Thursday, October 12, 2006

Moore on scepticism

[By Alex Gregory]

I volunteered to do a little guest blogging over here whilst Richard has work to do. I usually post at my blog over at atopian.org, on a combination of ethics and political philosophy, with some (mainly British) current affairs stuff mixed in. I may well do a post or two on some of that philosophy here as well, but for today I wanted to explore some other philosophy type stuff that I don't feel my own blog is the place for.

Here's a pair of apparently sound and valid arguments:

The no-knowledge argument:
1) No matter how small my odds of winning the lottery, I do not know that I will not win.
More generally: 2) I cannot know something if there is any possibility that I could be wrong.
3) I cannot rule out the (unlikely) possibility that I am dreaming, or continually misled by some supernatural entity, etc.

From (2) and (3):
Conclusion: 4) I do not know anything.

That's apparently compelling, and somewhat worrisome. But it gets worse. Here's argument number two:

The no-justification argument:
a) In order for any belief to be justified, the evidence must uniquely favour that belief.
b) All of the evidence available fails to distinguish the belief that reality exists roughly as we experience it from the belief that I am dreaming, or am continually being misled by some supernatural entity, etc.

From (a) and (b):
Conclusion: c) I have no justified beliefs.

Both of those argument are pretty worrying once you really understand the power they hold. There's an easy Moorean response though. Here are two equally valid arguments:

The knowledge argument (I've omitted the original premise (1) for brevity):
1') I know some things.
2') I cannot know something if there is (any) possibility that I could be wrong.

Conclusion: 3') I can rule out the possibility that I am dreaming, or continually misled by some supernatural entity, etc.

The justified argument:
a') I have some justified beliefs.
b') All of the evidence available fails to distinguish the belief that reality exists roughly as we experience it from the belief that I am dreaming, or am continually being misled by some supernatural entity, etc.

From (a') and (b'):
Conclusion: (c') It is not the case that in order for any belief to be justified, the evidence must uniquely favour that belief.

Since these arguments are equally valid, the real question is whether the no-knowledge and no-justification arguments are more sound than the knowledge and justified arguments. Further, since they share a premise each, then the question is which of the remaining premises we have more confidence in. For the first:
3) I cannot rule out the (unlikely) possibility that I am dreaming, or continually misled by some supernatural entity, etc.
or:
1') I know some things.

And for the second:
a) In order for any belief to be justified, the evidence must uniquely favour that belief.
or:
a') I have some justified beliefs.

And, of course, your standard Moorean will claim that our confidence in (1') and (a') is higher than our confidence in (3) and (a). Therefore, so they claim, it's the anti-sceptical arguments we should hold, not the sceptical ones.

Sorry to those familiar with this stuff - I just wanted to make clear both the power of the original arguments, and the simplicity and apparent power of the Moorean reply.

One way in which some people dismiss the Moorean reply is as some kind of conversational inpropriety. There is at least some form of oddness about Moore's response. However, either I'm missing something, or this is a very weak defence of scepticism. The way the worry is being phrased suggests that the only problem for Moore is simply the temporal order that the arguments have come in. Why, if only he'd been born earlier, then we'd be able to accuse the sceptic of that same conversational inpropriety! No, this objection to Moore will not work.

But there's a better reason to be worried about the Moorean reply. In both of the sceptical arguments, the premises provide some degree of an explanation for the conclusion (more than simple entailment). They show us why it is that we don't know anything, or why we don't have justification. In both cases, it's the worry about the possibility of alternatives ways that reality could be that would generate a similar appearance for our point of view.

In contrast, the Moorean anti-sceptical arguments don't really explain their conclusions, they merely state them. In the first, we haven't been given any explanation of why it is that we can rule out these possibilities. In the second, we haven't been given an explanation of why justification doesn't require our being able to distinguish reality from alternatives.

This doesn't make Moore wrong as such, but merely means that we need some further discussion of how exactly the conclusion is true. (Some people, of course, try to do exactly this. When Putnam provides a transcental argument against scepticism, it's intended to provide an explanation (based on semantics) as to why we can rule out certain sceptical possibilities as true.)

4 comments:

  1. That's interesting. Here's a related suggestion: faced with a conflict between two related claims, we should generally seek to resolve it at the more fundamental level. (I'm reminded of rational process requirements. Faced with the "state requirement" that our intentions match our beliefs about what ought to be done, it seems like the latter have priority. Quoting from the linked post: "In response to this conflict, rationality surely requires you to revise your intentions, not your beliefs about what action is best supported by reasons. Rationality requires us to go where our assessment of the evidence takes us, rather than revise our assessments to match the conclusions we’d like to reach. The latter sort of revision amounts to wishful thinking, not reasoning.")

    Applied to the present case: because infallibility is here assumed to be a component of knowledge, and hence more fundamental, we need to resolve the question of infallibility before the knowledge question, rather than vice versa. Moorean responses violate this principle, which is why we find them suspicious.

    ReplyDelete
  2. In contrast, the Moorean anti-sceptical arguments don't really explain their conclusions, they merely state them. In the first, we haven't been given any explanation of why it is that we can rule out these possibilities. In the second, we haven't been given an explanation of why justification doesn't require our being able to distinguish reality from alternatives.


