Anna Szabolcsi

This paper highlights a small selection of cases where crosslinguistic insights have been important to big questions in the theory of semantics and the syntax/semantics interface. The selection includes (i) the role and representation of Speaker and Addressee in the grammar; (ii) mismatches between form and interpretation motivating high-placed silent operators for functional elements; and (iii) the explanation of semantic universals, including universals pertaining to inventories, in terms of learnability and the trade-off between informativeness and simplicity.

(English translation of ...) This paper points out naturally-occurring examples, primarily in Hungarian but also to a more limited extent in English, in which disjunction (i) has a conjunctive force but (ii) its use highlights that the list is not intended to be exhaustive. The preliminary analysis is in terms of recursive proposition strengthening by exhaustification without a scalar alternative, assimilating exemplifications to known cases of conjunctively interpreted disjunctions in other languages.

According to the classical description of obviation, the subject of a subjunctive is disjoint in reference from the attitude-holder subject of the immediately higher clause. Inspired by Ruwet (1984/1991) and Farkas (1988; 1992), I present data from Hungarian where obviation in certain subjunctives is plainly lifted, and data where obviation occurs in indicatives. I argue that obviation is not the result of competition with another construction, and point to promising potential accounts in terms of semantics or pragmatics. My aim is to contribute desiderata for a theory of obviation and exemptions from obviation. The discussion centers on new data from Hungarian and does not attempt to provide final answers or a novel formalization. However, it pays close attention to pertinent semantic and pragmatic literature (Searle 1969; Shoemaker 1996; Schlenker 2005; 2011; Kaufmann 2019; 2020) and makes critical use of it.

Rawlins (2013: 160) observes that both unconditionals and more classical free choice can be meta-characterized using orthogonality, but does not actually unify the two. One reason may be that in English, different expressions serve in these roles. By contrast, in Hungarian, AKÁR expressions serve as NPIs, FCIs, and unconditional adjuncts, but not as interrogatives or free relatives. This paper offers a unified account of the Hungarian data, extending Chierchia 2013 and Dayal 2013. The account produces the same unconditional meanings that Rawlins derives from an interrogative basis. This result highlights the fact that sets of alternatives arise from different morpho-syntactic sources and are utilized by the grammar in different ways, but the results may fully converge.

The central topic of this inquiry is a cross-linguistic contrast in the interaction of conjunction and negation. In Hungarian (Russian, Serbian, Italian, Japanese), in contrast to English (German), negated definite conjunctions are naturally and exclusively interpreted as `neither’. It is proposed that Hungarian-type languages conjunctions simply replicate the behavior of plurals, their closest semantic relatives. More puzzling is why English-type languages present a different range of interpretations. By teasing out finer distinctions in focus on connectives, syntactic structure, and context, the paper tracks down missing readings and argues that it is eventually not necessary to postulate a radical cross-linguistic semantic difference. In the course of making that argument it is observed that negated conjunctions on the `neither’ reading carry the expectation that the predicate hold of both conjuncts. The paper investigates several hypotheses concerning the source of this expectation.

The additive presupposition of particles like "too"/"even" is uncontested, but usually stipulated. This paper proposes to derive it based on two properties. (i) "too"/"even" is cross-linguistically focus-sensitive, and (ii) in many languages, "too"/"even" builds negative polarity items and free-choice items as well, often in concert with other particles. (i) is the source of its existential presupposition, and (ii) offers clues regarding how additivity comes about. (i)-(ii) together demand a sparse semantics for "too/even," one that can work with different kinds of alternatives (focus, subdomain, scalar) and invoke suitably different further operators.

Surányi (2006) observed that Hungarian has a hybrid (strict + non-strict) negative concord system. This paper proposes a uniform analysis of that system within the general framework of Zeijlstra (2004, 2008) and, especially, Chierchia (2013), with the following new ingredients. Sentential negation NEM is the same full negation in the presence of both strict and non-strict concord items. Preverbal SENKI `n-one’ type negative concord items occupy the specifier position of either NEM `not' or SEM `nor'. The latter, SEM spells out IS `too, even’ in the immediate scope of negation; it is a focus-sensitive head on the clausal spine. SEM can be seen as an overt counterpart of the phonetically null head that Chierchia dubs NEG; it is capable of invoking an abstract (disembodied) negation at the edge of its projection.

The typical habitat of overt nominative subjects is in finite clauses. But infinitival complements and infinitival adjuncts are also known to have overt nominative subjects, e.g. in Italian (Rizzi 1982), European Portuguese (Raposo 1987), and Spanish (Torrego 1998, Mensching 2000). The analyses make crucial reference to the movement of Aux or Infl to Comp, and to overt or covert infinitival inflection. This working paper is concerned with a novel set of data that appear to be of a different sort, in that they probably do not depend on either rich infinitival inflection or on movement to C.

