A Priori Bootstrapping

Many philosophers have thought that the following proposition wasa priori:

  • If I don’t have any positive evidence that I am being deceived by any Cartesian evil demons, I am not being deceived by any such demon.

But how could this be a priori? Here’s a suggestion, which I’ll call the a priori bootstrapping argument.

Suppose that the following principle of rational belief is true: If one has a conscious experience as of its being the case that p, and no special strong reasons for regarding one’s current experience as unreliable with respect to propositions like the proposition that p, it is rational to form the belief that p. For short, let say that according to this principle, it is rational to “take experience at face value, in the absence of defeaters.”

Let us also suppose that there are some other similar principles of rational belief: a principle according to which it is rational to form introspective beliefs in which we self-ascribe our current conscious experiences; a principle according to which it is rational to form beliefs in the obvious logical consequences of propositions that we already rationally believe; and a principle according to which it is rational to form beliefs by means of inference to the best explanation.

Now imagine a Platonic soul waiting to beam down from the intelligible world into the sensible world. This Platonic soul knows all these principles of rational belief. So this soul can predict that so long as her experiences contain no defeaters, it will be rational for her to take her experience at face value. So she knows that so long as her experience contains no such defeaters, it will be rational for her to come to believe what Roger White, in his forthcoming paper “Problems for Dogmatism”, calls a “Track Record Proposition”, i.e. some proposition of the form:

  • I have an experience as of its being the case that p1, and p1; I have an experience as of its being the case that p2, and p2; … and I have an experience as of its being the case that pn, and pn.

If it were rational to believe any such “Track Record Proposition”, it would surely also be rational to believe the following proposition (“No Errors”):

  • The contents of my experiences have been true.

And if it were rational to believe that, surely it would also be rational to believe what the following proposition (“Reliability”) which seems to be the best and simplest explanation of “No Errors”:

  • My experiences are generally reliable guides to the truth.

Moreover, we may presumably assume the Platonic soul knows that if it is rational for her to believe “Reliability”, it is also rational for her to believe that she is not being deceived by a Cartesian demon.

So the Platonic soul already knows – even before she’s beamed down to the sensible world – that if her experiences don’t contain any defeaters, it will be rational for her to believe that she isn’t being deceived by a demon. But then surely it will already be rational for her to believe that if her experiences don’t contain any such defeaters, she is not being deceived by a demon. (This would follow by means of a principle like Roger White’s “meta-justification principle” – i.e. roughly, the principle that evidence that one will acquire evidence that p is already evidence that p.) If it is already rational for the Platonic soul to believe this, even before she beams down into the sensible world, then it seems that her justification for believing this is a priori.


Comments

A Priori Bootstrapping — 11 Comments

  1. “if her experiences don’t contain any defeaters, it will be rational for her to believe that she isn’t being deceived by a demon. But then surely it will already be rational for her to believe that if her experiences don’t contain any such defeaters, she is not being deceived by a demon.”

    There’s a shift in the scope of the rationality operator here that might be worth thinking about. We go from:
    If there are no defeaters then it’s rational to believe p
    to:
    It’s rational to believe that if there are no defeaters then p.

    This sort of shift might be thought to be a mistake for various reasons. For instance, one could argue that consideration of the second – conditional – proposition raises to salience situations that cannot be ruled out by the subject (e.g. the possibility that she has seriously misleading evidence about p), so it’s irrational for her to believe that conditional proposition. Yet consideration of the proposition p itself doesn’t raise such things to salience, so it might be rational to believe p in the absence of defeaters.

  2. Strictly speaking, you’re misrepresenting my argument here. My argument doesn’t base the conclusion “It’s rational for you to believe that if there are no defeaters then p” on the premiss “If there are no defeaters it’s rational for you to believe p”, but on the stronger premiss “You know that if there are no defeaters then it’s rational for you to believe p”.

    The argument that you criticize would certainly be unacceptable. Not everyone who is sensitive to the presence or absence of defeaters to the content of her experience even has concept of a defeater, or of experience. But the Platonic soul in my little fantasy is already supposed to know all the facts about what the fundamental principles of rational belief are, etc. So although I concede that my argument is distinctly sketchy and needs to be sharpened up in many ways, I don’t think that it suffers from the defect that you criticize.

  3. I have a worry about how the defeater is imagined to work. What is it to know that a defeater condition is satisfied? Some theories of defeasible reasoning insist on a weaker condition, namely that the agent not know that a known defeater holds. This would give you a non-monotonic defeater, since the agent could learn something later to trigger the condition.

    Could you say more about how you are thinking about defeaters working?

  4. Wedgwood’s argument relies on the supposition:

    (A) If one has a conscious experience as of its being the case that p, and no special strong reasons for regarding one’s current experience as unreliable with respect to propositions like the proposition that p, it is rational to form the belief that p.

    Now, why do so many epistemologists today accept (A)? If we accept (A), then why should we not also accept the following supposition:

    (B) If one has a conscious experience as of its being the case that p, and no special strong reasons for regarding one’s current experience as unreliable with respect to propositions like the proposition that p, then that very same conscious experience provides compelling a posteriori evidence that A is true (for that very same value of p).

    And if (B) is true, then can we still argue for the a priority of: “If I don’t have any positive evidence that I am being deceived by any Cartesian evil demons, I am not being deceived by any such demon”…?

