SOUTH ASIA SYNTAX-SEMANTICS NEWSLETTER Volume 6, 1998 Published by Rutgers-The State University University of Pennsylvania 18 Seminary Place 619 Williams Hall New Brunswick, NJ 08903 Philadelphia, PA 19104-6305 USA USA Dear Colleagues, We have another edition of the SASS Newsletter. As you will see the South Asian Linguistics Community has been very active. SALA 18 was held at JNU, Delhi in 1997. This year, CIEFL hosted a special GLOW, Delhi University is hosting two international seminars on Multilingualism and Agreement, University of York in hosting SALA 19. On a sadder note, South Gujarat University has closed down their Linguistics Department. We extend our sympathies to our colleagues there. Please do continue to send us your contributions. We prefer to get contributions by email since it cuts down on typing but if you do not have access to email, send it by regular mail. Others want to know about your work and you may benefit from feedback on your research. As before the newsletter is also on the web at http://www.ling.upenn.edu/sassn.html; A postscript version of the newsletter is also be available at the above URL. The html and the postscript versions are significantly more readable than this text version and we encourage you to use them. Veneeta Dayal Rajesh Bhatt Editor, SASSN Associate Editor, SASSN Department of Linguistics Department of Linguistics Rutgers University University of Pennsylvania 18 Seminary Place Room # 619, Williams Hall New Brunswick, NJ 08903 Philadelphia, PA 19104 Phone: (908) 932-6903 Phone: (617)-253-2618 email:dayal@rci.rutgers.edu bhatt@babel.ling.upenn.edu BOOKS Hindi Morphology: A word-based description Rajendra Singh, University of Montreal and R. K. Agnihotri, University of Delhi Published by Motilal Banarsidas Publishers Private Limited, Delhi as vol. 9 of the MLBD series in Linguistics ed. by Dhanesh Jain, 1997. This book provides a fairly comprehensive description of the morphology of Hindi. This description is located in the theory proposed by Ford and Singh. They question some of the most celebrated concepts of morphology and build a theory of morphological relatedness around the word as the basic unit and a set of bidirectional Word Formation Strategies. Morphology is essentially regarded as the study of relationships obtaining among formally and semantically related words. These Word Formation Strategies constitute extremely complex networks of word-relatedness. Access to a single member of a given network can activate the whole network. This book examines critically not only the concepts used in traditional morphology but also the work done in Hindi morphology during the nineteenth and early twentieth century. In addition to examining intra- and intercategorial relationships among Hindi nouns, verbs, adjectives, and verbs, the book includes sections on morphophonemic changes, minimization of morphological marks, non-morphemic morphemes and multiple affixation. The Yearbook of South Asian Languages and Linguistics 1998 Chief Editor: Rajendra Singh, University of Montreal ISSN:0971-9539 ISBN:0-7619-9231-6(US-hb);81-7036-685-2(India-hb) Contributors: Abel, Agha, Ayyar, Bhatia, Bhattacharya, Bright, Comrie, Dasgupta, Jayseelan, Kidwai, Kiparsky, Krishnamurti, Mesthrie, Nara, Peterson, Rahman, Saleemi, Shah, Tiffou, and Vasishth. Can be ordered from Sage Publications in Delhi( M-32 Greater Kailash Part-1,New Delhi,110 048,India),London(6 Bonhill Street,London EC2A 4PU,UK),or Thousand Oaks( 2455 Teller Road,Thousand Oaks,California,91320,USA) Kashmiri: A Cognitive-Descriptive Grammar Kashi Wali, Cornell and Syracuse Universities Omkar N. Koul, Central Institute of Indian Languages Published by Routledge,1997 ISBN/ISSN: 0-415-05868-6 in the Routledge Descriptive Grammars series. 408 pages Kashmiri, spoken in Kashmir, India, challenges every field of linguistics, be it synchronic, diachronic, areal, comparative, typological, modern or generative. Unlike other members of the Indo-Aryan language family, to which it is claimed to belong, its syntax, similar to Germanic and other verb second languages, has raised many significant issues within current generative theories proposed by Chomsky and other prominent linguists. The book contains extensive descriptions of Kashmiri syntax, morphology, agreement and pronominal clitics. It is invaluable as a reference and source book. Its originality lies in the fact that it presents a wealth of information on a relatively unknown verb second language. It will help to clarify certain issues in current theories Punjabi Grammar Tej Bhatia, Syracuse University Published by Routledge 1993 ISBN/ISSN: 0-415-00320-2 in the Routledge Descriptive Grammars series. Punjabi is the language of the Punjab - the land of five rivers - of northern India and Pakistan. Primarily written in three distinct scripts, a unique feature of the language is that, along with Lahanda and the Western Pahari dialects, it is the only modern Indo-European language spoken in South-East Asia which is tonal in nature. It is recognized as one of the several national languages of India and Pakistan, and approximately forty-five million people speak Punjabi as either a first or second language. This descriptive grammar accounts for the linguistic and sociolinguistic properties of Punjabi and Lahanda/Multani. It explores the standard language, giving a comprehensive account of syntax, morphology and phonology. With a descriptive, typological andc ognitive examination of the language, this is the most up-to-date, comprehensive and authoritative description of modern Punjabi to date. This volume will be invaluable to students and researchers of linguistic theory and practice. Kannada Grammar S. N. Sridhar, State University of New York, Stony Brook Published by Routledge, 1991 ISBN/ISSN: 0-415-00317-2 in the Routledge Descriptive Grammars series. 