Object Permanence as Relational Stability or How to Get Representation from the Dynamics of Embodiment

Object Permanence as Relational Stability or How to Get Representation from the Dynamics of Embodiment Jun Luo (jun.luo@utoronto.ca) Cognitive Science Program, University College, University of Toronto 15 King’s College Circle, Toronto, ON M5S 3H7 Canada reaches correctly. The same procedure is then done at the other well (called ‘B’). The child makes an “A-not-B error” if, during the ‘B trial’, they go back to search at A. Because the child repeats their A-trial response in a B trial, the A- not-B error is also called the “perseverative response”. The child outgrows such perseveration at about one year of age. For Piaget this phenomenon marks a transitional stage be- tween the beginning of a sense of “object permanence” (that occluded objects persist) and a robust grasp of it: the child can already retrieve a toy from a single hiding location, but they cannot yet cope with its “visible displacement” among multiple hiding locations. Many contemporary researchers used this phenomenon to “showcase” their research pro- grams: for Diamond (1990) it taps into the involvement of the prefrontal cortex, for Munakata (1998) it reveals the interplay of graded “active” and “latent” representation, for Marcovitch & Zelazo (1999) it is a window into conscious control, for Newcombe & Huttenlocher (2000) it illustrates the developmental interaction of locomotion and spatial coding. Thelen & Smith (1994) themselves already at- tempted an informal account of this phenomenon in their book that launched the DSA in developmental psychology. But, with the exception of Thelen & Smith, all such views are couched in explicitly representational terms: spatial code, location representation, conscious control, and so on. Against this background, we may understand the weight of the DSM for the DSA (Thelen et al 2001, p.2, e.a.): Abstract The dynamic systems approach is sometimes embraced in conjunction with causal coupling and in opposition to repre- sentation-based explanation. The dynamic systems model of the A-not-B phenomenon (Thelen et al, 2001) supposedly epitomizes this anti-representation alternative. It is argued here that this model fails to capture the crucial effects of body rotation in the A-not-B task and that we need a notion of rela- tional stability to account for these effects. Relational stability outstrips causal coupling but may ground a renewed under- standing of object permanence where representation is neces- sitated by and achieved through the dynamics of embodiment. Keywords: embodiment; representation; dynamic systems approach; A-not-B; object permanence; relational stability. Introduction Some advocates of the dynamic systems perspective in cog- nitive science take it to be an anti-representation alternative (Thelen & Smith, 1994; van Gelder, 1998; Thelen et al. 2001). This anti-representation stance, however, is not shared by many dynamic systems theorists of neural sys- tems (e.g. Seung 1996, Zhang 1996, Eliasmith & Anderson 2002) or even by theorists explicitly embracing the “dy- namic systems approach” (e.g. Elman, 1995; Beer, 1998). For discussion I will use the acronym “DSA”, with defer- ence to Thelen and Smith (1994), to mean the dynamic sys- tems perspective (i) restricted to psychological explanation solely in terms of causally coupled dynamic processes and (ii) taken to be anti-representational, but reserve the un- abridged “dynamic systems approach” for a general and possibly inclusive emphasis on the value of dynamic sys- tems analysis in cognitive science. The restriction to causal coupling is a central tenet of the DSA: it is to both provide a positive alternative to representation and steer it clear of the oft-laid accusation that dynamic systems modeling describes behavior without explicating the underlying mechanisms. To deliver on the promise of the DSA, Thelen et al (2001) offered a dynamic systems model of the A-not-B phenome- non (henceforth the DSM) from child development. Since Piaget (1954) identified it, this phenomenon has attracted the attention of generations of psychologists. Contemporary A-not-B experiments typically use a hiding device with two wells. The child (8- to 12-month-old) sits on the floor or in the caregiver’s lap watching the experimenter hide a toy in one of the wells. After hiding the toy, the wells are covered and the experimenter withholds the device to impose a delay of several seconds. The child is then allowed to reach and retrieve the toy. Such a trial is repeated a few times at the same well (called ‘A’ by convention) and the child typically Our message is: if we can understand this particu- lar infant task and its myriad contextual variations in terms of coupled dynamic processes, then the same kind of analysis can be applied to any task at any age. If we can show that “knowing” cannot be separated from perceiving, acting, and remember- ing, then these processes are always linked. There is no time and no task when such dynamics cease and some other mode of processing kicks in. Body and world remain ceaselessly melded together. The A-not-B task is representative of “any task at any age” because it had been invariably taken to mark the beginning of representational capacities for knowing that objects per- sist, remembering where they are, etc. Against this, the DSA proposes to treat it solely in terms of causally coupled dy- namic processes. The hope is: if the DSA were to succeed on this task, it could deal with, solo, any task later in life, whether or not it is traditionally viewed as representational. I shall note that the DSA’s take on the A-not-B task itself is also revisionary. While accepting the canonical design,

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.