Proceedings Published by the LSA (Linguistic Society of America)
Not a member yet
4323 research outputs found
Sort by
A dynamic neural field model of asymmetric interference effects in code-switching
Full text linkDuring instances of code-switching, bilingual speakers rapidly transition between languages. Following a transition from one language to the other, a code-switched word can exhibit phonetic differences from the same word produced in a single-language context. Observation of such ‘interference’ effects depends on language dominance. Interference effects are reported when a speaker switches from their non-dominant language into their dominant language, as shown, for example, in measurements of voice onset time (VOT) from Spanish-English bilinguals. We propose a neurocognitive model of such effects using Dynamic Field Theory (DFT). Interference arises from the interaction of separate language inputs into a single VOT planning field. Following principles from the inverse frequency effect, the amplitudes of the two language inputs are modulated by the frequency of language use, deriving the asymmetry. A key assumption underlying this result is that bilinguals’ speech representations interact in a shared phonetic space
Echo questions as anaphoric phrases in English
Full text linkEcho questions have proven to be a difficult question type to explain using syntactic generalizations. This analysis explains why both syntactic and semantic concepts should be used to interpret echo questions. It will describe the basic properties of echo questions, including how they typically match an antecedent. This is explained by applying focus semantics, where focus is interpreted on the echo question and requires an alternative. The antecedent is determined as this alternative, and results in an anaphoric relationship between the two utterances. This analysis further examines this topic by discussing in detail the types of mismatches that arise between echo questions and their antecedents. Entailment is shown to be a proposed solution as to why certain types of mismatches are acceptable
Variation and gradience in non-standard Turkish affix ordering
Full text linkSemantically vacuous affix order variation is a cross-linguistically rare phenomenon that promises to shed light on the principles underlying affix order in general. This paper discusses a case of free affix order in Turkish: if a verb contains more than one TAM morpheme, any or all of them can be followed by an agreement morpheme. I seek to determine which factors influence the acceptability of a given order for a given set of morphemes by collecting new data from 19 native speaker consultants. The results reveal a sharp bifurcation between some orderings that are perceived as categorically grammatical by all speakers and others that result in gradient and variable judgments without any discernible patterns. I discuss these findings against the background of previous approaches to affix order variation, highlighting the extent to which analyses are influenced by methodological choices about data collection
An exploration of principled mappings between English adjective order and subjectivity
Full text linkAdjective order in English is both very strict and largely intuitive. Recent research proposes subjectivity as an underlying mechanism for adjective order. The present studies aim to examine whether subjectivity is a productive cue for adjective ordering through two separate grammaticality judgment tasks. In the first task, participants rated grammatical and ungrammatical sentences containing color (typically closer to the noun) and size (typically further from the noun) adjectives paired with images of novel objects, kertunks. In the second task, participants completed the same design but with tables instead, and grammatical and ungrammatical sentences containing material (typically closer to the noun) and shape (typically further from the noun) adjectives. In each task, the images contained some combination of clear and ambiguous traits. If participants are sensitive to subjectivity in the moment, they should rate the ungrammatical sentences (e.g., color-before-size and material-before-shape) higher when the presented image is either of an ambiguous color or material and of a clear size or shape. We find that, in both tasks, participants are not more accepting of the ungrammatical order even when that order aligns with visual subjectivity in a given scene. This suggests that adjective order is not flexible based on context, and that any systematic relationship between subjectivity and order is learned and codified rather than relied upon in the moment
A mixed-quotational account of indirect discourse: Evidence from self-pointing gestures
Full text linkThis paper argues that indirect discourse is a form of mixed quotation. It further posits that self-pointing gestures in indirect discourse, when aligned with a third-person pronoun co-referent with the matrix subject, constitute a character viewpoint gesture quoted from the matrix subject (cf. Ebert & Hinterwimmer 2022). To formally model this, Davidson’s (2015) demonstrational account of quotation is combined with Ebert & Ebert’s (2014) approach to gesture semantics. This analysis also readily explains observations that certain indexicals can shift in indirect discourse (Plank 1986; Anderson 2019), by reinterpreting them as quotations from the matrix subject
Future-less-vivid conditionals and the modal past
Full text linkFuture-Less-Vivid conditionals (FLVs) are conditionals that display the typical morphological marking of counterfactuals, but whose antecedent has future reference time. An example is in (i):(i) If Ada took semantics next term, she would take logic next year. The literature has coalesced on a near-consensus that FLVs cannot be contrary-to-fact. In this paper, I argue that the near-consensus is wrong. FLVs can be genuinely counterfactual: in particular, FLVs are counterfactuals about the future, i.e. can involve suppositions that contradict settled future facts. This has an interesting theoretical upshot. The behavior of FLVs is challenging for all theories on which tenses affect root modals by backshifting the time index of the modal base. These theories include all so-called Past-as-Past theories of X-marking. Conversely, the behavior of FLVs can be accommodated by Past-as-Modal theories
