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.