384 pages Sridhar provides a complete grammatical description of Kannada, a language of Southern India belonging to the Southern Branch of the Dravidian Family. It is the official language of Karnataka (formerly Mysore) State, India, and has more then twenty million speakers. (The Routledge writeups have been taken from http://www.routledge.com.) A reference grammar of Maithili Ramawatar Yadav, University of Kathmandu Publisher: Berlin ; New York : Mouton de Gruyter, 1996. 400 p. in the series Trends in linguistics. Documentation ; 11 Chapter 19 of `Handbook of Second Language Acquisition' Bill Ritchie and Tej Bhatia eds. "Handbook of Second Langauge Acquisition" (1996: Academic Press); Chapter 19 deals with UG and Code-Mixing with special reference to Hindi-English. Chapter 9 of `Double Case': Inflecting Postpositions in Indic and Kashmiri John R. Payne (Manchester) in `Double Case: Agreement by Suffixaufnahme' edited by Frans Plank, Oxford University Press, 1995. A form of Suffixaufnahme (roughly speaking multiple case on one nominal) involving inflecting postpositions can be found in the majority of modern Indic languages and dialects. In most of these languages, this involves a genitive markers that inflects for case cf. the Hindi ka `GEN.direct.MSg' vs. ke `GEN.oblique.MSg/GEN.MPl'. The papers is a typological survey of this phenomena in the Indo-Aryan languages. DISSERTATION ABSTRACTS A Computational Study of Transitivity Tanmoy Bhattacharya, University of Hyderabad, 1996 The central issue that the dissertation addresses is: In what form is language available to the language user? One of the claims is that: clauses stage events (or actions) like a camera staging a film/ frame. The utterance/ understanding of a sentence is a spectacle. A major part of the thesis, therefore, is concerned with the presentational aspect of a clause. Another aspect is the connection that a language user makes with the staged spectacle. We capture this through the notions of accommodation and field which constitutes a modified DRT. This is the formal tool that is used to capture this connection. The main tension of the thesis is of using a particular formal method to capture concepts that lie beyond the boundaries of the formalism.The answer to the question: How does the user get a grip on the presented clause? has been the major thread of discovery in this dissertation - the notion of salience. We claim that salience is a general cognitive apparatus through which the user computes the clause and thus gets a grip on it. We capture salience of clause through various asymmetries that are part and parcel of a clause - like topic/ focus, AGRs/ AGRo, etc. Our claims are the following in this regard: (i) asymmetries can be subsumed under a general notion of a new versus given opposition (ii) asymmetries are reflected at each level of abstraction Transitivity is the clearest of the asymmetries which represents the cognitive/ perceptive notion of salience. Psycholinguistic evidence show that for a child, the basic conceptual structure is that 'persons perform actions and things are affected by actions'. We read this as transitivity. We construct a syntactic account of transitivity where certain syntactic configurations and operations reflect extra-sentential notions like staging, scening, and event. Regarding (ii) above, a slow reading of the dissertation displays a general narrowing down of the scope from discourse structure to clausal structure to phrasal structure (chapters 1-4). Among the topics dealt with are the given/new distinction, syntactic agreement, ergatives, transitives and unaccusatives, the Split-VP hypothesis, the Obligatory Case Parameter, the three-layered Case Theory and Principle-Based Parsing. Complement Clauses in Hindi and Gujarati Ara Shah, University of Hyderabad, 1995 Thesis Supervisor: Probal Dasgupta The syntactic approach chosen for this dissertation seeks to describe complement clauses, their structure, their idiosyncrasies, and attempts to understand their behaviour in terms of wider linguistic principles. This dissertation is within the generative paradigm. However, this dissertation maintains two attitudes. The first ensures that the dissertation provides a useful account for translators in the form of an exhaustive compilation of complements selecting verbs in Hindi and Gujarati and a thorough description of the types of complement constructions. The other attitude channelizes the focus of this dissertation in a direction which attempts to raise certain theoretical issues regarding the Hindi/Gujarati language pair. Chapter 1 is the Introduction. It spells out the approach and the attitudes underlying this dissertation as well as the motivation behind them. Chapter 2 deals with finite complement clauses in Hindi and Gujarati. The major issues taken up in this chapter are (i) the nature of ki and (ii) the non canonical position of the finite complement clause. This is a phenomenon common to several Indo Aryan languages, as well as to certain Germanic languages, as is evident from the discussion. We report a number of accounts regarding this phenomenon. I argue that the complement clauses in Hindi/Gujarati are extraposed to the right in order to be licensed by the matrix verbal complex. Issues of adjacency, directionality of government and theta marking will be discussed in the course of this chapter. Chapter 3 deals with non finite complement clauses, the three main sections dealing with gerunds, infinitivals and participials. We situate our discussion of gerunds within the minimalist framework, which we will modify in order to account for the Hindi/Gujarati kaa naa constructions. We then discuss infinitivals, that is, complement clauses with a postpositional complementizer. Using Kayne (1984) as a point of departure, we account for the null subject in infinitivals and postulate a phonetically null P/C in Hindi and Gujarati. This chapter also throws light on certain difficult to classify constructions, thereby contributing to the debate on "nominal clauses". Small clause complements are discussed in chapter 4. The interesting fact about small clauses in Hindi/Gujarati is that the subject of the construction is assigned Accusative