Maverick Philosopher

Footnotes to Plato from the foothills of the Superstition Mountains

On a Putative Counterexample to ‘Ought Implies Can‘

I have long subscribed to Kant's famous meta-ethical principle according to which our moral obligations cannot outrun our abilities. 'Ought' implies 'can.' If I am under a moral obligation to do X, then I must be able to do X. We are concerned here with moral not legal oughts, and we understand 'ought' in accordance with the principle that if one morally ought to do X, then one is morally obliged/obligated to do X. 

Roughly, if you ought to do something, then it must be possible for you to do it, not just logically, and not just nomologically; it must be possible for you to do it given your actual abilities at a particular time and in definite circumstances.  With a bit more precision:

OC. Necessarily, if agent A ought to do X at time t in circumstances C, then A is able to do X at t and in C.

So if I ought to come to your aid, then I am able to do so.  By contraposition, if I am unable to come to your aid, then it is not the case that I ought to, and I am not subject to moral censure if I fail to. 

Kant moralityNote the logical difference between 'It is not the case that A ought to do X' and 'A ought not to do X.'  To confuse those two would be to commit an operator shift fallacy by importing the negation operator into the negatum.  So the contrapositive of 'ought' implies 'can' is not 'cannot' implies 'ought not,' but 'cannot' implies 'not ought.'  Better still: 'not can' implies 'not ought.'

Now suppose I promise to drive you to the airport at six in the morning. So promising, I morally obligate myself to so doing, i.e., I ought to drive you to the airport at six.  It follows by (OC) that I can drive you to the airport in a very concrete sense of 'can,': I know how to drive; I know how to get to the airport; I have access to a car, no one is preventing me from driving, etc.  Obviously, a carjacking would absolve me of my moral obligation.

My ability in this concrete and specific sense is a necessary condition of my being morally obligated to drive you to the airport.

Putative Counterexample 

Suppose that the night before the airport run I get drunk, sleep through the alarm, wake up late and hungover, and forget to fill up the gas tank in my vehicle.  As a result we run out of gas and you miss your flight.  I am unable to deliver on my promise, and do what the promise obligated me to do, but it seems that I am nonetheless morally responsible and indeed open to moral censure. In this case it seems that 'not can'' does not imply 'not ought.'  It seems that my inability to get you to the airport on time does not absolve me of my moral obligation to perform than very action. For I did something blameworthy by getting drunk the night before.

I am not impressed by counterexamples of this sort.  Touching only the letter, but not the spirit of Kant's great principle, they merely invite a reformulation thereof.  To wit,

OC*. Necessarily, if agent A ought to do X at time t in circumstances C, then A is able to do X at t and in C subject to the proviso that around t and in C A has not done anything to impair his abilities or factors contributing to his abilities.

Posted

in

,

by

Bill Vallicella

Tags:

Comments

11 responses to “On a Putative Counterexample to ‘Ought Implies Can‘”

  1. Wesley Avatar
    Wesley

    I still haven’t had time to carefully look at your divine simplicity post. Having read this post, however, I thought I would ask what you think of the following:
    If A ought to do X at t, then either A is able to do X at t or, if A cannot do X, A is responsible for its being the case that A cannot do X.
    Think of it as a tracing condition, one that is similar to the one you gave. The *can* traces to the act in question, or to some past act of omission or commission. I might not be able to avoid hitting the house when I am inebriated, but I was able to avoid being inebriated. My culpability traces back into the past.

    Reply
  2. BV Avatar
    BV

    >>If A ought to do X at t, then either A is able to do X at t or, if A cannot do X, A is responsible for its being the case that A cannot do X.<< This doesn't hold in general. You ought to feed your kids. If so, you are able to. But on a particular occasion you are abducted against your will and cannot do your duty. This absolves you of your duty. It cannot be said that you are responsible for its being the case that you cannot feed your kids.

