Lexicogrammar of the Bihari Languages: a metafunctional survey

Lexicogrammar of the Bihari Languages: a metafunctional survey ABHISHEK KUMAR KASHYAP1 Department of English, School of Foreign Languages Sun Yat-Sen University, Guangzhou, P.R. China 1. Introduction This chapter is concerned with a survey of lexicogrammatical features of the Bihari languages. The Bihari group is a sub-set of the Eastern Indo-Aryan languages and these languages are primarily spoken in the present Bihar and the eastern part of Utter Pradesh in India and a few parts of Nepal (see Kashyap 2014). The language group received the name “Bihari” in the works of Sir G.A. Grierson (1881–87, 1968[1903]) after the geographical spread of the speech communities in Bihar, where these languages were spoken during that time. The speech communities are still primarily located in Bihar with a small community in Uttar Pradesh, an exception being Maithili that has a sizable number of speakers in Nepal.2 The Bihari group include the following five languages: Bhojpuri, Maithili, Magahi, Bajjika, and Angika. The first three of these are well-known among speakers among linguists and policy makers, and they are accepted as distinct speech forms. Of these, Maithili is one of the 22 languages listed in the eighth schedule of the constitution of India. It is also a major language of Nepal, as noted above. The later two, Bajjika and Angika, are new inclusion in the Bihari family. There is a controversy whether these two are languages on their own or they are dialects of one of the former three. Linguists have written about this issue and the language politics involved therein (see, for example, Burghart 1993; Kashyap 2014). The language issue involving these languages is still unresolved, however. In this paper I take a descriptive approach and focus on the grammatical structures at clause rank; essentially, the chapter is a brief overview of the clause types of Bihari languages and their realization statements. More specifically, I sketch here a general linguistic picture covering its lexicogrammar, which is a grammar of clause. The term lexicogrammar is used in Systemic Functional Linguistics. In other theories of linguistics, the grammar of clause is known as syntax. Understandably, the work presented in this article is preliminary its kind on this language group. My purpose here is to inform readers general lexicogrammatical properties of the Bihari languages. I believe this paper will serve as a foundation for more detailed and fine-grained description of these languages. 1 2 Email: [email protected] Before the 1912 division of Bengal, Bihar was part of the Bengal province. 2 2. General observation The Bihari languages are classified as Eastern Indo-Aryan branch of the Indo-Aryan language family: Indo-European > Indo-Iranian > Indo-Aryan > Eastern Indo-Aryan > Bihari. As yet, the only systematic survey of the languages of Bihar is by Sir G.A. Grierson (1883–87, 1968[1903]) who identified three Bihari dialects — Maithili, Bhojpuri, and Magahi. Sir Grierson provided detailed descriptions of the three dialects in his works, first in Seven grammars of the dialects and sub-dialects of the Bihari language (Grierson 1883– 87), and later as part of his monumental work, for which is known, Linguistic Survey of India (Grierson 1968[1903]). Although there has been no linguistic survey in Bihar after Grierson’s, research of Maithili, Magahi, and Bhojpuri have continued to research on these languages. Important point to note, however, is that some researchers have come up with the study of later two Bihari languages, namely Bajjika and Angika: for example, see Arun (1972), Jha (1994); Yadav (2003); Arun and Sharma (2008); Kashyap (2012, 2014, 2016). These languages follow OV word-order3. The Subject tends to appear at different places (i.e. clause-initially, medially, and finally), with no rigid fixed position in the clause, although the unmarked positon of Subject is clause-initial. These languages have their language-specific linguistic features by virtue of which they stand out from other IndoAryan languages. For example, length of vowel in all Bihari language have an allophnic feature, e.g. in the front vowel /i/ and the back vowel /u/. A most striking linguistic trait of the Bihari languages is the ability of multiple agreement. That is, the languages allow to co-index in their verbal morphology more than one referents. In this respect, Maithili is the richest and the most interesting; it allows to encoded in its agreement paradigm up to three referents (speaker, addressee, and a third person referent) at the same time. In contrast, Bajjika allows up to two referents to be simultaneously indexed in the verb agreement (cf. Bickel et al. 1999; Kashyap 2012; Kashyap and Yap in press). I illustrate these features in Section 4.1 below. 3. Theoretical perspective The linguistic theory on which this article is based is that of Systemic Functional Linguistics (SFL). SFL views language as a complex system organized at several strata: below clause level (i.e. phonology/morphology), at the clause (i.e. lexicogrammar), and above the clause (i.e. semantics). While semantics involves speech functions of statement, question, command and offer, they are realized at the level of lexicogrammar. And language is described in relation to context (Halliday 1978; Matthiessen 2005; Halliday and Matthiessen 2004). This principle is represented diagrammatically in Figure 1. In this paper, our focus is on the stratum of lexicogrammar, i.e. at the level of clause. 