Case. In this chapter we attempt to reformulate the hypotheses offered in Mahajan (1990) and Sinha (1991) in order to account for the alternative range of interpretations that are available due to factors of animacy, specificity and definiteness. Chapter 5 aims to provide a working bilingual dictionary for a closely related language pair. In this chapter we will present the agreement patterns available for verbs in Hindi/Gujarati. The main purpose of this chapter is to collate information for designing a specific purpose dictionary, a sample of which will be presented. An index of complement selecting verbs in Hindi and Gujarati is provided at the end of the chapter. The DP Analysis of English and Bangla Noun Phrases Rajat Ghosh, M Phil Thesis, CIEFL, 1995. The focus of this thesis is on a group of nominal suffixes called "classifiers". These classifiers seem to play an important role in definiteness marking. A phrase structure tree for Bangla DP is proposed keeping in mind recent developments in syntactic theory; assuming e.g. Kayne's (1994) proposal that the Spec-Head-Complement order is universal, the thesis tries to derive the Bangla surface order by morphologically motivated movement. One of the main findings is that the apparent change in word order in some definite expressions is due to the requirement of a null definite determiner that its SPEC position be lexically filled. The DP Analysis of English and Oriya Noun Phrases Kalyanamalini Sahoo, M.Phil Thesis, CIEFL, 1996. Oriya classifiers are distinguished into two types: the -Taa type, which is generated higher than the numerals in the phrase structure tree, and the other type which is generated lower than the numerals. One characteristic of the Oriya number system is that number is realized either as a numeral or as nominal inflection but not as both, i.e, the plural marker and a number word cannot co-occur. Besides the definite article -ka, Oriya has a null definite article which has all the properties of -ka. These two forms have two features: a number feature which can trigger classifier-raising (through head-to-head movement of the classifier) and a definiteness feature which can trigger NP-raising (through SPEC-to-SPEC movement of the NP). A demonstrative and an article can co-occur in Oriya; hence it is proposed that the Demonstrative (Deix) projects to DeixP, which is generated immediately above DP. ARTICLES IN JOURNALS/PROCEEDINGS OF CONFERENCES Null Elements in Discourse Structure To appear. Butt, M. and T.H. King, `Null Elements in Discourse Structure'. In K.V. Subbarao (Ed.), Papers from the NULLS Seminar, Moti Lal Banarasi Das (talk given in the NULLS Seminar at the University of Delhi, January 1997). In this paper we address the role of null elements (pro-drop) in Urdu/Hindi as seen from a discourse structure point of view. In previous work, based on Gambhir's (1981) pioneering analyses (see also Kidwai 1997 for a more recent approach in line with our thinking), we proposed that topics be situated in SpecIP, (non-contrastive) focus be always immediately preverbal through an adjunction movement, backgrounded (de-emphasised in Gambhir's terms) information be right adjoined to IP, and that preverbal completive information simply stays in situ. This four way distinction adds an extra subdivision to Vallduvi's (1992) tripartite distinction in which focus vs. ground are distinguished, and the ground is then further subdivided into a link(=topic) and a tail(=background). Based on Urdu word order facts, we argue that in addition to two types of old information, two types of new information must be distinguished: that which is focused, and that which is merely present in order to give as complete information as possible. In showing that Topic Dislocation (external topics) involves an NP forming a chain with a null element in the clause, Dwivedi (1994) observes that only referential NPs may be dropped. This observation is confirmed by an examination of a series of discourses taken from Hindi movies such as Dilwale Dulhania Le Jayenge or Silsilla. However, the dialogs show that not simply all referential NPs are dropped: only some referential NPs are null. Within our framework the generalization as to which NPs may be realized as null in discourse is made more precise in that only old information is predicted to be dropped. In fact, only continuing topics (Yokoyama 1986) and the background information that is related to a continuing topic may be dropped. In contrast with pro-drop languages like Italian in which the licensing of null elements has been associated with the possibility of agreement markers acting as pronominals, no such explanation is possible for Urdu/Hindi, where ergative (non-agreeing) subjects may also be dropped. An analysis in terms of discourse structure as presented here, on the other hand, yields exactly the right results. International Yearbook of South Asian Languages and Linguistics vol. 1 (1998) Knowing the word trikkhe: against purism in the study of language Probal Dasgupta, University of Hyderabad Linguistics has been practising an antipurism that this paper argues should be developed explicitly. Scientific study of language presupposes the untenability of any purism, religious (sacred initial languages), historicist (past purer than decadent present), or codal (homogeneous codes purer than hybrids), forcing us into idealizations based on social assumptions about real or pure languages or items. As generativism outgrows the structural-formalist purism of an inner Form that would be defiled by Substance, generative representations begin to provide an extended interface between expression and content, not a unique interpoint. That interregion's dynamics reflects the duality of substances. Reference becomes impure not only because the chain spreads it out wide, but also because - it is argued here - paradigmatic ladders (which do for Delegation what chains do for Displacement) spread it deep down. Our case study works this out for the word /trikkhe/ which occurs only