From K. Vossler, Positivismus und Idealisms in der Sprachwissenschaft (1904). Quoted by Polanyi in Personal Knowledge (p. 102):
The true artists of speech remain always conscious of the metaphorical character of language. They go on correcting and supplementing one metaphor by another, allowing their words to contradict each other and attending only to the unity and certainty of their thought.
I think this is right, provided one is clear that good artistic writers do not actually contradict themselves. A contradiction that exists at the level of words is merely an apparent contradiction, not a real one. Real contradictions exist at the level of thought and destroy its “unity and certainty”.
It can be an enlightening exercise to reflect on the pervasiveness of metaphors in language. Try expunging all metaphors and you’ll have a tough time saying much of real significance.
Does he deal at all with the purpose of metaphores? It has always seemed to me that metaphores express a type of truth that traditional logic cannot touch. Exactly what type of truth it expresses is a question I have often wondered about.
Polanyi is extremely interested in inarticulate knowledge. He views much of what we know to be inarticulable and hence, words become the tools that convey knowing (contra to much of epistemology, where the words ARE the knowledge).
So in one sense, Polanyi views all language as metaphorical in that it is always trying to grip on to what we truly know.
If I understand Polanyi right, he thinks that an inarticulate or ‘tacit’ background is a necessary precondition for articulate thought.
What I find surprising is the extent to which Polanyi’s views cohere with two other profound thinkers: Bernard Lonergan and Charles Peirce. Lonergan speaks of how any articulate ‘insight’ into reality is an abstraction from a field of experience and thereby always leaves behind an inarticulate ’empirical residue’. Peirce makes basically the same point in his discussions of the relations between abductive inference and perception. The very possibility of abduction presupposes both a body of already possessed background knowledge and a real continuity or ‘synechism’ in our field of experience that our articulate conceptions cannot fully express.
What I find even more surprising is that Polanyi, Lonergan, and Peirce seem to have arrived at their respective views independently of each other. That would seem to be a confirmation that they’re on the right track.