Funded Grants

More than Mere Colouring: A dialogue between philosophy and neuroscience on the nature of spectral vision

In its broadest form, the aim of my research is to bring together two disciplines with a shared intellectual territory - to see what the combined resources of the neurosciences and philosophy can tell us about the nature of mind and its relation to the world. As I understand it, the relation between these two disciplines is, or at least ought to be, an interactive one. In one direction, the new science of the mind can bring fresh empirical insight to a variety of traditional philosophical problems. First, and most obviously, insofar as the traditional questions are empirical, the neurosciences will eventually provide empirical answers. Second, and more controversially, I expect that the neurosciences will change the nature of the questions asked, both by re-drawing the boundaries between the empirical and the conceptual as presently conceived, and by re-structuring the ontology of the mental as we commonly understand it, i.e. by changing our understanding of our own psychology. In the other direction, neuroscientifically informed philosophy can contribute to the analysis of neuroscientific methodology and, more importantly, given its intellectual history, participate in the interpretation of neuroscientific experimental data.

To illustrate just how this interaction between philosophy and the neurosciences might work, I present one example, colour vision. By appealing to only widely accepted neuroscientific results, I want to show how one might re-interpret the recent colour research in a way that differs from the widely accepted story, and do so in a way that causes our understanding of our own phenomenology of colour to change.

For the most part, both neuroscientists and philosophers have assumed what seems to be a tautology: that the purpose of colour vision is to see the external world "in colour". What colour vision buys us, in other words, is the ability to see the material objects of the world (as well as various media such as water and light) as coloured, as blue, red, yellow, etc. As a result of this view, the common expectation is that we must find in the brain a "colour centre", a place in the brain where "the colours" are mapped in neural space. On the contrary, I suspect that if there is a story to be told about the function of the colour system, it would show that we see a world of coloured objects only fortuitously, in virtue of the following chain of evolutionary events. First, spectral information proved to be useful to a large variety of visual tasks in primate vision such that a system of general encoding for luminance and spectral information developed at the early stages of vision. Second, fortuitously, the neurophysiological principles of general encoding shared by both the luminance and spectral systems were such that a rough approximation to the spectral reflectance "profiles" of object surfaces were encoded. Finally, the perception of object surfaces as coloured - as having an "inherent" property, independent of illumination - turned out to be advantageous to the visual system in a number of ways.

If something like the above view is true, then it may well turn out that colour, as phenomenologically presented, rides on the coat tails of a host of other visual processes. It is possible that: (a) the colour system computes the spectral reflectances of objects only "on the fly", using the opponent colour information already encoded for other visual purposes, and hence that; (b) the nature of phenomenal colour space reflects the informational requirements or the spectral "packaging" of these other visual tasks not the co-ordinates of a "colour centre", and; (c)the visual system does not compute, as a matter of course, the spectral reflectance of each object within the visual field. All of these are results that would change both our conception of colour v and of our own colour phenomenology.