The Need for a Diverse Audiovisual Corpus of Infant-Directed Behavior

• In Event: 2-011 - Interactive Poster Session and Reception

Abstract

In interactions with infants, adult spontaneously modify their behavior in ways that engage attention and promote learning. Adults often use "motherese" (infant-directed talk), modifying the speech they address to infants (e.g., higher pitch, shorter utterances, repetition). "Motionese" or infant-directed action displays analogous modifications (e.g., exaggerated range of motion, shorter activity bouts, repetition). Likely, both motionese and motherese are part of a larger suite of infant-directed behavior modifications sometimes termed "natural pedagogy" (Csibra & Gergely, 2009). While some knowledge exists about the nature of natural pedagogy and its benefit to infants, we have not yet achieved a comprehensive understanding of the phenomenon. To what extent does natural pedagogy vary cross-culturally? How do individual elements of natural pedagogy (i.e., motherese and motionese) relate to each other? Is the extent to which these modifications are used consistent within the individual? There has been some investigation of cross-linguistic differences in motherese (e.g., Grieser & Kuhl, 1988; Fernald, Taeschner, Dunn, & Papousek, 2009) and, regarding motionese, researchers in the United States (e.g., Brand, Baldwin, & Ashburn, 2002), Europe (e.g., van Schaik et al., 2017), and Japan (e.g., Fukuyama et al., 2014) have been independently studying these infant-directed action modifications. However, to answer big-picture questions regarding the form and function of natural pedagogy, one would ideally have access to audiovisual data that integrates across this research. Thus, my goal is to assemble a set of audiovisual corpora that will enable researchers to address fundamental questions about the structure of infants' input and thus their learning circumstances.

Authors

• Jessica Elizabeth Kosie, University of Oregon
  Presenting Author

• Dare Baldwin, University of Oregon