch
badai- v to cross beyond the limit, to
pass a certain point Changba
getheng songmi baiaw badaiok.
Somebody crossed the border of
his village. Dolong khagabaaw
badaiwachi, ramchi agal sawgaba
ganang. When you will have
passed the hanging bridge, there
will be a forest fire along the road.
badal- v to unfold
badym n PLACE paddy field, wet rice
field
badyng- v to trade, to deal in, to do
business in Dakangmi chasongdo
rangdarangaw, rykdarangawsa
barudarangawsabadynga. As for
the past era/generation, they traded
brass gongs and all kinds of
ornaments.
bagan n PLACE garden
bagu n ART cloth for man worn
around the waist
bagukhawa n ART turban with a knot
on the front side of the head
baguriwa n GEO rainbow
bai n PLACE border
bai n PERS friend, kin
bai- v to break Balwa rakaimyng wa
baiok. Because of the hard wind
the bamboo has broken.
baikhop n PLANT type of big broad
green and purple bean
baisiga ~ baisega n KIN friend
baibai adj2 the same Ang hanep
baibai khadi khanphinni.
Tomorrow I will wear the same
clothes again.
baidam n PERS some (people)
Ytykyisa dyngthangdyngthang
songchina hapchina jaltokna
gaakoknowa. Baidam
wathaigyrymchi muok, uawdo
wathaigythym myngok. Baidam
Rongsa thyikhalmi hawaichina
jalangok, uawdo Rongsa Hawai
myngok. Thats why they were
forced to run away to different
villages and different places, it is
said. Some stayed in
Wathaigyrym; that village is now
called Wathaigythym. Some
stayed in the plains of the river
Rongsa; that village is not called
Rongsa Hawai.
baik n ART bike
baisykyl n ART bicycle
baji n TIME hour Atong baji
teewe? Tin baji dongok.
What time is it now? Its past
three oclock. Note that hours are
counted with numerals borrowed
from Hindi.
baju n KIN friend
bak- v to run after someone or
something Kyi masuaw kakna
bakaidong. The dog is running
after the cow to bite it. Banggaldo
tharapna guduk takwachiba,
tarakai jalariano magachake.
Ytykyimyng bakrawraw bakrawraw
janangoknowa. When the Bengal
almost reached it, the deer just ran
fast, it is said. So then, chasing it
more and more, they got far away,
it is said.
bak- v to attempt, to try Rawna
bakwachym ytykchiba manancha.
He attempted to catch it but he
could not.
bak- v to catch up with, to overtake
bak- v to scrape with a spade Samsiaw
bakaidonga. She is scraping away
the weeds with a spade.
bak- v to make barren, to weed out all
the plants Khudalsang ha bakwa.
We weeded the land with a
chopper.

623
bakbak adv easily
baket n ART bucket
baki n ACT credit Baki hynchawa. I
dont give credit.
bal- v to speak, to tell, to say Naa
atong khuchukaw balaidonga?
What language do you speak?
Nangmi jorae chang?
Balchawa angdo. Who is your
lover? I will not tell. Teewe ie
myngsa khelegabamynggymyn
anga choisa balna sykaidonga.
Now I want to tell a little about this
game. Bydyi myngsa balai
hynaimyng, baju takphinokno.
After an old man gave advice, they
became friends again, it is said.
Angmi balwami ichian jametwa.
Walnam. I will finish my story
telling now. Good night.
Ang nangaw balni. I will tell about
you.
balaga n PLACE outside
balgyto n PLANT orchid
balphak- v to blow away
Balphakram n PLACE land of the
spirits of the dead, national park in
the South Garo Hills District
balpisa n PLACE place to piss
balsem- v to talk very long
balsyruk- v to whisper
balwa n SUBST wind, air
balwa- vS1 to blow (of the wind)
bam- v to brood, to sit on an egg Taw
kurungchi bamaidong. The chicken
is brooding in her nest.
bam- vdat to obey, to surrender Ie
sagyrai angnado bamcha. This
child does not obey me. Arong
nokmae Duraaw Dorenggo
Wadachongaw panchi jap khaaimu
Englanmi Britis gobormen sason
kagabana bamchano. Having tied
a trap in a tree, Arong headman did
not surrender to the reign of the
British government.
ban- v to trap, to catch in a trap Jaga
saakno uchie, taw pangai
banokno. They set traps and then
caught many birds, it is said.
ban- v to flow (of rivers) Symsang tyi
Nongal dolongtakai banaidong.
The water of the Symsang river
flows under Nongal bridge.
bando n ART tree house
banga num five
bangbang adv empty
bangbol n ANIM type of fish
Banggal n PERS Bengali, non-
Garo/Atong person
bangganai n ANIM type of fish
bangka n ART fan
Banglades n PLACE Bangladesh
bangphak n ART posts at the entrance
of the bachelors house
bangsi n ART flute
banthai n PERS bachelor, unmarried
man
bapai- v to drop
bara- v to put in a hole, pan, wasung,
bag etc.
-barai evsp V always
baram- adj1 rough
barat n ART string that pulls the skin
of a drum tight
barat- vdat to be ashamed, to be shy
Ie gawi nangna barataidong. This
girl is feeling shy towards you.
Nawang naa! Ang nangna
barataidong. You idiot! I am
ashamed of you.
barata n FOOD paratha, flatbread
barata phelsa one paratha
baratwami n ACT shame
bari n garden
bas n ART bus
basak- v to burn and cause a rash
Thamat ~ thamotba najekwaba
khichido basaka. If you touch the
thamat/thamot plant and the najek
bamboo they cause irritation.
basak- v to cause irritation or itching
Thamat basaka. Ta pyi! The
thamat plant causes irritation.
Dont touch it!
basnengtakgaba n PLACE bus stop
basu n ANIM crown feathers of a bird
-bat evsp most
bat- v to stick in Kun habykungchi
batbo. Stick the stick in the sand.

624
batdyl n ART slingshot
bathan adj2 lying on his back
Bathanai juwbo. Lie down on your
back.
batkynyng- v to smash
baton n ART button
batphai- v to throw hard
batpyret- v to smash by throwing
something to the ground
baw- v to dry: to make jerky, to dry
vegetables
bawang clf length of the widely
stretched arms and hands
bawbyl n PERS enemy
bawbyl chambyl n khjyks PERS
enemy
bawen n SHAPE circle bawen sene
seven circles
bawen- v to move in a circle, to make
a circle around something
bawra n ACT arrogance bawra tak- to
be arrogant
bebe adv truly bebe ra- to believe
Ang nangaw bebe racha. I dont
believe you.
bebe- v to believe Memangaw bebea.
I believe in ghosts.
bejaw- v to experience the sensation
of being tickled Nang angau
thebajauwa, ang bejawok. You
tickled me and I feel tickled.
bek n ART bag
bel- ~ bil- v to retract the foreskin
from the glans penis Nang riaw
belbo. Retract your foreskin!
belcha n ART spade
bengblok n ANIM toad, type of frog
bera n ART a fence
bering n FOOD food cooked in a
wasung
bering-~ bereng- v to cook in a
bamboo cylinder (wasung) which
is sealed with banana leaves and
placed in the fire
beringwa ~ berengwa n FOOD food
cooked in a wasung
betyri n ART battery betyri thong
byryi four batteries
bewal n ACT tradition, habit
-bi evsp very
bichamchym n ACT fragment
bithyn ~ pithyn n BODY liver
bia n ACT wedding bia kha- to
marry, to have a wedding bia
khaak to be married
biambong n BODY biceps
biawthang n KIN brother-in-law:
wifes elder brother
biba interr when? Naa bibasa
raiani? When exactly will you
come?
biba n BODY breath, vapour, steam
bibasa adv wherever
bibyrokhon ~ bibakoron adv some
day
bichi interr where?
bichiba adv sometimes, somewhere
Ang chabiaw bichiba thagalok. I
lost my key somewhere.
bichiba bichiba adv sometimes,
seldom
bichylap n BODY abdominal
membrane
bie interr which?, where? Sam
Manama. Bie same? The medicine
stinks. Which medicine? Bie
nang jongdyrange? Nang jonge
bie? nookno janggaklchi
syngokno. Where is your younger
brother? Your younger brother,
where is he? she said, it is said.
bigaba ~ biga interr which? Bigaaw
biskut ranima? Which biscuits
shall I buy?
biji n ART injection, injection needle
biji su- to give an injection biji
phong ni two injections, two
injection needles
bijyrang- v to hang to dry Khadi
bijyrangbo. Hang the clothes to
dry.
bikha clf classifier for surfaces of 80
by 80 pit
bil- ~ bel- v to retract the foreskin
from the glans penis Nang riaw
belbo. Retract your foreskin!
bilding n ART house built with
cement, building
bimang n ABSTR body, appearance

625
bimung ~ bimyng n ABSTR name
Angmi bimung Samrat myngwa.
My name is Samrat.
bins n PLANT type of green bean
bipha n PERS lad, man, male Gawi
khaketsaan songchi muariok,
bipha kakhetdo palyngsang
jalangok. Only the women stayed
in the village, all the men ran away
to the jungle. masu bipha bull
mongma bipha male elephant
biphagaba n PERS husband
biri n ART cigarette
bisang interr to where?, from where?
bisangba ilocprof somewhere
bisangba prof somewhere Getheng
bisangba reengok. He has gone
somewhere.
bisangmi ~ bisangmyng interr from
where?
bisi n SUBST poison
biskut n FOOD biscuit biskut kep sa
one biscuit (focus on small size and
flatness) biskut phel sa one biscuit
(focus on the fact that it is a baked
thing)
biskyn interr how much? how many?
bistibal n TIME Thursday
bisyl n ART coin
bitykyi interr by which way? Bitykyi
reengnima? Ie ramtykyima utykyi?
By which way shall we go? By this
road or by that one?
-bo encl.cl imperative mood clausal
enclitic
borang n ART tree house
boba n PERS crazy man, idiot
bobi n PERS crazy woman, idiot
bobylawthok n PERS fool
bochi (Siju dialecht), jachung (Siju
dialecht) n KIN sister-in-law: elder
brothers wife
bodol- v to change
boiom n ART a jug boiom thai sa one
jug Getheng boiom thai tham
baiok. He broke three jugs.
bokbok n PERS liar
bol- v to cause irritation or itching
Thamat bola. Ta puy! The thamat
plant causes irritation. Dont touch
it!
boli n ACT offer to a spirit
Songgumukan ue mongmawana
wai khurutaisa boli hynaisa
manai sathokwano. Because the
whole village prayed and offered to
the elephant tusks, they all became
very rich, it is said.
bonduk ~ bondyk ~ byndyk n ART
gun, shotgun
-bongbong evsp V more than
necessary, V in abundance, V
scandalously much
bongbong ~ bong n PERS liar Ta
bong! You liar!
bonyng n KIN brother-in-law: the
relation of a man and his younger
sisters husband or a man and his
wifes elder bother
borong n PLANT cob, part of the fruit
where the seeds are set in
abongborong cob of corn. Jackfruit
also has a cob which is called
panchungborong jackfruit cob
bosok- v to itch, to be irritated, to
experience the sensation of
irritation or itching Najekwa
khiaimu cha bosokaidonga.
Having touched the najek bamboo
my leg is itching.
bostu n ABSTR thing bostu myng
tham three things
bot- v to court, to woo, to flatter
Nang Turachi nawmyl botwama?
Did you court the girls in Tura?
botol n ART bottle or its volume,
bottleful
breket ~ brekyt n bracket
Breketmyng nyngchi chipgaba
katha pangai gamchatcha. The
words in brackets are not very
important.
Britis n PERS British
bu- adj1 to be sharp (of pointed
things)
buchok- v to be sharp (of pointy
objects)
buchot n PLANT mango
buchotpan n PLANT mango tree

626
budbal n TIME Wednesday
bugyryk n PLANT type of vegetable
bui- adj1 murky, turbid Tyi buia. The
water is murky/turbid.
bukalang adj2 to have holes in it (of
clothes)
bul- v to stir
bul- v to dig up, to unearth
Noksangsamsang khudal
paiaimyng, tangkaaw bulai
Thengthonna hynetokno. Having
carried a chopper to the side of the
house, he dug up the money and
gave it to Thengthon, it is said.
buna n ANIM big black and yellow
flying insect
burbok ~ bulbok n PERS idiot
but- v to squeeze in, to penetrate, to
go inside a hole Saphawba
hangkhalnyngsang butai
jalangokno. The rabbit runs away
and squeezes into a hole. Bandi
palyng butangwachi matsa
chunggaba gorongokno. When
Bandi penetrated the jungle, he met
a big tiger, it is said. Nekat
wahangkhalnyngsang
butangaidonga. The bees are going
into the bamboo hole.
butang n PERS fucker (swearword)
butbal n TIME Wednesday
buthu- v to seal, to close a receptacle
by putting something in the
opening Wasung rekchaksang
buthuok. The bamboo cylinder is
sealed with banana leaves.
buthu- ~ buthyw-~ bythyw- v to boil
(of water)
butsa n ANIM type of big red ant
-butung sfx concomitant action suffix
bychym- v to pull up/out Una
myngsagaba sabanthai
myngsagaba bychymokno, uchiba
patangphaariok,
dangangphaariokno. Then one
son pulled the other out [from the
water], it is said, but then they just
crossed and they all just drowned,
it is said.
bydyi adj2 old (for persons)
bydyi n PERS old man
bydyi badai n PERS old couple
byira n ANIM cat byira amanthong
jungle cat (the pattern on the skin
of this cat is in the shape of an
aman) byirakhem type of bee
byirakhem n ANIM type of bee
byisa- ~ bysa- v to dance
byisyk interr how much? how many?
San byisyk muni? How many days
will you stay? Nangchi rong
byisyk ganang? How much money
have you got?
bykbyk adv quickly
bykot- v to unsheathe, to take out
Galaimuna kynsangdo phylgymaw
uan rykjolaimuna kukuri
bykotaimuna tokyrengaw
tanthongokno. After the eagle had
fallen to the ground, he ran and
unsheathed his knife and cut of its
head, it is said.
bykphyl adj2 inside out Nang
jama/chola bykphyl. Your shirt is
inside out.
byl n ABSTR/BODY strength, muscle
Ido sagyraido hambundo
chungwachido alamyla byldo
bylnikhon. In the future that child
might really become a bit stronger.
byl- v to make a drum and cover it
with skin
byl- v to cut and kill a big animal or
person, to slay
byl n ACT strike Masu tanna bylsa
nangni gethengo. To slaughter the
cow he needs one strike.
bylak- adj1 strong Uchi mutyngabae
bylakbatgabae Arong nokma
donganoa. The strongest one who
lives there is headman Arong, it is
said.
bylbang n ART tie beam
bylet n ART razor blade
-bylok evsp V into pulp
bylong- v to be too much Naa
bylongdugaai thelnabyi. Dont tie
it too hard. Bylongok! Its too
much!
bylongen adv very

627
bylongok interj So stupid! Bylongokte
nangdo angaw taksakchagado.
You are so stupid if you dont help
me.
bylsi n TIME year bylsi thinian every
year Bylsi chykhywdyrang
dongphinokno. Nine years have
passed, it is said.
bylu n QUAL blue
byryi num four
byrymbyrym adj2 multicoloured
byryp adj2 lying on his belly Byrypai
juwbo. Lie down on your belly.
byt- v to pull, to drag, to drive, to ride,
to transport, to lead, to haul, to
draw, to shock (electricity) Naa
gari bytna sapama? Do you know
how to drive (a vehicle)? Odek ang
khaw bytai thetok. The baby pulled
my hair and pulled it out. Mai
bytwamyngdo pungchina
songchina khairata. We carry the
rice harvest down to the granary, to
the village.
bytai- v to lead here, to bring here (by
driving)
bytchirit- v to draw a line
bytganggang- v to drive a vehicle
over a bumpy road
bythyi n ANIM porcupine
bythyn n GEO shade Ichi mubo. Ichi
bythyn galok. Sit here. The shade
is here. (lit. The shade has fallen
here.)
bythyw- v to be blocked
Wasung/paip bythywok. The
wasung/water pipe is blocked.
bythyw- v to close the wasung or
such containers with raichak so
that the vapour cannot come out
bytjekjek- v to give short jerks
bytphin- v to rewind
bytphuruk- v to tear out with the
roots Getheng waaw
bytphurukwa. He tore out the
bamboo.
bytsek- v to take away a person, to
steal a person gawi bytsekgaba a
person who steals somebody elses
girlfriend
bytsorok- v to pull out
bytwa ~ bytwami ~ bytwamyng n
ACT harvest
bytwami n ACT tug-of-war
bytym n ACT good smell
bytym n BODY/FOOD fat (of human
or animal), grease
bytym- v to smell nice Palengma
bytyma. The flower of the
palengma smells nice.
bytyw- v to boil (of water)
bywsa- v to dance
cha n FOOD tea
cha n FOOD tea
-cha sfx negative polarity suffix
cha n BODY leg, foot cha kantara
barefoot
chaang- v to set (of the sun) Rangsan
chaanga. The sun sets.
chabykung n BODY instep
chachak n PLANT tea leaf
chachok n BODY sole of the foot
chadok ~ chatok n BODY heel
chadyl n BODY/PLANT root, vein
chadylmorong n PLANT main root
of a tree
chadylsaphek n PLANT small root
chagyl n ACT footstep Nokhapalchi
chagyl kyryngaidonga. Footsteps
are making noise outside.
chagywgyw- v to kneel
chagywgyw- v to kneel down
chakereng n BODY shinbone, shin
chakhawak ~ chakhok n BODY
hollow side of the knee
chakhok ~ chakhawak n BODY
hollow side of the knee
chakhop ~ jahop n ART shoe
chakok n PLACE hollow between the
roots of a tree Kynsangdo thik ue
napite phep chakoknyng .sang
galatwa. Later the barber fell
exactly into a hollow between the
roots of the banyan tree.
chakyw ~ chaku n BODY knee,
length from the knee to the foot
Uchie Thengthon khudalsang
haaw sawaidongano.
Sawaidongano, thywangaidokno,
chakyw chyigykdarangdo. Then

628
Thengthon is digging in the
ground with a chopper, it is said.
He is digging and he is getting
deep, it is said, about four metres
deep.
chama n PLACE lower side,
downstream, bottom, below
chaman n ACT footprint
chamuk n BODY medial malleolus
chamyn n BODY leg hair
chapa n BODY sole of the foot
chapakithyk n BODY heel
chapathai n BODY calf
chapha clf a foot chapha tham three
feet
chaphak n BODY groin
chaphong ~ caphung n BODY thigh
chaphung n BODY upper leg
chapungdym n BODY hip
chari n PLANT seed for planting
chasi n BODY toe
chasijywbydyi n BODY big toe
chasyrong- v to stretch your leg
chatok ~ chadok n BODY heel
chawek n PLANT chaff
chawekdam n PLACE place where
the chaff is thrown after
winnowing the rice
chabak- v to fall (of water in a
waterfall)
chabi n ART key
chachek n ART tea strainer
chachura n BODY hair on top of the
head
chaduk- v to bump
chagak n BODY palate
chagak- v to hit, to crash Uchi,
pherudo panchi chagakai thyiokno.
Then the fox hit a tree and died, it
is said.
chagodot- v to stumble Gandichi
chagodotwa. I stumbled over a log.
chai- v to look (at), to watch Tedo
biphagaba Naweng aina
chaiaimyng tangabasang
biphagaba kamyn rokaimyng aina
tanmangabaaw kynaimyng
biphagaba niwachi chaiai
chyichie phalthangaw nukwanoaro.
Now when she picked up the
mirror which her husband had put
down after looking into it to shave
his beard, she saw herself, it is said.
Angba piktjyr chaina. I also want
to see the photos. Ang nangaw
chaikhuni. I will take revenge on
you.
-chai ~ -chyi evsp try to V, V and see
chaisi- v to hate
chaikhaw- v to spy
chaira n ACT a traditional song
chairuru- v to look around
chaisi- v to be annoyed by the looks of
something
chaithum- v to guard, to watch over
chaitunggaba n PERS watchman
chak clf classifier for leaves panchak
chaksa one tree leaf
chak n BODY arm, hand
-chak n PLANT leaf
chak- v to ignite wal chak- to make
fire Walthum chakbo. Light the
walthum!
chak chok n khjyksai BODY hand
Tykywtokreng chak chok
dangchagabachi mai
rymetaidongano. She is cooking
rice in a water pot of which the
neck is so narrow that you cannot
stick your hand in it.
chaka n ART wheel
chakchuk n BODY elbow
chakgydok n BODY wrist
chakgytok n BODY underarm
chakkhawak n BODY hollow part of
the elbow, elbow pit
chakkhop n ART glove
chakol n PERS servant
chakpha n BODY palm of the hand
chakphakhung n BODY back of the
hand
chakphong ~ cakphung n BODY
arm, upper arm
chaksi n BODY finger chaksi goi
banga five fingers
chaksigysep n BODY space in
between the fingers
chaksijotram n BODY index finger
chaksijywbydyi n BODY thumb
chaksikhol n BODY fingernail

629
chaksikhum n BODY back of the
hand
chaksirengma n BODY little finger
chaksiweng n BODY knuckles
chaksyrong- v to stretch your arm
chakwak clf classifier for handfuls
rong chakwak chit sa eleven
handfuls of stones
chal- v to support
chal- v to sow or plant by making a
hole in the ground with a stick and
putting the seed into the hole. Ha
khynmanwamungsa maisi khita.
Umung abongdarang chala,
dachangdarang chala. Only after
collecting the unburnt remains of
the jungle from the land, we sow
millet. Then we plant maize and we
plant dachang.
chalak adj2 cunning, clever Song
damsachi Thengthon mynggaba
morot myngsa ganangno. Ue
bylongen chalakno. In a village
lived a man called Thengthon, it is
said. He was very cunning, it is
said. Chalak morotaimyng udo
teewchinaan khengaidongano.
Because he is a cunning man, he is
now still alive, it is said.
chalgaba n ART a support
chamai ~ chame n KIN 1.
marriageable female cousin, 2 the
relation of female cousins from
intermarriageable families, 3. the
relation of the parents of a married
couple, 4. girlfriend, lover,
sweetheart
chame ~ chamai n PERS/KIN
sweetheart, female cousin,
daughter of mothers brother
(mama)
chanchi- vB to think (about/of) Ie
alsia raja atykyi khengaidok?
Atykyian jykaw haldunna
manaidok? noai morotdyrang
chanchiphinaidoknoro. How does
this lazy king live? How does he
feed his wives? thought the
people, it is said. Bandiaw watetna
chanchiaidokno. He thought about
sending Bandi, it is said. Jesang
ang reengchiba, mancha
nangaw awana. chanchia ang
nangawrarasa. (Aristo J Momin)
Wherever I go, I cannot forget you,
I think only of you.
chanchichyp- v to suppose Morot
chanchichypai thik dongokodo,
uchian rajaan uaw ajot
nosawnaka. Suppose someone gets
it right, then the king will tell him
ajot.
chanchok- v to lean on
chanchora ~ chanchura n ANIM
sparrow
chanet- v to put on the fire
chang interr who?
-chang bound multiplied by, times.
This morpheme is only used in
compound numerals with khol
twenty. It can be seen as a bound
morpheme and written together
with khol in numerals, viz.
khokchang byryi rong sa eighty
one.
-chang evsp V suddenly
changai n GEO the moon
changba prof somebody, someone
changchon n BODY waist
changgaba prof whoever Changgaba
manai saa changgaba nokdang
takga, umi bimyng gumukawan
thalai myngaimusa, wai
khurutaimu, saai ryngaimu,
nemkhalchiba nemkhalchachiba ue
morotnado dykdyksa chaisakni.
Whoever is rich, whoever is
wealthy, having called all their
names clearly, having performed
the incantation of the spirit, having
eaten and drunk, the person has to
wait for a short while to see
whether or not he has got better.
chanpat- v to build a bamboo bridge
chanpheng- vdat to defend Ang
hasongna chanphengni. I will
defend my country.
-chap evsp V along with
chap- v to stand (be in standing
position)

630
chapchap adv close together (as in a
crowd)
chara n PERS mothers brothers, the
chara come together when
important decisions concerning the
family have to be made
chara n PLANT sapling
charamong n PERS mothers eldest
brother
charanga num fifteen
chasong n ABSTR generation, era
Dakang achu ambi chasongchido
balphakramaw myngsynga. Tee
chasongsa balphakramaw
myngsynganchak aro
memangsong donganchakno. In
the era/generation of our ancestors
Balphakram was well known. This
era/generation does not know
Balphakram well any more and it is
no longer the land of the ghosts, it
is said.
chat- adj1 thick (of substances and
things), bulky
chat- v to be fixed together (like a
stapled pile of paper or a pile of
wood etc.)
chat- v to promise Ang una sot bonga
hynna chatwa. I promised to give
him fifty rupees.
chatgyk num eight
chaw onom splash! the sound of
something plunging into the water
Ytykyimyng magachakdo biskutaw
tyisamchi tanaimyng chaw!
thorokangokno. So then, having put
the biscuits by the side of the
water, the deer splash! jumped into
the water, it is said.
chaw- v to go by boat, to stream (of a
river) Rung chawchiba rung
bytrongreng. When you row the
boat, the boat spins.
chaw- v to winnow Awanchi mai
chawa. Rice is winnowed in a
winnowing basket.
chaw- v to winnow
chaw- v to stream (of water in a
river), to drown Ie morot tyi
hungna sapchaaimu tyi chauwa.
Because this person did not know
how to swim, he drowned. / This
person, not having known how to
swim, drowned.
chawki ~ chawkyi ~ chengkui n
ART big knife with a curled blade
used in the kitchen to prepare food
as well as in the field to cut plants
and weeds
chek n ART net, fishing net
chek num ten Rong chek hynbo. Give
me ten rupees Mityr chek howwa. I
jumped ten metres.
chek- ~ chyk- adj1 cold Aiaw!
Angdo chykaidonga. Atakwa?
Teewmangmangsa tyiruwa
naa. Jeez, Im cold. What did
you do? I just took a bath, man!
cheknai adv the day after tomorrow
chel n BODY bosom of a man
chelbak n BODY chest
chelku n BODY rib cage
chem- v to melt away, to burn up Pan
walchi chemok. The wood burned
up in the fire. Choklet khuchukchi
chemok. The sweet melted away
in my mouth.
chen n ART chain, zip fastener, zipper,
zip
-cheng evsp V first
cheng- adj1 light
chengkhu ~ chengkhyw n PLANT
ginger
chengchangbengchang n ACT noise,
racket
chengcheng n PLANT tamarind
chengchengmachok n ANIM spider
with a long stomach with yellow
stripes which can be fried and eaten
chenggang- v to be upright, to be
erect Chosa rypaimyng
jarawachian mynan chenggang
takariano pherumi myndo. He
stayed a bit in the water and after a
long time, his fur was still upright,
the foxs fur.
chengkhyna n BODY jaw
chengkhyw ~ chengkhu n PLANT
ginger

631
chengkui ~ chawki ~ chawkyi n
ART big knife with a curled blade
used in the kitchen to prepare food
as well as in the field to cut plants
and weeds
-chep evsp V alone
chep- v to milk Masudut cheparong.
Shes milking cows milk.
chep- v to release contained air or
water, to leak Robol balwa chapok.
The football has deflated./The
football has leaked air. Maityk tyi
chepaidok. The rice-cooking pot is
leaking water.
chep- ~ chip- ~ chyp- v to be
imprisoned, to be caught
(Thengthone):Nangtym angaw
wetsado khema khakhubo. Yhy!
Khema manchak noaimyng
koksep wataimyng koksepchi
chypangokno. Please forgive me
one more time, (said Thengthon).
No! You cannot get any more
forgiveness, they said and they
made a big bamboo basket and
imprisoned him in it, it is said.
Uchie Nepale: Ytykchido ang
reengsigama nangmyng phal?
nowano. Ma ytykchido
dongarini, ang chakdyrangaw
dengbo nooknoro. Khaakno,
Thengthonawdo. Ytykyimyng
chepgaba dengaimyng getheng
hongkotokno. Then the Nepali said:
So them I will go instead of
you?, it is said. Very well, in that
case, its all right. Untie my
hands, he said, it is said. He untied
Thengthon. So then, having untied
the prisoner, the prisoner came out,
it is said. Breketmyng nyngchi
chipgaba katha pangai
gamchatcha. The words in brackets
are not very important.
chep- ~ chyp- v to close Ue Hadura
waie songchi morot thyinaakodo
rongkhalmi nokapaw
chepchangano. Ue nokap
chepachian songgumukmi
morotdyrangan naano. As for that
spirit of Hadura, when a person
had died, the door of the cave
would suddenly close, it is said.
When that door is closes, the
people in the village hear it, it is
said. Khasinai chypangsa dawang
takaidonga. She is slowely closing
and opening her eyes. (Gostar R
Sangma)
chepchap chepchap onom the sound
of a mouse Abeknyngchi muchot
sagyrai mang byryi chepchap
chepchap parawthokaidonga.
Inside the abek are four baby mice
squeaking eek eek.
chepchep chepchap onom squeak
squeak (sound of a mouse) Muchot
chepchep chepchap parawa. A
mouse says squeak squeak.
cherym- ~ chyrym- adj1 heavy
chet- v to tear clothes
chetpyrak- v to tear apart
chewkhyi n ART big knife
chi num ten. This word is only used in
compound numerals ci byri
fourteen.
-chi encl.phr.cl locative enclitic
chi bri num fourteen
chi chat num eighteen
chi dok num sixteen
chi ni num twelve
chi sykhu num nineteen
chi syni num seventeen
chi tham num thirteen
-chichi evsp V with force
-chichi evsp to V into pieces
chichot n PLANT small inedible
jackfruit
chichu- v to blister
chichugaba n BODY a blister
chichugaba n BODY a blister
chigi n PLANT type of plant
-chik ~ -chyk evsp V as long as you
can
chikarak- v to joke
-chikchak evsp swarming
chikchak wekwak adv swarming
around something like fish around
bait
chin n ACT a sign

632
chinthai n PLANT melon
chinara n PLANT lemon
ching clf classifier for bamboo shoots
maiwa ~ maiwa chingsa one
bamboo shoot
chingpheng adj2 aslant, slant Nok
bydyiaimu chingpengok. The
house, having become old, is
aslant.
chingchongphyrot ~
chomchomphyrot n PLANT type
of white edible mushroom
chini n FOOD sugar
chinik n BODY dirt on your body
chinkak n PLANT type of plant
chip- ~ chep- ~ chyp- v to be
chirokhana n PLACE zoo
chisat- v to vomit, to throw up, to barf
chisol n ART a cross
chit- v to tear, rip
chit sa num eleven
chithong- v to tear cloths to shreds
chiti n ART letter
chiwal- v to trade, to deal in, to do
business in
chochep- v to be crumpled Wetsa
reengrawrawwachian dobachi
amakba gasorotaimu bunduk
baithongsyrangokno. Bonduk
baithongaimu kokchengba
chochepokno nemanchakno. Once
when they went a little further, the
monkey slipped and fell in the mud
and broke the gun in pieces, it is
said. The gun was broken and the
kokcheng was crumpled and not
good any more, it is said.
choisa ~ chosa adv a little bit
chomot ~ chongmot adv/evsp
actually, really Anga Ketketa Bura
dongcha. Ketketa Bura kanjota,
anga mela chaibataw noaimyng,
pheruna Ketketa Bura balwano.
Ytykchiba pherue: Nangan
chomot ketketa Bura nookno.
Im not Ketketa Bura. Ketketa
Bura is thin, I look much fatter,
said Ketketa Bura to the fox, it is
said. But the fox said: You are
really Ketketa Bura, it is said.
chosa ~ cho isa adv a little bit
chogop- v fully bent but not touching
the ground (used only with plants)
Rek chogopok. The banana tree is
bent.
chogyp- v to break off and fall down
(for branches and big leaves)
Balwana narykhelchak chogypok.
Because of the wind the branch of
the coconut tree has broken off and
fallen down.
chok clf classifier for bunches or
small heaps jaryt choksa one small
heap of chillies rasunok choksa one
bundle of spring onions
chok- v to scoop, serve up, dish up,
dish out khaw chok- to comb ones
hair
choka- v to be taken apart, to be
disassembled, to be torn Pen
chokaak. The pen is disassembled.

