Why many concepts are metaphorical.

The study of concepts in psychology has historically had little to say about the possible influence of metaphor on mental representation. This silence is not terribly surprising given the long-standing view that metaphor is an ornamental, even deviant, form of language bearing no relation to how people actually conceptualize of objects, individuals, events, and abstract ideas. But in the last 20 years there has been an explosion of research in cognitive science on metaphor and related topics. A significant part of this research, most notably studies from cognitive linguistics and psycholinguistics, suggests that metaphor is not merely a figure of speech, but is a specific mental mapping that influences a good deal of how people think, reason, and imagine in everyday life (Gibbs, 1994; Johnson, 1987, 1993; Lakoff and Johnson, 1980; Lakoff, 1987; Lakoff and Turner, 1989; Sweetser, 1990; Turner, 1991). One of the most provocative claims of this work is that many concepts, especially abstract ones, are structured and mentally represented in terms of metaphor. In his paper "On metaphoric representation," Gregory Murphy argues that metaphor might only have a limited role in the acquisition and representation of many concepts. He suggests that a theory of structural similarity provides a more satisfactory account of the linguistic and psychological data, thus greatly reducing, or eliminating, the need for metaphorical concepts. Murphy deserves credit for raising important questions about the possible role of metaphor in a psychological theory of conceptual structure. However, Murphy misunderstands aspects of the evidence for metaphorical concepts, accepts unquestionably certain assumptions about what a psychological theory of concepts must be like, and ignores the role of human embodiment in theories of metaphor, all leading him to underestimate the significance of cognitive linguistic and psycholinguistic work on metaphorical thought. Space limitations prevent me from commenting in this reply on all of Murphy's arguments, but I will address some of his main points against metaphoric representations.