Today’s post is adapted from an open-access review I wrote for Ethics.
Paul Schofield’s book opens with a Kantian lament—that “no part of morals has been more defectively treated than this of the duties to oneself” (1). By the end, I expect that many readers will agree that the complaint is outdated.
Duty to Self is the best book on the topic written this century. It is also the first since Kant’s Metaphysics of Morals to put forward an original theory. Schofield’s goal is not to pin down the details of what we owe to ourselves, nor does he try to derive them via airtight arguments from awesome moral premises—Kantian, utilitarian, or otherwise. The book is less a forced march and more a pleasant tour, as Schofield guides us through highlights from the history of duties to self and opens up promising lines for future inquiry.
The book’s main goal is to get readers to take duties to oneself seriously, to treat them as a central part of ethics rather than a peripheral puzzle. In this task, Duty to Self is a triumph. It is hard to read this book without getting excited about its questions. Those skeptical of its answers will still likely find them fascinating.
As one such fascinated skeptic, I’ll lay out what I take to be the big objections. But first, let me go through the central line of argument.
Schofield begins with the question of what it means to have a bona fide duty to oneself (chap. 1). First is the “duty” part. Duties to self are moral, giving us a “peremptory” command to comply (29) simply for their own sake (31). I cannot shirk moral duties willy-nilly, or just for the sake of a few goodies. Next is “to self.” The reason to comply with a duty to self is second-personal, as with any duty to anybody. The person to whom the duty is owed is the “you” to the agent’s “I,” the same dyadic predicament involved in addressing a demand, giving permission, or saying “Hey you.” There is no way to reduce this “second-personal” element, according to Schofield. The best we can do is to link it to a circle of related concepts. When I have a second-personal reason with respect to you, I owe you; not following the reason would wrong you, which means that I would violate your right against me (25)—and in the aftermath you would be justified not only in being upset but also in resenting me (34). So we can’t reduce the “to self ” in “duty to self,” but we can describe some upshots: self-wronging, self-resentment, and even rights against oneself.
The middle chapters (chaps. 2–4) take on the book’s fundamental question. Duties to self are moral duties—binding obligations—and yet, typically, when someone owesme something, I can release themat will. So how could I be obligated by a duty tomyself ? If I can waive away the duty whenever I wish, in what sense could it bind me?
Schofield’s answer is that waivable duties to self are not binding, and so would not be duties in any interesting sense (46–48). A true duty to self must be unwaivable.
Here’s an example to illustrate. I have a duty to myself not to smoke, since smoking will cause me a lot of problems later and gives me only a little pleasure now. If the duty is waivable, I can release myself at will, just as I can release others when they ask to get out of their promises to me. But then the “duty” to myself wouldn’t really be a duty. What makes it a duty is that I cannot simply allow myself to smoke when I wish. Even if I would now prefer the present pleasure, I may not choose it, because I cannot waive the duty not to smoke.
But why can’t I waive it? What could prevent me from waiving a duty to myself? Schofield’s answer—the most original idea in the book, and one of the most interesting ever said on the topic—is that I cannot waive the duty now because the pain it causes me comes later. To waive a duty, I must occupy the perspective of the right person at the right time: the perspective of the person who is affected when they are affected. But when I choose to smoke, I do not occupy the perspective of myself in the future, when I will suffer the noxious effects. So I cannot now release myself.
At this, one might wonder whether Schofield is assuming that the present “me” and future “me” are different persons, or different time slices. He makes no such assumption. His views are not driven by any such metaphysics of personal identity.
Instead, Schofield is drawing on concepts from moral psychology, especially from the work of Stephen Darwall. Darwall’s key idea, for Schofield’s purposes, is that waiving and demanding are actions done not just by a person to a person but by a person from a perspective to a person at a perspective.
What’s a perspective? The analogy is with vision (60). Just as a coin might look like a circle from one angle and an oval from another, one event might impinge differently on two people in different ways, or impinge differently on one person at two times. These are differences of perspective. Blowing smoke in your face might be fun from my perspective though not from yours. Smoking now might also be more pleasant for me now than in hindsight. Different people occupy different perspectives, and a person changes perspective year to year, moment to moment.
