The naming of colours: investigating a psychological curiosity using AI

The paper chronicles a case study investigation into colour naming decisions, a phenomenon at the boundary of language and perception. Colours in the region between blue and green are found difficult to name reliably, and it is suggested that decisions may be based primarily on intentional, context dependent language processes, rather than on simply perceptual mechanisms. Various models and codes for colour naming exist which make different assumptions based on theories of colour vision. A challenge for AI is to provide normative models to better understand aspects of colour vision, by using techniques such as fuzzy logic and artificial neural networks. This is based on the assumption that classifications of colour stimuli can be established and corroborated with case data, thus enabling clarification of psychological models. The paper outlines a preliminary research proposition, and some initial findings.