WCCFL 28 USC, February 19-21, 2010 Plural mass nouns and the morpho-syntax of Number Artemis Alexiadou Universitt

Stuttgart
[email protected]

1. 1.1

Introduction The mass vs. count noun distinction

As is well known, in a number of languages nouns fall into two main groups, see Jespersen (1909):1 (1) a. book b. water count noun mass noun (CN) (MN)

The two differ from each other in a variety of ways: Morphology: CNs are compatible with plural marking, MNs are not:2 (2) a. the books b. #the waters Syntax: a) CNs are compatible with numerals, MNs are not; b) CNs can co-occur with definite and indefinite articles, MNs only with the definite article, see Gillon (1992): (3) numerals a. two books b. #two waters articles c. the/a book d. the/# a water

CN MN

CN MN

NB: Of course there is a considerable amount of 'elasticity' that enables both CNs to occur as MNs and MNs to function as CNs: (4) There are three waters in the fridge 3 types/bottles of water

To explain the above distribution, the following has been proposed:

Note that several researchers have argued that this distinction does not hold universally. For example, Indonesian has been argued to lack this distinction (Dalrymple & Mofu 2008), and Halkomelem Salish as well (Wiltschko 2009). 2 See also Greenberg (1963), Mufwene (1981).

a) The morphological realization of plural has a semantic contribution: in particular, it makes reference to sums of individuals. Concretely, nouns with plural morphology qualify as atomic/individual (Borer 2005, Chierchia 2008). MNs are not atomic. b) Numerals introduce quantity readings. This means that they count concrete individuals. Since MN have cumulative reference, they are out in the context of a numeral (Quine 1960, Krifka 1989). c) Number and numerals are part of the functional information within the extended projection of NP, see Grimshaw (1991), van Riemsdijk (1998), Abney (1987), Ritter (1991), Heycock & Zamparelli (2005), Borer (2005), Chierchia (2008), among many others (5a-b), but cf. Bouchard (1998), Dobrovie-Sorin (2009). (5) a. the three book-s b. DP (Determiner phrase) 3 Def Quantity phrase 3 numerals Number phrase 3 Plural marking nP individuation/atomicity

This crucially means that MNs are incompatible in combination with plural morphology and/or numerals, unless type-shifting takes place, as in (4). 1.2 Two puzzles concerning MNs3

First Puzzle: In Greek, plural MNs can preserve their MN semantics (no type shifting). (6) epesan nera sto kefali mu fell-3pl water-pl on head my "Water fell on my head" Greek Tsoulas (2006)

(6) is unexpected given the above remarks. Second puzzle: Grimshaw (1990) argued in detail that argument structure nominals (ASNs) disallow numerals, plural marking and are only compatible with the definite article. This is illustrated below for English, but holds for Greek as well: ASNs behave like MNs (7) Plural: a. *{the, some, a lot of} jumpings of the cow b. one jump, two jumps ASN

Early discussion of the first puzzle can be found in McCawley (1979), see also Corbett (2000), Ojeda (2005), Acquaviva (2008), Tsoulas (2006). The second puzzle has not been discussed in the literature, as far as I can tell.

Numerals: c. *One jumping of the cow was interrupted by the fireworks. d. One jump was disqualified. Article e. the jumping of the cow was interrupted by the fireworks f. *a jumping of the cow was interrupted by the fireworks g. a jump

ASN

ASN ASN

Prediction: If plural marking on MNs is possible in a language (Greek), then ASNs in this language should also allow plural marking. Observation: This does not hold. ASNs cannot pluralize. Pluralization is possible only under type shifting. When this happens, they are interpreted as R(eferential)-nominals. Questions: What does this all show about the function of plural marking? What does this show about the category Number, the morpho-syntactic structure of MNs and of the noun phrase in general and its compositional semantics? 1.3 The idea in a nutshell

Split analysis of plural Number: there are two types of plural Grammatical plural A: is on the Number head and realizes a plural semantic feature: in combination with a nP X it gives the interpretation more than one X. lexical Plural: is on n (categorizing head)4 it combines only with certain roots, and does not give rise to the interpretation more than one. Plural MNs are simple nPs. Nominal derivational affixes and lexical plura both realize n; hence they are in complementary distribution.