633
Gore mangsa gethengmi aluaw
jamai saakno. Una aludarangaw
rydymabutung sokaw
sawamynggymyn teew manap
chaiwachido gumukan
chokarumokno. A horse had eaten
all his potatoes. Therefore, when he
looked in the morning, the potatoes
were all torn, because all the
sprouts were eaten while they were
sprouting, it is said.
choka- v to cut off
chokchok- v to sharpen (a pointy
object)
chokdeng n BODY throat
chokdeng n PLACE the end of a
pointy object
choket- v to scoop (for solid
substances)
chokhoi n ART fishing basket made of
bamboo
choki ~ chuki n ART chair
chokida n PERS warden
choklet n FOOD a sweet, chocolate
chokset- v to scoop away
chol clf classifier for ways, roads,
paths and rivers tyikhal chol ni two
rivers ram chol tham three roads,
paths sorok chol byryi four roads
chol n ACT idea Nanangdo myngsa
cholawdo taknaka. We will execute
one idea.
chola n ART shirt
cholwat n ABSTR a space
chom- v to stack, to pile up, to fuck
chom clf classifier for little piles of
fruit Narang chom ni hynbone.
Give me two little piles of oranges.
chong clf classifier for iron nails khil
chong ni two iron nails
chong n ANIM insect, bug, lice
chongsu n ANIM caterpillar
chongchang n ART bird cage made of
bamboo
chongchyro- v to squat
-chongmot ~ - chomot evsp V
determinedly, V certainly, V
definitely San nidyrang
dongphinaidok, nang noksang
raianado
pachongmotchaaidokkhon
nookno. It has been two days and
maybe he really does not dare to
come to your house.
chonnyk- v to look down on
Ytykyimyng kynsang phalthangaw
chonnykgabaaw naaimyng alsia
rajae: Na anga ytykyi cholie
cholisemchaaidok noaimyng
teewba jykmyng jalaidokno. So
then, later, having heard the ones
that looked down on him, the lazy
king said: Well, I certainly have
not succeeded at all and then he
ran away from his wives, it is said.
chot- v to tear (off) Sendel chotok. My
sandal is broken. Aia thetnabai!
Ang chak chotni! Ouch! Dont pull!
My arm will tear off!
chu- v to wrap into something
churet adv stuck Phepchi pheru
ytykyi muaidonoaro. Muwachie
rido churet takangokno
chamasang napitsang. The fox
was sitting in the banyan tree like
this, it is said. While he was sitting
there, his penis was stuck, it is said,
downward, toward the barber.
chusok- v to succeed Nangtyme
iawan phalthangthangna
hyngaawan kamtykyi
chusoketchachido nangtyme
atongtykyi phylgym kawna mana?
If you cannot succeed in the job
that I gave to yourselves, how can
you shoot the eagle?
chuduk- ~ chyduk- v to turn upside
down, to turn over
chugup v on its side Rung chugup
paitanbo. Turn the boat on its side.
Chugupai tanwa. I put it on its
side. Rung chugupok. The boat is
lying on its side.
chugup- v to cover with a lid
chui interj interjection to chase away
a pig
chui interj interjection to chase away
a domestic animal
chuki ~ choki n ART chair

634
chuli- v to be useful Tangka poisaba,
kamba janggina chulia. Money and
wealth are useful in life.
chultet- v to shake off
chun n BODY trump
chun n FOOD limestone (in ground
form)
chung- adj1 big
chunggalgal- v to grow up, to become
an adult Sa myngsa baaimung,
mandykarok. Mandykasola
hoong. ie jenkonparaba
raiasyrangchak. Welang
welang chunggalgalwasa
ganakachym. jengkonparaba
raiakhuchakhon? After one
child has been born, it is difficult.
Difficult indeed, yes. This Jenkon
and those associated with him
never come any more. He will
almost certainly be compelled to
grow up quickly.
chungtaw- v to grow
chungthai n BODY big bosom Samsa
mylthai samsa chungthai. One big
bosom, one small bosom.
chup ~ chyp adv fully dressed, with
all your clothes on, wearing
whatever it is you are wearing Ang
chyp tyruok. I took a bath with all
my clothes on. Ang chup
reengariok. I just went wearing
the clothes I was wearing at that
time.
churi n ART knife (Hindi )
churu n FOOD very little food
chuwil chuwal adv spinning Thot
thyngthot takwachina dabat
sykromaimyng
khanetsigaaidongno. Bandi
chakwatwamian chuwil chuwal
takjolangokno. He (Bandi) grasped
her (Sore) and poured the liquor
into her mouth to the last drop.
When Bandi let go of her (Sore),
her head was spinning.
chuwyng chuwang adv with a
spinning head, dizzily
chybym n BODY forehead
chyduk- ~ chuduk- v to turn upside
down, to turn over
chygyl n ANIM type of eel that comes
out of the river when it rains a lot
chygyp- v to fall face down on the
ground
chyhyl n ANIM type of snail
chyi- adj1 vexing, irritatingly boring,
irritatingly tiring
chyi- v to try Chaiai chyini
gorongnima gorongcha. Lets try
to meet him. (literally: We will try
by seeing if we will meet him or
not meet him.)
-chyi ~ -chai evsp try to V, V and see
chyi- adj1 tired
chyigyk num ten
chyk- ~ chek- adj1 cold
-chyk ~ -chik evsp V as long as you
can
chykyw num nine
chym encl.cl/prtcl irrealis enclitic or
particle Mura taisa ganangchym,
teew niwa. There was supposed
to be a small stool here, now its
gone. Jongkene: hm! kamba atong
kamaw khaaimunaka ie?
banthaichiba tangka poisa nang
nangarokona.
Nongkene: Hoong, chym. Jonken
says: Hm! And then that work,
what work will he be doing while
he stays here?When you are a
bachelor, because you need
money Nongken says: Yes,
supposedly.
chym- v to chew Goiaw nemai
chymaimu dakbo. Chew the betel
nut well, then spit it out.
chymbuk n ART magnet
chympyret- v to hit with your fist, to
crash head-on
chyn- v to offer to the dead Ie taw
mama thyigabana chynkhuni. We
will offer this chicken to our dead
uncle.
chyndyk n ANIM domestic water
buffalo
chyng- adj1 bright

635
chyng- v to burn Ie pan nemai
chyngni. This wood will burn well.
chyngchet- v to glitter
chyngaba n ACT offer to a dead
person
chyngmat n BODY comb of a rooster
-chyp evsp V wastefully, V
unsuccessfully, V completely
chyp- v to close Teewchinaan ue
waimi bimungsang songawba Siju
Duramong noai teewchinaan
myngairongkhua. Ytykyimung ue
rongkhal teewdo chypok. Now
still, the village is called Siju
Duramong after the spirits name,
still today. But the cave is closed
now.
chyp- v to imprison, to lock up Uchie
Thengthon balokno: Nangtym
angaw wetsado khema
khakhubo. Yhy! Khema
manchak noaimyng koksep
wataimyng koksepchi
chypangokno. Then Thengthon
said: Please forgive me one more
time. No! We cannot forgive you
any more, they said and having
woven a big bamboo cage, they
locked him up in the bamboo cage,
it is said.
chyp- ~ chep- v to close Ue Hadura
waie songchi morot thyinaakodo
rongkhalmi nokapaw
chepchangano. Ue nokap
chepachian songgumukmi
morotdyrangan naano. As for that
spirit of Hadura, when a person
had died, the door of the cave
would suddenly close, it is said.
When that door is closes, the
people in the village hear it, it is
said. Khasinai chypangsa dawang
Sangma)
chyp- ~ chip- ~ chep- v to be
chyp ~ chup adv fully dressed, with
all your clothes on, wearing
whatever it is you are wearing Ang
chyp tyruok. I took a bath with all
my clothes on. Ang chup
reengariok. I just went wearing
the clothes I was wearing at that
time.
chyrym- ~ cherym- adj1 heavy
chys interj interjection of disapproval
chyw n FOOD rice beer, alcohol, wine,
liquor chyw chek- to scoop the
chyw out of the gora with an abek
chyw n PLANT the new young leaves
of a tree
chyw- adj1 high, steep hakha
chywa the mountain slope is steep
chywgyn n ACT the festival of the
dead at which the soul of a dead
person is sent out of the house to
rest in peace. The festival is held
around the end of February or the
beginning of March. During
chywgyn people indulge in
different activities such as chyw

636
rynga to drink liquor, khata
juwkyna to tell stories, chaira
rynga to sing songs and Waljan
byta to tell the love story about
Waljan.
chywgyn- v to celebrate the festival of
the dead
da rat- v to fall down (for person)
daba n PLANT coconut
dabat postp since, from, until
(indicating a limit in time)
Ytykchiba naa angna aro
angmyng jykna nang khengwa
dabat ang thyicha dabat angaw
muai sana hynbo nookno.
However, you have to keep giving
me and my wife food as long as
you live until I die, he said, it is
said. umyng ~ umi dabat since that
time, from that time onward uchina
dabat until then, until that time
Tainimyng dabat nangmyngan
baju takchaka. As from today I will
not be your friend any more.
dabogos n ART skewer
dachang n PLANT type of shrub of
which both the leaves and the
flowers are eaten
dachang ~ datchang n PLANT type
of shrub of which the leaves and
flowers are cooked and eaten
dada n KIN elder brother. Is also used
to speak about or address a related
older male relative of your own
generation: cousin, or to address an
unrelated man older than you.
dagi n BODY scar
dai- v to be bigger, greater daiaiok
over, finished
dai- v to wash away (as in a
landslide) Rrang wawana, ha
nomaimu habyri daiok. Because
of the rain the ground had become
soft and therefore the mountain
washed away.
daijol- v to overstay
dainingrum n PLACE dining room
dairamphin- v to work overtime
dairukruk- v become more and more
dak n BODY freckle
dak- v to spit
dakal ~ takal n PERS witch
dakan- v to dress someone else
dakang adv previously Ie hapaw
atong myngnaka noai, Gandrung
songchamchi ramai chyichie
dakang mynggaba Songma
Songgni Khychu Badri nogaba,
bimungsang, khata jyksaisang,
Badri myngnaka noai,
Gandrungawsa Badri
myngchengwano. What shall we
call this place? they said and when
trying to search in the old village of
Gandrung, which was the
previously so called Songma
Songgni Khychu Badri, that was its
name, with two words with the
same meaning, and they called it
Badri, and so Gandyrung was first
called Badri, it is said.
dakang postp ago, before Bylsi sana
dakang jyk khymok gethengdo. He
got married one year ago.
Nangtymmi nanggabaaw
nangtymmi piaidongabaaw,
nangna dakangan phetangok,
nangna dakangan udo
reengsawok. That which you
needed, that which you were asking
for, had arrived before you and it
has certainly left before you. Sana
dakang chaksua. Before eating I
wash my hands.
dakang tw past, in the past, before,
earlier Dakangdo, mamung khem
niwachido dymchyrangsangsa
chywgyn ryngwano. In the past,
when there were no drums, they
celebrated the festival of the dead
only with the dymchyrang, it is
said. Gam manni udo uan, tangka
poisa. Uan gam mynga, dakangmi
chasongdo. Teewsa kepasyti noai
myngaidonga. Chasongna kri gam
myngariaro, tangka poisa. He will
obtain wealth, money. Earlier
generations called that wealth.
Now they call it capacity.

637
According to my generation this
money is called wealth.
dakanggaba adj1 first Uchi
thymaimyng, dakanggaba bobaan
dirichengokno. Then, having lain in
ambush, the first crazy person got
hold of (the horses tail) first, it is
said.
dakanggaba adv first, the first time
Dakanggabado jineral
mitingchengni. Umungsa song
gumuk thomaimung haba haryn
harynaw sowalni. First they will
start with a general meeting. Then
the whole village comes together
and they will divide the haba plot
by plot. Dakanggaba Turachi
muwachi Mobbinaw gorongwa.
The first time I stayed in Tura I met
Mobbin.
dakham n ART very small wooden
stool consisting of one rectangular
wooden board to sit on and two
small rectangular wood blocks
attached underneath as supports
dakmanda n ART long womens dress
tied around the waist, skirt
dala n ART round bamboo mat made
of watyng for drying papol or
chillies in the sun, also called
damplak
dala n PLANT branch of a tree not
directly attached to the trunk,
young plant dalasa one branch
dala pheksa one branch
daldi n PERS beloved person, love,
darling
dalibibi n ART doll dalibibi goisa
one doll
dalim n PLANT pomegranate
dam clf classifier for villages Song
damsachi alsia raja myngsa
ganangchym. In a certain village
there was supposedly a lazy king.
dam n ABSTR price Ie masugari
dame biskyn? What is the price of
this bullock cart?
dam n ART bamboo mat
-dam bound PLACE place jaboldam
rubbish heap chawekdam place
where the chaff is thrown
-dam evsp V truly
dam ~ dym onom bam! Thud!
sound of something heavy hitting
the ground Ytykyimuna tokyrengaw
manaimunga hachina wuuuuuuuk
dym! takramphinoknotyi phylgym
galwaan. So then, having got him
in the neck, the giant eagle fell to
the ground wooooosh bam!
dama n art drum Raja! Nang
damaw chosa ie ang
baisigathang pheru tamnano
noai takaidongano. Ytykyimyng
kantaraaw Tambono
tambono noai tanangarioknoe
magachakdo. Ytykyimyng tamai
chaichie tedo byirakhem
hongkotruruaimu
kaksyrangokno pheruawdo.
Nomangaidokno udo.
Ytykyimyng jalangthiriokno
magachakdo. O King, can my
friend the fox play a bit on your
drum?, he pretended to say, it is
said. Go ahead and play, go ahead
and play!, the deer pretended that
the king said, it is said. So then,
when he tried to play it, the bees all
came out and bit the fox all over, it
is said. The fox became weak, it is
said. So then the deer ran away
again, it is said.
-damdam evsp V in different places,
V one after the other, V
continuously
damdyl n ART bamboo mat that is
used as the side of a house damdyl
khapsa one damdyl
damplak n ART round bamboo mat,
also called dala, made of watyng
for drying papol or chillies in the
sun
damthol n ART a rolled up mat
dan- v to spread out, to lay out (mats
etc.) Naaw khantongai danwa.
She laid the fish down and cut it in
pieces. Palongchi kombol danbo.

638
Spread a blanket over the bed.
(when preparing it to go to sleep)
dan- v to spread Kombol palongchi
danbo. Spread the blanket out over
the bed.
dandan- v to be pressed with your
back against something, to lean
against something Dandanai
mubo. Sit with your back against
the wall (or anyu other supporting
object). Panchi dandanaidonga.
Hes leaning against a tree.
dang- v to enter, to go/come in
Ytykyisa ue Arong nokma
thyiwamisa saepe bondyk paiaimu
sipaidyrang dangna manokno.
Sipaidyrang dangwachie kantyra
gulinyi kawphetphetai
raiaaknokhon. Uchian songchi
dangok. Thats why, after
headman Arongs death, the gun
carrying sahibs were able to come
in, it is said. When the sahibs
entered, they might only have fired
without bullets. Then they entered
the village, it is said. Sansado ue
harataimyng hajambutungchi
umyng khuchuksang sotmai
dumna dangthokokno. One day,
while he was yawning, a swarm of
flies entered his mouth. Noksang
byk dangjolai jalangoknoai. He
quickly ran into the house. Uchie
pagongmachi sa gataimu uchie
dangangoknoro
gethengdo.Ytykyimu tyinyngsang
dangangokno Then, having put the
children on his shoulders, he went
in, it is said. So then he entered the
water, it is said. Ma baba,
atykyimu walawa? nookno
amakaw, amakmi sadyrange.
Niwa. Ue nang awangpara
nokchi dangphakawa naa"
noatakokno. But daddy, why are
you so late? It is already night, the
monkeys children said. Dont
worry. I visited your uncle he
said, it is said. Reenwachian
rangsan dangaimu walokno.
When he left, after the sun had set,
it was night, it is said.
dang- vphase to enter into a mental
state, to start Anga nangaw
nukjyryngaria uchian anga
nangaw nukjyryngwachian
nangna khagalwa dangok. I just
saw you every day, then, when I
saw you every day, I started loving
you.
dangkhym- v to collapse
dangthym- v to collapse (of a road or
bridge), to go into a hole
danyl n ART shield
-dap evsp V on top, V more, V and
add
dap- ~ dep- v to be on top, to press,
keep together by force, pinch
together, to pinch, to crush, to stack
dapet adv insipid, not tasty Jabek
dapet dapet takaidong. The curry is
not tasty.
darai n ART sword
-darang ~ -dyrang encl.phr plural
enclitic
darangba prof anybody, anyone,
nobody, no one Ang songchi
darangba Atong khuchuk olna
mancha. In my country there is
nobody to talk Atong with.
dareng n PLACE edge
dari- v to commit adultery, to have
sex, to be a bad person, to behave
badly
dausik n ANIM parrot
daw- v to open, to peel Kelkhi dawai
tanaimu, dawkha nokmi ruti
sakhawokno. Because somebody
had left the window open, a crow
had stolen bread from the house, it
is said. Nokhap dawbo! Open the
door! Khophylak dawarok. Shes
peeling the skin of a fruit. Narykel
dawaidong. Hes peeling an
orange. Tawti dawbo. Peel the
egg. Khasinai chypangsa dawang
Sangma)

639
daw- n ANIM bird. This is the bound
form of the word taw chicken,
bird that appears before the name
of the bird.
dawblok n ANIM bulbul bird
dawgamdot n ANIM eagle
dawgep n ANIM duck
dawkha n ANIM black crow
dawkharasun n PLANT crow onion,
type of onion
dawkruhasym n ANIM green
pigeon
dawkyru n ANIM pigeon
dawphaw n ANIM owl
dawphylgym n ANIM type of big
eagle
dawreng n ANIM eagle
dawsik n ANIM parrot
dawel- v to be circular
de inter o.k. then, well
deet- ~ diit- ~ diet- v to shit, to do
number two Udo deetna re'ewa.
He went for a shit Nakhung
dietsetaronga. He is picking his
nose.
detheng ~ getheng ppron he/she,
third person singular pronoun
referring to animates
dethengtheng ~ gethengtheng
ppron they, third person plural
pronoun referring to animates
dekdek- v to shiver, to tremble
Dekdekai thyiok. He died shivering.
dekoresyn n ART decoration
del- ~ dyl- v to sting (of a bee etc.)
delang ~ dylang n ART little house for
the spirit of a dead person built
close to the house where the dead
person is burnt to keep his remains
and ashes. The spirit of the
diseased will live in this little house
until it is burnt in the ceremony
called memang saweta about one
year after his death and the spirit
will go to Balphakram.
dem- v to fold
demdong- adj1 weak, soft
dempharai n ART lengthwise cut long
bamboo strip used in the
construction of a house
deng- v to untie Ang chakaw dengbo.
Untie my hands.
dengga n PLANT type of small leafy
green
denggu n ACT extortion, naughtiness
dep- ~ -dap- v to be on top, to press,
keep together by force, pinch
together, to pinch, to crush, to stack
deppyleng- ~ deppyleng ~
deppyleng v to flatten, to make
flat Gari bengbylokaw depylengok,
ytykyimu bengbyloke pylengok. The
car flattened the toad, so the toad
was flat.
di n BODY shit
dibut n ANIM dung beetle
dichongkhanthyi n BODY pygostyle.
The pygostyle is the main
component of the structure
colloquially known as Popes nose,
parsons nose or sultans nose. This
is the fleshy protuberance visible at
the posterior end of a bird (most
commonly a chicken or turkey) that
has been dressed for cooking.
dichyrak- v to have diarrhoea
dicongkhamai n ANIM cloaca
diit- ~ deet- ~ diet v to shit, to do
number two Nakhung
dietsetaronga. He is picking his
nose.
dikhal n BODY/PLACE arse, anus,
bottom
dikyntyk n PLACE toilet
dimai n BODY tail
diphathai n BODY buttock
dipyru- v to have diarrhoea
dipyryw- v to shit your pants
disep n BODY arse crack
disepra n BODY arse crack
dithap n MSRE/ART half (of a
volume), diper Gylas dithapan
phingancha. The glass is not half
full Gylas dithaptharaan. only
half a glass. Jywgaba sagaraiaw
dithap phaetaidonga. The mother
is putting a diper on the child.
dithom n BODY gizzard. The
gizzard, also referred to as the
ventriculus, gastric mill, and

640
gigerium, is an organ in the
digestive tract found in birds,
reptiles, earthworms, some fish,
and other creatures. This
specialized stomach constructed of
thick, muscular walls often
contains swallowed sand or grit,
which helps in the mechanical
breakdown of food.
digi n well, ditch
dikirin- v to tear (clothes, paper etc.)
diksyneri n ART dictionary
dil n BODY body smell Mongmadil
manama. The body smell of an
elephant stinks.
dile n ACT delay
din n PLACE bedroom
dinggarai n ART fish trap
diphing- v to fill Gylaschi tyi
diphingbo. Fill the glass with
water. Gylas phingok, diphingna
manchaka. The glass is full; you
cannot fill it any more.
diphu n BODY a fart
diphu- v to fart
dipot n ART teapot dipot thai ni two
teapots
diprin n PLANT type of vegetable
diri- ~ dyri v to hold
dirikhap- v to catch
diritat- v to hold firmly
disembyl n TIME December
disko n PLACE disco
distrik n PLACE district
disu- v to piss, to pass urine, to do
number one, to urinate
disutyi n BODY piss
disutyitup n BODY urine bladder
-do ~ -odo encl.phr.cl topic enclitic
dode n ANIM peacock
dokhakhu n ART carved, ornamented
and colourfully painted king post
of the bachelors house above the
entrance in between the tie beam
(bylbang) and the peak of the roof
dopho n ANIM owl
doba n SUBST mud
doi- v to catch
doi- v to hold, to grasp
doi- v to scoop into a receptacle
dok- v to take off (clothes), to take
apart, to disassemble, to unblock
dok- v to weave
dokhan n PLACE shop
dokra n ART bag
doksylok- v to be detached
dol n MSRE group dol ni two groups
dolromrom- v to roll up
dolong n ART bridge
dong- v to arrive Ytykyimyng
raiakno raiakno , nokthangchina
dongokno. So then, he went and
went, it is said, and arrived at his
own house, it is said.
dong- ~ dong- cop to be,
identity/equation copula Ue hape
Chigachak teew Kol India kolani
hapan dongwachymno. That place
Chigachak is now supposedly the
Coal India Colony place, it is said.
dong- ~ dong- v to be enough, to be
sufficient, to be OK, to be
convenient, to have passed, to be
past (Nepale) Ytykchido ang
reengsigama nangmi phal?
(Thengthone) Ma ytykchido
dengbo nooknoro. But shall I go
instead of you? Very well then,
in that case, it will be most
convenient, untie my hands, he
said, it is said. Aia! tangka
dongtawanchakthai angdo rong
chyigyksaan raariwa. Damn! To
my surprise the money is not
enough any more, I only brought
ten rupees. Bylsi chykhywdyrang
dongphinokno. Nine years have
passed, it is said. No baji dongok.
Its past nine oclock.
dongwa n ACT event
dongang- v to arrive
dora clf weight of 5 kg.
dorai n PLANT type of vegetable
dorma ~ dolma n ART salary
dosi n ACT blame Aca, naa
angmyng goreaw dosi hynok
nowano rangramyng rajado. So,
you blame my horse, said the king
of the sky, it is said.

641
dot clf classifier for cylindrical objects
like candles and bananas and logs
(but not for batteries) wa dotsa
one culm of bamboo kendel dotsa
one candle pan dotsa one log
drakha n PLANT grape
dram n ART drum, barrel
-duga sfx excessive suffix, V too
much, too V
duk n ACT sorrow, sadness
dukhup- ~ dykhyp- v to put clothes
on someone else
dukung- v to dam, to make circular a
wall of stones in the water in the
river to trap fish.
Baisigathangmaran tyi
dukungokno. Nado
ramramanchakno. The friends
dammed the water. There was
plenty of fish.
dum- v to gather, to swarm
Hajambutungchi umyng
khuchuksang sotamai dumna
dangthokokno. When he was
yawning a swarm of flies entered
his mouth, it is said.
duma n PERS crowd
duma- v to gather (of people)
dumut- adj1 moulded
dumuta n PLANT type of edible
mushroom
dung- v to put something in
something
dung- v to climb Amakdo welang
welang pankambaisang
dungkhatai jalangokno. Pherudo
pan dungna mancha. The
monkey quickly ran away,
climbing to the top of a tree. The
fox cannot climb trees. Bildo
teawba gore dungna sapchanotyi.
Bil does not know how to ride a
horse, it is said, to our surprise.
dupliket n ART a fake
durrrmeme onom sound of a bleating
goat: eeeeee! Pronunciation of this
English word in Atong orthography
would be eeeeee.
dykdyk adv for a short while, quickly
-dykdyk evsp about to V Ransan
songdykdykangaidok. The sun is
about to set.
dykhyp- ~ dukhup- v to put clothes
on someone else
dykyl n PERS Khasi person
(pejorative)
dykyl n PERS cannibal
dykym n BODY/PLACE head, upside,
top
dykymphak n PLACE side where the
head is, space above the head
Dokra dykymphakchi syithaiwa.
The bag hangs above your head.
dykyret- v to threaten
dykyryng- v to make noise on
purpose
dyl n PLANT root, vine
dyl- v to lead Songmongaw dylgabae
Dilbangkongdang
Umangchalmang mutynwano.The
leaders of Songmong were
Dibangkongdang and
Umangchalmang, it is said.
dyl- ~ del- v to sting (of a bee etc.)
dylang ~ delang n ART little house for
the spirit of a dead person built
close to the house where the dead
person is burnt to keep his remains
and ashes. The spirit of the
deceased will live in this little
house until it is burnt in the
ceremony called memang saweta
about one year after his death and
the spirit will go to Balphakram.
dylgaba n PERS leader Songmongaw
gylgabae Dibangkongdang
Umangchalmang mutynwano.
Songgadalaw dylgabae
Thometsangrepha
Rangkhaimadopha mutynwano.
The leaders of Songmong village
were Dibangkongdang and
Umangchalmang, it is said. The
leader of Songgadal village were
Thometsangrepha and
Rangkhaimadopha, it is said.
dym- v to grow (of plants), to sprout
dym ~ dam onom bam! Thud!
sound of something heavy hitting

642
the ground Ytykyimuna tokyrengaw
manaimunga hachina wuuuuuuuk
dym! takramphinoknotyi phylgym
galwaan. So then, having got him
in the neck, the giant eagle fell to
the ground wooooosh bam!
dymbyl n PLANT leaf of a tree which
can be dried and smoked like
tobacco
dymbyra dymbyra adv scattered
about
dymchyrang n ART type of snare
instrument played by plucking
dymdam adj2 naked
dymdam adv gratuitously, simply
dymdym damdam adv carelessly,
just, any way Alaga morotna
dymdym damdam hynna bai.
Dont just give it to someone else.
dyngdai- v to dangle
dyngdang adj1 alone Biphagaba
thyiokno. Kynsangdo gawigabado
dyngdanganokno. The husband
died, it is said. Then the wife was
alone, it is said.
dynggyni n ANIM type of fish
dyngthang adj2 different
dyngthangmancha adv especially
dypyw n ANIM snake
dyra- v to rape
-dyrang ~ -darang encl.phr plural
enclitic
dyri- ~ diri- v to hold
dytyi n KIN uncle: fathers elder
brother
dyw- v to add
-e ~ -ai encl.phr.cl focus enclitic,
occurs on NPs and on locative
clauses.
echaluk n ANIM snail
edres n PLACE address
ek- v to separate Pheruna hyncha
sawana amak, pherudo jalokno.
Baju ekokno. Ytykyimuna
kynsangdo amakdo dyngdanganok.
Because the monkey gave nothing
to the fox, the fox ran away, it is
said. The friends separated. So
then, later the monkey was alone, it
is said.
elong n ANIM type of fish
Endia n PLACE India
engkal ~ ingkal n ART handkerchief
epril n TIME April
epyl n PLANT apple
era n ANIM type of fish
-et sfx causative suffix, on transitive
verbs this suffix indicates that the
action is manipulated, more intense
or emphasises that the O argument
is affected
ga- adj1 good
ga- v to trample, to trod mai ga- to
thresh rice
gaak- vsec to be compelled to, to be
forced to Mongma wa niwamian
manai sachak, khanggal dongok.
Ytykyimu hapsan nukhung raja sa
muchido manai sana nengok.
Ytykyisa dyngthangdyngthang
songchina hapchina jalthokna
gaakok. Because the elephant
tusks were gone, they [the people
of Badri] were not rich any more,
they became poor. So then, if they
would stay together in the hundred
houses, they would run out of
wealth/food. Therefore they were
all compelled to run away to
different villages and places.
gadap- v to step on
gadukduk- v to prod with your legs
or feet Gore jalna rakbebeokno.
Khasinkhalai jalkhalna noaimyng
gadukdukchiba rakkhalai
rakkhalai jalariokno. The horse ran
really quick, it is said. Having told
it to run slower, whenever he
prodded it with his legs, it just ran
faster and faster, it is said.
gajonong- v to trample on, to crush,
destroy Uchi rupekba: Hai angba.
Ang habilchi nok takai
mugabaaw phangnan mongmae
gajononga. Then the frog said:
Come on, me too. The elephant
always crushes my earthen shelter
like a house in which I live.
gajyret- v to crush with your foot

643
gakhat- v to climb Amakdo pan
gakhatna mana. Monkeys can
climb trees.
gakynyng- v to trample on, to crush,
to destroy Uchi rupekba: Hai
angba. Ang habilchi nok takai
mugabaaw phangnan mongmae
gakynynga. Then the frog said:
Come on, me too. The elephant
always crushes my earthen shelter
like a house in which I live.
gaphak- v to hit with your foot while
walking
gaphynek- v to stamp to death
gapyret- v to stamp to death, to crush
with your foot
gapyryw- v to stamp through
something, to pierce by stamping
Thikthak saphaw butangga
rongkhalawan haiba mongmaba
gapyrywmanoknote. The elephant
stamped exactly through the hole
where the rabbit had squeezed in, it
is said.
gareret- v to tread on, to step on
something
gasu- adj1 splendid, cool, terrific
gasylek- v to sprain ones foot
gasyrot- v to slip and fall Wetsa
reengrawrawwachian dobachi
amakba gasorotaimu bunduk
baithongsyrangokno. Once when
they went a little further, the
monkey slipped and fell in the mud
and broke the gun in pieces, it is
said.
gatha n PERS idiot
gathymbylong v to make a hole in a
road or bridge by stamping
Getheng dolongchi
gathymbylongok. He made a hole
in the bridge by stamping on it.
gathyng v to kick
-gaba ~ -ga ~ -gyba encl.cl 1.
Attributive clausal enclitic:
marking a clause as an attributive
clause. Iskyn jangaba songsang
jalangok. He has run away to such
a far country. Phangnan rupek
mugabachido tyi ganang. At
places where frogs live, theres
always water. Rongdyngmaharimu
takruknagaba bostuaw tansetai
jalphinangoknowa. They ran away
leaving behind the things with
which to fight with the Rongdyng
clan, it is said. 2. Lexical
nominalisation: creating deverbal
nouns that can function as
argument in a clause. Tibimi
kyrynggaba rakdugabutuchi
changba nietok. When the sound
of the television was too loud,
someone turned it off.
Phalthangaw chonykgabaaw
naaimyng alsia rajae jalangokno.
Having heard those who despised
himself, the lazy king ran away, it
is said.
-gaba ~ -ga ~ -gyba sfx 1.
Derelationaliser of nouns of close
human relationship, making certain
inherently relational nouns as
derelational. Morot
sabanthaigabaaw kynchi baaimu
dawreng kawwano. A man
carrying his son on his back, shot
the eagle, it is said. 2. Attributiviser
of numerals, interrogatives and the
time noun dakang, making the
word function as a modifier to an
NP. Gynigaba song Badri
Maidugytym. The second village is
Badri Maidugytym. Bigaaw biskut
ranima? Which biscuits shall I
buy? Dakanggaba morot
jalangaimyng, gumukan
jalangthokwa. After the first person
ran away, everybody ran away. 3.
Adverbialiser of the time noun
dakang. Dakanggaba Turachi
The first time I stayed in Tura, I
met Mobbin.
gada n ANIM donkey
gadak- v to cut in pieces, to cut up,
Phylgym chunggaawdo
gadakaimu raakno,
kokchenggumuk. Having cut up the

644
big eagle, they took it with them, a
whole kokcheng full.
gadang n ART shelf
gajol n PLANT type of red carrot
-gak evsp V accidentally
gakat- v to climb
gakji n PLANT lemon
gal n ACT pride, arrogance Phalthang
khuchuk dumgaba sotmaiaw
hongkotna manchaaimyng
thygabaaw gal takokno. He was
proud of the flies which had
gathered in his own mouth and had
died not being able to come out.
gal- v to fall down Kynsangdo
raiwachie napitdo mongma
matsana nekarawrawna
kyrethyngaimyng phepmyng
galsyrangokno napitdo. Later,
when (the animals) were coming,
he feared the tigers, the elephants,
the ones that were continuously
coming closer, so much, he fell out
of the banyan tree, it is said, the
barber.
galruru- v scatter all over the place
galat- v to fall Tyikhal patwachi rong
rimylaimu gasokhokaimuna,
kokcheng galatokno, saphawba
galatokno. When they were
crossing the river, because the
stones were slippery, the kokcheng
fell and the rabbit fell too, it is said.
galcha- v to boast
galdai n PLANT star fruit, carambola,
averrhoa carambola
galjak ~ kaljak n ANIM catfish
galon n ART/MSRE jerry can
gambiri n PLANT type of tree of
which traditional drums called
khem were made
gamchat- v to be valuable, to be
important Breketmyng nyngchi
gamchatga(ba) n ABSTR value
gamsa n ART a cloth
gamsili n PLANT type of tree of
which traditional drums called
khem were made
ganthong n ART stick, handle (of
knife etc.), stump (of a tree)
ganthong thong ni two sticks
ganang v locative/existential verb, to
exist, to be Nang sagyrai
ganangma? Do you have children?
Tanka nichiba ganangchiba ang
nangaw nemnuka. Whether you
have money or not, I like you. Ie
songchi nok kola chit sa ganang.
There are thirty one houses in this
village. Song dam sachi alsia raja
myng sa ganangchym. In a village
supposedly lived a lazy king.
gandalak n ANIM type of frog which
says gagagagaga
gandi n PLANT a log
gandurian n BODY umbilical cord
gandyrui n BODY bellybutton, navel
gandyrui goi korok six
bellybuttons
gang- v to be erect, to have an
erection, to have a hard on, Nang
ri gangama? Do you have an
erection?/Do you have a hard on?
ganggawa n ANIM mosquito
gangma n BODY pimple
gangphu- v to swell, to blow up (like
a chapatti on the fire)
gangthai n BODY fin (of fish)
gantai n ANIM type of beetle
ganthai n ANIM small brown insect
that makes a loud whistling sound
gantheng n PLANT stalk
gantirengreng n ANIM type of beetle
that makes a very loud and high
pitched sound
gapsan ~ hapsan adj2 the same,
together
garamak n ART storage rack
garan n FOOD jerky
gari n ART vehicle, car
Garo n PERS/ACT Garo (person and
language)
garu n PLANT mustard
gasam n TIME afternoon, evening,
later part of the day Gasam tin

645
bajichi reengni. We will leave this
afternoon at three oclock. myia
gasam yesterday
evening/afternoon, taini gasam
today in the evening/afternoon, this
evening/afternoon Taini gasam
reengphinni. I will go back in this
evening/afternoon.
gasam- v to be evening Gasamok. It
has become evening. Gasamnaka.
It will soon be night.
gasam gasam adv sometimes, seldom
gasamphang n TIME afternoon,
evening, later part of the day
-gat evsp V up onto, to start V-ing
gat- v to dig
gat- v to put in/on, to load into/onto
Phagongmachi sa gataimyng
tyinyngsang dangangokno.
Having put the child on his
shoulders he entered into the water,
it is said.
gatdap- v to stack, to put on top
gatha n PERS fool, crazy person
(masculine)
gathi n PERS fool, crazy person
(feminine)
gawak n BODY disease
gawang ~ guwang n ANIM spider
gawangsyryng n ANIM spider web
gawasu n BODY rib gawasu tyn tham
three ribs
gawi n PERS female, girl (unmarried)
gawigaba n PERS wife
gawsu n BODY rib
getheng ~ detheng ppron he/she,
third person singular pronoun
referring to animates
gethengtheng ~ dethengtheng
ppron they, third person plural
pronoun referring to animates
gebeng n ABSTR width, breadth
geng clf classifier for long vegetables
rasunok gengsa one spring onion
genji n ART tank top genji khung/jora
ni two tank tops
ger n ART gear
gesep ~ gysep ~ gisep n PLACE space,
interval Bandi nochie atongaw
balkhunano? Samung sagyraichie
kanwani chungwani kalai
kharutchungaban thengchidokno.
Dykymchi khyrykrara, dikhal
gesepchi dirara, pipuke moina
chongchang takariokno. When I
talk about Bandi, what can I say?
He had no clothes; he wore a
loincloth tied together with sixteen
knots. On his head swarmed the
lice, his ass crack was full of shit,
his belly looked like a bird cage, it
is said. Panga bylsidarang
reengok umi gesepchian
thyikhal goisachian rongkhal
khalsa ganangchym. Ue
rongkhalaw Durakhal
myngwachym. Many years ago,
in a river, there was a cave, but
not any more, it is said. That
cave was supposedly called
Durakhal.
giching ~ gyching adj2 aslant, slant,
diagonal
ginggang adj2 having, with
gisep ~ gysep ~ gesep n PLACE space,
rongkhalaw Durakhal
Durakhal.