in the line /tin trikkhe nOy/ 'three threes are nine' in the Bangla multiplication table. Its referential labour co-involves but is not enslaved by /tin/ 'three'. Such a retooling of lexical reference takes the unexamined default purism out of a current debate between positions here called NC (lexical expression-content interfaces are uniquely provided for the computation so as to appear for LF as well as PF) and AM (given that much of the work is going into the PF-ward track, most or all lexicalization can occur there, feeding only skeletal scope information into LF), turning it into a more interesting debate which this paper tries not to resolve, but to highlight as a discussion-site we all need to examine more carefully, bringing the best of our methodologies to bear. Negative contexts in Hindi Shravan Vasishth, Ohio State University This is an empirical, questionnaire based pilot study of Hindi NPIs that examines van der Wouden's and others' claim that NPIs cross-linguistically exhibit sensitivity to antimorphicity (strong negative contexts), anti-additivity (medium negative contexts) and downward monotonicity (weak negative context). Using 89 informants of Hindi to provide judgements on NPI data, it emerges that although van der Wouden's claim is in principle correct, the nature of the sensitivity of Hindi NPIs to the above three kinds of licensing contexts differs significantly from the case in English and Dutch, among other languages. One conclusion this study reaches is that the stronger the negative context, the greater the tendency among speakers to affix the focus particles `bhii' and `tak', and the weaker the context, the greater the tendency for the NPI to appear without any focus particle. Constraining Argument Merger through Aspect In press. Butt. Miriam, `Constraining Argument Merger through Aspect'. In E. Hinrichs, A. Kathol, and T. Nakazawa (Eds.), Complex Predicates in Nonderivational Syntax, Syntax and Semantics Volume 30, Academic Press. In this paper, I summarize my LFG analysis of Urdu complex predicates such as the permissive and the "aspectual" light verbs, and show how it applies to the morphological causative. In particular, I formulate the conditions on argument fusion as parallel to those of syntactic control and take up Alsina and Joshi's (1991) proposal of crosslinguistic parametrization over causativization as a potential challenge to my strict formulation of argument fusion. Unlike in many other proposals, argument structure in this paper is viewed as a syntactic level of representation that captures the lexical properties of a predicate which help determine the syntactic realization of the predicate's arguments, i.e., their grammatical functions and case markings. In recasting Alsina and Joshi's analysis of causative alternations across languages, I argue (basically following Ramchand 1997) that the linking of thematic information to grammatical functions is sensitive to aspectual information associated with each argument of a predicate in the a-structure. Thus, an alternation in the case marking of the causee found in Romance, Bantu and Urdu/Hindi, which is generally associated with the degree of ``affectedness'' of the causee is accounted for in terms of the aspectual properties of the arguments. Interfaces as the Locus of Historical Change Butt. Miriam, 1997 `Interfaces as the Locus of Historical Change'. In M. Butt and T.H. King (Eds.) Proceedings of the LFG97 Conference, University of California, San Diego. CSLI Publications. See http://www-csli.stanford.edu/publications/LFG2/lfg97.html Here, I explored the origin of the "aspectual" complex predicates and traced them back to (based on Hook 1991) adverbial participle constructions formed with -tva in Sanskrit. These constructions still exist side by side with the complex predicates in modern Urdu/Hindi and Bengali and are now formed with kar and the perfect form of the verb, respectively. In the first part of the paper, I chart the proposed path of change from Sanskrit down to the current forms. In the second part of the paper, I propose that the historical change be viewed in terms of event structure in that the emergence of complex predicates should be understood in terms of a progressive bleaching of a predicate's event structure. A Lexical Semantic Account of ``Quirky'' Case in Hindi Bhuvana Narasimhan, to appear in Studia Linguistica The case canonically assigned to subjects across languages is Nominative or Ergative case. In a number of languages including Tamil, Russian, Finnish, Icelandic, Malayalam and Hindi, subjects can receive Dative case marking. This phenomenon, labeled 'quirky' or 'lexical' case-marking, is generally accounted for in terms of the association of Dative case with an argument bearing a particular thematic role in the lexical entries of individual verbs (Zaenen, Maling & Thrainsson 1985). Such an account does not account for all the data, neither does it explain why certain thematic roles should get Dative case marking. In this paper, I show that the case-marking patterns in Hindi can be accounted for in a principled manner in terms of the interaction of the aspectual characteristics of the construction, its adicity, and the relative prominence of the arguments of the verb on the Thematic Hierarchy. My account of this phenomenon is formulated within the framework of Role and Reference Grammar (Foley & Van Valin 1984; Van Valin 1991, 1993). PAPERS PRESENTED AT CONFERENCES The LSA in Chicago, January 1997 Modern Indo-Aryan direct discourse marking and the role of Persian influence Patrick Marlow, University of Illinois-Urbana The high degree of phonological similarity among many of the Modern Indo-Aryan direct discourse markers (viz. ke, kai, ki, ki, ka , ka, ak, ake, and -k) and nearly identical function and syntax makes it tempting to attribute all k-initial forms to the borrowing of Persian k. This approach is supported by the distributional facts which suggest (1) quotatives have given way to complementizers and (2) Indo-Aryan k-initial