Exhaustification

In this thesis, I focus on scalar implicatures, presuppositions and their connections. In chapter 2, I propose a scalar implicature-based account of neg-raising inferences, standardly analyzed as a presuppositional phenomenon (Gajewski 2005, 2007). I show that an approach based on scalar implicatures can straightforwardly account for the differences and similarities between neg-raising predicates and presuppositional triggers.
In chapters 3 and 4, I extend this account to “soft” presuppositions, a class of presuppositions that are easily suspendable (Abusch 2002, 2010). I show how such account can explain the differences and similarities between this class of presuppositions and other presuppositions on the one hand, and scalar implicatures on the other. Furthermore, I discuss various consequences that it has with respect to the behavior of soft presuppositions in quantificational sentences, their interactions with scalar implicatures, and their effects on the licensing of negative polarity items.
In chapter 5, I show that by looking at the interaction between presuppositions and scalar implicatures we can solve a notorious problem which arises with conditional sentences like (1) (Soames 1982, Karttunen and Peters 1979). The main issue with (1) is that it is intuitively not presuppositional and this is not predicted by any major theory of presupposition projection.
(1) I’ll go, if you go too.
Finally, I explore in more detail the question of which alternatives should we consider in the computation of scalar implicatures (chapter 6). Traditionally, the answer has been to consider the subset of logically stronger alternatives than the assertion. Recently, however, arguments have been put forward in the literature for including also logically independent alternatives. I support this move by presenting some novel arguments in its favor and I show that while allowing new alternatives makes the right predictions in various cases, it also causes an under- and an over-generation problem. I propose a solution to each problem, based on a novel recursive algorithm for checking which alternatives are to be considered in the computation of scalar implicatures and the role of focus (Rooth 1992, Fox and Katzir 2011).

Bare conditionals show an unexpected quantificational variability contingent on

whether they are embedded in an Upward Entailing context (universal import) or a Downward

Entailing context (existential import). Contra Herburger (2015a, b)’s ambiguity theory, we

argue in favor of a unified semantics for bare conditionals based on their behavior in VP ellipsis

constructions and in non-monotonic contexts. We show that a similar pattern exists with Free

Choice phenomena, and consequently suggest a parallel analysis to Fox (2007)’s treatment of

such phenomena. We propose that bare conditionals have a basic existential semantics which

is obligatorily strengthened into a universal meaning in UE contexts, while being preserved in

DE contexts. Our claim that bare conditionals are underlyingly existential is further supported

by Conditional Perfection data with bare and non-bare conditionals.

[POLARITY SENSITIVE ITEMS, FREE CHOICE, ALTERNATIVES, EXHAUSTIFICATION] Logic in Grammar reflects the results of Chierchia's research on Negative Polarity Items (NPIs), n-words, Free Choice Items (FCIs), Scalar Implicatures (SIs) and related phenomena dependent on the logical properties of the contexts such expressions (must) appear in. An approach to semantics and pragmatics where the computation of pragmatic implicatures is, in many cases, anchored in the syntactic structure (and thus not post-syntactic) is at the core of this system. The ungrammaticality of NPIs and other expressions is explained as a result of logical conflicts that arise when locally generated implicatures contradict the assertion of the whole sentence. Chierchia takes plain indefinites, such as a and sometimes, and NPIs, such as any and ever, as semantically equivalent. Nevertheless, the latter always introduce scalar and domain alternatives that must be exhaustified by some covert operator, whereas the former may optionally do so. To exhaustify means to negate all non-entailed alternatives and this is performed by a covert operator. The variation within the landscape of Polarity Sensitive Items (PSIs) can be attributed to different types of exhaustification. This mechanism fits into one of Chierchia's main projects, which is to account for the distribution of PSIs without positing the existence of licensers.

We propose a modification of the exhaustivity operator from Fox (2007) that on top of negating all the Innocently Excludable (IE) alternatives affirms all the `Innocently Includable' (II) ones. The main result of supplementing the notion of Innocent Exclusion with that of Innocent Inclusion is that it allows the exhaustivity operator to identify cells in the partition induced by the set of alternatives (assign a truth value to every alternative) whenever possible. We argue for this property of `cell identification' based on the simplification of disjunctive antecedents (SDA) and the effects on free choice (FC) that arise as the result of the introduction of universal quantifiers. We further argue for our proposal based on the interaction of only with free choice disjunction.

In this paper, I give an analysis of neg-raising inferences as scalar implicatures. The main motivation for this account as opposed to a presupposition-based approach comes from the differences between presuppositions and neg-raising inferences, noticed by and Homer (2012). In response to this issue, argues that neg-raising predicates are soft presuppositional triggers and adopts the account of how their presuppositions arise by . However, I argue that there is a difference between soft triggers and neg-raising predicates in their behavior in embeddings; a difference that is straightforwardly accounted for in the present approach. Furthermore, by adopting Abusch's (2010) account of soft triggers, inherits the assumptions of a pragmatic principle of disjunctive closure and of a non-standard interaction between semantics and pragmatics -assumptions that are not needed by the present proposal, which is just based on a regular theory of scalar implicatures. I also show that the arguments that presents in favor of the presuppositional account can be explained also by the scalar implicatures-based approach proposed here. Finally, while the main point of the paper is a comparison with the presuppositional account, I sketch a preliminary comparison with the revived syntactic approach by . the audiences at the Presuppositions in Context Workshop in Göttingen, at SuB 17 in Paris, and at NYU. The paper has benefited critically from the comments by the Editor and the two anoymous referees. All errors and omissions are mine.

I provide arguments in favor of an implicature approach to Homogeneity (Magri 2014) where the basic meaning of "the kids laughed" is "some of the kids laughed", and its strengthened meaning is "all of the kids laughed". The arguments come from asymmetries between positive and negatives sentences containing definite plurals with respect to (i) children’s behavior (Tieu et al. 2015), (ii) the availability of non-maximal readings, and (iii) the robustness of neither-true-nor-false (‘gappy’) judgments (Križ & Chemla 2015). I propose to avoid some problems of Magri’s analysis by modelling the Implicature account of Homogeneity after the Implicature account of Free Choice, based on a hitherto unnoticed analogy between the two phenomena. The approach that emerges has the advantages of Magri’s Implicature account of Homogeneity (predicting asymmetries), while at the same time bears a close resemblance to recent approaches to Non-maximality (Malamud 2012; Križ & Spector 2017), which enables restating their account of Non-maximality as following from the context-sensitivity of implicature calculation.

The aim of this paper is to present a new account for imperatives and their interaction with free choice item any and disjunctive or. Specifically, I argue that free choice inferences are a conversational implicature rather than an entailment. To that, I show that indefinite any and
disjunctive or are overt exhaustifiers of the set of alternatives introduced by imperatives and that the derivation of free choice inferences for any and or depend on different types of
exhaustification: recursion and iteration, respectively.

Two different readings are possible in English when a conjunction of definite objects is negated. It is argued in the present paper that different sets of alternatives are active in the two cases, but that this is only possible with the right interplay of the preceding context and the intonation contour of the sentence. Both readings result from exhaustification. However, the number of applications of the EXH operator is different in the two cases. This fosters discussion about the constitution of alternative sets, the exact role of focus, as well as triggers for (re-)exhaustification.

The goal of this talk is to provide a global account of universal Free Choice (FC) inferences (argued to be needed in Chemla 2009). We propose a stronger exhaustivity operator than proposed in Fox (2007), one that doesn’t only negate all Innocently Excludable (IE) alternatives but also asserts all “Innocently Includable” (II) ones, and subsequently can derive universal FC inferences globally. We further show that Innocent Inclusion is independently motivated by considerations that come from the semantics of only (data from Alxatib 2014). Finally, the distinction between Innocently Excludable and Innocently Includable alternatives allows us to capture differences between FC inferences and other scalar implicatures.

Warlpiri (Pama-Nyungan, Australia) has a single coordinator, manu. This coordinator occurs in constructions of the form P manu Q. In the following paper, I argue that manu has a non-strengthened disjunctive denotation (∨) which can undergo pragmatic strengthening to conjunction (&). I propose that the absence of a conjunctive coordinator in Warlpiri results in the set of scalar alternatives {P, Q, (P ∨ Q)} for P manu Q. This is identical to the set of scalar alternatives for English P or Q proposed by Singh, Wexler, Astle, Kamawar, & Fox (2013) for English-speaking children. I follow Singh et al. in using a recursive application of Fox's (2007) exhaustivity operator, Exh, to derive a strengthened conjunctive reading of P manu Q that is categorical in unembedded contexts. I also address strengthening under the scope of negation and in conditionals and Wh-questions, and place some pragmatic constraints on Exh application.

According to the 'grammatical account', scalar implicatures are triggered by a covert exhaustification operator present in logical form. This account covers considerable empirical ground, but there is a peculiar pattern that resists treatment given its usual implementation. The pattern centers on odd assertions like #"Most lions are mammals" and #"Some Italians come from a beautiful country", which seem to trigger implicatures in contexts where the enriched readings conflict with information in the common ground. Magri (2009, 2011) argues that, to account for these cases, the basic grammatical approach has to be supplemented with the stipulations that exhaustification is obligatory and is based on formal computations which are blind to information in the common ground. In this paper , I argue that accounts of oddness should allow for the possibility of felicitous assertions that call for revision of the common ground, including explicit assertions of unusual beliefs such as Most but not all lions are mammals and Some but not all Italians come from Italy. To adequately cover these and similar cases, I propose that Magri's version of the Grammatical account should be refined with the novel hypothesis that exhaustification triggers a bifurcation between presupposed (the negated relevant alternatives) and at-issue (the prejacent) content. The explanation of the full oddness pattern, including cases of felicitous proposals to revise the common ground, follows from the interaction between presupposed and at-issue content with an independently motivated constraint on accommodation. Finally, I argue that treating the exhaustification operator as a presupposition trigger helps solve various independent puzzles faced by extant grammatical accounts, and motivates a substantial revision of standard accounts of the overt exhaustifier "only".

In this talk I discuss the obstacles for the analysis of the Russian coordinator ni... ni ‘(n)either... (n)or’ as a disjunction within the scope of negation. Those include the existence of contexts where ni... ni is—at least superficially—merged higher than negation; wide scope readings of QPs conjoined by ni... ni w.r.t. negation; and a rejoinder to (the Russian version of) Jovana Gajić's argument for disjunctive semantics based on negated deontic contexts.
In the final part of the talk I sketch a licensing and anti-licensing mechanism for ni... ni, conjunctively understood; the mechanism is based in part on Zeijlstra's syntactic analysis of Negative Concord and in part on Chierchia-style semantic analysis of polarity items.

The presuppositions inherited from the consequent of a conditional or the (second) disjunct of a disjunction oscillate between a conditional and a non-conditional inference, depending on the context ; the so-called 'proviso problem' (Geurts 1996 a.o.). The general solution to this problem is assuming that both presuppositions are generated and that some mechanism selects between them. We discuss data for which this mechanism makes the wrong predictions. Similar data have been taken by van der Sandt (1992) and Geurts (1996) to motivate the DRT-approach to presuppositions. Schlenker (2011), however, has raised various arguments against such an approach. We propose an alternative analysis within a 'satisfaction' approach to presuppositions, which doesn't run into these problems. In our analysis, the differing presuppositions are the result of a systematic ambiguity involving exhaustification: a non-conditional presupposition obtains with exhaustification, and a conditional one without. Independently motivated plausibility considerations decide which reading is chosen with no direct selection of presuppositions needed. We discuss how this approach deals with the cases of proviso and the predictions it makes for biconditional sentences.

Two different readings are possible in English when a conjunction of definite objects is negated. It is argued in the present paper that different sets of alternatives are active in the two cases, but that this is only possible with the right interplay of the preceding context and the intonation contour of the sentence. Both readings result from exhaustification. However, the number of applications of the EXH operator is different in the two cases.  This fosters discussion about the constitution of alternative sets, the exact role of focus, as well as triggers for (re-)exhaustification.

Some languages make use of special forms for coordination markers in negative environments, known as negative coordination (‘Nina will neither sing nor dance’). This dissertation explores negative coordination in English and in BCMS (Bosnian / Croatian / Montenegrin / Serbian). The primary goal is to account for the logically strong interpretation (‘it is not the case that Nina will sing and it is not the case that Nina will dance’) of sentences with neither...nor, ni...ni, and niti...niti. The secondary goal is to relate these constructions, in which two coordination markers are present, to the ones where only a single neither/nor, ni or niti appears as an additive particle. Negative coordination constructions concern two major topics in formal semantics and syntax/semantics interface – negation and (binary) connectives. This perspective occupies a central place in the empirical description offered in the first half of the thesis, as well as in the theoretical discussion of the analyses proposed. Chapters 3, 4, 5 and 6 focus on the negative aspect of the construction, whereas Chapters 7, 8, 9 and 10 deal with the connective underlying negative coordination. In Chapters 3 and 4, it is shown that paired negative coordination markers have a wider distribution than combinations of a single marker with a plain connective or with sentential negation only. Chapter 5 shows that constructions with single markers can also materialize as independent sentences, in which case an additive particle use is invoked. English and BCMS diverge in the interaction of negative coordination with other negative or polarity sensitive expressions, as revealed in Chapter 6. Nonetheless, neither, nor, ni, and niti conform to the general system of negation in their corresponding languages, since neither and nor are found to be inherently negative, whereas ni and niti are not. The empirical investigation is narrowed to constructions with iterated coordination markers further on in Chapters 7, 8 and 9, where the logico-semantic nature of the connective component is determined. Using diagnostics with a third scope-taking expression, it is shown that negative coordination in both languages is underlyingly disjunctive. After dismissing an approach in terms of plural denotations and homogeneity, scopal ordering between disjunction and negation is discussed in Chapter 10. An LF with a disjunction scoping over negation is adopted for neither...nor, despite absence of alternative question readings and ignorance inferences with negative coordination in both languages. However, such an LF does not reflect the strong readings of negative coordination constructions, which are stable. On the other hand, ni...ni and niti...niti correspond to disjunctions in the scope of negation, which is directly compatible with their interpretation. The starting point for the analyses in the second half of the thesis is thus that both ni...ni and niti...niti in BCMS are underlyingly semantically non-negative disjunctions, which need to stay in the scope of a negative operator. This is implemented in Chapter 11 through a syntactic agreement mechanism between the coordination markers ni and niti and a c-commanding silent negative operator (Zeijlstra, 2004). It is further argued that this is not an option for neither...nor, which is why Across-The-Board movement of negative operators at LF is briefly considered. The form-meaning mismatch in English is given a better-motivated account in Chapter 12, where exhaustification of disjunct alternatives is used to strengthen the interpretation, in two versions – recursive Innocent Exclusion (Fox, 2007) and Innocent Exclusion followed by Innocent Inclusion of alternatives (Bar-Lev and Fox, 2017). This prompts ample discussion of issues which arise in the derivation of alternatives for negative coordination in English and its exhaustification in different environments. As the problems are not satisfactorily resolved, Chapter 13 offers an alternative approach, this time employing focus alternatives of the disjuncts to stipulate presuppositions on neither and nor, which not only brings about the strong meaning of negative coordination, but also accounts for their use as single additive particles. Chapters 14 and 15 reproduce strengthening by exhaustification and by presuppositions, respectively, for negative coordinations in BCMS, where it is vacuous, showing that the latter rather than the former can provide a useful add-on to the analysis with syntactic agreement from Chapter 11.

Bare conditionals show an unexpected quantificational variability contingent on whether they are embedded in an Upward Entailing context (universal import) or a Downward Entailing context (existential import). Contra Herburger (2015a, b)’s ambiguity theory, we argue in favor of a unified semantics for bare conditionals based on their behavior in VP ellipsis constructions and in non-monotonic contexts. We show that a similar pattern exists with Free Choice phenomena, and consequently suggest a parallel analysis to Fox (2007)’s treatment of such phenomena. We propose that bare conditionals have a basic existential semantics which is obligatorily strengthened into a universal meaning in UE contexts, while being preserved in DE contexts. Our claim that bare conditionals are underlyingly existential is further supported by Conditional Perfection data with bare and non-bare conditionals.