646
gisep gisep ~ gysep gysep adv from
time to time Nanage song
janrukok. Umi gymyn bichiba
gisep gisep chiti saietrukarinaka.
Our countries are very far from
each other. Therefore we will
sometimes write each other letters
from time to time.
git n ART music, music with lyrics, a
song
githing adj2 unripe
githyng ~ gythyng ~ githing adj2
unripe, uncooked, raw
gobormen n ACT government
godot- v to bump Cha rongchi
godotwa. I bumped my food on a
stone
gogak n ANIM beetle
gogat- v to carry on the shoulders
gogylek n ANIM cock, rooster,
cockerel
goi n PLANT betel nut, areca nut
(Areca catechu)
goi clf non-specific classifier
goichara n PLANT a young betel nut
tree
goichara n PLANT betel nut sapling
goilapan n FOOD betel nut and
paan/pan
Goira n GEO the god of thunder goira
kawa the god of thunder shoots /
the thunder roars goira byl tanok
the god of thunder has struck /
lightening has struck
gol ~ gool n ACT goal Getheng gol
saak. He got a goal (in football).
golap n PLANT rose
golmal ~ gormal n ACT a fight, a
quarrel, chaos
golmen n ACT government
golpho n ART story
golpho- v to talk extensively
gom n PLANT wheat
gom- v to bend gomga(ba) leech
gomagundai n PLANT thick type of
banana
gompyra n ANIM poisonous black ant
gomynda n PLANT pumpkin
gomynthyri n PLANT type of
vegetable
gondu n ANIM rhinoceros, rhino
gong- v to be willing, to agree,
Atakaronga? nookno
mongmado. "Niwa. Muariaronga
ytykyian" nookno amake. Atongba
sakhawarongkhonne nangdo
nookno. Sakhawcha naa. Niwa
naa, gongwanasa balwa sakai
muarong noatakokno amakba.
What are you doing?, said the
elephant? Nothing. Im just sitting
here like this, said the monkey.
Maybe you are stealing
something! said [the elephant]
Im not stealing! Theres nothing
going on. Im just sitting here
enjoying the wind because I want
to, said the monkey. Ha ambi
ang chakaw khenetkhu nowano.
Gongchak angdo sa jywtynnaka
tedo. Nangna myia khenetokte
nowano. Hey grandchild, please
scratch my arm! she said. I dont
want to any more, I am going to
put my child to bed now. I
scratched you yesterday, she said,
it is said.
gongchit n ANIM type of black beetle
gongdang adj2 bent Ytykyimyng
tedo magachakdo, ytykyi pan
gongdang takgabachi ne
nangwanote. Ne nanggaba
okkumachi jywsawtheriaidokno.
So then, now, as for the deer, so
there was a bees nest hanging
from a bent tree branch, it is said.
He was fast asleep under the bees,
it is said.
gonggong- v to bend over
gop- v to bury, to hide Nang baba
noksamchi tangka gopgaba
ganangno. Under your fathers
house lies buried money, it is said.
Morot thyigabaaw hanep gopnaka.
Tomorrow they will bury the dead
person. Thengthon morot tangka
bisyl pangai khaigabaaw nukokno.
Ytykyimyng hap damsachi syruk
syruk gopaidongano. Uaw
nukaimyng gethengdo thymai

647
chaioknoro. Thengthon sees a man
who was carrying a lot of coin
money, it is said. So then, he is
hiding secretly, it is said. Having
seen him, he lay in ambush and
watched him, it is said.
gopram n PLACE grave
gora n ART large earthen pot in which
rice liquor (chyw) is made.
gorai ~ gore n ANIM horse gore
dung- to ride a horse
gorial n ANIM crocodile
gorong- v to meet
-gorop evsp V with a whole group, V
together
gorothop n PLANT type of small leafy
green
grem clf gram, gr.
gremyr n ART/ABSTR grammar
guchung n ART ladder
guduk- v to wiggle, to be unstable, to
wobble, to move (unstably) guduk
tak- almost [verb] Nalam
gudukwachie teewdo tyi
thangpytpytaimyng jyksaiaiawan
Nawengawmu Kumiribaawma
khamoknowa. When the nalam
(type of fish) wiggled, water
splashed on the married couple
Naweng and Kumiri and burned
them, it is said. Rong gudukaimu
galatok. Because the stone moved,
I fell. Aia! Udo magachakdo
khorate noaimyng rykoknowa.
Tharapna guduk takwachiba
Hey, this deer is lame! he said
and chased after it, it is said. When
he almost caught up with the deer,
it run away fast, it is said, the deer.
gugyreng n ANIM type of grasshopper
gukchepchep n ANIM grasshopper
gukmadym n ANIM grasshopper
gulgulgalgal onom growling noise
that the stomach makes Pipuk
gulgulgalgal takaidonga. My
stomach is growling.
gumi n KIN brother-in-law: elder
sisters husband or husbands elder
brother
gumuk n MSRE all, whole,
everybody, everyone, everything
-gumuk encl.phr all, whole
gumuksangan adv everywhere
gumuksangan ~ gumuksang prof
everywhere Gumuksangan ganang
ukching. There are leeches
everywhere. Getheng gumuksang
reengok. He went everywhere.
gun montyro manga(ba) n PERS
person who can control the spirits
gunda n PERS brawler, fighter
guri n GEO mist, fog Guri thupa. The
fog is thick.
guruchup- v to be shrouded in clouds
Waimong nukcha, guruchupok.
Waimong mountain is not visible,
it is shrouded in clouds.
gurum- v to collapse, to break off and
fall down Banggyriaimu nok
gurumok. Because of the
earthquake the house has collapsed.
Narykhelchak gurumok. The leaf of
the coconut tree has broken off and
fallen down.
gusu- v to cough
gusum- v spoiled (only used with
meals) Mai jabek gusumok. The
rice and curry are spoiled
guthini n ART spear
guthini n ART bamboo spear which is
part of an elephant trap
guthyni n ART walking stick
gutum ~ gytym ~ gytum n ART
village
guwang ~ gawang n ANIM spider
gycheng n PLACE side, near Jyksaian
phonggychengchian muaidonga.
The married couple are sitting near
the cooking place.
gyching n LOC/ABSTR vicinity,
angle, inclination gychingching
mu- to be tilted, to make an angle
gyching ~ giching adj 2 aslant, slant,
diagonal
gyl- adj1 strong Usang, songga
Manggagremi banthaidarangba
raiaaithokaidongano,
Rakarelwakmadare,
Gyrynggyrang, Saljapang,

648
Aragundi, Motbanda,
Asyngduraparaba gumukan
raiathokaidongano. Chakphong
gylgabasano, kara
khyrynggabararasano,
alamylachagabasano. They are all
coming to there, the young man
from the strange village of
Manggare: Rakarelwakmadare,
Aragundi, Motbanda,
Asyngduraparaba, they are all
coming, it is said. They are men
with strong arms and tight veins all
over, it is said, they are not
ordinary men, it is said.
gyl- v to collect, to gather
gylas ~ gilas n ART glass or its
volume, glassful cha gylas ni two
glasses of tea. Gylas goi tham
baiok gethene. He has broken
three glasses.
gylgyl- v to roam
gylja n PLACE church
gyljanok n PLACE church
gymyn postp cause, reason, because
of, about Ue gam pangwami
gymyn kam pangwami gymyn
gethengtheng mykbyrukokno.
Because of this wealth and these
riches they had became jealous of
one another, it is said. Unmi
gymynsa ie hapawe Badri
Rongdyng Hawai noyi
techinakhyngkhyng myngwano.
Thats precisely why this place is
still called Badri Rongdyng Hawai
up till now. Uan jorami gymyn
chosa golpho kaetwa. I have told
a bit about that love match.
gynching ~ giching adj2 aslant, slant,
diagonal
gyp onom hitting sound: thunk!, tap!,
bam! Ue uawdo kunsang gyp
satetok. He hit him bam! with a
stick.
gyryp- v to cover
gyryw- v to shake (an object that you
can pick up, a non-fixed object)
gysep ~ gisep ~ gesep n PLACE space,
rongkhalaw Durakhal
Durakhal.
gysep gysep ~ gisep gisep adv from
time to time Nanage song
janrukok. Umi gymyn bichiba
gisep gisep chiti saietrukarinaka.
Our countries are very far from
each other. Therefore we will
sometimes write each other letters
from time to time.
gythyng ~ githyng ~ githing adj2
unripe, uncooked, raw
gytym ~ gythym ~ gutum ~ gytum n
ART village
ha n SUBST soil, earth
ha procl Take this! Take this from
me.
ha haw- v to cut/clear the land to
make a haba Phasgaba
hahawchenga. Umungsa ha
hawaimungsa wacham tana.
First we clear the jungle. Then,
having cleared the jungle, we cut
the old rice stalks.
ha kam- v to clear the field, to cut the
jungle to make a field, to tear out
weeds

649
haba n PLACE dry rice and vegetable
field on the slope of a hill made by
cutting away and burning the
jungle haba tym ni two dry rice
and vegetable fields on the slope of
a mountain
habacheng- vB to start, to begin
Uchisa matsana makbulna
mongmana paichaaimung byldyng
byldang jalna habachengok. Then,
not bearing the tigers and elephants
any more, they started to run all
over the place.
habachenggaba n ABSTR beginning
habanok n PLACE rice field house
habykung n SUBST sand
habyreng n PLACE old haba
habyri n GEO hill, mountain habyri
thut tham three hills, mountains
hachak n ACT wages
hachepchep n ANIM grasshopper
Hachyk n PERS/ACT Garo (person
and language)
hadawak n PLACE lower side of a
hill, low ground
hagun n PLACE old plot of land in a
haba Baidamdo
hawangman gaba hagun
saangmangaba
harynthangthangaw kanga. Some
people occupy their old already cut
plot, their own parcel which is
already used completely.
hagyrsak ~ hagylsak interj
interjection of astonishment Naa
sagyrai mylthengtheng
bisangreengaidong naa bunduk
pairama takaimu? nookno. Myla
hagylsakno sagyraido. Naa
bisang reengaidong naa sagyrai
mylthengtheng? You child which
is still very much too small, where
are you going carrying that gun in
your hand? he said. Good Lord,
that child is small! Where are you
going, you child who is still much
too small?
hagyrsak ~ hagylsak n ABSTR
everything, all, world Mekalaia
hagelsakgumukchi
wabatsyranggaba. Meghalaya is
the rainiest place on earth.
hajagyra n ACT the first weeding of
the haba Mai kaimanwamungsa
hajagara kama. Hajagara
kamaisa kamaimung kynsange
jakun kama. Jakun kamaimungsa
nobembyl, oktobylsomaichi saigyn
khana. Umungdo mai mynokodo
maidan syla toka. Having planted
the rice, we weed the land for the
first time. Having cleared the
weeds for the first time, we will
clear them for a second time.
Having weeded the land for a
second time, in October or
November we do a third weeding.
Then, when the rice is ripe, we
celebrate the new rice festival.
haka n PLACE upper side of a hill,
high ground Hakasang tawangbo.
Climb to a higher part of the hill.
hakha n PLACE a mountain slope, a
steep slope
hakha- v very tight Ian hakhaai
khabo. Tie this very tightly.
hakhong n PLACE valley
hakhung n GEO river bank
hakhyng n PLACE area
hamai n ANIM type of white
earthworm
hamang n SUBST soil, earth, clay
hamangkyrang n ANIM scorpion
hamangkyrang n ANIM scorpion
hamangrong adj2 brown
hamat adv troublesome
hapal n PLACE outside
haphal n PLACE field
harongrong n PLACE lower side of a
hill, low ground Harongrongsang
wylangbo. Go down to a lower part
of the hill.
haryn n parcel (of land), plot (of
land) Songgumuk thomaimyng
haba haryn harynaw sowalni.
The whole village gathers and will
divide the haba parcel by parcel.
haryn n PLACE plot of land, parcel
Songgumuk thomaimyng haba
haryn harynaw sowalni. The

650
whole village gathers and will
hasaw n ANIM type of black snake
with red neck and head
hasel adv for no reason, uselessly,
troublesome
hasong n PLACE country, village and
surrounding lands
hathywkong n GEO a puddle
hatykylok n GEO a puddle
hawai n PLACE plain area
hachi- v to sneeze
hai procl Lets go! Come on!
hai determiner the, this, that (thing or
person just mentioned)
hai- v general verb, to do whatever,
to happen, to do this/that, such and
such happens, it is like Uchie
karydyl chunggaba haiwano. ha
ambi ang chakaw khenetkhu
nowano. Then the hanging root did
this, it is said. Hey grandchild,
scratch my arm! she said, it is
said. Phylgymsa haiwa naa ue.
Garu ramgabachi nanang garu
ramtananggachi dietdapai
tanangwa nookno. The giant
eagle did this, oh you! In the dried
mustard, in our dried mustard, he
left a big shit, she said, it is said.
Ytykyi phetaaimungna haiokno,
nokchina janggalan: Bie nang
jongdyrange? Nang jonge bie?
nookno janggaklchi syngokno. So
when they had arrived, it was
like, at home, all of them:
Where is your younger brother?
Your younger brother, where is
he? she said, it is said.
Uchiansega haiokno, pheru
nuksegaakno sawamiaw. Then in
turn, this thing happened, the fox
spotted some food, it is said. Rymai
sawachie, amakdo pan gakhatna
manano. Khusumdo haiokno.
Khusumdo gakhatna manchano.
Having cooked and eaten it, the
monkey is able to climb into a tree,
it is said. As for the turtle, he did
whatever, it is said. The turtle
cannot climb trees, it is said.
haie ~ hai- determiner
determiner/general noun, Let me
see; uh; whatchamacallit
Utykyi dongwano ie
Wiliamnagarmi haie Dorenggo
Wadachongmi histyri. Thats how
it is, it is said, Williamnagars
uh Dorenggo Wadachongs
history. San sachi amak aro pherue
ytym myng ni bajunoro, mo.
Ytykyimuna kynsangdo atakokno?
Ytym myngnie sangumuk
gylgylarongno glgylarongno
gylgylarongno. Uchie amakan
haiaw rekthai myngaaw
nukoknoro. One day the monkey
and the fox were friends, it is said,
OK. So then, then, what happened?
They, the two of them roamed
around the whole day. Then the
monkey saw these ripe bananas.
haida procl I dont know.
hajal ~ hajar num thousand hajal sa
one thousand.
hajam- v to yawn
hal- v to feed muthai hal- to
breastfeed
hala kha- v to wake someone up, to
disturb someone
haldun- v to feed, to maintain Ie
alsia raja atykyi khengaidok?
Atykyian jykaw haldunna
manaidok? noai morotdyrang
chanchiphinaidoknoro. How does
this lazy king live? How does he
feed his wives? thought the
people, it is said.
halsia ~ alsia n PERS lazy person
ham- v to build, to construct
hama n PLACE underneath, below,
space between the floor or the base
of something and the ground
Tausagrai nok hamaaw
jalphakangaidonga. The chicks are
running under the house (from one
side to the other).
hambun tw later (but not today), in
the future

651
hampyi tw in the late afternoon, in the
evening
hancheng n SUBST sand
hantung- adj1 to be a dangerous
place Mungma pangwachi palyng
hantunga. When there are a lot of
elephants the jungle is a dangerous
place.
hantung- v to feel secure, to feel safe
Morot pangai raibo,
hanthungkhala. Go with lots of
people, you will feel safer.
hanep n tomorrow Hanep sansaanok.
Tomorrow there will be one day
left.
hang- v to warm your hands by the
fire
hangkhal n GEO cave, hole
hangkhal khal chatgyk eight caves
hangkhyn n ANIM small insect that
lives inside wood and eats it
hangkyn n ANIM type of ant
hangkyn raja n ANIM type of ant
hanseng- adj1 happy, joyful
Ytykyimyng nado amak di saaimu
hansengtokaidoknoa. So then, after
the fish had eaten the monkeys
shit, they were all very happy, it is
said.
hansenga ~ hansinga adj2 beautiful
hanthi- v to divide Angdo
dyngthangmancha nang
khagalchido nangmi gamaw
angna hathiphabo. If you love me
especially, divide your wealth for
me.
hanthi- v to divide, to share Umi
gesepchian gamaw hanthiokno
wagabae. In the meantime the
father had divided his fortune, it is
said. Songgumuk thomaimyng
haba haryn harynaw hanthini.
The whole village gathers and will
Je haryn nigababado uan
hanthirukai hawa. As for those
whoever does not have a plot, those
mutually share and clear the land.
hap n GEO place Umigymynsa ie
hapawe Badri Rongdyng Hawai
noanowa, aro rangawba mykha
badri myngwanowa. Thats why
this place is called Badri Rongdyng
Hawai, it is said, and the rain is
also called mykha Badri, it is said.
hap n QUANT half
haphu- v to blow
hapjyt- v to move house
happen n ART short pants, shorts
hapsan ~ gapsan adj2 the same,
together Ang nangmi/nangmyng
hapsan chunga. I am as big as you.
hat- v to fuck Teraka krismassomaichi
ue gawiaw babylsichi nokwengchi
hatok angdo. Last year at
Christmas I fucked that girl on the
floor in the kitchen.
haw n KIN uncle: mothers brother
haw- v to clear/cut the jungle to make
a rice field
hawnokhol n KIN father-in-law,
addressed as mama
hawchi dem over there, yonder
hawe ~ haw- dem that over there,
very far, non-visual demonstrative
hawtyi adv for some time
Thorokaimyng hawtyi rypokno
magachake. Having jumped down,
he stayed in the water for
sometime, it is said, the deer.
heng- v widely spaced, sparse
henraiting n ACT handwriting
het- v to clean an orifice or hole
Nakhal hetbo. Clean your ears!
hijra n PERS gay person, homosexual
hil n ART heel (of a shoe) Teew
reenggaba gawi, longpen kanai
juta hilaw The girl who just went
by wearing trousers and shoes with
high heels
hira n ART diamond
hit- v to command
hmm ~ mm procl no
ho- v to jump
hoong procl yes
hogol n snoring Juwchenwachi nang
hogol rawa. For the first part that
you were asleep you snored
hojokjok- v to jump up and down
hok- v to call loudly

652
Holen n PERS/PLACE Holland, Dutch
Holenmorot Dutchman /
Dutchwoman
hongkhot- v to come out Ytykyimyng
amak gawigaba biphagabaaw
kynaw thup, thup tokaidonganoa.
Uchian amakmu disa chat chat
hongkhotaidongano. So then, the
monkeys wife beats her husbands
back tok, tok. Then the monkeys
shit comes out squirt! squirt! it is
said.
hongkot- v to come out, to ejaculate,
to cum
hot- v to extract
huraw n ANIM gibbon
huk- v to sweep together
huksetgaba ~ huksetga n ART
dustpan
hung- v to swim
hup- v to suck
huraw n ANIM gibbon, Hylobates
hoolock
husyring n ANIM rabbit
hy interj no
hyiawchi dem over there, yonder
hyiawe ~ hyiaw- dem that over there,
very far, emphatic non-visual
demonstrative
hyn- v to give Angna tangka ratja
banga hynetbo. Give me five
hundred rupees. Nangba happy
new year hynetaidong. We wish
you a happy new year too.
hyt interj interjection to chase a
person or animal away, interjection
of anger
ie ~ i- dem this, proximal
demonstrative
ilektrisiti n ART electricity
inchi clf the width of the upper joint
of the thumb, i.e. the joint under
the nail
India n GEO India
inggeech
n ACT engagement Inggeech khaak. I
am engaged to be married.
Inggylan n PLACE England
Inggylis n PERS English
ingkal ~ engkal n ART handkerchief
insuren n ACT insurance
isor ~ isol n ABSTR God
isykyn adv this much, this many
itha ~ ita n ART brick itha thut sa one
brick
itihas n ACT history
ja interj interjection to chase a cow
away
ja n TIME month, moon ja sa- to
wake up ja phetok the moon has
risen
jabek n FOOD curry
jachung n KIN 1. (Siju and Badri
dialects) the relation of a man and
his wifes elder sister, or of a
woman and her younger sisters
husband, 2. (Badri dialect) sister-
in-law: elder sisters of ones wife
jaga n ART a trap Jaga saakno
uchie, taw pangai banokno. They
set traps and then caught many
birds, it is said. jaga sagaba
someone who sets traps, an enemy
jagaru n PLANT type of vegetable
jajol ~ jagol n PERS person with
long legs
jakhop n ART shoe
janaw n KIN elder sister. Is also used
to address an older female cousin
or a woman older than the speaker.
japhang n PLACE foot of a tree
Uchie panjaphangchi
thymsawaimu, khusum amakmyng
rikaran kakhchotokno. Then,
having lain in ambush at the foot of
the tree, the turtle bit and tore off
the monkeys balls, it is said.
jaraw- adj1 for a long time
jaryt n PLANT chilli pepper
jaboldam n ART garbage heap
jabyra n PERS fool, crazy person
jada n PERS stupid person, idiot
jadu n ACT magic
jagydok n BODY biceps, strength Ido
sagyraido hambundo
chungwachido alamyla byldo
bylnikhon, jagydokdo
jagydoknikhon. In the future that
child might really become a bit

653
stronger, it might really get
strength.
jagyra n PLACE right, right hand,
right hand side
jagyryng n ACT shadow cast by a
person
jagysi n PLACE left, left hand, left
hand side
jahas n ART ship
jai- v to scold someone
jai- v to oppose, to refuse Unasa
Ketketa Burae pheruna jaisakna
chol mancha
chomotaimyng:Acha, ytykchido
anga ang nokkhuthaisa anga
dykdyksa chaigamaimyng kepai
mukhuna noaimyng
tholthiriokno, pheruna Ketketa
Burae. Because Ketketa Bura could
really not come up with an idea to
oppose the fox, he said: Ok, in
that case I will quickly look out
from the top of my house and sit
and cry, he lied again to the fox, it
is said, Ketketa Bura.
jajong n GEO moon
jajyreng n ACT confusion
jajyreng- vdat to worry Ang nangna
jajyrenga. I am worried about you.
jakhal- v to use
jakhalthaw- adj1 to be very useful
jakhep clf as much as is contained in
the palm of the hand when
clenched, the quantity contained in
the closed palm
jaksithem n ART ring
jakun n ACT the second weeding of
the haba Mai kaimanwamungsa
jal- v to run away Iskyn jangaba
songsang detheng jalangok. He
ran away to such a far country.
Uchisa matsana makbulna
mongmana paichaaimung byldyng
byldang jalna habachengok. Not
bearing the tigers and bears any
more, they started to run away all
over the place. Tharapna guduk
takwachiba tarakai jalariano
magachake. When [the Bengal]
almost catches, [it] just runs away,
it is said, the deer.
jaljeng n ART cupboard
jalphakang- v to run from one side to
the other
jam- vphase to complete, to finish
Sawa jamkhucha. I have not
finished eating yet. Rang
nemchengama nanang chyw
jamchenga noai rangmu chyw
ryngsusai chyichie, range san chi
byri wawano. Will our liquor
finish first or will the rain stop
first? they said and while they
were trying to compete with the
rain in drinking, the rain fell for
fourteen days, it is said. Gasamchi
rymai sawa jamchagabaaw
Nawengpara Kumiriparae
sawaimyng garan bawai
tanoknokhon. The food that they
didnt finish in the evening, having
fried it, they might have dried it
above the fire, it is said. Ang
amaparami nokchi randai sana
jamcha. At my mothers house,
there is always meat to eat. ja
jamangwasang at the end of the
month
jama n ART shirt
jamang- v to set (of the sun) Rangsan
jamanga. The sun sets.
jamkhamwa n ABSTR the last one
jamura n PLANT pomelo
jan- adj1 far Haba janrukwaan
nukruketchawa. The rice fields are

654
very far apart from each other, you
will not see each other.
jang- adj1 quick
jangjot adj2 biconcave, curved on
both sides like the inner surface of
a sphere, narrow in the middle
Bandi nochie atongaw
chongchang takariokno. Chae
dabakun tykyi ympong jangjot
takarioknotyi. What more can we
tell about this so called Bandi?
When he was a child he was half
naked, wearing only a loin cloth
tied together with sixteen knots, it
is said. His head was all lice, his
arse crack was full of shit, his belly
was big like a birds cage, it is said.
As for his legs, they looked like
coconuts on sticks: bulgy in some
parts and very thin in other, it is
said to our surprise.
janggal n QUANT everybody,
everything, all, all of them, all of it
Sagyrai mylgabami
dadadarangawdo janggalawan
monokokno. Phylgym chunggaba
monokrumokno myng korokawan.
As for the brothers of the small
child, they were devoured, it is
said. The big eagle had devoured
them all, the six of them.
janggi n ACT life Kynsangdo matsado
morotsyn manaimyng raiwilokno
alsiado. Raiwilwilokno. Matsae
atakna? chanchiaidongano ue.
Kyreaidokno kyrewaba. Anga
thyie thyimanok. Ichian: Anga
janggiba thyimanok. Later, the
tiger caught the scent of the human
and walked in circles around the
lazy person, it is said. What does
the tiger want?, he thought. He
was very afraid, it is said. Im as
good as dead! Then: My life is
already gone.
janggi khenwa n khjyks ACT life Ang
janggi khengwa gumuk Atong
khuchuk balwa. I have spoken
Atong all my life.
janira n ART mirror
jantg- v to be quick
janti n ART filter for rice beer (chyw).
Woven cylindrical filter made of
reed that stands in the jug (gora) to
form a permeable membrane
between the fermented rice on the
outside of the filter and the
alcoholic water inside the filter.
Water is poured onto the fermented
rice and the alcoholic liquid is
collected inside the filter and
scooped out with an abek.
januari n TIME January
jap n ART trap to drive away enemies.
A pile of rocks is stacked on a hill
behind a plank. The plank is tied to
a tree. When the enemy comes, the
rocks are released, roll down and
crush the enemy.
jap- v to pile up
japang n PLANT tree trunk
japrukruk adv one on top of the
other, in a pile
jarambong n GEO full moon
jari- v to be startled
jasa- v to wake up, to get up, to get
out of bed
jaseng- v to shine
jat n PERS tribe, race
jatha n ART a spear
jatram n PLANT type of medicinal
plant
jaw- v to fry
je prof any, whichever, whatever Jeen
sanchi morot thyiok. One day
somebody died. Jemi sanchi
Dibangkhongdangaw matsa kakok.
On a certain day a tiger bit
Dibangkongdang.
jechiba prof anywhere
jekhai adv for example, for instance,
as Rongkhaisang jalanggaba
Thometsangrepha
Rangkhaimadophae nukkhung sot
dokmyng, jekhai Atongsang

655
balchido nokkhung rum thammang
ganangno. Thometsangrepha
Rangkhaimadopha, who ran away
to Rongkhai, had about sixty
houses, as you would say in Atong,
about sixty houses.
jel- v to increase, to multiply, to be
numerous Mahari jela. The family
is big.
jenetene adv somehow
jenetne adv somehow Ytyktyimyng
jenetne rajamyng noksang
phetangokno. So then, somehow
they reached the kings house.
jeng n PLANT plant of which brooms
are made
jero num zero
jesangba prof wherever, somewhere
Getheng jesangba reengok. He
has gone somewhere.
jesykyn prof Jesykyn nangchi
ganang chynaribo, kamalna.
However much you have, just offer
it to the priest.
jineral general Dakanggabado
jineral mitingchengni. Umungsa
song gumuk thomaimung haba
haryn harynaw sowalni. First
they will start with a general
meeting. Then the whole village
comes together and they will divide
the haba plot by plot.
jingjong- adj1 wiggly, unstable
jingonget- v to shake
jinka n PLANT type of vegetable
jinma n QUANT group, herd
Bajudyranggumukan jinmami
palyngsang sikhal khana
reengwa. All the friends went to
the jungle in a group to hunt.
mongmajinma a herd of elephants
jit- ~ jyt- v to move Rongawan jitna
jamchano. He could not move the
rock, it is said.
joba n PLANT Chinese rose
joi- v to drag, to catch (by dragging a
net through the water)
jojong n KIN younger brother. Is also
used to talk about or address a
related younger male of your own
generation: cousin, to address a
young male unrelated person
younger than the speaker.
jok- v to escape, to be freed, to come
out, to leak out, to jump because
something startled you Atakna
reengwa. Balchachido tokni,
madame. Jora chaiwa. Sala!
mylteng teeuan jora chaina,
roalan jokkhucha! Why did you
go? If you dont tell me I will hit
you, said the teacher. I saw my
lover. Damn! you are still too
small now to see your lover; you
have not even finished primary
school yet! Uching kakaimu thyi
jokok. Because he got bitten by a
leech, blood came out. Naa ie
sastiaw rakna manchido jokangni.
If you can endure this punishment,
you will be freed. Khabakaimyng
wangwachie kynsange: Ai! ido
alsiae kakate! Sokchakate angdo
noaimyng, matsado jenethene
jokaimyng jalangoknoro. After
having grabbed him firmly and
bitten him, the tiger said: Ouch!
This lazy person bites, Im telling
you! and having somehow
escaped, he ran away, it is said.
Robolmi balwa jokok. Air has
leaked out of the football.
jokal n PERS/ART comic strip,
cartoon, anime; a character from
one of these categories
-jokjok evsp V up and down
jokset- v to drain
-jol evsp V quickly
jol- v to roll up
-joljol evsp V quickly
jolpi n ART bamboo fish trap
jom- v to sneak, to sneak up on
somebody
jomphol n ART hoist, crow bar, pry
bar
jong n KIN younger brother. Is also
used to address a younger male
cousin or an unrelated man

656
jongsyri n KIN brother-in-law:
spouses younger brother
jonja n PERS twin
jonong- ~ jorong- v to dissolve Chini
tyichi jorongok. The sugar has
dissolved in the water.
jora clf classifier for things that occur
in pairs sendel jora sa one pair of
sandals muthai jora sa one pair of
breasts mykren jora sa one pair of
eyes
jora n PERS partner, love (person),
match in love
jorong- ~ jonong- v to dissolve Chini
tyichi jorongok. The sugar has
dissolved in the water.
jot- v to prod, to point, to fidget Jong,
naa reengaribo,chaksi jotetgaba
thongthong reengaribo. Brother,
you just go. Just go straight in the
direction of the finger with which I
point. Waiphinwami gesepchi
baratdugaaimu, wana jywna
baratai, baina tyngna baratai,
nawna baina baratai, dethengdo
danganaan chaithylaisa,
muarongno, gopjyrujyrutykyi.
hasa bamai, chaksi jotai, chaksi
phaiai nemen chanchiaidongno. In
the time before he returned, having
become very ashamed (when he
was thinking of what his parents
would think of him), when he
entered (his fathers house) he
looked away and looked down,
bending his head, fidgeting with his
fingers and wringing his hands he
was in deep thought, it is said.
jothat- v to prod
jotkhyngkhyng- v to mash
joton n ACT attempt, try joton kha- to
try, to make an attempt Hanep
raina jotong khani. I will try to
go tomorrow.
jotpyryw- v to pierce
juk- v to wink
jul- v to walk through the jungle with
difficulty
jul- v to jack up, to lift up Rukwakdo
pan mylgaba paiatakaimu rongaw
julokno. The toad took a small stick
and jacked up the stone, it is said.
jul- gul- v to walk through the jungle
with difficulty.
julai n TIME July
jumang ~ jywmang n ACT dream
Atong jumang nukwa? What did
you dream? / What dream did you
see? Taija walchi jywwachi
jywmangsang banggirigaba nukwa.
Last night at night when I was
sleeping, I saw an earthquake in
my dream.
jumu- v to collect
jun n TIME June
jut- v to encourage Getheng angaw
sakhawkhalna jutwa. I encouraged
him to steal.
juta n ART shoe Teew reenggaba
gawi, longpen kanai juta hilaw
The girl who just went by wearing
trousers and shoes with high
heels
jyk n KIN spouse
jykjak- v to be noisy, to make noise
jykmong ~ jykmongma n PERS first
wife of a man who has two wives
jyknyi n PERS widow, widower
jykrat- v to accuse of adultery
jykri n PERS widow, widower
jyksai n PERS husband and wife,
married couple
jyktyi n PERS second wife of a man
who is already married
jykyryi n PERS widow, widower
jyngjang adj2 dense
jyryk- v to have a nice taste
-jyryng evsp V daily, V all the time
jyryng jyryng adv always
jyryngnam adv always
jyryp adv quietly jyryp mubo sit
quietly
jyrypet- v to shut somebody up, to
make someone be quiet
jyt- ~ jit- v to move Rongawan jitna
jamchano. He could not move the
rock, it is said.
jyw- v to lie down (both the
movement and the position), to
sleep

657
jyw n ART a flattened bamboo used
to make mats jyw sa one flattened
bamboo damdyl khaw sa one jyw
of a damdyl
jyw n KIN mother
jyw wa n PERS parents
jywbydyi n PERS old woman,
woman with children
jywpara n PERS mothers house,
mothers household
jywri ~ jywryi n PERS child who
lost his mother
jywdap- v to lie on
jywmang ~ jumang n ACT dream
Atong jywmang nukwa? What did
you dream? / What dream did you
see? Taija walchi jywwachi
jywmangsang banggirigaba nukwa.
Last night at night when I was
sleeping, I saw an earthquake in
my dream.
-ka ~ -naka sfx imperious future or
certain future suffix
ka- adj1 bitter
kadymbai n BODY chin
kamyn n BODY beard
karan- ~ kharan- v to be thirsty for
Tyiba karanok baisiga angdo.
Ang tyi chosa ryngna. I am thirsty
for water, my friend. I want to
drink a little water.
kabal ~ kabar n ART cover
kabin n ANIM type of big black ant
kai- v to plant Ha
khynmanwamungsa maisi khita.
Umung abongdarang chala,
dachangdarang chala.
Ytykyimungsa chalmanwa
machotwamungsa mai kaichenga.
Only after collecting the unburnt
remains of the jungle from the
land, we sow millet. Then we plant
maize and we plant dachang. Then,
only after we finish planting these
do we plant/sow rice.
kak n ART lid potolkak lid of a bottle
kak n ART lid
kak onom the sound of something
hitting or slapping
kak- v to bite Jemi sanchi
Dibangkhongdangaw matsa kakok.
On a certain day, a tiger bit
Dibangkongdang. Ang nangaw
kakai sani nowano pherue. I will
bite and eat you, he said, it is said,
the fox.
kak- v to close with a lid
kakdep- v to bite on something
kakhet adv all, very much, really
kakhirok n ANIM head lice, pubic
lice, crabs
kakmyn n BODY antenna (of insect),
feeler
kakpyret- v to crush by biting
kal n ART horn (traditional
instrument)
kaltek n AMIM type of big red ant
kalthek n ANIM big red ant
kala n PERS deaf person
kalai n ART loin cloth
kaljak ~ galjak n ANIM catfish
kalthek n ANIM type of big red ant
kaltyk n PERS person who never
washes
kam n ACT work, wealth Kam khana
harataidong angdo. Im reluctant
to work. Kam niwa. Worthless.
kam- v to clear the field, to cut the
jungle to make a field, to tear out
weeds
kam- v to suffer a penalty
Sa'khawgabaaw jurimana kamna
nangni. Thiefs have to suffer a
penalty.
kamal n PERS priest
kambai ~ khambai n PLACE top,
upstream
kan- v to wear Teew reenggaba
gawi, longpen kanai juta
hilaw(Wilseng) The girl who just
went by wearing trousers and shoes
with high heels
kan n BODY body (of human)
kan- v to last Bigaba nygylmi rawa
ie? Tyngen kanokte ido. Angba
ytykgaba botolaw ranichymte.
From which market did you buy
this? It lasts very long. I should buy
one such a bottle too.

658
kanjot- adj1 slim, skinny, thin (of
person)
kanpeng n BODY side of the body
kana n PERS blind person
kanggal n PERS poor person, pauper
kangguru n ANIM Kangaroo
kangkang n ANIM type of edible frog,
green with black spots, which lives
in caves and the hollows of stones
at the side of a river
kangkylek n ANIM type of lizard with
red neck, said to drink human
blood
kantara n PLACE emptiness
kanting- v to tear spontaneously
kap- v to catch, to close
kapangsi n ANIM a clamp
kapkap- v to lie flat on your belly
Ytykyimyng rukpeke rongphelang
sylgabachi kapkapai hynoknowa.
So then the frog presented himself
on a beautiful flat stone lying flat
on his belly, it is said.
kapkung n ANIM snail
kar- v to peel off Abong karai saa
angdo I eat the corn while peeling
off the seeds with my hand.
kara n ART rope, vein
karan n PLANT seed, kernel, fruit
stone
karang n BODY wing rikarang
testicle, balls, scrotum
karat ~ karat n ANIM squirrel
karaw n ACT debt, obligation, trouble
Nangdo uaw takchido karaw
mannine. If you do that, youll be
in trouble. Nangaw ang karaw
balni nang angmi bostu
sakhawchido. Ill tell you what
your debt will be if you steal my
things.
karen n ART electricity San thamok
karen niwa. It has been three days
[and/that] there is no electricity.
kata ~ khata ~ katha ~ khatha n
ACT word kata ~ khata ~ katha ~
khatha jywkhynwa to tell long epic
stories during the festival of
chywgyn, one story usually takes
one night or longer to tell. kata ~
khata ~ katha ~ khatha jyksai
coordinate compound that consists
of two synonyms
katha n ART shallow bamboo basket
katua ~ khatua n ANIM turtle,
tortoise
kaw n PLANT type of fruit
kaw- v to shoot
kawwarai n BODY gill
kawrawraw adv easily, without effort
Ama, angdo mai sanaka. Ym,
kawrawraw gawa, teen sabo.
Mom, I will eat rice. Yes, that
will be easy, eat later.
keret n BODY gall
kebyl n ART cable
keji clf kilogram, kg
kek n FOOD cake
kek- adj1 blunt (of pointed things)
kek- v to grow
kek- v to chop wood
keko n ANIM type of large brown
tokay gecko with narrow white
stripes on its back and white-and-
brown ringed tail and brown eyes
kel- v to hide Mykhang baketchi
kelaidong. Shes hiding her face in
a bucket.
kelki ~ khelki n ART window
kemyra n ART camera
kendyl n ART candle kendyl dot sa
one candle
kensi ~ kesi n ART scissors
kep clf classifier for small flat things
biskut kep sa one biscuit
kep n PLACE cave
kep- v to cry
kepleplep ~ kepreprep adv stretched
out on your belly Ytykyimyng
teedo amak gethengdo rong
pelang sylgabachi kepleplep bamai
hyntakkonoa. So then, now the
monkey, as for him, he willingly
lay down stretched out on his belly
on a flat stone, it is said.
kereng n BODY bone
keset ~ kheset n ART cassette, tape
kesi ~ kensi n ART scissors
ket- v to be tight Jama keta. The shirt
is tight.

