A defense of Davidson's theory of metaphor

Date of Award




Degree Name

Doctor of Philosophy (Ph.D.)



First Committee Member

Leonard Carrier - Committee Chair


Metaphorical meanings are best understood as the effects of the speakers who utter them. As such, metaphors are not semantic, but pragmatic, causal, extra-linguistic phenomena of language. Among the pragmatic theories of metaphor that have been proposed, Davidson's theory most clearly distinguishes metaphors as the causal effects of metaphorical utterances. Other pragmatic theories (theories proposed by Robert Fogelin, Susan Haack, John Searle) falsely rely on semantic rules of one kind or another to explain the effects of metaphors. But if metaphorical meanings are the causal effects of metaphorical utterances, then no appeal to semantic rules will help to explain these effects. According to Davidson's theory of metaphor, dead metaphors are not metaphors at all, for their former metaphorical meanings (their effects), through repeated use, have become secondary literal meanings of these expressions. According to Davidson, therefore, dead metaphors have no effects. While this thesis that dead metaphors have secondary literal meanings is true, nevertheless Davidson overlooks the fact that a great many dead metaphors do retain certain metaphorical effects. This fact can be accommodated within Davidson's theory of metaphor by allowing that these metaphorical effects belong to the extra-linguistic intentions that a speaker of a dead metaphor intends her utterance to convey; extra-linguistic intentions which accompany the secondary literal meanings (the linguistic intentions) that it is the speaker's primary purpose to convey. Thus, Davidson's theory of metaphor is compatible with the notion that some, indeed, a great many, dead metaphors have metaphorical effects.


Language, Linguistics; Philosophy

Link to Full Text


Link to Full Text