Keeley has recently argued that the philosophical issue of how to analyse the concept of a sense can usefully be addressed by considering how scientists, and more specifically neuroethologists, classify the senses. After briefly outlining his proposal, which is based on the application of an ordered set of individually necessary and jointly sufficient conditions for modality differentiation, I argue, by way of two complementary counterexamples, that it fails to account fully for the way the senses are in fact individuated in neuroethology and other relevant sciences. I suggest substantial modifications to Keeley’s account which would both solve the problem cases and make better sense of the actual classifications made by scientists. I conclude by noting some limits to the way of classifying the senses that I suggest. I conclude by suggesting that the problem I identify in Keeley’s account has arisen from a confusion that sometimes arises in the philosophical literature between how we individuate the senses and what constitutes a sense.