Infants’ use of Shared Utility to Infer Unidirectional and Bidirectional Affiliation

• In Event: Investigating the Development of Social Affiliation Across Infancy and Early Childhood

Abstract

This presentation will argue that a core component of infants’ concept of affiliation is the probability that one individual will adopt the utility of their social partner. Utility adoption, or sharing, refers to one social partner taking on another’s goal as their own, and underlies helping and cooperation (Hamlin, Ullman, et al., 2013). If infants also consider utility adoption to be part of lasting social affiliation, then the more strongly they think A is affiliated with B, the more likely they will think A is to adopt B’s utility. This concept may allow inferences that are either unidirectional (e.g. A likes B, and will adopt B’s utility) or bidirectional (e.g. A and B like one another, and will share utility), depending on the evidence infants receive.

First, I will argue that infants’ expectations and evaluations following observations of social imitation are best explained as arising from the inference that imitators have a prosocial or affiliative propensity to adopt their social partner’s utility. This proposal makes sense of a number of findings: infants’ unidirectional expectations that imitators will approach their targets but not vice versa (Powell & Spelke, 2018a), infants’ preference for imitators as good social partners (Powell & Spelke, 2018b), and the disruption of this preference when imitation is not based on perceptual access to the model (Powell & Spelke, 2017).

Second, I will present novel data from an experiment testing if infants use shared utility to infer mutual affiliation (i.e. social group membership), in a domain-specific manner. Infants habituate faster to sequences made up of entities they perceive as more similar vs. more different (e.g. Pun et al., 2017). We used this aspect of infant visual attention to test the hypothesis that shared utility, expressed through joint behavior, is a cue to shared category membership selectively in the context of animate entities. Six- to 10-month-old infants observed an introductory sequence featuring two round, red characters and two quadrilateral, blue characters. Infants in the one-group condition saw all four characters moving together in close proximity, while infants in the two-group condition saw the characters grouped in two separate pairs (Figure 1). All infants subsequently saw a sequence of habituation trials featuring one character at a time, alternating between red and blue characters. Completion of data collection is delayed due to Covid-19 (and will be completed online), but with 42 of 50 intended participants, a mixed linear model finds evidence that infants’ looking declined more across trials in the one-group condition than the two-group condition (t(498) = 1.9, p = .06; Figure 2). A second group of infants (N=40 of 50) saw the same events, but carried out by shapes with a black-and-white stripe, rather than a face; when the entities were thus presented as physical objects, there was no interaction between the introductory grouping and habituation rate (t(474) = .272, p > .7). These findings suggest that infants use shared utility--exclusively the property of animate entities--as a cue to similarity of social affiliation.

Author

• Lindsey Powell, University of California - San Diego
  Presenting Author