The Value of Knowledge and Being in a Position to Know

Here’s some thoughts on an idea about the value of knowledge. The idea comes from John Turri, who posted the idea here over at Fake Barn Country. It’s a really interesting idea, one I hadn’t thought of, but well worth thinking about. A word of caution though: biologists say that the optimal strategy for predators and predatees(?) trying to get food or avoid being food is stochastic, and my travels through the logical space of the proposal will be optimal in precisely this sense.

To lead into John’s proposal, consider the Plato account: knowledge is more valuable than true belief because mere true belief is more likely to get up and wander off tomorrow than is knowledge. Not exactly the most careful formulation, but for the best formulation along these lines, see Tim Williamson’s formulation in K&IL.

I think all such Plato accounts fail, on grounds having to do with misleading defeaters. The world can conspire against you so that you will, or are likely to, encounter misleading defeaters in the future. And the true beliefs you have that aren’t knowledge might be, we might put the point, fundamentalist beliefs–entrenched by non-epistemic factors so that any further evidence acquired will be explained away rather than attended to.

OK, enough stagesetting, I think. The central point here is the diachronic nature of the Plato account, for which John substitutes a synchronic account. The details of it have to do with being in a position to know truths other than the particular truth known. John focuses on Sosa’s account, where the known claim p has to be in a field of propositions, where the field is such that, in one’s normal environment, you’d most likely be correct with respect to propositions in that field. As John glosses this account, you are in a position to know the other truths in this field. Since I think counterfactual accounts are nearly always mistaken, I’ll just focus here on John’s gloss rather than on the basis for it in Sosa’s thought.

Notice that instead of Plato’s diachronic view, John’s is synchronic. As such, it provides an interesting alternative in trying to explain the value of knowledge over true belief. So in what follows, I’ll call it the synchronic suggestion; ‘SS’ for short. Can it succeed?

First, some issues about formulation. Take any true belief that p that doesn’t count as knowledge. One one way of understanding SS, we’ll need to focus on true beliefs where such a true belief doesn’t put one in a position to know other truths. That’s going to be hard to find, I think. For example, normal human beings are always in a position to know what they believe and what they don’t, what they desire and what they don’t, etc. And any true belief will put one in a position to know a number of things: that you hold that belief, that you hold a belief not identical to other of your beliefs, that you hold a belief that was formed after you were born, etc. Call this the “vice-versa problem for SS”.

So let’s clarify the proposal a bit differently. Suppose S knows that p and S’ only believes truly that p. S is thereby in a position to know things that S’ is not in a position to know. The proposal is that this difference explains the difference in value between knowledge and (mere) true belief.

We can’t get around the vice-versa problem this way. S’s true belief that p puts S’ in a position to know things that S is not in a position to know, too–things about the mental states of S’, for example.

Maybe we could say this: knowing that p requires being in a position to know a range of truths that merely believing and being correct that p doesn’t. This finally avoids the problem because since knowledge implies true belief, anything you’re in a position to know in virtue of believing the truth you’re also in a position to know in virtue of knowing.

Here’s a deeper concern, however. Take the range of claims you’re in a position to know in virtue of knowing p. Suppose that you know all of these truths. Then compare knowing all of these truths with only believing them and being right.

There are two possibilities here. One is that the class of things one is in a position to know is insular: that is, that the class stays the same regardless of which propositions in the class one comes to know. In such a case, knowing the entire class and correctly believing it will no longer address the value problem. One will not be in a position to know anything that one doesn’t already know, and the true believer will believe all those truths. So if there’s an issue of how knowledge is better than mere true belief, being in a position to know won’t help with such a case.

The other option is that the class in question is not insular. So coming to know something that one is in a position to know may enlarge the class of things one is then in a position to know. In such a case, even when the class of things known vs. truly believed is enlarged beyond a single proposition, there will be further things that one will come to be in a position to know that the true believer won’t be in a position to know.

So, the first point to notice is that the proposal won’t work unless the class of things one is in a position to know is non-insular. Since this post is already pretty long, I’ll stop it here, and take up the non-insularity point in another post.


The Value of Knowledge and Being in a Position to Know — 4 Comments

  1. It seems to me that a limiting case shows that whatever
    complaint you have about the insular version applies to
    the non-insular version.

    Take the range of truths that are knowable by you.
    Compare knowing all these truths with merely believing them.

    The insular version doesn’t appear to add anything of value because
    you will already know what you are in a position to know.

    But, if so, then the non-insular version doesn’t add anything of value either:
    you will already know what you are in a position to know,
    what you are in a position to be in a position to know,
    what you are in a position to be in a position to be in a position to know,
    and so on.

  2. Patrick, welcome to the blog, and this is an excellent point! In fact, it was going to be my next post. The only philosophers who can escape the point are anti-realists who hold that all truths are knowable. Once we get to the point of omniscience, I don’t think they’ll be any difference between true believers and omniscient ones (more carefully, I should say there’ll be no difference between justified true believers and omniscient ones).

    Besides, if one is going to hold that all truths are knowable, there better be a solution to the knowability paradox…

  3. Jon,

    I really appreciate you posting on this. A few things in response.

    On the diachronic/synchronic distinction, I intended my suggestion to complement the credit approach (which, for those reading the thread unfamiliar with this, is the view that knowing p is more valuable than a mere true belief that p because when one knows, one deserves (intellectual) credit for believing truly, which is a good thing). So the overall approach has both diachronic and synchronic elements. Synchronic: It’s at the very time of knowing p that you are in a position to know other propositions in the field. Diachronic: The way in which your belief was formed and sustained determines whether you deserve credit.

    On the vice-versa problem, I agree with your suggestion for resolving it.

    The insularity/non-insularity issue is intriguing. I had pondered this myself when formulating the proposal. However, I wasn’t able to come up with an example of an insular set of propositions. Have you been able to come up with a plausible example? My hunch was that the best strategy would be to focus on a small set of simple arithmetical truths, or of simple introspectable properties of one’s occurrent conscious experience, but I was stumped.


    There might be something to your suggestion, but here’s what worries me about it. I don’t think there is a single, stable group of truths knowable by you. Suppose there are n truths knowable by me at time t, and call the set of them “{n}”. Once I come to know all n truths, the range of truths knowable by me will have increased dramatically. For example, take any two propositions in {n}, p and q, which I previously merely truly believed, but which I now know. I’m now in a position to know the material biconditional p iff q, whereas before I wasn’t.

    You might reply that p iff q was already in {n} in virtue of my being in a position to know p and q. To head this off, two things in response. First, it’s implausible that being in a position to know is closed under deduction. Second, even supposing it is closed under deduction, it seems clear that once I actually know p and q, I am in a much stronger position to know p iff q. My claim would then be that knowledge enahnces the strength of one’s position to know various things, rather than adding to the store of truths that one is in a position to know. Since an enhanced epistemic position is still epistemically valuable, it still speaks to the value problem. (This second response is just backup; I think the first response works just fine.)

  4. John, in your reply to Patrick, I think you are identifying what is knowable with what one is in a position to know. In the example you give, you are not in a position to know the extra items initially, but they are knowable truths. A simple proof of this: what is necessary, in the broad sense, and what is possible, in the broad sense, can’t change over time. And there is no reason to interpret Patrick’s suggestion as involving a notion of possibility stronger than broadly metaphysical.

    In this sense of possibility, the class of knowable truths for an individual is guaranteed to be insular.

Leave a Reply

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