Tanmoy Bhattacharya CASL/ University of Delhi
[email protected]
Structural Contextual Perspective The presence of a classifier when studied within its syntactic and semantic contexts, reveals a number of larger patterns that obtains significant descriptive typological and syntactic generalisations
PROJECTION [Syntactic] How the presence of classifier may define the modified structure of the noun phrase MICRO MACRO [Typological] How the internal structuring of the noun phrase may derive hitherto unexplored typological parameters REFERENCE [Semantic] How the positioning of the classifier and the micro structure derive larger semantic-pragmatic issues of nominal reference
The very first work suggesting a DP structure for noun phrases in Indian languages, based on the presence of the classifier, was Bhattacharya and Dasgupta (1992/ 1996), where the following syntactic typological finding was reported: GENDER-CLASS PARAMETER Languages are bifurcated into Gender or Class languages A term Badge was suggested to capture the correspondence between the two; in class languages it houses the classifier and in Gender languages it holds the Gender and Number information
DP
Spec
BP
Spec
D
B NP phul ‘flower’
ʈa CLA
Definiteness
Other research followed soon after: Bhattacharya, Tanmoy.1995. A Computational Study of Transitivity. University of Hyderabad [PhD] Sahoo, Kalyanmalini. 1997. Classifiers in Oriya. CIEFL [MPhil] Bhattacharya, Tanmoy. 1999. Structure of the Bangla DP. University College London [PhD] Thakur, Anil. 2000. The DP in Hindi. University of Delhi [PhD] Ghosh, Rajat. 2001. DP in Bangla and Asamiya. Tezpur University [PhD] Borah, Gautam. 2008. The Indeterminacy of the Bare Noun: The Case of Assamese. Trondheim [PhD] And others ...
In Bhattacharya (1998) 0nwards, another parameter , Greenbergian in nature, comes to light: Relating OV/VO with DP-internal Movement Head-final languages show PHRASAL movement inside the DP whereas Head-non-final languages show HEAD movement inside the DP: du-ʈo gach gach du-ʈo two-CLA tree tree two-CLA ‘two trees’ ‘the two trees’
The Three-layered DP Structure DP D XP (= QP) X (=Q) NP X = NumP (Ritter 1988 for Hebrew, French) X = QP (Giusti 1991 for Romanian, Löbel 1989 for German ) X = KP (Sigúrdsson 1993 for Icelandic; Tang 1990 Chinese) X = ArtP (Santelmann 1993 for Swedish) X = BP (Bhattacharya and Dasgupta (1992) X= AgrGENP (Siloni 1997 for Hebrew)
DP D
QP
Spec
Q’
Q NP du-ʈo AP two-cla notun kichu new some
NP saɽi Saris
notun saɽi du-ʈo ‘the 2 new saris’
Summary of Part 1 and 2: Syntax: a noun phrase with a Classifier projects a DP
The classifier itself is a head or a part of a head but not
the D D interacts with the classifier –holding head syntactically and semantically Typology: Class Vs Gender languages Parameter Phrasal (OV) Vs Head (VO) DP-internally Semantics?
Dasgupta (1983) noted that addition of a classifier
makes the noun definite:
boi-ʈa gach-ʈa baɽi-ʈa book-CLA tree-CLA house-CLA ‘the book’ ‘the tree’ ‘the house’ Although this is true, the indefinite versions makes the picture complicated: æk-ʈa boi etc. one-CLA book ‘a/ one book’
Bhattacharya (1999) proposed a finer distinction of the semantic effect obtained when the Classifier is to the right and to the left of the N, by showing that Bangla makes use of Specificity rather than a course-grained Definiteness, and that, it is obtained configurationally: (i) du-ʈo [notun saɽi ] 2-CLA new saris ‘two new saris’ (ii) [notun saɽi] du-ʈo the (specific) two new saris
Specificity presumes an identified discourse referent (Heim (1982), Fodor and Sag (1982), Enç (1991), Ludlow and Neale (1991)
Dayal (2013) recently argued for a definiteness reading of the [NP-CLA] order on the basis of certain tests as follows: (i) Partitive Specificity tin-ʈe chatro eʃe chilo. du-ʈo chatro boʃlo [CLA-NP] 3-CLA student come.cp was. 2-CLA student sat ‘3 students had come. 2 students sat down’ #chatro du-ʈo boʃlo [NP-CLA] Maximality, a sign of definiteness, blocks the partitive reading in the second alternative.
Two reasons why more needed to be said: (a) However, here partitivity is forced by the saliency of sequence of the events chosen (‘coming’ followed by ‘sitting’), a non-salient sequence allows the specific order: tin-ʈe chele elo. bẽʈe (chele) du-ʈo geye uʈhlo 3-CLA boy came. Short (boys) 2-CLA sing.cp rise.pst 3 boys came. The two short ones started to sing. (b) This is in fact the test used to favour specificity in Heusinger (2011)
(ii) Referential Specificity Indefinite Specific Indef
a. jodi du-ʈo chatro aʃe, ami pɔrabo if 2-CLA student comes, I teach.will ‘If the 2 students come, I will teach’ b. jodi chatro du-ʈo aʃe, ami pɔrabo (exactly 2) Def ‘If the 2 students come, I will teach’
(a) The regular indefinite reading with [2-CLA NP] is possible only if the classifier is jɔn, ʈo obtains a particular reading (b) The specific indefinite reading is not available: jodi du-jon chatro aʃe, ami pɔrabo. ... #tader nam holo jonaki ar robi Their name be Jonaki and Robi (c) The definite order giving the “exactly” meaning to work the NP must have anaphoric/ discourse reference; also if there are more than two, the sentence does not become infelicitious, and in fact is the stronger interpretation as long as the 2 specific students are a part of the group of >2.
It is claimed that (ii) shows maximality, giving an exhaustive reading (only 2 red books); whereas no such reading obtains in (i): (i) oi du-ʈo [ lal boi ] that two-CLA red book (ii) oi [lal boi] du-ʈo that red book two-CLA
However, not enough attention has been paid for the constituents of the noun phrase. For example, a focus on the Dem will neutralise the difference between the two orders, and obtain a contrastive reading between proximality/ distality of the objects for both as follows: oi du-ʈo [ lal boi ] oi [lal boi] du-ʈo ‘those 2 red books as opposed to these 2 red books’ Maximality is not settled beyond doubt
Heusinger (2011): i. Referential Specificity ii. Scopal Specificity iii. Epistemic Specificity iv. Partitive Specificity v. Discourse Prominence vi. Noteworthiness Specificity vii. Topical Specificity None of these tests work in Bangla, showing that indefinite DPs cannot provide specific interpretation
This test claims that a paritive reading will be derived if a specific description of the DP is at work, this test works for Bangla even if we use a NP-Num-CLA order: 50-jon chatro klas-e ɖhuklo, meye-du-ʈo boʃlo 50-CLA student class-loc entered, girl-2-CLA sat down ‘50 students entered the class, the two girls sat down’ This is ok only if the set of 2 girls is a maximal set, thus maximality does not block specificity, as we still obtain a partitive reading: Maximal yet Partitive
Summary of Part 3:
Specificity in Bangla involves a pragmatic notion of “referential intention” of the speaker, it shows up in various ways as partitivity, presuppositionality, or maximality within the scope of a partitive expression.
SYNTACTIC The DP structure of languages with Classifiers is an extended projection of the classifier head The constituency provided by the DP derives invisible typological generalisations TYPOLOGICAL Class versus Gender languages DP-internal Head/Phrase movement related to VO/OV parameter SEMANTIC Classifiers induce a pragmatically driven referential intention of the speaker
