Language based criticisms of the intelligibility of the programme of neuropsychology have made use of the principle that words the meaning of which is established in the context of descriptions of aspects of whole persons cannot be used in that sense to ascribe properties to parts of human bodies. In particular neither human brains nor their parts think, are conscious, imagine, suffer and so on. Recently, Bennett and Hacker have presented the error as a mereological fallacy, because brains are parts (...) of persons. However, while brains are parts of human bodies it is not clear that they are parts of persons. I restyle the argument in terms of fields of family resemblances, in such a way that it makes sense to describe the hippocampus as an organ for remembering, but does not support the claim that neuroscience is core psychology. Such fields are networks of meanings linked by two principles. (1) Taxonomies of relevant body parts are determined by the psychological role they play in everyday human life. (2) Many body parts are also identified by the role they play as tools in human activities including psychological tasks. Arguments are developed to show that objections to the idea that brains and their constituent organs are tools are misplaced. Hybrid psychologies are possible. (shrink)