complementizers closely parallel the borders of the Mughal empire. Despite the similarity between forms, however, not all can be related to Persian k. Many derive from verbs of speaking (e.g. ak and ake). By deriving problematic forms from absolutives, my analysis explains them in a straightforward way: they are part of an earlier pattern of DD-marking which has given way under Persian influence. This analysis is supported by the distributional facts: complementizers form an innovative core; shifted forms provide a transition area, and true verbal quotatives remain on the periphery. This analysis has the consequence of reducing all South Asian DD-marking to three types, viz. verbal (e.g. Bengali bole), deictic (e.g. Sanskrit iti, Marathi asa), and relative (e.g. Bengali je, Persian k). SALA 18 in Delhi, at Jawahar Lal Nehru University, Delhi, India, January 1997 Event Structure: Internal and External Miriam Butt, University of Konstanz In this talk I take a further look at the "Aspectual complex predicates" in Urdu/Hindi as described in Butt 1995 and try to account for the differences between light verbs like paR `fall' and le `take' on the one hand, and verbs like lag `begin' on the other. I suggest that the distinction can be understood if one views the light verbs as "event modificatory" and the lag `begin' as "event embedding". That is, the light verbs do not introduce events of their own, but serve to focus the points of inception and completion of the main verb's event. On the other hand, the effect of verbs like lag `begin' is to take an event and embed it under a beginning event. In order to model this contrast, a distinct separation between the notions of an internal event structure that may be modified, and the placement of the event within a larger situation is needed. Taking Smith's (1991) proposals for Discourse Representation Theory (DRT) as a point of departure, I incorporate distinct representations of internal and external event structure. Monotonicity constraints on Negative Polarity in Hindi Shravan Vasishth, Ohio State University This study addresses the issue of admissible licensing constraints on Hindi NPIs in terms of downward monotonicity and classifies NPIs into three distinct types, based on whether the NPI in question can be suffixed by the particles `bhii' (even, also) or `tak' (even, until/by). Further, three distinct licensing environments, constituting increasingly stronger negative contexts, are identified, and extensive empirical support is provided to claim, inter alia, that NPIs marked by `bhii' are allowed in the weakest possible negative context, while those marked by `tak' appear only in a relatively stronger negative context (anti-additive contexts). (To appear in the proceedings of SALA 18) Annual meeting of the Linguistics Association of Great Britain, University of Edinburgh, Edinburgh, April 1997 Aspectual Complex Predicates, Passives and Disposition/Ability Miriam Butt, University of Konstanz In these two talks (a version of this paper was also presented at Sinn und Bedeutung the Annual Meeting of the German Society for Semantics, at Humboldt-Universitaet zu Berlin, Berlin, December 1997), I looked at some Urdu complex predicates which receive an entirely unexpected reading in the imperfect. (1) Nadya gaarii calaa le-tii hai N.Nom car.Nom drive take-Impf.F.Sg be.Pres.Sg `Nadya does/will drive a car.' This reading entails not only that Nadya is able to drive a car, but that she actually does so. It contrasts with the run-of-the-mill modal construction (``sak" `can') in the following way: while the modal could be used to describe a context in which Nadya in principle knows how to drive a car, but hasn't done so for about 20 years, this could not be the case for (1). The ``le"-construction entails that Nadya does in fact excercise her ability. The closest analog in the literature to the semantics of this ``ability" construction are found in Lawler's (1973a,b) descriptions of existential or ``dispositional" generics in English, such as in `My pet toad will eat flies', which predicates not only the toad's potential to eat flies, but also asserts that the toad does in fact engage in this activity. A further twist to the story is provided by passives (most commonly used in the negative) which appear to provide a reading very close to the disposition/ability reading of (1). In the passives, furthermore, this reading is not limited to the imperfective, but also appears in perfects and futures. Lawler argued for a modal interpretation of these readings, and I follow him in presenting an analysis of the constructions in terms of Kratzer's (1981) theory of modality in which the complex predicates are analyzed in terms of conditional necessity, while the ``passives", which actually turn out to be complex predicates as well, receive an analysis in terms of a modal of ability. Ordered Concept routines for non-conceptual meaning: the case of hi Richard Breheny, University College, London richardb@ling.ucl.ac.uk The Hindi particle hi produces the meaning of `only'. Treating it as the morphological realization of focus and then using an analysis like Rooth (1992) gives us the correct semantics. In association with negation, however, hi is interpreted like `even'. This reading is not predicted by an account such as Rooth's. Breheny uses `ordered concepts' to derive the unexpected `even' reading from the semantics of hi and negation (a version of this paper was also presented at the 9th Student Conference in Linguistics, Eastern Michigan University, Ypsilanti, May 1997). The 33rd meeting of the Chicago Linguistics Society, 1997 Fluid Ergativity in Gujarati and Kashmiri and the Notion of Suspension Peter Edwin Hook, University of Michigan and Omkar N. Koul, Central Institute of Indian Languages An ergative-absolutive pattern in subject and object case-marking and in verb agreement is found in the tensed clauses of most Indo-Aryan languages spoken west of Bihar. In all such languages the choice of subject case is an automatic consequence of