No Pro Tanto Duty to Keep A Promise

I have a very short argument to the effect that the promising-relevant pro tanto moral obligation is not a pro tanto moral obligation to keep one's promises but, rather, a pro tanto moral obligation to not break one's promises.

[The argument crucially depends on the notion of a conditional release from a promise. Whereas an unconditional release (or a simple release) is a promisee's saying to a promisor something like "You don't have to phi", where phi-ing is something the promisor had promised the promisee to do, a conditional release is a promisee's saying to a promisor something like "If C, then you don't have to phi". I take it that both simple and conditional releases are altogether ordinary moral phenomena.]

Here's the argument: Suppose that on Monday A says to B "I promise to lend you my recording of Don Giovanni on Wednesday" and on Tuesday, B says to A "If you let me hold onto this copy of Cosi fan Tutte which you lent me the other day and were planning to take back today, then you don't have to lend me your Don Giovanni tomorrow." In this case, it seems that the following two courses of action are both permissible: (1) A takes back his Cosi fan Tutte on Tuesday and on Wednesday lends B his Don Giovanni, and (2) A does not take back his Cosi fan Tutte and on Wednesday does not lend B his Don Giovanni. However, option (1) involves the keeping of a promise and the breaking of no promises and option (2) involves neither the keeping nor the breaking of any promises. If there were a pro tanto moral obligation to keep a promise, as opposed to merely not breaking one, in this case (1) should be obligatory and (2) impermissible. If, on the other hand, the only promising-relevant pro tanto moral obligation were to not break one's promises, then, as neither involves the breaking of a promise, both courses of action would be permissible, which, I contend, is in fact the case. 

14 Replies to “No Pro Tanto Duty to Keep A Promise

  1. That seems right to me. What hangs on this, though? Are there those who insist that there is a pro tanto moral obligation to keep a promise, as opposed to merely not breaking one? And if so, what leads them to do so?

  2. Not sure. Maybe not a lot.
    Holly Smith, “A Paradox of Promising” (Phil Review 1997), argues that denying that there is any moral value to the keeping of a promise (or as I’d prefer to put it, there is no prima facie duty to keep a promise) and holding only that there is disvalue to (there is a prima facie duty against) the breaking of promises is necessary to avoid certain problems associated with the morality of promising. (Earl Conee, “The Moral Value in Promises”, (Phil Review 2000) disagrees.)
    Also there might be interest among those sympathetic to deontic pluralism (neo-Rossianism) about how exactly to understand the fundamental pro tanto moral obligation vis-a-vis promising. This might just be a small part of working out how all the deontic elements of such a pluralism work together. For instance, Ned Markosian in “Rossian Minimalism” (JESP 2009) argues that the best way of carrying out the neo-Rossian program is to hold that an action is right iff it minimizes pf duty violations instead of maximizing pf duty utility or anything like that. Getting a little clearer on what the pf duties are will be important, I think, to carrying out any such program. (Ross, in listing his prima facie duties, I believe, presents the prima facie duty of fidelity as a matter of keeping one’s promises.)
    Last, some have argued that negative prima facie duties (duties not to do things) are more stringent, in some sense (ack! don’t ask me to try to say exactly how), than positive duties (duties to do things). (I believe that this was Philippa Foot’s view. (She tried to appeal to that difference to explain the Trolley Problem. Thomson showed that that explanation won’t work (even though Thomson has subsequently disowned her refutation of Foot and now accepts the Foot solution to the Trolley Problem by way of an appeal to the distinction in stringency between positive and negative duties).)) I’ve even (in print ) flirted with the idea that negative duties are not subject to an ability constraint whereas positive ones are. Anyone who thinks there is some morally significant difference (whatever it might be) between positive and negative duties will have some stake, I’d think, in the question whether the prima facie duty vis-a-vis promising is to keep one’s promises or to not break them.

  3. The target is supposed to be the view that there is a pro tanto moral obligation to keep one’s promises. I think this means, for any promise of yours, P, you have an obligation to keep P.
    But then the point about the example goes like this:

    However, option (1) involves the keeping of a promise and the breaking of no promises and option (2) involves neither the keeping nor the breaking of any promises. If there were a pro tanto moral obligation to keep a promise, as opposed to merely not breaking one, in this case (1) should be obligatory and (2) impermissible.

    This seems to be an argument against a different view, namely, the view that we are obligated to perform a certain action instead of an alternative if the action keeps a promise and the alternative doesn’t. This is different because whether there is any promise to be kept may depend on the choice itself… as it does in the example.
    So, as I see it, if A lets B hold on to the Cosi, there is no promise for A to keep or break, whereas if A demands the return of the Cosi, there is a promise for A to keep or break. The keep-a-promise view does not imply that A has an obligation to perform the action that will result in the keeping of a promise, because that action also determines that there is a promise.
    It’s confusing. I wonder if adding time indices to the obligations would help clarify.

