Tuesday, January 13, 2009

Does 'Has Reason' Imply 'Can'?

I'm in the process of reading through Mark Schroeder's latest (fascinating) book, Slaves of the Passions. Therein, he makes the following argument about Brett, who is in a position where he can only satisfy one of two conflicting desires:

". . . there simply can't be a reason for Brett to do everything that promotes his desires. His very situation shows that it is impossible for him to do everything that promotes his desires. So on a generalization of the principle that 'ought' implies 'can', we simply shouldn't accept that there is a reason for Brett to do this." (53)

What I'm curious about here is this generalization of 'ought' implies 'can'. Do we need to think that the generalization holds, that 'has reason' also implies 'can'? I will make a few preliminary remarks about my thoughts so far, and then I'd like to see what people think.

First, a bit about why we might think this is true: Of course, one might think that 'has reason' implies 'can' is just intuitively true (perhaps in the same way that 'ought' implies 'can' is). For a more principled reason, consider the fact that many (including myself) take it that oughts and reasons can be understood in terms of one another, allowing for the following platitudes (my thanks to Christian Coons for mentioning the relevance of the latter): (A) x ought to φ just in case x has most reason to φ; (B) If x has reason to φ then, ceteris paribus, x ought to φ.

If one accepts (A) and (B), it does seem prima facie plausible that having a reason to φ means one can φ. After all, consider a case in which one's only reason were to do some impossible thing. In such a case, one would (it seems) have most reason to do that thing, in which case—by (A)—one ought to do that thing. But then, of course, 'ought' implies 'can' has been violated.

But I find none of this particularly persuasive. First, I do not find the idea that 'has reason' implies 'can' intuitive. It seems to me that if I want to experience what it is like to be a bat, I have some reason to turn into a bat.* Now, of course, it is not possible for me to turn into a bat, and so it will never be the case that I ought to do so, but this doesn't make me think that I have no reason to do so.

But what of the connection to 'ought'? I'm not sure how I feel about this, but let's try it on for size: What if I take the very fact that it is impossible for me to turn into a bat as a very good reason for me not to do so? What's more, I take it that any act's being impossible is going to be a very good reason for me not to perform that act. Then (A) and (B) don't seem to pose a problem any longer. It will never turn out to be the case that my reason to do something impossible is what I have most reason to do (or that all else is really equal) because I will always have a stronger reason not to do it—namely that it is impossible.

Then again, I must admit I'm a bit queasy about the idea that something's being impossible is a reason not to do it. There is something odd about the idea that I take the fact that I can't φ as a consideration against my φ-ing. Then again, there might be other solutions in the same area. Perhaps the impossibility isn't itself a reason against, but there is some other explanation for why it will never turn out that in the balancing of reasons, those to do impossible things rise to the top (or that everything else is ever equal when one has reason to do an impossible thing).

Anyway, I'd like to know what people think. Where do intuitions lie? Is this just an intuitive matter? Or is there some more principled reason to accept 'has reason' implies 'can', perhaps stemming from acceptance of 'ought' implies 'can'?

*Actually, I don't believe this at all, but insofar as I'm playing with this Humean framework, I think it plausible that if desires were connected with reasons in the relevant way, then this desire would mean I had this reason.


  1. For what its worth, there does seem to be something odd about the thought that there could be a class of acts (the impossible) which one could "have reason" to do yet which one could never, even under the wildest hypothetical circumstances consistent with the act's remaining in the class of "the impossible", have most reason to do. I'm not sure why I find that odd. But I'm certain that I do.

  2. Maybe it's that a reason to do something impossible is a reason that one could never (rationally? or just never simpliciter? maybe both?) act on under even the wildest hypothetical circumstances.

  3. P.M. Jaworski1/15/2009 6:52 PM

    If you're looking for intuitions, here's mine: We can have reason to do what is impossible. I have a reason to fly under my own powers, because that would make getting from my office to my apartment a cinch. I think we can see this if, in the future, we figure out a way to actually fly under our own powers. Then someone could say, "I always had reason to fly, I just didn't have the means."

    Apart from intuitions, I think the idea that "has reason" implies "can" makes certain "common sense" expressions impossible. Of two things that I could have done, but can't do simultaneously, I want a way to explain that there were benefits for choosing the less choiceworthy of the two, even if I don't choose it. One way to do this is by saying that I had reason to do it.

    "Didn't you have a reason to x?" someone might ask.

