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.
