In some discourses, a given inference would be natural at one point, but impossible at the next. We focus here on such cases with inferences of temporal and causal order usually described as ambiguities of discourse coherence. The fact that these inferences come and go over the course of a discourse raises challenges for their representation within theories of dynamic semantics: representations must either include coherence-related meaning in a way that can be selectively edited later, or else persistently underspecify it. As theories of humans' actual cognitive states, these approaches make different predictions for patterns of difficulty during real-time comprehension. Borrowing a standard assumption from sentence processing, true incremental representation should be associated with some detectable "Regret" when a comprehender meets input incompatible with a previously-preferred analysis. Previous work on the processing of coherence has failed to demonstrate Regret, leaving underspecification on the table. In a new study, we probe further, by constructing discourses with a strong, verified bias towards an initial inference, and testing the effects of cues which can reverse this inference early in the discourse, or much later. We observe that these discourses provoke Regret in self-paced reading, in particular when the critical cue comes late, in a discourse unit following the initial ambiguity. Among many formal accounts which could explain this pattern, we elaborate in particular on a version of Segmented Discourse Representation Theory where analyses are ranked at the offset of each discourse unit.
We describe a novel interpretive difference between x wonders hopefully ?p and x wonders fearfully ?p, where the former is only compatible with x hoping p, but the latter, with x either fearing p or not p. Extending a goal-based analysis of root polar questions by Tabatowski (2022) to embedding under inquisitive predicates, we argue that this difference arises due to hopefully imposing general constraints on an agent’s goals, in particular on their attitudes in entertaining questions, that fearfully does not.
We propose a formal model for dealing with what we call outlook markers, a subtype of discourse markers that govern discourse management pertaining to negotiations about ‘agreeable outlooks’ (on possibly subjective matters). For this purpose, we take up the semantic analysis of an ‘emphatic’ discourse marker nanka in Japanese. We argue that nanka is essentially a counterstance marker in that it imposes a constraint on the prior discourse that a contextually salient ‘counterstance’ is at issue, and that the speaker’s own stance embodied in the prejacent is incongruent with this counterstance as a subjective judgment. The model we develop combines Coppock’s (2018) outlook-based semantics for the treatment of subjectivity and Farkas & Bruce’s (2010) Table model for discourse management. We demonstrate that this unified model illuminates the way in which the notoriously elusive evaluative implication of nanka arises from the very nature of this particle as an abstract, meta-level discourse marker: it explicitly identifies the discourse conflict as a conflict of in principle irreconcilable opinions (which could threaten the whole point of ‘agreeable outlook’ management).
This work investigates three patterns of polarity (in)sensitivity in the domain of adjectival degree modification: modifiers such as fairly are positive polarity items, those such as (all) that are negative polarity items, and those such as very are not polarity sensitive. Modifiers like fairly and (all) that belong to the broad class of attenuating polarity items (Israel 1996), which are widespread crosslinguistically but not well accounted for by leading semantic theories of polarity sensitivity. We demonstrate that their behavior can nonetheless be captured in an alternative-based semantic framework, by supplementing strength-based competition with a second mechanism, namely simplicity-based competition. We demonstrate that this approach can provide a general account of the three classes in question, and discuss potential extensions within the degree domain and beyond, as well as some open questions that remain.
The interaction between modified numerals and modals offers important insights into their interpretation. As pointed out by Buring (2008), not all superlative numerals interacting with modals give rise to the same interpretations: while both at least and at most interacting with universal modals give rise to authoritative/bounded-interpretations, only at most interacting with existential modals does so. In the case of interactions with at least, the readings are of speaker-insecurity/ignorance. This paper offers a new perspective on this interaction, building on a focus-based Krifka (1999)-style analysis for superlatives as well as a view of context structured by questions under discussion. The pragmatic component of the proposal makes a case for enrichment driven by considerations of speaker cooperativity and authority. The account thus differs from proposals in the literature that appeal to scalar alternatives and disjunction-style reasoning.