    "Sure I did," might be the response, "I just had more reason to y."

    Separately, there seems to be something a bit fuzzy about the claim "ought implies can," at least with the way that's phrased.

    I ought to keep my promises, but I probably ought to stop and help this guy who just careened off the road, flipping his Jeep Liberty a couple of times, uhm, more. I can't do both (suppose). But it's still true that I ought to keep my promises, it's just that, then, my helping this guy at the side of the road has more "ought" power than the promise-keeping ought. (Which is why I owe, at the very least, an explanation for why I missed our movie date. Otherwise, if the latter ought obliterated the former ought, I'd owe you nothing). It's why it makes sense to speak of "all-in" oughts.

    By the way, I think something fruitful might come out of looking at wide-scope oughts here too, but I can't think of it myself. Just a gut feeling, is all.

  4. P.M. Jaworski1/15/2009 6:54 PM

    Oh, just ask Corwin about whether or not we owe an explanation if we break a promise because we helped some guy get out of his Jeep Liberty after a car accident. He'll tell you that that's true.

  5. Corwin,

    "Maybe it's that a reason to do something impossible is a reason that one could never . . . act on under even the wildest hypothetical circumstances."

    This addresses the particular case that I was discussing, but 'has reason' implies 'can' seems to be stronger than this. I take it that 'ought' implies 'can' is about nomological possibility. If I can't actually save Bob's life, then it can't be the case that I ought to. I guess my reading of 'ought' implies 'can' stemmed from some sort of intuition about its being "unfair" if morality (or any system of genuine normativity) were to expect things of us that we could not do. But your finding it odd to say that one has reason to do something impossible because it is inconceivable seems to be about some other kind of possibility. So, I just wonder what happens when we look at cases over different kinds. Do all of the following seem problematic?:

    1. Alicia has reason to fly under her own power. (nomologically impossible)
    2. Alicia has reason to travel back in time and kill herself as an infant. (metaphysically impossible)
    3. Alicia has reason to be married to a bachelor. (conceptually impossible)
    4. Alicia has reason to do everything that promotes her (sometimes inconsistent) desires. (logically impossible)

    My intuition seems to be that all of these are things one might have reason to do, except perhaps (3). Again, I guess my thought was that there is a difference between what a normative system expects of one (what one ought to do and thus what one must be able to do) and what one "merely" has some (perhaps very weak) reason to do.

  6. Peter,

    I certainly agree with what you say in the first paragraph, and I think this highlights part of my point in my comment to Corwin. Whether or not 'has reason' implies 'can' in some way, it doesn't seem like it could be a mere extension of 'ought' implies 'can' because the latter is about what we can actually do and it seems plausible that I could have a reason to do something I am as yet unable to do (and not merely have a reason to try to become able to do it).

    I don't see, though, why 'has reason' implying 'can' makes these "common sense" claims impossible. Surely, I can have reasons to do two competing things, and be able to do each, though I am not able to do both (that is I can do the thing I have lesser reason to do, just not if I do the thing I have greater reason to do). I don't see why this is incompatible with 'has reason' implies 'can' unless I have a reason to do both, as opposed to reason(s) to do each.

    Your last concern, I think, just marks a terminological difference. You talk about 'oughts' and 'all-in oughts'; I just talk about 'oughts' and add ceteris paribus clauses to some of them. Really, I think, what we mean when we say that one ought always keep one's promises is just that one always has a (very strong) reason to keep one's promises. And then, as I said in my original post, I take it as a platitude that if one has a reason to φ then, ceteris paribus, one ought to φ.

    I'm not going to touch that wide-scope thing.

  7. Dave:

    I think you're right that the intuitions depend on the kind of impossibility you're talking about. I'm with you that there's no reason to do the conceptually impossible. On the other hand, I see no reason why one couldn't have a reason to do the nomologically impossible.

    Re: the metaphysically impossible - I'm worried by the example you use. I'm not sure it's an intuitively clear enough case of "metaphysical impossibility" to generate clear intuitions. A lot of people will probably think it's just a matter of nomological impossibility. In fact, I'm not even sure how to intuitively respond to "metaphysical impossibility" at all where this is clearly distinct from logical or conceptual impossibility. But that's mostly because I'm metaphysically ignorant.