    Reply
  3. Angra Mainyu Avatar
    Angra Mainyu

    Hi Bill,
    I think a key question is: did the person who made the promise (say, A) incur moral blame (related to the matter at hand, of course) after he was already incapable of taking the other person (say, B) to the airport on time?
    If the answer is negative, then it seems to me this isn’t a counter example to the (proposed) principle. I think the answer is negative, because in the scenario, A’s immoral behavior consists in:
    1. Getting drunk the previous night. He should have known that there was a significant risk that he would not be able to deliver his friend to the airport (apart from other potential reasons this might be wrong).
    2. Driving while hungover. A does not know what the level of alcohol is, so based on the information available to him, driving might or might not be a crime. But he knows or should know there is a risk that it is. More importantly, aside from the morality of risking breaking the law, he knows or should know it’s dangerous: even if he hasn’t read any studies on the subject, he surely can feel the hangover.
    So, it seems to me once A woke up hungover, A did not have a moral obligation to drive B to the airport, but rather, a moral obligation to refrain from driving until he felt better. He also had a moral obligation to call B to apologize and tell him that he was in no condition to drive safely, due to his own wrongful behavior the previous night. We could modify the scenario and say it happens 20 years into the future, and A has a safe, fully autonomous car. But then, the car would let him know that there is insufficient fuel – or much more likely, that the batteries need to be recharged -, so that’s not going to work, either. There are plenty of other potential variants, but I haven’t found one that works.
    Granted, it might look as though this is a way to block the (purported) counterexample without addressing the crux of it, but at least in the cases of the purported counterexamples I’ve encountered, there is always one factor or another that makes the counterexample fail, so there might be something to that. In any case, I think the question I mentioned in the beginning would very likely be an effective way of testing any such scenarios.

    Reply
  4. Wesley Avatar
    Wesley

    Bill,
    I have just one follow up, or rather pushback. I don’t see how your example is counter to the principle I gave. You are using ‘ought’ in a general way; I was not so using it. That is why I specified t (at a time). Suppose someone gave the following line of reasoning.
    1. You ought to feed your kids.
    2. This week you were unable to feed your kids because you were abducted on Sunday and held through Saturday against your will.
    3. Whenever you fail to do what you ought to do, you do something wrong.
    4. Therefore, you did wrong this week by failing to feed your kids.
    Something is wrong, for sure, but it would seem be with the line of reasoning. I think the first premiss is the culprit. It is general, but the person in question, so far as the premisses imply, has not failed in general to feed his kids; he has failed to feed his kids at some particular time (or for some particular time span) span. However, when you adjust that premiss to render it specific in the relevant way, it no longer seems true, given premiss two. The principle I gave is supposed to shed light on why this is the case.

    Reply
  5. BV Avatar
    BV

    Wesley,
    First tell me whether you think the putative CE I gave is a genuine CE to Kant’s principle.

    Reply
  6. Wesley Avatar
    Wesley

    Bill,
    Yes, I think it is genuine.

    Reply
  7. BV Avatar
    BV

    Well then, that’s where we differ. Thank for your comments.

