Wittgenstein on Metaphor

In this work, I examine Wittgenstein’s possible contributions to an elucidation of the grammatical status of certain metaphors – often found in theoretical and speculative texts – which resist an approach based on the assumption of a clear split between the fields of pragmatics and semantics. I take as examples of works that depart from this assumption Elizabeth Camp’s Contextualism, Metaphor and What is said (which explores the lines suggested by Paul Grice), and John Searle’s Expression and Meaning. Both rely on a distinction between speaker’s meaning (utterance meaning) and sentence meaning to explain the nature of metaphor. They assume that the very metaphorical operation involves meaning something instead of saying something. But it is anything but obvious that, when we consider, e.g., the following metaphor of Philosophical Investigations: “A picture held us prisoners” (§115), we can assume that we are facing a non-descriptive use of language. I argue that Wittgenstein himself can provide us with tools to examine a possible descriptive function of this kind of sentence when he develops his grammatical research methods which: a) are not focused on the linguistic dimension of a sentence but on the linguistic dimension of discourse; b) bring up the issue of language learning; c) lead us to ask if certain metaphors could not work as modifiers of convictions, i.e., if they could not act directly on what Wittgenstein once called Weltbild