Metaphor and Thought

Abstract

One of the central problems in philosophy of language is discovering the relation of language to thought. It has been said that language makes our thought accessible to ourselves and others. For instance, Frege argued that signs are necessary for our fleeting impressions and signs we would hardly rise to conceptual thinking.1 The operation of thoughts by signs, however, can be thought of as metaphorical. According to this metaphor, thought is unstable and has a tendency to escape, while the sign is some kind of physical tether which can curb these tendencies. This could mean that any statement, however literal it may seem, functions only because of this underlying operation of thought with words which is metaphorical. It could also be argued, however, that using a sign to a thought is a literal operation and that speaking of fixing the thought with the sign is a dispensable metaphor. Derrida challenges those who think this and asks how we can imagine what it is that we do when we use signs to fix thought without resorting to some kind of metaphor. While it is widely accepted that metaphor is a common feature of natural language, its relation to language and thought is disputed. There are those who believe metaphor is to all linguistic transaction. Derrida, for one, believes that language only works because of the basic mechanism of metaphor: roughly, every time we exchange a word for a concept we are operating metaphorically. A word is not of the same order as the concept that the word represents, and when the word transfers us to the concept with which it is associated, this may be understood as a metaphorical transfer. From this point of view, the phenomenon of metaphor is essential to language and therefore to thought because even literal statements rely on this metaphorical operation. There are those who believe metaphor is not foundational. According to Frege's and Davidson's view, for instance, metaphors are the inessential bows and ribbons. From their perspective, it is possible to make literal utterances that are free of metaphorical contamination and that are not dependent on a covert metaphorical mechanism. On this view, it is possible for all metaphors to be restated in literal terms. Davidson also believes metaphors have no cognitive content, and this would seem to mean that metaphor cannot represent a thought in the same way that a literal statement can. Because of the idea that metaphor has no cognitive content, we can conclude that from Davidson's point of view, metaphor is not fundamental either to representing thought or to language. If one thinks, however, as Derrida does, that language is based on a pervasive metaphorical operation-the basic exchange of a word for a concept-one can conclude that metaphor is a mode of thought and is indispensable to language. It is thus possible to think of metaphor in at least two ways. What we will call the foundational takes metaphor as the basis for language and what we will call the gives metaphor a small place in a theory of meaning for a language. Those who take the containment approach seem to be system builders, while those who take the approach seem not to have systematic reasons for doing so. This is significant, and although it may incidentally say something about the interests of the different theorists, that there can be such an opposition about the same phenomenon seems to indicate something about the nature of metaphor as well. We will see that those who favor the approach for metaphor do not, or maybe cannot, give systematic reasons for giving it this role and as we will show, those advocates of a limited role for metaphor have difficulties, in certain instances, with imposing limitations on metaphor. The project of discovering the relationship of metaphor to thought would be much easier if one did not have to try to define thought or cognitive content. …