Pre-publication version (not quite identical) of Borschev, Vladimir and Barbara H. Partee (2002) The Russian genitive of negation: Theme-rheme structure or perspective structure? Journal of Slavic Linguistics 10, 105-44. The Russian Genitive of Negation: Theme-Rheme Structure or Perspective Structure?* Vladimir Borschev and Barbara H. Partee 0. Introduction..........................................................................................................................................................2 1. Babby on “Declarative” and “Existential” Sentences..........................................................................................6 1.1. The Scope of Negation. Theme and Rheme. Problems..............................................................................6 1.2. Sentences with the Verb byt’ ‘be’ and Referential Subject..............................................................................10 2. Existential Sentences and the “Genitive Rule”..................................................................................................12 3. Existential Sentences: LOC as Perspectival Center...........................................................................................16 3.1. THING and LOC.............................................................................................................................................17 3.2. Perspective................................................................................................................................................18 3.2.1. The Notion of Perspective and Its Role in Existential Sentences........................................................18 3.2.2. Perspective vs. Theme-Rheme Structure..................................................................................................19 3.2.3. Issues about the Sources of Presuppositions............................................................................................23 3.3 Notes about Syntax...........................................................................................................................................24 4. Concluding Remarks..........................................................................................................................................25 References...................................................................................................................................................................26 Abstract. In recent work we have come to challenge assumptions that we shared (Borschev and Partee 1998a) with Babby (1980) concerning the role of Theme-Rheme structure in accounting for the nominative-genitive alternation in negated existential sentences (the NES construction, in the terms of Babby (1980), the classic work which we are building on). The challenge is exemplified most clearly in our “kefir example”: (i) [Ja iskal kefir.] Kefira v magazine ne bylo. [I looked-for kefir.] Kefir-GEN in store NEG was -m-sg -n-sg ‘[I was looking for kefir.] There wasn’t any kefir in the store.’ It is an important part of the explanatory structure of Babby 1980 that in sentence (i), the Theme is v magazine and the Rheme is kefir- [byl-]. Babby takes Theme-Rheme structure to be crucial for determining the scope of negation, and scope of negation to be a necessary condition in licensing the occurrence of the genitive of negation. But arguments from word order, intonation, and pragmatics have convinced us that kefira in example (i) must be considered (part of) the Theme, and not the Rheme. We * We are grateful to Eva Hajičova, Petr Sgall, Jiří Hana, and Tomáš Hoskovec for permission to take parts of this paper from Sections 1, 2, and 4 of Borschev and Partee 2002a. Fuller acknowledgements appear in that paper. In this paper our first acknowledgement is to Leonard Babby, both for inspiring this work with his foundational work on the genitive of negation in Babby 1980 and for both encouragement and critical suggestions for our work on the topic over the last several years. Thanks to Ji-yung Kim for assistance in her role as Research Assistant on the NSF grant mentioned below and her help with the formatting of this paper. This work was supported in part by a grant to the first author from the Russian Foundation for Basic Research, Project No. 96-06-80315a, in part by a COBASE grant from the National Academy of Sciences in 1997 to the second author for collaboration with the first author on the project “Towards an Integration of Formal and Lexical Semantics: Meaning Postulates and Fine-grained Sortal Structures”, and in part by the National Science Foundation under Grant No. BCS-9905748 to the both authors for the collaborative project “Integration of Lexical & Compositional Semantics: Genitives in English and Russian”. BorschevParteeJSL2002.doc 1 now argue that independent of Theme-Rheme structure there is a relevant “perspective structure”, a kind of diathesis choice, allowing a proposition involving a suitable verb to be structured with either of its two arguments as “Perspectival Center”. In a locative DS, the sentence predicates “being in a certain location” of the “thing” argument, whereas in an ES, the sentence predicates “having a certain thing in it” of the “location” argument. The theoretical status of such a layer of structure remains in need of further investigation. 