659
kewal n ART a peddle
kha interj interjection to threaten
somebody and to warn that you
might fight, war cry, Beware!
Beware for X! This interjection
precedes a clan name and in some
cases the proper name of a
mythical person in a story. Kha
Marak! Beware Marak! This
interjection can also be used before
the clan name of the person who
says it as a way of self support, i.e.
Beware of me! Kha Bandi
Goira! noangthiriaidonga
Bandiba. Beware of Bandi and
the god of thunder! Bandi is
saying again, it is said.
kha- v to tie Nokbanthai dokhakhuchi
khachapai tangaba mongmawa
dora byryi dongabaaw raiai
jalangokno. They took the elephant
tusks weighing twenty kilos which
were kept tied to the dokhakhu of
the bachelors house and ran away,
it is said. Angawdo gorechi
chaaw nemen khabone nookno.
As for me, tie my legs well to the
horse, he said it is said.
kha n ABSTR fighting spirit
Gethengdo kharara taka, angba
kha ganang, gethengnado
kyrecha, takrukarini. He has
fighting spirit, but I also have
fighting spirit, I am not afraid of
him and will just fight with him.
kha- v to be bitter
kha- v to do, to work Kam khani. I
will work. Angna phone
khaetboto! Call me (on the
phone)!
kha- v to pour liquid into a jug
khaat- v to work with, to handle
Getheng koila khaata. He works
with/handles coal.
khadang- vdat to care for with great
love Ama thyiaimu akai sagyraina
khadanga. After the mother died,
her elder sister took care of the
child with great love.
khadi n ART clothes
khadong v to hope
khadong- v to be courageous, to be
hopeful Teewe ningan
kyryiphina. Alsia rajado
khadongaria. Thoroksyrangok
una, ningdo jalgabaak. Now it is us
who are afraid [of him]. The lazy
king is just courageous. He jumped
out [of the banyan tree] and so we
became the ones who ran away.
khagal- vdat to love Ang nangna
khagala. I love you.
khapak- vdat to miss Ang songna
khapaka. I miss my village.
khapet- v to be angry with Ang
bajuaw khapetaidong. I am angry
with my friend.
khaphak n BODY chest
kharan- ~ karan- v to be thirsty for
Tyiba karanok baisiga angdo.
Ang tyi chosa ryngna. I am thirsty
for water, my friend. I want to
drink a little water.
kharek n PLANT yard long bean,
yardlong bean, also known as the
long-podded cowpea, asparagus
bean, snake bean, or Chinese long
bean. The subspecies name is
sesquipedalis.
kharekrek- v to vomit, to barf
kharongthai n BODY chicken heart
khasi- v to not like and ignore
Gethengthengrara gorongrukokno
gorongaimyngdo tedo,
khasirukarokno. They met each
other and having met, they did not
like each other and ignored each
other, it is said.
khasin adj2 slowly
khasin-kadym adj2 khjys slow Uaw
badaiangwachian bean bebe
darairaragabasang dolong
khagabachina phetangoknowa,
Badido. Dykhimi balgabatykyi
khasin kadymai reengcha.
Jaljoljolangaidongano. When he
crosses beyond that point, truly
Bandi arrives at a bridge made
entirely out of swords, it is said. As
Dykhi had said, he does not go

660
slow. He is running quickly, it is
said.
khasop n BODY lung
khathol n ANIM wattle (of a chicken)
khathong n BODY heart O chame,
angmi nangna khagalgabaau
nangmi khathongchi dangetna
manphanima? (Sandish M Sangma) O
sweetheart, will you be able to
insert also my love for you into
your heart?
khawa n PERS lover
khawak khuwak adv with open
mouth Memangma morotma ie
sagyraido? noaimu khawak
khuwak chaisawthokaidongano.
Is that child a ghost or a man?
they said and all were surely
watching him with open mouth, it
is said.
khabak clf as much as the arms can
encompass, an armful
khabak- v to embrace, to grab firmly
as in an embrace Alsia rajado
matsami chaphungaw
wangjoloknoaro.
Khabakaimyng wangwachie
kynsange: Ai! ido alsiae kakate!
Sokchakate angdo noaimyng,
matsado jenethene jokaimyng
jalangoknoro. The lazy king bit the
tiger on the leg, it is said. After
having grabbed him firmly and
bitten him, the tiger said: Ouch!
This lazy person bites, Im telling
you! and having somehow
escaped, he ran away, it is said.
khachol n ANIM type of fish
khadok n ANIM type of fish
khagymyk n ANIM type of fish
khai- v to carry on the body
Ytykyimyng tedo gethengthengdo
na khynaimyng
baisigathangmaran rukpekba
tangsa, amakba tang
sakhaiaknowa. So then, now, as
for them, having collected the fish,
the frog and the monkey carried
one basket each, it is said.
khaithyi- v to hang oneself Ie nokchi
morot phalthangaw khaithyiwa. In
this house somebody has hanged
herself.
khakhudyl n PLANT type of plant
khal clf classifier for orifices, holes
and caves hangkhal khal ni two
caves nakhungkhal ni two nostrils
khal n PLACE hole khal sene seven
holes
-khal sfx intensifier suffix used in
comparative and superlative
constructions, more than, -er as in
bigger, larger and greener, most, -
est as in biggest, largest, greenest
Mamyngawan nangchawa raja
naa angna nangmyng gore jalna
rakkhalgabaaw hynetaribo
nookno. I dont need anything, o
king, you just give your fastest
running horse, he said, it is said.
Gore jalna rakbebeokno.
Khasinkhalai jalkhalna noaimyng
gadukdukchiba rakkhalai
rakkhalai jalariokno. The horse ran
really quickly, it is said. Having
told it to run slower, whenever he
prodded it with his legs, it just ran
faster and faster, it is said.
khalbong n PERS person who eats
scandalously much
khali adv only, exclusively
khaljong n ANIM type of fish
khalpak n ART belt that goes around
the head to carry a basket
khalput n PERS dirty person
khaltyi n SUBST soda
kham- v to burn Rangsan khama. The
sun burns/it is hot Wal nokaw
khamok. The fire burnt the house.
khambai ~ kambai n PLACE top,
upstream
khambykthai n PLANT type of edible
tuber
khamphung n PLANT type of edible
tuber
khampyryw- v to have a hole in a
cloth or paper as the result of
burning

661
khamthymbylong- v to have a hole in
a road or bridge as the result of
burning Dolong walsang
khamthymbylongok. The fire
burned a hole in the bridge. / The
bridge was damaged by the fire.
khamynkhap n ANIM type of fish
khan clf classifier for objects like log
boats rung khan ni two boats
khan n PLANT cassava
khan- v to suckle
khan- v to slaughter, to chop, to
mince, to cut
khanchot- v to cut Nang ang khaw
khanchotbo. Cut my hair.
khanperet- v to split, to cut open
khana n PLACE port, harbour, station
khanchot- v to cut (hair)
khang- v to solidify
khang- v to occupy Baidamdo
hawangman/gaba hagun
saangmangaba
harynthangthangaw khanga.
Some people occupy their own
parcel which is already cleared and
used up completely.
khanmychyw n PLANT cassava
khanmynchyw n PLANT type of
edible shrub
khanphyt- v to cut a solid object in
half lengthwise
khanpyrak- v to cut a hollow object
in half lengthwise
khansynen n PLANT type of edible
tuber red on the outside and white
on the inside
khansyrui n ANIM earthworm
khanta n QUANT hour
khanthong- v to cut in half
khantongthong- v to cut up in pieces
khap clf classifier for flat materials
tota khap sa one plank tin kahp sa
one sheet of corrugated iron
damdyl khap sa one damdyl
khap clf classifier for flat pieces of
hard material like stone or metal
sorekhap khap ni two pieces of
mica
khap n ART cup, teacup or its volume,
cupful khap goi sa one teacup
khap thai sa one teacup Cha khap
tham hynbo. Give three cups of
tea.
khap n SUBST flat piece of hard
material
khap- v to be cooked without maityi
~ maiti jabek khapgaba curry
without maityi ~ maiti
khapeng- v to hinder
kharok n ANIM type of very small
fish
kharongthai n BODY kidney
Khasi n PERS Khasi
khasi- v to castrate, to remove the
testicles
khasin adj2 slow
khasot clf classifier for bundles
rasun khasot sa a bundle of onions
khat interj interjection to chase a dog
away
khat- v to slaughter Biana wak khatna
rawaidonga. We are catching a
pig to slaughter for the wedding.
khata ~ khatha ~ kata ~ katha n
ACT word kata ~ khata ~ katha ~
khatha jywkhynwa to tell long epic
stories during the festival of
chywgyn, one story usually takes
one night or longer to tell. kata ~
khata ~ katha ~ khatha jyksai
coordinate compound that consists
of two synonyms
khatdep- v to wrap, to wrap up, to
fold
khatom clf classifier for bagsful
rasunok khatom sa one bagful of
spring onions
khatua ~ katua n ANIM turtle,
tortoise
khaw clf classifier for teeth wa khaw
sa one tooth
khaw n BODY hair (of the head)
-khaw evsp V secretly, V
surreptitiously
khaw clf classifier for teeth, planks,
sheets of corrugated iron for roofs
and flattened bamboos used to
make mats (jyw) when they are in
a mat damdyl khaw sa one jyw of
a damdyl wa khaw ni two teeth,

662
two tusks (of elephant) tota khaw
tham tree planks, tin khaw byryi
four sheets of corrugated iron
khaw- v to catch water in the palms
of your hands Paipmi tyi getheng
khawaimu ryngok. He caught the
water from the pipe in his hands
and drank it.
khawakwak- v to vomit, to barf
khawcharyng n BODY sideburn
khawchi n BODY grey hair
khawchyryng n BODY scalpel hair
khawdam- v to put down
khawkhai ~ khawkhi n BODY grey
hair
khawkham n ART pillow
khawkhirok n BODY dandruff
khawkhuthuk n ART cloth for men
worn around the head
khawphyng n ART turban
khawsuk n PLACE source of a river
khe- v to be proper, suitable Nang
mykchagaba biphae khecha. The
boy you fancy is not suitable (to
marry).
khel n ACT care
khele- v to play
khelegaba n ACT game khelegaba
myng sa one game
-khelek evsp V for fun
khelhi ~ kelki n ART window
khem n ART drum (traditional
instrument)
khema n ACT forgiveness Nangtym
angaw wetsado khema khakhubo.
Please forgive me one more time.
khen- v to scratch Machok kan
panchi khenaronga. A deer is
scratching his body against a tree.
khen n ANIM river crab
khenkhorong n BODY claws of a
crab
kheng- v to be alive
khengchek adj2 green, blue
khengkhang adj2 eternal
khengsyryk adj2 dark green
khengwa n ACT life
-khep evsp V firmly Rawkhepbo!
Hold it firmly!
khep- v to cry
khep- v to pinch, to cut with scissors
Rongkhalchi khonokaimu khen
chak khepok. When I felt under the
stone, a river crab pinched my
hand.
khep ~ khup ~ khyp- - v to close, to
cover, to spread out, to put on
clothes
khereng- v to struggle, to make a
great effort
khereng- v to resist
khewal n ART oar khewal phong sa
one oar
khi- v to count
khi- v to hit (a target), to touch Myng
sa them! Khawoknotyi.
Khianchano. One person shot,
pow! He did not hit it, it is said.
khiil n ART nail (iron) khiil chong sa
one iron nail.
khingcheng adj2 aslant, slant
khirip ~ kyiryp n PLANT type of
edible plant of which the leaves are
mashed and dried and then cooked
to pulp
khit- v to sprinkle, to sow seeds
Hakhynmanwamungsa maisi
khita. Only after collecting the
unburnt jungle material from the
land, we sow millet.
khok- v to remove (skin, bark, peel,
dress etc.)
khokalang n PERS bold person
khol n BODY skin (of human, animal
or plant), hide (of animal), scale (of
fish) masukhol cow-hide, cow-
skin
khol num twenty
kholchang num twenty, used only in
compound numerals kholchang
byryi rong ni eighty two.
khole num twenty. This word is only
used in compound numerals khole
rong sa twenty one.
khole chyi ~ kola chi num thirty
kholgyk ~ kholgryk num twenty. The
variant kholgryk is a loan from
Garo but is used overwhelmingly
more frequently than kholgyk.

663
kholjisop n BODY infection of the
inner ear, labyrinthitis
kholthyrai- v to shed skin, to come
off (of skin) Rangsan khamaimu
ang nakhung kholthyraiok. After
the sun burnt it, the skin on my
nose came off.
khom- v to sit with your head in your
lap and your legs pulled up
khon encl.cl/prtcl speculative
modality enclitic or particle
Rangsan rangbyrymaidonga,
wainikhon. The sun is blocked by
clouds, it might rain. Nang ama
nygylsang reengwama?
Hoong, khon. Did your mother
go to the market? Yes, maybe.
khong- v to bark
khonok- v to search by feeling Ichi
rongkhalchi khen ganangthel
mang sa mangnido,
ganangthelnaba ganang. Hai,
nang usang khonokbo ang isang
khonoknaka. Here in these holes
under the stones there are crabs for
sure. Lets go, you feel and search
over there and I will feel and
search over here.
khopalak n PLANT skin of onion,
garlic, corn etc.
khophylak n PLANT/ANIM skin of
fruit, eggshell
khophynga n ART cloth for women
worn on the head with a knot at the
back of the head
khorat n ART a saw
khori n ART watch
khorong n BODY horn (of animal)
khoryndachong n ANIM silkworm
-khu sfx incompletive aspect suffix
khu hamgaba n ART grammar
khubisi- v to hate, to dislike
khucheng- v to bite your teeth firmly
together Ytykyimyng pherudo
rypangthiriokno. Phalthang sokwa
dabatdo tyinyngchi rongchi
pyiaimyng wa khuchengphinai
sakchikaidokno. So then the fox
soaked in the water again, it is said.
Until he could not hold out any
longer, he sat under water as long
as he could bear it, holding on to a
stone and biting his teeth firmly
together, it is said.
khuchi- v to dislike Mongma ranai
sana ang khuchia. I dont like to
eat elephant meat.
khuchuk n ACT mouth, language
khuchul n BODY lip
khuma n PERS dumb person,
someone who cannot speak
khumong- v to conspire Ramchi
hampyi nanangdo watchaka
gethengawdo, sala! Gethengaw
watkhuna sootthelarinaka noai
khumongangokno. This evening
we will seize him on the road, the
bastard! We will kill him after all
to banish him once more, they
conspired freely, it is said.
khurang n BODY voice
khurasak- v to promise. Rongdyng
maharimu Jaksongram
matsanokphandaimi matsamu
takrukaisa Rongdyng
maharidyrange dokra Habeng
khungsa ripan patsai sympak
khungsa hynnaka noyi
khurasakokno. Because the
Rongdyng family fought with the
tigers from the bachelors house of
Jaksongram,the Rongdyng families
promised to give one Habeng bag,
one womans dress and one
sleeping mat, it is said.
khusak- v to answer, to reply, to
respond
khusum ~ kusum n ANIM tortoise
khusymang n BODY facial hair
khuthi khuthyraiga(ba) n
mumbling
khuti khuthyraiga(ba) n ACT
mumbling
khutip- v to close your mouth
Hongkhotphinna manchaaiyng
khuthipwachie sotmai mang sene
mansymokno. Not being able to
come out again, when he closed his
mouth, he swallowed the seven
flies, it is said.

664
khutyisot- v to spit
khuchia n ANIM type of fish
khuchina n ANIM eel
khudal n ART hoe, chopper
khugyri n ART small basket made of
bamboo and reed
khul n ART pillow stuffing
khuli n FOOD opium
khuli- v to open
khung clf classifier for flat things (and
photos even when displayed on a
computer screen) tangka khung sa
one banknote piktiyr khung sa one
photograph, picture
khung n BODY shell of a crab,
tortoise etc., carapace
khup- ~ khep- ~ khyp- v to close, to
cover, to put on clothes
khuru clf length from the top of the
thumb to the top of the middle
finger when one puts ones hand
down on the table on these points
khurung adj2 wanting to lay an egg
khurut- v to perform an incantation,
to summon a spirit
khusep n BODY corner of the mouth
khusi dong- ~ dong- v to be happy
Gethengtheng khusi
dongthamakaimyng gore
dimaichi phalthang cak diriga
sangwalaimyng watokno. They
were so excessively happy that
they forgot their own hands which
were holding the horses tail and
they let go, it is said.
khutai n top (of a house)
khuthym- ~ kutum v to kiss
khutyi ~ khuti n BODY spittle, saliva,
spit
khyi- adj1 sour
khym- v to marry Ue alsia rajae jykba
myng ni khymanoro. That lazy
king married two wives, it is said.
khymgaba n PERS spouse
khyn- v to pick up, to gather, to
collect, Tyipaichi sukyrung khynok
ningdo. We have gathered river
snails in the Tyipai river. Sawaisa
hakhamchido khynna nangcha.
Hakhamchachido hakhynchenga,
hakhynmanwamungsa maisi khita.
After burning the jungle, if the land
is completely burnt, it is not
necessary to collect the remaining
unburnt jungle material. If the land
is not burnt, we collect the unburnt
jungle material from the land first,
only after collecting it, we sow
millet.
-khyngkhyng adv still, continuing Ue
habyriawe teewchina khyngkhyng
Atong khuchuksang Matsa
Chawkyi Asetram myngwano. We
call that mountain up till now still
Matsa Chawkyi Asetram in the
Atong language, it is said. Angdo
teewrawrawdo isykynmi dukna
khyngkhyngdo nemai
muphaaidonga. I am still well
despite my continuing suffering.
khyp- v to take a bite
khyp- ~ khup- ~ khep- v to close, to
cover, to put on clothes
khyryithang ~ khyrythang n KIN
nephew
khyryk n ANIM louse (plural: lice)
khyw- v to drain, to shake out fluid
kilip ~ kylip n ART clip
kilomytyr clf kilometre
kingreng kingcheng adv like a
chicken without a head, like crazy
Thot thyngthot takwachina dabat
sykromaimyng
khanetsigaaidongno. Bandi
chakwatwamian chuwil chuwal
takjolangokno. Taw tokai
watetwatykyi usang kingreng usang
kingcheng takjoletarioknoti. He
(Bandi) grasped her (Sore) and
poured the liquor into her mouth to
the last drop. When Bandi let go of
her, her head was spinning. Like
letting go of a beheaded chicken
she ran around like a chicken
without a head.
kirin adj2 to be torn (of cloth and
paper)
kirin- v to tear Longpen kirinok. The
trousers are torn.
kitap n ART book

665
klas n PLACE class Ando klas wan
mangmangsaan
dongphaarikhuwa. I have only
gotten as far as class one.
korot n PLANT sugarcane
kobi n PLANT cabbage
koila n SUBST coal
koilagari n ART coal truck
kok n ART basket
kokalang n PERS a bold person
kokbal n ART biggest basket used to
store food in
kokbal n ART enormous basket made
of bamboo and used to store rice
and vegetables in the kitchen.
kokcheng n ART type of basket made
of bamboo carried on the back with
a strap around the head, smaller
than a koktang.
koke n ANIM type of large gecko that
lives in trees
kokpylak n PLANT chaff
koksep n big woven bamboo cage
used to keep chicken in when they
are sold on the market
koktang n ART big basket made of
bamboo used to carry goods and
worn on the back with a strap from
the head
kol n ACT telephone call
Kol India n PLACE Coal India
kolachi ~ kolechyi num thirty
kolachita n PLANT Momordica
charantia, bitter gourd, bitter
lemon
kolani n PLACE colony
kom- v to feel like a loser Angdo
barataimyng komok. Me, being
ashamed, I felt like a loser.
kombol n ART blanket
komok- v to feel insulted
kompiutyr n ART computer
komputer n ART computer
kompyl ~ kongpyl adj2 to be bent
kun kompyl a bent stick/the stick is
bent
komyla n PLANT orange
kon adj1 winding
kong- v to have rabies
kongkenaken adv zigzag, winding
kongpyl ~ kongpyl adj2 to be bent
kun kompyl a bent stick/the stick is
bent
kongtoksi n PERS used in the
expression halsia kongtoksi lazy
person (pejorative)
kontrektyr n PERS contractor
kopi n FOOD coffee
kopja n ART a hinge
koplak n BODY egg shell
korea n ART big metal pan
korok num six
korong ~ kyrong n BODY horn (of an
animal)
koros n ACT expenses
korot n ART a saw
kri postp in accordance with,
according to Gam manni udo uan,
tangka poisa. Uan gam mynga,
dakangmi chasongdo. Teewsa
kepasyti noai myngaidonga.
Chasongna kri gam myngariaro,
tangka poisa. He will obtain
wealth, money. Earlier generations
called that wealth. Now they call
it capacity. According to my
generation this money is called
wealth.
krismas n TIME Christmas
kristan ~ kristen n PERS Christian
krymkraw adv together, in unison, in
co-operation
kusum ~ khusum n ANIM tortoise
kukuri n ART type of knife with a
blade with an obtuse angle used to
survive in the jungle
kulal ~ kular ~ kural n ART axe
kuli- v to open
kun- adj1 curly
kun clf classifier for stick-like things
nokwek kun banga five brooms
kun n ART a stick
kural ~ kular ~ kulal ~ kurar n ART
axe
kurasak- v to promise
kusymang n BODY facial hair, beard,
moustache
kutuuukutukutuk interj interjection
to call a dog
kyi n ANIM dog

666
kyiwa n BODY canine teeth
kyimang n ANIM fruit fly
kyiryp ~ khirip n PLANT type of
edible plant of which the leaves are
mashed and dried and then cooked
to pulp
kykgul n BODY eyelid
kyl- v to hide, to avoid
kylchap n PLANT cotton
kylip n ART a clip
kylip ~ kilip n ART clip
kyltuk n PLANT type of tree which, in
the dry season, brings forth
bunches of bright red and yellow
beans with black seeds. The seeds,
when eaten, make you sleepy.
kymkha n PLANT type of berry
kyn n BODY back Ang nangaw kyn
kyn symni. I will follow you
closely.
kyndam n PLACE land behind a
village
kyngjung- v to turn your back to
someone Angsang kynjungbo. Turn
your back to me.
kynkyreng n BODY spine
kynokhol n KIN son-in-law
kynpha- ~ kympha- v to be last, later
Ie sagyraido kynphaai
phetdangok. This child arrived last
kynphak- v to sleep in. Rang
wawamigymyn manaphci
kynphakwa. Because of the rain I
slept in this morning.
kynsang n PLACE behind nokapmyng
kynsang behind the door
kynsang postp Ytykyimyng walchi
raiaphyinokno. Raiaphinaisa
beanbebe phalthangmyng nokaw
gethengdo masu dimyng
phirinaimyng ue sona
bichamchymaw nok ryphiokno.
Nok ryphiwamyng kynsangdo teew
getheng nokawan alaga
morotdyrangdo tengchypchypai
nukariokno. So then he came back
at night, it is said. Having come
back, having mixed it with cow
dung, he plastered his house with
the golden flakes, it is said. After
plastering his house now, other
people saw how his house was
shiny, it is said. Rang
nemchengama nanang chyw
jamchenga noai rangmu chyw
ryngsusaai chyichie, range san
chibri wawano. Umi kynsang san
khole san sa chyw ryngkhuanowa.
Will the rain stop first or will we
finish our liquor first? they said
and while competing in drinking
with the rain, the rain fell for
fourteen days, it is said. Thereafter
they still drank liquor for twenty
one days, it is said.
Phetangaidokno, dongaidoknotyi,
sane san chidok wal chidok
reengwami kynsangsa. He has
arrived, it is said, the had reached
his destination, to our surprise, it is
said, after going for sixteen days
and sixteen nights.
kynsang tw PLACE later afterwards,
Kynsangdo jywchangna
nangokno. Uchie jywchangna
nangwachie matsa
gorongtatoknotyi, maikapchi
jywwachi. Kynsangdo matsado
morotsyn manaimyng raiwilokno
alsia rajado. Later he had to stay
the night somewhere, it is said.
Then, when he needed to stay the
night, he met a tiger when he was
sleeping on the hay, it is said.
Later, after the tiger had smelled
the sent of a human, it walked in
circles, it is said, around the lazy
king.
kyp- v to fit tightly, to fit and close off
kyrewami ~ kyriwami n ABSTR
danger
kyrong ~ korong n BODY horn (of an
animal)
kyrydyl n PLANT type of liana that
looks like a long arm with elbow
joints as it hangs between trees
kyryi n ACT fear
kyryi- vdat to be afraid of Ang
mongmana kyryia. I am afraid of
elephants.

667
kyryiwa n ACT fear Bil! Bil! Ningdo
nangaw phina raiwachym!
nookno. Uchie Bile: Atakna?
nookno. Ytykyimyng phina
reenggabadyrangba kyryiwa
ganang. Bil! Bil! We have come
to invite you! they said, it is said.
Then Bil said: Why?, it is said.
So then in turn, the ones that came
to ask were afraid.
kyryk- v to be clear, transparent
-kyrym evsp to V in a group Taw
dangkyrymangok, tawnokchi. The
chickens have all entered their
coop.
kyrymkyraw adv united, together, in
cooperation
kyryng- adj1 tight Usang, songga
Manggagremi banthaidarangba
raiaaithokaidongano,
Rakarelwakmadare,
Aragundi, Motbanda,
Asyngduraparaba gumukan
raiathokaidongano. Chakphong
gylgabasano, kara
khyrynggabararasano,
alamylachagabasano. They are all
coming to there, the young man
from the strange village of
Manggare: Rakarelwakmadare,
Aragundi, Motbanda,
Asyngduraparaba, they are all
coming, it is said. They are men
with strong arms and tight veins all
over, it is said, they are not
ordinary men, it is said.
kyryng- v to make noise, to make a
sound
kyrynggaba n ABSTR sound, noised
kyryngwa n ACT sound, noise
kyryw n ART thin strip of bamboo
used to make rope, bamboo rope
kyw interj I am here! (answer to a
search call)
laha n SUBST resin
lain clf classifier for a collection of
objects lined up on shelves
laisak n PLANT cabbage
laisen n ART licence
lait n GEO/ART light
laklak- v to prod in an orifice or hole
for pleasure, to nag
-langlang evsp very
lap n ACT profit, interest, gain, value
Hagylsakaw chol takai chyichiba
nangaw khymaido angdo mamyng
lap niokte. When you try all sorts
of small jobs, it will not benefit me
to be married to you.
lap- v to gain, tomake profit, to be
profitable Ang taisa rajasa
lapokchym, thyiok. I just made a
hundred rupees profit but I lost the
game. Atakgaba raja naa angna
gore lapchagabaaw watetwa
nookno. What kind of king are
you to send me a worthless
horse?, he said, it is said.
lapan n PLANT pan/paan leaf
lapchagaba n PERS a good-for-
nothing
lapchagaba adj good-for-nothing,
unprofitable, worthless
las n ABSTR the last one
lathia n ANIM type of fish
law n PLANT cucumber-like vegetable
lechu n PLANT lychee
lekadawreng n ART kite
lekat- v to waste time Naa mai
sywkhalna balwachymte. Naa
teewchinaba lekataidongkhua.
You said that you would pound
some more rice. Until now you
have been waiting time and you are
still wasting time.
lekha n ART book, paper
leklek- v to prod in an orifice or hole
lengla adv crippled
lepstik n ART lipstick
letrin n PLACE toilet
lityr clf litre
lolal n ART roller
longpen n ART a pair of pants,
trousers, long pants, long trousers
Teew reenggaba gawi, longpen
kanai juta hilaw The girl who
just went by wearing trousers and
shoes with high heels longpen

668
khung sa a pair of pants, trousers,
long pants, long trousers
loskor n PERS highest rank in the
system of customary law of Garos,
judge
lukchok ~ rukchok n ANIM type of
frog
lukchokchok n ANIM type of small
gecko that creeps up the walls of
houses at night
lukpekpek ~ rukpek n ANIM type of
frog
lukwak ~ rukwak n ANIM toad
mm ~ hmm procl no
ma encl.cl/prtcl question enclitic or
particle
ma- v to lose Sunglas palyngchi maak
angdo. I lost my sunglasses in the
jungle.
ma inter o.k., o.k. then, (very) well
then (Nepale) Ytykchido ang
reengsigama nangmi phal?
(Thengthone) Ma ytykchido
dengbo nooknoro. (The Nepali)
But shall I go instead of you?
(Thengthon) Very well then, in
that case, it will be most
convenient, untie my hands, he
said, it is said.
maam- v to moan
machong n PERS clan
machot- v to be an orphan
masu n ANIM cow
masubolot n ANIM ox
masugari n ART bullock cart
machak- v to be vengeful Ang
nangaw machakaidonga I am
vengeful towards you.
machirit- v to scratch, to be scratched
machok n ANIM type of large deer
machot- vphase to finish, to end, to
stop Mongmaaw chaina sawa
machotok. I stopped eating to look
at the elephant.
madam n PERS female teacher
magachak n ANIM type of small deer
magal n ANIM type of fish
magana- v to lose Amakan
raiatkaimuna: Mykhangba syma
dymbrubru. Dimaiba rawa
dymbrubru noaimuna
balatakokno. Rasong
manatakarokno haiaw
dawkhaaw, amakansega. Uchido:
Hoong, hoong, aak! aak!
noatakarongnote. Uchie byk
galatokno rutido. Uawdosega
pherusa kakkhypaimu saakno.
Ytykyimuna kynsangdo dawkhado
maganaak, sana mananchakno.
The monkey cam: Your face is
black and shiny and you have a
black and shiny tail. he said, it is
said. The crow became proud
because of the monkey, it is said.
Then the crow said: Yes, yes!
Caw! Caw!, it is said. Then the
bread fell down, it is said. The fox
then grabbed it with its mouth and
ate it, it is said. And so the crow
had lost its bread and could not eat
it any more, it is said.
magyna adv in vain
mahari n PERS relatives, family
mai n FOOD rice (cooked)
mai ~ mei n TIME May
maicheng n PLANT edible shrub
with scented (nice smelling) leaves
and soft thorns, (maybe related to
Zanthoxylum oxyphyllum)
maityi ~ maiti n FOOD juice from
jabek
maiwek n ANIM type of bird that is
believed to call every time
somebody comes to visit the
village
maichek~ maichyk n FOOD cold rice
maidan n PLANT new rice (just
harvested) Mai mynokodo maidan
syla toka. When the rice is ripe, we
maigasam n FOOD meal eaten in the
later part of the day or evening,
dinner
maijyk ~ maimijyk n ANIM
dragonfly
maijyreng n FOOD dried rice for the
pigs

669
maijyreng n FOOD leftovers of
cooked rice dried in the sun used to
feed the pigs
maikap n PLANT hay
maikholnang n PLANT unpeeled rice
maikhyt n FOOD burned rice
maikung n ACT second rice harvest
(in November)
maimanap n FOOD meal eaten in the
morning, breakfast
maimijyk ~ maijyk n ANIM
dragonfly
mainym clf length from the elbow to
the top of the fist
maip ~ mep n ART map
mairong n PLANT husked rice,
uncooked rice
mairugu ~ meringgu ~ meringgaw ~
merenggaw n PLANT mushroom
(edible)
maisan n FOOD meal eaten in the
middle of the day, lunch
maisen n FOOD sticky rice in a
banana leaf
maisi n PLANT millet
maityk n ART pot for cooking rice
maiwa ~ maiwa n PLANT bamboo
shoot maiwa ~ maiwa ching sa
one bamboo shoot
maja tw the day before yesterday,
some time ago, in the recent past
makbul n ANIM bear
mal- v to be familiar, easy to deal with
mama n KIN mothers brother. Is also
used to address an unrelated man
older than the speaker in a
respectful way, and to address my
father-in-law.
mamylet n FOOD omelette
man n ACT respect Gethengaw man
rana nanga. You have to respect
him.
man- v to crawl, to creep Khyryk
khawchi manaidonga. Lice are
crawling in my hair. Atongbatykyi
gasokok aksokok
hunthamakaimuna saphawba
hakhungchina mangatokno.
Somehow the rabbit, stumbling and
barely swimming, crawled onto the
river bank, it is said.
-man ~ -man evsp already V-ed
man- vB to be able, to get, to obtain,
to succeed manai sa- to eat in
great amounts, to be rich Sagyrai
raina mana. The child can walk
manai in great amounts Morot
bilding chunggabaci mugaba
manai saak. The man who lives
in that big house is rich. Uchian
magachakdo biskutaw mana
mana mana mana mana
saaidokno. Then the deer ate the
biscuits in great amounts, it is said.
Ang baju kam manok My friend
got work/a job. Kawbutungchi thik
thokyrengaw manokno. When he
shot [the giant eagle] he got it
exactly in the neck.
manai adv in great amounts, see
man-
mandapwami n ACT profit, interest,
gain
mandyk- adj1 difficult, complicated,
troublesome Angdo
gylgylrongchawanasa teew nokchi
rang waaimu kam khana haratok,
mandykok kam khana. As for me,
precisely because I usually roam
around now that it is raining, Ive
become reluctant to work, its very
difficult to work. Alternatively: It
has become difficult to work.
mansymrukruk- v to inherit
manak- adj1 dark
manam n ACT bad smell, stench
manam- adj1 to stink Di manama.
Shit stinks.
manap n TIME morning
manap- v to be morning Manapok.
It has become morning.
Manapnaka. It will soon be
morning.
manapmi adv very early in the
morning Kynsangdo manapmi
sirimynmyn reengaimungna
Dabatwarisang dinggarai saiakno.
Then, having gone to Dabatwari
very early in the morning at the