3 In my opinion the use of “word-order” problematic, but I retain this term here, as it is the most widely known term to refer to the order of major elements of clause. I address this issue below in Section 4.2.2. 3 context (of culture) above the clause semantics lexicogrammar at the clause phonology / graphology below the clause photetics Figure 1: Stratification in language Like the grammar of any other language, the grammar of Bihari languages has a number of systems at the ranks of morphology and lexicogrammar, e.g. the systems of TENSE, ASPECT, MOOD, PROCESS TYPE, THEME/RHEME, and so on. And an overall system of any of the Bihari languages would include all these systems. I do not intend to draw a system network here, but if you would like to see a how a overall picture would emerge if we drew a system network including all salient grammatical features of a language, you are advised to consult Kashyap (in press), which provides a preliminary system network of Bajjika with all salient grammatical resources of the language. As we focus on lexicogrammar in this article, I provide a system network that includes PROCESS TYPES, MOOD and THEME later in the following section: see Figure 2. 4. Lexicogrammar In this section I focus on the following aspects: verb-agreement, clause types and how the clauses are organized for making meaning. The verb-agreement is feature of verbal morphology, but because the agreement morphemes refer to referents that are nuclear elements of clause, it has strong lexicogrammatical implications. We first explain the system of verb-agreement in these languages. 4.1 Verb-agreement In general the verbal morphology of the Bihari languages is very elaborate. Particularly interesting and probably one of the most complex in Indo-Aryan languages is the paradigms of verb-agreement, which co-references multiple participants. The verbagreement allows simultaneously encoding in it more than one referents in different case 4 relations (Bickel et al. 1999; Kashyap 2012). For example, in the following examples taken from Kashyap (2012: 1669), the verb agreement in (1a) refers to the third person Subject referent bābuji ‘father.3h.NOM’, while in (3b) the agreement morphemes marked on the verb simultaneously co-indexes subject as well as non-subject referent. (1). Bajjika (Kashyap 2012: 1699) a. bābu-ji-ø khet me kām kar-ait ha-thin. father-h-NOM field LOC work do-PROG be.PRS-3h.NON ‘Father is working in the field.’ b. dosar kisān-ø appan dos-ke uṭhā ke other farmer-NOM self friend-ACC wake CONV bata-lak-ai. inform-PST.3nhNOM-3nh.NNOM ‘The other farmer woke his friend up and informed him.’ This kind of multiple agreement is also found in other Bihar languages, for example, in the following example from Magahi (Verma 1991: 132): (2). Magahi ham toh-rā dekh-li-o 1.NOM 2h-DAT see-PST.1-2h ‘I saw you’. And the situation in Maithili is even more interesting. Maithili verbal agreement allows encoding up to three referents at the same time, as in the following example taken from Bickel et al. (1999: 482): (3). Maithili ham to-rā kaniyā-ke 1.NOM 2nh-DAT bride-DAT dekh-au-l-i-au-nh see-CAUS-PT-1NOM-2nh/hNNOM-3hNNONM ‘I showed you the bride.’ In the case of single agreement, the agreement can be either with the nominative referent or the non-nominative referent, but in the case of double agreement, the first agreement is with the nominative referent and the second agreement with the nonnominative referent. The verb-agreement in PERSON is closely associated with the phenomena of honorification, which in turn is consistent with the hierarchical social order of Bihari speech communities. For the details on agreement paradigms of Bajjika and Maithili and their interpretation, see Kashyap (2012) and Bickel et al. (1999), respectively. A side effect of complex and multiple verbal agreement paradigm in the Bihari languages is absence of agreement in number from the verbal morphology: despite highly elaborate verbal morphology, verbs of Bihari languages do not show the distinction of number and the gender distinction is very marginal. The following examples of Bajjika attest the absence of verb-agreement with reference to number: 5 (4). Bajjika a. ham 1.NOM ‘I ate.’ b. ham-ni 1.NOM-PL ‘We ate.’ kh-ail-i. eat-PST-1.NOM kh-ail-i. eat-PST.1.NOM Gender is expressed only in reference to second person honorific and the third person honorific nominative referents, as in (5a)–(5d). The non-nominative agreement does not allow gender to be encoded at all. (5). Bajjika a. baiā a-el-an. elder.brother.NOM.M come-PST-3h.NOM.M ‘Elder brother came.’ b. bhauji a-il-in. sister.in.law.NOM.F come-PST-3h.NOM.F ‘Sister-in-law came.’ c. panḍi-ji a-el-an? priest.NOM.M-h come-PST-3h.NOM.M ‘Did the priest arrive?’ d. panḍit-āin a-il-in. priest-F come-PST-3h.NOM.F ‘Did the priest’s wife arrive?’ As it can be seen in the above examples, some verbs receive the inflection for feminine gender without any phonological change in the stem, whereas some other, for example aelan ‘came [male]’ versus ailin ‘came [female]’ in (5a) and (5b), respectively, make some morphological modification. The change depends on the internal phonological environment of the verb. These languages show a system of three tenses: present, past and future, with a binary distinction of non-future and future in non-copular constructions. The non-future and future are distinguished by the presence or absence of the morpheme -l- in the agreement morphology. The presence of -l- refers to non-future and its absence denotes that the tense is future. In the non-future scenario, the present and past tenses are distinguished by the sound preceding -l-. The following examples illustrate this principle. (6). Bajjika a. ḍholak-bā git Dholak.NOM-DEF song ‘Dholak sings a song.’ b. ḍholak-ba git Dholak-DEF song ‘Dholak sang a song.’ gāw-a-l-e. sing-PRS-NFUT-3nhNOM ga-e-l-ak. sing-PST-NFUT-3nhNOM 6 c. ḍholak-ba git ga-t-au. Dholak.NOM-DEF song sing-FUT-3nhNNOM ‘Dholak will sing a song.’ Each of these Bihari languages also have their own language-specific features (cf. the languages described in Cardona and Jain 2003). For example, two consonants in Bajjika are often geminated, e.g. two nasal /m/ in the genitive pronoun /hammar/ ‘my’ and in the adjective /nimman/ ‘good’, and two /j/ in the noun /majjil/ ‘death procession’. This feature of Bajjika is not yet reported for any other Bihari language. 4.2 Metafunctional organization Language has evolved to serve needs of mankind. If we observe carefully, and as previous research findings have shown, they serve us three primary functions, known as mentafunction in SFL (Halliday and Matthiessen 2004; Matthiessen 2004). Firstly, language serves us as a resource for construing our experience of the reality of life. This function is known as the ideational function of language; secondly, language functions as our resource for exchanging information and goods-&-services as well as establishing and maintaining interpersonal relationships with members of various social groups by recourse to deploying in speech various linguistic forms, e.g. politeness and honorification. This function is known as the interpersonal function. And thirdly, we use language for constructing discourse where we deploy the experiential and interpersonal function in our narratives. This third function is known as the textual function of language. The grammatical systems deployed for studying these three functions of language are those of TRANSITIVITY (experiential), MOOD and MODALITY (interpersonal), and THEME (textual). Seen from the metafuncitonal point of view, like in any other language all these three functions are expressed in the Bihari languages at the level of clause. In other words, the grammatical unit called clause embodies in it the three modes of meaning described above as metafunction: function of language for construing our experience, the function as an exchange of meaning in dialogue as an interactive event, and the function in the construction of text. In this section we will deal with the function of clause in different modes of meaning. The following sections focus on the three modes of meaning express by the clause where I describe how the Bihari languages organize different elements of clause for expressing the three modes of meaning. 4.2.1 Experiential In the experiential domain of grammar, clause of Bihari languages can be distinguished in two ways: based on structure and based on meaning. In terms of structure, they are either copular (i.e. one that contains a copula), as in (7), or non-copular, as in (8). (7). Bajjika kuch din pahile ke bāt h-ae. 7 something day before GEN matter ‘It’s a matter of a few days ago.’ COP (8). Bhojpuri rauā ai-ni. 2mh come-2mh ‘Your came.’ The copula verbs are different in each of the Bihari dialects, e.g. the forms ha in Bajjika (9a) and Maithili (9b), and forms of bā in Bhojpuri (9c)4. These languages have rich system of verbal agreement, each copula forms agree with subject and/or complement to encode the status of the speaker, addressee or a third person referent. (9). a. Bajjika i hun-kar bhul h-ae. . 3nh 3h-GEN mistake COP-3nh ‘This is his mistake.’ b. Maithili (Yadava 1996: 112) u nik ae-ch. . 3nh good 3nh-COP ‘He is good.’ c. Bhojpuri i hamār sārh. bā-ran. 3h 3h-GEN brother-in-law COP-3nh ‘He is my brother-in-law (i.e. wife’s brother).’ These copula forms also serve as an auxiliary verb in non-copular constructions, such as ha in the following example: (10). Bajjika ham ab ja-it ha-ti. 1 now go-PROG AUX-1 ‘I am leaving now.’ I describe the copular constructions as relational clauses below. 4 Note that there are other verb stems that serve as copula in each of the Bihari languages, for example, Bajjika uses rah- in non-present scenarios. Likewise, other Bihari languages also have another set of stems and forms that function as copula/auxiliary. I do not intend to go into detail in this paper. 