2.

Plural-MNs

Plural MNs are not typologically rare (see Corbett 2000, (8-10) from Acquaviva 2008: 109ff., Kramer 2009): (8) (9) The river discharges its water/waters into the lake xood har-ramlaat mn hoon get the sand-pl here out "get this sand out of here"

Syrian Arabic

Following Acquaviva (2008), I use the term lexical plural here, cf the distinction between"inherent vs. inflectional number" in Acquaviva (2004b). See also Kramer (2009) who develops and then rejects a similar split analysis of Number in Amharic.

(10) b--ro az kaf-e pazxune jam kon water-pl-acc from floor-prt kitchen gathering do "Wipe away the water from the kitchen floor" Plural-MNs are not pluraliatantum: (11) 2.1 scissors, trousers, ... The properties of plural MNs in Greek

Persian

Plural MNs have cumulative reference, pretty much like singular-MN. Still, they agree with the verb in plural (6). Are they really MNs? Yes. In combination with numerals and articles they behave exactly like their singular counterparts: (12) (13) #Dio nera peftun apo to tavani (possible only under type shifting) two waters fall from the ceiling mazepse ta rizia apo to patoma/#mazepse ena rizi apo to patoma get the rice- pl from the floor/get one rice from the floorn "Wipe away the rice from the floor"

Is plural morphology then expletive? (as argued in Tsoulas 2006) No: Plural MNs do receive a different interpretation; a great amount of MN (see the notion of "greater plural" in Corbett 2000: 30f., Acquaviva 2008: 109ff.): (14) nero nera water-pl = water-sg = a lot of water vs. vivlio = book-sg (Greek) vivlia = more than one book book-pl

Distributional restrictions, based on Acquaviva (2004a): i) Plural-MNs are found with predicates such as fall, spray, drip, gather (mostly spray/load predicates, see Tenny 1987, Rappaport & Levin 1988). Such contexts facilitate the conceptualization of a greater amount, (15a-b):

(15) a. ipia nero/#nera (only under type shifting) drank-1sg water/water-pl "I drank water" b. hithikan nera / hithike nero sto patoma dripped water-pl/ dripped water on the floor "A lot of water/water dripped on the floor" ii) Some plural MNs are better than others: water, oil, mud, sand, rice, snow are better than honey, juice and traffic.

iii)

The change in interpretation is not constant. For example, in Greek ksilo (wood) receives an MN interpretation only under pluralization: c. ksilo vs. wood-sg "piece of wood" ksila wood-pl "wood"

I take (i)-(iii) to indicate properties normally associated with idiosyncratic behavior.

The above properties are unexpected and raise a number of questions concerning the nature of plural morphology. 2.2 Plural-MNs and the structure of the noun phrase

As mentioned above, plural marking is standardly taken to be associated with a particular projection in the structural representation: (16) a. the ta b. three tria book-s vivli-a

DP (Determiner phrase) 3 Def Quantity phrase 3 numerals Number phrase 3 Plural marking nP

Semantics of Plural: Plural denotes that a certain amount is greater that one. This is why plural marking presupposes individualization/atomicity, Mufwene (1981), Chierchia (2008). Morphology: it surfaces on the noun. Note that in Greek the article, numeral and noun all agree in number. NumberP is the input to Quantity.5 Meaning-form relationships: Plural marking on CNs: Singular marking on MNs: x + Plural = more than one x Default

Plural marking on MNs seems ambiguous (14-15b): MNs + Plural = more than one (of type) x nP has been shifted, but plural is compositional MNs + Plural = a lot of/a great amount of MN Plural is interpreted as part of the nP

This explains the incompatibility of indefinite articles with MN, since they are mostly numerals.

