ABSTRACT
A philosopher’s claims about ‘what we say’ are based on personal reflection. He considers relevant examples and concludes that he would apply a given word to such and such cases and refuse to apply it to others. Discussion with others may lead him to change his mind: they may draw his attention to other examples, or to features of examples, that he had overlooked. But the appeal will still be to his personal reflection: he must be able to see that he would or would not be prepared to apply the word, in ordinary conversation, to those examples. How does he acquire this knowledge and how can he generalize from what he would say to claims about what we – the languagecommunity – would say?1
