As people increasingly interact with large language models (LLMs), a critical question emerges: do humans process language differently when communicating with an LLM versus another human? While there is good evidence that people adapt comprehension based on their expectations toward their interlocutor in human–human interaction, human–computer interaction research suggests the adaptation to machines is often suspended until expectation violation occurs. We conducted two event-related potential experiments examining Chinese sentence comprehension, measuring neural responses to semantic and syntactic anomalies attributed to an LLM or a human. Experiment 1 revealed reduced N400 but larger P600 responses to semantic anomalies in LLM-attributed text than human-attributed one, suggesting participants anticipated semantic errors yet required increased composition/integration efforts. Experiment 2 showed enhanced P600 responses to LLM-attributed than human-attributed syntactic anomalies, reflecting greater reanalysis or integration difficulty in the former than in the latter. Notably, neural responses to LLM-attributed semantic anomalies (but not syntactic anomalies) were further modulated by participants’ belief about humanlike knowledge in LLMs, with a larger N400 and a smaller P600 in participants with stronger belief of humanlike knowledge in LLMs. These findings provide the first neurocognitive evidence that people develop mental models of LLM capabilities and adapt neural processing accordingly, offering theoretical insights aligned with multidisciplinary frameworks and practical implications for designing effective human–AI communication systems.

Pater's (2019) target article builds a persuasive case for establishing stronger ties between theoretical linguistics and connectionism (deep learning). This commentary extends his arguments to semantics, focusing in particular on issues of learning, compositionality, and lexical meaning.

This article proposes a syntax and a semantics for intonation in English and some related languages. The semantics is ‘surface-compositional’, in the sense that syntactic derivation constructs information-structural logical form monotonically, without rules of structural revision, and without autonomous rules of ‘focus projection’. This is made possible by the generalized notion of syntactic constituency afforded by combinatory categorial grammar (CCG)—in particular, the fact that its rules are restricted to string-adjacent type-driven combination. In this way, the grammar unites intonation structure and information structure with surface-syntactic derivational structure and Montague-style compositional semantics, even when they deviate radically from traditional surface structure.

The article revises and extends earlier CCG-based accounts of intonational semantics, grounding hitherto informal notions like ‘theme’ and ‘rheme’ (a.k.a. ‘topic’ and ‘comment’, ‘presupposition’ and ‘focus’, etc.) and ‘background’ and ‘contrast’ (a.k.a. ‘given’ and ‘new’, ‘focus’, etc.) in a logic of speaker/hearer supposition andupdate, usingaversion of Rooth's alternative semantics. A CCG grammar fragment is defined that constrains language-specific intonation and its interpretation more narrowly than previous attempts.

This article presents four experiments that investigate the meaning of English and Italian statements containing the epistemic necessity auxiliary verb must/dovere, a topic of long-standing debate in the philosophical and linguistics literature. Our findings show that the endorsement of such statements in a given scenario depends on the participants' subjective assessment about whether they are convinced that the conclusion suggested by the scenario is true, independently from their objective assessment of the conclusion's likelihood. We interpret these findings as suggesting that English and Italian speakers use epistemic necessity verbs to communicate neither conclusions judged to be necessary (contrary to the prediction of the standard modal logical view) nor conclusions judged to be highly probable (contrary to the prediction of recent analyses using probabilistic models) but conclusions whose truth they believe in (as predicted by the analysis of epistemic must as an inferential evidential). We suggest that this evidential meaning of epistemic must/dovere might have arisen in everyday conversation from a reiterated hyperbolic use of the words with their original meaning as epistemic necessity verbs.

This article reports on the results of a broad crosslinguistic study on the semantics of quantity words such as many in the superlative (e.g. most). While some languages use such a form to express both a relative reading (as in Gloria has visited the most continents) and a proportional reading (as in Gloria has visited most continents), the vast majority do not allow the latter, though all allow the former. It is argued that a degree-quantifier analysis of quantity words is best suited to explain why proportional readings typically do not arise for quantity superlatives. Based on morphosyntactic evidence, two alternative diachronic pathways through which proportional quantifiers may develop from quantity superlatives are identified.