8 transitive material + Actor mental + Senser TRANSITIVITY verbal + Sayer + Goal intransitive projecting ^ projected clause non-projecting existential Existent relational copula: ho-verb clause attributive + Attribute + Carrier identifying THEME marked + Identifier + Identified unmarked declarative MOOD declarative imperative falling intonation interrogative polar Negotiator: ki elemental Figure 2: Primary systems of Bihari clause in three metafunction The non-copular clauses are traditionally described as either transitive or intransitive. A clause of this type can have an auxiliary, as in (11a), or it can be without an auxiliary, (10b). The auxiliary is typically the same verb forms to those of copula, but in non-copular constructions, the copula/auxiliary is part of a multi-word verbal group, as hathin in inkār ka dele hathin in (11a). (11). Bajjika a. lekin hun-kar pati dilip thākur hun-kā rakhe but 3h-GEN husband Dilip Thakur 3h-DAT keep se inkār ka de-le ha-thin. ABL refuge do give-PST be.AUX.PRS-3h.NNOM ‘But her husband Dilip Thakur has refused to keep her (with him). b. pancaiti bh-el-ai. trial happen-PST-3nh.NNOM ‘The trial took place.’ This kind of analysis primarily focuses on forms and undermines meanings; as a result, several important and delicate semantic distinctions are lost. For example, the traditional interpretation of transitive/intransitive constructions ignores the potential impact on the affected participants and generalizes this distinction applying to all types of clauses, e.g. in 9 the clause that construes the experience of material action (such as waking and kicking) as well as the ones that construe the experience of inner workings expressed by mental process, e.g. thinking and knowing. Such are some basic problems with description in terms of copular and non-copular distinctions. Therefore, a meaning-based characterization of clause is imperative. In the meaning-based characterization in functional grammar (more specifically, in systemic functional grammar), we find four types of clauses in the transitivity system of the Bihari languages — relational, material, mental, and verbal. Each of these four types comprises a process typically realized by a verbal group, participants involved in the process, and optional circumstances.5 The participants (e.g. Actor, Goal, and so on) are realized by nominal groups, and circumstances by adverbial groups and postpositional phrases.6 Relational clauses typically have a copula that realizes the Process. There are three subtypes of this clause type: attributive, identifying, and existential (Figure 2). The existential clause has only one participant, the Existent — something that exists. A typical example of an existential clause is the opening clause of a typical fairy tale, given in (12), where rājā ‘king’ is the Existent, meaning ‘there existed a king’. (12). Bajjika e-go rājā rahe. one-NUM king be.PST Existent Process ‘There was a king.’ The identifying and attributive clauses construe the relationship between two entities that are of the same nature related by the copula verb. Crucially, while an existential clause construes the existence of an entity, attributive and identifying clauses establish a relationship of the two participants of the clause. Apparently, they differ from an existential clause in terms of the number of participants. The attributive clause describes the class-membership or quality, the Attribute, attributed to a participant and the grammatical category assigned to the participant that carries the ascription of quality or class-membership is Carrier: (13). a. rādhe barā nimman hae. Radhe much good be.PRS Carrier Attribute Process ‘Radhe is very good.’ b. Maithili u nik ae-ch. 3nh good 3nh-COP Carrier Attribute Process 5 Although in some behavioral clauses the circumstances are not optional; rather they are obligatory, as in this clause: u ta gajbe beohār karait hai ‘he is behaving strangely’ (Bajjika). 6 In an attributive type of relational clause, the Attribute can be an adjectival group, e.g. barā nimman “very good” in the example (13a). 10 ‘He is good.’ In the identifying type of relational clause, one entity serves to serves to identify another entity. The two entities of the clause are of identical properties and related in such a way that one identifies the other; thus one entity functions as Identifier and the other as the Identified. This typical configuration of this clause type is in the form of x is y.: (14). a. Bhojpuri i hamār sārh. bā-ran. 3h 3h-GEN brother-in-law COP-3nh Identified Identified Process ‘He is my brother-in-law (i.e. wife’s brother).’ b. Bajjika bar-kā ke nām rahe rahmān. big-DEF of name be.PST Rahman Identified Process Identifier ‘The name of the elder one was Rahman.’ Material clauses construe the experience of the outer world around us — a world of action and event — of doing and happening, e.g. running, raining, and dancing. They can be distinguished as ‘happening’ and ‘doing’ constructions in all these Bihari languages, as subtypes of material clauses. The happening clause type (which is also characterized as an event clause) construes the experience that seems to have self-actualized, as in (15a). The material clauses expressing meteorological experiences are of the eventive type, (15b). (15). Bajjika a. kām ho ge-lak. work happen go-pst.3nh ‘The work was done.’ (Lit. The work happened.) b. megh bars-ait hai. rain pour-PROG AUX-PRS ‘It is raining.’ The ‘doing’ clauses (i.e. action clauses) express that the material change in the process is brought about by some external agency, for example by hunkā ‘3h.DAT) in (16a) and chãorā ‘lad’ in (16b). (16). a. Maithili (Bickel et al. 1999: 492) hun-ka ciṭhi likh-ai-ke ̃ cha-l-ainh. 3h-DAT letter write-IP-DAT AUX-PST-3h.NNOM ‘He had to write a letter.’ 11 b. Bajjika chãorā gen pokhrā me phẽk de-lak. lad ball pond LOC throw give-PST.3nh. NOM ‘The boy threw the ball in the pond.’ A material clause typically contains an Actor that brings about the change in the process. A transitive material clause also contains a Goal to which the effect of process carries over or extends, (17b) and (17c); an intransitive clause has only one participant, Actor, as in (17a): (17). Bajjika a. ā tin-o ādmi jaorahi bhittar ge-lan. and three-also man together inside go-PST.3nh Actor Manner Place Process ‘And all the three people entered together (inside the theatre).’ b. u paenā khĩc-lan. 3nh stick pull-PST.3h Actor Goal Process ‘He pulled the stick. c. baki paenā na khĩc-a-el. but stick NEG pull-PASS-PST Goal Process ‘But he couldn’t pull the stick.’ (Lit. ‘But the stick did not get pulled.’) The mental clause construes the experience of inner consciousness — what goes within us, e.g. thinking, feeling, and knowing. In Bihari languages, mental clauses have four subtype: perceptive — the construal the experience of perception, (18a); emotive — the construal of the experience of emotion, (18b); cognitive — the construal of the experience of cognition, (18c) and desiderative — the construal of the experience of need, (18d): (18). Bajjika a. rauā ham-rā ke cinh-li? 2mh I-DAT of recognise-PST.1/2mh Senser Phenomenon Process ‘Did you recognize me?’ b. ham-rā kobi na sohā-le. 1-DAT cauliflower NEG like-PRS Senser Phenomenon Process ‘I don’t like cauliflower.’ c. tu ta sārā kahāni janbe kara-la. 2h TEX all story know do-PRS.2h Senser Phenomenon Process ‘You already know the whole story.’ d. e beri sāikil ke ki jaruri h-aw? this TIME bicycle GEN what need be.PRS-2h.NNOM 12 Time Phenomenon Reason Process ‘Why do you need a bicycle at this time?’ One of the participants of the mental clause is a conscious entity, the Senser. The Senser is usually a human being who senses — perceives, thinks, or knows, but it can also be a nonhuman entity represented as one that is endowed with consciousness, e.g. some nonhuman participants in folk stories or fairy tales. In the examples given above, the participants rauā ‘you’ in (18a), hamrā ‘to me’ in (18b), and tu ‘you’ in (18c) are all conscious participants. The Senser in (18d) is elliptical, recoverable from agreement morpheme -aw, which indicates that the participant is second person honorific. The Senser of desiderative mental clause is always in the dative case. In (18d) it is reflected by the non-nominative agreement marking in the verb. (Note that the nominative and non-nominative distinction is made in the gloss only when it is needed for explanation. We continue to ignore the nominative and nominative marking in the gloss unless it is needed otherwise.) Another feature of Bihar mental clause is the capacity to project another clause. A mental clause can have a Senser + Phenomenon + Process configuration like in the examples of mental process clauses given above, or it can have a configuration of Senser + Process + projected clause, as in the examples given in (19): (19). Bajjika a. α raghunāth dekh-lan Raghunath see-FUT.3h Senser Process: mental ‘Raghunath thought (lit. saw).’ b. β je ijor ho-i REL light happen-FUT Actor Process: material ‘It is bright.’ c. γ ta bhaiā dekh li-han. then elder brother see take-HON.FUT Senser Process: mental ‘Then the elder brother will see (us).’ The clauses (19a) through (19c) make a clause complex, in which the alpha (α) clause (19a) construes a mental process, a projecting clause that projects two dependent clauses, (19b) and (19c). The projected clause can be any of the four clause types (i.e. material, mental, relational, and verbal). For example, of the two projected clauses above, (19b) is a material clause, whereas (19c) is a mental clause. There are two main grammatical properties that distinguish mental clauses of Bajjika from material ones: 13 (i). (ii). Consciousness: at least one participant in a mental clause is a conscious entity or an entity that is endowed with consciousness. A material clause does not need to have a conscious or endowed-with-consciousness participant. Projection: A mental clause can project another clause in the logical clause combination. A material clause cannot do so. There is another type that partially shares features of both material and mental processes, as in the following Bajjika example: (20). Bajjika okk-ar deh sihar g-el. 3nh-GEN body shiver go-PST ‘S/he shivered.’ (Lit. His/her body shivered.) This clause has one conscious participant, the one that shivers (or whose body shivers). Here the experience of inner working of the participant is reflected by physical behaviour. In the description of English, this type of clause is described as a ‘behavioural’ process (Halliday and Matthiessen 2004). However, in Bihari, in addition to one feature of mental process, i.e. the consciousness of the participant, it has no distinct grammatical property of its own, nor does it have all properties of mental process. For example, it does not have the ability to project another clause. Rather it is more close to a material clause. Therefore, I interpret this type of clause as a material clause. The verbal clause construes the experience of ‘saying’. Here the experience of inner consciousness is enacted though speech in the form of language, as in (21): (21). Bajjika rādhā ro-it-kan-ait appan dukh kah-lak. Radha cry-PROG-RDP-PROG self grief say-PST.3nh Sayer Manner Verbiage Process ‘Radha cried and narrated her grief.’ The verbal clauses can be either the projecting type or the non-projecting type and typically contain the Sayer (i.e. the participant who speaks or says) as an obligatory participant. A non-projecting verbal clause has another participant called the Verbiage — what is said — realized by a nominal group, as appan dukh ‘his/her grief’ in (21). And the projecting ones have a projected clause instead of a Verbiage. 4.2.2 Interpersonal The description of order of constituent elements of unmarked clause is a significant factor taken into language description. This is well-known as word-order in traditional grammar. Here I would like to raise two issues involved with the use of word-order and this need to be understood for the precision of category we use. Firstly, the entities described as “wordorder” are not order of words in their real sense; it is the order of major constituent 14 elements of clause that are composed of a word or a string of words. For example, Subject and Object in the order SOV are not words; they are functional categories that can be composed of a word or a string of words. Secondly, the terms Subject, Object, and Verb are categories of different nature. While Subject and Object are functional constituents, Verb is a word class and should be grouped and compared with other members of word classes such as noun and adjective but not functional constituents. We known that we do not compare apples with oranges; we compare apples with apples and oranges with oranges. In traditional sense, the sequence of unmarked declarative clause of Bihari languages is (S)OV. In naturally occurring texts, the Subject in Bihari languages appears in different locations in the clause, which is consistent with most Indo-Aryan languages as flagged in several other works (for example, see (Shapiro 2003) for Hindi). In particular, in spoken discourse S is very freely movable; in the written discourse S appearing in the beginning of the clause seems more likely. Since we identify the problem with the use of S, O and V above, we will reinterpret the O and V. As noted above, the labels S(ubject), O(bject), and V(erb) are categories used in traditional grammar. In the grammatical theory we use in this paper, i.e. SFL, the categories Object and Verb (of the “word-order”) are reinterpreted and re-introduced. In SFL the categories equivalent to Object and Verb are Complement and Predicator, respectively. This is in keeping with a distinct category at different levels, but they also have different roles in grammar. As noted above, Verb is a category in the word class, i.e. below the clause rank, while the Subject is a category of clause. Thus, to say that a language has SOV as its unmarked sequence is to confuse categories of different ranks. As we are describing the lexicogrammar, i.e. at clause rank, we use a distinct category to maintain this distinction. This is shown in the analysis presented in Figure 3. Subject, Complement, and Predicator are clause-rank categories. So, to sum up, an unmarked clause of a Bihari languages follows (S)CP order (that is, Subject ^ Complement ^ Predicator). In actual text the occurrence of C and P is fixed, while that of S is freely movable. It can occur anywhere in the clause. Clause amit gari-ā ke Amit Vehicle-DEF ACC Traditional characterization Subject Object SFL characterization Subject Complement ‘Amit stopped the car.’ rok de-lak. stop give-PST.3nh Verb Predicator Figure 3: Analysis of constituent elements of an unmarked declarative clause of Bajjika In the interpersonal domain of grammar, clauses of Bihari languages, like those of most other languages, are interactive moves in the ongoing dialogue as an exchange of meaning. They are deployed for exchanging either information or goods-&-services and the commodities exchanged are in the form of demand or supply. This exchange (of information and goods-&-services) are served by four primary speech functions — statement, question, command, and offer, which are grammaticalized in the system of MOOD (see Figure 2 above) The exchange of information (statement and question) is grammaticalized in the indicative mood: interrogative — demanding information, (22a); declarative — giving 15 information, (22b); and the exchange of goods-&-services — command and offer — is grammaticalized in imperative and interrogative mood, respectively, (22c) and (22d). Note that while statement, question, and command have a distinct mood for realizing each of the three speech functions, offer is associated with interrogative mood; it does not have a distinct mood of its own. This is the situation in most languages of the world (Halliday 1984: 19–20) and Bajjika is not an exception. As exemplified in (22a), Bajjika does not have a distinct way of formalizing ‘offer’. It is formalized in the same way as a polar interrogative clause. (22). Bajjika a. a-elan ki? come-PST.3h MOD: INT Predicator Negotiator ‘Did (he) come?’ b. na, hun āj nanna a-i-han. no 3h today NEG come-FUT-3nh Subject Adjunct Predicator ‘No, he will not come today.’ c. u-hãi dha d-a. DEM-PLACE put give-2h Adjunct predicator ‘Put (that) there.’ d. mamā pāni ān di-aw? uncle water bring ASP-2h Vocative Complement Predicator ‘Uncle, should (I) bring water for you?’ ‘question’ ‘statement’ ‘command’ ‘offer' The following are an example of statement from Maithili, Magahi and Bhojpuri each: (23). a. Maithili ahã dhanik ā nāmi dunno ch-i. 2mh rich and famous both COP.PRS-2mh ‘You are both rich and famous.’ b. Magahi ham-rā se dekh-dākh na howa ho. 1-DAT ABL see-RDP NEG happen AUX.PRS ‘I cannot manage to look after (these things).’ c. Bhojpuri bajār jā-it bā-ni. market go-PROG aux.prs-1 ‘(I) am going to the market.’ 16 The interrogative clauses are of two types: elemental and polar. The polar interrogative clause contains a mood key ki, labeled as the Negotiator in (22a). The elemental interrogative (that is traditionally known as the WH-interrogative), (24), has a question element that typically begins with the sound /k/. (24). a. Bhojpuri i ke 3h/nh what ‘Who is he?’ b. Bajjika i ke 3h/nh what ‘Who is he?’ bā-ran? be.PRS-3h ha-tan? be.PRS-3h There is a close relationship between speech function (or mood) and intonation in these languages. In fact, intonation is the most powerful resource for exchanging different types of meaning and different intonation contours are deployed for conveying different meanings. Statements and commands have falling intonation; questions have rising intonation. The elemental interrogative contains a question element, e.g. ke ‘what’ in (24) confirms the meaning that is being exchanged and specifies the element of information that is being sought. The polar interrogative has a mood key, the Negotiator discussed above. However, a polar interrogative clause without a Negotiator is equally frequent in speech. The resource that serves to exchange meaning in such a scenario is rising intonation. 4.2.3 Textual The third function of language, as theorized in SFL and introduced in the beginning of this section, is the organization of information to enable the speaker or writer to construct text. Whether unconsciously, as in most spoken texts, or consciously, as in most written texts, the speaker/writer carefully organizes the information that s/he wishes to communicate in discourse. The whole message is created as smaller units at the clause rank and put together in such as way that the text forms a unified whole. The information in a clause flows as a wave of information giving prominence to certain (not all) chunk of information. A clause of the Bihari languages begins, in textual terms, with the peak of the wave of information. The first segment of the clause contains the prominent part of information — the message that the speaker is concerned with. As the clause moves on, the textual prominence gradually declines making the whole wave of prominence like a trough. The clause thus has two parts: Theme (i.e. the concern of message) and the Rheme (i.e. the rest), with the order of Theme followed by Rheme. This is exemplified in the following examples of Bajjika: (25). Bajjika a. thorā bahut kheti-bāri ta little much agriculture-RDP PT rah-be kar-ain hun-kā. be-DEF do.PRS-3h he.HON-DAT 17 Theme Rheme Given New ‘Little bit of agriculture he already had.’ b. mālo-jāl rakh-le rah-ath. cattle-RDP keep-PST be-PST.3h Theme Rheme Given New ‘(He) also had cattle.’ The clauses also have a close relationship between the structure of Theme and that of information. The information already available (i.e. the given information) is mapped on to the Theme, and the new information is mapped on to the Rheme. More specifically, the information that is already available to the reader is usually placed in the beginning of the clause followed by the new information placed in the later part of the clause. To understand this principle of THEME and INFORMATION, we need a detailed analysis of a text or (an extract of text), but at this point it will suffice to flag the observation that in Bihari languages usually the information available to the speaker/writer is mapped on to the Theme. 5. Summary and Conclusion In this article I have provided an exploratory survey of the lexicogrammatical features of the five Bihari languages, namely Maithili, Magahi, Bhojpuri, Bajjika and Angika and have tried to show that these languages share common lexicogrammatical features. The description in this paper included observations with respect to clause types, their structural arrangements and an explanation to how each clause of these languages convey three modes of meaning. Essentially, the description covers detail of the system of verbagreements, the function of clause in construing our experience of reality of daily life (experiential function), its function in exchanging information and goods-&-services (interpersonal function), and its function in creating discourse/text (textual function). In terms of transitivity (i.e. experiential function), the Bihari languages have four PROCESS TYPES: material, mental, verbal, and