| Summary: | Virtually all currently discussed accounts advert to a shift or replacement of a property or properties in describing what happens to the ordinary words in metaphors. And the mechanism of this shift tends to involve an overt or sometimes hidden appeal to similarity, or to some notion that is essentially connected to it. In the first part of the paper, I argue that this route is a dead end, and in the second part I offer my own preferred alternative. That alternative is not argued for, or developed in detail – that is done in my book Objects of Metaphor – but my main aim in the paper is simply showing how radically it differs from the property<br />route.
|
|---|