Neg-raising predicates show a special behavior with respect to propositional anaphora. When negated, a neg-raising predicate makes a discourse referent available for the negation of the proposition denoted by its embedded clause. Non-neg-raising predicates do not allow such anaphora. The behavior is puzzling since no constituent of the sentence denotes the negation of the embedded clause. I suggest that an excluded middle presupposition-based approach, cf. Bartsch 1973, to neg-raising offers a solution. I propose that one disjunct of the excluded middle presupposition introduces a discourse referent for the negation of the embedded clause. This discourse referent is accessible conditionally on the negation of the other disjunct. Such an interaction between disjunction an anaphora is observed by Rothschild 2017.
This paper examines three prominent approaches to Free Choice (FC) and Dual Prohibition (DP): the implicature account relying on exhaustivity (EXH; Fox 2007; Bar-Lev & Fox 2017, 2020), the cognitive-bias account based on Neglect Zero (NZ; Aloni 2022), and a recent development of the implicature account grounded in presuppositional exhaustification (PEX; Bassi, Del Pinal & Sauerland 2021; Del Pinal, Bassi & Sauerland 2024). While all three approaches capture the FC-DP alternation in simple sentences (e.g., Leo can/cannot buy the cupcake or the pear), they diverge in their predictions regarding the strength of FC and DP in non-monotonic environments (e.g., Exactly one child can/cannot buy the cupcake or the pear). We tested these predictions in two experiments investigating the possible interpretations of simple and non-monotonic sentences varying in polarity. The results align with the predictions of the PEX approach, while posing challenges for the other two. We conclude by discussing potential theoretical refinements for the EXH and NZ accounts and by highlighting remaining open questions for PEX.
Frustrative particles modify a clause by conveying the non-realization of an associated expectation. The semantics of frustrativity are relatively understudied, complicated by variation in whether the relationship between the marked clause and the ‘target’ of frustration is lexically or only pragmatically specified. We contribute to the emerging picture by reporting on O’dam, which is typologically unusual in having two contrastive frustratives: t1i and t1ipup. Where most existing analyses treat frustratives as clausal markers of non-inertia, we argue that O’dam frustratives are also sensitive to event-based non-maximality. We propose a unified approach on which the differences are derived from a parametric presuppositional contrast: the weaker t1i commits its speaker to the non-inertial status of the evaluation world, while t1ipup imposes a stronger form of non-inertia by blocking maximal instantiation of the marked-clause event. Our analysis suggests that the non-inertial approach is appropriate for weak frustratives (which do not fully preclude ‘frustrated’ outcomes; Kroeger 2024), but opens the door to a reanalysis of strong frustrativity (Copley & Harley 2014; Davis & Matthewson 2022: a.o.) in the crosslinguistic landscape.
The standard of precision (SoP) is the discourse parameter governing how much imprecision is tolerated in a given context. Speakers can implicitly negotiate the SoP through metalinguistic disagreements (MDs)—moves that challenge the assertability of an (im)precise utterance in order to signal that the SoP should be updated. Prior work has claimed that such negotiations are asymmetric: MDs can effectively raise the SoP but cannot lower it (e.g., Klecha 2018; Lewis 1979). However, recent experimental work has failed to find evidence for this purported asymmetry (Wu & Aparicio 2025a), motivating further empirical investigation into the validity of this claim. We present results from two studies in which participants engaged in metalinguistic (dis)agreements over (im)precise utterances. Our findings reveal a bidirectional update pattern: MDs can shift the SoP both upward, as previously claimed, and downward, contrary to earlier accounts (Klecha 2018; Lewis 1979). However, these two update types differ qualitatively: participants making imprecise utterances tend to abandon their original stance in favor of the disagreeing view, whereas participants making precise utterances tend to accommodate the disagreement without abandoning their own position. We suggest that this contrast may underlie earlier intuitions about the unidirectionality of SoP updates.
