Implicit causality biases and thematic roles in American Sign Language

Anne Therese Frederiksen,Rachel I Mayberry

doi:10.3758/s13428-021-01561-1

Anne Therese Frederiksen, Rachel I Mayberry

Open Access

https://doi.org/10.3758/s13428-021-01561-1

Copy DOI

Abstract

Implicit causality (IC) biases, the tendency of certain verbs to elicit re-mention of either the first-mentioned noun phrase (NP1) or the second-mentioned noun phrase (NP2) from the previous clause, are important in psycholinguistic research. Understanding IC verbs and the source of their biases in signed as well as spoken languages helps elucidate whether these phenomena are language general or specific to the spoken modality. As the first of its kind, this study investigates IC biases in American Sign Language (ASL) and provides IC bias norms for over 200 verbs, facilitating future psycholinguistic studies of ASL and comparisons of spoken versus signed languages. We investigated whether native ASL signers continued sentences with IC verbs (e.g., ASL equivalents of ‘Lisa annoys Maya because…’) by mentioning NP1 (i.e., Lisa) or NP2 (i.e., Maya). We found a tendency towards more NP2-biased verbs. Previous work has found that a verb’s thematic roles predict bias direction: stimulus-experiencer verbs (e.g., ‘annoy’), where the first argument is the stimulus (causing annoyance) and the second argument is the experiencer (experiencing annoyance), elicit more NP1 continuations. Verbs with experiencer-stimulus thematic roles (e.g., ‘love’) elicit more NP2 continuations. We probed whether the trend towards more NP2-biased verbs was related to an existing claim that stimulus-experiencer verbs do not exist in sign languages. We found that stimulus-experiencer structure, while permitted, is infrequent, impacting the IC bias distribution in ASL. Nevertheless, thematic roles predict IC bias in ASL, suggesting that the thematic role-IC bias relationship is stable across languages as well as modalities.

Highlights

The present study investigates how implicit causality (IC) biases are distributed in American Sign Language (ASL) verbs and provides norming data that can be applied in future studies
We investigate whether the phenomenon of body-anchoring affects the relationship between IC biases and thematic roles, how thematic role is lexicalized in signed as compared to spoken languages, and how this might influence the distribution of verb biases in ASL
We focus on previous findings related to realization of thematic roles in sign languages. These findings provide some expectations about IC biases in ASL and whether thematic roles should impact the distribution of biases in implicit causality verbs in the manual-visual modality

Summary

Introduction

The present study investigates how implicit causality (IC) biases are distributed in American Sign Language (ASL) verbs and provides norming data that can be applied in future studies. IC biases are at the basis of many psycholinguistic processes, but at this time, no published norms exist for IC biases in ASL or any other sign language. The prominent role of implicit causality (IC) biases in psycholinguistic research is due to their role in many processes in language comprehension, such as reading time, sentence processing and most notably in the resolution of pronominal anaphora (Garvey et al, 1976). Due to IC biases, sentences such as (1) and (2) do not cause comprehension problems, despite the fact that the pronoun is temporarily ambiguous. The pronoun is interpreted in line with the bias, effectively resolving its ambiguity

Methods

Results

Conclusion