Hierarchical and parallel mechanisms in the organization of visual cortex.

We argue that it seems fruitful to regard the retino-geniculate-cortical pathway, and perhaps the visual pathways in general, as comprising distinct neuronal channels which begin with the major groupings of ganglion cells, and subserve distinct functions within the overall operation of the visual system. One problem for future work is to determine the extent and, equally importantly, the limitations of the idea of independently functioning neuronal channels operating within the visual system. Some evidence of those limitations is already available. Kulikowski and Tolhurst have provided evidence suggesting that pattern detection is mediated by the X-like system at high spatial frequencies and by the Y-like system at low frequencies, but that at intermediate frequencies, both systems are likely to contribute to this function. Again, there is already physiological and psychophysical evidence of inhibitory interaction between X- and Y-cell systems, which may contribute to their functioning. That is, although there is little evidence of excitatory interaction between W-, X- and Y-cell systems, at least up to the first cortical synapse, the functioning of, say, the X-cell system may depend on the inhibitory influences impinging on it from Y-cell activity. Further, it may prove to be the case that one cell 'system' may be involved in several distinct functions and considerable work may be required to establish whether or not these functions can be considered constituent parts of an overall function, such as 'ambient' or 'foveal' vision. In the following section we suggest a classification and terminology for visual neurones which may provide a framework for future work on these lines.