670
break of dawn, [he] put up his fish
traps, it is said.
mandal n PLANT erithrina
superosastricta, type of tree with
thorns and very red flowers which
blossom in the late part of the dry
season
mang- clf classifier for animals,
knives and tools bythyi mang sa
one porcupine chawkyi ~
changkui mang sa one big knife
manggywak n ANIM millipede
mangka n ANIM type of fish
mangkung n ANIM type of cricket
mangkyrang ~ mankyrang n ANIM
scorpion
mangmang adv only, just, exclusively
-mangmang evsp only
mangneng- v to whine Ytykyimu
isangdo jykdo sado, sado jongdo
mai okhiaimu mangnengaidokno
mangnengaidokno. Babado biba
rainaka?" noaidokno.So then at
home, because they got hungry, his
wife and children are whining and
whining, it is said. When is dad
coming back? they are saying, it is
said.
mangsong- v to plan
mani n KIN aunt: fathers sister. Is
also used to address your mother-
in-law.
mani- v to worship Ning songsyrekdo
ning atongdo dakangdo mamyng
thoromaw niwami somaichido
waiaw mania. We heathens, we the
Atong, in the past, in times when
there was no religion, we
worshipped spirits.
manjuri ~ manjyri n ART post,
supporting post
mankyrang ~ mangkyrang n ANIM
scorpion
mansylang n BODY spleen
mantaw n PLANT type of gourd
mantawbylati n PLANT tomato
mantawthai n PLANT type of
vegetable
mars n TIME March
mastel n PERS male teacher
mat n ANIM wild animal
mat n BODY a wound
mat- v to be sharp, to be wounded, to
(be able to) wound, to (be able to)
cut
matdi n ANIM wild water buffalo
mathai ~ mathai n ANIM bachelor
elephant
matji ~ maji n PLACE middle Ang
Sandishmyng Bittermyng matjichi
muaidonga. Im sitting in between
Sandish and Bitter.
Ang gumukmyng matjichi muaidonga.
Im sitting in the middle of
everybody.
matpalyng n ANIM wild animal
matrong n ANIM jungle goat
matsa n ANIM tiger
matsadu n PERS/ANIM creature
which is human during the day and
becomes a tiger at night
mausa n PERS friendly name to call a
person
mawkhol n PLANT bark (of a tree)
mawsa ~ mosa n KIN 1. marriageable
male cousin: the child of mama
mothers brother and anay
fathers sister or of awa fathers
younger brother and asoy mothers
younger sister, 2. the relation of
male cousins from
intermarriageable families, 3. a
male friend belonging to an
intermarriageable family
meama n PERS married woman
meapha n PERS married man
mema kyi ~ mimangkyi n ANIM
type of small frog that says
pekpekpekpek
memang ~ mimang n PERS ghost,
spirit of a dead person memang
sawet- ceremony performed a year
after someones death. The spirit of
the dead person then leaves the
house and goes to Balphakram.
memangguchung n PLANT type of
liana, woody vine that grows in the
jungle as winding branch with an
undulating pattern twirling itself
around other trees for support. The

671
name of this plant translates as
ghost ladder
memangkereng n ANIM stick insect,
walking stick, phasmadotea
memangkoksi ~ mimangkoksi n
PLANT pitcher plant
memesi n ANIM flying insect
mei ~ mai n TIME May
mejakbal n ANIM alligator
Mekalaia n PLACE Meghalaya
mel- adj1 fat (of person)
melanggaw n ANIM poisonous red or
black ant
memaboro n PLANT type of nice
smelling rice
memaboro n PLANT nicely smelling
type of rice
menpart n PERS most important or
most salient person
mep ~ maip n ART map
meringgu ~ meringgaw ~
merenggaw ~ mairugu n PLANT
mushroom (edible)
mes n ANIM sheep
-mi ~ -myng encl.phr genitive enclitic
mimang ~ memang n PERS ghost,
spirit of a dead person memang
sawet- ceremony performed a year
after someones death. The spirit of
the dead person then leaves the
house and goes to Balphakram.
mimangkoksi ~ memangkoksi n
PLANT pitcher plant
mimangkyi ~ mema kyi n ANIM
type of small frog that says
pekpekpekpek
mili- v to assemble, to meet, to come
together, to be appropriate
milimityr clf millimetre
mimi- v to laugh at someone Ie biphae
gawigumukaw mimia. This guy is
laughing at all the girls.
miniksuru- v to be flat-haired (of
animals) Magachakmi myndo
tyisiwachian miniksuru
takjolarianoro. When the deers fur
is wet, it just quickly gets flat-
haired, it is said.
mirang n ANIM neck feathers of
chicken
mistyri n PERS mason, house builder
and painter
miting v to hold a meeting
Baisigathangmaran myngtham
mitingaidoknowa. The friends are
holding a meeting, it is said.
Dakanggabado jineral
mitingchengni. Umungsa song
gumuk thomaimung haba haryn
harynaw sowalni. First they will
start with a general meeting. Then
the whole village comes together
and they will divide the haba plot
by plot.
mityr clf metre
mmmm mmmm onom the sound of
an eagle
mo encl.cl/prtcl confirmative tag
enclitic or particle. Tyt!
diphuramama? nookno.
hoong manamaidongmo Hey!
did you just accidentally fart? he,
said, it is said. Yes, it stinks,
doesnt it. Tanmanokona thyiok
udo, mo. Because they had cut him
up, he died, that one, isnt.
mobail n ART mobile phone
mobil n SUBST motor oil, engine oil
mochok n PLANT sapling
moila n SUBST dirt, filth
moina n ANIM type of bird
mojekjek- v to shake (a fixed object)
mojet- ~ mojot- v to suck
mon clf unit of 40kg mon sa one unit
of 40kg
monggolbal n TIME Tuesday
mongma n ANIM elephant
mongmachong n ANIM caterpillar
mongmachongsu n ANIM giant
caterpillar
mongnal n PLANT lotus
mongyreng n ART knife
monok- v to swallow, to devour
Goilapan chymaimu monokbo.
After chewing the betel nut and
paan, swallow it. Sagyrai
mylgabami dadadarangawdo
janggalawan monokokno. Phylgym
chunggaba monokrumokno myng
korokawan. As for the brothers of

672
the small child, they were
devoured, it is said. The big eagle
had devoured them all, the six of
them.
montyri n PERS minister
morot n PERS person, human, human
being, man
morot- v to grate
mosa ~ mawsa n KIN 1. marriageable
male cousin: the child of mama
mothers brother and anai
fathers sister or of awa fathers
younger brother and asyi mothers
younger sister, 2. the relation of
male cousins from
intermarriageable families, 3. a
male friend belonging to an
intermarriageable family
mot- v to shake a fixed object
-mu evsp durative: to keep V-ing,
mu- v to stay, to sit (be in sitting
position), to sit down, to be at, to
live somewhere Phylgymdo
nukanchano. Atongba saai
muarongno. The giant eagle did
not see [him]. He was sitting and
eating something. Dakangdo Dawa
maharisa ichi muwanokhone.
Perhaps in the past the Dawa
family lived here, it is said.
Chokichi mubo. Sit down on the
chair. Dakanggaba Turachi
The first time I stayed in Tura I met
Mobbin.
muchonchyro- v to squat Gari
hapalchi muchonchyroai chaibo.
Look under the car squatting.
mudap- v to sit on
mukhuchok n BODY nipple
mupeng- v to prevent
mupyret- v to crush by sitting on
something
murong- v to shit on
muten- v to look after, to watch, to
keep company
muthai n BODY breast (of woman),
bosom muthai hal- to breastfeed
muchi n ANIM type of fish
muchot n ANIM mouse, rat
Abeknyngchi muchot sagyrai
mang byryi chepchap chepchap
parawthokaidonga. Inside the abek
are four baby mice squeaking eek
eek.
mudu n PLANT papaya
muja n ART sock
muk- v to smoke
mukthai n ABSTR asperity,
protrusion
mula n PLANT white radish
muluwa n ART type of bamboo
mungmawa n BODY elephant tusk
muni n ACT magic spell Ma
pynwasama muni ma ang
mykrenaw, ma nangchi ganang
atongba jadu. (Wylseng S Marak)
Whether my eyes are covered by a
magic spell, you have something
magical.
mura n ART stool (to sit on)
muri n FOOD popped rice
musuri n ART mosquito net
mychym- v to smile at someone Gawi
angaw mychymaidok. A girl is
smiling at me.
myia tw yesterday
myk clf the length from the elbow to
the top of the middle finger
myk- v to tell lies
mykbu- v to be jealous
mykbyryw- v to have itchy eyes
mykcha- v to like somebody
mykchagaba n PERS sweetheart, girl
or boy that you fancy
mykchel- v to shine in the eyes
mykchep- v to look down upon, to
despise, to scorn Atakna nangdo
angaw mykchepa? Why do you
despise me?
mykep n BODY temple
mykgythal n ABSTR reality
Mykgythaldo dongcha jywmangsa.
Its not reality, its just a dream.
mykha badri n GEO long period of
incessant heavy rainfall Ue hapchi
muwachi rangmu chyw
ryngsusawanasa Hachyk
khuchuksang mykha badri noyi

673
myngwano.
Teechinakhyngkhyngba Hachyk
songgumukdo mykha badri noaria
je rangawba. Ytykchiba ie Badri
joldo asingkatiba septembyr
oktobolmi rangawsa mikha
badriwa asing badri kati marsatwa
noyi mynga. Ramram rangawdo
mykha badri myngcha.
Umigymynsa ie hapawe Badri
Rongdyng Hawai noanowa, aro
rangawba mykha badri
myngwanowa. When they stayed in
that place to hold the drinking
competition with the rain, they
called it mykha badri in Garo. Until
now all Garo villages call any rain
mykha badri. But the Badri area
calls only the rains in September
and October mykha badri marsat.
Ordinary rain is not called mykha
badri. Thats why this place is
called Badri Rongdyng Hawai, it
is said, and the rain is also called
mykha badri, it is said.
mykhal- v to be older than someone
Getheng ang mykhalgaba. He is
my elder.
mykhang n BODY/PLACE face, front
Kyi nok mykhangchi muaidonga.
The dog is sitting in front of the
house. Arong nokma
chaikhawwachi, Arong nokmamyng
mykhangaw khiaimyng thyiokno.
When headman Arong peeked,
headman Arongs face was hit and
he died, it is said.
mykhang- v to face Isang mykhangbo.
Face this way. Mykhangrukbo.
Face each other. Ang gethengsang
mykhangaidong. Im facing him.
mykhi ~ mykkhi n BODY slime from
the eyes
mykjywm- v to doze off
mykphylyp- v to blink with your eyes
mykrak- v to hold a wake (often used
with the incorporated noun wal
night) Wagaba thyigabana
sadyrangba jywgabamyng wal
mykrakaidonga. The children and
the mother are holding a wake for
the dead father.
mykraket- v to warn Ang nangaw
mykraketarong. Im warning you.
mykren ~ mykyren n BODY eye
mykren wathok songruk- to look
attentively
myksep n BODY corner of the eye
myksolkhare n BODY ring finger
myksong- v to plan, to intend Ang
kymna myksongarongchym
ytykchiba manni ma manchabai
kymna. I intend to get married but
maybe I will not be able to.
myksu- v to wash your face Angdo
phangnan ja saaimyng myksua. I
always wash my face after getting
up.
myksymyl n BODY eyebrow
myksyram n BODY eyelash
mykthoram n BODY middle finger
myktoksi n PLANT plant with
beautiful white flowers with a
yellow heart that look like big
jasmine flowers
myktyi n BODY tear myktyi thothak ni
two tears/two teardrops
myktyiwatram n BODY side of the
hand under the index finger
mykyren ~ mykren n BODY eye
myl- adj1 small
mylthai n BODY small bosom Rong
sa mylthai, rong sa chungthai. One
big bosom, one small bosom. Phak
sa mylthai, phak sa chungthai. On
one side a big bosom on the other a
small bosom.
mym clf classifier for fists and things
that are like a fist
mym n BODY a fist (counted without
classifier)
mym- v to be like a fist
myn- v to be ripe, to be cooked, to be
ready Mai mynokodo maidan syla
toka. When the rice is ripe, we
celebrate the new rice festival. Ie
panchung mynkhucha. This
jackfruit is not yet ripe. Jabek
mynok. The curry is ready.

674
myn n BODY body hair (of human),
fur (of animal)
mynsym- v to be hairy with small
hairs
myntyi n BODY/PLANT pus; rosin;
latex of jackfruit, thick fluid of
various fruits Myntyi gumukan
takapa. All thick fluids of plants
are sticky./All pus is sticky.
myng- clf classifier for spoken things,
games and for the word bostu
golpho myng sa one story khata
myng ni two words bostu myng
tham three things
myng- v to call someone/somebody a
name Wiliamnagalaw symsanggre
noai mynga. Williamnagar used to
be called Symsanggre. Angmi
bimung Braiton myngwa. My name
is Braiton. Angmyng amaaw Goje
M Sangma myngwa. My mother is
called Goje M Sangma.
-myng ~ -mi encl.phr genitive enclitic
-myng ~ -mung ~ -mu encl.phr
comitative enclitic
-myng ~ -mung ~ -mu ~ -mungna ~ -
muna encl.cl sequential clausal
enclitic
myng- clf classifier for persons
myng thamkhua there are still six
persons left
mynga- to call upon someone or
something Kambaisangmi wai
nukchido dyngthang khuruta.
Chamasangmi wai nukchido
dyngthang khuruta. Mani
myngwaan, hapawan
dyngthangdyngthang myngaa,
thokthok myngaa. Ie chamasangmi
wai khurutchido ue hyisangmiaw
Banggladesmi thyl Kongosmi jaria
hagyrsakgumukawan myngani. If
you see the upstream spirit, a
different incantation if performed,
if you see the downstream spirit, a
different incantation is performed.
As for what we call the
worshipping, different places are
called upon, they are called upon
accodring to the devision. When
[he] summons the downstream
spirit, that [priest] will call upon
the influence of all those far away
[places] up till Bangladesh [and]
the influence of Kongos, all of
them.
myngkhelek- v to call somebody by a
nickname Ang nangaw Matsumoto
myngkheleka. I call you by the
nickname Matsumoto.
myngkheleka n ACT nickname
myrumyru adv not clearly
myryng myryng adv barely Hyiawchi
habanokba ganang, myryng
myryng nuketaria. There is also a
rice field house, it is barely visible.
myt- v to extinguish Wal mytok. The
fire is out.
myte n ABSTR deity, god
mythel- v to thank, to appreciate Anga
nangtymaw mythelbiok. I thank
you very much.
na- v to hear
-na ~ -ona encl.phr dative enclitic
-na evsp V in a beautiful or nice way
na n ANIM fish
nachan n ANIM firefly
nacheng n ANIM river shrimp
nagarang n ANIM type of electric
fish
nagungphel n ANIM type of fish
najek n ANIM type of fish
nakha n ANIM type of fish
nalam n ANIM type of blue, purple
river fish that tastes particularly
good when prepared in a bamboo
cylinder (see bering-)
naluk n ANIM tadpole
namatsa n ANIM type of fish
nanang ppron we, first person plural
inclusive
nanyl n ANIM electric eel
napat n ANIM type of fish
naphok n ANIM type of fish
narong n ANIM type of fish
naru n ANIM type of fish
narym n ANIM type of fish
narymkhu n ANIM type of fish
nasak n ANIM type of red fish
nasaw n FOOD fermented fish

675
nawachak n ANIM type of fish
nawak n ANIM type of fish
nabak n ART knot
nadanggorot n BODY oesophagus
nadekaram n BODY earlobe
nagok adj2 deaf
nai n KIN aunt: fathers sister
nainokhol n KIN mother-in-law,
addressed as mani
naija tw next year, at some time in the
far future
nak- adj1 black
-naka ~ -ka sfx imperious future or
certain future suffix
nakamai n ART small basket to sow
rice from
nakhal n BODY ear nakhal sam sa
one ear
nakhal chadan n BODY part of the
head behind the ear
nakhong n BODY backside of the ear
nakhung n BODY nose nakhung goi
sa one nose
nakhungkhal n BODY nostril
nakhungkhal ni two nostrils
nakhungmyn n BODY nose hair
nakung n BODY nose; snot (liquid)
nakung di n BODY hard piece of
snot
nal- v to gorge, to stuff your face
nalangtaupal n ANIM type of fish
nalbas adv nervous
nalsasang n LOC the other side
namakai ~ nakamai n ART small
basket used to sow rice out of
nambal ~ nombol n ABSTR number
namchyk n KIN niece
namgaba n KIN niece
namnokhol n KIN daughter-in-law
nang- v to bear fruit
nang- v to need, to have to, must
nang ppron you (singular), second
person singular
nangtym ppron you, second person
plural
nangchomot- adj1 important
nanggandai adv naked
nanggandai n PERS naked person
nanggodolong n PERS naked person
nangthaigaba n BODY swelling
nangthaigaba n BODY abscess
-nap evsp V with all your heart
napit n PERS hairdresser
narang n PLANT orange
narot n PLANT type of edible tuber
narykel n PLANT coconut
narykeltyi n PLANT coconut milk
narykhel n PLANT coconut
nasi- v to suffer from a loud noise or
sound Kyi parawchido ang nasia.
When the dog barks I suffer from
the loud sound.
nat- v to scrub, to scour, to clean by
scrubbing, to remove by scrubbing
Wa natbo. Brush your teeth. Nang
nonoe tyigatchi nataidok. Your
younger sister is washing the
dishes at the water place.
natheng n BODY side of the head
nathym- v to listen to
naw n KIN younger sister. Is also used
to address a younger female cousin
or an unrelated woman younger
than the speaker
naw- v to scold Angawe amaparae
nawba nawnaka tokba toknaka.
The people in my mothers house
will certainly scold me and beat
me. Angdo Tura reengni,
dongchachido angaw baba nawni.
I will go to Tura, otherwise father
will scold me.
nawang n PERS retard, half-brain,
fool, stupid, confused person
nawchak n ANIM type of fish
nawmyl n PERS marriageable girl
nawsyri n KIN sister-in-law: younger
brothers wife
ne encl.cl/prtcl affirmation seeking tag
enclitic or particle Uchi Nepaldo:
Naa ang masu mang rajasaaw
tynangsegabone nookno. Ym
noaimyng Thengthonba
tynangokno. Then the Nepali said:
You lead my hundred cows away,
ok? Yes, he said and
Thengthon led them away, it is
said. Nemai reengbo baisiga, ne.
Go carefully, my fiend, ok?
ne n ANIM bee

676
nekat n ANIM type of bee
nekattuhp ~ nekhathup n ART hive,
bees nest
newal n ANIM type of bee
Nedyran n PLACE The Netherlands,
Dutch
nek- adj1 close, near Rame tyi
nekokno. The road is very close to
the water. Literally: As for the
road, the water is very near.
neksem- adj1 to be very near
nem- adj1 good
nemen adv very Teew wen sa rypa
nangdo nemen sylnaka
noaidongano pherue. If you go
into the water once more you will
certainly be very beautiful said the
fox, it is said.
nemgyni n ABSTR advantage, good
fortune, good luck
nemnuk- v to like
neng- adj1 tired
neng- vsec to lack, to fail to Ha
chamai Bandi, byl nengchiba chak
nengchiba iaw ryngetphabo! Take
this sweetheart Bandi, when you
lack strength, when your hands are
tired, drink this! Wak rakhiwami
gesepchian dethenge maimynawan
manai sana nengokno. When he
was herding pigs, he could not eat a
lot of rice.
nengtak- v to rest
nengtak- ~ ningtak- v to take a rest,
stop for a while
Nepal n PERS/ACT/PLACE Nepali
(person and language), Nepal
(country)
nesynyl haiwe n PLACE national
highway
net n ART medium size basket worn
on the waist to put in the harvested
rice
ni num two
-ni sfx non-certain future modality
suffix
-ni ~ -nyi encl.phr without
Chininyicha takbo. Make tea
without shugar.
ni- ~ nyi- v negative
locative/existential verb, to not
exist, to not be naa mykcha
khagalwa niwama ganang? Do
you have someone you fancy,
someone you love or not? Ning
songsyrekdo ning atongdo
dakangdo mamyng thoromaw
niwami somaichido waiaw mania.
We pagans, we the Atong, in the
past, in times when there was no
religion, we worshipped spirits.
Somaido nyiok Theres no time
left.
niet- v to switch off, to turn off Lait
nietbo. Switch off the light!
niwa interj nothing
niam n ACT custom, law, tradition
ning ppron we, first person plural
exclusive
ning ppron we, us, first person plural
nisi- v to poison Sanarai sachido ang
nangaw nisina manchaka. If you
eat a centipede, I will certainly not
be able to poison you.
-no encl.cl quotative clausal enclitic
Song damsachi morot myng sa
manai sabigyba ganangnochym.
In a village was supposedly a very
rich man, it is said.
no- v to say
nobembyl ~ nobembol n TIME
November
noga n ART tree-house
nok n PLACE house nok tham three
houses
nokap n ART door
nokbanthai n PLACE bachelors
house. Before Christianity each
village had a Bachelors house for
every clan that lived in the village.
In this house lived young,
unmarried men. They would
practice fighting, hunting, singing,
story telling and all kinds of things
that young men would have to
learn before getting married.
Women and members of other
clans were not allowed to enter the
bachelors house.

677
nokchama n KIN the relationship of
the parents of a married couple
nokchol n ART door, entrance
nokdang n ABSTR the family that live
together in one house
nokgaba n PERS landlord, house
owner, God
nokhap n PLACE level piece of land
on which a house is built
nokhap n ART door
nokhol ~ nokhor n PERS slave
nokhung ~ nukhung ~ nokkhung n
ART roof
nokma n PERS village headman, rich
man Dakangmi somaido ning
sagyrai mylbutungchido nokma
nogado manai sagasa, gam
panggasa nokma mynga. As for
the past, when we were small
children, a so called nokma was a
wealthy person, only someone with
a lot of wealth was called nokma.
nokphandai n PLACE bachelors
house
nokphin- v to return home
Nokphinniba utymdo. They will
return home.
noksam n ART wall of a house, the
piece of ground where a house is
built on Noksamchi simen, tota, tin
pirinai hama. They build the walls
of their houses with a mix of
cement, planks and corrugated iron.
Nang baba noksamchi tangka
gopgaba ganangno. Under your
fathers house lies buried money, it
is said.
noktapa n ANIM type of small gecko
that creeps up the walls of houses
at night
nokthai n ART small house next to the
main house
nokwek n ART broom (for sweeping)
nokwek kun tham three brooms
nokweng n ART floor
nol n ART fenced enclosure
nom- adj1 soft, weak, easy Dam
noma. Its cheap. Raja! Nang
damaw chosa ie ang
baisigathang pheru tamnano
noai takaidongano. Ytykyimyng
kantaraaw Tambono
tambono noai tanangarioknoe
magachakdo. Ytykyimyng tamai
chaichie tedo byirakhem
hongkotruruaimu
kaksyrangokno pheruawdo.
Nomangaidokno udo.
Ytykyimyng jalangthiriokno
magachakdo. O King, can my
friend the fox play a bit on your
drum?, he pretended to say, it is
said. Go ahead and play, go ahead
and play!, the deer pretended that
the king said, it is said. So then,
when he tried to play it, the bees all
came out and bit the fox all over, it
is said. The fox became weak, it is
said. So then the deer ran away
again, it is said.
nombok- v to be unconscious, to be
tired after eating a lot Taisa
nombokok. A little while ago he
was unconscious. nombok thyibok
almost dead
nombol ~ nambal n ABSTR number
nong- v to apply, to put (on the skin or
body, like a cream or medicine), to
smear, to spread, to crush and
smear out Khuchul pisak nongwa
lepstik She has put lipstick on her
red lips. Jaryt chamussang
nongaidong. Im crushing and
smearing out the chillies with a
spoon.
nono n KIN younger sister. Is also
used to talk about or address a
related younger female of your
generation: cousin, to address a
young unrelated female person
norok n LOC hell
nosto dong- v to be damaged
nuk- v to see, to look like, to find
Uchie phylgym chunggabaaw
nukokno. Then he saw a very big
eagle, it is said. Ie kuy matsatykyi
nuka. This dog looks like a tiger.
Sabanthaigaba noaian
tynganchakno. Kanjotokno.

678
Morottykyi nukanchakno. He did
not recognise his so called son, it is
said. He was very skinny, it is said.
He did not look human any more.
Nang Atonggawiaw sylai nukama?
Do you find Atong girls pretty?
nukcham- v to see into the future
nukhu n PLACE courtyard
nukhung ~ nokhung ~ nokkhung n
ART roof
nygyl n PLACE market
nygyltyi n TIME week nygyltyi sa tyi
ni one week or two Nygyltyi
raiagadyrangchi raiani. He will
come sometime next week.
nygyktyityi every week
-nyi ~ -ni encl.phr without
Chininyicha takbo. Make tea
without shugar.
nyi ~ ni- v to not exist Mamung
tangka niwa aro sana ryngnaba
niwa I dont have any money and
nothing to eat or drink. Ang sagrai
niwa. I dont have children.
nyng n KIN aunt: fathers sister
nyng n PLACE inside
nyngtyw- v deep
o interj Oh! This interjection can be
pronounced long to indicate
acknowledgement.
obosta n ACT event
odek n PERS baby
-odo ~ -do encl.phr.cl topic enclitic
ogynang- ~ oknak- ~ oknang- v to be
pregnant
ogynanggaba ~ okgynanggaba n
BODY pregnancy
oi interj interjection to draw
someones attention
ok n ACT hunger
-ok ~ -ak ~ -k sfx perfective aspect
suffix
okgynanggaba ~ ogynanggaba n
BODY pregnancy
okha- v to be full after eating
okhi- v to be hungry Mai okhiedok
angdo. Im hungry. Alternative
spelling: Mai okhiaidok angdo.
okhynyng- v to break a round hollow
object in half (not lengthwise, i.e.
the result is a cross section, see
thong)
okma n BODY the front of the body
Nawgabaaw janawgaba okmachi
baaidok. The elder sister is
carrying her younger sister on the
front of her body.
okmyng- v to starve
oknak- ~ ogynang- ~ oknang- v to be
pregnant
oktobyl n TIME October
ol- v to speak, talk
oltho ~ ortho n ABSTR meaning
-ona ~ -na encl.phr.cl dative enclitic
ong n ANIM wasp
ongang n ANIM big edible frog
ongang n ANIM type of frog
opis ~ ophis n PLACE office
opiser ~ ophiser n PERS officer
ortho ~ oltho n ABSTR meaning
Ostyralia n PLACE Australia
oto n ART auto rickshaw
otorewain n ACT auto rewind
otyk n PLACE bottom of ravine or
cliff
pa- adj1 low, plain, flat, thin (of
things)
pa- v to perch Sympak cunggabachi
phylgym paai musawaidongano.
The eagle is perching in a sympak
tree.
pa- vsec to dare San nidyrang
dongphinaidok, nang noksang
raianado
pachongmotchaaidokkhon
nookno. It has been two days and
maybe he really does not dare to
come to your house.
pai- v to carry by hand Banggal myng
sa biskut chyrymbiai
paiaidonganote. A Bengal is
carrying a heavy load of biscuits.
Ue alsia rajae jykba myng ni
kymanoro. Ytykyimyng jykba myng
nian sanaba jyk paithumna
nangano, jywnaba jyk paina
nangano. That lazy king was
married to two wives, it is said. So
then, as for these wives, the two of
them, when he eats his wives have

679
to carry him on their hands, and
when he sleeps his wives have to
carry him on their hands, it is said.
pai- vdat to support, to tolerate Ang
khol rangsanna paicha. My skin
does not tolerate the sun. Uchisa
matsana makbulna mongmana
paichaaimung byldyng byldang
jalna habachengok. Then, not
bearing the bears and elephants any
more, they started to run away all
over the place.
paila n ART scale (for weighing)
paip n ART water pipe
paitaw- v to lift up
pakara ~ pakyra n PLANT stalk of a
fruit, cord for kukuri
pakyl n ART centre strap of a sandal
pal n PLANT flower Pal manok. The
flower is blossoming. pal mochoka
a flower bud
pal- v to bloom Pan palaidonga. The
tree is in bloom.
palak n ART bamboo spoon: piece of
bamboo split in half and used to
stir
palengma n PLANT Barebina-
xariegata, tree with beautiful white
flowers that smell very nice like
magnolia and are edible
palong n ART bed
pan clf classifier for apparatus,
appliances, mechanical and
electrical things, cars, bikes,
bicycles, mortars and umbrellas
gari pan sa one car redio pan sa
one radio satha pan sa one
umbrella thep pan sa one tape, tibi
pan sa one TV, asam pan tham
three mortars
pan n PLANT tree, firewood pan phan
sa one tree pan dot sa one log
panpyrak- v to cut breadth wise
panachol n PLANT mushroom (not
edible)
panbai n PLANT firewood
panchak n PLANT leaf
panchoka n PLANT small log
panchong n PLANT tree trunk
panchung n PLANT jackfruit
panchungchongsu n ANIM type of
black hairy caterpillar that lives on
jackfruit trees
panchyksi n PLANT twig panchyksi
goi sa one twig
pang- adj1 a lot, many
pangwami n ABSTR quantity,
abundance
pangkol n PLANT guava
pangkywal n PLANT guava
pangyrym n PLANT jungle thicket
panju n PLANT firewood
pankhol n PLANT bark (of a tree)
pannok n wood shed
panphek n PLANT sapling, young tree
pansok n PLANT end of a branch
where the tree grows
panthai n PLANT type of fruit
pantiki n SUBST wood chip
pape n ANIM tokkeh, type of gecko
papol n FOOD pasta
papret- v to throw to death
-para encl.phr associative plural
enclitic, X and company, X and
those associated with him/her
parang n PLANT/ART reed, thatch
parang v to wander, to go astray
-parang evsp V without destination, V
without goal, V aimlessly
parap- v (too) salty
paraw- v to call (of animal), to shout
(of animal and human)
parawchyrik- v to shout loudly Morot
sorokchi kepai parawchyrikaidong.
The man was shouting loudly on
the road because he was angry.
-pat evsp V across
pat- v to cross Sikhar khana
reengokno reengokno
reengokno. Tyikhal goisaaw
patna nangokno. They went
hunting, they went and went. Then
they had to cross a river.
patal ~ phatal ~ phathal ~ pathal n
GEO stone
patyl n ART slingshot
paw- v to agree
pawjong n KIN elder brother. Is also
used to address an older male

680
cousin or a man older than the
speaker.
pawai n ART bowl or its volume,
bowlful classifier for curries
Atong jabek saak? Alu
nasaw pawai sa, taw khirip
pawai sa. What curry did you
eat? Potatoes with fermented fish
and chicken with khirip.
pawdyr n SUBST powder, baby
powder
peel ~ pheel dong- ~ dong- v to fail
Getheng lekha nemai poreancha,
ytykyimu poreka peel dongok. He
did not study the book well, so then
he failed his exam.
-pek encl.phr distributive enclitic
pek- v to be drunk pekok drunk
peket clf classifier for packets Sigyret
peket sa ganangkhuama? Do you
still have a packet of cigarettes?
peking ~ pheking n ACT luggage,
packing
pel- v to fuck
pelang adj2 flat
peleng- ~ peleng- v to deflate
pen n ART pen
peng- v to prevent, to block, to hinder
peng- v to curse Takgaba
Rywgabasang Phatigaba
Rarongabasang phalthang pengai
tananggabaaw raphinkhana
dengetkhana. The supreme god
wanted to lift the curse that he
himself had put [on the village].
pereng- adj1 straight
-peret avsp to V open, to V so that it
splits (open)
peret- v to split, to cut in half, to
crack, to explode
pering- v to be straight
pering tongtong adv straight
-pha evsp V also, V in addition, V
along with, V together (S/O
quantifier)
pha- v to dare Noksang rainaba
phaphinchaaidok. Jebadong anga
takruksyrangarinaka. Matsami
chaphungaw wangjoloknoaro,
alsia rajae. I dont dare to go
home. Anyway, I will just fight to
the end He bit the tiger on the
thigh, it is said, the lazy king.
phaat- ~ phaet- v to apply, to put
on, to put on a wound, to apply to a
wound Sambanggyri akaiokno,
tokdepdepaimu phaatokno. He
plucked sambanggyr, crushed it
and put it on the wound, it is said.
Jywgaba sagaraiaw dithap
phaetaidonga. The mother is
putting a diper on the child.
phagongma ~ phagungma n BODY
shoulder
phai- v to break, to translate
phaithong- v to break a solid object
in half (not lengthwise, i.e. the
result is a cross section, see thong)
phaikana ~ paikhana n PLACE toilet
phaithawa n BODY cheek
phak clf classifier for halves of
objects cut lengthwise
phak n SHAPE side, half which is the
result of a longitudinal section or a
cut along the length
-phak evsp V lengthwise, V for a little
while, V by the side of something
phak- v to throw out, to empty
phakphaklak- v to spill
phakset- v to throw away (for solid
substances and things)
phakwal n BODY armpit
phal n PLACE share, shift of work,
instead of Ang reengsigama
nangmi phal? Shall I go instead of
you?
phal- v to sell Ang ie narykhel teen
nygylsang raangaimyng phalni. I
will bring these coconuts to the
market and sell them later.
phalak n ART piece of old cloth used
to clean things
phalap n PERS whore, prostitute
phalthang ppron self
phalthangthang ppron selves
phan clf classifier for trees; classifier
for food packed in bundles in
raichak pan phan sa one tree
phang clf classifier for grass, trees and
flowers narang phang sa one

681
orange tree narang rong sa one
orange
phangnan adv always, never
Phangnan rupek mugabachido tyi
ganang. There is always water
where there are frogs. Thawgaba
symgaba phangnan
sarongchagaba jilami
bostudyrangaw raai hynaimung
khasin khasin gumukawan
palyngchi jalgabadyrangaw
jykthangthangaw jumuphynaakno.
Having brought and given tasty and
sweet things from the district
which are usually never eaten, they
slowly recollected all their
husbands who had run away into
the jungle, it is said.
phangphyl adj2 upside down
phangphylok to be turned over, to
be upside down
phanthai n PLANT type of sour fruit
phas n ABSTR the first one
phasgaba adv first Phasgaba
the old rice stalks.
phat clf classifier for clothes
phat- v to chuck away, to throw out
phathi- v to bless, to bestow upon
phatsai n ART womans dress
phe- v to disembowel, to gut (fish),
pheep ~ phep n PLANT banyan tree
phebaw n PERS person with a swollen
cheek
phebuari n TIME February
pheel ~ peel dong- ~ dong- v to fail
Getheng lekha nemai poreancha,
ytykyimu poreka peel dongok. He
did not study the book well, so then
he failed his exam.
phek clf classifier for smaller
branches of trees dala phek sa one
branch
pheking ~ peking n ACT luggage,
packing
phel clf classifier for flat baked things
and coins barata phel sa one
paratha, biskut phelsa one biscuit
tangka phel sa one coin
phep ~ pheep n PLANT banyan tree
pheru n ANIM fox
-phet sfx V detrimentally
phet- adj1 to be thin
phet- v to swell up
phet- v to arrive (at), to reach, to come
out of the water, to emerge
Ytykyimyng ue raja nygylchina
phetokno. So then the king arrived
at/reached the market, it is said.
Bewal rypaimyng phetaakno.
Having been in the water for some
time, he emerged, it is said.
Ytykyimyng jenetne rajamyng
noksang phetangokno. So then he
somehow reached the house of the
king, it is said.
-phetphet evsp V repeatedly
phi- v to invite Beanbebe
montyridyrngba Bilaw phina takyi
hongkotangthokokno. The
ministers truly all went out to invite
Bil, it is said.
philm n ART film, movie
phing- v to be full Gylaschi tyi
phingok. The glass is full of water.
Gylas phingok, diphingna
manchaka. The glass is full; you
cannot fill it any more. Nangmi
khathong bangbang dongchido
ang phingetni. (Samrat N Marak) If
your heart is empty, I will fill it.
phing- v to fill
phingpyryt- v to be overfull
phoot ~ phot n ANIM mythical black
amphibian like a salamander
phok- v to swell, to lift up
phok- v to uproot, to pluck
phone ~ phoon n ART/ACT
telephone, telephone call Angna
phone khaetboto! Call me (on the
phone)!
phong clf classifier for cylindrical
objects and for long sharp or pointy
objects
phong n ART wooden handle of big
knives, axes and spears
phong n ART fire place for cooking

682
phongthu n ART fire place for
cooking
phoren n PERS/PLACE white person,
foreigner, foreign country
phot ~ poot n ANIM mythical black
amphibian like a salamander
phuchul n ANIM big water reptile
that can eat humans Kha Dawa
nochachido phuchul raarianoro.
If you dont say Kha Dawa the
phuchul will get you, it is said.
phuchul n ANIM type of big water
lizard
phulis n PERS police
phulkobi n PLANT cauliflower
phuruk- v to become uprooted Bildo
kyryiaimyng pandyrangchi
pyichiba panba baiariokno,
wachi pyichiba waba
phurukariokno. As for Bil, because
he was afraid, when he held on to
the trees, the trees would break,
when he held on to bamboo, the
bamboo would just become
uprooted, it is said.
phuthi n ANIM type of fish
phyl- v to transform, to change into
phyl- v to transform, to change into
Imi wa juw wak phylwa. Her
father and mother have changed
into pigs.
phylgym n ANIM type of big eagle
phyltawtaw adv jerkingly (over a
rough road)
phylyp- v to blink (with your eyes)
Mykrenmi phylyp chaiwaan. I
looked at her with blinking eyes.
(Gostar R Sangma)
phylyp- v to blink once with your eyes
phyn- v to cover Ang kombol
phynaidonga. A blanket is covering
me./I am lying under a blanket.
-phyn ~ -phin ~ -phyn ~ -phin evsp
V backward, V back, over-V, V
overtime, V fully
phyt- v to slice
phyw- v hollow
phywra n FOOD rice powder
pi- v to ask, to beg, to pray
pisa n ABSTR childhood
pithyn ~ bithyn n BODY liver
piti ~ pityi n FOOD rice beer (gold
coloured)
pityi ~ piti n FOOD rice beer (gold
coloured)
pityng n ART thread, necklace
pibok adj2 white, unripe, very light
green
picham adj2 old (of things)
pidan adj2 new
pidio n ART video
pido n ACT game played with small
stones. The game can be played by
just one person or a small group of
people. Every round consists of a
series of jugglings that become
more difficult each time a round is
completed.
pijyw n PLANT rice seeds for sowing,
newly harvested rice
piktiyr n ART picture
pinak adj2 black
ping- v to block the way
pipuk n BODY belly, intestines,
bowels, stomach
pirin- v to mix
piryt n BODY gall bladder
pisak adj2 red, blond
pit clf the length of two fists and two
thumbs when one joins them
potolong n PERS person with a naked
chest
pok- v to swell
pokotia n PERS freeloader, sponger,
person who takes advantage of the
kindness of others
poop n FOOD triangular pastry eaten
with tea
porai- ~ pore- v to read; to study
poram- v to fly over
pore- ~ porai- v to read; to study
porika n ACT exam, examination
puksuk n BODY waist, side of the
body
puktyng n BODY small intestine
pung n ART granary, rice stock house
Mai bytwamyngdo pungchina
songchina khairata. We carry the
rice harvest down to the rice stock
house, to the village.

