Portal de Periódicos da CAPES

“The Thing To Do” Implies “Can”

2013; Wiley; Volume: 50; Issue: 1 Linguagem: Inglês

10.1111/nous.12037

1468-0068

Nicholas Southwood,

Philosophical Ethics and Theory

Countless interpretations of (OC) have been proposed to respond to this familiar complaint. Some have focused on the "ought" (e.g. Streumer 2007; Vranas 2007), some on the "can" (e.g. McConnell 1989), some on the "implies" (e.g. Hare 1963, pp. 53–4; Sinnott-Armstrong 1984), some on the "ϕ" (e.g. Baltzly 2000; Brennan and Southwood 2007), and some on the "t" (e.g. Howard-Snyder 2006). I suspect that all these interpretations either a) fail to block the complaint in a plausible and non-ad hoc way or b) throw out the baby with the bathwater,2 though I won't make any attempt to defend this sweeping claim here. I shall argue that (TDC) is entailed by important but underappreciated truths about practical reason, and that it is not vulnerable to the familiar complaint. This is interesting in its own right. Moreover, I shall suggest that it has interesting (though by no means straightforward) implications for (OC). Precisely what it implies for (OC) depends on the relation between claims about what we ought to do and claims about the thing to do—in particular, whether the former imply the latter. If they do, then the argument for (TDC) that I shall offer also suggests a novel argument for (OC) and a novel response to the familiar complaint. If they don't, then (TDC) naturally suggests a novel error theory of (OC). To explicate (TDC), we must say something about the core notion of a "claim about the thing to do." I am going to assume that there are such claims, and that they are (or are made true by) claims about the correct answer to the question of what to do. There are two important ideas here that require some elaboration: a) the idea of the question what to do, and b) the idea of a correct answer to the question of what to do. Let us briefly consider each in turn. The question of what to do is the question that confronts us in practical deliberation. When we engage in practical deliberation we ask ourselves and try to resolve the question of what to do: whether to go on holiday to Corsica or Sicily or Sardinia; whether to send one's children to a private or a public school; whether to have a sixth pain au chocolat; and so on. Following Pamela Hieronymi (2009; 2011), I propose that the question of what to do is not the same as the question of what one ought to do.3 One way to bring this out is to notice that we resolve the question of what we ought to do by forming a normative belief, whereas we resolve the question of what to do by forming an intention or decision (see Hieronymi 2009; Owens 2011). For example, I might resolve the question of whether I ought to have a sixth pain au chocolat by forming the belief that I ought not to. This doesn't yet mean that I have resolved the question of whether to have a sixth pain au chocolat. To resolve the question of whether to have a sixth pain au chocolat would be to decide to have a sixth pain au chocolat or to decide not to.4 Another way to bring out the contrast between the question of what to do and the question of what one ought to do is to note that it seems that we can ask ourselves and try to resolve the question of what we ought to do without asking ourselves and trying to resolve the question of what to do. This may happen, for example, if we have already resolved to our satisfaction the question of what to do. Suppose that I have received a substantial Christmas bonus and intend to spend it on a new car. This does not rule out asking myself the question of what I ought to do with my bonus—say, whether I ought to give it to Oxfam. In asking myself the question of whether I ought to give the money to Oxfam, I needn't be reopening the question of what to do with my bonus. Furthermore, it seems that we may also ask ourselves and try to resolve the question of what to do without asking ourselves and trying to resolve the question of what we ought to do. Again, this may happen if we have already resolved the question of what we ought to do. Perhaps I have already resolved the question of whether I ought to have a sixth pain au chocolat, yet the question of whether to have a sixth pain au chocolat remains open. Or it may happen if one is a consistent normative nihilist who holds that there are no truths about what one ought to do and who is consistent inasmuch as one refrains from having any beliefs about, or even interrogating the question of, what one ought to do. Being a consistent normative nihilist of this kind obviously doesn't mean that one is unable to interrogate the question of, say, whether to have a sixth pain au chocolat or whether to go on holiday to Corsica or Sicily or Sardinia. Claims about the thing to do are (or are made true by) claims about the correct answer to the question of what to do. What does it take in order for an answer to the question of what to do to be correct? On the assumption that I ought not to have a sixth pain au chocolat, (1) implies that if I answer the question of whether to have a sixth pain au chocolat by deciding to have one, then I have incorrectly resolved the question of what to do. Nothing that I shall say in what follows is inconsistent with (1).5 Suppose that one resolves the question of whether to send one's children