    Re: the logical impossibility case. Since this is the case of interest with respect to Schroeder. I'm actually not sure that this is a case of logical impossibility, at least depending on how we understand the key phrase "everything that promotes her (sometimes inconsistent) desires". You might think that at T1 I have desires for X,Y, and Z that cannot all be satisfied. Satisfying any two of them will prevent me from ever satisfying the third. Nevertheless, at T1 you might think that I have reason to satisfy X, reason to satisfy Y, and reason to satisfy Z. And there's nothing, at T1, logically impossible about my satisfying X, satisfying Y, or satisfying Z. So I've got reason to satisfy each (and thus all) of my desires at least at one level of description since at T1 I can, nomologicaly, logically, conceptually, metaphysically, act so as to satisfy any of them. So if you think of "having reason to satisfy all of your desires" as equivalent to "for every desire, I have reason to satisfy it" then it's certainly not logically impossible until, in the hypothetical case described, you choose which two to pursue, thus ruling out the third. But if "having reason to satisfy all of your desires" is supposed to mean "having reason to do that which will leave none of my desires unsatisfied" then that's clearly logically impossible even at T1 prior to the choice.

    So: The fact that I cannot satisfy all of my desires doesn't necessarily imply that, for any and every desire, I cannot have reason to satisfy it. But this might be because, at the relevant level of description, it's not a case of logical impossibility. I can have reason to do satisfy each of my desires, even though I can't possibly satisfy them all (as a set), since, until I start picking, there's none of them that I can't satisfy.

    When we render "having a reason to satisfy all of my (sometimes inconsistent) desires" as "having a reason to leave none of my (sometimes inconsistent) desires unsatisfied", this looks like a clearer case of logical impossibility, and it seems that in such a case there can be no such reason.

    So here's the take-away:
    Nomological Impossibility - Sure, I can have reason to do these things.

    Metaphysically Impossible - ::channeling Frey - though he'd never be as ignorant as me of such matters:: "I don't understand what you mean"

    Conceptually Impossible - No reason to do that I can see.

    Logically Impossible - No reason to do that I can see.

  8. P.M. Jaworski1/16/2009 11:30 AM

    This is great stuff. David, the distinctions between kinds of impossibility is very useful. I think I've found one use for wide-scope vs. narrow-scope reason.

    When Corwin questions your example of the logically impossible, you could simply rephrase your sentence to make your meaning more plain. Like this: Alicia has reason to (satisfy preference x, satisfy preference y, satisfy preference z). That would make "has reason" range over all three, and it seems clear that that was your meaning.

    Oops, I'm late. More (possibly) later.

  9. Corwin,

    Schroeder discusses the ambiguity in the "reason to do everything that promotes one's desires" case. I should have been more clear that I meant this to be read de dicto; that is, it is not merely that one has reasons to do each thing that would promote one's desires, but that one has reason to do all of them. On that reading, it is logically impossible, and that is the case where I am curious whether one can have a reason. Obviously, your intuitions are that one cannot. I suspect many would agree with you, but I'm still curious whether there are any more principled arguments. And notice, importantly, that the fact that you are willing to countenance reasons to do that which is nomologically impossible indicates (I think) that your position is not just a strict generalization from 'ought' implies 'can'.

    As to metaphysical impossibility: I'm actually with you here; I tend to be suspicious of the distinction between metaphysical and conceptual possibility. I admit, though, that I continue to use the notions because I'm not sure how else to deal with cases like water and H20.

    Anyway, as to the specific case I offer: This is supposed to be metaphysically (and thus nomologically) impossible, but both logically and conceptually possible. Why? Well, certainly there is no logical contradiction in my saying that I am going to go back in time and kill myself as a baby. Likewise, there doesn't seem to be anything in any of the concepts themselves that generate a contradiction (or so it is claimed). However, when we found out something about the way time works, it became apparent that there is no (metaphysically identical) possible world in which one can alter the past. Anyway, I take it this is supposed to be the argument, though as I said I'm somewhat suspicious the impossibility isn't merely conceptual. If you prefer, question whether Alicia can have a single reason to drink water but not drink H20.

    Finally, I think I'm willing to admit that we can have reasons even for conceptual impossibilities. All we have to do is generate the usual silly cases in which someone offers Alicia $1 million per year as long as she stays married to a bachelor. Or where someone offers her $1 million to drink water without drinking H20.

  10. Peter,

    Yes, sorry, I thought you meant something else with the wide-scope comment. As I said to Corwin, I should have done a better job of disambiguating. And Schroeder himself does, so the fault was mine.