Friday, May 23, 2008

A quick question about reasons

Suppose there's a reason, call him 'Randy', and it speaks in favor of pushing the red button. Suppose there's a reason, call her 'Griselda', and it speaks in favor of pushing the green button. Suppose it's impossible to push both buttons. Question. Does Griselda _also_ speak against pushing the red button? I'm not asking whether it could. I'm asking whether it _must_ simply in virtue of (a) that Griselda speaks in favor of pushing the green button and (b) you cannot push the green and the red.

I'm really not sure what to think. I can see Randy speaking in favor of pushing the red button without thereby speaking against pushing the red button and the same for Griselda. Of course, pushing the red means missing the opportunity to conform to the demands Griselda made. Pushing the green means missing the opportunity to conform to the demands made by Randy.

Here's a little argument that I _think_ is a decent little argument against the view that Griselda _must_ speak against pushing the red button. Changing examples should help. Jones is out for his morning constitutional when he sees a man fall into the river. Seeing that the man was struggling to pull himself out and is in serious danger of being pulled away from the bank, he rushes over to lend a hand. Smith is heading out to meet Jones the next day for lunch when he sees a different man fall into the same river. Seeing that this man was struggling to pull himself out and in danger of being pulled away from the bank, he rushes over to lend a hand knowing full well that his assisting this stranger would prevent him from keeping his appointment with Jones. Now, the strength of the case for Φ-ing is a function of both the reasons to Φ and the reasons to refrain from Φ-ing. It seems that the case for Smith to assist and the case for Jones to assist is equally strong. It follows that the reasons that favored keeping the meeting were not reasons to refrain from helping the man.

Thoughts? I first became interested in this issue when I was trying to explain Ross' view in class and it struck me that I wasn't quite sure whether Ross would say that the reasons associated with each of his prima facie duties _both_ asked us to perform acts of certain types and act against the other duties on the list. (Of course, he believed in real conflict but I don't think it is obvious that believing in real conflict and rejecting the specificationist view on which moral principles 'anticipate' each of the circumstances under which they fail to determine what ought to be done all things considered and treats them as exceptions you have to take a stand one way or the other.)


Errol Lord said...

Although I think it's obvious that both Jones and Smith have overriding reason to help, I'm not convinced that 'the reasons that favored keeping the meeting were not reasons to refrain from helping the man.' I'm not even sure that the weight of the set of reasons to assist is equal for both Jones and Smith. But I kinda sorta feel the intuition in that direction. Nevertheless, I think that that intuition can be explained without concluding that Smith's reason to keep his appointment doesn't count against helping. For example, you might think that Jones's reason to have a pleasant walk is roughly equal in weight to Smith's reason to keep his date. The latter reason might speak against helping just as the former does. Nevertheless, one might still have the intuition taht the weight of reasons is equal for Jones and Smith. I actually think that this is what guides my intuitions in this case.

In order to really test what you want, I think you need a case where Jones has absolutely no reason to do something other than helping and Smith has some fairly trivial reason to do something other than helping. If one's intuition is that the weight of Smith's reasons to help is equal to Jones's reasons to help, then Smith's trivial reasons to do something else won't count against helping. I doubt you can construct such a case because in almost every situation there will be some at least very weak reason for Jones to do something other than help.

Clayton said...

Hey Errol,

That's helpful, thanks.

As I was imagining the case, there really was nothing that spoke in favor of taking the morning constitutional. It was just a habit he'd fallen into. If you are worried about contamination, I think we could follow your suggestion by making the reason to perform the action that would prevent helping stronger. We'd have two cases where there is a reason to assist coming into conflict with reasons where in C1, the opposing reason is weaker than in C2. Shouldn't that work?

Also, part of what I want to do is suggest that there's a subtle difference between the overall weight and the strength of the case in favor of what is all things considered right. That there are (defeated, as it happens) that speak in favor of B-ing when A-ing is the thing to do does not diminish the strength of the case for A-ing. It would have been just as strong even if there were no reasons that favored B-ing. At least, that's what I'm trying to motivate.

