Abstract
David Cooper has argued that any acceptable theory of metaphor should account for ‘indeterminacy’: the sense that many metaphors admit of multiple acceptable interpretations, none of which can be uniquely demonstrated to be correct. He further argues that the ‘speaker's meaning’ model of metaphorical content cannot meet this constraint and, thus, should be rejected. In this paper I argue that Cooper's characterization of the proposed constraint is imprecise as stated and give my own characterization of the problem. There is a general tension between the authority granted to first-person ascriptions of intentions and facts concerning the phenomenology of metaphor production, given that it seems to misrepresent the latter to ascribe to the speaker special access to a cognitive content, which their metaphorical utterance then expresses. I argue that one way of resolving this tension is by following Crispin Wright in viewing facts about intention as essentially response dependent.