Against Egoism and Subjectivism




Derek Parfit was, in my view, the greatest moral philosopher of (at least) the twentieth-century, and Reasons and Persons may be the greatest philosophy book of all time. So I was excited to get to write the Cambridge Elements volume on Parfit’s Ethics (and to retain the copyright, so you can download the entire pre-print PDF for free).

In the spirit of encouraging broader appreciation of Parfit’s work, I’m planning a short series of posts, Parfit in seven parts, drawing from my book. This first post will introduce Parfit’s arguments for thinking that we objectively ought to care about others’ well-being (no matter what desires we might have to begin with).

Against Egoism

Non-philosophers often identify rationality with self-interest. That is, they accept Rational Egoism: the view that what each person ultimately has most reason to do is whatever would make their own life go best. But Parfit argues that this view is wildly implausible, on both intuitive and theoretical grounds.

Intuitively, it needn’t be irrational for an agent to heroically sacrifice their life to save others (even if they know they would otherwise live a happy life, unhaunted by remorse, such that the early death is unquestionably worse for them). Given that the agent is fully informed, thinking clearly, and acting in a way that is morally admirable, it is difficult to see any fair, non-dogmatic basis for insisting that their choice is irrational, just because they choose to prioritize others’ interests over their own. Unless supported by some incredibly compelling theoretical rationale, the implausibility of Rational Egoism’s verdicts in cases like this gives us good grounds to reject the view in favour of some more permissive alternative.

But in fact theoretical considerations simply undermine Rational Egoism even further. Compare the following three principles:

(A) No individual preference is intrinsically irrational (just in virtue of its content), not even preferring a lesser benefit over a much greater one.

(B) It’s irrational to prefer a lesser benefit over a much greater benefit, merely on the grounds that the former occurs now whereas the latter occurs later.

(C) It’s irrational to prefer a lesser benefit over a much greater benefit, merely on the grounds that the former accrues to you whereas the latter accrues to another.

Rational Egoists accept principle (B) but reject both (A) and (C). Parfit argues that this is an unstable position, as there are good theoretical grounds for treating (B) and (C) alike. Parfit’s basic idea is that there is a kind of formal analogy between ‘I’ and ‘now’, or between agent relativity and temporal relativity. When Rational Egoism dictates that we must be temporally neutral (giving equal weight to our interests at all times) but agent relative (giving more weight to ourselves than to others), it reveals itself to be what Parfit calls an “incompletely relative” theory. A theory is on sounder structural ground, Parfit believes, when it is either fully relative or fully neutral, treating both these dimensions of variation alike.

Why does Parfit think this? One way to understand his core insight is to notice that choices are made not only by particular agents but also at particular times. (It may be helpful to think of the deliberating agent as a “momentary self”, distinct from the various “future selves” that will replace them at later times.) Just as a deliberator may ask, “Why should I sacrifice my interests just so that some others may benefit?”, so we may imagine one asking, “Why should I now sacrifice my current interests just so that my future selves may benefit?” If the former question is thought to raise a serious challenge to altruistic requirements, parity of reasoning would suggest that the latter question should be considered similarly challenging to requirements of prudence.

Rational Egoists might seek to defend requirements of prudence by appeal to the objective features of normatively significant phenomena such as pain. Pain matters because of how it feels, and the felt badness of pain is not affected by mere differences in timing. This is, Parfit suggests, an excellent defense of (B). But it is not one that the Rational Egoist can comfortably appeal to, for analogous reasoning would equally support principle (C). After all, the felt badness of pain is likewise unaffected by mere differences in who feels it.

Rational Egoism is thus undermined on both intuitive and theoretical grounds. We should instead accept a theory of practical rationality that is either more subjective or more impartial. Parfit’s arguments here provide a nice demonstration of the power of philosophy to force a rethinking of prevalent assumptions. As a result of such arguments, philosophers now overwhelmingly reject Rational Egoism.

Against Subjectivism

More philosophers remain drawn to normative subjectivism, the view that we have reason to do whatever will fulfill our ultimate (non-instrumental) desires—desires that may, in principle, have any contents whatsoever. Where Hume declared, “’Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger,” Parfit argues that desires can be irrational just as beliefs can.

To illustrate, Parfit imagines an agent with Future-Tuesday Indifference, who “would choose a painful operation on the following Tuesday rather than a much less painful operation on the following Wednesday.” The imagined agent knows he will subsequently regret it, but simply doesn’t care—about either his future agony or the associated regret. Such an agent seems less than perfectly rational. Many of us would probably describe such a pattern of concern as “senseless” or even “crazy”. As Parfit sums up his case: “Preferring the worse of two pains, for no reason, is irrational.”

Future Tuesday Indifference shows us that there’s more to practical rationality than just taking the effective means to whatever your ends may be. Our ends themselves are open to rational evaluation. At a minimum, there’s some rational pressure to treat like cases alike, or avoid arbitrary distinctions: if pain is worth avoiding on other days, and it feels no different on those calendar days arbitrarily designated to be ‘Tuesdays’, then we rationally ought to regard Tuesday-pain as similarly worth avoiding.

But Parfit’s objection to subjectivism can be pressed further: Avoiding arbitrary distinctions by becoming indifferent to all future agony would simply compound upon the error of the Future-Tuesday-Indifferent agent. To restore rationality, it isn’t enough to be consistent. If sufficiently wrong-headed, that might just make you more consistently irrational. To do better, we must respond to evaluatively significant features of the world in the ways that they actually merit. (And agony merits avoidance, assuming it offers no countervailing instrumental benefits.)

Normative subjectivists have trouble accommodating the datum that all agents have reason to want to avoid future agony. For their view seems to imply that agents never really have reason to want anything: our wants are simply taken as given, and the subjectivist instead focuses on what we have reason to do, namely, effectively pursue whatever it is that we antecedently want.

To my mind, this raises a puzzle: Why would we have reason to pursue some end that we have no reason to want? Hypothetical imperatives of the form, “If you want X, you should do Y,” present relations of normative inheritance: given that X is worth pursuing, then Y is too. But a view on which there are only hypothetical imperatives is effectively a form of normative nihilism—no more productive than an irrigation system without any liquid to flow through it. Or so it seems to me.


If we reject both egoism and subjectivism, we’re left with some form of broadly altruistic objectivism: some things—including the well-being of others, and not just ourselves—are objectively worth caring about. If we fail to care about them, we’re not merely not nice (though we may be that too!), we’re also making an error of judgment, failing to respond to genuine reasons. So claims Parfit’s normative objectivism.

While one is by no means forced to share Parfit’s conclusions here (subjectivists, especially, have offered sophisticated responses to try to escape his arguments), I think there’s a lot more going for them than many would otherwise realize. For any readers who haven’t come across Parfit’s work before, I warmly encourage you to look more into it!

[For further reading on this topic, see sec. 2 of Parfit’s Ethics, and the works cited there.]

Originally appeared on Good Thoughts Read More