Experimental Paradigms on Scalar Implicature Estimation
Full text linkExperimental research on the processing of Scalar Implicatures (SIs) relies on behavioral tasks that purport to measure the rate at which scalar implicatures are computed within an experimental paradigm. Two paradigms, the Truth Value Judgment Task (TVJT) (Gordon, 1998; Crain & Thornton, 2000) and the Picture Selection Task (PST) (Gerken & Shady, 1998) have dominated the experimental pragmatics literature; yet the effects of task choice on implicature rate have remained underexplored. Here we report the results of three studies testing participants in the TVJT and the PST using three different linguistic scales in English: “ad-hoc”, “or-and”, and “some-all”. In the first experiment, the task variation was manipulated within subjects while in the second experiment, it was manipulated between subjects. The third experiment examined a variant of the PST called the Hidden Card Task (HCT) which is increasingly used in the context of priming research (Bott & Chemla, 2016). We found that the estimated rate of scalar implicature computation varied noticeably between different tasks. This suggests that the experimental paradigm itself has a significant impact on our estimates of the implicature rate for a given linguistic scale, and thus, researchers studying scalar implicatures need to carefully consider the effect of experimental paradigms in experimental design and the interpretation of their results
Linguistic and Social Meaning Match: An experiment on modal concord in English
Full text linkModal concord (MC) refers to the phenomenon where two modal elements of the same flavor and force in a sentence yield an interpretation of single modality (SM). In this paper, we report on an experimental study on MC in English, addressing their linguistic and social meaning. Our results show a strengthening effect by necessity MC and a weakening effect of possibility MC in that significantly higher speaker commitment ratings were received for necessity MC vs. SM constructions (i.e., must certainly vs. must) with the reverse pattern for possibility modal constructions (i.e., may possibly vs. may). Furthermore, MC and SM were shown to differ in social meanings, suggesting a correlation between the meaning strength of a linguistic expression and the social perception of the speaker
Mechanistic Support Language in Colombian Spanish-speakers
Full text linkBeyond basic spatial relations (e.g., teddy on table), we know little about how children learn to talk about Mechanical Support events (e.g., objects attached/hung from a surface via tape) and map them onto linguistic structures. Moreso, the majority of the research that has been done focuses on children learning English - a language that has several verbs that lexicalize support via a specific mechanism (Levin, 1993; e.g., glue, tape, clip, etc.). The current study seeks to deepen our understanding of spatial language acquisition by diversifying the populations that have been studied. Specifically, 4- to 6-year-old monolingual Spanish-speaking children and adults in Colombia viewed Mechanical Support events (e.g., girl puts paper on door via tape) and were then asked, ‘Can you tell me what my sister did with my toy?’. Both children and adults used Non-Mechanism (e.g., poner = ‘put’, colgar = \u27hang\u27) and Mechanism Verbs (e.g., pegar = \u27stick\u27); the use of Mechanism Verbs increased from 4- to 6- years of age. In addition, whether the mechanism was visible in the event influenced how it was mapped to language; when the mechanism was visible (vs. when it was hidden), children and adults were more likely to encode the mechanism in a prepositional phrase (e.g., lo colgó con un gancho = ‘she hung it with a clip’). These findings shed light on the development of Mechanical Support language in Spanish- speaking children, the influence of context - specifically, visibility of mechanism - on language, as well as the lexicalization patterns for encoding physical support in Spanish more generally
Missing the cues: LLMs’ insensitivity to semantic biases in relative clause attachment
Full text linkWe investigate whether large language models (LLMs) replicate English speakers\u27 well-established preference for low attachment in relative clause (RC) ambiguities, and how they respond when semantic cues such as world knowledge and stereotypical associations (e.g., age or gender plausibility) conflict with this preference. Eight commercial LLMs spanning the Claude-3/3.5 and GPT-3.5/4o families were evaluated using structurally and semantically ambiguous stimuli, alongside items that introduced plausibility-based biases toward either high or low attachment. In the absence of disambiguating cues, all models showed a strong preference for low attachment, consistent with human tendencies in ambiguous contexts (i.e. no semantic bias cues). However, models varied in their sensitivity to semantic information: newer Claude-3.5 models frequently shifted toward high attachment when the LA interpretation was implausible, while GPT-based models rarely did so. Attachment preferences were also affected by prompt format, suggesting that LLMs do not consistently integrate syntactic and semantic information in a stable, human-like way. These findings highlight both convergence and divergence between LLMs and human sentence processing, offering insight into the limits of current pretraining paradigms in handling structural ambiguity and world knowledge