Thursday, August 6, 2015

Fear and epistemic probability

If I resent your doing A and you didn't do A, then my resentment was perhaps justified (if I was justified in thinking you did A) but it was nonetheless misplaced. On the other hand, if I am crossing the road and I notice a car speeding towards me, and I fear it will run me over, but then the driver brakes and stops just barely in time, my fear was entirely appropriate and not at all misplaced.

The proper object of resentment, thus, is an event (or action) taken as actual (and wrongful), and when that action doesn't occur, the resentment is misplaced. But the proper object of fear is an event merely taken to be a serious chance. What kind of chance? An objective chance or a merely epistemic probability?

I will argue that it's an epistemic probability. Suppose that I fear that my investments will fail. I get into a time machine, travel to the future, and notice that my investments won't in fact fail. I go back in time and it would be appropriate for my fear to go away. Nonetheless, there is an objective chance of the investments failing: the chancy processes that make investments go up and down continue to run despite my knowledge. But there is no longer a serious epistemic probability. So it looks like epistemic probability is what is relevant. Moreover, I think it can be appropriate to have fears about things that are in fact necessarily false. For instance, if I have an answering a multiple choice exam in calculus, and I the question asks whether the definite integral of some function over some range range is 2, 3, 5, or π/2. I think it's probably 5, but there is something in my calculation that I am not confident of, and I realize that if I got that wrong the answer is π/2. My fear that the definite integral might be equal to π/2 is in fact appropriate, even if it is necessarily true that the answer is 5.

This makes fear very different from resentment: fear is made appropriate by epistemic probabilities--either the actual ones or the ones my evidence justifies (which one?), while resentment is made appropriate by what people have actually done.

I wonder if this focus on the epistemic dimension isn't partly responsible for the notorious way that fears resist rational thought. No matter how much I reflect on the very good statistics for indoor wall climbing injuries (the chance of injury during a session is about the same as that while driving 26 miles) and what I know about the stringency of Baylor's training of my belayer, when I look down from 50 feet up, I feel fear. This fear is misplaced: my epistemic probability for a fall is tiny (and justifiedly so given the evidence). Why? Because it looks dangerous. Now, in the absence of defeaters, appearances yield epistemic probabilities. Moreover, many times even though a defeater to an appearance of an impending bad is sufficient to defeat belief, a sufficient epistemic probability will remain (after all, we may be wrong about the defeater), and it could take quite a bit of time to evaluate whether the defeater is complete or only partial. Given that physical danger may require a quick response, and the examination of defeaters takes time, it makes sense for us to be wired in such a way that appearances have a strong tendency to directly drive fear. So in cases like my climbing case, while the fear is misplaced, inappropriate and unjustified, it is nonetheless understandable (unlike my pathological fear of dogs!).

(Well, when I reflect on the fact that an indoor climbing session has equal injury probabilities to a 26 mile drive, this actually makes me feel a bit afraid. For I do think driving (or being driven) by an average driver is genuinely dangerous. And so perhaps my fear is justified, just as I would be justified to be afraid of a 26 mile drive (even if in fact I don't always feel afraid). If so, then change the example, say to standing on a five inch thick glass floor above a precipice.)

9 comments:

SMatthewStolte said...

“My fear that the definite integral might be equal to π/2 is in fact appropriate, even if it is necessarily true that the answer is 5.”
I’m not sure this is what I’m really afraid of. Maybe it is. But for some reason, I feel like fear should be future-directed. It’s at least possible to characterize the math test fear in this way. I am afraid that I will be rightly marked off for getting the answer wrong or that I will be shown to have less than optimal math skills. (Or perhaps I have a disjunctive fear: I am afraid that either (a) I will be shown to have less than optimal math skills or (b) (what would be worse) my mistake will not be noticed and I will not be able to correct myself.) These are still fears that are justified by epistemic possibilities, of course.

Maybe there’s no good reason for trying to make fears seem future directed. “My friend is late! I’m afraid he’s been in an accident,” sounds like a perfectly normal thing to say.

Maybe the reason I feel like fear should be future-directed is that I know that the past is already determined and so not subject to objective probabilities. This knowledge might interfere with my thinking about subjective probabilities.

Fatalists can advertise their doctrine as a way of alleviating fear, even if they don’t offer any knowledge of future events. It seems at least somewhat plausible. Why is that?

Alexander R Pruss said...

