The K-B Implication and the Value of Knowledge

Suppose you are tempted to agree with Lewis et. al. that knowledge doesn’t imply belief. Lewis’s example is of a timid student, uncertain of a given claim who will only hesitatingly answer, but answers correctly nonetheless. Not the most persuasive example: belief is compatible with some level of uncertainty, and it isn’t clear that the student knows that, e.g., Lincoln was assassinated in 1865.

Here’s a more interesting worry, however. Suppose Lewis is adamant about the example or some suitable refinement of it: the student knows the answer, and thus has propositional knowledge of some sort; and the student doesn’t believe the claim in question because he takes himself to be guessing (again, not a compelling inference, but I’ll let it go). Given such a position, true belief has something going for it in terms of a connection to action that knowledge doesn’t. To use Socrates’s example, you’re better off hiring a guide with true opinion about how to get to Larissa than hiring one with knowledge when that knowledge isn’t accompanied by belief. For, as in Lewis’s student example, we can expect hesitation and indecision from those who know but lack belief, and we have no such similar expectation concerning those with true belief. In fact, in the student case, if Lewis is right, the student might know and simply refuse to answer at all because of the lack of belief. The analogue in Socrates’s example would be a knowledgeable guide who won’t go any further at some point on the journey because he is uncertain which way to go. So, if you agree with Lewis, it looks like you should downplay the idea that knowledge is better than true opinion. Meno wondered why we prize knowledge over true opinion. If Lewis is right, such puzzlement is a confusion–the only general truth in the area is that when practical affairs are in view, true belief is preferable to knowledge. Sometimes knowledge is just as good as true belief, but in other cases it is not as good.

That implication strikes me as a mark against Lewis’s claim, though of course not a conclusive one.


The K-B Implication and the Value of Knowledge — 8 Comments

  1. Jon,

    I think that the right way to think about the Meno cases, at least if one agrees with Lewis, is by comparing cases of knowledge to cases of mere true belief when these are the respective explanations of the same action. In other words, the question becomes “Is knowledge on the basis of which one is ready to act more valuable than a mere true belief on the basis of which one is ready to act?”. I take it that if these are the comparison classes, then one can still make the same, or similar, arguments that knowledge is more valuable than mere true belief, while eliminating cases like that of the timid student from affecting the answer.

    For what it’s worth, I think that the right analysis of Lewis’s case (at least if you are among those who reject the KK principle) is the following. The timid student knows AND believes that the answer is what it is, but does not believe THAT he believes it or that he knows it. If one acts, in these kinds of cases, on the basis of one’s first-order beliefs (where the order of the belief is relative to the action), then one will answer the question correctly (should one choose to answer it at all) and at the same time deny that one believes the answer, since both of these actions involve beliefs directly about the matter appropriate to the two actions (answering the question and reporting one’s beliefs respectively).

  2. I think you have a good point here.

    If I recall correctly, Lewis’ rejection of the K-B implication follows from his account of knowledge. That account of knowledge goes something like this:

    S knows that p if and only if S’s evidence eliminates all possibilities incompatible with p (except for those we are properly ignoring.)

    Now, according to this account of knowledge, someone could know that p if and only if they were firmly and decisively convinced of not-p. Imagine a creationist who studies evolutionary biology to a high level. This creationist has evidence which eliminates all possibilities incompatible with evolution being true. So by Lewis’ lights, the creationist knows that evolution is true, even though he firmly (and irrationally) rejects it.

  3. On the other hand, true belief does not in general have something going for it in terms of action that knowledge does not. Suppose your guide believes truly that the old bridge is safe to cross, but admits he doesn’t know. If he knew, I would cross; as it is I wouldn’t. It always looks to me like true belief is as good as knowledge from the third person: not from first or second person.

  4. An aside, regarding the case itself – here is, perhaps, a better case where the subject doesn’t believe but still could be said to know (one I initially heard from Keith DeRose): Imagine the daughter of a German man, who himself lived through WWII but has since died. Some evidence comes to light which overwhelmingly supports the fact that her father was an active Nazi before and during the war, but after the war was able to hide this about his past. The daughter, though faced with this evidence, cannot bring herself to believe this about her father; yet she knows, “deep down,” given the compelling evidence…

    Re: what Omar says above. The explanatory gloss given of Lewis’s account isn’t quite right – “Now, according to this account of knowledge, someone could know that p if and only if they were firmly and decisively convinced of not-p.” But one might have evidence which eliminates all (properly unignored) incompatible possibilities but nonetheless not be “firmly and decisively convinced of” it. According to the gloss the creationist *doesn’t* know, though according to Lewis’s account it seems he does… similarly, I think, for the Nazi’s daughter example. And it’s easy to drum up problematic cases in part because Lewis doesn’t tell us much about what it is to “have evidence” (can one have evidence tacitly or only occurrently? Can I have evidence but nonculpably not heed it? Can I have evidence but not recognize all that it supports? etc.)

  5. If we’re going to suggest that the daughter knows ‘deep down” that her father was an active Nazi, on what grounds do we resist the claim that ‘deep down’ she believes it too? Of course, she won’t assent to certain things we’d usually expect her to assent to if she had such a belief, and she may even go so far as to adamantly deny them. But that doesn’t matter, since we’re talking about ‘deep down’ belief, rather than occurrent belief.

    Here’s perhaps a better way of putting the worry. What are the ground-rules for discussing such cases when we allow attributions of ‘deep down’ beliefs and knowledge? What’s supposed to block the kind of appeal to a species of ‘deep down’ belief that accompanies ‘deep down’ knowledge that I gestured at in the previous paragraph? How can we tell when we’ve got a genuine example of ‘deep down’ knowledge without ‘deep down’ belief on our hands? (As I’ve suggested, it doesn’t seem enough that she has, or claims to have, a contrary occurrent belief).

  6. Good point, Aidan. I wasn’t meaning to lean heavily on the term “deep down” as a way of invoking some kind of nether realm in which knowledge could be had but not belief; it was more of heuristic metaphor…

    For the person who thinks that the Nazi daughter example is a case of knowing without believing, it’d probably be better not to resort to such idioms, and rather maintain that she just knows, though doesn’t yet believe it. I’d guess this is related to some cases of being in denial regarding some truth that has been encountered…

  7. Thanks to all for the comments. Some quick replies:

    Pavel, I agree with your assessment of the case. I also think that the additional language you use (“on which one is ready to act”) doesn’t eliminate the disconfirming power of the original issue.

    Omar, I wrote a post about this issue a couple weeks ago, though I think your “iff” connective in the last paragraph needs to be replaced with “even if”.

    Matt and Aidan, my initial reaction to the case was just what Aidan wrote. It looks to me like a case where the person resists the evidence, and thus neither knows nor believes.

    Mike, it’s hard to get the examples right when checking the value of true belief vs. that of knowledge. In your example, we need to assume that the guide has a true belief, but doesn’t know. If he says, “I have a true belief, but do not know,” the example gets sidetracked by the quasi-paradoxical character of the assertion. In general, imagining situations in which assertions are used to generate the epistemic status in question will have this problem. But if you assume that the guide has a true belief, but lacks knowledge, there’s no good reason not to cross. I suspect your hesitation arises from considering a different case: one where the guide recommends crossing, but admits he doesn’t know whether the bridge is safe. But that case compares knowledge with belief, not knowledge with true belief.

  8. Yes, that was meant to read “even if”, not “if and only if” in my last paragraph. Thank you for pointing that out!

Leave a Reply

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