 Problem: One form (plural), more than one function. This behavior of plural MNs is reminiscent of the distinction between syntactic (and exception-less) vs. lexical (and limited/restricted) properties, see di Sciullo & Williams (1987); Wasow (1977), or of the idea that word formation makes reference to two distinct domains/levels (word vs. root level) in Marantz (2001). clear separation: idiosyncratic processes are specially marked or rather they result from an operation of affixation at a very low height. For Distributed Morphology, that is the root level. (17) a. root-cycle 3 morpheme Root b. outer-cycle attachment 3 morpheme functional head 3 X Root

Merger with root implies: 1. negotiated (apparently idiosyncratic) meaning of root in context of morpheme 2. apparent semi-productivity (better with some roots than others) Merger above functional heads implies: 1. compositional meaning predicted from meaning of stem 2. apparent complete productivity Proposal: split-analysis of Plural number: Grammatical plural realizes atomization/individualization features and is on Number, see also Barner, Wagner & Snedeker (2008), Barner & Snedeker (2008). It combines with an nP. Plural marking via NumberP happens outside the nP In principle it is insensitive to the type of nP it combines with. lexical plural is on n (the categorizing head). The plural form should be separated from plural meaning (Beard 1995). Both types of plural are plural in form, but lexical plural is relatively limited in its distribution (i.e. it does not combine with all roots, but only in certain contexts and with a particular meaning): (18) a. Vivlia/books: Number 3 Pl n book Semantics: [PL (book)] b. Nera/water-pl: n 3 Pl water Semantics: [water & Pl]

The plural form in (18b) does not take the reference of the predicate as its domain, it is rather part of the nP, see Acquaviva (2004a-b).
6

Lexical plural is a realization of n, which competes with other possible n-Root combinations. The same root can combine with another [n] which is not intrinsically plural, and simply carries gender/class diacritics. The above explains: i) the distributional restrictions; ii) the non-transparent/identical interpretation; iii) the fact that certain MN are better than others The hypothesis that number can be involved in word formation processes explains the existence of pluralia-tantum as well as the fact that certain MNs receive an idiosyncratic interpretation, e.g. holidays. If plural can function as part of the nP predicate, then we expect a certain amount of interaction with other word formation processes. This becomes clear, when we turn to ASNs. 3. 3.1 (19) ASNs, Aktionsart and plural marking Aktionsart

Aktionsarten 3 Situations Properties 3 to be blond +resultative -resultative 3 3 +durativ e -durative +durative -durative build a house enter dance cough
Accomplishments Achievements Activities Semelfactives

cf. Vendler (1967)

Activities and properties are atelic, accomplishments and achievements are telic. Krifka (1989), Mourelatos (1978), Bach (1986) among others: The count-mass distinction in the domain of nouns is comparable to the telic distinction in the domain of verbs (related to cumulative reference). How do the derived nouns from these two classes behave? 3.2 ASNs

Several authors have pointed out that ASNs derived from telic predicates are CN; on the other hand, nouns derived from atelic predicates are MN, see Borer (2005), Alexiadou, Iordachioaia & Soare (2009), Barner, Wagner & Snedeker (2008): (20) a. b. (21) a. b. John pushed the cart for 2 hours/*in two hours John pushed the cart to New York in two hours/*for two hours There was one pushing of the cart to New York by John There was (*one) pushing of the cart by John Borer (2005: 239f.)
7

3.3

Atelic ASNs and plural marking

Prediction: since plural MNs are in principle possible, we expect that ASNs derived from atelic predicates (which are MNs) should be able to pluralize and still maintain their cumulative reference, i.e. there should be no type shifting involved. Observation: This is not borne out. 3.3.1 English -ing in English is incompatible with achievements and compatible with atelic predicates; accomplishments can combine with -ing, but then they receive a process-like interpretation, see Englehardt (2000), Borer (2005), Alexiadou, Iordachioaia & Soare (2009): (22) a. b. c. d. e. f. g. the sinking of the ships the falling of the stock prices the jumping of the cows *the arriving of the train *the erupting of the Vesuvius *the exploding of the balloon the building of houses

process

-ing nouns are MNs: they cannot combine with numerals, they do not tolerate plural marking and they can only occur with the definite article: (23) a. b. *the fallings of the stock prices *a/one falling of the stock prices

3.3.2 Greek Kolliakou (1995): -m- affixation is a productive process to build nouns out of verbs, but it is sensitive to the type of verbal predicate it attaches to. It namely requires a durative verb.6 (Note here that all nouns that contain -m- have neuter gender):7 -m- does not combine with stative predicates: (24) a. agapo (love-1sg) c. *agapima (loving) b. pistevo d. *pistepsimo (believe-1sg) (believing) Kolliakou (1995: 211f.)