0. Introduction We agree with Comrie (1980) in his assessment of the significance of Babby’s (1980) account of the interaction among the formal properties of the subject genitive of negation construction and its semantic and pragmatic properties: where others had given lists of properties that facilitate or impede the use of genitive of negation, “it is in providing a more general solution to these problems that Babby’s main contribution lies.” (p. xiii) After reviewing the principal generalizations identified by Babby in that landmark work, Comrie notes that “related semantico-pragmatic properties that are often treated as arbitrary in earlier analyses are shown to be special cases of, or at least correlated with, the above-mentioned generalizations.” (p.xiv) Our own work on this topic, starting with Borschev and Partee 1998a, has been built on the foundations of Babby’s work; our principal goal has been to try to formalize some of the semantic and pragmatic insights found in Babby’s work and to integrate them with some of the insights into lexical semantics that have come from the work of Padučeva and other members of the Moscow semantic school. As our work has progressed, we have found ourselves making modifications in certain points of Babby’s analysis. One such point of difference concerns the role Babby attributes to Theme-Rheme structure in determining the scope of negation. We now believe that the crucial distinction is not to be found in Theme-Rheme structure but needs a new structure which we call Perspective Structure. The nature of this structure and the differences between our current position and Babby’s are described below. We consider the issues far from settled, and in any case we consider our modifications as friendly amendments to Babby’s account, with which we have far more points of agreement than of disagreement. We will follow Babby’s (1980) terminology of “negated declarative sentences” (NDS), for the sentences with nominative subjects, and “negated existential sentences1” (NES), for those with genitive “subjects”, as in (1a) and (2a) (his (81a-b), from Ickovič 1974). We give the corresponding affirmative sentences (ADS and AES) in (1b) and (2b)2. 1 The most common term in Russian corresponding to Babby’s “existential sentences” is bytijnye predloženija; see Arutjunova (1976, 1997), Arutjunova and Širjaev (1983). The contrasting term “declarative” is not widespread in the English-language literature, and is perhaps not the best choice, but there is no standard alternative term. The Russian alternative preferred by Arutjunova and Padučeva is “sobytijnye” predloženija, ‘eventive sentences’, but that does not apply comfortably to the stative “locative” sentences which frequently form minimal pairs with ES’s. 2 In glossing our examples, we use the following abbreviations: NOM nominative SG singular GEN genitive PL plural ACC accusative 1 first person M masculine 2 second person F feminine 3 third person N neuter We use boldface to highlight the relevant occurrences of NOM and GEN on nouns and N.SG on non-agreeing BorschevParteeJSL2002.doc 2 (1) NDS (a) Otvet iz polka ne prišel. Answer from regiment NEG arrived NOM.M.SG M.SG ‘The answer from the regiment has not arrived.’ ADS (b) Otvet iz polka prišel. Answer from regiment arrived NOM.M.SG M.SG ‘The answer from the regiment has arrived.’ (2) NES (a) Otveta iz polka ne prišlo. Answer from regiment NEG arrived GEN.M.SG N.SG ‘There was no answer from the regiment.’ AES (b) Prišel otvet iz polka. Arrived answer from regiment M.SG NOM.M.SG ‘There was an answer from the regiment.’ As Babby notes, the affirmative ADS and AES sentences differ (obligatorily) in the order of subject and verb, while in the negative sentences, where the difference between NDS and NES is marked by case, the word order can vary; we return to this important point later. Here are some more standard examples.3 (3) a. NDS: Stok talyx vod ne nabljudalsja. Runoff melted water NEG was.observed NOM.M.SG M.SG ‘No runoff of thawed snow was observed.’4 b. NES: Stoka talyx vod ne nabljudalos’. Runoff melted water NEG was.observed GEN.M.SG N.SG ‘No runoff of thawed snow was observed.’ (= There was no runoff.) (4) a. NDS: *(#) Somnenija ne byli.5 Doubts NEG were NOM.N.PL N.PL b. NES: Somnenij ne bylo. Doubts NEG were GEN.N.PL N.SG ‘There were no doubts.’ (5) a. NDS: Lena ne pela. Lena NEG sang NOM.F.SG F.SG ‘Lena didn’t sing.’ verbs. We do not gloss irrelevant morphology. 3 Example (Error! Reference source not found.) is from Ickovič (1974) (cited in Babby 1980), and example (Error! Reference source not found.) from Babby 1980. After many years of work on GenNeg, a large collection of examples has in a sense become common property. We largely take our examples from this collection (sometimes with variations), drawing principally on examples cited by Ickovič (1974), Babby (1980), Apresjan (1980, 1985), and Padučeva (1992, 1997). 