    Unfortunately, I don’t know whether (B) is true. But I also don’t know whether (A) is true. What is there to recommend (A) over (B)?

  5. I’ll have to think about Ram Neta’s point a bit more, but on the face of it (B) seems to me much less intuitively plausible than (A). My experience (at least in the absence of defeaters) seems to give me evidence for the truth of propositions like p itself, but not (except in some very indirect or circuitous way perhaps) for normative propositions about what it is rational to believe like (A). Anyway, even if (B) is true, it would surely be possible for someone to have both a priori justification and empirical justification for certain propositions. So I don’t immediately see how (B) would prevent the Platonic soul from being able to achieve a priori justification for believing that if her experience contains no defeaters, then she is not being deceived by a demon.

    As for Gregory Wheeler’s question, I was hoping that an intuitive grasp of defeaters would be enough to make the argument reasonably plausible. E.g. suppose that you have the experiences that the character Neo has in the first Matrix movie: I take it that these experiences constitute strong special evidence that he is being radically deceived, and so also constitute defeaters with respect to the ordinary propositions that he would normally come to believe by taking his experience at face value.

  6. Can I ask what “one has a conscious experience as of its being the case that p” means, exactly? In particular, am I now having a conscious experience as of being systematically deceived by a Cartesian evil demon? Because, after all, my experience is exactly like the experience I would be having if I were being so deceived. I have no special defeaters for this hypothesis (it seems to me), so is believing that I am now being systematically deceived a rational belief? I don’t think that’s where we were supposed to end up.

    Maybe it means “one is inclined to believe p because of one’s present conscious experience”. But then I think the principle is questionable. For I might have been wrongly trained or brainwashed or whatever to believe some absurd p whenever I am confronted with some particular kind of experience. That doesn’t seem like rational belief.

  7. Heath — I’m assuming that every normal conscious sensory experience has a propositional content. (Obviously this is a controversial assumption, but the basic idea behind my “a priori bootstrapping” argument could probably be adapted to other conceptions of experience as well. It may not be quite as controversial as its sounds because I mean to take no stand for present purposes about whether the content of sensory experience is conceptual content, as McDowell thinks, or nonconceptual content, as Peacocke thinks.) Anyway, by talking about a “conscious experience as of its being the case that p“, I meant a conscious experience the propositional content of which immediately logically implies that p.

  8. Ralph, let me try to press my worry. Suppose that all my life I have been taught that dingoes are flamingoes and vice versa; I mix these two animals up systematically. I see a flamingo, up close, in good light, and immediately form the belief, “There’s a dingo.” Have I had an experience as of it’s being the case that there’s a dingo, or as of it’s being the case that there’s a flamingo?

    Option 1, my experience is as of a dingo. I have no reason to think there is anything unveridical about this experience. But is it rational for me to believe that there’s a dingo there? I’m unsure. And surely we would not want to say a priori that, if there’s no positive evidence that I’m not seeing a dingo in such a case, then I’m seeing a dingo. Because I’m not.

    Option 2, my experience is as of a flamingo. Then my perceptual beliefs need not match the propositional contents of my experiences. In that case, I could accept (even know!) “Reliability”, that my experiences are reliable guides to truth, but it would not follow that my perceptual beliefs were reliable guides to truth. The demon’s deception could intervene between experience and belief.

    Last comment. I don’t see how some non-conceptual content can logically imply a proposition. If X implies a proposition, I would have thought that, ipso facto, X was conceptual.

  9. Heath, in the case you describe, doesn’t Ralph just need that it’s a priori justifiable that:
    if there’s no positive evidence that you’re not seeing a dingo, then _it’s rational for you to believe_ there’s a dingo?
    (And not, that is, that it’s a priori justifiable that if you’re having that experience then there _is_ a dingo, which surely everyone will reject.)

  10. Well, except that the conclusion the post defends is that it is apriori justifiable that “If I don’t have any positive evidence that I am being deceived by any Cartesian evil demons, I am not being deceived by any such demon.” Maybe biting the bullet is the right answer: it is a priori justifiable [for me to believe] that (if I am having this experience, with no defeaters, I am seeing a dingo).

    If this is the right answer, then I think the whole argument would be better put in terms of the contents of perceptual beliefs, which are uncontroversial, rather than the contents of conscious experiences, which are controversial.

    If this is not the right answer, then I think it tells against the principle of rational belief the argument begins with, namely that it is rational to take experience at face value, in the absence of defeaters.

  11. Yes, Heath, that’s exactly right — thanks! I am in fact willing to bite the “bullet” that you describe: of course my having the experience in question is necessary if it is to be possible for me to refer to it as “this experience”, but it doesn’t follow that this experience is what justifies me in believing the conditional, which I am indeed inclined to regard as a priori.

    You’re also right that I could rephrase the argument in terms of perceptual beliefs, which might well be preferable. (Then the argument would start from the premiss that if I have a perceptual belief that p, and no defeaters, then it is rational for me to believe that p; I would also need to claim that — not necessarily always, but often enough — one is in a position to arrive at rational introspective beliefs about what perceptual beliefs one has; and that it is also rational to assume that if one’s perceptual beliefs are generally reliable one is not being deceived by a demon or the like. With those adjustments, pretty much the same argument seems to go through.)

Leave a Reply

Your email address will not be published. Required fields are marked *