Indeed, Schofield thinks you might occupy two perspectives at once, if you have two clashing “practical identities,” two ways of conceiving “who or what” you are that carry different implications for how you ought to act (104). For example, if you are a parent and a citizen, you might be torn as to whether you ought to treat your children impartially (107). In some cases, the clash is serious enough that, from one perspective, you can issue a demand to yourself in another perspective, generating a duty to yourself within a single moment.
At this point, the book shifts from moral to political philosophy, as Schofield applies his theory to the issue of paternalism—state interference with a citizen’s liberty for his or her own sake (chaps. 5–6).
Paternalism is easy to justify if we are illiberal and think that the state’s business is to enforce morality where it can. But if we are liberals and think that “state coercion must be strictly limited” to the aim of enforcing duties and promoting freedom (139), paternalism seems perverse. If the goal is to promote freedom, how could we achieve that by infringing on a person’s free choices?
Schofield’s answer: by enforcing duties to self. If I am going to damage my future body irreparably by smoking, the state can stop me—not because I am foolishly harming my long-term interests, but because I am violating a duty to myself, a duty that I cannot now waive (169). The state may also interfere in my finances to prevent me from forcing “myself into labor later in life for the sake of activities completed much earlier on” (173)—for example, the state may restrict how much money I can borrow. And the state may outlaw slavery contracts because a person has a duty to herself “not to waive the duty preventing others from assuming toward her the role of master” (178). (I think that Schofield should have treated this last example as a clash of practical identities: the slave and the free person. Instead, he focuses on the duty not to waive, which is not so clearly linked to his theory.)
Is it because
There’s a little part of you
That wants to be dead?
Schofield then explores other political issues like punishment and distributive justice.
One is left with the striking image of a person as vulnerable to domination, at each moment, by the unfettered hands of his or her past self. The state then must choose what to do with its fetters. This is liberalism founded not on the separateness of persons but on the separateness of perspectives. I found this part of the book wonderfully creative, even if it leaves many cutoff points uncut. I agree that “it’s possible to hold that the ingestion of protein shakes is rightful, whereas allowing tobacco use is not” (188). But one does wonder about the in-between cases—junk food, face tattoos, anabolic steroids, unflossed teeth. Duty to Self is not so interested in litigating these line calls. It is more about showing you that this is a line worth drawing.
So much for the book’s argument. Now my objections. Each touches on a different aspect of Schofield’s view on perspectives—in particular, (1) accessing other perspectives, (2) destroying perspectives, and (3) replacing perspectives.
First, access. In Schofield’s view, I have a duty to myself not to smoke because I cannot now waive a duty that harms myself in a later perspective. The principle here is that one cannot access other perspectives in one’s life. One cannot consent to things that befall one in the future, or issue and rescind demands from the perspective of one’s past.
This seems to me—at first glance—too restrictive. Can’t I consent now to surgery that will take place in a few moments, or later this afternoon? Can’t I forgive you now for a wrong done last year? What about promising to be faithful in the future, signing a yearlong contract, or selling property rather than leaving it for my future self?
In reply, Schofield could say that my perspective persists over time, and that I inhabit one enduring perspective during my present consent and this afternoon’s surgery—but this does not seem true to his view that perspectives last only a moment.
A better move is to point to the priority of the future self’s will. If, later today, I decide that I am not willing to undergo surgery after all, that overrides my prior consent. True. But this is not quite strong enough to defend Schofield’s view. The view is not that past consent is only presumptively valid, with an exception for future vetoes; the view is that past consent is always as such automatically invalid—like an attempt to consent on behalf of somebody else. That view is quite strong, and it needs a bit more defense.
The second objection has to do with suicide. Most writers who believe in duties to self believe in a duty not to end one’s life when the future is bright. Schofield, too, is open to such a duty (though perhaps not in the case where one’s “rational faculties” will soon degrade or be overwhelmed by pain (138).
But how could Schofield’s theory forbid suicide? If I end my life, I destroy my perspective in the future. Duties to self, for Schofield, involve a conflict of perspectives. If there is no second perspective, whence the conflict?
Schofield would reply that no actual conflict has to take place to generate a duty; it is enough that someone from some perspective could legitimately demand that I, in my current perspective, refrain. Duties flow from “hypothetical demands from an idealized standpoint” (63). His example is the burglar who eludes a homeowner who nevertheless could have demanded that the burglar not burgle. By analogy, Schofield might say that I could, from a future happy perspective, demand that I now refrain from suicide.