Formal semantic analyses often take words to be minimal building blocks for the purposes of compositionality. But various recent theories of morphology and syntax have converged on the view that there is no demarcation line corresponding to the word level. The same conclusion has emerged from the compositional semantics of superlatives. In the spirit of extending compositionality below the word level, this paper explores how a small set of particles (Japanese KA and MO, Chinese DOU, and Hungarian VALA/VAGY, MIND, and IS) form quantifier words and serve as connectives, additive and scalar particles, question markers, and existential verbs. The main question is whether the meanings of these particles across the varied environments are highly regular, or they are lexicalized with a variety of different meanings that bear a family resemblance. This paper does not reach definitive conclusions, but it raises analytical possibilities using Boolean semantics and the semantics of alternatives. It also draws attention to systematic similarities and some differences between MO and DOU that have not been studied in the literature.

Modifying the descriptive and theoretical generalizations of Relativized Minimality, we argue that a significant subset of weak island violations arise when an extracted phrase should scope over some intervener but is unable to. Harmless interveners seem harmless because they can support an alternative reading. This paper focuses on why certain wh-phrases are poor wide scope takers, and offers an algebraic perspective on scope interaction. Each scopal element SE is associated with certain operations (e.g., not with complements). When a wh-phrase scopes over some SE, the operations associated with that SE are performed in its denotation domain. The requisite operations may or may not be available in a domain, however. We present an empirical analysis of a variety of wh-phrases. It is argued that the wh-phrases that escape all weak islands (i.e., can scope over any intervener) are those that range over individuals, the reason being that all Boolean operations are defined for their domain. Collectives, manners, amounts, numbers, etc. all denote in domains with fewer operations and are thus selectively sensitive to scopal interveners—a “semantic relativized minimality effect”

Semantics plays a role in grammar in at least three guises. (A) Linguists seek to account for speakers‘ knowledge of what linguistic expressions mean. This goal is typically achieved by assigning a model theoretic interpretation in a compositional fashion. For example, *No whale flies* is true if and only if the intersection of the sets of whales and fliers is empty in the model. (B) Linguists seek to account for the ability of speakers to make various inferences based on semantic knowledge. For example, *No whale flies* entails *No blue whale flies* and *No whale flies high*. (C) The wellformedness of a variety of syntactic constructions depends on morpho-syntactic features with a semantic flavor. For example, *Under no circumstances would a whale fly* is grammatical, whereas *Under some circumstances would a whale fly* is not, corresponding to the downward vs. upward monotonic features of the preposed phrases. It is usually assumed that once a compositional model theoretic interpretation is assigned to all expressions, its fruits can be freely enjoyed by inferencing and syntax. What place might proof theory have in this picture?

We show in rather informal terms how witness sets can be useful in both explicating some basic intuitions about scope and understanding how particular denotational semantic differences between noun phrases affect their abilities to bear out certain scopal patterns. More generally we suggest that the usual notion of scope needs to be factored into variation distributivity and maximality. This part lays some groundwork for several of the subsequent chapters and is thus of interest to all readers. The second part shows that already in this initial raw form the above insights can be applied to make a novel claim concerning the availability of so-called branching readings. In logical terms a branching reading can be defined for any sentence with a subject and a direct object. However speakers of English accept only a fraction of these readings, so the question arises how the data can be predicted from the meanings of the participating quantifiers and the syntactic structure of the sentence. We propose that thinking about the behavior of quantifiers along the lines introduced in the first part leads to a simple answer to this question.

The de Morgan laws characterize how negation, conjunction, and disjunction interact with each other. They are fundamental in any semantics that bases itself on the propositional calculus/Boolean algebra. This paper is primarily concerned with the second law. In English, its validity is easy to demonstrate using linguistic examples. Consider the following: (3) Why is it so cold in here? We didn’t close the door or the window. The second sentence is ambiguous. It may mean that I suppose we did not close the door or did not close the window, but I am not sure which. This `I am not sure which’ reading is irrelevant to us because it has disjunction scoping over negation. But the sentence may equally well mean (and indeed this is the preferred reading) that we didn’t close the door and did not close the window. This `neither’ reading bears out de Morgan law (2).