  4. I think you are quite correct that we can absolve people of promises they have made to us and we can do so conditionally. If you let him keep the Cosi your obligation to return the Don G goes away.
    I think what is perplexing you here is the problem of how this conditional absolution can make doing something that was wrong (not doing what you promised to do) morally okay.
    My view is that the mistake here is in thinking that promising to do something ever makes it wrong to fail to do it.
    Promising X that you will A entails that you have an obligation to X to do A but that does not entail that it would be wrong for you to refrain from doing A only that it would be permissible for him to force you to A.
    This does assume the Retributive Theory of promising, which treats promising as a conditional permission to harm. But then, so far as I know, there are no other theories of promising on offer.

  5. Hi Jamie,
    You write:
    “So, as I see it, if A lets B hold on to the Cosi, there is no promise for A to keep or break, whereas if A demands the return of the Cosi, there is a promise for A to keep or break. The keep-a-promise view does not imply that A has an obligation to perform the action that will result in the keeping of a promise, because that action also determines that there is a promise.”
    I think I disagree. I think I’d prefer not to reify promises. (Or, if I were reifying promises I’d say that there is a promise just in case someone promised someone else to do something. And on such a reading there is a promise in the case no matter what A ends up doing. I’d say that the promise comes into existence on Monday with A’s utterance.) I took the question of what the promising-relevant prima facie moral obligation is to be a question of what conditions give rise to there being a prima facie moral obligation to do or not to do something in virtue of (past) promising-activity. So, the theory I took it I was arguing against is one that held that the fact that, in a certain situation, A’s phi-ing would count as A’s doing that which A promised B to do (and from which A had not been released by B) is a sufficient condition for its being the case that A has a promising-relevant prima facie duty to phi. I was thinking that the argument showed that to be false.
    I think maybe I should have simplified things along the temporal dimension by making the time of the promised action identical to that of the condition of the release. Altering the case in this way would just be a matter of B’s saying on Wednesday just before A is about to lend him the DG: “If you let me hold onto this copy of Cosi fan Tutte which you lent me the other day and were planning to take back today, then you don’t have to lend me your Don Giovanni.” In response to this, though, you might want to take a similar line to that which you suggest above, just that in this case in option (1) the promise comes into existence at the moment of its fulfillment and in option (2) no promise ever comes into existence at all. So I guess I just have to rest with my response in the previous paragraph above.

  6. FWIW, one going account, Scanlon’s, has the obligation include a no-releases clause. The relevant part of his Principle F (which covers more than promising, but explains our obligations with respect to promising) is:
    . . . in the absence of special justification, A must do X unless B consents to X’s not being done. – Scanlon, (1990) p. 208
    I’m not sure what turns on this, but it seems worth noting that some versions of the duty to keep the promise view include the release exception.

  7. Hi Mark,
    Thanks for this. I meant to be attacking a view that builds a release clause into the pf obligation. I wanted to use a conditional-release case because with conditional release, on some options the promise is unreleased and in others it is released (by engaging the condition of the conditional release). However, I guess the Scanlon proposal could be amended as follows:
    …in the absence of special justification, A must do X unless B consents to X’s not being done, or B consents to X’s not being done on condition C and C obtains.
    (Like releases from promises, consent can, of course, come in conditionalized form.)
    Perhaps that’s a fine amendment, in which case we may have an account of the promising-relevant pf obligation as a positive duty that is immune to the argument I offered. I’ll have to think whether there might be any problems for it. But perhaps there is one thing going for the negative-duty account over this positive-duty account–it seems simpler.

  8. Hi Peter,
    I guess on my way of doing books, I think that B consents when B consents to X’s not being done on condition C, if condition C is met, so I did not think the extra clause is needed to cover conditional consent. But I don’t think I’d object to including it explicitly either. But once I say that I see the reason why Jamie thinks time-indexing might be helpful, since you might think that the person’s obligations change at the time condition C is met.
    I’ve got to admit to not being entirely sure what turns on the overall issue or on whether we add a specific clause for conditional releases as opposed to understanding the Scanlon-style release clause to include those when the condition is discharged. I may have a tendency to think of some of these positions as involving different ways of saying essentially the same thing, rather than as disagreements over substantive matters. But I expect that given the right further theoretical assumptions they might become issues of substance.

  9. Hi Mark,
    Thanks again for this. I am sympathetic to what you say in the second paragraph; I’m open to this all just being much of a muchness. Though, as you say, with additional theoretical assumptions, how the pf obligation is characterized (either as a positive duty or as a negative duty) may become philosophically significant.

