Social contingency facilitates infants’ vocabulary growth above and beyond word input

• In Event: The Diversity of Social Contingency and its Contributions to Social and Communicative Development

Abstract

In this project we investigate the role of socially contingent interactions in facilitating infants’ understanding of the communicative, referential, and intentional nature of language. We hypothesize that by observing a parent responding to their actions with language or producing language in response to their babbling, infants may infer that language is the means of intentional communication and that specific words trigger specific outcomes. By observing specific behaviors, attention shifts, or object manipulation following utterance of specific words by one’s interlocutor, infants may infer that words are referentially linked to mental representations or mental states that trigger behaviors. These inferences lead to a qualitative shift in infants’ understanding of the nature of words and lead to advances in language acquisition, such as more rapid vocabulary growth, which occurs around 12-14 months of age (see Bergelson, 2020).
Previous research provides tentative support to this hypothesis, showing that infants’ experience with socially contingent interactions in the first year of life predicts their subsequent vocabulary size (e.g., Bigelow & Rochat, 2006; Bornstein et al., 2015). However, these investigations did not control for the quality and quantity of word input. This leaves open the possibility that infants’ vocabulary growth was due to the increased word input received in the context of socially contingent interactions and not because of a qualitative shift in infants’ understanding of the nature of words. To address this concern and investigate whether social contingency predicts language development above and beyond word input we conduct a video corpus investigation. We measure infants’ experience with socially contingent interactions in the first year to predict their subsequent vocabulary growth, while controlling for the quality and quantity of word input.
Additionally, in contrast with prior research that focused on parents’ responses to infants’ vocalizations, we investigate both bidirectionally and unidirectionally verbal contingent interactions, including parents’ responses to infants’ behaviors and infants’ responses to parents’ behaviors. Adults’ non-verbal reactions to children’s vocalizations, as well as adults’ speech in response to infants’ actions, provides critical insight into the connection between speech and other social actions.
We used the Rollins corpus (Rollins, 2003) and analyzed 20-minute video-recordings of naturalistic parent-infant interactions, in which parent-infant dyads played with toys and books when infants were 9, 12, 18 months. We coded infants’ and parents’ verbal and non-verbal behaviors, including object-directed behaviors, and identified socially contingent exchanges as semantically related behaviors, directed at the interlocutor, which occur within 2 seconds from the interlocutor’s action. Infants’ receptive vocabulary was measured at 12, 18, and 22 months.
Preliminary results (n=10) show that the frequency of socially contingent responses to object-directed behaviors at 9 months predicts infants’ vocabulary size at 12 months (p<.05), controlling for word types, and tokens, and parents’ mean length of utterance. These preliminary results support our hypothesis and clarify earlier findings, suggesting that the effect of social contingency on infants’ vocabulary cannot be reduced to increased word input. Video annotation and analyses for 10 additional dyads at 9 months and all dyads at 12 and 18 months are in progress.

Authors

• Elena Luchkina, Northwestern University
  Presenting Author

• Fei Xu, University of California - Berkeley