Von Prince, Krajinović, and Krifka (2022) argue that irrealis is a crosslinguistically legitimate semantic category, and they define it in terms of a domain encompassing both future possibility and counterfactuality. In this response, we argue that this definition is too narrow, because it excludes past and present possibility and necessity. We suggest instead that the correct characterization is that irrealis expressions correlate with quantification over possible worlds—or in simpler terms, with modality. We then ask a compositional question: do irrealis expressions signal the presence of modality contributed by other morphemes in the clause, or do they contribute modality themselves? Based on a comparison between the languages in von Prince et al.'s sample and preliminary data from Lutuv (Lautu) Chin (South Central Tibeto-Burman, formerly called Kuki-Chin), we suggest that the answer to this question may vary from one language to the next, thereby contributing to a richer picture of how modal meaning is reflected and encoded crosslinguistically.

While published linguistic judgments sometimes differ from the judgments found in large-scale formal experiments with naive participants, there is not a consensus as to how often these errors occur nor as to how often formal experiments should be used in syntax and semantics research. In this article, we first present the results of a large-scale replication of the Sprouse et al. 2013 study on 100 English contrasts randomly sampled from Linguistic Inquiry 2001-2010 and tested in both a forced-choice experiment and an acceptability rating experiment. Like Sprouse, Schütze, and Almeida, we find that the effect sizes of published linguistic acceptability judgments are not uniformly large or consistent but rather form a continuum from very large effects to small or nonexistent effects. We then use this data as a prior in a Bayesian framework to propose a small n acceptability paradigm for linguistic acceptability judgments (SNAP Judgments). This proposal makes it easier and cheaper to obtain meaningful quantitative data in syntax and semantics research. Specifically, for a contrast of linguistic interest for which a researcher is confident that sentence A is better than sentence B, we recommend that the researcher should obtain judgments from at least five unique participants, using at least five unique sentences of each type. If all participants in the sample agree that sentence A is better than sentence B, then the researcher can be confident that the result of a full forced-choice experiment would likely be 75% or more agreement in favor of sentence A (with a mean of 93%). We test this proposal by sampling from the existing data and find that it gives reliable performance.

Previous research on the acquisition of noun classification systems (e.g. grammatical gender) has found that child learners rely disproportionately on phonological cues to determine the class of a new noun, even when competing semantic cues are more reliable in their language. Culbertson, Gagliardi, and Smith (2017) use artificial language learning experiments with adults to argue that this likely results from the early availability of phonological information during acquisition. Learners base their initial representations on formal features of nouns, only later integrating semantic cues from noun meanings. Here, we use these same methods to show that early availability affects cue use in children (six- to seven-year-olds) as well. However, we also find evidence of developmental changes in sensitivity to semantics; when both cue types are simultaneously available, children are more likely to rely on phonology than adults are. Our results suggest that both early availability and a bias favoring phonological cues contribute to children's overreliance on phonology in natural language acquisition.

The present study uses naturally occurring conversational data from various dialects of Spanish to examine the role of second-person (T/V) reference forms in the accomplishment of social action in interaction. I illustrate how the turn-by-turn progression of talk can occasion shifts in the linguistic means through which speakers refer to their hearers, an interactional commonality between dialects (and possibly languages) that are otherwise pronominally dissimilar. These shifts contribute to the action of an utterance by mobilizing the semantic meaning of a pronominal form in order to recalibrate who the interactants project they are, and who they project they are to one another—not in general, but rather at that particular moment in the ongoing interaction. The analysis posits a distinction between identity status and identity stance to argue in favor of a more microlevel conceptualization of identities and contexts as emergent features of moment-by moment discourse, co-constructed through the deployment of grammatical structure.