But talk of “hypothetical demands” is ambiguous when my actions affect which perspectives will exist. Is Schofield saying that duties are generated by hypothetical demands from actual perspectives, or from hypothetical perspectives as well? There are two principles here:
Actualism. I have a duty to do X if there is a perspective from which X could be demanded.
Possibilism. I have a duty to do X if there could be a perspective from which X could be demanded.
When suicide snuffs out the possibility of a future perspective—one from which I could have legitimately demanded that I not commit suicide—the principles come apart. Actualism sees no problem with tragic suicide; Possibilism rules it out.
To my mind, the Possibilist principle is more plausible (at least, in the idealized cases of tragic suicide at issue here; real suicides are more complicated). But Possibilism has an unpalatable consequence: I owe it to merely possible people to bring them about if they would have happy lives, since there is a possible perspective—the one where they are happy—from which they could legitimately demand to exist. This is not a popular view, outside of utilitarian circles [for a sense of the debate, see Frick or Gustafsson]. Perhaps Schofield could avoid it by only counting the possible perspectives of actual people, rather than the possible perspectives of merely possible people. But why insist on actuality for persons but not for perspectives?
There is a dilemma here: only Possibilism can ground the duty not to commit suicide, and only Actualism can avoid the duty to make happy people.
There is no easy way out of this dilemma. Schofield might want to say, “So much for the duty not to commit suicide.” But he needs that duty to exist, since, on his view, such duties are the bulwark against the demands of duties to others:
A person can owe herself the performance of particular actions; her passions, projects, and relationships can generate, through self-directed duty, normative force that canmeet the force which emanates from duties owed to others, making possible a life worth living. (197)
I agree. But if that’s right, then the duty not to commit suicide is essential, or else nothing will “meet” the obligation to sacrifice one’s life for others’ good. Schofield needs Possibilism to protect personal freedoms.
Finally, there is a wonderful objection due to Yuliya Kanygina. It is a version of the nonidentity problem.
Suppose I would like to smoke now, but I know that smoking will cause my life in the future to be painful, though still well worth enduring. There is a hypothetical conflict between two perspectives (“P” for present, “F” for future):
P: wants to smoke.
F: will be harmed.
I owe it (from P) to myself (at F) not to smoke.
But refraining from smoking does not make me better off at F. It replaces F with another, happier perspective:
F+: will be helped.
For Schofield, then, I have no grounds at F to complain about smoking at P. Smoking would not make me worse off at F, but would replace F with F+. From the point of view of F, I should, if anything, be grateful to have been brought into existence in that good-enough perspective. So, for Schofield, it seems that there is no duty here not to smoke.
How does this objection relate to the one about suicide? It seems Schofield could get out of Kanygina’s objection by appealing to Possibilism and saying that I can demand from F+ that I not smoke. If that’s right, then Kanygina’s objection, like the suicide objection, only has force if Schofield accepts Actualism. But Kanygina’s point—that I cannot complain from F—raises a new problem. It suggests that smoking wrongs me at F+, when one might have expected the wrong to fall on F. (Is the distinction obscure?)
A final point about this argument: one could reply by saying that F = F+. Though different qualitatively, perhaps they are numerically the same. But this does not fit what Schofield says about perspectives elsewhere, so it seems just as ad hoc as the claim that I inhabit the same perspective before my surgery as I do while under the knife.
What can we learn from these objections? I think they reveal a disanalogy between persons and perspectives. For one thing, we identify with future perspectives. The perspectives in a life form a natural unity, not a mere aggregate, and I may make certain decisions on behalf of my future self. We also identify with other possible perspectives. When I look back on past mistakes, I wish I had chosen otherwise, so that I would be better off in some other possible perspective; I do not see this as a loss, only an upgrade.
I find all of this fascinating. Schofield’s framework is inventive and fruitful, and though it gives rise to problems, they strike me not as philosophical sand traps but as genuine issues worth puzzling through.
Pleasantly surprised by the fact that you’re Mount Eerie fan!
This was really interesting, as I came in with no idea of how y’all approach this topic in the modern day. The perspectives (rather than temporally different selves) stuff is cool! If you’re willing to offer spoilers, how does your upcoming book differ from Schofield’s takes, in broad strokes?
(I actually have a bunch of questions on this topic, but they are still very much in the process of formulating themselves….)