  10. Hi Mark,
    I’m sorry for being muddle-headed yesterday. I should have been a bit more clear about my target. The view against which I was arguing was that which accepts, among others, the following theses:
    A. What one is morally obliged to do in a situation is to take that of her options which has the highest ratio of prima facie rightness to prima facie wrongness.
    B. An action is prima facie right just in case it consists in the satisfying of a prima facie duty.
    C. An action is prima facie wrong just in case it consists in the violation of a prima facie duty.
    D. An action satisfies a prima facie duty just in case the action consists in doing what the prima facie duty says to do when the conditions in which it says to do it are satisfied.
    E. The prima facie duty with respect to promising is as follows: If phi-ing would be doing what you promised someone you would do and you have not been released from doing it, then phi.
    This picture, or something like it, is, I believe, at play in many Rossian and neo-Rossian accounts of moral philosophy. As far as I can tell the Scanlon view on promising may or may not be consistent with this picture. I’m not sure because the passage you quoted doesn’t talk of prima facie duties at all. (I should also not have said that the revised Scanlon view I suggested is immune to my argument, for I’m not sure whether the revised Scanlon view is consistent with this picture against which I was arguing.) To make things less confusing temporally, I should also have made the time of the condition on which B’s release was conditioned simultaneous with the phi-ing that A promised B to do (in the manner I suggested above in my reply to Jamie). Suppose I had presented it that way. Then, given that B, D, and E of the above picture are true, it seems that option (1) has some prima facie rightness and no prima facie wrongness (in option (1), A does keep a promise from which he has not been released (for in option (1) though B has offered a conditional release, the condition is never satisfied)) and option (2) has neither any prima facie rightness nor any prima facie wrongness. And thus, given A from the picture above, it should follow that option (1) is obligatory in the case. However, intuitively, option (1) is not obligatory in that case.
    This is all by way clarifying what I meant to say. Sorry for being confusing about all of this.

  11. Let me try one more time.
    You promise someone you will do X, and she relies on you to keep your promise. And you know that if you fail to keep your promise she will be harmed because she is counting on you to do X. Surely then it would be wrong for you not to do X because, other things equal, it is wrong to knowingly lead other people to harm.
    But suppose that instead of promising the other person that you will do X, you deliberately lead her to believe that you are going to do X by some other means (it is not hard to tell stories like this). And suppose you know that she is counting on you to do X. Now if you don’t do X you will still be doing something wrong won’t you? And won’t it be wrong for the same reason as before: other things equal, it is wrong to knowingly lead other people to harm?
    Is causing harm in the first way somehow different than harming people in the second way because the words, “I promise” are involved? Don’t you always have a prima facie duty not cause other people harm? Does my duty not to gratuitously harm you some how get augmented or added to if I add “And, moreover, I promise not to harm you”?
    In your story it is clear that B is not going to be harmed by A’s not doing what was initially promised because, after B gets the Cosi, B is no longer expecting A to do what was initially promised, so won’t be harmed by this failure. No harm, no foul. That is, since there was no harm done, nothing morally wrong has been done.
    This will only be mysterious if we think that promising itself some how magically endows the act itself with a special kind of promise breaking badness that needs to dispelled in some way (e.g. by somehow some sort of built in “release-clause” which, when triggered, makes the special promising juju go away).
    Isn’t this just a reductio of a mistaken theory of promising: one which supposes that promising to do X somehow magically makes it wrong to do X.

  12. Hi tomkow
    I wasn’t here trying to get drawn into a sustained defense of there being a promising-relevant pro tanto moral obligation. To do so would involve a lot more argument than there is space for here. Also, it would take a fair amount of time to deal with all of the strands of your argument. So, instead, simply please understand my argument as implicitly qualified by a “given that there is a promising-relevant pro tanto moral obligation…”
    That said, I should note that:
    (1) I don’t think that all of the moral force of promises can be accounted for by your other principles. I’m obliged to keep my promises even if my promisee doesn’t rely on my promise (say, because she knows I’m a horrible louse).
    (2) Even if my breaking my promise would cause you no harm, I’m still morally obliged (other things being equal of course) to do what I promised you to do.
    (3) I act wrongly if my promise breaking doesn’t harm you but instead deprives you of a benefit you would have had had I not broken my promise. But ordinarily, though I ought not to harm you, it is not the case that I ought to benefit you.
    (4) I’m open to the suggestion that word-giving and promising are pretty much the very same thing morally speaking and so are governed by the same moral principles (though of even this I’m not very sure (the very act of binding oneself may indeed be special)), I certainly do not think the morality of these phenomena is to be accounted for simply in terms of harm.

  13. I agree that A’s Promising B to do X gives A moral obligation to B to do X. I take that to mean that it would be morally permissible for B to force A to do X , or enlist others to help her force A to do X, if needs must.
    And I take it that it means that A will be morally obliged to compensate B for any harm B suffers as a result of A’s not doing X.
    And I agree these obligations don’t go away even if B doesn’t really expect A to do X. It would still be permissible, if she felt like it, to force A to do X whether on not she would suffer from A’s not doing X. And she would still be entitled to compensation for harms done, even if she did not anticipate them.
    But none of this is entails that it is morally wrong, even prima facie, for A to refrain from doing X.

  14. Hi Pete,
    You didn’t seem muddle-headed yesterday, but this elaboration does make your target clearer. It looks to me like the culprit here may be the view in A, or (perhaps more accurately) combining A with the rest of the claims given the way conditional releases can work insofar as they can remain an option up til the time of performance. That’s a very quick reaction and I may be missing something.

Comments are closed.