relational. Interpersonally, the basic distinction in the enactment of speech functions is based on the exchange of information and goods-&services that are grammaticalized in indicative and imperative moods, respectively. In relation to textual function, a Bihari clause begins with the peak of information, and as the clause develops, the prominence gradually declines. The later part of clause is an area of interpersonal prominence in that the clause has a number of interpersonal indicators there, for example, in relation to speaker’s and addressee’s social status such as politeness / honorification formalized in their very elaborate agreement paradigms. In the absence of real and adequate data, I was unable to provide examples from Angika, and the examples of Magahi and Bhojpuri are fewer than sufficient. I must confess this. However, as a native speaker of the Bihari languages and a student of linguistics, I am convinced that the observations I have made in this paper are valid and can be generalize 18 across languages of the Bihari group. Needless to say, detailed analyses of texts from Angika and other languages of the Bihar family are on my agenda for future research. Abbreviations 1 2 3 ACC ADJ ASP AUX DAT DEF DEM DIM F FUT GEN h hh IMP first person second person third person accusative adjective aspect auxiliary dative definiteness demonstrative diminutive female future tense genitive honorific high-honorific imperative INT LOC M mh NEG NFUT nh NNOM NOM PFV PL PN PROG PRS PRT PST SG interrogative locative male mid-honorific negative non-future tense non- honorific non-nominative nominate perfective plural proper name progressive present tense particle past tense singular References Arun, Awadheshwar. 1972. Bajjikā, Hindi aor Bhojpuri kā tulnātmak adhyayan. Muzaffarpur: Kumud Prakashan. Arun, Awadheshwar and Ram Niwas Sharma (eds.). 2008. Bhārti Bajjikā Hindi koś. New Delhi: Konark Prakashan. Bickel, Balthasar, Walter Bisang and Yogendra P. Yādva. 1999. Face vs. empathy: the social foundation of Maithili verb agreement. Linguistics 37(3). 481−518. Burghart, Richard. 1993. ‘A Quarrel in the Language Family: Agency and Representations of Speech in Maithili.’ Modern Asian Studies 27(4): 761–804. Cardona, George and Dhanesh Jain (eds.). 2003. The Indo-Aryan languages. London and New York: Routledge. Grierson, G.A. 1883–87. Seven grammars of the dialects and sub-dialects of the Bihari language, Parts I–VIII. Delhi: Bharatiya Publishing House. Grierson, G.A. 1968[1903]. Linguistic survey of India, Vol. V. Indo-Aryan family, eastern group, Part II: Specimens of the Bihari and Oriya languages. Calcutta: Office of the Superintendent of Government Printing. Reprinted in 1968 by Motilal Banarasidass. 19 Halliday, M.A.K. 1978. Language as social semiotic: the social interpretation of language and meaning. London: Edward Arnold. Halliday, M.A.K. 1984. ‘Language as code and language as behaviour: a systemic-functional interpretation of the nature and ontogenesis of dialogue.’ In Fawcett, R.P., M.A.K. Halliday, S. Lamb and A. Makkai (eds.), The semiotics of language and culture. London: Frances Pinter. 3–35 Halliday, M.A.K. and Christian M.I.M. Matthiesen. 2004. An Introduction to Functional Grammar. 3rd edition. London: Hodder Arnold. Jha, Shailjanand. 1994. ‘Maithili in the Indian Census.’ Journal of Multilingual and Multicultural Development 15(5): 385–397. Kashyap, Abhishek Kumar. Forthc. A functional grammar of Bajjika: A systemic functional perspective. Leiden: Brill. Kashyap, Abhishek Kumar. In press. ‘On the linguistic resources of Bajjika.’ In: Chauhan, Vibha (ed.). The Languages of Bihar. New Delhi: Orient Blackswan. Kashyap, Abhishek Kumar. 2012. ‘The pragmatic principles of agreement in Bajjika verbs.’ Journal of Pragmatics 44(13): 1668–1687. Kashyap, Abhishek Kumar. 2014. ‘The Bajjika language and speech community.’ International Journal of the Sociology of Language 227: 209–224. Kashyap, Abhishek Kumar. 2016. ‘The representation of gender in Bajjika grammar and discourse.’ In: Baider, Fabienne and Julie Abbou (eds.), Gender and the Periphery: Grammatical and Social Gender from the Margins. Amsterdam/Philadelphia: John Benjamins. Kashyap, Abhishek Kumar and Foong Ha Yap. In press. ‘Epistemicity, social identity and politeness marking: A pragmatic analysis of Bajjika verbal inflections.’ Linguistics. Matthiessen, Christian M.I.M. 2004. ‘Descriptive Motifs and Generalizations.’ In Caffarel, Alice, J.R. Martin and Christian M.I.M. Matthiessen (eds.), Language Typology: A Functional Perspective. Amsterdam: John Benjamins. 537–673. Matthiessen, Christian M.I.M. 2005. ‘The ‘architechure’ of language according to systemic functional theory: development since the 1970s.’ In Hasan, Ruqaiya, Christian Matthiessen and Jonathan Webster (eds.), Continuing discourse on language vol 2. London: Equinox. 505–561. Shapiro, Michael C. 2003. ‘Hindi.’ In Cardona, George and Dhanesh Jain (eds.), The IndoAryan Languages. London and New York: Routledge. 250–285. Verma, Manindra K. 1991. Exploring the parameters of agreement: the case of Magahi. Language Sciences 13 (2). 125–143. Yadav, Ramawatar. 2003. ‘Maithili.’ In Cardona, George and Dhanesh Jain (eds.), The IndoAryan languages. London and New York: Routledge. 523–546.