to a public or a private school by deciding to send them to a private school, and that one's decision is based on the belief that sending them to a private school will give them an advantage over others later in life. Perhaps this belief is false. Perhaps sending them to a public school would in fact confer upon them the desired advantage. Suppose, however, that on the false assumption that it is true, it follows that one ought to send them to a private school—because it is true that one ought to do what one will give one's children an advantage over others later in life. Under these circumstances, deciding to send them to a private school may count as the correct answer to the question of what to do, even if it is not true that one ought to send them to a private school. Alternatively, suppose that one ought not to do what one will give one's children an advantage over others later in life. (Perhaps doing so is contrary to some valid egalitarian principle of justice.) Under these circumstances, deciding to send them to a private school amounts to incorrectly resolving the question of what to do—since, assuming that one's (false) belief that sending them to a private school will give them an advantage over others later in life is true, it follows that one ought not to send them to a private school. This is so, irrespective of whether, in fact, one ought to send to send them to a public or a private school. I won't try to resolve the question of which (if either) of these accounts of correctness is the right one. It suffices to note that what I shall say in what follows is consistent with a wide variety of views about the correctness of answers to the question of what to do. I now want to suggest two important correctness constraints. Notice that an answer to the question of what to do may fail to be correct in two ways. First, it may be the case that some rival answer is the correct one. This is the kind of failure that happens, for example, if one resolves the question of whether to have the sixth pain au chocolat by deciding to have it when, in fact, the correct answer to the question was to decide not to have it. Call this "the correctness of answer constraint." What kinds of acts are such that it is not correct for an agent to ask herself the question of whether to perform them? Take past acts, such as my kissing my wife yesterday morning. I cannot correctly ask myself the question of whether to kiss my wife yesterday morning. That involves a kind of mistake. Past acts are beyond the jurisdiction of an agent's practical deliberation. (I can ask myself the question of what past acts I ought to have done, but not what past acts to do.) Or take others' acts, such as your buying a new kayak. I cannot correctly ask myself the question of whether you are to buy a new kayak (absent some special authorization). Again, that would involve a mistake. Others' acts lie beyond the jurisdiction of an agent's practical deliberation. (I can certainly ask myself the question of what others ought to do, or what they will do, but not what to do for others, as it were.) Or take acts that we lack the ability to perform. If I don't speak Russian, then it would be incorrect for me to ask myself the question of whether to have a conversation in Russian with my next-door neighbor about the finer points of Dostoyevsky's pre-Siberian novels. If I don't play the piano, then it would be incorrect for me to ask myself the question of whether to play Ravel's "Gaspard de la nuit" this evening. Or take acts that we have the ability to perform in general but where our ability is impaired or blocked by external obstacles. I may have the ability to keep promises in general. (I am not subject to any kind of promissory pathology.) Yet it may be that I cannot keep my promise to be at your wedding in Dunedin, because this would require me to take a flight from Australia, and all flights out of Australia have been grounded due to severe tropical cyclones. Under these circumstances, it would be incorrect for me to ask myself the question of whether to attend your wedding. Call this "the correctness of question constraint." We are now in a position to give the argument for (TDC). What this means is that if one cannot perform a particular act, then deciding to perform that act cannot be the correct answer to the question of what to do. Thus, for example, deciding (now) to kiss my wife yesterday, deciding that you are to buy a new kayak, deciding to have a literary chat in Russian, deciding to play "Gaspard de la nuit" this evening, and deciding to attend one's friend's wedding in New Zealand cannot be correct answers to the question of what to do. (5) and (6) imply (TDC). Therefore, (TDC) For any agent a∈A, at time t, ϕing is the thing for a to do only if, at t, a can ϕ (from (5) and (6)). This argument establishes the truth of a non-trivial claim, namely, (TDC). It also helps to see why (TDC) is not vulnerable to the familiar complaint, namely, that it licenses impermissible inferences in situations where agents intentionally make it the case that they are unable to perform certain acts. This inference can be seen to be perfectly unproblematic, once we spell out in full the chain of inferences by which we are entitled to infer (8) from (7). Notice that this inference is perfectly unproblematic. Returning your prized first edition of Newton's