logically prior choices in predicate type (transitive vs. intransitive), tense (past or perfect vs. others), and aspect (non-durative vs. durative): That is, the subjects of all transitive predicates in a non-durative aspect of the past or perfect get the ergative case and the verb agrees with the object (or assumes a default third person masculine singular form): In some of these languages, however, in a restricted set of constructions, the automatic operation of ordinary subject-marking rules is suspended and the use of ergative case or postposition is determined by the interaction of syntactic, semantic, and pragmatic factors. The Gujarati inceptive provides an example of such a suspension. The transitive predicate ma:ND 'build' has been grammaticalized as an auxiliary meaning 'begin to'. When the predicate complement of ma:ND is itself intransitive normal rules are suspended and a series of nested conditions comes into play to determine the subject's case-marking: (3) a. If the complement's predicate is a passive, use nominative for the subject, or b. if the subject refers to a source of experience or a sensation, use the nominative, or c. if the predicate is one that selects only "unaccusative" subjects, use the nominative. d. Otherwise, if the subject is sentient and acts volitionally, use the ergative, but e. if the subject is sentient and acts non-volitionally, use the nominative. f. If the subject is non-sentient and acts independently or unexpectedly, use the ergative. g. Otherwise, use the nominative. In the paper we show that an etymologically unrelated construction in Kashmiri also leads to the suspension of its automatic rules of subject marking and that the set of conditions and constraints governing the appearance of ergative or nominative case that comes into play is largely parallel to the Gujarati set. The LSA in New York, January 1998 Grammaticalization through areal & typological pressure: The case of Meithei Shobhana L. Chelliah, University of North Texas Grammaticalization of the second verb in verb-verb sequences, through the semantic bleaching of the second verb, is an areal feature of South Asia that has spread from the Dravidian to the Indo-Aryan languages. I will show, through examples from the Tibeto-Burman language Meithei, that the diffusion of grammaticalization has further spread from the Indo-Aryan to the Tibeto-Burman languages. Unlike Indo-Aryan languages, however in addition to semantic bleaching, the grammaticalization has gone further in that verbs have become reduced phonologically and have become affixes. I argue thet this `decategorialization' is due to the typological tendency for agglutinative or polysynthetic suffixing languages to contain little or no compounding. In other words, typological pressure regulates contact induced change. On the role of Tense in NPI-licensing Rita Bhandari, SUNY-Stony Brook The behavior of NPIs in Hindi vs. English contrasts along two dimensions. First an NPI may be licensed in subject position in Hindi, but not, in general, in English. Second, negatives in subordinate infinitives can take matrix scope in Hindi and license matrix NPIs, unlike in English. I propose that these distinctions between English and Hindi follow from two differences in the respective grammars. First, they differ in the hierarchical ranking of functional projections - AgrSP is outside TP in English while in Hindi all Agr projections are inside TP. Following Zanuttini (1995)'s proposal that NegP is parasitic on TP predicts that subject NPIs are licit in Hindi, but unlicensed in English. Second the Hindi negative has head properties and undergoes covert movement yielding wide-scope negation. AgrP in UG: Elimination or special dispensation Rakesh M. Bhatt, University of Tennessee Chomsky 1995 eliminated the Agr projection from the clause structure of natural language grammars. On his view, SpecTP is where case/EPP is checked by the a subject DP in the overt syntax, and the object checks its features at LF in the [Spec,vP]. In this paper, it is argued that AgrP does exist in the inventory of thos languages that make use of it. I present evidence from the distribution of subjects and objects in Kashmiri to argue for the existence of AgrP, required for nominative case checking. I further argue that the existence of AgrP is also warranted to account for quirky constructions in Icelandic and Hindi, which is accomplished by dissociating case and D-feature licensing. Tulu relative clause acquisition in a cross-linguistic perspective Shamitha Somashekar (Cornell University), Claire Foley (Morehead State University), James W. Gair (Cornell University) Earlier acquisition work has shown that free relative clauses may be developmental precursors to lexically headed forms. In this paper, we investigate this trend in a study of Tulu structures like (a) and (b) (the English counterparts are shown) (a): the boy who ate the apple is standing (b): which boy ate the apple, he is standing Sixty children acquiring Tulu as a first language were tested in a controlled production study. Results show that despite the surface `transparency' of forms like (b), children show an early preference for structures like (a); Children productively convert forms like (2) into forms like (1). Results lead us to propose a new notion of transparency; Children more quickly assemble structures that are more transparent not in surface form, but in their reflection of UG principles. The Extraordinary GLOW at CIEFL Hyderabad, January 1998 Crosslinguistic Variability of Laws and Their Interaction: Parameter Setting or Constraint Prioritization K. P. Mohanan, National University of Singapore The paper discusses anaphora facts from Kannada and argues that prioritization of principles as in Optimality Theory is not sufficient to express the interaction of the laws governing the Kannada anaphoric system. It further argues that differences in anaphoric systems are best expressed as variations in the way universal building blocks of anaphora are assembled in