The Brain’s Best Guess

Abstract

In this final chapter, I propose that behavior-based calibration of perceptions and abstract representations are constrained by a preconfigured brain. The nervous system may have evolved to mimic the statistical probabilities of the physical world and the behavior of already existing species and thus become an efficient predictor of events. Because of their high diversity, neurophysiological and perceptual brain dynamics, both spanning several orders of magnitude, share a common mathematical foundation: the log rule. The tails of these wide and skewed distributions have apparently distinct qualitative features that we describe by discrete words, such as familiar and novel, rigid and plastic, good-enough and precise. Yet every novel situation contains elements of familiarity. Brain correlates of newly acquired experience are not created in the sense of adding new neuronal words to an ever-expanding vocabulary. Instead, the preconfigured brain is a dictionary in which the behavioral significance or meaning of initially nonsense neuronal words is acquired through exploration.