The scalar approach to negative polarity item (NPI) licensing assumes that NPIs are allowable in contexts in which the introduction of the NPI leads to proposition strengthening (e.g., Kadmon & Landman 1993, Krifka 1995, Lahiri 1997, Chierchia 2006). A straightforward processing prediction from such a theory is that NPI’s facilitate inference verification from sets to subsets. Three experiments are reported that test this proposal. In each experiment, participants evaluated whether inferences from sets to subsets were valid. Crucially, we manipulated whether the premises contained an NPI. In Experiment 1, participants completed a metalinguistic reasoning task, and Experiments 2 and 3 tested reading times using a self-paced reading task. Contrary to expectations, no facilitation was observed when the NPI was present in the premise compared to when it was absent. In fact, the NPI significantly slowed down reading times in the inference region. Our results therefore favor those scalar theories that predict that the NPI is costly to process (Chierchia 2006), or other, nonscalar theories (Giannakidou 1998, Ladusaw 1992, Postal 2005, Szabolcsi 2004) that likewise predict NPI processing cost but, unlike Chierchia (2006), expect the magnitude of the processing cost to vary with the actual pragmatics of the NPI.

The focus of this paper is the syntax of inverse scope in Hungarian, a language that largely disambiguates quantifier scope at spell-out. Inverse scope is attributed to alternate orderings of potentially large chunks of structure, but with appeal to base-generation, as opposed to nonfeature-driven movement as in Kayne 1998. The proposal is developed within mirror theory and conforms to the assumption that structures are antisymmetrical. The paper also develops a matching notion of scope in terms of featural domination, as opposed to c-command, and applies it to otherwise problematic cases of pied piping. Finally, the interaction of different quantifier types is examined and the patterns are explained invoking morphological considerations on one hand and A-bar reconstruction on the other.

Combinatory logic (Curry and Feys 1958) is a “variable-free” alternative to the lambda calculus. The two have the same expressive power but build their expressions differently. “Variable-free” semantics is, more precisely, “free of variable binding”: it has no operation like abstraction that turns a free variable into a bound one; it uses combinators—operations on functions—instead. For the general linguistic motivation of this approach, see the works of Steedman, Szabolcsi, and Jacobson, among others. The standard view in linguistics is that reflexive and personal pronouns are free variables that get bound by an antecedent through some coindexing mechanism. In variable free semantics the same task is performed by some combinator that identifies two arguments of the function it operates on (a duplicator). This combinator may be built into the lexical semantics of the pronoun, into that of the antecedent, or it may be a free-floating operation applicable to predicates or larger chunks of texts, i.e. a typeshifter. This note is concerned with the case of cross-sentential anaphora. It adopts Hepple’s and Jacobson’s interpretation of pronouns as identity maps and asks how this can be extended to the cross-sentential case, assuming the dynamic semantic view of anaphora. It first outlines the possibility of interpreting indefinites that antecede non-ccommanded pronouns as existential quantifiers enriched with a duplicator. Then it argues that it is preferable to use the duplicator as a type-shifter that applies “on the fly”. The proposal has consequences for two central ingredients of the classical dynamic semantic treatment: it does away with abstraction over assignments and with treating indefinites as inherently existentially quantified. However, cross-sentential anaphora remains a matter of binding, and the idea of propositions as context change potentials is retained.

Section 1 provides a brief summary of the pair-list literature singling out some points that are particularly relevant for the coming discussion. -/- Section 2 shows that the dilemma of quantification versus domain restriction arises only in extensional complement interrogatives. In matrix questions and in intensional complements only universals support pairlist readings, whence the simplest domain restriction treatment suffices. Related data including conjunction, disjunction, and cumulative readings are discussed -/- Section 3 argues that in the case of extensional complements the domain restriction treatment is inadequate for at least two independent reasons. One has to do with the fact that not only upward monotonic quantifiers support pairlist readings, and the other with the derivation of apparent scope out readings. The reasoning is supplemented with some discussion of the semantic properties of layered quantifiers. The above will establish the need for quantification, so the question arises how the objections explicitly enlisted in the literature against quantification can be answered. Section 4 considers the de dicto reading of the quantifiers restriction, quanticational variability, and the absence of pairlist readings with whether questions, and argues that they need not militate against the quantificational analysis. -/- Section 5 summarizes the emergent proposal -/- Finally, section 6 discusses the significance of the above findings for the behavior of weak islands.

[...] I will only investigate [Austin's] claims as challenges to present-day model theoretic semantics. My main point will be to draw a sharp line between the semantic and pragmatic aspects of performatives and thereby discover a gap in Austin’s treatment. This will in my view naturally lead to the proposal in Section 2, that is, to treating performatives as denoting changes in intensional models. The rest of Section 2 will be concerned with the status of felicity conditions and a tentative extension of Montague’s PTQ.

This paper uses a partially ordered set of syntactic categories to accommodate optionality and licensing in natural language syntax. A complex but well-studied data set pertaining to the syntax of quantifier scope and negative polarity licensing in Hungarian is used to illustrate the proposal. The presentation is geared towards both linguists and logicians. The paper highlights that the main ideas can be implemented in different grammar formalisms, and discusses in detail an implementation where the partial ordering on categories is given by the derivability relation of a calculus with residuated and Galois-connected unary operators.