particular languages. To concretize the idea of law assembly, a list of universal building blocks is suggested. Aspect and Event Structure in Vedic Paul Kiparsky, Stanford University The Vedic tense-aspect system can be accounted for, and its tense categories identified with those known from other languages, if we make two refinements of the basic Reichenbachian approach. The first is to specify, as part of the representation of certain tense/aspect categories, a particular mapping of the verbal predicate's event structure into the parameters that define their temporal relations. The second is to adopt from morphological theory the principle that general categories are blocked by specific categories. This approach solves some classic problems of the English tense/aspect system as well, including tense shift in subordinate clauses and the ``present perfect puzzle''. Only in Hindi Shravan Vashisth, Ohio State University The semantics of `only' in Hindi is examined and evidence is provided for the monotone decreasing analysis of only defended in Horn (1996) but rejected by Atlas (1996). It is argued that Horn's analysis cannot be abandoned totally, as it fully accounts for the Hindi facts. Causation and Reflexivity in Kannada: Evidence for Post-syntactic Morphology Jeffrey Lidz, IRCS, University of Pennsylvania This paper examines a valency puzzle among change of state (COS) verbs in Kannada whose solution implicates a model of grammar in which syntactic structure is not projected from the meaning of a verb, but rather is generated independent of lexical information. The syntax and the lexicon must ultimately meet certain correspondence conditions in order for a sentence to be grammatical. The analysis of COS verbs also leads to an analysis of the verbal reflexive which entails that morphology applies to the output of the syntactic derivation. Morpheme-Internal LF Representations Tara Mohanan, National University of Singapore This paper argues that, contrary to the general position in GB models extending to Minimalism, the LF of single morphemes must contain a certain degree of compositionality, allowing for the expression of grammatically relevant meanings such as causation, state, change of state, and so on. This proposal entails projecting part of the LCS into LF. The substance of the proposal is that the calculation of the meanings of morpheme combinations is sensitive to certain aspects of morpheme internal meanings. The LF representation of the adverbial almost and its counterpart -aar in Malayalam is discussed in the context of this proposal. Kannada Clause Structure R. Amritavalli, CIEFL, Hyderabad The paper sketches the functional architecture of the Kannada clause, with a focus on negation. Kannada has moved from `synthetic' to `analytic' verb forms, dissociating a Tense-Agr complex from a Neg/Modal one. An earlier verb sequence [Vstem-infix(T/Neg/Modal)-Agr] has split into [Vstem-T-Agr] or [Vinf/gerund-Neg/Modal]; i.e. `matrix infinitives/gerunds' occur with main clause negation. Tense, though covert, is argued to be present. The semantics of Kannada gerunds and Kannada infinitives is investigated and Stowell (1982)'s differentiation of gerunds and infinitives w.r.t. tense is argued not to yield the right results. Compounding in Hindi and Bangla Rajendra Singh and Probal Dasgupta, University de Montreal and University of Hyderabad Bangla and Hindi provide particularly strong evidence that perhaps the only way to make sense of the apparent compositionality and the nityatva of words described in both traditions (classical Indian generative, contemporary generative) as compounds is to give theoretical recognition to the fact that in words of this type the freedom attributed to their constituents is remarkably limited, if not non-existent. The theoretical premises for the argument sketched in the paper include (a) a reading of lexical integrity associating it with the core generative principle that a word is a minimal free expression capable of supporting interpreting and (b) an emphasis on the context-sensitive use of commutation as a structure detector. The empirical argument works from morphologically clear to unclear examples in the manner of the parametric carry-over of notions and results from implicit to explicit morphological subsystems. Light Verb Raising, Empty Prepositions, and Zero Derivation P. Madhavan, CIEFL, Hyderabad The paper notes the absence of certain kinds of zero derivations in Malayalam and hypothesizes that this follows from the lack of zero [-N] categories. This approach predicts that Malayalam will have no instances of dative shift in double object constructions, which is indeed the case. Questions, Quantifiers, and Polarity in Malayalam K. A. Jayaseelan, CIEFL, Hyderabad Malayalam has two suffixes -um and -oo, that signify conjunction and disjunction respectively. When these suffixes are added to question words, they yield negative polarity quantifiers and existential quantifiers respectively. The explanation for the wh + `OR' yielding an existential quantifier is as follows: the wh-word is simply a variable. A variable plus disjunction is interpreted as an infinite disjunction of variables, which yields the meaning of an existential quantifier. As for the negative polarity quantifier, it is noted that -um, besides being a conjunction marker, also means `even' and `also'. The author uses Lee & Horn (1994) and Israel (1996)'s analysis to derive the appropriate semantics for wh + -um. Rightward Ho! Steven Schdufele, Soochow University The author offers a typology of structural focus and complement-head order and provides a critical discussion of antisymmetry, Rightward movement, and the syntax/pragmatics interface. It argues that South Asian languages are best described as head-final and for a rightward movement in which a focussed constituent moves rightward from its base position into a V-adjoined/ V-incorporated slot. The Verb Typology Workshop at the Hyderabad GLOW, January 1998 In Support of AGR as a Functional Category