683
purun n ANIM goat
pusipusi interj interjection to call a
cat
puspus interj interjection to call a cat
puyikhyrep v to crush with your hand
pyi- ~ pyi- v to touch, to grasp Uchie
songmyng morotmyng jyrym
thymaimyng Thengthonaw
rawokno pyigoropokno. Then the
people of the village, having
quietly lain in ambush, caught
Thengthon, they grasped him all
pyikhap- v to catch
pyiram- v to feel for, to search by
feeling
pyikhep- v to hold firmly
pyiru- v to collapse
pyjyw- v to sow seeds by scattering
them
-pyl evsp V rapidly
pylang adj2 flat Ytykyimyng teedo
amak gethengdo rong pelang
sylgabachi kepleplep bamai
hyntakkonoa. So then, now the
monkey, as for him, he willingly
lay down stretched out on his belly
on a flat stone, it is said.
pyleng- ~ pyleng ~ pyleng v to be
fat Gari bengbylokaw depylengok,
ytykyimu bengbyloke pylengok. The
car flattened the toad so the toad
was flat.
pyn- adj1 dense, thick
pyn- v to pack, to wrap up, to pack in
a banana leaf, to cook in a banana
leaf
pyndap- v to cover
-pyrak evsp V and cut
pyrap- v to be too salty
pyru- ~ pyryw- v to pierce, to make a
hole in something
pyryi- v to be mature
-pyryt evsp over-V
pyryw adj2 to have a hole in it (of
walls)
pyt- v to wrap neatly as a present
pyw- v to fly
pywgak- v to crash (in flight)
pywtaw- v to jump over something
Muraaw ang pywtawa. I jump over
the small stool.
ra procl Give!
ra- v to bring, give
ra- v to get, buy, take Nang ie tupi
bimi raak? Turami raak.
Where have you bought that cap?
I bought it in Tura. Ang nangaw
bebe racha. I dont believe you.
Wami jywmi balgabaaw katha
rachagabae anga ytykgachina
dongok. Because I did not listen to
the words of my parents, I have
become like this.
raang- v to take away
rasak- v to welcome Nangtym
angaw tyichi typratwaba nemariok
aro koksep chungkhuna
nangachym. Watyng tyngphekna
masu mangphek hynwa, gumuk-
gamak angna masu mang raja sa
hynetwa angnado aro
tyinyngsangba nawmyl
sylsylgabasa rasaksawa. You
threw me into the water and that
was good, and I should have had a
bigger koksep. For every bamboo
strip they gave me a cow and in all
they gave me one hundred cows
and inside the water only beautiful
girls welcomed me.
rasek- v to snatch
rasek- v to accept, to receive Angdo
myngsa agrai rasakchawa. I will
not accept/receive more than one
person.
raani ~ rani n PERS queen, also used
to call your daughter when she is a
little child, like in English little
princess
Rabuga n PERS god who created the
world according to ancient religion
rai n PLANT reed
rai- v to go; to come Hai, rainaka
Come on, lets go. Kynsangdo
matsado morotsyn manaimyng
raiwilokno alsia rajado. Later,
after the tiger had smelled the scent
of a human, it walked in circles, it

684
is said, around the lazy king.
Maisanaw sana reengbutungchi
sabutungchi Nepal myng sa boba
takgaba raiawanoro. While they
were leaving to eat lunch, a crazy
Nepali came, it is said.
raia- v to come Phorenmi morot
raiadonga, phorensangmi
raiaidonga. Foreign people are
coming, they come from foreign
countries. Angdo hanep
nangsang reengni. Raiabo.
I will go/come to your place
tomorrow. Come.
raibyt- v to carry around
raichak n ART big leaf used to pack
food
raiganggang- v to go/drive/ride over
things on a bumpy road Rongaw
raiganggangwa. I bumped over a
stone while going.
raiphak- v to hit with your elbow
while walking
raisotwa n ACT shortcut
raityng n ART washing line, clothes
line raityng tyng tham three
washing lines
raiwil- v to walk around something
raidi n PLANT turmeric
raithai n PLANT tree with thorns on
its stem
raityng n ART line (to dry clothes on)
raityng n PLANT cane
raja n PERS king
raja num hundred raja sa one hundred
rak- adj1 hard, difficult, loud, glottal
stop
raka n ACT glottal stop
rakhi- v to guard, to keep
rakhigaba n PERS caretaker
raki- vdat to protect, to guard against
Ning habachi mongmana amakna
mai saniwana rakiaronga. We are
protecting our dry rice and
vegetable field against elephants
and monkeys so that we will eat
rice.
ram n PLACE road, way, path ram
chol tham three roads, paths, ways
-ram evsp V inadvertently, V
unintentionally, V fortuitously, V
because of the situation
ram- v to dry in the sun, to put in the
sun to dry Garu balagachi ramai
tanaimuna, habasang ha kamna
reengokno. After she had put the
mustard leaves outside to dry in the
sun, she went to work in the rice
field, it is said.
ram- v to search, to want
ramram adj2 ordinary, normal Ie
ramram dongcha. This is not
normal. Baisigathangmaran tyi
dukungokno. Nado
ramramanchakno. The friends
dammed the water. There was
plenty of fish.
-ramram evsp V normally, V
naturally, V commonly
ramramtyi n BODY sweat
ran- adj1 to be dry
randai n FOOD meat, flesh, body
rang n ART type of drum (instrument)
rang n GEO rain Rang waaidong. Its
raining. Rang nemok. The rain has
stopped. Rang nemchengama
nanang chyw jamchenga noai
rangmu chyw ryngsusaie range san
chi byri wawano. Will the rain
stop first or will we finish our
liquor first? they said and while
competing in drinking with the
rain, the rain fell for fourteen days,
it is said.
rangset- v to breathe
rangbrym n GEO cloud
rangbyrym- v to be shrouded
in/blocked by clouds Rangsang
rangbyrymaidonga, rang
wanikhon. The sun is blocked by
the clouds, it might rain.
rangchinek n GEO cloud
rangdylekpa n GEO lightning
ranggorai n ANIM macaque. Monkey
with a long tail, brown body and a
red face.
rangra n GEO sky
rangsan n GEO/TIME sun, day

685
rani ~ raani n PERS queen, also used
to call your daughter when she is a
little child, like in English little
princess
rap- v to thatch, to roof
-rara encl.phr exclusively, only,
among, amongst
rasong n ACT boasting, praise Rasong
manai takokno usa,
chungchunggarangsa, udo
phylgymawdo jonggaba kawwano.
The eldest ones boasted a lot (about
themselves) although it had been
the younger brother who had shot
the eagle, it is said.
rasun n PLANT onion rasun pibok
garlic rasun pisak red onion rasun
tyisuk type of onion
-rat evsp V downward
rat- v to throw Matsa rong ratwa. A
tiger threw a stone.
ratat- v to take out
ratsok- v to miss the mark
raw- adj1 tall, long
raw- v to catch Changba bydyi myng
sa khen rawarong Somebody, an
old man, is catching river crabs.
Uchie songmyng morotmyng jyrym
thymaimyng Thengthonaw
rawokno pyigoropokno. Then the
people of the village, having
quietly lain in ambush, caught
Thengthon, they grasped him all
rawreng- adj1 slender and long
rawsoksok- v to fail to catch
-rawraw evsp continue to V
rawsykot- v to slip out of the hand
reeng- v to go, to go away, to leave
Bisang reengwa naa? Usang
nalsasang reengwa. Where do
you come from? I come from the
other side of the river. Kynsange
nygyltyi ni reengwachi thikthak
jahas kanachina dongangok.
Later, when she had been going for
two weeks, she arrived exactly at
the ship and the harbour. Garu
balagachi ramai tanaimuna,
habasang ha kamna reengokno.
Having put the mustard leaves
outside to dry, she went to the
haba to weed. Hanep nanang
myng ni Turasang reengnine.
Tomorrow the two of us will go to
Tura, OK? Angdo hanep
nangsang reengni. Raiabo.
I will go/come to your place
tomorrow. Come.
redio n ART radio
reel n ART train, rail
rek n PLANT banana tree
rekhep n PLANT type of huge beans
rekhep- v to be dry (of plants), to be
wrinkled (of person)
rekkun n PLANT banana flower
rekphang n PLANT banana tree
rekphul n PLANT non-edible banana
flower
rekthai n PLANT banana
relgari ~ reelgari n ART train
rens
n ART wrench
repa chepa adv in various places
ret n ACT childrens game played in a
grid. There are hunters who may
only move along the lines of the
grid. The other children have to try
to cross the grid without being
touched by a hunter.
rewet n PLACE riverside, riverbank
-ri ~ -ryi encl.phr without, privative
enclitic
ri n BODY penis ri goi ni two
penises
riambanthai n BODY gland of the
penis
ribaw n PERS person with one
testicle bigger than the other
riganthong n BODY erect penis,
erection, hard-on Nangchi
riganthong ganang. You have an
erection/ a hard-on.
rigol n BODY penis (used as
swearword for men), dick
rikaran ~ rikeren n BODY testicle,
balls, scrotum rikaran rong ni two
balls, testicles
rikhuchul n BODY foreskin
rikun n BODY glans penis, dickhead

686
rimyn n BODY male pubic hair
ripan n ART a short dress that
women wear around the waist
riros n BODY sperm, semen
risokop n BODY scrotum
rityi ~ riti n BODY sperm, semen
rijap n PLACE forest reserve
rimirimi adv used in the expression
Mykren rimirimi takaidong. My
eyes are closing because Im so
tired.
rimyl- adj1 slippery
rin- v to keep as domestic animal
ring n PLANT taro, type of edible
tuber
ringaba n ART place where you keep
a domestic animal
ringaba n ART fishery
riphi- ~ ryphi- v to plaster (with a
mix of clay and cow dung)
riprip- v to rub
roal n PLACE (lower) primary school
robol n ART/ACT football, soccer
robolphil ~ robolpil n PLACE soccer
field, football field, playground
rochok n ART picket, pole
rochong n PLANT tree stump
-rogoi encl.phr alternative enclitic
rok- v to shave Kamyn rokai matok.
I cut myself while shaving my
beard.
rokhom n ABSTR shape, type
rokset- v to wipe off
romthom- v to be spherical Robol
romthoma. A football is round.
rong clf classifier for small round
objects, money, small stones,
seeds, stones in a game (when they
have a value) and fruits, default
classifier for counting buchuot
rong sa one mango tangka rong
chek ten rupees
rong n GEO colour
-rong evsp usually V
rong n GEO stone rong thut sa one
stone rong rong sa one small
round stone in a game
rongcheret n GEO pebble size stone
rongchun n SUBST lime stone (in
rock form)
rongchung n GEO big rock
rongchung thut sa one rock, one
big stone
rongchyret n GEO very small stone
rongdep- v to crush with a stone
ronggyrym ~ rongrymrym adj2
GEO being full of big rocks, stony
land Ie ram ronggyrymrara, angdo
raichawa. This road is full of big
stones, I will not go.
rongka n PLACE cliff
rongkhal n PLACE space under a
stone Ichi rongkhalchi khen
ganangthelnaba ganang. Here in
the spaces under the stones there
are river crabs for sure.
rongkhol n GEO cave
rongmisi n GEO very small stone, to
be covered by a stone
rongpatal n GEO big rock
rongphek n GEO a grain of sand or
very small stone
rongrymrym ~ ronggyrym adj2
GEO being full of big rocks, stony
land Ie ram ronggyrymrara, angdo
raichawa. This road is full of big
stones, I will not go.
rongsa n ART whetstone, flat stone
for sharpening knives or edged
tools
rongsyrek n GEO small stone
rongthai n GEO a rock
rongthyk n GEO a big rock
rongmyng- v to shuffle cards
-rongreng evsp V while spinning
around
rongthal- v to clean, to clarify, to
explain
rongthala- adj1 clean
rongtyk n ART large clay pot to keep
rice in, rice pot
ronok- adj1 smooth
rophil- ~ rophyl- v to joke
ros n BODY/PLANT/
FOOD sperm, semen, juice (of meat
and fruit)
rot- v to boil (something in water)
rubibal n TIME Sunday
ruchut- v to join, to connect Sala
burbok sagyrai naa ningaw

687
halakhagabaai noaimungna
bunduk raasetetaimungna uaw
sagyraiaw gadakchichiokno.
Gadakchichiaimuna
singsingkholongsang typsetyi
tanangokno, typsetyi tanangokno.
[] Ytykyimuna uba
ruchutethiriaimungna
raiaronganoro. Damn you stupid
child who disturbed us! they said
and they took out their guns and
smashed the child into pieces, it is
said. Having smashed him up, they
threw him into a deep hole in the
ground and left him there, it is said.
[] But then it came (home) back
after it had joined together again, it
is said.
ruda n PLANT type of cactus
rugung n PLACE edge
-ruk sfx reciprocal suffix
rukchok ~ lukchokchok n ANIM type
of frog
rukpek ~ lukpekpek n ANIM type of
frog
rukwak ~ lukwak n ANIM toad
-rum evsp all (S/O quantifier)
rum num twenty. This word is only
used in compound numerals rum
tham sixty.
rumal n ART head band
rung n ART river boat made out of a
hollowed tree trunk rung khan sa
one boat
rungkhut n FOOD broken rice
runi n BODY brains
rupek ~ rukpek n ANIM type of small
frog which says pekpekpekpek
-ruru evsp more and more V around,
V all over the place
ruru- v to make liquid come out
rydym- v to sprout leaves
rygyn n PLACE side rygynchi near,
next to Ang choki rygynchi
chapaidonga. Im standing next to
the chair.
-ryi ~ -ri encl.phr without, privative
enclitic
ryk n ART necklace
ryk- v to chase, to herd Sagyrai
masu rykarok. The children are
herding the cows. Taini kakai
sachongmotnaka noaimyng
rykaidokno magachakaw banggale.
Today I will really devour it, the
Bengal said and chased the deer, it
is said.
rym- v to cook
rymkhap- v to cook without maityi ~
maiti Angdo jabek rumkhapni. I
will cook curry without maityi ~
maiti.
rymreng rymreng adv dazed
rymyl n PERS 1. marriageable female
cousin, 2 the relation of female
cousins from intermarriageable
families, 3. the relation of the
parents of a married couple, 4.
girlfriend, lover, sweetheart
rymyt adj2 yellow, orange
ryng- v to drink Tyi ryngbo. Drink
water. Sigyret ryngbo. Smoke a
cigarette.
ryng- v to sing
ryngchyw ~ ryngchu n FOOD
flattened rice
ryngkhaw- v to drink sneakily
ryngkhele- v to drink for fun
ryp- v to dive, to be/stay under water,
to immerse, to submerge
thorokangokno. Thorokangaimyng
hawtyi rypokno magachake. Bewal
rypaimyng phetaakno. So then, the
deer, having put the biscuits on the
river bank, splash! jumped in, it is
said. Having jumped in, he stayed
under water for some time, it is
said, the deer. Having stayed under
water for some time, he emerged, it
is said.
ryphi- ~ riphi- v to plaster (with a
mix of clay and cow dung) Nok
riphiaidong. Im plastering the
(floor of) the house.
sa interj interjection to chase away a
chicken

688
sa num one, a/an Combined with the
attributive suffix, the form sagaba
~ saga can mean first, or
onethe other. Ue gawichi sa
myng korok ganangnoro aro
detheng pipukchi ganangkhua
myng sa. That woman had six
children, it is said, and in her belly
she had one more. Uchie ramchi
pheru mang sa gorongwano. Then
they met a fox on the road. Unasa
boba myngsagaba teew abun
boba nukaisigaakno. Then the first
fool saw another fool, it is said.
Songsagabaaw Songmong
myngwanowa, songsagabaaw
Songgadal myngwanowa. One
village was called Songmong, it is
said, the other village was called
Songgadal, it is said.
-sa encl.phr.cl delimitative enclitic
sa- v to be ill, sick dykym sa- to have
malaria
sa- v to set in place as a trap
sa- ~ sai- v to set as a trap, to set in
place, to do Jaga saakno uchie,
taw pangai banokno. They set
traps and then caught many birds, it
is said.
sa n PERS child
sa- v to eat Maijyreng sacha, wakna.
You dont eat dried rice, its for the
pigs. Darai warem saak. The
sword has rusted. Getheng gol
saak. He has got/kicked a goal (in
football). Mamathanggaba
samynchykgana khyrethangaw
raai sanaka. Mothers brother
will marry his daughter to her
cousin. Song damsachi morot
manai sagaba ganangnochym. In
a village lived a rich man.
(Literally a man who eats in great
amounts.) Gethengdo morot wa
saagaba. He is a tough person.
(Literally He is a person who eats
bamboo). Nange bichi krismas
sanima? Where will you celebrate
Christmas?
sabanthai n KIN son. Is also used to
talk about the male children of my
elder or younger brother.
sadap- v to spill, to take more and
more
sagyrai n PERS child, sagyrai odek
baby Sagyraiwana khymchawa.
Because shes a child I will not
marry her.
sakhaw- v to steal Hai baisiga
biskut sakhawna noaidongano.
Come on, my friend, lets steal the
biscuits, he said, it is said.
sakhele- v to eat for fun
salak- v to lick
samynchyk n KIN daughter. Is also
used to talk about the male children
of my elder or younger brother.
sanala ~ sanyl- vdat to be jealous of
Getheng angna sanala. Hes
jealous of me. Getheng angna
jama sanyla, Hes jealous of my
shirt.
sarong adj2 to be of the same age
sathup n BODY uterus, womb
satyra n PERS child without father
sabun n ART soap sabun thut ni two
bars of soap
sadu n KIN brother-in-law: the
relation of men who married
women that are sisters, sadu
chunggaba the elder brother of a
sadu sadu mylgaba the younger
brother of a sadu
sagal n GEO sea
sai n PERS husband
sai- v to write
sai- v to choose, to select, to elect
sai- ~ sa- v to set as a trap, to set in
place, to do Jaga saakno uchie,
taw pangai banokno. They set
traps and then caught many birds, it
is said.
saido n ART fishing line
saigon n PLANT teak tree
saigyn n ACT the third weeding of the
haba Mai kaimanwamungsa

689
sainokga(ba) n PERS author
saip ~ saep n PERS European, white
person, British military commander
sajin n BODY illness that makes
everything taste bitter
-sak evsp V appropriately
sak- adj1 red
sak- v to fit
sak- v to make a rope by rubbing
thread between your hands
sak- v to bear, to persevere, to endure,
to enjoy, to hold out, to be patient,
to suffer Rangsan sakna
manchaaimyng nokchi dangok.
Not being able to bear the sun any
more, he went into the house.
balwa saka to enjoy the wind Naa
ie sastiaw rakna manchido
jokangni. If you can endure this
punishment, you will be freed.
Gongwanasa balwa sakai
muarong" noatakokno amakba.
Im just sitting here enjoying the
wind because I want to, said the
monkey. Aiaw! Biskynba bylsi
nidyrang dongphinai duk
sakwachido dethengna mamyng
tangka poisa, dethengmi duk
sakwana, wak rakhiganaba, tangka
poisa hynchano. Oh!
Approximately two years have past
in which he suffered from sorrow,
he got no money for his suffering
and for the pig keeping they gave
him no money either.
sak- vdat to depend on Ang maharina
sakaidong. Im depending on my
family.
sakchyk- v to behave well Ie sagyrai
sakchykna mancha. These
children cannot behave well.
saket- v to insert, to plug in
sakhi n PERS witness
sakhyna- v to be wounded
saknaram ~ salnyram n PLACE east
sakrem- v to twist
salam interj hello
salam- ~ selem- ~ serem- ~ saram- v
to break/tear easily, to be easily
damaged Ie mudupan serema. This
papaya tree breaks easily.
salgypeng n PLACE south
salgyro n PLACE north
Saljong n PERS sun god
salniram n PLACE west
salnyram ~ saknaram n PLACE east
sam clf classifier for hands, arms,
legs, feet, ears and tires nakhal sam
sa one ear cha sam sa one leg/foot
taiyr sam ni two tires
sam n PLANT/FOOD medicine, plant,
weed
sam- v to wait Mosa naa sambota.
Hynniba nangna teen. Hey
friend, wait! I will give it to you,
ok, later.
samalmaisirong n ANIM very small
type of ant
samanggyri n PLANT type of small
plant
sambanggyri n PLANT medicinal
plant that stops bleeding. Uching
chaaw kakaimu, sambanggyri
tokdepdepaimu phawa. Because a
leech bit him on the foot, he
crushed sambanggyri and put it on
the wound.
samchak n FOOD vegetable
samkong n PLANT high grass
samsai n PLANT low grass
samsi n PLANT grass
samsin n BODY abscess, boil
san n TIME day Range san chi bri
wawano. The rain fell for fourteen
days, it is said.
san- v to put in a bag
sanarai n ANIM centipede

690
sandi- v to inquire, to search for Abu,
angdo dadaparaaw
sandiedongachym nookno. Ha?
Dadaparaaw? Nang dadaparado
usang phylgym chunggaaw kawna
reengwanote. Grandma, I am
searching in vain for my elder
brothers he said. Huh? Your
elder brothers? Your elder brothers
went that way to shoot the big
eagle!
sang bound side, place (see also
sangphak) chamasang
downstream, bottom of a hill
kambaisang ~ khambaisang
upstream, top of a hill
-sang encl.phr
mobilitative/instrumental enclitic
sang- v to burn Ie pan nemai sangni.
This wood will burn well.
sangori n GEO fog
sangphak ~ samphak bound side
isangphak this side usangphak that
side habyrisangphak the side of
the mountain Ytykyimuna
kambaisangmi dinggaraimi
chaichiba, uchiba naan niokno.
Uchiba matdam saakno uawba.
Ytykyimuna san thongsachinado
sathiriokno. Sathiriaimungna umi
chaithirichiba, ba, matdam
saakno, aro kynsang
gasamsangphak chaithirichi
uawba matdam saakno. So then,
whenever he went to inspect the
fish trap, then there were no fish, it
is said. An otter had eaten it. So
then, he set his traps out in the river
again for half the day, it is said.
Having set out the traps, when he
later inspected them again, well,
the fish had all been eaten, it is
said, and later, when he looked
again in the evening, an otter had
eaten them again. it is said.
sangwal- v to forget
sanmaji n TIME noon, midday
sanyrai n ANIM centipede
sap- v to swoop down (of birds of
prey)
sap- vsec to know a skill Bildo
teewba gore dungna sapchanotyi.
Bil does not know how to ride a
horse, it is said to my surprise.
saphairam n PLANT type of
medicinal plant
saphang n ACT first rice harvest (in
August)
saphaw n ANIM rabbit
sapset- v to drain piseri sapsetbo drain
the fish-tank
saram n ACT new rice offering
festival in which the first rice is
offered to the gods or spirits.
saram n FOOD dry rice grains Saram
sywai saa. Dry rice grains are
flattened (by pounding them with
an asam in an aman, and eaten.
saraw- v to borrow Bengmi tangka
sarawni angdo. I will borrow
money from the bank. Ang nangna
tangka sarawai hynni. I will lend
you the money.
sasti n ACT punishment Sakhawchido
naa sasti manni. If you steal, you
will be punished.
sat clf classifier for bundles garu sat
tham three bundles of mustard
sat- v to spill
sat- v to hit with a stick or bat, to cut
with a sword
sat- v to flush out
satha ~ sytha n ART umbrella satha
khung byryi four umbrellas
sathup n PERS sick person
satkhap- v to box, to slap
satpyret- v to hit with the open hand
-saw evsp V expectantly, V and wait,
keep V-ing, V and stay, V
patiently, V certainly
saw- v to be rotten
saw- v to curse at (use bad words)
saw- v to dig Nokdanggumuk gopram
sawwa habyri nalsasang. The
whole family dug graves at the
other side of the hill. Uchie
Thengthon khudalsang haaw
sawaidongano. Sawaidongano,
thywangaidokno, chakyw
chyigykdarangdo. Then

691
Thengthon is digging in the
ground with a chopper, it is said.
He is digging and he is getting
deep, it is said, about ten knees
deep.
saw- v to burn, to roast on the hot
ashes of the fire
Nokphandaidyrangaw sawaimung
nok phandai dokhakhuchi
khachapai tangaba mongma wa
dora byryi donggabaaw raai
jalangokno. Having burnt the
bachelors houses, they took the 20
KG weighing elephant tusks which
were tied to the dokhakhu and ran
away, it is said. Ramchi agal
sawgaba ganang. On the road is a
burning forest fire. Jaryt sawai,
mantaw sawai, maichengmung
nalammung thiksa berengai saa.
We roast the chilli pepper, we roast
the brinjal and cook it until it is
well done with maicheng (a type
of leafy green) and nalam (a type
of fish) in a bamboo cylinder and
eat it.
sawmyk- v to smell rotten, to smell
foul
sawsaw- v be able to cause a burning
sensation
sawel n PLANT type of vegetable
sawkun n ANIM vulture
sawn n ACT sound
sawthal n PERS dirty person, person
who never washes
sawyl n PLANT type of vegetable
-sega ~ -siga evsp V in turn,
(alternative suffix)
-sek evsp to V and steal
sekari n ART pin lock
sel- v to leak Tenkimi mobil selarong.
Oil is leaking from the tank.
sel- v to pour
selsoksok v to masturbate, to wank, to
jack off, to jerk off, to whack off
Ri selsoksokni angdo. I will wank.
selu n ANIM cockroach
-seme evsp V reluctantly
sendel ~ sendyl n ART sandal, sendel
jora ni two sandals
sene num seven
seng- adj1 clever, intelligent
seng- v to shine, to dawn, to become
light Wal sengwachi Sijunygylsang
reengni. At dawn we will go to
Siju market.
seng- v to shine
seng- v to bother by misbehaving
sengsot- v to abbreviate Ue
habyriawe sengsotai Matsa
Changkui myngsigaariok. That
mountain is just called Matsa
Changkui for short.
sengki n PLANT type of fruit
sengsyp n ANIM type of small fish
sentimityr clf centimetre
sep- v to be stuck Wangai sachido
abongrandai wachi sepni. If I eat
the cob windingly the corn will get
stuck between my teeth.
sep- v to wring, to squeeze out
sepjyrot- v to wring
sepsep- v to masturbate, to wank, to
jack off, to jerk off, to whack off
Ri sepsepni angdo. I will wank.
septembyl n TIME September
serabera n SUBST dirt serabera tak-
to be dirty Ang longpen raityngmi
galaimu serabera takthiriok.
Because my trousers have fallen
off the clothes line they have
become dirty again.
serek n ABSTR surface,
serek n PLACE balcony of a rice field
house
serekmyk n MSRE the length of one
forearm
serem- ~ salam- ~ selem- ~ saram v
to break/tear easily, to be easily
damaged Ie mudupan serema. This
papaya tree breaks easily.
serembut n ANIM type of fish
-set evsp V and do away with, V and
dispose of, V away
si- v to starve Ang pisachi amapara
babapara kanggal dongwana saa
siwa. When I was a child, because
my mother and father and their
families were poor, we starved for
food. khathong si- to feel pity

692
Angdo ue sagyraina khathong
sia. I feel pity for that child.
si- v to peel
si- v to sharpen (a pointy object)
siwil- v to carve, to sharpen a pointy
object
sidikeset n ART CD, compact disc
-siga ~ -sega evsp V in turn,
(alternative suffix)
sigyret n FOOD cigarette
Sijyw n PLACE Siju
sik- v to scratch, to pinch
sikol n ART a chain
siksik- v to scrape, to rub
silongket n PLANT Shillong tree
sima n ABSTR boundary, limit
simen n SUBST cement
singho n ANIM lion
singsip n ANIM type of fish
sinthong- v to cut/break in two pieces
,to cut/break in half
sintongtong- v to cut up in many
pieces
sip- v to smell
sipai n PERS soldier
sipyling ~ spyling n ACT spelling
Atong khuchukmyng
sipyling/spyling rakancha. The
spelling of the Atong language is
not difficult.
siri ~ suri n SUBST snow
sirong n BODY scrotum
sit interj interjection to chase a cat
away
sit- ~ syt- v to take out the shit from
an animals intestines Angdo masu
pipuk sytaidong. Im taking out the
shit from the cows intestines.
Nangdo na pipuk sitbo. Take the
shit out of the fishs intestines.
sithi n FOOD fermented rice from
which chyw is drawn by adding
water
sithi n FOOD fermented rice
skrin ~ sykrin n ART screen
skul n PLACE school
soot- v to kill, to murder
Chigachakchi Dibangkongdang
Umangchalmangsa mongmaaw
sootai matsaaw sootai
mutynwano. At Chigachak
Dibangkogdang Umangchalmang,
killing the tigers and killing the
elephants, lived as the leader.
sore n SUBST mica
sosorot- v to slip Ramchi sosorotok.
I slipped on the road.
soal- ~ sual- v to divide, to share
soalrukai hawa. As for those who
do not have a plot, those mutually
share and clear the land.
sojana n PLANT type of long thin
vegetable
sok n PLANT the new young leaves of
a plant (but not a tree) or vegetable,
a shoot, sprout
sok- v to succeed, to hold out Aia!
ido alsiado kakatyi! Sokchakatyi
angdo noaimyng matsado
jenetene jokaimyng jalangokno.
Jeez! That lazy person bites to my
surprise. I will not be able [to
fight]/ I will not succeed [to fight
with him], the tiger said and
having escaped somehow, he ran
away, it is said. Ytykyimyng
pherudo rypangthiriokno.
Phalthang sokwa dabatdo
tyinyngchi rongchi pyiaimyng
wa khuchengphinai
sakchikaidokno. So then the fox
soaked in the water again, it is said.
Until he could not hold out any
longer, he sat under water as long
as he could bear it, holding on to a
stone and biting his teeth firmly
sokhop n ART cover, sheath
soksek- v to shake something without
picking it up
soksok- v to masturbate, to wank, to
jerk off, to whack off
soldi n BODY a cold Soldi manok. I
have caught a cold.
sombal n TIME Monday

693
somphi n ACT a joke, a riddle
song n PLACE village
song- v to set up post, to dig a hole
and stick something in it so that it
keeps standing up, to raise
Wasung habykungchi songbo.
Dig a hole in the sand and put the
bamboo stick in it. Myngsene
bytwa motchagabaaw
manchagabaaw gudukchagabaaw
chaiaimu, Bandiba kawraw
bytjasaaimyng phalthang
phagongmathangchi phaiai
raiaaidonganote, Bandiba. Bandi
paianggabaaw mykren wathok
songphinai Gyrynggyrang
chaisymaidongano. Having
watched the seven unable men and
the pillar that does not move, Bandi
easily pulled the pillar out and is
carrying it on his shoulder, it is
said, Im telling you!
Gyrynggyrang is watching the
carrying Bandi with eyes raised on
bamboo sticks (i.e. attentively), it is
said.
song- v to elect Songchi nokchi raja
songna angawtara nukariokno.
Angaw bytangaidonga, song
damsachi angaw raja songnino. In
the village the people wanted to
elect a king and they just saw only
me. They are carrying me away;
they will elect me king in a certain
village.
songkhot- v to come out of a narrow
space
Songdu n PLACE the Brahmaputra
river Symsangdo Gohatichigaba
Songduna kylkhala. The Symsang
is smaller than the Brahmaputra in
Guwahati.
songga n PLACE another village
songgamyng morot a person from
another village
songkhot- v to come out of a small
opening, to squeeze out of
songmong n GEO main village
songrai- ~ songre- v to travel
songrat- adj1 to be bent
songre- ~ songrai- v to travel
songsal n ACT society
songsyrek ~ songsarek n PERS
pagan, heathen Dakangmi
pichammi kamdyrangdo ie
Garohils Habeng Harot Chisak
Rangtak Badrijol Kychujol
gumukan songsyrek
dongbutungchido bylongen
hansenga. In ancient times, when
the Habeng-Harot Chisak
Rangtak, Badri areas and the Koch
areas were all pagan, we were very
happy.
sorok n PLACE road, path, way sork
chol ni two ways, roads, paths
sorok- v to re-pound the rice
sorong adj2 straight
sorong- adj1 straight khaw soronga
straight hair
sorot- v celebration in
commemoration of a dead person
one year after this person died Ning
achuaw sorotaidong. We are
celebrating in commemoration of
our dead grandfather.
sosila n PLANT plant of the Arum
family with a pink inflorescence
consisting of an elongate or ovate
spathe (a sheathing bract) which
envelops the pink spadix (a flower
spike with a fleshy axis). This plant
looks remarkably like the
Amorphophallus bulbifer.
-soso evsp V to/on the ground
Nengdugaaimyng
musosoangokno. Having gotten
tired, he sat down on the ground.
sot n ANIM very small type of fly that
comes out in the evening and at
night and cause itchiness
-sot evsp V directlysot- v to spit
Ainachi sagyrai khuti sotjaak.
The child has spat on the mirror
again.
sotmai n ANIM housefly
sotok ~ sot dok num sixty
spit n ABSTR speed
spun n ART spoon

