What's embodied and how can we tell?

This paper by Daum, Sommerville, and Prinz (2009) presents an intriguing theory of the development of social cognition, distinguishing early developing embodied processing, based on perception and production of bodily states, from later emerging symbolic processing, based on language. While this distinction is discussed predominantly in relation to social understanding, it is broadly applicable to cognition more generally. In emphasizing the sensory and motor bases of infants’ and young children’s cognitive abilities, this proposal echoes classic Piagetian theory (e.g. Piaget, 1937/1954). Importantly, however, it avoids Piaget’s notion of qualitatively different developmental stages, suggesting instead that the embodied mode of understanding is essentially continuous from infancy through adulthood. This idea of continuity, shared with otherwise very different recent developmental theories (e.g. Spelke & Kinzler, 2007), is theoretically attractive in that it suggests that the study of infants and young children can provide direct insight into cognitive capacities which in adults are tightly integrated with other abilities and difficult to isolate experimentally. Specifically, Daum and colleagues suggest that preverbal infants may constitute a ‘pure model’ for embodied social cognition, uncontaminated (so to speak) by other types of processing. This is a fascinating proposal, with important methodological and theoretical consequences. In this commentary, I will discuss two major difficulties that pose limitations on this proposal as currently formulated, and end with a suggestion on how to conceptualize what’s embodied