OK so here's my problem with pragmatic encroachment. (It's a problem with moral encroachment, too.)

You might want to stand back, I'm gonna do some philosophy

Background: "pragmatic encroachment" in epistemology is the name for the idea that knowledge depends on certain kinds of practical considerations. Whether you know something depends on e.g. how important the question is.

In situations where the cost of being wrong is high — "high stakes cases", in the parlance — it takes more evidence to know.

Pragmatic encroachment is a pretty new idea; it's basically a 21st-century theory (though one can certainly draw connections to e.g. American pragmatists)

Moral encroachment was the next hot new thing. The idea here is, when it's morally quite risky to go around forming beliefs (say, because you might be forming a racist belief), the standard for knowledge goes up — it takes more evidence to count as knowing.

(Moral and pragmatic encroachment are different in subtle but crucial ways from contextualism, which is a view about language. I like contextualism. This thread isn't about contextualism.)

Right, so, moral and pragmatic encroachment are motivated in part by plausible intuitions about cases: when the stakes go up, you shouldn't count on the belief unless you have tons of evidence, so it requires tons of evidence to have knowledge.

But there's a problem.

How are you supposed to recognize high-stakes situations? How do you know how much evidence you need before you have knowledge?

You could reason about stakes based on what you know, but that just pushes the same problem down a level.

Here's another way to come at the problem:

Suppose I'm trying to figure out what to do. (Should I wait in the bank queue? Should I give my keys to this guy standing near the valet stand?)

I think the way to figure out what to do is to use evidence to figure out what's going on, then use values to figure out what's best.

Complex moral and practical problems, I think, can and often should be factorized into somewhat simpler epistemological problems and ethical problems.

But moral and pragmatic encroachers have a hard time saying this.

If you're an encroacher, you think that practical or moral considerations about what you should do partly determine what you should believe about what is the case.

So you can't start by using your evidence to figure out what's going on, and use that to make a good decision.

To speak metaphorically, if the facts about knowledge are floating along in response to the facts about what you should do, your knowledge isn't helping you act. You have to just figure out what to do, and then the questions about knowledge just sort of fall out of that.

I think knowledge is more practically important than that.

If you think we should make our decisions based on knowledge, and that knowledge actually makes it easier to make good decisions, then you can't think that knowledge depends on which decision is best.

This is part of why, for example, it's really valuable and important for a democratic public to be well-informed. We want them to use information to help make good decisions!

(And if you don't care as much as I do about knowledge, you can run exactly the same argument about "evidence" or "rational belief" or whatever epistemic state you do care about. This is a reason to think states you think are important, aren't encroached upon.)

So anyway, that's my concern with pragmatic and moral encroachment. Deciding what to think can be a good first step toward deciding what to do. So it has to be possible to decide what to think without first knowing what to do.

If you want to read a more careful version of it with cites, Ben Jarvis, Katherine Rubin, and I published this paper in 2012. For whatever reason I don't think it got a lot of notice/uptake but I think there's something important there. https://philpapers.org/rec/ICHPEA