This site has been permanently archived. This is a static copy provided by the University of Southampton.

TY - GEN ID - cogprints314 UR - http://cogprints.org/314/ A1 - Heylighen, Francis Y1 - 1994/// N2 - Given that knowledge consists of finite models of an infinitely complex reality, how can we explain that it is still most of the time reliable? Survival in a variable environment requires an internal model whose complexity (variety) matches the complexity of the environment that is to be controlled. The reduction of the infinite complexity of the sensed environment to a finite map requires a strong mechanism of categorization. A measure of cognitive complexity (C) is defined, which quantifies the average amount of trial-and-error needed to find the adequate category. C can be minimized by "probability ordering" of the possible categories, where the most probable alternatives ("defaults") are explored first. The reduction of complexity by such ordering requires a low statistical entropy for the cognized environment. This entropy is automatically kept down by the natural selection of "fit" configurations. The high probability, "default" cognitive categorizations are then merely mappings of environmentally "fit" configurations. PB - World Science, Singapore KW - fitness KW - default reasoning KW - requisite variety KW - decision-making KW - complexity KW - entropy KW - categorization KW - evolutionary epistemology KW - trial-and-error KW - cognitive complexity KW - natural selection KW - probability ordering TI - Fitness as Default: the evolutionary basis of cognitive complexity reduction SP - 1595 AV - public EP - 1602 ER -