K. V. Subbarao, University of Delhi The paper proposes that Agr projections are needed in UG to account for data in languages such as Mizo, Hmar and Paite to explain the agreement morphology in the V complex. It is further proposed that nonsubcategorized arguments too participate in adposition incorporation in Mizo, Hmar and Paite. Further incorporated postpositions have a `transitivizing' effect which has implications for agreement in split ergative constructions in these languages. Lexical Logical Form and Verb Typology: A case study from Malayalam KP Mohanan and Tara Mohanan, National University of Singapore The authors argue that the interaction between the meanings of morphemes (LCS) on the one hand, and their syntax and phrasal LF on the other, is mediated through Lexical Logical Form (LLF), the rest of the information in LCS being invisible to syntax and phrasal LF. The syntactically relevant verb classifications in Vendler 1967, Levin 1993 and others are encoded in terms of LLF. The Structure of Maithili Verb Morphology Yogendra P. Yadav, Royal Nepal Academy, Kathmandu The author provides a detailed description of the rich agreement morphology found in Maithili. In Maithili, the verb may agree with more than one argument at a time. Further it may also agree with non-arguments such as possessors of arguments. All the various possibilities are explored and a very complete, hitherto unavailable, picture of Maithili agreement appears. A typology of counterfactual marking in Modern Indo-Aryan languages Rajesh Bhatt, University of Pennsylvania/MIT This paper intends to account for certain generalizations about counterfactual morphology in the modern Indo-Aryan languages (henceforth MIA). It discusses the means of marking counterfactuality in MIA, develops a typology of counterfactuals in MIA, and extends the formal analysis of the semantic contribution of tense-aspect morphology developed in Iatridou (1996, 1997). The generalizations discussed are: 1. The imperfective participle is an ingredient of the morphology in counterfactuals in many modern Indo-Aryan language. 2. In a counterfactual environment, the imperfective participle does not contribute its usual set of interpretations. 3. There is no periphrastic tense marking (such as the Hindi hai/thaa `Prs/Pst') in the counterfactual. If a language uses periphrastic tense markers with the imperfective participle to form the present/past habitual, it does not use them to form the counterfactual. The Four Copulas in ODia Bibhuti Bhusan Mahapatra, CIEFL, Hyderabad ODia has four copulas - an `equative', an `existential', a past tense copula, and a present tense copula. The author discusses the factor that condition the appearance of one of these copulas over another. He argues that it is not the structural configuration alone which predicts the choice of a copula in a copular sentence, the time reference of the sentence also plays a role. WCCFL 17 at the University of British Columbia, Vancouver, February 1998 Ability Modals and their Actuality Entailments Rajesh Bhatt, University of Pennsylvania/MIT (a version of this paper was also presented at the Indian GLOW in Hyderabad) The paper argues that there are two kinds of ability attributions realized by can/be able: one that means something `manage to' and one that means something like `had the ability to'. The distribution of the two kinds of ability attributions is related to the availability of generic readings. The `had the ability to' ability attribution appears in generic environments, while the `managed to' ability attribution appears in non-generic environments. In languages where imperfective aspect appears on generic sentences (sentences in the perfective aspect lack generic readings), we find that when the ability modal occurs with imperfective aspect, there is no actuality entailment (cf. the Hindi sak-taa `CAN-Habitual'). When the ability modal occurs in the past perfective, there is an actuality entailment (cf. the Hindi sak-aa `CAN-Perfective'). Complex Predicates workshop at the DGfS in Halle, March 1998 An LFG View of the Syntax/Phonology Interface of Complex Predicates Miriam Butt, University of Konstanz The paper presents an integrated treatment of the syntax and phonology of complex predicates within LFG. The phonological properties of complex constructions are integrated with their syntactic analyses using LFG's projection architecture. A level of phonological structure is added to the pre-existing levels and this phonological structure is projected from the c(onstituent)-structure. The Phonology of Bengali Complex Predicates Jennifer Fitzpatrick-Cole and Aditi Lahiri, University of Konstanz The papers notes that certain word orders which might seem to be ambiguous between a complex predicate readings and some other syntactic construction are disambiguated by the phonology. So in Bengali, the verb sequence mere p^haela `beat throw' can be interpreted as either a beating followed by throwing or as a complex predicate, where `throw' does not contribute its meaning as a main verb. Instead the sequence means `beat to death'. The non-complex predicate case is assigned two phonological phrases, one for `beat' and one for `throw'. On the other hand, in the complex predicate case, both the verbs are part of the same phonological phrase. A detailed analysis of the interaction of this phenomena with segmental assimilation, focus, reduplication, and clitics is provided. Complex Predicates as Polarity Items Rajesh Bhatt, University of Pennsylvania/MIT The paper argues that certain complex predicates in Hindi-Urdu are polarity sensitive. Treating them as polarity sensitive nature enables us to explain facts about their distribution. Their licensing is most naturally expressed at a semantic level of representation. The passive/intransitive of inability behaves like a negative polarity item. Aspectual complex predicates behave like positive polarity items. The paper raises the question of exactly where the polarity sensitivity of these complex predicates marked.