Aaron Boyden said...

I'm willing to bite the bullet and say Jones has a minutely stronger case for helping. It does worry me a bit, of course, that I seem to be agreeing with Kant (the thesis of an analytic connection between means and ends).

Clayton said...


Thanks for stopping by and commenting. I think it's a first, no?

I was hoping you might say more about the Kantian thesis. I was worried that there might be something in the ballpark of Kant's thesis about willing the means and the ends that might conflict with the point I'm trying to make, but it might depend on its formulation and might just make the issue that much more interesting.

(*) Any reason R that is a reason for S to X in circumstances where S knows S cannot X without thereby Y-ing is thereby a reason to Y.

I take it that this is the sort of thing you had in mind. I'm not sure it is true. I cannot mix a drink without leaving impressions on the carpet as I move to the bar. That I'm thirsty, however, is no reason to make impressions on the carpet. At least, it doesn't seem to be. I'd be happy to use such a case to reject (*), but I don't know what to put in its place to cause trouble. As (*) is formulated, it seems that any reason to X is a reason to both pursue the necessary means and bring about the unavoidable consequences. I don't think that this is right, however.

Tim said...

I’m a new reader of your blog; so I don’t know the background of this discussion. Anyway, I hope my comment is relevant.

You reject the following principle:

(P) If phi-ing and psi-ing are incompatible, a reason for phi-ing is always a reason against psi-ing.

I’m not sure what to say about the Jones/Smith-case. But I think that there’re much simpler counterexamples. Here is one: Suppose I want to buy a book for my forthcoming weekend trip. I can afford only one. Book A is a pageturner. Book B is a pageturner and funny. That book A is a pageturner is a reason for choosing book A. However, this is not a reason against buying book B. Otherwise there would be a reason against buying this particular pageturner just because there are other pageturners.

Am I missing something? Or do I misunderstand the principle you’re attacking?

Clayton said...

Hey Tim,

Thanks for the comments and welcome.

There's a bit of background to the discussion, but I've found that if I fill in all the details no one will read them and comment.

I think the example you offer is perfectly good. I suppose there's a bit of complication because you have two options with a common good making factor and some will say there's nothing to the first that the second does not provide. But, I think this is likely incidental and another indication that I tend to do things the hard way.


Michael said...

It may also help to consider Buridan's Ass. In this case, since the two bales of hay are equally appealing, the ass has just as strong a reason to go for the one as the for the other. Does he also have reasons against going for each because of the benefit he would gain by going for the other?

An important question to consider here is how strong such putative reasons-against would be in comparison to the undeniable reasons-for. One seemingly natural view would be that reasons-against are just as strong as the reasons-for, if the benefits gained or missed out on are equal. But this leads to the conclusion that the reasons-against and reasons-for cancel out for each alternative, leaving the ass with no reason to select either bale.

This is a problem in party because it would make impossible the traditional solution to the "problem" of the ass. If the ass has, on balance, positive-but-equal reasons to go for either bale then he can appeal to a higher-order reason to justify choosing one arbitrarily. But if neither alternative has any "value" in itself, in the sense of there being a positive reason for pursuing it, then there would be no higher-order reason saying it is better to pick one arbitrarily rather than none at all.

This implies that either the reasons-against must be weaker than the corresponding reasons-for, or they don't exist at all. In part because it seems difficult to say why the reasons-against would be weaker when what is at stake is just the same, I'm tempted to conclude that the existence of multiple attractive alternatives doesn't necessarily imply reasons against each.

Errol Lord said...

So now I'm not sure what you want, and I am thus not sure if I disagree or not. Here is what you say you are trying to motivate in your first response to me:

That there are (defeated, as it happens) that speak in favor of B-ing when A-ing is the thing to do does not diminish the strength of the case for A-ing. It would have been just as strong even if there were no reasons that favored B-ing. At least, that's what I'm trying to motivate.

