Naming of Neurones; pp. 185–200

Abstract
Many schemes of ganglion cell classification seek to classify the cells by some particular characteristic, such as the time course of the cells'' physiological responses or their dendritic morphology. It is argued that such schemes are based on the Aristotelian concept of essences and raise the same difficulties as have arisen with the essentialist approach to animal taxonomy. A better approach to the classification of neurons, it is proposed, is to base the classification on as many features of the cells as possible, and to regard the classification as an hypothesis, subject to testing and modification by experience, about the functional niches occupied by the cell types distinguished.