ABSTRACT
In chapter 6, we sketched in broad outline a computational theory that claims that cognition-or at least re-cognition-can be caused by connecting up a network, even when the latter consists only of dumb information processor nodes. Such networks will, over a number of trials, produce a representation of rather complex environments, and thus, in one sense of the word at least, process information, or compute. Originally (McClelland et al., 1988; Rumelhart et al., 1986), this was a theory of how the equivalent of a brain (conceived of as a neural net) might think but we discovered that othersWeick and Roberts, and Hutchins-were soon prepared to go further and to see parallel distributed processing in a network not only as a metaphor for how cooperatively working groups of humans function, but also as a theory of collective organizing (at least in the small). Using data recorded in naturally occurring situations (“in the wild,” in Hutchins’ words), they have tried to show how groups composed of individuals with distributed-segmented, partial-images of a complex environment can, through interaction, synthetically construct a representation of it that works: one which, in its interactive complexity, outstrips the capacity of any single individual in the network to represent and discriminate events. This is a theory that says not only that people put their ideas together summatively to produce a common pool of knowledge, but also that, out of the interconnections, there emerges a representation of the world that none of those involved individually possessed or could possess.
