Embodied cognition, embodied regulation, and the Data Rate Theorem

Abstract

The Data Rate Theorem that establishes a formal linkage between linear control theory and information theory carries deep implications for the design of biologically inspired cognitive architectures (BICA), and for the more general study of embodied cognition. For example, modest extensions of the theorem provide a spectrum of necessary conditions dynamic statistical models that will be useful in empirical studies. A large deviations argument, however, suggests that the stabilization of such systems is itself an interpenetrating dynamic process necessarily convoluted with embodied cognition. As our experience with mental disorders and chronic disease implies, evolutionary process has had only modest success in the regulation and control of cognitive biological phenomena. For humans, the central role of culture has long been known. Although a ground-state collapse analogous to generalized anxiety appears ubiquitous to such systems, lack of cultural modulation for real-time automatons or distributed cognition man-machine 'cockpits' makes them particularly subject to a canonical pathology under which 'all possible targets are enemies'. More general dysfunctions of large-scale topology and connectivity analogous to autism spectrum and schizophenoform disorders also appear likely. A kind of machine psychiatry may become a central engineering discipline as the number of computation cores in real-time critical systems increases exponentially over the next few decades.