Categorisation as topographic mapping between uncorrelated spaces

Abstract

In this paper, I propose a neurophysiologically plausible account for the evolution of arbitrary, categorical mental relationships. Topographic, or structure-preserving, mappings are widespread within animal brains. If they can be shown to generate behaviours in simulation, it is plausible that they are responsible for them in vivo. One behaviour has puzzled philosophers, psychologists and linguists alike: the categorical nature of language and its arbitrary associations between categories of form and meaning. I show here that arbitrary categorical relationships can arise when a topographic mapping is developed between continuous, but uncorrelated activation spaces. This is shown first by simulation, then identified in humans with synaesthesia. The independence of form and meaning as sensory or conceptual spaces automatically results in a categorial structure being imposed on each, as our brains attempt to link the spaces with topographic maps. This result suggests a neurophysiologically plausible explanation of categorisation in language. © 2013 Springer-Verlag Berlin Heidelberg.