- The Names of Comedy by Anne Barton
Oxford, 221 pp, £22.50, August 1990, ISBN 0 19 811793 0
‘The French call it pain, the Germans call it Brot and we call it bread; and we are right, because it is bread.’ So wrote (I have been told though I have not been able to verify the reference) an English theorist of language in the 17th century. The thought is at once robust and lunatic. The writer believes that there are true words and false. His is an extreme case of what Anne Barton calls cratylism.
The word comes from Plato’s dialogue, the Cratylus, which is all about the question of whether language is naturally rooted in reality or is merely arbitrary. Within the dialogue the character called Cratylus maintains the former doctrine, one Hermogenes the latter, while Socrates (here as always the philosophic hero) proves elusive. Professor Barton chooses to confine Plato’s problem to proper nouns. In the original dialogue the thought is allowed to overflow from proper nouns to common nouns and thence to all linguistic structures. Indeed, without this overspilling there would be little philosophic interest in the work. At the level of nouns we are all hermogeneans now; French and English vocabularies stand equal in the sight of God (or Wittgenstein) and anything could, in principle, be called anything.
Everything becomes more difficult, however, when we move to the level of the sentence. It may be nonsense to call a particular word ‘true’, but most people believe that sentences may be true or false. The account given of ‘sentence truth’, however, may itself be either hermogenean or cratylic: roughly speaking, those who hold a ‘coherence theory’ of truth believe that those sentences are true which cohere with the relevant body of discourse, while those who hold a ‘correspondence theory’ believe that ‘the cow is in the meadow’ is true if the cow is in the meadow – that is, if it corresponds with reality (ta onta, in Plato). This time the hermogenean line, which denies any link with a reality outside language, is no longer obviously correct. As long as we stick to nouns hermogeneans can readily concede that, although there is no absolutely right word for pain, bread etc, within a given culture it makes perfectly good sense to say: ‘You’ve used the wrong word; that’s not bread; the right word for that is “cake”.’ Here ‘right’ means ‘cohering with usage’ (‘correct’ rather than ‘true’). But, as we saw, when we move to the sentence, our hermogenean confidence may falter. There is, after all, no immediate contradiction in the proposition that an entire community may produce a great number of observations which agree internally with each other but are all in fact false. We can all understand how this could be the case. But the converse ‘correspondence theory’ has its own difficulties. How is one to check the supposed detailed correspondence of the sentence with the reality it successfully represents? Sentences are grammatical, have subjects and predicates, but in the world there are no subjects and predicates; matter has no grammar.