This paper focuses on the claim by Schneider and Butcher (2000) that it makes little sense to criticise the use of performance-enhancing drugs as ?dehumanising? (as, for example, Hoberman does (1992)) because we are unable to give a satisfactory account of what it is to be human. Schneider and Butcher (2000, 196) put this as follows: ?The dehumanisation argument is interesting but incomplete. It is incomplete because we do not have an agreed-upon conception of what it is to be human. (...) Without this it is difficult to see why some practices should count as dehumanising.? The paper begins by considering J.L. Austin's (1962) treatment of the word ?real?. By transposing ideas from Austin to the terms ?dehumanise? and ?human? I argue that (a) In the pair ?dehumanise? and ?human?, the term ?dehumanise? is dominant; (b) We cannot understand ?dehumanise? and ?human? independently of either the context of their use or the contrast that is drawn in their use; (c) Either one of these is sufficient to understand the terms; (d) ?Dehumanise?, ?human? and their cognates are not univocal; we can have no recourse to exceptionless accounts of the meaning of such terms. The importance of context is developed further by consideration of an example from the work of Charles Travis (2005), and the issue of exceptionless accounts of the meaning of words is addressed through an application of Gordon Baker's (2004) characterisation of Wittgenstein's uses of the term ?metaphysical? to Miah's (2004) treatment of human-ness. I argue that Miah's conception of human-ness exhibits all the forms of metaphysical use of terms (in this case the term ?human?) outlined by Baker (2004). The article attempts to clarify some objections to the use of performance-enhancing drugs and the prospect of genetic modification of athletes by sketching an overview of possible concrete uses of ?dehumanise?. The focus of the paper, however, is on ?making sense of what we (are inclined to) say ? [rather than] making explicit what underlies what we say? (McFee, 1993/4, 115). (shrink)