4 This is Babby’s translation (Babby 1980: 59); in most cases he translates nominative subjects in NDS’s with the definite article, and we believe a definite article would be appropriate in this case as well. 5 The status of examples like (Error! Reference source not found.a) as “ungrammatical (*)” or “infelicitous or semantically anomalous without a special context (#)” is a theory-dependent matter that does not have a straightforward answer. Examples (Error! Reference source not found.a) and (Error! Reference source not found.b) are very hard to improve by manipulating the context; other “bad” examples are much easier to improve by that means. BorschevParteeJSL2002.doc 3 b. NES: *(#) Leny ne pelo. Lena NEG sang GEN.F.SG N.SG Among the many puzzles about GenNeg, we have been most concerned with the characterization of the semantics of the construction. One question in this area is why, under sentential negation, some sentences allow only genitive subject, some only nominative subject, and some both. When both are possible, we want to know what the difference is. It is to be hoped that if we understand the semantics of the construction and the lexical semantics of the various verbs, the answers to those questions will follow. Here is Apresjan 1985: The key to understanding the construction we are considering can be found in the position taken long ago in the works of S.O. Karcevskij and R. Jakobson, namely that in negative sentences with the genitive case, “the subject itself is negated” (Jakobson), and in negative sentences with nominative case, “what is negated is not the subject, but its activity” (Karcevskij), or, speaking in the words of R. Jakobson, “just the action”. (Apresjan 1985: 295) In a certain sense, it is that same claim which is sharpened and developed in the work of Babby (1980). Babby holds that the scope of sentential negation is different in “existential sentences” (ES’s) and “declarative sentences” (DS’s), and that that difference is in turn determined by the differing communicative structure of these sentences, the structure of Theme and Rheme. According to Babby, in an ES, the entire sentence falls within the Rheme, and hence when the sentence is negated, the subject as well as the verb falls within the scope of negation. In a DS, on the other hand, the subject is the Theme and the verb (and its complements, if any) is within the Rheme; therefore, under sentential negation, the subject is outside the scope of negation. Many other accounts have been offered from many points of view and in many different theoretical frameworks. In recent years the stream of literature on this theme has not diminished, and evidently not all “keys to understanding the construction we are considering” have been found yet. Thus for example E.V. Padučeva stresses the interaction of verbs of perception with this construction and discusses the role of “Observer6” (Padučeva 1992, 1997). Many works are dedicated to the syntactic aspects of the given construction, and there has been much discussion of the hypothesized centrality of “Unaccusativity” as a pivotal and explanatory property of the verbs whose “subjects” (being underlying Objects) appear in the genitive under negation.7 Other works focus in addition or instead on hypotheses concerning the position(s) of negation with respect to the verbal complex and its relation to the position of arguments in existential and 6 The linguistic notion of an observer is looked at in detail in the works of Ju. D. Apresjan, in particular in Apresjan (1986). The point of view of an implicit observer also plays a role in Fillmore’s analysis of English come and go (Fillmore 1971), and in a number of insightful works by Carlota Smith and by Padučeva about the manipulation by authors of fiction of the points of observation from which the reader is ‘shown’ the unfolding scene (see, for example, Smith and Whitaker 1985, Padučeva 1994, 1996.) 7 Pesetsky (1982) argued in favor the hypothesis that Unaccusativity is a crucial part of the explanation of GenNeg, as do Perlmutter and Moore (1999) in the framework of Relational Grammar. Babby (2001) argues against the hypothesis. BorschevParteeJSL2002.doc 4 declarative sentences (Bailyn 1997, Brown 1999, Abels 2000). While we agree with Babby (2001) that the assignment of case to the subject is almost certainly mediated by the syntax (since case assignment is unlikely to depend directly on semantics/pragmatics), in this article we remain uncommitted and agnostic about the syntax. We discuss some syntactic issues briefly in Borschev and Partee 2002a and mention just a few of them in Section 3.3 below. Our main goal is to discuss some proposals that have been made for the description of the semantics of this construction and the issue of “different ways of looking at the same situation”. We will concentrate particularly on Babby’s approach, plus our own approaches in Borschev and Partee 1998a, 1998b, considering some of the problems and apparent contradictions among the various structural schemes that have been proposed, and possible directions of resolution. Special attention will be focussed on Babby’s central use of Theme-Rheme structure, its attractions and its problems. In Borschev and Partee 1998a, we followed Babby in his use of Theme-Rheme structure, making certain modifications in his proposals for the sake of the semantics. But it has long been observed in the literature (for example, in Arutjunova 1976, 1997) that ES’s may be used with a wide variety of communicative structures, and many believe that the subject can well be the Theme even when it shows up with GenNeg. The acknowledgement of this possibility8 seems to create a major problem for the account of Babby 1980, and at least requires that the arguments be rethought. Our present position is that Babby did indeed carve out an aspect of structure, some kind of “marking” or distinguishing of some constituent(s) of the internal part-whole structure, which determines the choice of the genitive vs. nominative construction. But we no longer agree that the relevant structure is to be identified with Theme-Rheme (or Topic-Focus) structure in the classical Praguian sense; we suggest that it is a notion of “taking a perspective”, a notion we suggested (in slightly different terminology) in Borschev and Partee 1998b and will explore further here. In sentences (3a - 5a), the “marked” element is the first constituent, the imja bytujuščego predmeta ‘name of existing object’ in the terminology of Arutjunova, or the THING, as we have called it for brevity. In sentences (3b - 5b), on the other hand, it is the LOC(ation) (oblast’ bytija ‘region of existence’, in Arutjunova’s terms) that is marked, and the situation is looked at in terms of this LOCation. We call this structure of “marking” perspective structure, and the “marked” constituent the Perspectival Center; we introduce these notions more fully in Section 3. This paper is organized as follows: In Section 1 we describe Babby’s (1980) analysis, the core of which consists of two theses: (i) that the most crucial conditioning factor for GenNeg is that the (indefinite) subject and (weak) verb both fall under the scope of sentential negation, and (ii) that the scope of sentential negation is determined by Theme-Rheme structure. As a result, the existential sentences in which GenNeg occurs are a species of “Rheme-only” sentences. We introduce some problems that have led us to doubt the correctness of the proposed link between Theme-Rheme structure and GenNeg (which has been questioned by others as well). Also in Section 1, we discuss some problems raised by Babby (1980) concerning sentences with the verb byt’ ‘be’ with definite subjects marked with GenNeg. 8 This fact was brought to our attention by our colleagues Tanya Yanko, Elena Paducheva, and Sandro Kodzasov, and was confirmed by Jan Firbas and our Prague colleagues Eva Hajičová and Petr Sgall. Babby considered and rejected this “fact” in his 1980 book; we discuss his response below. BorschevParteeJSL2002.doc 5 In Section 2, we review several characterizations of “existential sentences”, and outline the kernel of the evolving analysis developed in Borschev and Partee 1998a,b, 2002a,b, Partee and Borschev 2002, and in the present work, particularly our notion of “Perspectival Structure”. Section 3 expands on our proposal that a LOCation is always the Perspectival Center in existential sentences. We discuss the roles we call THING and LOCation in both existential and locative sentences, and the additional perspectival structure which we posit to distinguish between those sentence types. We present arguments for favoring our use of perspectival structure over Babby’s use of Theme-Rheme structure in the analysis of GenNeg, while noting that our notion needs further support and that the issue is far from settled. We include brief discussion of the sources of presuppositions associated with existential and declarative sentences, and we close Section 3 with a few remarks about syntactic issues. The paper ends with some concluding notes on the continuing empirical and theoretical challenges of finding an integrated view of the fascinating subtleties of this important construction. 1. Babby on “Declarative” and “Existential” Sentences 1.1. The Scope of Negation. Theme and Rheme. Problems We noted above that Babby (1980) distinguished “declarative” from “existential” sentences. For Babby’s distinction between NDS and NES, the negative sentences, occurrence of nominative vs. genitive is criterial. For the corresponding affirmative sentences, ADS and AES, there is no absolute surface criterion; the usual basis for distinguishing between ADS and AES is to ask which case would be used on the subject if the sentence were negated. It is on this criterion that one can determine that there is normally an obligatory difference in word order in the affirmative sentences, as noted in the introduction. But since word order can be varied with accompanying changes in intonation in certain contexts, word order is not itself criterial in the affirmative case. We note that for almost every NES, one can construct a corresponding “declarative” sentence, although, as shown by example (4a) above, the result will not necessarily be acceptable. The converse is by far less valid; it is very easy to find NDS’s with no NES counterpart. Babby’s first main proposal about the distinction is shown in his chart (6) (Babby 1980: 72) below: DS’s and ES’s differ in their “scope of assertion/negation”. (6) AFFIRMATIVE NEGATED EXISTENTIAL [ VP NP] ⇒ [ VP NP ] Scope of A NEG ne gen DECLARATIVE NP [ VP] ⇒ NP [ VP] Scope of A NEG nom ne Thus the declarative sentence (3a) presupposes that there was some runoff of thawed snow and asserts that it was it was not observed, i.e. negates only that it was observed. The corresponding ES (3b) is used to negate the very existence of any runoff of thawed snow9. The ES also negates 9 Cf. the discussion of these examples in Apresjan 1985: “In contemporary terminology in cases of the type of Stoka BorschevParteeJSL2002.doc 6 “was observed”, i.e. it negates the whole sentence; but in this case nabljudalsja ‘was.observed’ functions as a “weak verb” (often described as “semantically empty”). The notion of “weak” or “empty” verbs was at the center of the work reported in Borschev and Partee 1998a and is briefly reviewed in Borschev and Partee 2002a. According to chart (6), a declarative sentence consists of two parts. The subject NP, which remains outside the scope of assertion/negation, and the predicate VP, which is in the scope of assertion/negation. An ES, in contrast, has just one “part” in this sense; the entire sentence is inside the scope of assertion/negation, and such a sentence typically asserts/denies the existence of what the subject NP refers to. Babby cites Kuroda’s (1972) discussion of Brentano and Marty’s distinction between categorical and thetic judgments, and notes that Kuroda lists existential sentences and impersonal sentences as examples of thetic judgments. Categorical judgments, corresponding to Babby’s “declarative sentences”, have two parts, one part involved with recognizing the subject, the other part predicating something about that subject; thetic judgments are “simple”, have just one part. Babby suggests that the Russian nominative/genitive alternation provides even stronger linguistic evidence for the linguistic reality of the categorical/thetic distinction than the Japanese wa/ga alternation discussed by Kuroda. Babby’s second main proposal is that the scope of assertion/negation can be equated with the Rheme of the sentence according to the division of the sentence into Theme and Rheme (or Topic and Focus). He notes that various authors define the concepts of Theme and Rheme differently, but the majority agree that Rheme is that which is asserted in an affirmative assertive sentence and negated in sentences that have normal sentential negation.10 He states the following principle: “When an assertive sentence is negated, its scope of negation is determined by its underlying rheme; the theme accordingly falls outside the scope of negation.” (Babby 1980: 103) He then expands on this: “In other words, what I am proposing here is that the rheme of a negated assertive sentence and the scope of negation in an assertive sentence are different terms for the same thing” (Babby 1980: 103). On Babby’s view, an ES, AES or NES, is a “rheme-only” sentence11. In Prague School literature and Russian linguistics these are known as “unpartitioned” utterances (nerasčlenennye vyskazyvanija in Russian). Babby’s final formulation of his rule of genitive marking in NES’s talyx vod ne nabljudalos’ [=(Error! Reference source not found.b)], Moroza ne čuvstvovalos’, there is no presupposition and one asserts the nonexistence or complete absence of the object spoken about: there was no runoff of melted snow at all, there was no frost. In contrast, the sentences Stok talyx vod ne nabljudalsja [=(Error! Reference source not found.a)], Moroz ne čuvstvovalsja most often carry the presupposition of existence of the object spoken about and assert only that it did not carry out the given action or was not in the given state: some runoff of melted snow in principle took place, but no observations were made of it; there was frost, but it could not be felt (because, for instance, everyone was warmly dressed).” (Apresjan 1985:295-96) 10 Narrow, or contrastive, “constituent negation” is a separate story; we leave it aside, since it does not trigger GenNeg effects. (See Babby 1980: 104-108.) However, there is room for debate about the status of the negation in a number of the examples here and in the literature, particularly examples with final ne ‘NEG’ + Verb, especially those with final ne ‘NEG’ + byt’ ‘be’. Examples of the last sort in this paper include (Error! Reference source not found., Error! Reference source not found., Error! Reference source not found. (=Error! Reference source not found.) , Error! Reference source not found., Error! Reference source not found., and the right-hand sides of Error! Reference source not found.a-d). This class of examples is discussed in Section 3.2.2. 11 Babby notes that an existential sentence may have an additional Locative constituent, which if present will be outside the Rheme. One of our modifications of Babby’s work is to give the Locative a more important role; see below. BorschevParteeJSL2002.doc 7 (his (160)) is given in (7) below. (7) NEG [ V NP] ⇒ [ ne V NP ] Rheme gen Conditions: (a) NP is indefinite (b) V is semantically empty Condition (a) requires that the subject of an ES be indefinite, not refer to a definite particular; Borschev and Partee (1998a) argued against this condition, and some of the conditions surrounding the use of definite NP’s in an ES were further discussed in Partee 2000. Condition (b), that the verb be “semantically empty”, was discussed in some detail in Borschev and Partee 1998a, where we offered an account in which the verb retains its literal meaning but the construction effectively reduces its assertional component to little more than “existence” (in a given LOCation.) In our earlier work, we agreed with Babby about the role of communicative structure in the interpretation of existential and declarative sentences. Our principal amendment was our endorsement of the principle that existence is always relative to a “LOCation”, which may be implicit12. We accept Jackendoff’s (1972, 1990) metaphorical-structural extensions of “being in a location” to include “being in some state”, “occurring in some spatiotemporal region”, “being in someone’s possession”, extending also to “being in the speaker’s (or an observer’s) perceptual field” (Padučeva 1992, 1997). Then whereas Babby analyzed ES’s as “Rheme-only”, with a possible optional Thematic Location, we argued that the LOCation, either given or contextually presupposed, is a semantically obligatory part of the construction and is the Theme. The assertion (Rheme) is that the/a “THING” described by the subject NP exists in that LOCation. The idea that all “Rheme-only” sentences have an implicit topic, understood as something like a situation or a spatiotemporal LOCation, occurs in various forms in the literature. Erteschik- Shir (1997), who introduces the notion of “stage topic”, corresponding to an explicit or implicit spatiotemporal argument (à la Kratzer 1995) for such sentences, notes that her distinction between sentences with stage topics and sentences with individual topics is equivalent to Kuroda’s distinction between thetic and categorical judgments, and to the distinction made by Guéron (1980) between presentation sentences and predicational sentences. She also notes that the idea that scene-setting expressions that specify temporal or spatial background for the sentence also function as topics may be found in Gundel 1974 and Reinhart 1981. Sgall et al. (1986: 202) claim that initial locatives are always topics. Others who claim that an initial locative in an otherwise unpartitioned sentence always functions as a topic are Dahl (1969:38), Arutjunova (1976: 210-11), Gundel (1974:34), all cited by Babby (1980:96), who agrees; he notes that others have claimed that it is part of a complex Rheme (Krylova and Xavronina 1976:26, [note 2]; Crockett 1976:241). So there is very wide agreement that an initial Locative in an existential sentence is generally a topic, and considerable support for the idea that when such 12 The claim that existential be-sentences always have an obligatory Locative argument is also made by Yokoyama (1986), Kondrashova (1996), Comorovski (1995). Chvany (1975) emphasizes that no such argument is syntactically obligatory, while Kondrashova (1996) goes so far as to make the location the subject at her level of NP-structure. BorschevParteeJSL2002.doc 8 a locative topic is not overt in an existential sentence, something like it is implicitly understood. Now, in the light of the observations of our colleagues mentioned in the Introduction and after reviewing Arutjunova (1976, 1997) and other works, we have doubts about the full correlation of the distribution of NES’s and NDS’s with the postulated difference in Theme- Rheme structure. Thus in examples (8-10) below, it appears to us that our colleagues are correct in claiming that the words sobaki ‘dog ’, kefira ‘kefir ’, and otveta ‘answer ’ GEN.F.SG GEN.M.SG GEN.M.SG are the Themes (or part of the Theme) of these sentences. Both their most natural intonation pattern and their (most likely) interpretation in the given contexts support this point of view, which argues against the generalization in (7). (8) Sobaki u menja net. (Arutjunova 1976) dogGEN.F.SG at IGEN not.is I don’t have a dog. [Context: talking about dogs, perhaps about whether I have one.] (9) [Ja iskal kefir. ] Kefira v magazine ne bylo. [I looked-for kefir- Kefir- in store NEG was- ACC.M.SG GEN.M.SG N.SG ‘[I was looking for kefir.] There wasn’t any kefir in the store.’ (10) [Ja napisal emu i ždal otveta.] Otveta ne prišlo. [ I wrote him and waited.for answer-GEN.M.SG.] AnswerGEN.M.SG NEG came-N.SG [I wrote to him and waited for an answer.] No answer came. Our conception of the communicative structure of sentences is based to a considerable extent on works in the tradition of the Prague School such as Hajičová 1973, 1974, 1984, Sgall et al. 1986 and works of Russian linguists. For us the Theme is, roughly speaking, what is being talked about in a sentence, which is presupposed to be familiar to the hearer, referring back to something which was either spoken about earlier or else simply well known. The Rheme is new information which the speaker wishes to communicate. We note that Babby, alongside the Theme-Rheme opposition, considers an additional opposition, between old (given) and new information. Judging by his discussion of similar examples, he would consider sobaki ‘dog ’, kefira ‘kefir ’, and otveta ‘answer ’ GEN.F.SG GEN.M.SG GEN.M.SG in (8-10) above to be old information but part of the Rheme. Although we consider this issue far from settled13, we are now inclined to treat them as (part of) the Theme of the corresponding sentences. Even the relation of GenNeg to the scope of negation is not as clear as it has generally been believed to be. Consider examples (11-12). (11) [My nadejalis’, čto na seminare budut studenty.] No ni odin [We hoped, that at seminar will.be students] But NI one NOM.M.SG student tam ne byl studentNOM.M.SG there NEG was-M.SG [‘We hoped that (some of the) students would be at the seminar. But not a single one of 13 We return to this issue in Section 3.2.2., where we mention a proposal by Erteschik-Shir according to which the sentences (Error! Reference source not found.-Error! Reference source not found.) are indeed Rheme-only as claimed by Babby, and kefira and the like have the status of a subordinate Topic (Theme) inside the main Focus (Rheme). This “subordinate Theme” could correspond to Babby’s “old information”. BorschevParteeJSL2002.doc 9 the students was there.’ (12) [My nadejalis’, čto na seminare budut studenty.] No ni odnogo [We hoped, that at seminar will.be students] But NI one GEN.M.SG studenta tam ne bylo. studentGEN.M.SG there NEG wasN.SG [‘We hoped that there would be students at the seminar.] But there was not a single student [or: not a single one of the students] there.’ The sentences (11) and (12) have the same truth conditions if looked at in simple extensional models; in both cases, the expression translated as “not one student” is necessarily under the scope of the sentential negation in its clause.14 Yet only in (12) does that expression occur in the genitive; (11), with the nominative, is equally well-formed, which seems unexpected not only under Babby’s account but under virtually all accounts of GenNeg15. There is one important difference in the interpretation of (11) and (12). In the sentence with the nominative subject, we must be speaking of a contextually definite group of students. The background sentence, which has the same surface form in both (11) and (12), contains the bare noun studenty ‘students’, which is indeterminate with respect to definiteness or specificity. Given its sentence-final position, studenty cannot easily be definite ‘the students’, but it can have a partitive interpretation ‘some of the students’, and must have the partitive interpretation for (11) to be felicitous. The nominative ni odin student is also understood partitively: ‘not a single one of the students’. In (12), on the other hand, there is no such restriction. A partitive interpretation is possible but a completely non-specific interpretation is normal. This problem and other issues concerning the scope of negation are discussed in Partee and Borschev 2002 and will not be discussed further here. 1.2. Sentences with the Verb byt’ ‘be’ and Referential Subject Babby (1980: 124) (like Arutjunova 1976: 225) does not include sentences like (13) among the class of NES’s, even though in all syntactic and morphological respects, including the manifestation of GenNeg, (13) looks like an ordinary NES. (13) Ivana ne bylo na lekcii Ivan NEG was at lecture GEN.M.SG N.SG ‘Ivan wasn’t at the lecture.’ The problem is the semantics: sentences like (13) whose subject is a proper name or other clearly 14 The ni ‘not’ of ni odin, ni odnogo ‘not one’ always (except in certain constructions that are irrelevant here) requires the co-presence of sentential negation ne in its clause. Russian is a “Negative Concord” language, and expressions like ni odin ‘not one’ and nikto ‘no one’ are generally analyzed as a negative polarity items (see Brown 1999). We therefore gloss ni simply as ‘NI’ rather than translating it as ‘not’. 15 We discuss examples of this kind, including some noted by Babby (1980), in Borschev and Partee 2002a and in Partee and Borschev 2002. BorschevParteeJSL2002.doc 10