    Reply
  8. Angra Mainyu Avatar
    Angra Mainyu

    Bill,
    Here’s a modification that avoids the hangover issue: A just forgets to set his alarm clock, and oversleeps. The tank is full. Does he have a moral obligation to take B to the airport?
    I would consider 2 different (sub)scenarios, and my assessment of them:
    S1: A wakes up 1 hour before the flight. He knows that B has to be at the airport 1 hour before the flight to check in, and the trip would take roughly 2 hours.
    In this case, A does not have any obligations to take B to the airport. He already did something wrong – he was negligent in not setting his alarm clock -, and he should apologize. But there is no further wrongdoing in not taking B to the airport.
    S2: A wakes up 3 hours before the flight. He knows that B has to be at the airport 1 hour before the flight to check in, and the trip would take roughly 2 hours. A reckons that depending on traffic, they have about a 50% chance of making it on time, and his assessment is epistemically rational given the information available to him. In this case, A does have an obligation to try to take B to the airport. However, it’s not the case that he incurs further blame if due to traffic (normal for that hour, though it might have been less) they don’t make it. In fact, as long as A does his best to get there, he is only to blame for forgetting to set the alarm clock the previous night, and nothing more. Now, B is human, so there is a good chance he would be more upset and more angry with A if they don’t make it than if they do, but that isn’t because A’s behavior was more immoral in one case than in the other. Rather, that’s because humans do tend to care more about wrongful behavior when they or those close to them suffer negative consequences.
    So, I think those variants also fail to present a successful counterexample.
    Side note: Some (maybe many) people believe that sometimes, immoral behavior is not blameworthy but excusable. I don’t believe this is so – I believe that necessarily, a behavior that is blameworthy if and only if it’s immoral. But on the view that sometimes behavior is immoral but not blameworthy, it’s hard to test whether ought implies can. For example, as far as I know, even a carjacking might not absolve A of a moral obligation, but only make him not blameworthy if he fails to meet his obligation. Under the assumption that there is immoral behavior that is not blameworthy, I think the principle one could test would be whether blameworthiness in case of failure implies can (i.e., the principle analogous to OC, but using blameworthiness instead of “ought”).

    Reply
  9. The Dark Ostrich Avatar
    The Dark Ostrich

    >>if agent A ought to do X at time t in circumstances C
    Here t is the time that A fails to turn up, right?
    What about ‘One should not make rash/reckless promises’?
    Suppose A knows pretty well that he goes on frequent benders and is liable not to turn up at promised time. So he ought not to make promises to turn up, with his acknowledged weakness, and t is the time of making the promise.
    This contrasts with the case where A rarely goes on a bender, and so is not making a reckless promise. But suppose he meets an old drinking buddy from olden days. They go out for ‘just one’ which turns into ‘just many’, as occasionally happens.
    Clearly the breach occurs as a result of the drinking bout. But there is still a problem. Do we blame A for ‘one/a few for the road’? But we all know (I assume) there is that point where responsible action ceases. After 8 beers, A sincerely believes, in an intoxicated way, that he will keep the promise. All seems rosy. In the morning, with a thundering hangover, it is very different.
    So, do we blame A in the latter case?

    Reply
  10. BV Avatar
    BV

    Ostrich,
    Yes, to first question.
    My example does not assume that A goes on frequent benders. Suppose A, who is sobriety itself normally, is an agreeable and easily influenced fellow who is collared by some old school chums that A hasn’t seen in ages and they haul him into a pub where they all get soused for old times’ sake.
    Having written this, I now see that this is your second case.
    Yes, we blame A in the latter case.
    The question, however, is whether this is a CE to Kant’s principle.
    What do you say about this?
    I say it is not a genuine CE. A promised to get his friend to the airport on time, and so he ought to do it. And he can do it. It is irrelevant that he disabled himself by getting drunk.
    Kant’s point is that what one is morally obliged to do one must be able to do. A is able to get the guy to the airport on time and able to keep himself sober for the job.
    So I say that the putative CE is not a genuine CE to the principle, but to an inadequate formulation of the principle.

    Reply
  11. The Dark Ostrich Avatar
    The Dark Ostrich

    >>Kant’s point is that what one is morally obliged to do one must be able to do.
    Agree. However there is a difference between the blame we attach to breaking a reckless promise, which we attribute to foolishness, youth or ignorance. If A made a non-reckless promise, i.e. one made with complete self-knowledge, then either (i) A knows his capacity for alcohol will not intefere with the promise, in which he goes on the bender or (ii) he knows he lacks the capacity, in which case he ought not go on the bender.
    An in-between case, they start drinking absinthe or some strong German beer that has an unanticipated effect. But even in that case, A should be self-aware enough to avoid unfamiliar drinking brands.
    In summary, neither case (reckless or non-reckless promise) is an exception, however we should attach less blame to the reckless promise (we generally don’t blame foolish or young people, but try and protect them).

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *