Proceedings of Sinn und Bedeutung
Not a member yet
1139 research outputs found
Sort by
Minimal sufficiency, plural predication, and scalarity
Full text linkMinimal sufficiency readings of exclusive modifiers (Just the thought of food makes me hungry) have resisted a comprehensive semantic analysis that accurately predicts their distribution. In this paper we show that the distribution of minimal sufficiency readings is directly correlated with the interpretation of plural arguments and that the distributional facts reflect the connection between plural predication and scalarity: sufficiency readings are licensed precisely in contexts where ordering relations over alternatives are reversed. We develop a semantics for exclusives that is capable of generating either exclusive or sufficiency readings depending on the direction of scalarity
Revisiting the role of structural complexity in symmetry breaking
Full text linkMany recent theories of strengthening assume following Katzir (2007); Fox and Katzir (2011) that the alternatives used to derive strengthened meanings may be at most as complex as the prejacent. We explore a novel response to several known problems for this view that maintains three of its core assumptions: the selection criterion is (i) structure-based, (ii) indifferent to whether the alternatives form entailment-based scales, and (iii) cannot be overruled by relevance. Contra Fox and Katzir, the crucial structural criterion is not complexity per se, but similarity to the prejacent. Moreover, this property only applies to break stalemates between symmetric alternatives; in the absence of stalemates, alternatives are structurally unconstrained. We suggest that while assumptions (i) and (ii) have good empirical consequences, this account is not entirely successful because (iii) is too strong: In certain special cases, which seem to have an information-structural characterization, stalemates can be broken by relevance
Unbinding equatives
Full text linkWhile traditional degree semantics posits that comparison constructions operate on scalar dimensions, property equatives–which compare individuals with respect to categorical properties–present a theoretical challenge to this view. I demonstrate that property equatives, despite lacking inherent scalar structure, exhibit parallel structural constraints with degree equatives, suggesting a unified semantic operation underlying both constructions. Based on novel data evidence, I propose a QUD-based semantics for equatives that accounts for both scalar and non-scalar interpretations. This analysis suggests that the scalar interpretation of degree equatives emerges from the interaction between question-based similarity assessment and the ordered nature of gradable predicates, rather than from inherent scalar semantics in the comparison operation itself
You can leave your dog outside, but do you have to?: An RSA approach
Full text linkA deontic possibility utterance can have a necessity interpretation in some contexts. For example, a café owner’s utterance of “You can leave your dog outside.” to a dog owner who is about to enter can be taken to mean that the dog owner has to leave their dog outside.This phenomenon, where a logically weaker item (possibility) implies a stronger item (necessity) on a scale, is not straightforwardly explained through standard theories of scalar implicature that predict that the assertion of a logically weaker item implies the negation of a stronger item. I attempt to account for this phenomenon of possibility-to-necessity inference using a theory of graded modality (Lassiter, 2011, 2017; Chung and Mascarenhas 2023) and the Rational Speech Act (RSA) framework (Frank and Goodman, 2012; Goodman and Stuhlmüller, 2013). I construct an RSA model where the pragmatic listener jointly infers about the state of the world and the value of the semantic variable as in Lassiter and Goodman (2013, 2017). The speaker utility includes social utility along epistemic utility, after Yoon et al. (2017). I consider two situations which differ in terms of whether explicit deontic necessity utterances are available to the speaker. The RSA model predicts a high probability of necessity interpretation in both. I further compare these results to the ones I obtain by controlling for components of the RSA model such as utterance cost and social utility, examining the effect that each has on the final result. The present study thus offers a way of systematically investigating a previously understudied phenomenon of deontic necessity inference arising from a deontic possibility utterance
Classifiers and comparison class: Evidence for cross-linguistic variation in the calculation of standards
Full text linkThe notion of comparison class is often invoked as a way to identify the standard against which vague adjectives are evaluated. However, the means by which such a standard is derived is the subject of long-standing debate. This paper brings new cross-linguistic evidence to bear on this question. In this paper we show that classifiers can attach to adjectives under certain conditions in the Shantou variety of Teochew (Southern Min), and that these ‘adjective classifiers’ are interpreted as expressing a comparison class. However, we argue that the means by which a standard is calculated from the classifier must proceed in a very different way from that proposed for more familiar comparison class expressions, such as English for-. We propose a novel analysis in which the dependency between the adjective and the modified noun – mediated by the classifier – is established via conventional implicature. The analysis takes cross-linguistic differences to derive at least in part from the morphosyntactic properties of the overt comparison class expression. The findings provide evidence that notions of comparison class may be established in very different structures and that the calculation of a standard from overt comparison class expressions may be subject to cross-linguistic variation
NPI licensing in attitude reports with content-bearing nominals
Full text linkNoun complement clauses as in the claim that pigs are dirty can host weak negative polarity items like any and ever that are licensed from the matrix clause. This property is shared with attitude verbs but contrasts with relative clauses to singular definite nouns, which is surprising in the light of recent predicate analyses of complement clauses. This puzzle has recently been observed by Sharvit (2024). In this paper, I show how a kind-based approach to attitudes can readily account for this
Pseudo-scoping out of relative clauses: A functional approach
Full text linkSentences where a definite DP is modified by a relative clause containing a universal quantifier (relativized DPs), like the supervisor that each volunteer reported to, license readings which carry separate presuppositions of uniqueness and existence for each volunteer—henceforth, ‘varying definite readings’. Barker 2022 argues that these readings involve the universal DP scoping out of the relative clause and above the definite determiner, proposing to analyze them using a non-local scope shifting mechanism, like quantifier raising (QR). In this paper, we argue for a functional interpretation of the relativized DP and propose that varying definite readings result from the definite DP denoting an ⟨e, e⟩ function, from volunteers to supervisors. We present two arguments in favour of the functional analysis, both of which involve readings a QR based approach fails to capture. Combining these arguments with the fact that the functional approach can also be minimally extended to derive varying definite readings, we conclude that non-local QR is not necessary to handle these examples. As a result, we can maintain the assumption that relative clauses are scope islands for universal quantifiers
Covert generic causatives in Korean: A dispositional ascription analysis
Full text linkThis paper introduces a rarely discussed type of causatives in Korean, i.e. covert generic causatives. The construction conststs of a sole causer subject which is strictly non-agentive and a verb which is strictly non-causative. These causatives describe a dispositional causative property of the subject that brings about a result state describe by the verb. I propose that the verb undergoes covert causativization (i.e. zero causativization), taking a non-agentive causer subject. This causativizing strategy is unexpected in two ways: (i) Typologically unexpected in that covert causativization shows a restriction on one semantic domain, namely non-agentive causers; (ii) Locally unexpected in that Korean already has an overt causativizing strategy. I argue that this construction fills in a paradigmatic gap for non-agentive causers, since overt causatives require agentive causers. I further propose dispositional ascription analysis to account for covert generic causatives, and further claim they are a type of dispositional middles,extending the typology of middles
Pseudo-incorporation, event kinds, and atelicity
Full text linkThis paper investigates how kind-referring arguments shape (a)telicity by examining aspectual contrasts involving pseudo-incorporation (PI) in Turkish. Building on Sağ\u27s (2019, 2022) analysis of Turkish PI and Chierchia’s (2023) approach to plural kinds and their effect on lexical aspect, I argue that PI involves singular kind reference at the event-kind level, establishing taxonomic event kinds. This approach provides an explanation for the puzzling atelicity of PI constructions, distinguishing them from caseless indefinites and canonical singular kind arguments. Additionally, it elucidates how PI relates to plural kind argumentation, highlighting their similar yet distinct characteristics. Ultimately, my analysis demonstrates that event-kind-level arguments play a fundamental role in determining the aspectual interpretation of verbal predicates, expanding our understanding of kind reference beyond the nominal domain
Modeling uncertainty, unawareness, and underspecification among Structural Causal Models
Full text linkWe argue that linguistic communication of information about causal relations frequently involves epistemic states of not only uncertainty about, but also unawareness of causal facts. We propose a simple model of Causal Beliefs with Unawareness (CBU), which augments Structural Causal Models with additional structure to model both uncertainty and unawareness