The Three Languages of the Brain: Quantum, Reorganizational, and Associative
Subhash C. Kak, Karl H. Pribram, Joseph S. King · 1996
this paper I consider the computational aspects of the problems of perception and adaptation in light of dual and associative processes. First, I summarize the limitations of computational models by considering questions raised by new researches in animal intelligence. The central insight obtained from the study of animal intelligence is that it is predicated on continual selforganization, as seen, for example, in superorganisms. That biological processing has a quantum basis has been argued by several authors. Quantum models provide a natural explanation for the unity of awareness in addition to explaining other puzzling features of brain behavior. In one class of such models, quantum behavior is postulated within neurons. But this does not resolve the question of the continuing self-organization of a biological system. My own proposal (Kak, 1992, 1995b) looks at organization and information as new quantum mechanical variables and I call this holistic view as quantum neural computing. This topic is reviewed and its implications are described. For any quantum phenomenon there should be classical approximate representations. Since self-organization is the basic feature of biological processing, one needs to consider an explicit signaling scheme for this. This additional signal provides a dual to the usual neural transmissions in parallel with the many-component vectors of a quantum description. This bottom-up dual signaling regime for the brain may be taken to complement the top-down quantum view. This paper also considers the associative learning problem, that deals with the most basic linguistic category of how associations are implemented. When a pattern is presented to a Mathematical World Mental World Forces Physical World