There are serious problems evident in both of these citations. Hobbes accuses metaphor of professing deceit, and he presumably has in mind here that thought, shared by all too many philosophers, that metaphors typically say the false: Juliet is most certainly not the sun; Tolstoy is long since grown out of infancy. In picking and choosing examples, one can make it seem as if metaphors are literal falsehoods, but a more cautious view is that, taken literally, they do not so much as possess sense. And, even if there weren't many other reasons to think so, this in itself should suggest that metaphors do not profess deceit, and therefore that they must be understood in some fundamentally different way. The passage from Nietzsche is typically opaque, but philosophers tend to understand him as saying that ordinary assertions —‘ Truths ’ , as he has it — involve dead metaphors. However, if this is all that Nietzsche is saying, it is not a very profound defence of metaphor. As noted earlier, those who fi nd the claims made on behalf of metaphor overblown can perfectly well concede that metaphorical processes have left some mark on expressions we currently use. But this is a far cry from saying that metaphor itself is in any sense primary. Cooper 1986: 258 does cite a source (Bolinger 1980: 143 – 5) who claims that ‘ every thought and utterance we make is