Principia simply isn't available to me. It lies beyond the jurisdiction of my practical deliberation. It doesn't matter why it isn't available to me. The fact that the explanation happens to involve intentional wrongdoing on my part is neither here nor there. Again, this inference is perfectly unproblematic. Given that it would be incorrect even to ask myself the question of whether to return your first edition of Newton's Principia, clearly answering it by deciding to do so cannot be the correct answer to the question of what to do. This inference is also perfectly unproblematic. Claims about the thing to do just are (or at least are made true by) claims about the correct answer to the question of what to do. So, the inference that (TDC) licenses from (7) to (8) is underwritten by a chain comprising three inferences: a) the inference licensed by the correctness of question constraint from (7) to (9); b) the inference licensed by the correctness of answer constraint from (9) to (10); and c) the inference licensed by the definition of claims about the thing to do from (10) to (8). Each of these three inferences is perfectly unobjectionable. So too, therefore, is the inference from (7) to (8). This is why (TDC) is not vulnerable to the familiar complaint. I have argued that (TDC) follows from certain important truths about practical reason and that it is not vulnerable to the familiar complaint. This is interesting in its own right. But we might also wonder: Where does this leave us with respect to (OC)? Some philosophers want to tie ought very closely to practical deliberation (see Kant 1998; Korsgaard 1996). For such philosophers, (11) may appear very natural, even truistic. Other philosophers insist upon sharply distinguishing ought and practical deliberation and are likely to be hostile to (11) (see Broome 2013). For my part, I am not going to take a stand on the issue of whether (11) is true or false.7 This is partly because I am genuinely unsure about whether (11) is true or false.8 More importantly, it doesn't matter for my purposes here. Either way, our argument for (TDC) has interesting implications for (OC). (TDC) For any agent a∈A, at time t, ϕing is the thing for a to do only if, at t, a can ϕ. Therefore, (OC) For any agent a∈A, at time t, a ought to ϕ only if, at t, a can ϕ. So we have a new argument for (OC)—or at least a new argument for a version of (OC) that is restricted to actions.9 This is interesting. If (11) is true, it also means that (OC) is not vulnerable to the familiar complaint. The familiar complaint holds that (OC) licenses impermissible inferences in situations where we wrongfully make it the case that we cannot perform certain acts. We saw that (TDC) is not vulnerable to the familiar complaint, since the apparently objectionable inferences licensed by (TDC) can be seen to be perfectly unobjectionable once we spell out the complete chains of inference that underwrite the apparently objectionable inferences. If (11) is true, then the apparently objectionable inferences licensed by (OC) will be underwritten by the same chains with an extra link including those licensed by (11). So we also have a new and interesting response to the familiar complaint to (OC).10 Therefore, (OC) For any agent a∈A, at time t, a ought to ϕ only if, at t, a can ϕ (from (5*) and (6*)). I am skeptical, however, that this analogous argument will end up doing any serious dialectical work. Take (4*). (4*) implies that it is incorrect to ask oneself the question of whether one ought to become a better person than one is capable of becoming. This is not obviously correct. I suspect, however, that insofar as (4*) strikes us as plausible, its plausibility is due to the evident truth of (4) proper (it is clearly incorrect to ask oneself the question of whether to become a better person than one is capable of becoming); and one's implicitly accepting the idea that there is a connection between the question of what we ought to do and the question of what to do, such that we ask and answer the former correctly only if we ask and answer the latter correctly. If we don't accept that there is this connection, then it is far from obvious why we should accept (4*). On the other hand, if we do accept that there is this connection, then it's hard to see how one could do so without also accepting (11). And if one accepts (11), then the analogous argument for (OC) is redundant, since, as we saw, (OC) follows directly from (11) and (TDC). Still, there is a different way in which (TDC) might be relevant to (OC) even if (11) is false. Suppose that we are antecedently skeptical about (OC)—perhaps because of the familiar complaint, perhaps for some other reason. It would be nice to be able to explain why (OC) has nonetheless mistakenly struck so many people as compelling. In other words, it would be nice to be able to provide an error theory of (OC). Notice that, if (11) is false, (TDC) may help us to provide such an error theory. Even if (11) is false, it is presumably not obviously false. Indeed, it may seem true, even truistic. So here is how the error theory goes: We fall into error because we recognize implicitly the truth of (TDC) and we mistakenly conflate claims about what we ought to do and claims about the thing to do. The apparent truth of (OC) is parasitic on the truth of (TDC) coupled with the apparent truth of (11).