Context-Sensitivity and Individual Differences in the Derivation of Scalar Implicature

Abstract

The derivation of scalar implicatures for the quantifier some has been widely studied to investigate the computation of pragmatically enriched meanings. For example, the sentence “I found some books” carries the semantic interpretation that at least one book was found, but its interpretation is often enriched to include the implicature that not all the books were found. The implicature is argued to be more likely to arise when it is relevant for addressing a question under discussion (QUD) in the context, e.g., when “I found some books” is uttered in response to “Did you find all the books?” as opposed to “Did you find any books?”. However, most experimental studies have not examined the influence of context on some, instead testing some sentences in isolation. Moreover, no study to our knowledge has examined individual differences in the ability to utilize context in interpreting some, whereas individual variation in deriving implicatures for some sentences in isolation is widely attested, with alternative proposals attributing this variation to individual differences in cognitive resources (e.g., working memory) or personality-based pragmatic abilities (e.g., as assessed by the Autism-Spectrum Quotient). The current study examined how context influences the interpretation of some in a story-sentence matching task, where participants rated some statements (“I cut some steaks”) uttered by one character, in response to another character’s question (QUD) that established the implicature as relevant (“Did you cut all the steaks?”) or irrelevant (“Did you cut any steaks?”). We also examined to what extent individuals’ sensitivity to QUD is modulated by individual differences via a battery of measures assessing cognitive resources, personality-based pragmatic abilities, and language abilities (which have been argued to modulate comprehension in other domains). Our results demonstrate that QUD affects the interpretation of some, and reveal that individual differences in sensitivity to QUD are modulated by both cognitive resources and personality-based pragmatic abilities. While previous studies have argued alternatively for cognitive resources or personality-based pragmatic abilities as important for deriving implicatures for some in isolation, we argue that arriving at a context-sensitive interpretation for some depends on both cognitive and personality-based properties of the individual.

On the Variability of Scalar Implicature Computation: Evidence from L2 English Adults

“…Our findings seem to converge with a recent study that did find a relationship between participants' personality traits and the proportion of scalar implicature interpretation [80]. Yang et al examined the comprehension of scalar implicatures triggered by the quantifier some in a story-sentence matching task and found that those with greater sociopragmatic abilities (as measured by the AQ) had higher sensitivity to pragmatic violations, and thus evidence that lends support to accounts arguing about the likelihood personality traits modulate the sensitivity to under-informative statements (e.g., [33,81]).…”

“…Second, this study also asked the participants to judge the infelicity of the under-informative statements in an artificial context. However, a general limitation of testing some in isolation is the absence of a more natural communicative setting that provides information about whether or not the implicature is relevant, and therefore a better approach in which one can examine how individuals would interpret some in scenarios akin to everyday language situations [80,94,95]. Artificial contexts may also discourage the participants to project their own personality characteristics and their own attitudes which may in turn modulate their comprehension and interpretation of critical utterances [76].…”

Early ERP Evidence for Children’s and Adult’s Sensitivity to Scalar Implicatures Triggered by Existential Quantifiers (Some)

“…In simple terms, it constitutes a very efficient way to enhance the scalar inference originated by some (cf. Yang et al, 2018, for a study that uses the same strategy to enhance the derivation of scalar implicatures).…”

Exploring the connection between Question Under Discussion and scalar diversity

“…It is a robust empirical finding that B's utterance, Mary ate some of the cookies, gives rise to the familiar some but not all SI at a higher rate in (4) than in ( 5) -see i.a. Degen & Tanenhaus (2015); Ronai & Xiang (2020); Yang et al (2018); Zondervan et al (2008) for converging results. Such findings can be given an explanation along the following lines.…”