I think I accept that. The way I look at the weights is like this: there is a set of reasons for A-ing (call it set A) and a set of reasons for B-ing (call it set B). The fact that B exists doesn't diminish the strength of what's in A. However, when we decide what the thing to do is, we must see whether there is more reason to prefer A to B or vice versa. In that case, the fact that B is non-empty and that B-ing is incompatible with A-ing does speak against A-ing. It might be that the thing to do really is A, but that doesn't change that there is a reason against A-ing provided by the facts that B is non-empty and B-ing is incompatible with A-ing.

I take the quote above to agree with me that the mere fact that B exists doesn't diminish the weight of the things in A. If that is all you are out to show, then I agree. But I thought that in the original post you wanted to show that there are cases where B is non-empty and B-ing is incompatible with A-ing, but that those facts don't provide a reason against A-ing. I think that is false.

In discussions like this I think we have to keep in mind the lessons to be learned from Schroeder's negative reason existential fallacy work. For example, I don't think that Tim's case works. To see this, imagine that you go into the book store looking for a pageturner to read. I have the intuition that you have a reason at that point to get any of the pageturners in the store. But if that's true, then why would that reason simply disappear when you decide you want a pageturner that's also funny? I think that your reasons for getting the pageturners that aren't funny are outweighed in that case, but I don't think they disappear.

I generally look at these cases like that. So, I doubt you will be able to construct a case that avoids these worries that clearly shows what you want it to.

Michael said...

So, we're asking whether a reason to B is always a reason not to A, where A and B are incompatible. It looks like we've got two sorts of candidate counterexample cases here. Both seem viable to me, and it might help to distinguish between them more clearly.

1. Cases where the reason to B is positive, but trivial (either absolutely, or in comparison to the reasons to A.) Clayton's case with Smith & Jones falls into this category.

The idea here is that a weak reason to B really can "disappear" completely in the presence of a strong reason to A, when A and B are incompatible. If the reason to B disappears in the presence of the reason to A, it cannot be a reason not to A. I think this is right. To alter the example a bit: while wanting to go on a pleasant walk is a good reason to take that walk if you've got nothing better to do, it is NO reason at all not to help a person whose life if in peril. (In the same bin: the fact that Bush likes cake was no reason at all for him to eat cake instead of doing his job the day after Hurricane Katrina made landfall.)

I think this could plausibly be tested by looking to see whether the agent would, after the fact, regret not having been able to B because he had to A instead. Imagine you have just saved someone's life. Do you regret that you did not get to go for a walk? Of course not; you can go for a walk any time. And, conversely, if you had still gone on a walk knowing that someone was in peril, you would have regretted it greatly -- regretted going on the walk, even, not just regretted not saving the person.

Nor is the fact that one would not regret not being able to B here a mere function of the fact that it is a better choice on balance to A; for there are many cases in which we still regret not being able to take the less attractive of two good options. For example, suppose you have two great job offers. Even if you end up completely satisfied with the one you took, you might always wistfully wonder how things would have turned out if you had taken the other. The present case is not like that. Smith will not wistfully wonder how things would have turned out if he had kept on his walk, for he knows they would have turned out very regrettably.

This is of course not to deny that there is some prima facie reason to B if we block A from consideration completely. But that is not what is at issue, since asking whether the reason to B is a reason not to A requires that they be compared in context.

And then we have...

2. Cases where the reason to B is positive but (at least possibly) redundant, given the context. Tim's example is like this.

Cases like this will probably ultimately need some sort of analysis in terms of sets of reasons, which I am probably not qualified to give. But consider this example. There is a jar filled with many identical candies. My mother, who is strict about sweets, allows that I may take one and only one piece. Since I love sweets, there is clearly a sense in which, for any candy, I have a reason to select that candy. Can we really think that the presence of each piece of candy counts as a reason against selecting every other piece? Surely this would be absurd. For, say there are 1,000 candies. Then for any candy there will be one reason for selection and 999 reasons against. On almost any weighting this would yield that I have much greater reason not to select any given piece than I have to select it. It also seems to imply that adding more and more candy to jar would make each individual candy progressively more dubious. That can't be right.