Why Do Child-Directed Interactions Support Imitative Learning in Young Children?

Abstract

Child-directed cues support imitation of novel actions at 18 months, but not at two years of age. The current studies explore the mechanisms that underlie the propensity that children have to copy others at 18 months, and how the value of child-directed communication changes over development. We ask if attentional allocation accounts for children's failure to imitate observed actions at 18 months, and their success at two years of age, and we explore the informational value child-directed contexts may provide across ontogeny. Eighteen-month-old (Study 1) and two-year-old (Study 2) children viewed causally non-obvious actions performed by child-directed (Study 1 & 2), observed (Study 1 & 2), or non-interactive (Study 2) actors, and their visual attention and imitative behaviors were assessed. Results demonstrated that child-directed contexts supported imitative learning for 18-month-old children, independent of their effects on proximal attention. However, by two years of age, neither directness nor communication between social partners was a necessary condition for supporting social imitation. These findings suggest that developmental changes in children's propensity to extract information from observation cannot be accounted for by changes in children's interpretation of what counts as child-directed information, and are likely not due to changes in how children allocate attention to observed events.

Child‐directed teaching and social learning at 18 months of age: evidence from Yucatec Mayan and US infants

“…Thus, Mayan infants displayed a very different pattern of results from the US infants described in previous studies (e.g. Brugger et al ., ; Király et al ., ; Matheson et al ., ; Nielsen, ; Sage & Baldwin, ; Shneidman et al ., ). Infants seemed to be learning generally about the lab setting, perhaps by becoming more comfortable with the room or with the experimenters, or gaining experience manipulating stimuli, and thus performed more imitative actions on their second visit than on their first visit; however, child‐directedness was not a relevant cue for informing this imitation.…”

“…Given these properties of infants' early social experiences, in Study 1 we consider whether Yucatec Mayan infants display the same pattern of responses to child‐directed and observed contexts as infants from the United States have demonstrated in prior studies (e.g. Brugger et al ., ; Király et al ., ; Matheson et al ., ; Nielsen, ; Sage & Baldwin, ; Shneidman et al ., ). Using a nearly identical procedure as Shneidman et al .…”

“…In several studies 11‐ to 18‐month‐old infants were more likely to faithfully copy others' actions when they were directly taught the action (e.g. when the actor looked at and talked to the infant), as compared to when they simply observed these actions (Brugger, Lariviere, Mumme & Bushnell, ; Király, Csibra & Gergely, ; Matheson, Moore & Akhtar, ; Nielsen, ; Sage & Baldwin, ; Shneidman, Todd & Woodward, ). For example, at 18 months, infants were more likely to imitate the particular means of a demonstrated novel action (like using the head to turn on a light, or an elbow to activate a switch) when an experimenter looked and talked to the infant while performing the action, as compared to when she talked to herself (Matheson et al ., ) or to another person (Shneidman et al ., ).…”

“…when the actor looked at and talked to the infant), as compared to when they simply observed these actions (Brugger, Lariviere, Mumme & Bushnell, ; Király, Csibra & Gergely, ; Matheson, Moore & Akhtar, ; Nielsen, ; Sage & Baldwin, ; Shneidman, Todd & Woodward, ). For example, at 18 months, infants were more likely to imitate the particular means of a demonstrated novel action (like using the head to turn on a light, or an elbow to activate a switch) when an experimenter looked and talked to the infant while performing the action, as compared to when she talked to herself (Matheson et al ., ) or to another person (Shneidman et al ., ). Differential imitation following child‐directed and observed events occurred even when infants deployed equal visual attention to these contexts, suggesting that child‐directed situations provide informational value beyond the ways in which they shape infants' attention in the moment (Sage & Baldwin, ; Shneidman et al ., ).…”

“…Using a nearly identical procedure as Shneidman et al . (), we ask if Mayan infants, like US infants, demonstrate more robust imitative behavior when directly addressed as compared to when observing an interaction. If, as has been suggested, there is something special about child‐directed interactions for informing infants' conceptual understanding of an actor's intentions, or for triggering a learning stance to treat information as generalizable, then one would expect child‐directed communication to be important for supporting imitative learning even in this community where child‐directed interactions are a relatively uncommon form of early social experience.…”