Traditionally, only is thought to negate non-weaker alternatives. Kratzer’s (1989) still life example, presented more than thirty years ago, challenges this purely logical approach to the semantics of only, appearing to show that the constitution of objects matters to what only negates. This paper reexamines Kratzer’s example in the light of later developments in the literature on only. We discuss four types of responses: (i) uphold the traditional view of only supplemented by lexical postulates (van Rooij 2005), (ii) replace entailment in the meaning of only with ‘lumping’, a variant of entailment defined in terms of situation semantics (von Fintel 1997; Bonomi & Casalegno 1993; Kadmon 2001), (iii) adopt an innocent exclusion semantics for only (Fox 2007), (iv) following Kratzer’s (1989) suggestion, appeal to a general no-overlap constraint on counting domains. We show that all approaches but (iv) face empirical challenges. Despite some loose ends, we favour (iv) in the comparison, as it faces no clear empirical challenges, though significant work remains to be done.
The Japanese classifier tsu is used for counting separable inanimate entities, and when combined with the numeral hito ‘one’, as in hito-tsu, it indicates that the number of entities is one. However, hitotsu can also function as a scalar particle, where hitotsu itself is interpreted as ‘even’. In this paper, I analyze the meaning and use of hitotsu as a scalar particle by comparing it with ordinary scalar particles, such as sae ‘even’, and show that the scalar particle hitotsu imposes a semantic restriction that the noun it attaches to must be non-specific and interpreted as the theme of an event. I argue that the scalar particle hitotsu is sensitive to event kinds and creates contrasts among events typically denoted by a VP. It will be shown that the meaning and distribution pattern of the scalar particle hitotsu can be explained based on the idea of pseudo-incorporation (e.g., Carlson 2003; Gehrke 2019). In studies on the semantics of EVEN, various approaches have been proposed– such as scope theory (e.g., Karttunen & Peters 1979), lexical ambiguity theory (e.g., Rooth 1985), and degree-based theory (e.g., Greenberg 2018). These approaches generally assume that EVEN functions as a propositional operator. This paper demonstrates that, unlike ordinary scalar particles, the scalar particle hitotsu ‘even’ is not a propositional operator. Instead, it represents a new type of EVEN, which is sensitive to thematic roles and event kinds.
Questions with nested wh-phrases like "which book by which author" cannot receive pair-list readings. We propose that this restriction falls out of a general constraint against questions whose answers are all Strawson-equivalent, together with certain independently motivated assumptions about the syntax-semantics of multiple wh-questions. Support for our proposal comes from Bulgarian, a language that has been argued to display key components of the LF of wh-questions overtly. If this is true, the language teaches us that there are two LFs that can in principle be associated with nested wh-questions, only one of which is subject to the restriction observed in English. This distinction is a direct consequence of our proposal.
Mandarin polarity-sensitive item rènhé, just like English any, is ungrammatical under a necessity modal; nevertheless, this may be ameliorated by the presence of the quantificational adverb dou, with the rènhé-DP carrying a universal construal. Crucially, the presence of dou alone is not sufficient to render rènhé grammatical. We suggest that the distribution of Mandarin rènhé in a modal environment is best captured by the two-layered approach by Crniˇc (2019a,b, 2022), according to which i) the license of any is subject to Strawson downward entailment, and ii) exhaustification merely plays a supplementary role by creating an environment hospitable to any. Along with these lines, we suggest that because of the universal force dou introduces, the scope of a necessity modal may be turned SDE and hence rènhé may be rescued in this environment.
Agreement morphology is often assumed to realize φ-features unin- terpreted at LF. We provide experimental evidence from French relative clauses showing that agreement can in fact have an interpretative effect, adding to the range of ‘semantic agreement’ effects (e.g. Charnavel & Sportiche 2025). The choice of variable person agreement turns out to be sensitive to the placement of focus. We provide an analysis that relies on the interaction between focus, binding and the distribution of focus-sensitive operators.