Activities are compatible with -m-: (25) a. perpatao walk-1sg b. to perpati-m-a the walking

Accomplishments are compatible with m- and receive a durative/atelic interpretation:

See Fassi-Fehri & Vinet (2007) for a similar distribution in Arabic nominalizations. -m- and -sim- are allomorphs of the same morpheme, see Malikouti-Drachman & Drachman (1995). According to Ringe (2006: 61), -m- has its origin in the affix that used to build collective nouns in Proto-Indo-European men. According to Rijhkoff (1991), collective affixes realize nominal Aspect.
7

(26) a. htizo ena spiti build-1sg a house c. zografizo ena kiklo draw-1sg a cricle

b. to htisimo enos spitiu the building of a house d. to zografisma enos kiklu the drawing of a circle

Kolliakou notes that even if the predicate is a telic one, as in the case of accomplishments, the -m- noun is interpreted as a process, i.e. it focuses on the +durative component of the event: Evidence: Modification via "for-some-time"- phrases: (27) to htisimo enos spitiu the building of a hous ja 10 hronia/* se 10 hronia for 10 years/in 10 years

Achievements are incompatible with m-affixation: (28) a. anagnorizo recongnize-1sg b. *to anagnorisma c. ftano the recognizing arrive-1sg d. *to afigma the arriving

-m- nouns are MNs: they are incompatible with numerals: (29) *ena htisimo enos spitiu one building of a house Importantly, however, unlike non-derived MNs, they cannot pluralize: (30) *ta htisimata ton spition kratisan 5 hronia the building-pl of the house took 5 years (31) a. to kuboma tu paltu kratise 3 lepta the buttoning of the coat took 3 minutes b. *ta kubomata tu paltu kratisan 3 lepta the buttoning-pl of the coat took 3 minutes Plural marking is possible, but then the noun receives an R-(object) interpretation: type shifting. In this case, plural marking does what is supposed to do, according to Grimshaw (1990). It individualizes the noun and as a result only the R-interpretation is possible. (32) Agorasame diafora kubomata bought-1pl various buttons "We bought various buttons" How can we explain this behavior within the same language? water (kinds/bottles of water) Plural-MNs a lot of water combine with NumberP outside nP Type shifting is on n limited distribution combine with Number

Plural-ASNs (=MNs) cf. (32; Type shifting)

4. The morpho-syntax of number Two parallels between plural MNs and -m-/-ing- nouns: The idiosyncratic interpretation of plural is found only with MNs. (33) Input: Root + [ n = Plural] nP

-m-/-ing affixation is only possible with +durative predicates: (34) Input: VP + -m-/-ing on n change of category, nP incompatible with lexical plural

Both lexical plurals and -m-/-ing build new words and are not compatible with all sorts of input. We can further observe a similarity between the 'a lot of/a great amount of' interpretation of plural MN and the [+durative] interpretation of ASNs (in the area of Aktionsarten, see Rijkhoff 1991, 2002). Two places for Plurality: Plural on Number, cf. (16b):8 (35) DP (Determiner phrase) 3 Def Quantity phrase 3 numerals Number phrase 3 Plural nP

Plural on n: (36) a. n 3 Plural a lot b. n 3 n VP (+durative)9 -m-/-ing durative

-m-/-ing as well as lexical plural both realize an n head. They compete for insertion; since they have different conditions on insertion, a double marking is blocked.

