First, I want to thank Sean for raising an interesting topic. Second, I want to endorse Stuart's observation that Wittgenstein was well aware that he wasn't likely to bring these thoughts to completion. That too should be considered. The most important reason to raise the issues surounding the writing of _On_Certainty_ is the simple virtue of getting the facts right. But as I've mentioned elsewhere, regarding the "cash value" of periodization schemes, what interests me about such distinctions is the use to which we're to put them. What is the relevance of these facts? I would assume that, hagiography and religious metaphors aside, no one here treats Wittgenstein's remarks as some sort of unquestionable authority in any case. So, distinctions between his most carefully considered thoughts and things have might have entertained but reconsidered (or would have reconsidered given the chance) aren't relevant in that sense. Whether what Wittgenstein has to say is true, insightful, or interesting is what matters, whether it is a remark recorded from casual conversation or a remark from his most finished work. In a related topic, we might consider whether some ideas ought to be properly called "Wittgensteinian". Here the dispute would be about classification not authority. I am unaware of any differences between professed Wittgensteinians that turn solely on the acceptance of ideas in OC, but it is at least remotely plausible that such a dispute might arise. Similarly, I could at least imagine a polemic against Wittgenstein that focused on works like OC while giving little attention to PI. Aside from the fact that such a polemic would be rather bizarre, it might be deemed relevant to point out that one is judging a philosopher by work he was unable to complete. So far these possibilities aren't very interesting but this is: Suppose someone were to bring to our attention remarks in OC that clearly conflicted with remarks in PI. (If it is conflicting theses, then the reading would be suspect by my lights but perhaps the conflict is methodological, i.e. he says something in OC that conflicts with his account of philosophical practice in PI.) Now, what might we say? 1. OC is a demonstration of the method and if it seems to conflict with what PI says, then when need to reconsider our reading of the PI's recommendations in light of the practice demonstrated in OC. 2. He was reconsidering his previous ideas about the role of philosophy and this makes a case for a "third Wittgenstein" that is not only a set of additions to but also a departure from the Wittgenstein of the PI. 3. He wasn't at his best and he would have spotted the conflict and revised accordingly had he had the chance, bringing them into conformity with PI. And it is for arguments like (3.) that the circumstances of writing OC would become significant. This is of course completely hypothetical but it is not implausible that such a debate might arise, on a larger or smaller scale. But without a clear case before us, without further context, I can't say that any of those approaches (or others I might not have considered) might not be the most sensible. JPDeMouy ========================================= Need Something? Check here: http://ludwig.squarespace.com/wittrslinks/