Most literature on too has focused on its "regular" uses, where, as a first approximation, too triggers a presupposition that one of the focus alternatives to its prejacent, distinct from the prejacent itself, is true (e.g., Axel likes cats. BROOK likes cats, too.). There has also been some discussion of "refutational too", used in positive polarity responses that refute the antecedent utterance (e.g., A: You didn't do your homework. B: I did too!). Recently, Thomas (2023) proposed that refutational too is a response particle, akin to yes and no—or even more pertinently, to French si and German doch. In this paper, I present novel data on what I will call "confirmatory too", which is used in positive polarity responses that confirm the antecedent utterance (e.g., A: It’s hot. B: It is too.), as well as some additional data on some other additive particles in responses. I argue against the response particle analysis of such uses, proposing instead that both refutational and confirmatory uses of additive particles share their core semantics with their regular uses, with the difference being the focus alternatives: alternative commitments with respect to issue p? (a subcase or a sibling of verum focus) in refutational cases vs. alternative performances of a speech act making the same commitment (a subcase of dictum focus) in confirmatory cases.
Dynamic semantics standardly treats negation as externally static. This is inconsistent with anaphora to doubly-negated indefinites, and Partee disjunctions. I propose a technique for lifting dynamic systems into richer meaning spaces that support dynamic negation. I argue that extant theories of dynamic negation instantiate this technique in various ways, prove that all such lifted systems are isomorphic, and identify a minimal/canonical enrichment of Dynamic Predicate Logic. I relate this perspective on dynamic negation to recent work on 'bounded' meaning (Mandelkern 2022). Finally, I consider how such lifted interpretations with dynamic negations can be extended further, into richer meaning spaces that support exceptional scope of indefinites (Charlow 2025).
A longstanding debate about conjunction concerns so-called ‘asymmetric’ interpretations with additional causal or temporal inferences. Some theories posit a semantic component of conjunction as part of their explanation (e.g. Gómez Txurruka 2003). In this paper, we explore an unstated prediction of such accounts: that the specific semantics of conjunction might vary (i) across languages, and (ii) across conjunction strategies within a language. Drawing on data from A’ingae (an indigenous language of Ecuador) and Gizey (an indigenous language of Cameroon and Chad) respectively, we show that both predictions are met, arguing that such semantic variation is constrained in ways anticipated by theories of discourse coherence such as SDRT (Asher & Lascarides 2003).
In this paper, we investigate disjunctions that combine an assertion with a question. We will explore a number of puzzles arising from such construction and propose an account which construes speech act disjunction as a device to introduce alternative discourse strategies.
Sentences with accomplishments in before-clauses are prima facie ambiguous between a strong reading and a weak reading, and two types of theories of these sentences have been proposed. Under-specification theories posit that the before-clauses in question are not actually ambiguous: their only LF corresponds to the weak reading, with the strong reading falling out as a subcase. Rett’s (2020) ambiguity theory, by contrast, claims that the strong reading surfaces by default, with the weak reading only arising if we coerce the embedded accomplishment into an achievement. Under standard psycholinguistic assumptions, coercion incurs a processing cost, and hence, I argue, Rett’s (2020) theory makes processing predictions that under-specification theories do not. I test these predictions via two self-paced reading experiments and find moderate evidence that they are borne out.
There are several ways of overtly marking distributivity across languages. In this paper, we focus on three distributive constructions in Highland Mayan Sign Language (HMSL): ONE-smooth-arc, ONE-punctuated-arc, and ONE-alt. We show that HMSL patterns with other sign languages in making use of reduplication and space to mark distributivity in these constructions. However, we also observe micro-variation in the way these constructions are interpreted. Accordingly, we propose that the strategies observed thus far can be decomposed and that HMSL’s three distributive constructions are the result of the pluralizing of an allocation construction (‘x-FOR-i’) at different nodes. Our data further suggests that iconicity alone does not determine meaning and that it must be constrained by the morphosyntax.
