The Cybersemiotic Roots Of Computation: A Critique of the Computational Model of Cognition

This paper challenges the prevalent metaphor of human cognition as a von Neumann- type (1945) computational process. This computational model of cognition is flawed because it fails to recognize the crucial role of an embodied observer's capacity for semiosis in any computational process. The paper argues against the computational model of cognition on epistemological, theoretical, practical, and ethical grounds. It affirms Brier's (1996) cybersemiotic framework , which states that semiosis is the organism's selection of environmental perturbations in the attempt to satisfy its own needs. The paper identifies the primary computational steps involved in the Turing (1936) machine and the von Neumann (1945) architecture, as well as those of three common applications of artificial intelligence. It then argues that each of these computational processes requires one or more of the human capacities for abstraction , purposive control of the physical environment , and judgment . It concludes that fully autonomous, self-adapting computers in some imagined utopian (or dystopian) future would diverge from human evolutionary relevance because they are incapable of semiosis. Keywords: cybersemiotics, semiosis, computation, artificial intelligence, second-order cybernetics, cognition.