Lakoff, George. 1987. The death of dead metaphor. Journal Article. pp. 143-147
Abstract
The term 'dead metaphor' is a holdover from a traditional folk theory of language that has turned out not to be workable. According to the old theory, the locus of metaphor was language not thought, ordinary everyday language was "literal," and only novel poetic or rhetorical expressions were candidates for being metaphors. Metaphorical "life" was seen in poetic novelty alone, and mundane unpoetic language supported no metaphorical life. A "dead" metaphor was defined relative to that theory as a linguistic expression that had once been novel and poetic, but had since become part of mundane conventional language, the cemetery of creative thought. As we saw in this column (Lakoff, 1986), the term literal is a cluster concept, defined relative to the old theory. When empirical research showed the old theory to be wrong, the term 'literal' split. The four senses making up the cluster had to be distinguished, and one could no longer use 'literal' in its old overly simplistic sense in discussions of metaphor. Like 'literal', 'dead metaphor' has been defined in a theory-dependent way. As that theory dissolves under the scrutiny of empirical research, the meaning of "dead metaphor" cannot remain constant. What were called dead metaphors in the old theory have turned out to be a host of quite disparate phenomena, including those metaphors that are most alive - the ones that we use constantly in everyday thought. If one wants to keep the term 'dead metaphor', it will have to come to mean something very different in contemporary theories.(George Lakoff)