There seem to be past-directed fears. "I'm scared that I made a fool of myself when drunk last night." And one can have that fear even if all one's companions were too drunk to remember. (Of course, getting drunk is constitutive of making a fool of oneself no matter what one does when drunk, but I suppose the fear is that one made a fool of oneself in some further way.)

Or: "I hope I gave her the right answer." This makes sense even if nothing further hangs on it: one may have a horror of having said something false, even inadvertently so.

SMatthewStolte said...

Right. I think those are like the car accident example I gave. But is the feeling that there ought to be something future directed about fear entirely a result of the fact that we have less knowledge about salient future events than about past ones?

Alexander R Pruss said...

That sounds right. I skipped over the accident case as I looked at your comment. Sorry. It's a better example than mine.

You can make the point even clearer with past pains. "I fear my child suffered a lot yesterday." This need have nothing to do with the repercussions for the future.

(It's interesting that we're unlikely to have such fears about ourselves. Suppose I might have been tortured yesterday while drunk, and now the memory is all gone. Does it make sense to fear that?)

Alexander R Pruss said...

A parallel point is that we have past-directed hopes. "I hope he didn't suffer much" is the sort of thing people often say, at least in movies.

SMatthewStolte said...

No, I don’t think it does. I might have concerns about the moral status of the torturer or the fact that I am now missing a memory or the possibility that the torture might have caused some other problems that have nothing to do with my memory, but it doesn’t seem right to fear the possibility that I might have suffered. Then again, this thought experiment involves abstracting from all of the causal connections that suffering is usually involved with. And that might be distorting my intuitions, here. I know that a lot of the fear I have of certain pains has to do with things other than the pains in isolation.

Heath White said...

Suppose I know with certainty that I will be torn apart by wild beasts in the Coliseum tomorrow. I might be afraid at this prospect, I think--no uncertainty required.

Curiously, though, one says, "I am afraid BECAUSE I will get torn apart" not "I am afraid THAT I will get torn apart". The latter invokes chance, the former does not. Also, you can put in "fear" for "am afraid" in the latter case but not the former.

Roughly, "I am afraid because..." seems to offer a causal (?) explanation of a largely (?) phenomenological mental state whereas "I fear that..." reports a propositional attitude. I think you can have this latter propositional attitude without much (any?) phenomenology ("I fear I have lost my keys again," said with a mild sigh of resignation) whereas the former is essentially phenomenological. Perhaps you can also have the phenomenological state without the associated attitude, as in Freud's free-floating anxiety.

Maybe there is fear1 which is a credence attached to a negative value judgment, and fear2 which is a phenomenological state of anxiety or dread. The two are often associated but they can come apart.

Alexander R Pruss said...

In the Rhetoric, Aristotle distinguishes fear, whose object is an uncertain future event, and dread, whose object is a certain future event.

In your example, typically there will be uncertainty about what it's like to be torn apart (and there may be uncertainty as to what happens after death). That said, Aristotle may just be wrong. In any case, this is still a case where there is a probability--maybe 1--attached to a negative state.

Angra Mainyu said...

Hi Alexander,

You say: "If I resent your doing A and you didn't do A, then my resentment was perhaps justified (if I was justified in thinking you did A) but it was nonetheless misplaced. On the other hand, if I am crossing the road and I notice a car speeding towards me, and I fear it will run me over, but then the driver brakes and stops just barely in time, my fear was entirely appropriate and not at all misplaced.

The proper object of resentment, thus, is an event (or action) taken as actual (and wrongful), and when that action doesn't occur, the resentment is misplaced. But the proper object of fear is an event merely taken to be a serious chance."

I'm not sure that there is such a difference proper resentment and proper fear. For example, what about the following scenario:

S1: Alice believes Bob is a serial killer who kidnapped, tortured and murdered dozens of people. Given the information available to her, she's epistemically justified in having that belief. Alice resents Bob for killing all of those people, and fears Bob because she believes he's a serial killer. But Bob didn't do any of the sort, and in fact is a good person who would not hurt anyone without good reasons.

In S1, is Alice's fear any more or less misplaced or any more or less proper than Alice's resentment?
I reckon she has no moral obligation or any other obligation not to resent him, given the information available to her. Granted, she resents an innocent person. But she has good reasons to resent him - just as she fears a person who has no intention at all of kidnapping her, etc., but she has good reasons to suspect otherwise.

So, I'm not sure in which sense her resentment would be misplaced, but her fear wouldn't be, or in which sense her fear would be proper, but her resentment improper.
What do you think?