Parents Fine-tune Speech to Children’s Vocabulary Knowledge

• In Event: The Diversity of Caregiver-child Coordination and its Contributions to Language Development

Abstract

Language development in children is uniquely predicted by the quantity and quality of child-directed speech they hear (Weisleder & Fernald, 2013). The linguistic tuning hypothesis posits that the power of child-directed speech comes from its adaptive nature—parents tune their speech to children’s developmental level, keeping children’s linguistic input at an optimal complexity for learning (Snow, 1972). How precise is this tuning? One possibility is that parents coarse-tune speech to their children’s general linguistic abilities. Alternatively, parents might also fine-tune their speech to their children’s knowledge of specific words.

To test the precision of caregivers’ linguistic tuning, we asked parents and children (n = 41; ages 2 to 2.5) to play an interactive iPad game, in which the parent’s goal was to direct their child to select a target animal from a set of three (Fig. 1). Each animal appeared as target twice during the study. Before the session, parents completed a vocabulary checklist for their child that included the target animals in the study.

We found three major results. First, our results show that parents coarse-tuned their language to the general difficulty of each animal word, producing longer referring expressions for animals that were reported to be known by fewer children (p < .034) (Figure 2). Second, parents fine-tuned their language based on beliefs about their child’s knowledge of specific animal words, using longer utterances to describe animals they thought their particular child did not know (p = .003). Third, linguistic tuning can occur in real time: when a child selected incorrectly on a trial where the parent thought they knew the target animal, parents then used a longer referring expression when the same animal appears as the target later on in the game (p < .001). Our results thus support a strong form of the linguistic tuning hypothesis—over and above coarse-tuning, parents fine-tune speech to their children’s specific vocabulary knowledge. Over the course of this tuning, parents also provide rich descriptive language: parents produced more descriptive speech for words they think their children do not know. Specifically, they used more adjectives (p < .001), comparisons to other animals (p = 0.001), and basic level category labels (p = .026) for these animals. Interestingly, canonical labels were almost always used, even for animals that parents thought their children did not know. By connecting what children know (e.g. color words) to what they do not know (e.g. unfamiliar animals), parents can scaffold word learning by providing opportunities to draw the connection between a referent, its label, and its identifying characteristics.

In daily interactions, parents can leverage their knowledge of children’s language development to provide linguistic input that is tailored to their specific children’s needs. These adaptations can facilitate communication in-the-moment, and may also play an important role in children’s long-term learning.

Authors

• Ashley Leung, University of Chicago
  Presenting Author

• Alexandra Tunkel, George Washington University
  Non-Presenting Author

• Daniel Yurovsky, Carnegie Mellon University
  Non-Presenting Author