Below the fold are six cases designed to tease out just how we think about the interaction of practical interests and epistemic justification. The main interest will be in looking at cases where agents have mistaken and/or unreasonable beliefs about their practical environment.

The basic form of case is taken from Jeremy Fantl and Matthew McGrath’s _Philosophical Review_ paper.

There are two types of train that leave from the station nearest Harry’s work, the Local and the Express. The Express goes by the station nearest Harry’s home, so he shouldn’t catch that one. (Though normally it isn’t a disaster if he does, because he can just catch a local train back the other direction if he overshoots his stop.) One day on the way home from work, he is running for a train that’s about to leave. He can’t see any sign as to whether it’s a Local or an Express, but he overhears some other passenger say it’s a Local. Passengers aren’t always reliable, but this is some evidence. The whistle has blown on the train and its about to leave so Harry doesn’t have time to get any more information before deciding whether to catch it. He can see that his regular local train will leave in ten minutes. He has a weak preference for arriving home ten minutes earlier than he would arrive on his regular Local. Just how much it will cost him if the train is an Express and he has to catch a train back the other way varies between the cases.

Harry and his wife, Harriet, have an important meeting scheduled at their son’s school about his truancy problem. It is very important to each of them that at least one parent attend the meeting, but not important which of them does, or whether both parents are there. (They trust each other enough to represent their son’s interests at the meeting and to report on the results of the meeting to each other.) The cases below all differ about whether Harriet will make the meeting, and what Harry’s belief is about Harriet’s attendance. Harry will make the meeting if he catches his regular train, but not if he catches this train, and it turns out to be an Express.

*Case One*
Harry believes, correctly and with excellent reason, that Harriet will be there, so there’s no special need for him to attend as well.

*Case Two*
Harry believes, correctly and with excellent reason, that Harriet will not be at the meeting, so he has to attend.

Fantl and McGrath’s core idea is that in case one, but not case two, Harry is justified in believing the train is a local. That sounds plausible to me as well, though I’d want to explain this fact by appeal to the pragmatics of belief not the pragmatics of justification. But for now I want to look at a few more cases.

*Case Three*
Harry has excellent reason to believe Harriet will not be at the school appointment. Harriet called him earlier in the day to say she’d be in a business meeting across town that wouldn’t end in time to make the school appointment. And in fact she is still in her meeting when Harry is running for the train, and unless she steals a helicopter, there is no way to get from the meeting to the appointment in time. Sadly, Harry has forgotten all this and believes Harriet will be at the meeting.

*Case Four*
Harry has just the same reason to believe Harriet will not be at the school. And again he (unreasonably) believes that she will be there. Fortunately, Harriet’s business meeting ended early, and she will be at the school on time.

I think in case three Fantl and McGrath would say that Harry’s belief that the train is a local is unreasonable. What is interesting about this case (and what makes it a hard case for my preferred way of dealing with the case) is that Harry maximises subjective expected utility by acting on this belief and catching the early train. In the paper I’m writing on this I talk about this case at some length. Anyway, I think the intuition most people have in such a case is that it is epistemically just like case two. (Someone who follows Harman in taking beliefs to have evidential value will disagree here, but most epistemologists won’t regard cases two and three as different.)

But this raises a hard question about case four. There are two options here, and neither of them is attractive. First, we could say that because Harriet’s business meeting ended early, Harry is in a low-stakes situation, so his evidence does justify believing the train is a local. This would be a fairly radical form of externalism about justification. (And one that has trouble with case five below.) Second, we could say that what actually happens at Harriet’s business meeting is irrelevant, what matters is what Harry should believe about it. We now are in danger of having a circular theory. Moreover, this seems to lead to puzzles about case six below.

*Case Five*
Harry had excellent reason to believe (as he does) that Harriet would make the school appointment. But due to a freak storm (as we’re having as I write this) Harriet is seriously stuck in traffic as Harry is running for the train. It is nomologically indeterminate whether she will make the school appointment. What determines whether Harry is in a high stakes situation? Whether Harriet actually makes it to the school 30 minutes later? The objective probability of her making it? It’s hard to say.

*Case Six*
Harry has evidence that would, in normal circumstances, justify him in believing that Harriet is running late, so he needs to get to the school on time. Unfortunately, he does not believe this. Fortunately, the proposition _Harriet is running late_ is for independent reasons very high stakes. So his evidence does not justify him in believing that. So he’s not unreasonable to not believe it. So now it looks like he’s back in case one. And it’s reasonable to believe the train is a local and catch it.

I don’t have a unified story about these cases, they’re just illustrations of how complicated things get when we start allowing interests into epistemology.