We argue that DISTRIBUTIONAL MODIFICATION is one strategy that language affords for composing propositions about the quantity of entities that participate in a given situation. Distributional modifiers apply to kind descriptions, contributing the entailment that the kind is instantiated by a SET of tokens with a particular distribution. As a case study, we analyze frequency adjectives (FAs, e.g. occasional). We show that previous work, including our own, has suffered for focusing on the paraphrases of FAs rather than on their morphosyntax. We argue for two subclasses of FAs: those that are intersective modifiers sortally restricted to events, and those that are not. The study reinforces two novel theoretical claims in Gehrke & McNally 2011: sometimes kinds are realized by sets of tokens, rather than individual tokens; and some clauses constitute descriptions of EVENT KINDS, rather than EVENT TOKENS.

The question of whether Irrealis is a meaningful concept in crosslinguistic comparison has been the subject of long-standing controversy. In this article, we argue that the semantic domain of irreality is split into two domains—the possible and the counterfactual—and that an ‘irrealis’ marker in a given language may refer either to only one of these domains or to both. A significant part of the crosslinguistic variation in what is referred to by the term irrealis can be traced back to this distinction. Other factors that obscure the realis/irrealis divide include functional subdivisions of the irrealis domain and paradigmatic competition within the TAM system of a language. We conclude that ‘irrealis’ is a crosslinguistically meaningful notion.

Eckert (2008) rightly points out that context, variation, and indexicality are inextricably bound. This work—an in-depth case study of the social significance of the English definite article—presents a picture whereby semantic meaning is part of that same web of interrelations. The primary empirical claim of this work is that using the with a plural NP (e.g. the Americans) to talk about all or typical members of a group of individuals tends to depict that group as a monolith separate from the speaker, and to an extent that using a bare plural (e.g. Americans) does not. I present two variationist, corpus-based studies that provide clear evidence of this effect. I then provide a principled account of the effect, building on the insights of sociolinguistic and pragmatic research and extending their collective reach. As I show, the effect is largely rooted in crucial differences between the semantic meaning of the-plurals and that of related alternative expressions. As with a broad range of associated phenomena, the exact interpretation of a particular the-plural on a given occasion of use depends importantly upon its indexical character, the beliefs of the speech participants, and myriad other contextual factors, but is nonetheless constrained in a principled way.

Serial verb constructions have often been said to refer to single conceptual events. However, evidence to support this claim has been elusive. This article introduces co-speech gestures as a new way of investigating the relationship. The alignment patterns of gestures with serial verb constructions and other complex clauses were compared in Avatime (Ka-Togo, Kwa, Niger-Congo). Serial verb constructions tended to occur with single gestures overlapping the entire construction. In contrast, other complex clauses were more likely to be accompanied by distinct gestures overlapping individual verbs. This pattern of alignment suggests that serial verb constructions are in fact used to describe single events.

The structural focus of linguistics has led to a static and modular treatment of meaning. Viewing language as practice allows us to transcend the boundaries of subdisciplines that deal with meaning and to integrate the social indexicality of variation into this larger system. This article presents the expression of social meaning as a continuum of decreasing reference and increasing performativity, with sociolinguistic variation at the performative extreme. The meaning potential of sociolinguistic variables in turn is based in their form and their social source, constituting a cline of ‘interiority’ from variables that index public social facts about the speaker to more internal, personal affective states.

We explore children's use of syntactic distribution in the acquisition of attitude verbs, such as think, want, and hope. Because attitude verbs refer to concepts that are opaque to observation but have syntactic distributions predictive of semantic properties, we hypothesize that syntax may serve as an important cue to learning their meanings. Using a novel methodology, we replicate previous literature showing an asymmetry between acquisition of think and want, and we additionally demonstrate that interpretation of a less frequent attitude verb, hope, patterns with type of syntactic complement. This supports the view that children treat syntactic frame as informative about an attitude verb's meaning.