A NEW ARGUMENT FOR OUGHT-IMPLIES-CAN Peter B. M. Vranas vranas@wisc.edu University of Wisconsin-Madison Central APA, 21 February 2019 Suppose that you have promised to play in tomorrow’s basketball game. You are the best player in your team, and your teammates are relying on you. This morning, however, you slipped over on the ice and broke your leg. Do you still have an obligation to play in tomorrow’s game, although you cannot do so? It seems clear that you do not. The moment you broke your leg and thus you lost the ability to play in tomorrow’s game, you also lost the obligation to do so. You have no obligation to do something that you cannot do. This is the idea that “ought” implies “can”: you ought to do something, in the sense of having an obligation to do it, only if you can do it.

“OUGHT” IMPLIES “CAN” (OIC) By virtue of conceptual necessity, if an agent at a given time ought (i.e., has an objective, pro tanto obligation) to do something, then the agent at that time can (i.e., has both the ability and the opportunity to) do the thing. Obligations, abilities, and opportunities are indexed to times. OIC relates obligations at a given time with abilities and opportunities at the same time. More precisely, here is the ought-implies-can principle that I plan to defend: by virtue of conceptual necessity, if an agent at a given time ought to do something, in the sense of having an objective, pro tanto obligation to do something, then the agent at that time can do the thing, in the sense of having both the ability and the opportunity to do it. I call this principle ‘OIC’ for short. It is important to note that obligations, abilities, and opportunities are indexed to times. For example, before you broke your leg, you had an obligation to play in tomorrow’s basketball game, but after you break your leg, you no longer have such an obligation. Similarly, before you broke your leg, you had the ability to play in tomorrow’s game, but after you break your leg, you no longer have this ability. OIC relates obligations at a given time with abilities and opportunities at the same time.

“INFEASIBLE” IMPLIES “LOST” (INFIL) By virtue of conceptual necessity, if an agent at a given time becomes unable to obey an obligation that the agent has until that time, then the agent at that time loses the obligation. If INFIL is true, then losing the ability to obey an obligation is sufficient for losing the obligation. Then the only plausible explanation is that lacking the ability to obey an obligation is sufficient for lacking the obligation: OIC is true. So, to defend OIC, I will defend INFIL. My plan is to defend OIC by defending the following consequence of OIC: by virtue of conceptual necessity, if an agent at a given time becomes unable to obey an obligation that the agent has until that time, then the agent at that time loses the obligation. In other words, “infeasible” implies ”lost”: obligations that become infeasible at a given time are lost at that time. I call this principle ‘INFIL’ for short. We have already seen an example: your obligation to play in tomorrow’s basketball game becomes infeasible at the time you break your leg, and the obligation is lost at that time. If INFIL is true, the only plausible explanation of why INFIL is true is that OIC is true. Indeed, if INFIL is true, then losing the ability to obey an obligation is sufficient for losing the obligation. But then the only plausible explanation is that lacking the ability to obey an obligation is sufficient for lacking the obligation. Equivalently, having the ability to obey an obligation is necessary for having the obligation, so OIC is true. So, to defend OIC, in the rest of this talk I defend INFIL.

THE FUNERAL SCENARIO You are vacationing in Honolulu, and on Monday at 9pm you are informed that your father has died. The funeral is scheduled to take place in Boston on Tuesday starting at 2pm. You can make it to the funeral, but only if you rush to catch the last flight, which leaves at 11pm. Although you had solemnly promised your father that you would attend his funeral, you do not feel like interrupting your vacation, so you decide not to attend. At 9:15pm, a friend who works at a local blood donation center calls you and asks you to donate blood on Tuesday at 2pm. You promise her to do so. On Tuesday at 1:55pm, you are about to enter the blood donation center. To defend INFIL, consider the following scenario, the funeral scenario.

TWO OBLIGATIONS Obligation 1 (to attend your father’s funeral) is incompatible with and stronger than obligation 2 (to donate blood). So if you don’t lose obligation 1 when it becomes infeasible and thus you still have it on Tuesday at 1:55pm, then on Tuesday at 1:55pm it is all-things-considered (atc) obligatory for you to attend the funeral, and thus it is atc forbidden for you to donate blood―which is false. In the funeral scenario, there are two obligations in play. First, on Monday at 9pm you acquire an obligation to attend your father’s funeral in Boston on Tuesday at 2pm. Second, on Monday at 9:15pm you acquire an obligation to donate blood in Honolulu on Tuesday at 2pm. The two obligations are incompatible, and let’s assume that the first is stronger than the second. Then, assuming that there are no further obligations or other normatively relevant considerations, at any time at which you have the first obligation, it is all-things-considered obligatory for you to attend your father’s funeral, so it is all-things-considered forbidden for you to donate blood. So if you do not lose the obligation to attend the funeral when this obligation becomes infeasible, namely when you become unable to catch the last flight, and thus you still have this obligation on Tuesday at 1:55pm, then on Tuesday at 1:55pm it is all-things-considered forbidden for you to donate blood. But this is false: at 1:55pm, given that you cannot attend your father’s funeral but you can donate blood, you must keep your promise to donate blood.

THE WRONG-ADVICE ARGUMENT The upshot: Those who claim that obligations that become infeasible are not lost (and so reject INFIL) would give you wrong advice: they would advise you (at 1:55pm) not to donate blood. This is not the familiar point that a morality which rejects OIC can give useless advice (“attend your father’s funeral, although you cannot”). It is instead the novel point that a morality which rejects INFIL (and thus rejects OIC) can give wrong advice (“do not donate blood, although you can”). The upshot is that those who claim that obligations that become infeasible are not lost (and thus reject INFIL) would give you wrong advice: they would advise you (at 1:55pm) not to donate blood. I call this the wrong-advice argument for INFIL (and, by extension, for OIC). This argument does not make the familiar point that a morality which rejects OIC can give useless advice (“attend your father’s funeral, although you cannot”); the argument makes instead the novel point that a morality which rejects INFIL (and thus rejects OIC) can give wrong advice (“do not donate blood, although you can”). One might object that, strictly speaking, moralities do not give advice: it is people who do so. In reply, although I chose to talk about advice for the sake of vividness, my main point can be made without talking about advice: regardless of the merits of the advice “do not donate blood”, my main point is that it is false that on Tuesday at 1:55pm it is all-things-considered forbidden for you to donate blood. A strength of the wrong-advice argument is that, in the funeral scenario, it is your fault that you become unable to obey an obligation (since you do not rush to catch the last flight); such scenarios are commonly considered particularly damaging to OIC, but the wrong-advice argument uses the funeral scenario to support OIC. But this feature of the funeral scenario, namely your being at fault, is inessential. The wrong-advice argument also works with a modified funeral scenario in which it is not your fault that you become unable to obey the obligation to attend your father’s funeral: for example, you do rush to catch the last flight, but you miss the flight due to a traffic jam.

THE GIFT CARD SCENARIO You have received (1) an electronic gift card that expires at 5pm and (2) a sheet with two ten-digit codes. To redeem the gift card, you must visit a website and enter code 1, to get a dishwasher, or code 2, to get a computer. Your father has asked you to get the dishwasher, but your son more urgently needs the computer, so you promised your son you will get the computer. At 4:55pm, as you are about to visit the website, your cigarette burns a hole in the sheet with the codes, and code 2 becomes illegible. You have no way to retrieve code 2 by 5pm, so you can no longer get the computer; but you can still enter code 1 and get the dishwasher. Another strength of the wrong-advice argument is that the argument can be adapted to cases in which an agent becomes unable to obey an obligation due to epistemic (rather than physical) limitations. To see this, consider the following scenario, the gift card scenario.

THE WRONG-ADVICE ARGUMENT AGAIN If (contrary to INFIL) you still have, right after 4:55pm, the obligation to fulfil your promise to your son and get the computer, then it is all-things-considered forbidden for you to get the dishwasher. But this is false: given that you can no longer get the computer but you can still get the dishwasher, you must (or at least you may) fulfill your father’s request to get the dishwasher. In the gift card scenario, after code 2 becomes illegible, I grant that in one sense of “can” you can still enter code 2 on the website: you are still physically able to type any ten-digit code. But in the sense of “can” that I take to be relevant to OIC, you can no longer enter code 2 (and thus you can no longer get the computer) because you do not know (and you have no way to find out) which ten-digit code is code 2. If (contrary to INFIL) you still have, right after 4:55pm, the obligation to fulfil your promise to your son and get the computer, then it is all-things-considered forbidden for you to get the dishwasher. But this is false: given that you can no longer get the computer but you can still get the dishwasher, you must (or at least you may) fulfill your father’s request to get the dishwasher.

OBJECTION: OBJECTIVE OBLIGATIONS Objection: It is atc objectively (not subjectively) forbidden for you to get the dishwasher: it is atc objectively obligatory for you to get the computer. My reply: Then a morality of objective obligations (understood as unaffected by epistemic limitations) gives wrong advice (“do not get the dishwasher”). My point is not that such a morality gives useless advice (“get the computer, although you cannot”). Instead, such a morality gives wrong advice (“do not get the dishwasher, although you can”). One might object as follows. Although it is not all-things-considered subjectively forbidden for you to get the dishwasher, it is all-things-considered objectively forbidden, because it is all-things-considered objectively obligatory for you to get the computer: your objective obligations are unaffected by your epistemic limitations (or so the objection goes). If so, I reply, then a morality of objective obligations (understood as unaffected by epistemic limitations) gives you wrong advice: it advises you not to get the dishwasher, although the gift card would expire unredeemed as a result. My point is not that such a morality gives useless advice (“get the computer, although you cannot”); my point is instead that such a morality gives wrong advice (“do not get the dishwasher, although you can”). So the above objection fails: after 4:55pm, it is not all-things-considered objectively obligatory for you to get the computer.