This is similar to outer Aspect in the domain of verbs/eventualities, cf. Rouveret (1994), Rijkhoff (2002). (i) TP (Tense phrase) 3 Aspect phrase 3 VP 9 The VP here could contain a complex internal structure, see Alexiadou (2009), Borer (2005), Ramchand (2008).

10

5. Summary In this paper I put forth a split analysis of plural Number. 10 Grammatical plural realizes a feature on the Number head, associated with the semantics of plurality. Lexical plural is on n, and gives rise to idiosyncratic readings of MNs. References
Acquaviva, P. 2004a. Plural mass nouns and the compositionality of number. Verbum XXXVI 387401. Acquaviva, P. 2004b. Constraining inherent inflection: number and nominal Aspect. Folia Linguistica XXXVIII: 333-354. Acquaviva, P. 2008. Lexical plurals: a morpho-semantic approach. Oxford: Oxford University Press. Alexiadou, A. 2009. On the role of syntactic locality in morphological processes: the case of Greek nominals. In A. Giannakidou & M. Rathert (Hrsg.), Quantification, Definiteness and Nominalization, 253-280. Oxford: Oxford University Press. Alexiadou, A., G. Iordachioaia & E. Soare. 2009. Plural marking in argument supporting nominalizations. In P. Cabredo-Hofherr & B. Laca (Hrsg.) Layers of Aspect. CSLI Publications. Bach, E. 1986. The algebra of events. Linguistics and Philosophy 9: 5-16. Barner, D. & J. Snedeker. 2005. Quantity judgments and individuation: Evidence that mass nouns count. Cognition 97, 41-66. Barner, D, L. Wagner & J. Snedeker. 2008. Events and the ontology of individuals: verbs as a source of individuating mass and count nouns. Cognition 106: 805-832. Beard, R. 1995. Lexeme-morpheme based morphology. A general theory of inflection and word formation. New York: SUNY Press. Bloomfield, L .1933. Language. New York: Henry Holt. Borer, H. 2005. In name only. Oxford: Oxford University Press. Bouchard, D. 1998. The distribution and interpretation of adjectives in French. Probus 10: 139-183. Chierchia, G. 1998. Reference to kinds across languages. Natural Language Semantics 6: 339-405. Chierchia, G. 2008. Mass nouns and vagueness. Ms. Harvard Universitt. Corbett, G. 2000. Number. Cambridge: Cambridge University Press. Dalrymple, M. & S. Mofu. 2008. Plural semantics, reduplication and numeral modification in Indonesian. AFLA XV.

de Belder, M., N. Faust & N. Lampitelli. 2009. On inflection and derivational diminutives. Paper presented at the Roots workshop, Stuttgart June 2009.
Di Sciullo, A.-M. & E. Williams. 1987. On the Definition of Word. Cambridge, M.A.: MIT Press. Dobrovie-Sorin, C. 2009. Number as feature. Ms. Universit de Paris 7. Englehardt, M. 2000. The projection of argument-taking nominals. Natural Language and Linguistic Theory 18: 41-88. Fassi-Fehri, A. & M.-T. Vinet. 2007. Verbal and nominal classes in Arabic and Chinese. In Recherches Linguistiques de Vincennes. Gillon, B. 1992. Towards a common semantics for English count and mass nouns. Linguistics and Philosophy 15: 597-639. Greenberg, J. 1963. Some universals of grammar. In J. Greenberg (Hrsg.) Universals of Language 73113. Cambridge, Mass.: MIT Press. Grimshaw, J. 1990. Argument Structure. Cambridge, Mass.: MIT Press. Grimshaw, J. 1991. Extended projection. Ms. Brandeis University.

Number is not the only category that shows this duality. Recently it has been argued that diminutives also show a similar partition (de Belder, Faust & Lampitelli 2009).