694
spyling ~ sipyling n ACT spelling
Atong khuchukmung
spyling/sipyling rakancha. The
spelling of the Atong language is
not difficult.
ss interj interjection to chase away a
chicken
stel n ABSTR haughtiness
Gethengchi stel panga. He/she is
very haughty. This word probably
comes from English style.
stulkhabar n ART tablecloth
su- v to scold
su n BODY vagina, cunt, pussy, slit,
snatch, twat
su- ~ syw- v to pound, to punch, to
prod, to inject, to crush Ang
khawchi mugaba khyrykaw
sywbone. Crush the lice in my
hair, will you?
subylok- v to mash Beringwa
mynwachido subyloka. When the
food cooked in the wasung is
ready, we mash it.
sugol n PERS vagina (used as a
swearword for women), cunt, bitch
sukherek- v to crash down Jariaimu
sukherekaimu wa khaw sa
baiokno. Wa khaw sa baiokno
mongmaba. Because he was
startled, he crashed down and
broke one tusk, it is said. He broke
one tusk, it is said, the elephant.
sumyn n BODY pubic hair (female)
sunadylep n BODY clitoris
suthat- v to prod, to poke
suut- adj1 damp
sua n ACT profanation
sual- ~ soal- v to divide, to share
soalrukai hawa. As for those who
do not have a plot, those mutually
share and clear the land.
suis n ART switch
suk n comfort Juwna suk dongancha.
We did not sleep enough.
suk- v to be well, to be comfortable, to
enjoy sexually Naa gawi hatchido
sukama? Do you enjoy it when you
fuck a girl?
suk- v to insert, to stitch
suksai n PLANT type of plant of
which traditional umbrellas are
made
sukulbal ~ sykulbal ~ sykubal n
TIME Friday
sukyrung ~ sykurung n ANIM type
of river snail
sul adj 2 next, neighbouring song sul
the next/neighbouring village
sun- v to move, to shift
sun ~ sundul n BODY trunk
sung clf classifier for hollow cylinders
wasung sung tham three bamboo
cylinders
sung n ACT/BODY remembrance,
thought, mind, brain, intelligence,
spirit, life sung ra- to remember,
to think (of, about), to keep in mind
Isolaw sung raai je Kristen
donggabado Isol phiaia
sachengna. Thinking of God,
anyone who is a Christian will pray
to God and start eating.
Gethengchi sung ganang. He is
intelligent.
sung- adj1 short (of time, person,
thing)
sungman- ~ suman - vdat/v to
remember. Some speakers mark the
second argument (the Theme) with
the dative and others with the
accusative. Nokchi ang dyngdang
muchiba, sungmaneta anga
nangna. (Aristo J Momin) When Im
sitting at home alone, I think of
you. Alternatively: Nokchi ang
dyngdang muchiba, sungmaneta
anga nangaw. When Im sitting at
home alone, I think of you.
sunibal n TIME Saturday
suri ~ siri n SUBST snow
-susa evsp V competitively, compete
in V-ing
suset- ~ susut- ~ susyt- v to wash
susu n BODY penis

695
suthul n BODY comb of a rooster
suting n TIME taking pictures, photo
shooting
sutuk- v to put over, to cover, to hide
Mykhang baketchi sutuka. Shes
hiding her face in the bucket.
swich n ART switch
syi n KIN uncle: mothers younger
brother
-syi ~ -si ~ -thai ~ -tyi sfx mirative
suffix, to (my) surprise
syithai- ~ syithyi- ~ syithi- v to hang
Syithai tanarong raityngchi. Its
hanging on the washing line.
syk- v to insert, to be inserted, to
press, to push
syk- vsec to want Jywna
sykaidongkhua. I still want to
sleep.
sykdep- v to press with a finger
syket- v to insert
sykhathang adv disorderly,
carelessly, simply, for nothing, for
free
sykhym- v to moan, to complain, to
feel sorrow, to mourn
syki- v to learn, to teach
sykjyret- v to crush with your hand
sykrin ~ skrin n ART screen
sykrom- v to grasp someone Thot
thyngthot takwachina dabat
sykromaimyng
khanetsigaaidongno. He (Bandi)
grasped her (Sore) and poured the
liquor into her mouth to the last
drop.
sykulbal ~ sykubal ~ sukulbal n
TIME Friday
sykup- v to fold
sykurung n ANIM river snail
sykurung ~ sukyrung n ANIM type
of river snail
syl n SUBST iron
syl- adj1 beautiful, pretty
Atongnawmyl sylate. Atong girls
are pretty!
sylet- v to pour in
sylasyng n ART necklace
syldangkhep n ART big pliers to take
pans off the fire
sylet- v to make beautiful
sylkeng n ART hoop, ring
sylkengkun n ART stick to drive a
hoop
syltyi n SUBST hail, ice
sym adj1 sweet
-sym evsp V and follow, imitate in V-
ing
sym- v to follow Ang nangaw kyn kyn
symni. I will follow you closely.
sym n SUBST salt, medicine
sym- v to build a fence
sym- v to soak, to make wet
symthap- v to taste
symgong n PLANT type of plant of
which the red flowers are edible
and produce a lot of honey which
you can shake out
sympak n PLANT type of tree
symphak n ART type of blanket
symsak- vdat to care for/about, to be
careful about Phalthangthangna
symsakaribo, jalthikaribo! Just care
for yourselves, just run away!
Symsang n PLACE the Simsang river,
also called Someswari
-symsym evsp V continuously
syn n ACT a smell Syn manaidong.
He smells something.
syng- v to ask
synggaba n ACT question
synggera n BODY moustache that
sticks out
synggi n ANIM type of fish
syngsyngkholong n PLACE deep hole
in the ground
synthi- v to suffer, to regret, to repent,
to lament, to moan, to whine
Phepchi synthibutungchi teewe
napit myng sa raiphaknoro.
While he was suffering in the
banyan tree, a barber came by.
sypsak- v to be scratched Ha
khamaimu chak sypsakarok. After
working in the field my arm is
scratched.
-syrang evsp V very much, V
strongly, V completely, wholly V,
V till the end
syrong- v to stretch

696
syrup- v to suck
syryng adv not clearly Angdo
gethengmi balgaba syrynh syryng
nawa. I did not hear clearly what
he said.
syryng n ANIM web (of spider)
syryng- v to stretch out (rope etc.), to
build a bamboo bridge
syt interj interjection to chase away a
cat
syt- ~ sit- v to take out the shit from
an animals intestines Angdo masu
pipuk sytaidong. Im taking out the
shit from the cows intestines.
Nangdo na pipuk sitbo. Take the
shit out of the fishs intestines.
sytha ~ satha n ART umbrella
syw n KIN grandchild
syw- ~ su- v to pound, to crush, to
punch, to prod, to inject Ang
khawchi mugaba khyrykaw
sywbone. Crush the lice in my
hair, will you?
ta interj interjection used in
accusations Ta bongbong! You
liar!
ta prtcl prohibitive particle Naa ta
dykyryngto! Dont make noise! Ta
ie nok dyngdang ham. Dont build
this house alone.
-ta ~ -to sfx emphatic imperative
suffix
tainep tw this morning
taini tw today Getheng tainidarang
raianikhon. He might come
sometime today.
taisa tw a little while ago (today), at a
certain time in the past today
taia- v to pull
taija tw last night Taija walchi
jywwachi jywmangsang
banggirigaba nukwa. Last night at
night when I was sleeping, I saw an
earthquake in my dream.
tairakrak adv not too big and not too
small Amakdo getheng
baisigathanggabamyng kynaw
rongpatal syltengbigabachi
kepreprep bamai hynbutungchi
pantong myk sa donggabasang
tairakrak takgabasang tep tep tep
tep tokaidoknoa. The monkey hit
his friend who was lying flat on his
belly on the very beautiful flat
stone with a stick of one myk,
which was made not too long not
too short, tap, tap, tap, tap on the
back.
taiyr n ART tire taiyr sam/rong/goi
tham three tires
tak- vB to do, to make, to pretend, to
act like Ie khamaw krymkraw takna
nangni. We will have to do this
work together. Magachake: Hai
baisiga biskut sakhawna
noaidongano. Hyt mancha
nangba atong budi nowano
pherue. Ang denggu takni naa.
Naa paiai jalbone noaidongano
magachakan. The deer said:
Come on, friend, I want to steel
those biscuits! What?! You
cant! What are you thinking?!
said the fox, it is said. I will do
some extortion. You carry the
biscuits and run away, ok? said
the deer, it is said. Tedo
magachakan khora takaidongano.
Now the deer is pretending to be
lame, it is said. Alsia rajae: Ie
gari biskyn? Hyw naa ranae
synge syngchagaba taknae synge
syngchagaba. The lazy king
asked: How much for this bullock
cart? Hey! As far as your buying
is concerned, you are just acting
like someone who asks but who is
not really interested.
Chigachakchi Dibangkongdang
Umangchalmangsa, mongmaaw
sootai matsaaw sootai
mutynwano. Uchi song damni
takwano. In Chigachak,
Dibangkongdang and
Umangchalmang, having killed the
elephants and tigers, lived as the
leaders, it is said. There they built
two villages, it is said. Ytykyimyng,
ha! wenni rypwachian miniksuru
takokno sylokno magachakmi

697
nyndo. So then, ha! when he had
bathed twice, his fur was flat, it is
said, it was beautiful, it is said, the
deers fur. Angdo masuba
pangphachaaimyng niwa,
jamaiawan tanawachym noai
golphoa takaidongano. Uchi
myngsagaba: Angdo jamaiawan
tanasyrangok mangthamawan
noai bala takaidongano. As for
me, I already had only a few cows,
and now there is not one left; I
slaughtered them all, he is telling,
it is said. Then another man says:
I slaughtered all three of them it
is said. Aia! Udo magachakdo
it run away fast, it is said, the deer.
takal ~ dakal n PERS witch
takap- v to stick
takbewal n ACT tradition
takruk- v to have an orgasm (of a
woman)
takruk- v to fight Noksang rainaba
phaphinchaaidok. Jebadong anga
takruksyrangarinaka. Matsami
chaphungaw wangjoloknoaro,
alsia rajae. I dont dare to go
home. Anyway, I will just fight to
the end. He bit the tiger on the
thigh, it is said, the lazy king.
Rongdyngmi oltoe, dakang somai
Jaksongram matsa nok phandaimi
matsamu Rongdyng maharimu
takrukwanowa. As for the meaning
of Rongdyng, in times long ago,
the tigers of Jaksonrams tigers
bachelors house fought with the
Rongdyng clan, it is said.
taksak- v to help
taksakgaba n ACT help
tala n ART a lock Tala thekbo. Lock
the lock.
tam- v to wait, to stop
tam- v to trim, to prune
tam- v to play an instrument tama
toka to play an instrument
tamo ~ tamaw interj Wait!
tan- v to put, to stop Thengthon
tangkaaw raaimyng, uaw
kerengaw palyngchi gopai
tansigaakno. Thengthon took the
money and left those bones buried
in the jungle. Balai tanangok. I
have already said it. Angdo ytykyi
balaimyng tanarinaka. As for me,
having spoken like this, I will just
stop now.
tan- v to cut, to cut up, to slay, to
slaughter Galaimuna kynsangdo
phylgymaw uan rykjolaimuna
kukuri bykotaimuna tokyrengaw
tanthongokno. Tokyreng
tanthongaimungna kynsangdo
dykymawdo jytsetetokno,
ytykyimungna pipukaw
tanpyrakokno. Uchie phalthangmi
dadadyrangaw nukokno,
phawjonggarangaw. After falling
down, having run quickly towards
the eagle, having drawn his knife,
he cut the neck off, it is said. After
decapitating the bird by cutting its
neck, he pushed the head away, it
is said. So then he cut the belly
open, it is said. Then he saw his
own brothers, his elder brothers, it
is said. Una phalthangmyng
masuthangthangaw tanaimyng
khaiangokno. Ramchi golphoanga
takaidongano: Angdo masuba
pangphachaaimyng niwa,
jamaiawan tanawachym noai
golphoa takaidongano. Then,
having slaughtered their own cows,
they carried them to the market, it
is said. On the way, they are taking,
it is said. I dont have many cows,
but I would have slaughtered all of
them anyway, he said while
chatting, it is said.
tanchekchek- v to cut into small
pieces
tancholeng- v to cut a piece out of
something

698
tanpyrak- v to cut open Tokyreng
tanthongaimungna kynsangdo
dykymawdo jytsetetokno,
ytykyimungna pipukaw
tanpyrakokno. Uchie phalthangmi
dadadyrangaw nukokno After
having decapitated (the eagle), he
pushed the head out of the way, it
is said. So then he cut its belly
open, it is said. Then he saw his
own elder brothers, it is said.
tanthong- v to decapitate, cut off the
head; to cut off Galaimuna
kynsangdo phylgymaw uan
rykjolaimuna kukuri bykotaimuna
tokyrengaw tanthongokno. After
the eagle had fallen to the ground,
he ran and unsheathed his knife and
cut off its head, it is said.
tandap- v to be on top, to cover
tang clf classifier for koktang baskets
Ytykyimyng tedo gethengthengdo
na khynaimyng
baisigathangmaran rukpekba
tangsa, amakba tang
sakhaiaknowa. So then, now, as
for them, having collected the fish,
the frog and the monkey carried
one basket each, it is said.
tangka n ART money tangka poisa
money tangka phel sa one coin
tangka kung sa one banknote
tangka rong chykhyw nine rupees
tankynyng- v to cut up in many pieces
tannet n ART measure basket
tanset- v to abandon, to leave behind
tap n time, turn tap sa once tap ni
twice tap tham three times
tap- v to hit, to beat-up
-tara encl.phr only, exclusively
tarai tw this year
tarak- v quick, fast. swift Hare! Ie
atakgaba magachake? Tedo
lengla lengla jaltherianoa. Una
ryktheriokno banggale. Tharapna
guduk takwachiba tarakai
jalariano magachake. Huh?!
What kind of a deer is that? Now
it is running as if it is lame again, it
is said. Then the Bengal chases
after it again, it is said. Whenever
he almost catches up with it, the
deer just runs away quickly, it is
said. Anga noksang tarakai raina
nangaidong. I need to go home
quickly. Tarakbo naa! Hurry up!
tarang n PLACE layer Angdo ie
rajami khemna jywthumaidonga,
damana. Ramramchagaba
kyryngwa ido. Hanyng tarang
chinina imyng kyryngwado rajami
dama. I am lying here guarding
the royal drum. It has an unusual
sound, this thing. The sound of it
reaches welve layers inside the
earth.
tarik n TIME date
tas n ACT cards (the game) Tas keleni
ningdo. We are going to play cards.
-tat evsp V compulsorily
tat- v to drive in (as with a nail in
wood)
-taw evsp V upward
taw- v to go up, to ascend
-taw ~ -aw encl.phr accusative
enclitic
taw n ANIM bird, chicken
taw sagyrai n ANIM chick
tawdimai n BODY tail feathers
tawgurung n nest
tawgylyk n ANIM type of jungle bird
tawkarang n BODY birds wing
tawkhasi n ANIM capon, castrated
rooster/cock
tawkurung n PLACE the nest of a
chicken
tawmyn n BODY fluff, body
feathers
tawnok n ART chicken cove, coop
tawpachi n ANIM swallow (Family
of Hirundinidae)
tawpak n ANIM bat, butterfly, moth
tawpakhal n GEO cave in Siju
tawpaksa n ANIM moth
tawpaktyi n ANIM caterpillar
tawpalyng n ANIM jungle fowl
tawpynchyrep n ANIM type of small
bird with green wings and tail
feathers, a white chest and
brownish read head and beak,

699
approximately seven centimetres
from head to tail
tawreksyrup n ANIM banana bird, if
translated literally its name is
banana tree sucking bird
tawsagyrai n ANIM chick
tawthup n ART nest
tawti ~ tawtyi n ANIM egg
tawa n ART frying pan
tawel n ART towel
-te sfx declarative suffix
teen tw later today
teew tw now teew mangmang just
now
tebyl n ART table
tek- v to tie
telephon n ART telephone
teng onom sound of falling money
-teng evsp still too V
tengchypchyp- v to shine, to glitter
Ytykyimyng walchi raiaphyinokno.
Raiaphinaisa beanbebe
phalthangmyng nokaw gethengdo
masu dimyng phirinaimyng ue
sona bichamchymaw nok
ryphiokno. Nok ryphiwamyng
kynsangdo teew getheng nokawan
alaga morotdyrangdo
tengchypchypai nukariokno. So
then he came back at night, it is
said. Having come back, having
mixed it with cow dung, he
plastered his house with the golden
flakes, it is said. After plastering
his house now, other people saw
how his house was shiny, it is said.
-tengteng evsp still much too V
tenki n BODY tank
teraka tw last year
tet- v to pour out
thagythyng n PLANT type of
vegetable
thalet- vdat to explain Ang nangna
Atong khuchuk saina thaletni. I
will explain to you how to write the
Atong language.
thamakhu n PLANT tobacco
thabisi n ART amulet, antidote
thagal- v to lose an object Ang
chabiaw bichiba thagalok. I have
lost my keys somewhere.
-thai ~ -tyi ~ -syi ~ -si sfx mirative
suffix, to (my) surprise
thai clf classifier for receptacles
boiom thai sa one jug dipot thai
sa one teapot khap thai sa one cup
Getheng boiom thai ni baiok. He
broke two jugs
thai n PLANT/FOOD fruit
thaimathaigundai n PLANT type of
bright orange fruit that grows in
creepers high in the jungle trees in
the rainy season. The round fruits
are about seven centimetres in
diameter. The outside consists of a
thick, uneven leathery rind while
inside there are about eight sweet
orange carpels each containing a
smooth stone.
thaisymphak n PLANT type of plant
thajyri- v to make trouble
thal- adj1 clear, explicit
thali n ART plate (for eating) or its
volume, plateful
tham num three
-tham evsp barely V
thama n ACT divination, thama chai-
to see the future, to practice
divination Kamalchi thama chaia.
At a priests house, divination is
practiced.
-thamak evsp V barely, V excessively
thamat n PLANT green plant that
grows in the jungle and of which
the side leaves, the young leaves,
and the fruits cause irritation when
touched
thamylang n PLANT type of
vegetable
thankhoana- v to gut lengthwise,
longitudinally
thang v to fall down on
-thang encl.p own ang nokthang my
own house
thang- v to throw away with great
force Matsa kherengwachido
wachungbyryidarangdo
thangaaidonga Bandiba. Bandi

700
kherengwachido wachu byryi
wawa wawa thangasigaaidoknote.
When the tiger makes a great
effort, he throws Bandi one
bamboo length away. When Bandi
makes a great effort, he throws the
tiger whoooosh! four bamboo
lengths away, it is said, Im telling
you!
thangchichat- v to drain
thangguduk adv suddenly
thangphytphyt v to splash Nalam
gudukwachie teewdo tyi
thangpytpytaimyng jyksaiaiawan
Nawengawmu Kumiribaawma
khamoknowa. When the nalam
(type of fish) wiggled, water
splashed on the married couple
Naweng and Kumiri and burned
them, it is said.
thangtaw- v to squirt out
thanthong- adj1 blunt (of pointed
things)
thanyng n BODY brain
thap onom the sound of something
hitting thok, thap! hit, slap!
thap- v to beat, to beat up, to destroy
Phangnan ning nokaw
thaparonga noai
balokno.Ytykyimyng myngtham
reengokno mongma mathaiaw
thapna. He always destroys out
houses he said, it is said. So then
the three of them went on their
way, it is said, to beat up the
elephant.
thaphu- v to blister
thapthap adv quickly
thapyra n SUBST ashes
tharai- v to change, to exchange, to
swap Getheng chola gachawana
nygylmyngaw tharaiok. He
changed the bad shirt for a new one
from the market.
tharap- v to catch up with, to be on
time Aia! Udo magachakdo
it ran away fast, it is said, the deer.
thari- v to prepare, to arrange, to
repair
thasa- v to wake somebody up
Thasabo uaw. Wake him up!
-that evsp V excessively
thatthongthong- v to tear to pieces
thaw onom bang! (sound of a gun
firing)
thaw- adj1 tasty
thawjyw n PLANT type of fruit
thawal n BODY scab
themet- v to fold
themyt n PLANT cucumber
thebajaw- v to tickle Nang angau
thebajauwa, ang bejawok. You
tickled me and I feel tickled.
thek- v to block off, to lock Tala
thekbo. Lock the lock.
thek- v to insert Waiyr karenchi
thekbo. Put the wire into the
electric socket.
-thel evsp surely V Ichi rongkhalchi
khen ganangthelnaba ganang.
Here in the spaces under the stones
there are river crabs for sure.
thel- v to tie
them onom sound of a gunshot, pow
Myngsagado them! kawokno. The
first one shot, pow! it is said.
themtaw- v to roll up
theng clf classifier for pieces of meat
thep clf classifier for heaps and small
packets
thet- v to pull out, to pull Ang
phakwalmyn theta. I pull out the
hair in my armpit.
thetchot- v to break by pulling Ning
kara thetchotok. We broke our
rope.
thik adv exactly, well cooked, well
done Kawbutungchi thik
thokyrengaw manokno. When he
shot [the giant eagle] he got it
exactly in the neck. Jaryt sawai,
mantaw sawai, maichengmung
nalammung thiksa berengai saa.
We roast the chilli pepper, we roast

701
the brinjal and cook it until it is
well done with maicheng (a type
of leafy green) and nalam (a type
of fish) in a bamboo cylinder and
eat it.
thik dong- ~ dong- v to be correct
Morot chanchichypai thik
dongokodo, uchian rajaan uaw ajot
nosawnaka. Suppose someone gets
it right, then the king will tell him
ajot.
thik kha- v to fix a date and time
Takrukna san somai thik
khawachym. They supposedly
fixed a time and a day to fight.
thikthak adv exactly, precise,
precisely Kynsange nygyltyi ni
reengwachi thikthak jahas
kanachina dongangok. Later,
when she had been going for two
weeks, she arrived exactly at the
ship and the harbour. Thikthak
phangnado chykhyw bajichi
saaidonga. He always eats exactly
at nine oclock. Thikthak
kawoknotyi ue sagyraie.
Kawbutungchi thik thokyrengaw
manokno. That child shot very
precisely. When he shot [the giant
eagle] he got it exactly in the neck.
thimini- v to make someone smile
thimini- v to make someone smile
thingthingthing onom klang klang!
sound of something falling made of
metal
thintaw- v to climb up
thiri n ART bow (of bow and arrow)
thiri n ART arrow (of bow and arrow)
-thiri evsp V again
thirikun n ART arrow (of bow and
arrow)
thiriphong n ART part of an elephant
trap
-thirithiri evsp V again and again
tho n FOOD mustard oil
tho- v to compare
thoma n ACT group
thogi- v to betray, to cheat (on), to
deceive Gawi angaw thogiok. The
girl has betrayed me/cheated on
me.
thojekjek- v to shake a fixed object Ie
panaw thojekjekchido thai
galkhalni. If you shake this tree,
fruit will fall down.
-thok evsp V together, everybody
(S/O quantifier)
thokbyrang adj2 multicoloured,
many coloured
thokbyrym adj2 multicoloured, many
coloured
thokthok adv precisely
thol- v to lie, to tell lies Anga
Ketketa Bura nogaawan
tyngkhucha noaimyng Ketketa
Burae phalthangawan pheruna
tholokno. I dont know this so
called Ketketa Bura yet, lied
Ketketa Bura about himself to the
fox, it is said.
tholam n PERS liar
thom clf classifier for things in heaps
or piles jyw thom sa a pile of
flattened bamboo used to make
mats
thom- v to come together Song gumuk
thomaimung haba haryn
harynaw sowalni. The whole
village comes together and they
will divide the haba plot by plot.
thom- v to make a heap Rong
thomaidonga. Hes making a heap
of stones.
thong clf classifier for cylindrical
objects betyri thong byryi four
batteries
thong n SHAPE half which is the
result of a cross section or a cut
across the width or a crosscut
-thong evsp V in half
thongthong adv straight
thop onom hitting sound: thok! thunk!
thop- v to gang up on Watnabai iaw
alagaaw! noaimyng
rykathokaidongano, Bandiaw
thopna. Dont let this stranger
go! they said and they were
chasing him, they wanted to gang
up on Bandi.

702
thorok- v to jump (down from / out
of) Alsia raja phepmyng
thorokokno. The lazy king jumped
out of the banyan tree, it is said.
thorokangokno. So then, having put
the biscuits by the side of the
water, the deer splash! jumped into
the water, it is said.
thorom n ACT religion Ning
songsyrekdo ning atongdo
dakangdo mamyng thoromaw
niwami somaichido waiaw mania.
We pagans, we the Atong, in the
past, in times when there was no
religion, we worshipped spirits.
thot- v to hit, to bump into something
or against something Cha ronchi
thotwa. I hit my foot on a stone.
thot thyngthot adv to the last drop
Thot thyngthot takwachina dabat
sykromaimyng
khanetsigaaidongno. He (Bandi)
grasped her (Sore) and poured the
liquor into her mouth to the last
drop.
thothak n QUANT/MSRE a drop,
classifier for drops myktyi thothak
ni two tears/two teardrops
mykrensam thothak ni two drops of
eye medicine.
thotphyret- v to smash by hitting
against or on something
thuk- v to enclose with a fence nol
thuka to fence
thuk- v to wicker a bamboo mat
Damdyl thukaimu nok hama.
Having made the bamboo mats,
they build the house.
-thum evsp V on behalf, instead of
someone else
thunuk- v to show
thup n PLACE nest Tawreksyrup
mang sa gethengmyng thup
phangnan mongma phaiai
sarongwana, mongma mathaiaw
thokna reengaidongano. Because
the nest of a banana bird always
gets broken and eaten by an
elephant, it is on its way to beat the
bachelor elephant up, it is said.
thup onom beating sound: thunk!,
slap!
thup- v to nest, to be thick (of fog or
mist) Teedo ue mongmaai rekchi
thupai thupai mugabaaw
phangnan phaiai phaai saronga.
Now this elephant always breaks
and eats the place in which I nest in
the banana tree. Guri thupa. The
fog is thick.
thut ~ thun clf classifier for big
spherical things, stones, bricks,
rocks, heads, hills, mountains and
bars of soap rong thut tham three
rocks habyri thut sene seven hills,
mountains sabun thut sa one bar of
soap, dykym thut sa one head
thyyk- v to have the hiccups
thyi- v to die
thyi n BODY blood
thyikhop n ART dried fruit in which
water is stored for consumption
thyiwami n ACT death
thyk n ART pan for cooking rice
thyk- v to be fixed sideways
-thyl evsp V and avoid, V ahead
thyl postp up to, until (spatial) Ie
chamasangmi wai khurutchido ue
hyisangmiaw Banggladesmi thyl
Kongosmi jaria
hagyrsakgumukawan myngani.
When he summons the downstream
spirit, that [priest] will call upon
the influence of all those far away
[places] up till Bangladesh [and]
the area of Kongos, all of them.
thyl- v to go very far
Gathyngaimuna thylangok.
Because [I] kicked [it], [it] went
very far.
thylapak ~ thylampak n BODY
tongue
-thylong evsp V nicely
thym- v to lay in ambush
thym- v to take revenge
thymbylong adj2 to have a hole in it,
damaged (of roads, bridges and
wooden planks)

703
thymyn v to ripen Panchung
thymytetbo. Keep the jackfruit so
that it can ripen.
thymyn- v to ripen
thymyt- v to put out (fire), to switch
off, to extinguish
-thyng evsp V so much
thyng- v to kick
thyngel- v to tilt
thyngpyret- v to kick
-thyngthyng evsp V so much, V
continuously
thyp- v to throw (sidearm) Wakaw
paiaimungna amake welan
welang pankambaichi
dungangaimungna sasigaaknotyi
dyngdang. Dyngdang sasigaakno.
Sala! Naa angna kholaw
hynatemo. nookno khusume.
Amake: Tambone hynni
nangnaba noaimuna Ha
kerengaw sabo! noai
thypratetokno. Having carried the
pig, the monkey quickly climbed to
the top of a tree and ate alone.
Damn you! You can at least give
me the skin! said the turtle. Wait,
OK! Ill give you something too,
the monkey said and then: Take
this, eat the bones! he said and
threw them down. Uchie getheng
nokhapchina tangkaaw thypai
thypai khiaidongano, Rong sa,
rong ni, rong tham noaimyng.
Then he goes home and is counting
the money throwing it on the
ground saying: One rupee, two
rupees, three rupees. Dada, anga
nang jongsaba nookono. Sala
burbok sagyrai naa ningaw hala
khagabaai noaimungna bunduk
raasetetaimungna uaw sagyraiaw
gadakchichiokno.
Gadakchichiaimuna
singsingkholongsang thypsetyi
tanangokno. Brothers, brothers, I
am your younger brother! he said,
it is said. Damn you stupid child!
You have woken us up! they said
and, having thrown away their
guns, they cut the child up in
pieces, it is said. Having cut him
into pieces, they threw him in a
deep hole in the ground, it is said.
thyrgyryw v to shake something large
and unmovable
thyw- adj1 deep
thywkhong adj2 globular, protruding,
bulging
Tibet n PLACE Tibet
tibi n ART television
tiiititi interj interjection to call a
chicken
tin n ART corrugated iron sheet used
to make roofs tin kap sa one sheet
of corrugated iron
tintyrin n PLANT tamarind
tiup n ART tube
-to ~ -ta sfx emphatic imperative
suffix
totheng n PERS the little forest spirit
of wealth
toilet ~ toilyt n PLACE toilet
tok- v to beat, to beat up, to play an
instrument Tawreksyrup mang sa
gethengmyng thup phangnan
mongma phaiai sarongwana,
mongma mathaiaw tokna
reengaidongano. Because the nest
of a banana bird always gets
broken and eaten by an elephant, it
is on its way to beat the bachelor
elephant up, it is said.
Ang phulistau khapeta bajuaw
tokwana I am angry with the police
because they beat up my friend.
Uaw Dorenggo Wadachongawdo
achu ambido tawnaan Dorenggo
Wadachong jatram saphairam
noaimu samaw chaaw itykyi
tokano. As for Dorenggo
Wadachong, in order to go up on
Dorenggo Wadachong, our
ancestors beat so called jatram and
saphairam medicinal plants with
their feet like this, it is said. Ue
atakwa jong? Madam tokwa.
Atongmaina? Ytykyian
tokariwa. What happened there,
son? My (female) teacher has hit

704
me. Why? Like that she just hit
me.
tokdepdep- v to crush, to grind
Sambanggyri akaiokno,
tokdepdepaimu phaatokno. He
plucked sambanggyri, crushed it
and put it on the wound, it is said.
tokgepgep- v to beat to pulp
tokhynyng- v to smash into pieces
tokkhyphu ~ tokybu n BODY throat,
area just under the chin
tokorot n BODY throat, glottal area
tokphyrong- v to take a powdered
substance in the palm of one hand
and softly tap on it with the other
hand
tokpyret- v to crush by hitting
tokset- v to cough, to have a cold
tokset- v to pull loose
tokta n SUBST type of wood
toktai- v to hang oneself
tokthining ~ tokthynyng n BODY
neck
tokthong- v to smash in half
tokthynyng ~ tokthining n BODY
neck
toktokylek n PLANT type of flower
tokyphu n BODY gullet, throat
tokyreng n BODY neck
tong- v to fuck
tota n ART plank tota kap sa one
plank tota khaw sa one plank
totakhaw n ART a plank tota khaw
sa one plank
totyp adj2 bent Bandi muetwachian
dakhamba hachina chaksi ni
dongna guduk totyp totyp
takaidonganote. When Bandi sits
on the dakham it bent almost
completely to but two fingers from
the ground, it is said.
tu- ~ ty- v to feed (by putting food
or drink into the mouth Sagyraina
mai tuwa. I fed the child rice.
Bylsi senemi chywaw pityi jyngjang
phinggabaw botol
chaksijywsykyngabaaw
kanetaidongano. Dada choisa
kawakbone noaimyng
hynetaidongano. Atong churu
ryngnaka? noaimu Botolgumuk
tyetsyrangbo. They are pouring
seven year old wine from a fully
filled bottle as small as a thumb, it
is said. Elder brother, open your
mouth a little they said and then
What little will I drink? he said
Feed me the whole bottle.
tuk- v overgrown, dense (of
vegetation) Ram tuka. The road is
overgrown. Palyng tuka. The
jungle is dense.
tum clf classifier for places and
packets Hap tumbyisyk? How
many places?
tun- ~ tyn- v to lead, to guide to lead,
to guide Naa ang masu mang
raja saaw tynangsegabone. You
lead my hundred cows away, OK?
tung clf classifier for objects like
bridges dolong tung ni two bridges
tung- adj1 hot, warm
tupi n ART cap, heat
ty- ~ tu- v see tu-
tyi n SUBST water; fruit juice
tyi n ANIM egg
tyi- v to lay an egg
tyibal n GEO wave
tyibasal n GEO whirlpool
tyibek n ART traditional bottle used to
drink water out of and made of a
dried vegetable also called tyibek
tyichabakram n GEO waterfall,
cascade
tyichang n PLACE island
tyigat n place place in a river or at the
end of a water pipe where the
people get drinking water, take a
bath and wash their clothes and
dishes.
tyigum n ART water container made
of metal and shaped like a big vase
used to store water in the kitchen.
Its place in the house is in the
tyinok
tyikaran ~ tyikharan ~ tyikaran
adj2 thirsty
tyikhal n GEO river thyikhal chol ni
two rivers
tyimong n GEO main river

705
tyimuk n GEO spring, source
tyimyk n GEO source
tyinala n PLANT algae
tyinok n ART place in the kitchen
where the water pots (tyigum) and
other utensils like plates, cups and
glasses are stored.
tyiphek n GEO tributary river, the
smaller one of two rivers that flow
together
tyisam n PLACE river bank, edge of
the water Tyi gagaba wari
thywgaba tyisamchi hap
sylgabachi myng ni
baisigathangmaran chang
tyrywchengnaka? noaidongano.
By the side of a deep wari with
good water on a beautiful spot the
two friends were arguing about
who would take a bath first, it is
said.
tyisi- v to be wet Magachakmi myndo
tyisiwachian miniksuru
takjolarianoro. When the deers fur
is wet, it just quickly gets flat-
haired, it is said.
tyisurung n GEO rainwater that
streams over the ground
tyithai n ART water scoop made of a
hollow, dried gourd.
tyksyl n ART pan for cooking rice
-tykyi encl.phr perlative/similfactive
enclitic
tykyw n ART water pot
tym clf classifier for fields haba tym
ni two dry rice and vegetable fields
on the slope of a mountain
-tym sfx personal pronoun plural
suffix used to form the second
person plural exclusive personal
pronoun from the second person
singular and the third person plural
from the distal demonstrative:
nangtym you (plural exclusive)
utym they
-tyn evsp lead/bring to V, be the
leader of the action, Keletynbo!
Play with the others as the leader.
tyn- ~ tun- v to lead, to guide Naa
ang masu mang raja saaw
tynangsegabone. You lead my
hundred cows away, OK?
tyng clf classifier for long thin objects
like ropes, chains, hairs etc. kara
tyng sa one rope
tyng- v to know, to recognise
Ketketa Bura? Anga Ketketa Bura
nogaawan tyngkhucha noaimyng
Ketketa Burae phalthangawan
pheruna tholokno. Ketketa Bura?
I dont know this so called Ketketa
Bura yet, lied Ketketa Bura about
himself to the fox, it is said. Ang ie
khata dakangdo tyngchachym,
teewdo nemen tyngok. I did not
know this word before but now I
know it well. Nang baletgaba
morote atongtykyi angawe
tyngsawnaka? How will the person
you talk about certainly recognise
me?
tyngcheng- v to know first, to
discover
tyngen adv very Ue raja kam
khanaba tyngen haratachym. That
king was supposedly very reluctant
to do work.
tyngetwami ~ tyngetwamyng n ART
announcement, notice
tyngkarang adv in one go
-tyngtang evsp V all over the place
tyngtet- v to hang someone Tyngtet
khaai tan! Hang him up!/Kill him
by hanging him!
tyngwami n ACT knowledge,
understanding
typ- v to throw down Sala burbok
sagyrai naa ningaw
halakhagabaai noaimungna
bunduk raasetetaimungna uaw
sagyraiaw gadakchichiokno.
Gadakchichiaimuna
singsingkholongsang typsetyi
tanangokno, typsetyi tanangokno.
Damn you stupid child who
disturbed us! they said and they
took out their guns and cut the
child into pieces, it is said. Having
cut him up, they threw him into a

