As we saw in the preceding chapter Wittgenstein did not, in his later years, abandon phenomenalism. What he abandoned was only the Tractatus view that the propositions of ordinary language are truth-functionally constructed from more elementary propositions. But this change required a modification of Wittgenstein's ideas about philosophical method. While he continued to think that philosophical problems arise from misunderstandings of the ‘logic’ of our language, he could no longer think that such misunderstandings can be corrected by logical analysis. The idea of analysis is the idea that a philosopher can dissect our ordinary propositions into those sense-datum propositions from which they are truth-functionally constructed and can thereby show, for example, that Moore was wrong to oppose Berkeley's account of propositions about material objects. But with analysis now out of the question, the discovery of a new method became an urgent matter for Wittgenstein.

The discovery was not long in coming. Drury reports that Wittgenstein said to him in 1930: “I know that my method is right.” And Moore, in his account of Wittgenstein's lectures of 1930–1933, reports him as saying that his particular results did not matter as much as the fact that “a method had been found.” What was this new method?