10

11

Heycock, C. & R. Zamparelli. 2005. Friends and colleagues: plurality, coordination and the structure of DP. Natural Language Semantics 13: 201-270. Jespersen, O. 1909. The philosophy of grammar. London: Allen & Unwin. Kolliakou, D. 1995. Definites and Possessives in Modern Greek: an HPSG Syntax for Noun Phrases. Ph. Diss. University of Edinburgh. Kramer, R. 2009. Definite markers, phi-features, and agreement: a morphosyntactic investigation of the Amharic DP. Ph.D. Dissertation, UCSC. Krifka, M. 1989. Nominalreferenz, Zeitkonstitution, Aspekt, Aktionsart: eine semantische Erklrung ihrer Interaktion. In W.r Abraham & T. Janssen (Hrsg.) Tempus - Aspekt - Modus. Die lexikalischen und grammatischen Formen in den germanischen Sprachen, 227-258. Tbingen: Niemeyer. Link, G. 1983. The logical analysis of plurals and mass terms: A lattice-theoretical approach. In R. Buerle, C. Schwarze, and A. von Stechow (Hrsg.), Meaning, Use, and Interpretation of Language, 302323. Walter de Gruyter, Berlin. Lbel, E. 1993. On the parametrization of lexical properties. In G. Fanselow (Hrsg.) The parametrization of Universal Grammar, 183-199. Amsterdam: John Benjamins. McCawley, J. 1979. Lexicography and the count-mass distinction. In Adverbs, Vowels, and other objects of wonder. Chicago: University of Chicago Press. Malikouti- Drachman, A. & G. Drachman. 1995. Prosodic Circumsription and Optimality Theory. Studies in Greek Linguistics 1994. 143-161. Auf Griechisch. Marantz, A. 2001. Words and Things. Ms., MIT & NYU. Mattieu, E. 2009. On the inherent use of inflectional morphology in Ojibwe. Ms., University of Ottawa. Mourelatos, A. 1978. Events, processes and states. Linguistics and Philosophy 2: 415-434. Mufwene, S. 1981. Non-individuation and the count/mass distinction. Proceedings of CLS 7: 221-238. Ojeda, A. E. 2005. The paradox of mass plurals. Ms. UC Davis. Quine, W. V. 1960. Word and Object. Cambridge, Mass.: MIT Press. Ramchand, G. 2008. Verb Meaning and the Lexicon: First Phase Syntax. Cambridge: Cambridge University Press. Rappaport, M. & B. Levin. 1988. What to do with Theta Roles. In: Thematic Relations. Wendy Wilkins (Hrsg.), 736. San Diego: Academic Press. Rijkhoff, J. 1991. Nominal Aspect. Journal of Semantics 8: 291-309. Rijkhoff, J. 2002. The Noun Phrase, Oxford University Press, Oxford Studies in Typology and Linguistic Theory. Ringe, D. 2006. A Linguistic History of English. Volume I: From Proto-Indo-European to ProtoGermanic. Oxford: Oxford University Press. Ritter, E. 1991. Two functional categories in noun phrases: evidence from Modern Hebrew. In S. Rothstein (Hrsg.) Syntax & Semantics 26,. 37-62. San Diego: Academic Press. Rouveret, A. 1994. Syntaxe du galois. Paris: CNRS ditions. Tenny, C. 1987. Grammaticalizing aspect and affectedness. Ph.D. diss., MIT Tsoulas, G. 2006. Plurality of mass nouns and the grammar of Number. Paper presented at the 29th GLOW colloquium in Barcelona. Van Riemsdik H. 1998. Categorial feature magnetism: the endocentricity and distribution of projectionjs. The Journal of Comparative Germanic Linguistics 2: 1-48. Vendler, Z. 1967. Linguistics in Philosophy. Cornell: Cornell University Press. Verkuyl, H. 1993. A Theory of Aspectuality. The Interaction between Temporal and Atemporal Structure. Cambridge: Cambridge University Press. Wasow, T. 1977. Transformations and the Lexicon. In Culicover, P., T. Wasow, and J. Bresnan (Hrsg.), Formal Syntax, New York: Academic Press, pp. 327-360. Wilhelm, A. 2008. Bare Nouns in Dne Suine. Natural Language Semantics 16: 39-68. Wiltschko, M. 2009. On the absence of a grammaticalized count/mass distinction in Halkomelem Salish. Ms. UBC.

12