Adults regularize variation when linguistic cues suggest low input reliability
Abstract
Children regularize inconsistent probabilistic patterns in linguistic input, yet they also acquire and match probabilistic sociolinguistic variation. What factors in the language input contribute to whether children will regularize or match the probabilistic patterns they are exposed to? Here, we test the hypothesis that low input reliability facilitates regularization. As a first step, we asked adult participants to acquire a variable plural marking pattern from a written (Exp 1) and a spoken (Exp 2) artificial language under different conditions, where they were led to believe input was more, or less, reliable. In both experiments, input reliability was manipulated through both information about the speaker (e.g., whether the speaker was likely to make mistakes) and linguistic cues (e.g., typos or pronunciation errors). Results showed that adults regularized the written language more only when they were told the speaker would make mistakes and the plural variants resembled typos (Exp 1), whereas they regularized the spoken language more when the plural variants resembled pronunciation errors regardless of the speaker’s said reliability in the spoken language. We conclude that input reliability is an important factor that can modulate learners’ regularization of probabilistic linguistic input, and that linguistic cues may play a more critical role than top-down knowledge about the speaker. The current study lays down an important foundation for future work exploring whether children are able to incorporate input reliability cues when learning probabilistic linguistic variation.
Keywords
Full Text:
PDFDOI: https://doi.org/10.3765/plsa.v7i1.5293
Copyright (c) 2022 Yiran Chen, Kathryn Schuler

This work is licensed under a Creative Commons Attribution 4.0 International License.

Linguistic Society of America
Advancing the Scientific Study of Language since 1924
ISSN (online): 2473-8689
This publication is made available for free to readers and with no charge to authors thanks in part to your continuing LSA membership and your donations to the open access fund.