Human beings are the most imitative creatures on the planet. Prior to language, human infants, use imitation to learn about physical objects and cultural practices [1]. Young children learn novel skills [2], abstract principles [3,4], and cultural customs from observing how other agents act upon the world. Imitation requires a mapping between action perception and action production. This mapping is now being delineated at both the psychological level [5] and at the neural level [6]. I will describe a theory of social-cognitive development-the "Like-Me" developmental theory [7]--that has emerged from studying how infants make sense of self, other, and their interaction. The theory proposes that one of the infant's first and most basic psychological acts is the recognition of others who act, move, and behave like the self. Preverbal children devote special attention to entities that are "Like-Me" and are predisposed to learn from "Like-Me" agents [7]. This theoretical framework, based on behavioral studies, is aligned with emerging findings in developmental neuroscience, including results using infant EEG [8], which is beginning to document the neural basis of "Like-Me" perceptions prior to language [5,9]. I will analyze infants' mapping between self and other, the role of experience in changing this initial state, and the impact of early self-other mapping on interpersonal functioning and cognitive development.
Topics covered include how infants understand the goals and intentions of other agents [10], and how they use imitation, gestures, and behavioral patterns--not simply morphological features--to recognize a person as the "same one again" after a break in perceptual contact [11], called the "social identity function" of imitation [12]. I will also discuss how infants come to understand the visual gaze of others [13] and use gaze to bootstrap language learning [14].
A foundation for infants coming to understand the gaze of others is their own self-experience with their own perceptual systems [15]. The "Like-Me" developmental theory describes this ontogenetic sequence, which is a crucial aspect of human interpersonal interaction, and a rare skill in the animal kingdom.
All of these findings with human infants-imitation, intention/goal understanding, person recognition, gaze following, and action coding and representation-have implications for robotics and human-computer interaction (HCI). At our Institute we have fostered several collaborations with computer scientists and roboticists, which have specifically analyzed how studies of infant behavior, brain development, and computer modeling can be combined to illuminate critical aspects of social learning and interpersonal understanding [16-19]. Most recently I have collaborated with a French team, and we have used computer modeling and robotics to examine the "social identity function" of gestural imitation [20]. Members of the French team will be at this conference. We will provide specific examples of the computer modeling and algorithms that were used in our social-interaction experiments with robots, adults, and children with autism. The overarching goal of our research program is to build bridges between human infant development, cognitive neuroscience, computer science, and social/ interpersonal psychology.
https://dl.acm.org/doi/10.1145/2823513.2830654 (2015)
Follow on work:
Peer Papers:
Old Papers: