What Needs Explaining about the Value of Knowledge

An issue that comes up in discussions of the value of knowledge (and which will come up this week at our session at the Central meeting on the value turn in epistemology) is the special value knowledge is supposed to have when compared with mere true belief. If knowledge is only sometime and somewhere more valuable than true belief, it won’t be as difficult to explain its special value. But if it is always and necessarily more valuable than true belief, things get harder. So which is it?
My proposal is that it is the latter, but to make the claim plausible, some qualifications are needed. To say that greater value is present always and necessarily should be understood as we understand the claim that probabilistic causes are probability enhancers. To this latter claim, counterexamples abound. One of my favorites is the relationship between taking birth control pills and thrombosis. Taking the pill causes thrombosis (as best we can tell at this point), but lowers your chances of getting thrombosis. The reason is causal interaction: taking the pill also prevents pregnancy, which is itself a cause of thrombosis. The trick to uncovering probabilistic enhancement for causal factors is in the controls: hold fixed the right factors and the enhancement shows up. Of course, it is hard to say exactly what needs to be held fixed, but once we do, we get the right result.
The same holds for the value of knowledge. If you threaten to kill my family if I should happen to have knowledge rather than mere true belief, we have a counterexample to the claim that knowledge is always and necessarily more valuable than mere true belief. We should suspect, though, that the underlying idea remains untouched. We should revise in just the way proposed for understanding probabilistic causation. So we should hold fixed other values that are independent of purely theoretical considerations. Once we do so, we should find the special value of knowledge, if it has such special value. As the death to the family example shows, the value is defeasible, but it is no less present always and necessarily because of that.
This result helps explain why it isn’t satisfying to explain the value of knowledge by pointing out that if you know something today you are more likely to continue to believe it tomorrow than if you happen to merely believe it and are correct. One problem with this proposal is that it isn’t necessarily true, and another problem is that it isn’t always true. So if this is all we can say about the special value of knowledge, we should view ourselves as having been defeated by the Meno problem (the problem Meno raises to Socrates of explaining why we prize knowledge over true opinion). We should view Meno as having been taken in by some special cases of knowledge, and as having illegitimately generalized from those special cases. We should walk away with our epistemologist head bowed low, sad to find out that knowledge isn’t as special as we wanted it to be. The account is, as Duncan put it in San Francisco, revisionary of our attitude concerning the value of knowledge, rather than explanatory of it.


Comments

What Needs Explaining about the Value of Knowledge — 10 Comments

  1. I don’t see why one should be disappointed to find that knowledge is not always and necessarily more valuable than mere true belief. Suppose knowledge usually (but not always and not necessarily) has a special value that mere true belief always and necessarily lacks. Wouldn’t that be enough to explain why one might believe that knowledge has a special value? Or suppose that knowledge always and necessarily has a special value that mere true belief sometimes (but rarely) has too. What would be disappointing or revisionary about that?

  2. Suppose causes sometimes enhance probability and sometimes don’t. That’s not good. Just so, Meno is rightly perplexed when Socrates points out that true opinion is as useful as knowledge. So, a first stab at addressing the concern is this: control for usefulness and we are on our way to finding out why we prize knowledge. But why control at all? Why not say, “sometimes one thing, sometimes another is better?” Because Meno is right. We do prize knowledge over mere true belief. Maybe that’s misplaced prizing, though.

  3. I’m not sure I understand. Let us control for usefulness as you suggest.
    What I am wondering is why we should adhere to

    (1) Knowledge is always and necessarily more valuable than mere true belief,

    as opposed to something slightly weaker. Here is another example:

    (2) Knowledge is always and necessarily at least as valuable as mere true belief, and indeed 99.99% of the time knowledge is more valuable than mere true belief.

    Why can’t (2) satisfactorily explain why we do and should prize knowledge
    over mere true belief? Why do we need (1)?

  4. Take a case where the revised account says they have the same value or that true belief is more valuable. What’s the explanation? Whatever it is, we can then control for it, and have a nice account of the stronger sort. Full generality is a good thing in a theory, so if we can have it, we should.

    So perhaps there is nothing to control for. It is just random. Nearly always knowledge is more valuable. On occasion true belief alone is. If that’s the story, we need more. We need a defense of the claim that there is no hidden variable explaining the difference between the two types of cases.

    So (2) isn’t a deep enough explanation on its own. But there is another point as well. When people compare things (ducks are better swimmers than humans, for example), we don’t expect the universal claim to be true until controls are in place to prevent counting, e.g., dead ducks, disabled ducks, etc. But once the controls are in place, it would be misleading to still insist that ducks are better swimmers, if the universal claim weren’t intended.

  5. Does the value problem arise mainly for objectivists about value? If one were to think that the only sense in which something is valuable is that it is valued, does the value problem dissolve? Becuase all I have to do, if I’m a subjectivist, to show that knowledge has a special value, is show that I (or we) value it. Socrates’ question becomes: how do we value knowledge, intrinsically or instrumentally (or in some other way)? We have a challenge to the idea that we value knowledge instrumentally, because true belief is just as useful as knowledge. Conclusion: we must value knowledge intrinsically. Now, given that, is there anything more, for the subjectivist? It seems like the question on everyone’s minds is the question of whether knowledge really has any intrinsic value, additional to that possessed by true belief. But for subjectivists, isn’t this a silly question? It has additional value if (and only if) we value it additionally. And it seems that we do – or at least some of us, or many of us, do, probably to varying degrees.

  6. Jon,

    Is this the view under consideration now?

    Defeasible value of knowledge (DVK):
    Necessarily, for all beliefs b: it is defeasibly, intrinsically better for b to be knowledge rather than mere true belief.

    (I think you want the ‘intrinsically’ in there, right?) Threatening examples are then viewed as introducting defeaters of the essential, defeasible, comparative intrinsic superiority of knowledge over mere true belief.

  7. Hi Allan, there’s something right about what you say on behalf of this kind of subjectivist. They are sad souls though–their theory commits them to being infallibilists about what is valuable!

    OK, back to the program. If you are a subjectivist and you identify value with valuing, you might still want an explanation for why you value one thing more than another. One such explanation is that you value it intrinsically, for its own sake, and not for the sake of something else. Even so, you might wonder what it is in virtue of which you value it this way. If the answer is “nothing”, that’s a bit perplexing. Why would I value something like this for no reason whatsoever? On any theory of value, we want a supervenience claim to be true: fix the non-values and you thereby fix the values. And if we can identify the supervenience base, that allows us to engage in learning about values, because we can notice inconsistencies in what we value. So even a subjectivist shouldn’t be completely content to answer every value question simply by recording present attitudes.

  8. Hi John, yes, that’s what I’m thinking. I don’t want to rule out extrinsic, instrumental explanations a priori, but the ones that come to mind aren’t plausible (e.g., the usefulness idea that Socrates rejects). So I think the answer will have to be that knowledge is intrinsically superior to true belief. But it can’t be indefeasibly superior, and your principle captures that point nicely.

  9. Dear Jon, I hope all is well, and that you’re taking your precious tar-heels’ hammering in stride:) I just wanted to run by you some swamp-related thoughts that have been puzzling me of late. First off, I think you’re spot on in your characterization of the ‘Meno Assumption’ as a prima facie claim about the comparative epistemic value of knowledge and mere true belief. Suppose now that an epistemologist wants to show that her account of knowledge is (as you say in your book) at least logically consistent with the Meno Assumption. One way to do this would be to do what Greco and Sosa do–by giving examples like this: Bill’s knowing p (i.e. a cognitive success through ability) is more valuable than were he to merely truly believe p (i.e. to form the true belief in a way that his ability does not explain the success). The thought at play here is that, if we buy the achievements-are-finally-valuable idea, we’d be inclined to think that the general strategy Sosa and Greco offer has preserved the Meno Assumption. What has been concerning me of late is not anything to do with the substantive parts of this kind of response, but rather, the strategy of trying to preserve what I think is as assumption about the comparative values of TYPES of epistemic standings by showing how a given theory implies that a token of one is more valuable than a token of another. This strategy seems to be the modus operandi among epistemologists trying to meet the demands of Meno; Goldman and Erik Olson (whose substnative approaches differ significantly from the Sosa/Greco approach), for example, also assume that such a strategy is without problem. Upon closer inspection, though, this strategy seems questionable. Consider that whatever properties knowledge has as a type of epistemic standing, on which value would supervene, would be properties shared by all knowledge tokens. However, the properties of the knowledge tokens that folks like Greco/Sosa/Goldman/Olson point to when claiming to preserve (on their respective theories) the Meno Assumption are ones that not all tokens share, and so wouldn’t be what makes one type of standing more valuable than another. For example, whatever it is that makes Bill’s knowledge of where his keys are more valuable than merely truly believing them seems to be some property which is epistemically valuable at least in part because Bill wants to know where his keys are. But whatever property would make knowledge as a type of standing more valuable than mere true belief wouldn’t be a property on which the value that supervenes on that property is constituted in part by Bill’s interests. A related thought that I think might help to motivate this sort of worry is that epistemologists who disagree significantly with respect to each other’s substantive accounts of knowledge tend to buy into the Meno Assumption. This suggests that whatever it is about knowledge by virtue of which epistemologists with radically disparate theories of knowledge are lead to think it more epistemically valuable than true belief is some theoretically-neutral feature of one epistemic standing not shared by the other, and something different from whatever property would make a knowledge token more valuable than a corresponding true belief token only if some particular theory of knowledge is correct. In sum, then, my two worries about the widely used modus operandi of responding to Meno is that it (i) Assumes, and I think unwarrantedly, that what makes knowledge comparatively more epistemically valuable qua type than true belief qua type would also be the salient value bearer of some K token that would serves to make it more epistemically valuable than some corresponding T token; and (2) The modus operandi is one that attempts to explain an intuition shared across widely differing substantive theories of knowledge in a way depends on a particular theory of knowledge being correct.
    These thoughts are at this point a bit inchoate, but I’d nonetheless be curious to know whether you share either of these concerns (or perhaps find them to be misguided). Cheers, Adam

  10. Adam, I’m not sure I follow your concerns here, but I *think* some of them I talk about in the new piece about the swamping problem that Duncan is publishing in the new collection on the value problem. If that’s not relevant then I’m not quite sure what you are after.

Leave a Reply

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