    I don't think this is quite right. Moorean anti-skeptical arguments are intended to show that the skeptical arguments are unjustified, and in a way that is self-defeating if they are taken seriously as arguments for their conclusions. For anyone to be able to take the justification argument seriously, for instance, they would have to be justified in believing (a) and (b). But, by the conclusion of the argument, they cannot be so. Likewise if I do not know anything, I do not know that the argument is sound (or even valid, or even apparently plausible). This immediate self-defeating character combines with a more general self-defeat -- our inability to reconcile the conclusions of either argument with the vast majority of things we already believe (including things like (1') and (a'), which allow for turning the tables on the skeptical arguments by, so to speak, reversing the polarity of the argument).

    Thus, I think the Moorean can simply say that the appearance of explanation on the part of the skeptical argument is simply illusory because if the arguments were right, nothing could be treated as an explanation of anything. Now, this is indeed a far cry from explaining or analyzing justification or knowledge; but the Moorean can respond that this is not necessary to handle the skeptical arguments -- all that's necessary is to show that the arguments that there is nothing to explain or analyze here (i.e., knowledge or justification) can't do what they're supposed to do unless there really is something to explain or analyze here.

    Similarly, the Moorean can concede to Richard that (assuming infallibility is a feature of knowledge, when it is analyzed) infallibility issues are prior to knowledge issues in the analysis itself. But the Moorean is not doing an analysis of knowledge; instead, he is arguing that there is something, which we call knowledge, to be analyzed. And in that sense the recognition that there is knowledge has to be prior to any analysis of knowledge.

    ReplyDelete
  3. The general self-defeat isn't that it has surprising results; a claim that implies that most of what you believe is false isn't a surprising result, it's a result that is indicated to be false by all the evidence you have. In other words, the Moorean won't let the skeptic get away with just looking at the conclusion of the argument in isolation, or perhaps contrasting it with one or two believes; he will, if he is a genuine Moorean, do what Moore did, and insist that the import of all your beliefs. If the skeptic can find a skeptical argument just conflicts with a small set of beliefs, he's safe from this sort of response; but his skepticism is also a minor sort of skepticism about a few beliefs, and not a major sort of skepticism at all.

    On the other ways of formulating skeptical arguments, this may be a situation where you'd just have to go through them argument by argument to see if the . But even then we have a region of skeptical arguments that can be destroyed by Moorean responses, which is significant; and in doing so we seem to have established that skeptical arguments that do certain things -- eliminate all justification, conclude that we know nothing, etc. -- are inadmissible. Given this, the skeptic is retreating and on the defensive. So I think the Moorean can call it a victory, even if there's a skeptical argument somewhere that can put up a better fight for a smaller region. (Given that all Mooreans, going back to Moore himself, are themselves skeptical about some things, this need not be considered a problem until it can be shown to be one.) It may just be details, but it may also be that skepticism can be made to collapse detail by detail; and Moorean responses seem to be successful on those terms.

    ReplyDelete
  4. "This can't be quite right, given that precisely what's in question is whether my beliefs constitute "evidence" at all. What is true is that the result contradicts many of our beliefs. But Moore's argument will have to reject some very plausible beliefs too, since it will have to reject at least one of the (plausible) premises in each of the sceptical arguments."

    There's a cascade effect that the Moorean will say the skeptic can't get around. For to make the argument at all, the skeptic must be conceding some things about our ordinary understanding of 'evidence' -- what counts, what doesn't, and so forth. So I don't think the skeptic can simply claim that the dispute is about what counts as evidence, and that therefore the Moorean is begging the question. The Moorean's point, in a sense, is that this is precisely not what the dispute is about: either the skeptic understands 'evidence', 'justification', etc. in much the way we ordinarily do, in which case his arguments are incoherent; or he doesn't, in which case they are irrelevant to what we ordinarily take to be evidence, justification, and so forth.

    I'm not sure that the premises of the skeptical arguments are so plausible, but even conceding that, there is no problem for the Moorean: the Moorean is rejecting one (or both) claims; the skeptic is rejecting an immense number of claims. Rejecting a few claims because they lead to a conclusion that conflicts with our sense of the plausibility of a vast array of other claims is just not parallel to rejecting an immense number of claims just on the basis of your sense of the plausibility of two claims.

    I think we need to distinguish two things: the fact that the skeptic is wrong and the account of how the beliefs in particular are justified, known, etc. The Moorean response certainly doesn't have the latter; but it doesn't really claim to. I suspect that this is because it's a sort of reductio; and reductios are always indirect proofs. If I prove by reductio that 2 + 2 do not equal 5, I still haven't provided an account of how 2 + 2 = 4. But this doesn't affect the quality of my reductio; it just points out that my reductio is not an account of numbers. And I think the Moorean is in much the same state: his reductio is not an account of justification (or knowledge, or any other such thing), and can't be treated as such. But in a sense that's not a problem, because the Moorean can just say he's clearing the ground for beginning the inquiry into such an account, by showing that there is indeed something to have an account of.

    ReplyDelete

Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.