706
deep hole in the ground and left
him there, it is said.
tyret- v to bathe someone else
tyru-~ tyiru- ~ tyiryw- v to bathe, to
take a bath, to wash oneself
tyt- v to pour Ue tyigummi tyi tytbo
depotchi. Pour water from that
tyigum into the teapot.
uchi disccon then
uchiba disccon but then Una
myngsagaba sabanthai
sabanthai myngsagaba
bychymokno, uchiba
patangphaariok,
dangangphaariokno. Then one
son pulled the other out [from the
water], it is said, but then they just
crossed and they all just drowned,
it is said.
uching ~ uching ~ ukching n ANIM
leech
ue ~ u- dem that, there, distal
demonstrative
umi ~ umido ~ umisa ~ umyng ~
umung ~ umyngdo ~ umyngsa
disccon then Naa wai
chunggabaaw nukoknoai
chanchibo, masu ranaka, purun
ranaka, taw rana nangni, wak
rana nangni, unado. Umi chywba
symna nangni, ue kamalna.
Suppose you see a big spirit. Youll
get a cow youll get a goat, youll
need to get a chicken, youll need
to get a pig, for him. Then youll
also need to brew some liquor for
that priest. Masu nanga, wak
nanga, taw nanga, chyw nanga
waikhurutna. Umido uaw kamal
sandini. You need a cow, you need
a pig, you need a chicken, you need
liquor to perform an incantation.
Then you will search for that priest.
Phasgaba hahawchenga.
Umungsa ha hawaimungsa
wacham tana. First we clear the
jungle. Then, having cleared the
jungle, we cut the old rice stalks.
umigymynchi ~ umynggymynchi
disscon for that reason, therefore,
because of that
umyng ~ umung~ umyngdo ~
umyngsa ~ umi ~ umido ~ umisa
disccon then Phasgaba
the old rice stalks. Naa wai
chunggabaaw nukoknoai
chanchibo, masu ranaka, purun
ranaka, taw rana nangni, wak
rana nangni, unado. Umi chywba
symna nangni, ue kamalna.
Suppose you see a big spirit. Youll
get a cow youll get a goat, youll
need to get a chicken, youll need
to get a pig, for him. Then youll
also need to brew some liquor for
that priest. Masu nanga, wak
nanga, taw nanga, chyw nanga
waikhurutna. Umido uaw kamal
sandini. You need a cow, you need
a pig, you need a chicken, you need
liquor to perform an incantation.
Then you will search for that priest.
una disccon therefore, then Jetakai
patangchiba rung
bytrongrengangariano,
sangkyningan. Unasa rung chawna
dakangan ytykyi rung dykymaw
gatyngaimuna kha Dawa!, kha
Dawa! noaimusa rung chawaimu
patronganoro. Whatever you do
whenever you cross, the boat will
spin. That is the water dragon.
Therefore, before you cross by
boat, because you stamp on the
head of the boat saying Kha
Dawa! Kha Dawa! and then
having gone by boat, you usually
cross, it is said. Only then does the
water dragon not get you, it is said.
utym ppron they, third person plural
personal pronoun
wa n BODY tooth, tusk (of elephant)
wa khaw ni two teeth, two tusks
-wa sfx factitive suffix

707
wa- vS1 to rain Rang waaidok. Its
raining. Literally: Rain is raining.
This verb can only take the noun
ra as its subject.
wa.tana n ART part of an elephant
trap
wa n PLANT bamboo wa dot sa one
culm of bamboo wakhaw sa one
long half of a bamboo morot wa
sagaba a strong and tough person
wa n KIN father
wacham n PLANT old rice stalk
which is left over after harvesting
the rice
wachu n MSRE one bamboolength
wachun n MSRE the length of one
bamboo stick
wachurek n ABSTR capacity
wachyrik- vdat to be startled Gari
horn khawanasa wachyrikok,
gethenge. Because the car blew its
horn he was startled.
wada n PLANT type of bamboo of
which each culm comes out of the
ground individually instead of in a
bush
wadokolong n ART water pipe made
of bamboo
wagat n ART bamboo shoulder yoke
wagatram n BODY shoulder,
literally: the place where you put
the bamboo shoulder yoke
wagydok n ART water pipe made of
bamboo
wajong n PLANT type of bamboo
wakai n PLANT type of big bamboo
wakhal n ANIM grasshopper-like
insect
wakhaw n PLANT one long half of a
bamboo split lengthwise wakhaw
sa one long half of a bamboo
wakhyntha n PLANT type of
bamboo
waphuk n ART white half of a strip
of bamboo used to make rope
wari ~ waryi n PERS child who lost
his father
warung n PLANT young bamboo
wasung n ART bamboo cylinder used
as container and used to cook
bering in wasung sung ni two
bamboo cylinders
wasyl n ART green half of a strip of
bamboo used to make rope
wathok n PLANT hollow bamboo
stick Myngsene bytwa
motchagabaaw manchagabaaw
gudukchagabaaw chaiaimu,
Bandiba kawraw bytjasaaimyng
phalthang phagongmathangchi
phaiai raiaaidonganote, Bandiba.
Bandi paianggabaaw mykren
wathok songphinai
Gyrynggyrang chaisymaidongano.
Having watched the seven unable
men and the pillar that does not
move, Bandi easily pulled the pillar
out and is carrying it on his
shoulder, it is said, Im telling you!
Gyrynggyrang is watching the
carrying Bandi with eyes raised on
bamboo sticks (i.e. attentively), it is
said.
wathyrai n PLANT type of bamboo
watyng n ART bamboo strip used to
make baskets, and other woven
utensils as well as rope. Nangtym
angaw tyichi typratwaba nemariok
aro koksep chungkhuna
nangachym. Watyng tyngphekna
masu mangphek hynwa, gumuk-
gamak angna masu mang raja sa
hynetwa angnado. You threw me
into the water and that was good,
and I should have had a bigger
koksep. For every bamboo strip
they gave me a cow and in all they
gave me one hundred cows.
wach n ART watch
wachyw n BODY incisors (the four
front teeth used for biting)
wadi ~ wakhi n BODY plaque
wagydok n ART bamboo water pipe
wagyleng n PERS person who is
missing one or more teeth
wai n PERS spirit wai khurut- to
perform an incantation, so summon
a spirit
wai- v to return, to go/come back
Rangsando saniarokno, sikharba

708
khachypanchakno, ytykthyngai
somai jamchypaimuna jyksang
sasang waiangokno. The sun was
stetting. The hunting had failed and
having done all this, having wasted
time, he went back to his family, it
is said.
wai- v to plough Bydyi myng sa
haphal khamaidongano,
masusang haphal waiaidongano.
An old man is working in the field,
it is said. He is ploughing the field
with a cow, it is said.
wai- v to scoop (of liquid) Angna
chyw gylaschi waibo. Scoop some
liquor into the glass for me.
Tyikhalmi tyi waiaimu getheng
ryngok. He scooped out some water
from the river and drank it.
waiseng n ART very big knife
traditionally used to kill tigers and
men waiseng mang sa one
waiseng
waiphin n ACT return Waisa waiphin
lak sa nanga. To go and come back
you need one hundred thousand
rupees.
waiphin- v to go back, to return
Nangmi sabanthai waiphinaakte.
Your son has returned.
waisa n ACT the going (to
somewhere) Waisa waiphin lak sa
nanga. To go and come back you
need one hundred thousand rupees.
waiset- v to drain a little bit of water,
to scoop out water Chamussang
waisetbo. Scoop it out with a
spoon.
waiyr n ART wire
wak n ANIM pig, pork
wakam n BODY molar (tooth)
wakeng n ART axe
wakhi ~ wadi n BODY plaque
wakhol ~ wakholong n PERS person
who is missing one or more teeth
wakhu n ART chopper
waknok n ANIM domestic pig
waknol n ART pigsty
wakpalyng n ANIM wild pig
wal n TIME night
wal- vS1 to be night Walni. It will be
night. San walok. It has become
night. (Literally: The day has
become night.)
wal n GEO fire, torch Nokhapalchi
wal chakbo. Light a fire outside
the house. Nanang walchi wal
netaimu khen rawna reengni. We
will light a torch and go catch river
crabs at night.
walbek n FOOD burnt curry Mai
sanaan jabek walbek
thawcawanaan. As far as eating
rice is concerned, the curry was
burnt, it was not tasty.
walbyt n ART match (to make fire)
walcham n ART bamboo torch
waldi n SUBST ambers, glowing
pieces of burnt wood
walkhu n SUBST smoke
walkhu- v to produce smoke
walkungki n SUBST black ashes
walsam n PLACE fireside
walthum n fire that is burned during
the winter outside the house to
keep warm
wala- v to arrive at night, to be late so
that it is already night Ma baba,
atykyimu walawa? nookno
amakaw, amakmi sadyrange.
Niwa. Ue nang awangpara
nokchi dangphakawa naa"
noatakokno. But daddy, why are
you so late? It is already night, the
monkeys children said. Dont
worry. I visited your uncle he
said, it is said.
walchak- v to kindle the fire with your
breath by blowing
walsymsym n GEO twilight
-wami ~ -wamyng encl.cl action/state
nominaliser clausal enclitic
wang n KIN 1. uncle: fathers younger
brother 2.stepfather
wang- v to bite a bit out of something
wang- v to turn, to wind
wanggala n ACT biggest Garo festival
warasak- v to defend, to shield, to
protect
warem n SUBST rust

709
wari n PLACE deep place in the river
where you can swim or take a bath
Tyi gagaba wari thywgaba
tyisamchi hap sylgabachi myng ni
baisigathangmaran chang
tyrywchengnaka noaidongano. At
the waterside of a place in the river
where there was nice and deep
water, in a beautiful place, the two
friends are arguing about who will
take a bath first, it is said.
waribul- v to fish at the festival of
waribula
waribula n ACT the Siju fishing
festival in the Symsang river at
Dabatwari
wat- v to weave things from reed or
bamboo, to make a mat or basket
from bamboo or reed Getheng
koksep watna mana. He can
weave a bamboo cage.
wat- v to send away, to banish, to get
rid of, to switch on an electrical
apparatus like a radio, TV,
computer etc., to let go Gethengmi
nokaw getheng watok, ytykyimu
dynthang nokchi muarok. His
family sent him away, so now he is
staying in another house. Git
watbo. Play some music. (on the
radio/tape/CD/etc.) Ram watbo. Go
out of the way. Sunyngchi rityi
watchawa. I will not cum inside
her vagina/cunt. In the negative this
verb can mean to seize, to
capture. Ramchi hampyi
nanangdo watchaka
gethengawdo, sala! Gethengaw
watkhuna sootthelarinaka noai
khumongangokno. This evening
we will certainly get rid of him on
the road, the bastard! We will kill
him after all to get rid of him once
more, they conspired freely, it is
said. Ytykyi pywtynangaimyng
kynsangdo dongna guduk
takwachian gethengtheng khusi
dongthamakaimyng gore dimaichi
phalthang chak diriga
sangwalaimyng khusi dongaimyng
watokno. Ytykyimyng galai
thyiokno. So then, having flown
away, later, when they almost
arrived, because they were so very
happy, they forgot to hold on to the
horses tail with their own hands,
and because they were so happy,
they let go, it is said. So then they
fell down and died, it is said.
watet- v to send (away), to post Ang
songthangchina dongangwachi
nangtymna chiti watetni. When I
have arrived in my own country I
will send you letters. Stem rong chi
dok tanaimu chiti wateta. You put
sixteen rupees worth of stamps and
post the letter.
watwa watwa adv scattered all over
the place
watyi n TIME rainy season
wawa onom throwing sound Bandi
kherengwachido wachu byryi
wawa wawa thangasigaaidoknote.
When Bandi makes a great effort, he
throws the tiger whoooosh! four
bamboo lengths away, it is said,
Im telling you!
wek onom the sound of a pig: squeal!
oink!
wek- v to sweep
wekwak- adj1 to be very soft (like
mud), sloppy Ram wekwakok. The
road is very soft like mud.
welet- v to flash
wen- ~ wen- v to wind around, to
wrap around, make as a coil
wen ~ wet n MSRE time, turn wet sa
~ wen sa once, wen ni twice wen
tham three times
wen ~ wyn ~ wyt- ~ wot v to
sharpen, to whet
weng n PLANT node (of bamboo),
joint
wenphak- v to wind around
something Dypyw ang chakaw
wenphakwa. The snake wound
itself around my hand.
wet ~ wen n MSRE time, turn wet sa
~ wen sa once, wen ni twice wen
tham three times

710
wetanchian adv every time
wetantian adv every time, time and
time again
-wil ~ -wilwil evsp V around
winwin- v to wind (something around
something)
wongong- v to stir
wongwet- v to dangle
wot- ~ wyt- ~ wen- ~ wyn- v to
sharpen, to whet
wungwung- v to stir
wyiset- v to wipe off
wyl- v to go down, to descend
wylang- v to go down, descend
wyn ~ wen ~ wyt- ~ wot- v to
sharpen, to whet
wynget- v to dangle
-wyngwang evsp V in a confused way
wyngwang- v to wag Kyi dimai
wyngwangaidong. The dog is
wagging its tail.
wyngwet- v to swing, to move back
and forth
wyt- ~ wot- ~ wen- ~ wyn- v to
sharpen, to whet
ym procl I agree, affirmative, OK,
thats right, yes
ymbyng n ART bamboo flute
ympong adj2 lopsided, convex,
having a surface or boundary that
curves or bulges outward, as the
exterior of a sphere Tyibekan
ympong. A traditional water bottle
is lopsided.
ymyi interj interjection of surprise
yndyn adv for nothing, in vain, for
free, simply
ytyk- v to do like this/that Ytykaria,
teewrawrawmi gawido. They do
just like that, the girls of nowadays.
ytykchiba disccon but, however
Acha babaji, angmi joraaw
chaina mannima? manniba.
ytykchiba raja sa nangnine. Ok,
fortuneteller, can you see my love
match? I can, but I will of course
need one hundred [rupees]. Angna
mamyngawan nangchawa,
ytykchiba naa angna aro angmyng
jykna nang khengwa dabat ang
thyicha dabat angaw muai sana
hynbo nookno. I dont need
anything. However, you keep
giving me and my wife something
to eat as long as you live until I
die, he said, it is said.
ytykchido disccon so, in that case
Nangtym ang nokaw sawwaba
nemariok. Anga nangtymaw
mythelbiok aro ang nok
chungkhuchido, ina daiai
mannichym anga tangka nookno.
Ytykchido ningba phalthang
nokaw sawaimyng hathapyra
phalchie mannima? It is really
good that you burnt down my
house. I thank you very much and
if my house had been bigger, I
would have gotten more money
than this, Thengthon said, it is
said. In that case, after we will
have burnt down our own houses
and sold the ashes, will we get
money too?
ytyken adv like this/that Atakna
raiawa? O, gylgylarong ytyken,
haratwanasa. Why have you
come? Oh, I am just roaming like
this, just because Im lazy.
ytykgaba adj2 this kind of, like this,
such Angba ytykgaba khadi
ranichymte. I would also buy
clothes like these.
ytykmachiba disccon but, however
Gadakchichiaimuna
thypsetthiriokno. Ytykmachiba uba
sagyraiba jumu
khathirithirioknotyi. He cut it into
pieces again and disposed of it
again, it is said. But that child
reassembled once again, it is said
to our surprise.
ytykyi adv like this/that Ma, manni
dongchido ie parang kunsaaw
kawanchyi noai hynokno. Ytykyi
songtawai hynokno. Well, you
will (be able to shoot the eagle), try
to shoot this culm of reed, she said
and gave one. She chose one and
put it upright like this, it is said.

711
ytykyimyng ~ ytykyimu ~
ytykyimuna ~ ytykyimung ~
ytkyimungna disccon so then,
having done that/this Teewe alsia
rajano song dam saci. Ytykyimyng
alsia raja song dam saci noaisa.
Kam khana nobo haratanoaro ue,
alsiae. Ytykyimyng jykba myng ni
khymanoro. Jyk myng ni khymano.
Ytykyimyng sanaba jyk paina
nangano, jywna jyk paina nangano.
Now, there is a lazy king, it is said,
in a certain village. So then, a lazy
king in a certain village, Im
saying, right. He is reluctant to do
work, it is said, that one, the lazy
person. So then, he is married to
two wives, it is said. He is married
to two wives, it is said. So then, his
wives have to carry the lazy king in
order to eat, it is said, and his wives
have to carry him in order to sleep,
it is said. Ytym myng korokan ha
kamarokno. Ytykyimungna
kamaimungna kynsangdo jywgaba
noksang raiaakno. The six of
them worked weeding the haba, it
is said. So then, after weeding the
haba their mother went home, it is
said.
ytykyisa disccon/adv therefore, like
that/this, then, thats why, so
Saepdyrang jalangokkhon noai
Arong nokma chaikhawwachi
Arong nokmami mukhangaw
khiemu thyiokno. Ytykyisa ue
Arong nokma thyiwamisa saepe
bondyk paiaimu sipaidyrang
dangna manokno. He said:
Maybe the white soldiers have run
away and when headman Arong
surreptitiously looked, having hit
headman Arong in the face, he
died, it is said. Thats why, because
of headman Arongs death, the
gun-carrying white soldiers were
able to enter the village, it is said.
Ytykyisa Bandie balaidongano []
Bandi spoke like this, it is said []

712
References
_____________________________________________________________________

Aikhenvald, Alexandra Y. 2003 a. A Grammar of Tariana, from Northwest Amazonia.
Cambridge: Cambridge University Press.
Aikhenvald, Alexandra Y. 2003 b. Classifiers: A Typology of Noun Categorization
Devices. Oxford, New York: Oxford University Press.
Aikhenvald, Alexandra Y. 2007. Typological distinctions in word formation, pp. 1-
65 in Timothy Shopen (editor). Language typology and syntactic description.
Vol. 3: Grammatical categories and the lexicon. Second edition. Cambridge
University Press.
Aikhenvald, Alexandra Y. 2008. The Manambu language from East Sepik, Papua
New Guinea. Oxford: Oxford University Press.
Aikhenvald, Alexandra Y. Forthcoming 2009. Versatile cases: Manambu and
beyond, in Journal of Linguistics, issue 45 vol. 1.
Aikhenvald, Alexandra Y. Forthcoming b: Causatives which do not cause: on non-
valency-increasing effects of valency-increasing derivations.
Aikhenvald, Alexandra Y. and R.M.W. Dixon. 1998. Dependencies between
grammatical systems, pp. 56-80 in Language, Vol. 74.
Anderson, Stephen R. 1992. A-Morphous Morphology. Cambridge, New York:
Cambridge University Press.
Andrews, Avery D. 2007. Relative clauses, pp. 206-236 in Language typology and
syntactic description, Volume II: complex constructions. Second edition.
Timothy Shopen (editor). Cambridge University Press.
Basumatary, Phukana Candra. An introduction to the Boro language. New Delhi:
Mittal Publications.
Benedict, Paul, K. 1972. Sino-Tibetan: a Conspectus. Contributing Editor, James A
Matisoff. Princeton-Cambridge Studies in Chinese Linguistics #2. Cambridge:
Bhat, D.N.S. 1999. The prominence of tense, aspect and mood. (Studies in Language
companion series, Vol. 49). Amsterdam, Philadelphia: John Benjamins
Publishing Company.
REFERENCES

713
Bhattacharya, Pramod Chandra. 1977. A descriptive analysis of the Boro language.
Gauhati University, Department of Publication.
Bickel, Balthasar. 1999. Nominalisation and focus constructions in some Kiranti
languages, pp. 271-269 in Topics in Nepalese linguistics, Yogendra P. Yadava
and Warren G Glover (editors).
van Breugel, Seino. 2006. The factitive category. An informal presentation held for
the Sino-Tibetan Special Interest Group at the Linguistics Department of La
Trobe University, on 4 July.
van Breugel, Seino. 2008. Similarities in verbal and nominal morphology in Atong,
pp. 169-190 in North East Indian Linguistics, Stephen Morey and Mark post,
(editors). New Delhi: Cambridge University Press India.
van Breugel, Seino. 2009 a. Atong-English dictionary.
van Breugel, Seino. 2009 b. Atong morot balgaba golpho.
Burling, Robbins. 1959. Proto-Bodo. pp. 433-453 in Language, Vol. 35, No. 3.
Burling, Robbins. 1961. A Garo grammar (Deccan College Monograph Series, 25).
Poona: Deccan College Post Graduate and Research Institute.
Burling, Robbins. 1963. Rengsanggri. Tura: Tura Book Room.
Burling, Robbins. 2003 a The Tibeto-Burman languages of Northeastern India, pp.
169-191 in Thurgood, Graham and Randy J. LaPolla (editors). The Sino-Tibetan
Languages. London, New York: Routledge.
Burling, Robbins. 2003 b. Garo, pp. 387-400 in Thurgood, Graham and Randy J.
LaPolla (editors). The Sino-Tibetan Languages. London, New York: Routledge.
Burling, Robbins. 2004. The Language of the Modhpur Mandi (Garo). Vol I:
Grammar. New Delhi: Bibliophile South Asian in association with Promilla &
Co., Publishers.
Comrie, Bernard. 1981. Language universals and linguistic typology. Oxford:
Blackwell.
Comrie, Bernard. 1998 a. Attributive clauses in Asian languages: Towards an areal
typology, pp. 51-60 in Sprache in Raum und Zeit, Band 2: Beitrge zur
empirischen Sprachwissenschaft. Winfried Broeder, Christoph Schroeder
(editors), Tbingen: Gunter Narr Verlag.
Comrie, Bernard. 1998 b. Rethinking the typology of relative clauses, pp. 59-86 in
Language Design. University of Granada, Spain.
REFERENCES

714
Comrie, Bernard. 2006. Syntactic Typology: just how exotic ARE European-type
relative clauses?, pp. 130-154 in Linguistic Universals. Ricardo Mairal and
Juana Gil (editors), Cambridge university Press.
DeLancey, Scott. 1986. Relativization as nominalization in Tibetan and Newari.
Paper presented at the Nineteenth Annual International Conference on Sino-
Tibetan Languages and Linguistics, Ohio State University, Columbus.
Dixon, R.M.W. 1994. Ergativity. Cambridge: Cambridge University Press.
Dixon, R.M.W. 2000. A typology of causatives: from and meaning, pp 30-83 in
Dixon R.M.W and Alexandra Y. Aikhenvald (editors). Changing valency: Case
studies in transitivity. Cambridge: Cambridge University Press.
Dixon, R.M.W. 2002. Australian Languages. Cambridge: Cambridge University
Press.
Dixon, R.M.W. 2003. Demonstratives, pp 61-112 in Studies in Language, 27: 61-
112. John Benjamins Publishing Company.
Dixon, R.M.W. 2004 a. Adjective classes in Typological perspective, pp. 1-49 in
Dixon, R.M.W. and Alexandra Y. Aikhenvald (editors). Adjective classes: a
cross-linguistic typology. Oxford University Press.
Dixon, R.M.W. 2004 b. The Jarawara language of Southern Amazonia. Oxford
University Press.
Dixon, R.M.W. 2006 a. The semantics of clause linking. Position paper for RCLT
International Workshop, 13-18 August 2007.
Dixon, R.M.W. 2006 b Complement Clauses and Complementation Strategies in
Typological Perspective, pp 1-48 in Dixon, R.M.W. and Alexandra Y.
Aikhenvald (editors). Complementation. Oxford University Press.
Dixon, R.M.W. 2006 c Comparative constructions in English, pp. 5-27 in Studia
Anglica Posnaniensia Nr. 41. Micha Jankowski, Ewa Kowalkowska and Anna
Hebda (editors), Poznan: The University of Adam Mickiewicz.
van Driem, George. 1987. A grammar of Limbu. Berlin, New York, Amsterdam:
Mouton de Gruyter.
van Driem, George. 2001. Languages of the Himalayas. Leiden, Boston, Kln: Brill
Elkin, A.P. 1970. The Australian Aborigines: How to understand them. Angus and
Robertson.
REFERENCES

715
Endle, S. 1881. Outline grammar of the Kachri (Br) language as spoken in the
district Darrang, Assam; with illustrative sentences, notes, reading lessons, and
a short vocabulary. Shillong: Assam Secretariat Press.
Genetti, Carol. 1992. Semantic and grammatical categories of relative clause
morphology in the languages of Nepal, pp. 406-427 in Studies in Language,
Vol. 16, No 2. John Benjamins Publishing Company.
Givn, Talmy. 1981. On the development of the numeral one as an indefinite
marker, pp 35-53 in Folia Linguistica Historica 2, 1.
Givn, Talmy. 1984. Syntax: A functional-typological introduction, Vol. 1.
Amsterdam, Philadelphia: John Benjamins.
Gordon, Raymond G., Jr. (ed.), 2005 a. Ethnologue: Languages of the World,
Fifteenth edition. Dallas, Texas: SIL International. Online:
http://www.ethnologue.com/. URL:
www.ethnologue.com/show_language.asp?code=aot
Gordon, Raymond G., Jr. (ed.), 2005 b. Ethnologue: Languages of the World,
Fifteenth edition. Dallas, Texas: SIL International. Online:
http://www.ethnologue.com/. URL:
http://www.ethnologue.com/show_language.asp?code=ato
Grierson, G. A. 1902. Linguistic Survey of India. Vol. 3, Tibeto-Burman Family; Part
2, Specimens of the Bodo, Naga and Kachin groups. Calcutta: Office of the
Superintendent of Government Printing.
Haig, Geoffrey. 2001. Diffusion in East Anatolia, pp.195-224 in Alexandra Y.
Aikhenvald and R.M.W. Dixon (editors). Areal diffusion and genetic
inheritence: Problems in Comparative linguistics. Oxford University Press.
Haiman, John. 1978. Conditionals Are Topics. pp 564-589 in Language, Vol. 54.
Hale, Kenneth. 1976. The adjoined relative clause in Australia, pp 78-105 in Dixon,
R.M.W. (editor). Grammatical Categories in Australian Languages. New
Jersey: Humanities Press Inc.
Haller, Felix. 2000. Dialect und Erzhlungen von Shigatse. Bonn: VGH
Wissenschaftlsverlag GmbH.
Haller, Felix. 2006. Verbal Valence in Shigatse Tibetan and Themchen Tibetan.
Linguistics of the Tibeto-Burman Area Vol. 29.2, pp 63-77.
REFERENCES

716
Heine, Bernd and Tania Kuteva. 2002. World lexicon of grammaticalization.
Hildebrandt, Kristine A. A grammar and glossary of the Manange language, pp.
193-324 in Tibeto-Burman languages of Nepal. Carol Genetti (editor).Canberra:
Pacific Linguistics.
Hopper, Paul J and Elisabeth Closs Traugott. 1993. Grammaticalization. Cambridge
University Press.
Hopper, Paul J. and Sandra A. Thompson. 1980. Transitivity in Grammar and
discourse Language, Vol. 56, pp. 251-299.
Jacquesson, Franois. 2005. Le Deuri: Langue tibeto-birmane d'Assam. Leuven,
Dudley (MA): Peeters.
Jacquesson, Franois. 2006. La rconstruction du pass: le cas des langues boro-
garo, pp. 273-303 in Bulletin de la Socit Linguistique de Paris, Vol. 101 no 1.
Jendraschek, Gerd. 2008. Questions on transitivity: Iatmul and beyond, a paper
presented in the Workshop on Transitivity of the Research Centre for Linguistic
Typology of La Trobe University, Bundoora, Victoria, Australia, 4 September
2008.
Joseph, U.V. 2007. Rabha. Leiden, Boston: Brill.
Joseph, U.V. and Robbins Burling. 2006. The comparative phonology of the Boro
Garo languages. Mysore: Central Institute of Indian Languages.
Keenan, Edward L. 1985. Relative Clauses, pp. 141-70 in Language typology and
syntactic description, Timothy Shopen (editor). Cambridge University Press.
Keenan, Edward L. and Bernard Comrie. 1977. Noun phrase accessibility and
universal grammar. Linguistic Inquiry 8: 63-99.
Kelly, Barbara. 2004. A Grammar and glossary of the Sherpa language, pp. 193-324
in Tibeto-Burman languages of Nepal. Carol Genetti (editor). Canberra: Pacific
Linguistics.
Lambrecht, Knud. 1994. Information structure and sentence form: Topic, focus and
the mental representation of discourse referents. Cambridge University Press.
LaPolla, Randy J. 1993 Arguments Against Subject and Direct Object as Viable
Concepts in Chinese, pp. 759-813 in Bulletin of the Institute of History and
Philology 63.4
Lass, Roger. 1984. Phonology. Cambridge University Press.
REFERENCES

717
Laver, John. 1994. Principles of Phonetics. Cambridge University Press.
Lehmann, Christian. 1984, Der Relativsatz. Typologie seiner Strukturen - Theorie
seiner Funktionen - Kompendium seiner Grammatik. Tbingen: G. Narr (LUS,
2).
Lehmann, Christian. 2007. Nexion Complex sentences. Universitt Erfurt. Online:
www.uni-erfurt.de/sprachwissenschaft/personal/lehmann/ling/lg_system/gramm
ar/nexion/index.html
Lehmann, Christian. 2003. Possession in Yucatec Maya. Second, revised edition.
Erfurt: Seminar fr Sprachwissenschaft der Universitt (ASSidUE, 10)
Lehmann, Thomas. 1989. A grammar of modern Tamil. Pondicherry: Pondicherry
Institute of Linguistics and Culture.
Lyons, J. 1977. Semantics. 2 volumes. Cambridge: Cambridge University Press.
Matisoff, James A. 1975. Rhinoglottophilia: the mysterious connection between
nasality and glottality. pp. 265-287 in Nasalfest (Papers from a Symposium on
Nasals and Nasalisation). Charles A. Ferguson, Larry M. Hyman and John J.
Ohala (editors). Stanford: Department of Linguistics, Stanford University.
Matisoff, James A. 1978. Variational semantics in Tibeto-Burman. Philidelphia:
Institute for the Study of Human Issues.
Matisoff, James A. 1982. The Grammar of Lahu. Berkeley, Los Angeles, London:
University of California Press.
Matisoff, James A. 2000. On the uselessness of glottochronology for the subgrouping
of Tibeto-Burman, pp. 333-71 in Time depth in historical linguistics, Colin
Renfrew, April McMahon and Larry Trask (editors). Cambridge: The McDonald
Institute for Archaeological Reseach.
Matisoff, James A. 2003. Handbook of Proto-Tibeto-Burman. Berkley, Los Angeles,
London: University of California Press.
Matisoff, James, A. 1972. Lahu nominalization, relativization and genitivization,
pp.237-57 in John Kimball (editor). Syntax and Semantics, Volume I. Studies in
Language Series. New York: Seminar Press.
Matsumoto, Yoshiko. 1997. Noun modifying constructions in Japanese: a frame
semantic approach. Amsterdam, Philadelphia: John Benjamins Publishing
Company.
REFERENCES

718
Mithun, Marianne. 1984. The evolution of noun incorporation. pp. 847-894 in
Language, Vol. 60, No. 4.
Moravcsik, Edith. 2003. A semantic analysis of associated plurals. pp 469-503 in
Studies in Language 27:3. John Benjamins Publishing Company.
Nengminza, D.S. 2001. The School Dictionary Garo to English. Tura: Sushil Kr. Das.
Noonan, Michael. 1997. Versatile nominalizations, pp. 373-394 in Essays on
Language Function and Language Type, Dedicated to T. Givn, Joan Bybee,
John Haiman & Sandra Thompson (editors). Amsterdam, Philadelphia: John
Benjamins.
Nordlinger, Rachel and Louisa Sadler. 2004. Nominal tense in crosslinguistic
perspective. pp. 776- 806 in Language, Vol. 80 Nr.4.
Okell, John and Anna Allott. 2001. Burmese/Myanmar dictionary of grammatical
forms. Curzon Press.
Okell, John. 1969. A reference grammar of colloquial Burmese. London: Oxford
University Press.
Parker, Steve. 1996. Toward a Universal Form for Yes: Or, Rhinoglottophilia and
the Affirmation Grunt. pp. 86-95 in Journal of Linguistic Anthropology 6(1).
1996. American Anthropological Association.
Peterson, David A. 2007. Applicative constructions. Oxford University Press.
Playfair, Alan. 1975. The Garos. Guwahati: United Publishers; reprint of Playfair,
Alan. 1909. The Garos London: David Nutt.
Rutgers, Roland. 1998. Yamphu. Grammar, texts & lexicon. Leiden: CNWS
Publications.
Sadock, Jerrold M. and Arnold M. Zwicky. 1985. Speech act distinctions in syntax,
pp. 155-196 in Shopen, Timothy (editor). Language typology and syntactic
description. Vol. 1: Clause structure. Cambridge University Press.
Sapir, Edward. 1938. Glottalized continuants in Navaho, Nootka and Kwakiutl (with
a note on Indo-European), pp.248-274 in Language, Vol. 14, reprinted in David
G. Mandelbaum (editor). 1951. Selected Writings of Edward Sapir in Language,
Culture and Personality. Berkley, Los Angeles: University of California Press.
Scancarelli, Janine. 1992. Aspiration and Cherokee orthographics, pp. 135-151 in
Pamela Downing, Susan D. Lima and Michael Noonan (editors). The linguistics
of literacy. Amsterdam, Philadelphia: John Benjamins Publishing Company.
REFERENCES

719
Shafer, Robert. 1953. Classification of the northern-most Naga languages, pp. 225-
264 in Journal of the Bihar Research Society, XXXIX (3).
Shafer, Robert. 1974. Introduction to Sino-Tibetan. Wiesbaden: Otto Harrassowitz.
Shin, Yong-Min. 2004. Possession und Partizipantenrelation. Eine funktional-
typologische Studie zur Possession und ihren semantischen Rollen am Beispiel
des Deutschen und Koreanischen. Bochum: Universittsverlag Dr. N.
Brockmeyer.
Sohn, Ho-Min. 1999. The Korean language. Cambridge University Press.
Thurgood, Graham and Randy J. LaPolla (editors). 2003. The Sino-Tibetan
Languages. London, New York: Routledge.
Van Der Mark, Sheena Chantal. 2003. The phonetics of Blackfoot Pitch Accent.
University of Calgary, Department of Linguistics: Calgary, Alberta.
Van Valin, Robert D. 1984. A typology of syntactic relations in clause linkage, pp.
542-558 in Proceedings of the tenth annual meeting of the Berkley Linguistics
Society. Claudia Brugman and Macaulay, (editors). Berkley: University of
California.
Van Valin, Robert D. and Randy J. LaPolla. 1997. Syntax: Structure, meaning and
function. Cambridge University Press.
Watters, David E. 2002. A Grammar of Kham. Cambridge: Cambridge University
Press.
Zubin, D. and M. Shimojo. 1993. How general are general classifiers? with special
reference to ko and tsu in Japanese, pp 490-502 in J.S. Guenter, B.A. Kaiser
and C.C. Zoll (editors). Proceedings of the Berkeley Linguistics Society:
Proceedings of the Annual Meeting 19. Berkeley: Berkeley Linguistics Society.

Seino Van Breugel-A Grammar of Atong PDF

Uploaded by

Copyright:

Available Formats

Seino Van Breugel-A Grammar of Atong PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Seino Van Breugel-A Grammar of Atong PDF

Uploaded by

Copyright:

Available Formats

A Grammar of Atong

higher pitch than previous syllable

ns [phaltaaw] cnkgbw n'ym

mcydok noymo te'ewb, jokmo jlydkn.

m cgb, d'ty tanaokno.

koym dgry sakno. ko

nsd, mn: pmi srmonmon

y ja'bek?, nokn. t doww

/. The allomorphs of the privative enclitic <=noy ~ =ni>

You might also like