OBJECTIVE OBLIGATIONS VS EPISTEMIC LIMITATIONS If OIC is true, then your objective obligations are affected by what you can do. What you can do is affected by your epistemic limitations (e.g., you cannot get the computer because you do not know which ten-digit code is code 2). Thus: If OIC is true, your objective obligations are affected by your epistemic limitations (e.g., you have no objective obligation to get the computer). It is common in the literature to understand objective obligations as unaffected by epistemic limitations. As a proponent of OIC, formulated in terms of objective obligations, I reject that understanding. If OIC is true, then your objective obligations are affected by what you can do. But what you can do is affected by your epistemic limitations: for example, you cannot get the computer because you do not know which ten-digit code is code 2. Therefore, if OIC is true, then your objective obligations are affected by your epistemic limitations: for example, you have no objective obligation to get the computer. This conclusion is plausible: there are convincing arguments in the literature against understanding objective obligations as unaffected by epistemic limitations, and the wrong-advice considerations that I adduced provide a further argument against such an understanding—in addition to providing an argument for INFIL, and thus for OIC. I conclude with the immodest assertion that, to my knowledge, the wrong-advice argument is the strongest argument in the literature in support of ought-implies-can.

REPLY TO TOMAS BOGARDUS PART 1 I agree that feasibility facts are normatively relevant; this is compatible with premise 7. Suggestion 1: After you miss the last flight, you still have the obligation to attend your father’s funeral, but this obligation is inert: it does not count towards determining your atc obligation. My reply: If infeasible obligations are inert, then (1) atc obligations are feasible, and (2) infeasi-ble obligations do not bind you (i.e., you do not need to do what they prescribe), so you do not have them (and thus OIC is true). I am grateful to Professor Bogardus for his interesting comments. To start with, let me clarify that I agree that feasibility facts are normatively relevant. This is compatible with what Professor Bogardus calls premise 7, which says that there are no other normatively relevant considerations, beyond those specified in the funeral scenario. I will address three suggestions that Professor Bogardus makes. The first suggestion is that, after you miss the last flight, you still have the obligation to attend your father’s funeral, but this obligation is what I call inert: it does not count towards determining your all-things-considered obligation, which is to donate blood. I have two points in reply. First, if infeasible obligations are inert, then only feasible obligations count towards determining your all-things-considered obligations, so all-things-considered obligations are feasible; in other words, “ought-implies-can” holds if “ought” is restricted to all-things-considered obligations. Second, if infeasible obligations are inert, then they do not bind you; in other words, you do not need to do what they prescribe. But then you do not have infeasible obligations, and this is precisely OIC. OIC is not the thesis that infeasible obligations do not exist: they might exist, as impersonal obligations, but OIC is the thesis that no agent has them, they do not bind any agent.

REPLY TO TOMAS BOGARDUS PART 2 Suggestion 2: Some legal obligations are infeasible. My reply: “Legally required” (i.e., required by a law) does not entail “legally obligatory” (i.e., required by a legitimate law). A law can require you to square the circle. Suggestion 3: Your infeasible obligations might guide you to apologize. My reply: To explain why an apology is in order, there is no need to say that you have an infeasible obligation: say instead that you rendered infeasible an obliga-tion that you previously had but no longer have. A second suggestion made by Professor Bogardus is that one may have infeasible legal obligations: if you have signed a legally binding contract to be at your father’s funeral, then, even after you miss the last flight, you still have a legal obligation to attend the funeral. I reply that “legally required”, namely “required by a law”, does not entail “legally obligatory”, namely “required by a legitimate law”. If a law requires you to keep a contract that you cannot keep, for example a contract to square the circle, then the law is not legitimate, and you have no corresponding legal obligation. Finally, a third suggestion made by Professor Bogardus, following an article by Brian Talbot, is that your infeasible obligations can provide guidance: they might guide you to apologize. I reply that to explain why an apology is in order we do not need to appeal to infeasible obligations: we can appeal instead to the fact that you rendered infeasible (by deciding not to catch the last flight) an obligation that you previously had but you no longer have. To conclude, let me thank again Professor Bogardus for his thoughtful comments.