This volume brings together research presented at the workshop Existence: Semantics and Syntax, held at the University of Nancy 2 in 2002, along with some invited contributions. The papers are organized into three thematic sections, preceded by a brief introductory chapter.
Ronnie Cann (‘Towards a dynamic account of BE in English’) seeks to capture the flexible distribution of be under a unified analysis. Instead of assuming homophony for be, or deriving a family of meanings from a basic one, Cann analyzes be as an underspecified monadic predicate, the content of which is determined by linguistic and extralinguistic context. In Dynamic Syntax, the incremental construction of utterances restricts the pragmatic and linguistic resources available for determining content. Cann exemplifies how such restrictions can be used to derive standard copular constructions, constructions in which a postcopular predicate is elided, existential sentences, and so-called existential focus constructions (Neuroses just ARE). The analysis raises various interesting issues, such as a potential derivation of some of the well-known properties of existentials from architectural properties. The idea that be is radically underspecified intriguingly invites comparison with have, for which underspecification has also been claimed, and the formulation of a general theory of underspecification within the principled account of the interaction between form, meaning and context outlined.
Ileana Comorovski (‘Constituent questions and the copula of specification’) analyzes French constituent questions of the form [quel cop DP] as specificational clauses. The postcopular D(eterminer)P(hrase) is argued to be a specificational subject denoting an individual concept; quel-questions function to inquire about the extension of this concept relative to a context. Comorovski argues for a lexically specified copula of specification, and, consequently, against analyses of specification as inverse predication. The proposed semantics explains some of the characteristics Comorovski identifies for specificational subjects, for example, the fact that they cannot be referential or quantificational. An open question Comorovski points to is the apparent requirement that specificational subjects contain a discourse-familiar referent (‘indirect discourse anchoring’). This is not explained by type-theoretical choices. An answer to this question would presumably also explain other interesting observations made in the paper, such as the fact that incomplete definite descriptions cannot be specificational subjects. Beyond analyzing a class of French interrogatives, the paper demonstrates that individual concepts are relevant to understanding ‘specification’, converging with evidence from other areas of research, such as the semantics of concealed questions (e.g. Romero Reference Romero2007).
Ljudmila Geist (‘Predication and equation in copular sentences: Russian vs. English’) offers an analysis of Russian predicational, specificational, and equative copular sentences which maintains a unified basic interpretation for the copula as an identity function on predicates. She thus argues against positing a separate copula of identity, contra, for example, Higgins (Reference Higgins1973), instead correlating the distinction between the predication and identity readings in part with a property- vs. entity-type denotation (respectively) for the postcopular nominal. She then makes novel use of type-shifting to achieve semantic composition when the copula – which unambiguously needs a property-type complement – has to combine with an entity-type complement to yield an identity reading: rather than adopting the standard move of type-shifting the nominal, Geist combines the relevant nominal type-shifting function with the copula via function composition, combining that result with the nominal. This seemingly innocuous change, combined with the assumption that type-shifting must apply to explicit linguistic expressions, cleverly accounts for a variety of facts in both English and Russian.
Claudia Maienborn (‘On Davidsonian and Kimian states’) discusses what she calls statives, a class of predicates including those expressed by postcopular constituents. In previous work, Maienborn has argued that statives do not involve a Davidsonian event variable but instead express K(imian)-states. Unlike eventualities, K-states are realized only in time, not space, and hence are not perceptible. Here, Maienborn defends this thesis from two potential counterarguments. One is that if statives lack a Davidsonian variable, they should not permit manner modification, yet they sometimes do. Maienborn explains away such cases as involving noncompositional interpretation. The second counterargument, based on Parson's ‘time travel’ argument, also involves manner modification. Since statives do not allow manner modification, she claims that this argument does not carry over to them. The success of this response therefore depends directly on the strength of her reply to the first counterargument. Another issue is whether Maienborn's distinction requires abandoning Davidsonian logical forms, rather than just the recognition of non-spatial, abstract truth makers. For example, could the distinction be incorporated into a Davidsonian analysis as a sortal distinction in the domain of eventualities, on which the relevant ontological distinctions are pinned? Assuming that manner modifiers are sort sensitive, the adverbial modification facts might then be derived.
Jianhua Hu & Haihua Pan (‘Focus and the basic function of Chinese existential you-sentences’) analyze the discourse function of existentials in Mandarin Chinese. The main aim of the paper is to account for existentials in which a discourse-old, formally definite noun phrase occurs in an existential with a focus marker. Hu & Pan's claim is that in such cases, the existential sentence functions not to introduce a new discourse referent but rather to introduce the new information that a given entity is an element of a given set (on the assumption that the given set is not given by an enumeration of its members but by a description). As Hu & Pan point out, this is a more precise characterization of so-called ‘list’ readings of existentials with definite NPs. The main intuition of the paper seems correct; unfortunately, however, the paper stops short of an explicit analysis. The formalization given (see, e.g., 25) is not coherent, one of several places where more careful copy-editing would have helped.
Barbara H. Partee & Vladimir Borschev (‘Existential sentences, BE, and the genitive of negation in Russian’) consider whether there are necessary or sufficient conditions for identifying existential sentences in Russian. The prototypical Russian existential looks like a copular sentence containing a locative predicate, only with the constituent order reversed so that the locative expression is sentence-initial rather than postcopular. However, Partee & Borschev point out that there are other contrasts between typical existentials and locatives, including the absence of both definite subjects and null present tense copulas in existentials, and the presence of genitive of negation marking on the subjects of negated existentials but not on the subjects of negated locatives. They present a useful overview of the genitive of negation and Russian existentials literature, including their own analysis in terms of what they call Perspectival Structure. They then show how the distributional correlates of existentials vs. locatives fail to align as predicted. Partee & Borschev outline several different approaches to this problem, tentatively suggesting that, for Russian, ‘existential sentence’ is a cluster concept associated with a set of independent but frequently coinciding characteristics.
In ‘Negative quantification and existential sentences’, Lucia Tovena discusses two nonverbal constructions in Italian involving the negative elements (n-words) niente ‘nothing’ and nessuno ‘no/nobody’, which correspond roughly to negative existentials:
- (1)
(a) Nessuntestimoneintornoalei.nowitnessesaroundtoher‘(There were) no witnesses around her.’
(b) Nienteprocessoperlatruppa.nothingtrialforthetroops‘(There will be) no trial for the troops.’
Tovena suggests that these constructions are interpreted as tripartite quantificational structures, the n-word contributing the quantifier, the adjacent noun the restriction, and the rest of the clause the scope (essentially as in Keenan's 1987 analysis of English existentials, where the expletive and copula are semantically inert). The paper provides a rich and intriguing discussion of various semantic and pragmatic properties differentiating the two constructions. For example, niente constructions assert that the intersection of the restriction and scope is empty, whereas the nessuno construction allows that it is not. Tovena speculates that this reflects an information-structural distinction related to conservativity. Since niente is second-argument conservative, its first argument is NOT topical, but rather part of the assertion, and nothing is assumed in the common ground about its extension. This is appealing, but as Tovena herself notes, the link with conservativity is weak. Niente and nessuno are both conservative on both arguments, and conservativity can therefore not determine their information structural differences.
Francis Corblin (‘Existence, maximality, and the semantics of numeral modifiers’) presents a novel analysis of nominals containing the modified numerals at least n, at most n, and exactly n, on which these expressions introduce a comparison between two sets: one having the cardinality n; the other, the maximal set of individuals satisfying the conditions expressed by the sentence. For example, the satisfaction conditions for At least two girls sang require that the set of girls who sang is at least as large as a set of two girls. Corblin then argues that an utterance of a nominal containing a modified numeral introduces two discourse referents: one for each of the compared sets. He supports this analysis with data involving discourse anaphora and appositive modification, which show that the cardinality conditions relevant for truth are distinct from those relevant for discourse dynamics. Finally, Corblin points out a problem his analysis encounters with existential sentences lacking coda phrases. Interestingly, this problem would not arise if Corblin assumed a relational semantics for existentials such as that proposed by Partee & Borschev, pointing to the relevance of modified numeral data for debates over the semantic analysis of existentials.
Klaus von Heusinger (‘Referentially anchored indefinites’) discusses the notion of specificity pertaining to specific indefinites. He argues that the relevant notion for the analysis of all specific indefinites is what he calls ‘relative specificity’, which involves anchoring to an entity familiar in the context. Familiarity is modeled as in File Change Semantics (Heim Reference Heim1982). Specific indefinites and definites differ in that the latter are directly anchored to a familiar entity, whereas the former introduce a new discourse referent which is ‘linked’ or identifiable relative to a familiar referent. According to von Heusinger, unlike the case of definites, the anchoring referents for specific indefinites must be introduced within the same clause as the indefinite. This assumption predicts that specificity is dependent on configurational facts, a prediction argued to be borne out in languages like Turkish. However, he also shows that anchoring can be to entities which are not syntactically represented in the sentence, such as the speaker. It thus seems that specificity is not so much clause-bound as utterance-bound. This unified analysis of specificity is intuitively very appealing; however, the formalization is slightly confusing, as it is unclear whether index variables stand for sets of indices (as indicated by notations such as {i}⊆j) or sets of individuals (as indicated by notations such as i⊆j). This is another instance where better copy-editing would have helped.
Bart Geurts (‘Existential import’) examines an asymmetry between strong quantifiers such as every and weak quantifiers such as some: the former, but not the latter, carry what Geurts calls existential import, i.e., the implication that their domain of quantification is not empty. Geurts considers, and rejects, two previous explanations for why some quantifiers have existential import and others do not. He then defends the view that strong quantifiers presuppose a nonempty domain, while weak ones do not. He reviews the evidence for this view and replies to the criticism that strong universal quantifiers cannot presuppose their domain because universal quantifications can be true when their domain is empty. The discussion leads to a more general reflection on the nature of presupposition in which Geurts argues that the existential import facts support Strawson's pragmatic conception of the phenomenon. In addition to serving as an argument for the binding theory of presupposition (van der Sandt Reference van der Sandt1992), which Geurts shows is one way to formalize a Strawsonian view of presupposition, the paper provides thoughtful commentary on the relation between definedness, accommodation, givenness and existential import.
Finally, Roberto Zamparelli (‘On singular existential quantifiers in Italian’) examines the system of singular indefinites in Italian, with emphasis on qualche ‘some’ and un qualche, literally ‘a some’. The puzzle is that while un qualche is always interpreted as singular and with a free choice implication, qualche can be interpreted as strictly singular or as entailing more than one. The latter interpretation is the default; the former, restricted to contexts which Zamparelli loosely describes as intensional. Another difference is that singular qualche and un qualche imply intedeterminacy of identity, while plural qualche implies indeterminacy of quantity. Zamparelli accounts for these facts under a uniform semantics for qualche by appealing to a ‘layered’ DP structure, which affords two distinct landing sites for qualche, and by interpreting indefinites with respect to Horn scales which yield different implicatures depending on which position qualche occupies. In a very interesting move, he attributes the free choice interpretation of un qualche not to domain widening, as is standard, but to a blocking of domain restriction.
We close with a few comments on the formal aspects of the volume. There is a useful subject index that includes reference to the different languages discussed in the contributions. Although each section is reasonably cohesive, the papers vary considerably in length, resulting in a certain unevenness. As mentioned above, the volume would have benefitted from more thorough editing and proofreading: there are inconsistencies in the presentational details of the contributions and typographical errors in all of the chapters, and in some cases the papers have stylistic elements that are more characteristic of an expanded handout than a published article. Nonetheless, the volume presents a number of novel, thought-provoking proposals and makes for worthwhile reading.
