Moore appears to have been the leading spirit in the rebellion against idealism with which he and Russell inaugurated analytic philosophy. Moore and Russell developed quite different ideas about what kind of analysis was desirable, and it is Moore's which was arguably the more influential, leading ultimately to the dominant preoccupation of twentieth century analytic philosophy with various kinds of analysis of ordinary language.

Moore's early work was partly the critique of idealism and scepticism, and partly more positive work on ethics and on the theory of perception. In the paper "A Defence of Common Sense", published in 1925 Moore is no longer concerned with detailed critiques of idealism, but is now primarily presenting an alternative philosophical point of view respecting ordinary language and aknowledging common sense and its claims to certain knowledge.

The principle assertions of this paper are:

• that many ordinary beliefs about the existence of objects and other people are true and certain knowledge
• that idealism is false
• that scepticsm is incoherent
• that ordinary concepts can usefully be subjected to philosophical analaysis which will be informative even to people who fully understand their meaning

What I had not myself noted, but is here pointed out by Stroll, is the insistence by Moore that it is unecessary for there to be any clarification of the meaning of his commonsense propositions. Later in the century, not only ordinary but also philosophical discourse is treated in the same way (the example of this which struck my most is the concepts a priori and a posteriori in Kripke's "Naming and Necessity"). I can accept this in relation to Moore's examples while deploring it in Kripke's. I guess I didn't note it because when confined to ordinary discourse it did not seem to me controversial. The connection with the shift in expectations of philosophical discourse which seemed to take place in the second half of the century perhaps didn't then occur to me.

This point about meanings must be taken into account in relation to Moore's discussion of "analysis" later in the paper. Moore holds that, though the sample propositions are true and can be known with certainty to be true, and that we do ordinarily fully understand the meaning of the propositions (without the help of philosophers), there remains uncertainty about the correct way in which the propositions should be analysed. He does not state that it is the business of philosophy exclusively to indulge in such analyses, but its clear that he does think it a sensible thing for a philosopher to do, and in considering the ways in which they might be analysed he is primarily seeking a way to analyse them in terms of sense data.

I found Stroll's account here helpful in crystallising my attitude towards the later philosophy of Wittgenstein so I will try here a condensed extract with my own reactions.

First note Stroll's most concise summary.

The new philosophy thus turns on three features:

• an appeal to everyday language,
• an appeal to a gamut of cases and the contexts in which they occur,
and
• an appeal to human practices.