What Ethical Theory Is

And why low-decouplers can't handle it

Consider the popular inference:

P1. Ethical theory X implies that one objectively ought to φ in circumstances C.

P2. That sounds bad. We shouldn’t want people to be so disposed to φ.

∴ C. Ethical theory X is false and dangerous.

This inference is invalid. It falsely assumes that we should always want people to be disposed to perform any objectively right action. But we shouldn’t always want this, for at least three reasons:

(1) Some such dispositions might be too costly, for example due to also disposing the (fallible) agent to act disastrously wrongly in nearby, difficult-to-distinguish circumstances (C*, where the agent falsely believes they are in C).

(2) Even just considering C, we may not want the agent to φ due to moral uncertainty: the upside given X may not be worth the potential downside given other credible views. (That doesn’t show that X is false, though: people, not theories, should be uncertain.)

(3) Even some knowable true claims that could be safely and reliably acted upon might “sound bad” to assert for contingent social reasons (leading low-decouplers—who view everything through the lens of coalitional signals—to associate the claim with Nazis or the like). For example, many true claims about which lives it would do more good to save—due to differences in life expectancy, expected quality of life, or social/instrumental value to others—are socially unassertable.¹

As a result of all this, ethical theorists (competent ones, at least) are careful to distinguish criteria of rightness from recommended decision procedures.² We should expect the two to sometimes come apart—even for deontologists—as there’s no empirical guarantee that naive instrumentalist pursuit of moral ends will be the most reliable means of securing them. So it’s vital to be clear about which question one is trying to address.

Fake Moral Theories

Ethical theory concerns criteria of rightness (or objective preferability). People talking about ethics on social media and public forums are almost always instead talking about something more practical: “recommended decision procedures”. This leads them to invent a range of totally fake theories. Consider:

Fake virtue ethics: Virtues are important for guiding action! [I recommend that:] people should try to inculcate and act upon good character traits.

Fake rule consequentialism: Rules are important for guiding action! [I recommend that:] people should follow good rules rather than attempting to discern case-by-case which act would make things go best.

Fake act consequentialism: [I foolishly recommend that:] people should naively attempt to calculate and maximize expected value on a case-by-case basis, without any regard for higher-order evidence about how their reliability at attempting this compares to the reliability of other (rule or virtue-based) heuristic guidance.

NONE OF THESE ARE MORAL THEORIES!³ They’re just practical recommendations. (And the last one is obviously bonkers.) Don’t get me wrong: practical recommendations are important, and worth discussing.⁴ If you’re trying to catch a ball, keep your eye on the ball rather than on pen-and-paper calculations involving Newtonian formulas. Good advice, but no threat to Newtonian physics. In the same way, attending to rules and virtues may be good moral advice, but is also no threat to (real) act consequentialism as an ethical theory.

Ethical theories, like scientific theories, simply aren’t in the business of offering practical advice. Even so, they may (in the right hands, in the right circumstances, when used appropriately) prove extremely helpful for determining what the best advice would be.

So why care about moral theory?

The first thing to say to this challenge is: you don’t have to! Not everyone has to care about physics. Not everyone has to care about ethical theory. You could be an excellent athlete and a good person without having a single theoretical thought (about either physics or ethics) in your entire life. Still, I think we should all want some people to do theoretical research, for both intrinsic and instrumental reasons.

Intrinsically: fundamental theorizing may be among the noblest of human activities. There’s just something awesome about humanity’s collective scientific progress in understanding the physical world; and while philosophical progress is harder to verify as such, I think it’s no less awesome when it does happen. Insofar as you think intelligent people are more morally significant than sheep, you should probably agree that we ought to be exercising that distinctively valuable capacity in order to realize our full and distinctive value. (The unexamined life may still be worth living, but there’s something a bit pitiful about it.)

Instrumentally: as hinted at above, improving our grasp of fundamental truths can prove extremely useful! See, e.g., bioethics, for a sense of how vibes-based ethics can lead us badly astray. Personally, I think my theoretical background gives me some distinctive and valuable insights into practical ethics that vibes-based applied ethicists tend to miss. Others may disagree with me, of course. But I would guess that any who think my practical ethics contributions are overall bad will also think that my fundamental theorizing is misguided (and that this latter fact explains why my practical recommendations are so bad). More generally, as I replied to John Quiggin in a recent thread:

If you’re only interested in applied ethics and public policy, then a lot of ethical theory may look at first glance like “pointless theoretical rabbit holes”. But fundamental (“ideal theory”) disagreements may have important downstream implications for what we ought to do in practice. After all, it’s hard to know how to respond appropriately to uncertainty if you can’t even answer the easier question of what would be preferable in the absence of any uncertainty!

Ethical Theory is Decoupled from Politics

Ethical theory is extremely theoretical. It is an attempt to give a systematic explanation of moral properties and how they apply across every possible world. A complete moral theory is like a function that takes (as input) a complete description of a world’s natural/descriptive truths, and returns a complete account of its moral truths. This characterization brings out why ethical theories aren’t directly practically applicable: one of the most important facts about our practical predicament is that we don’t know what possible world we are in. Empirical uncertainty is ubiquitous, and how to deal with it is the central challenge of practical ethics. But that’s simply not what ethical theory is about. It concerns what God would judge (and on what basis) if he were to survey the entirety of modal space and form moral judgments about everything he saw. (That’s why we can get away with stipulating details of thought experiments. It doesn’t matter, for the purposes of ethical theory, which possible world is actual!)⁵

One important upshot of this is that moral truths are radically “decoupled” from things like perceived political valence or dog-whistle associations. As noted at the start of this post, there may be literal moral truths that “sound bad” to assert in our social context. Because low decouplers conflate truth and social assertability (or “sounding good”), they struggle to engage competently with ethical theory. You instead get absurdities like Emile Torres accusing Nick Beckstead of “white supremacy” for mentioning in his Ph.D. thesis the possibility that saving a life in a rich country may have greater positive ripple effects, and thus be all-things-considered “more important”, than saving a life in a poor country, other things being equal.⁶

As with the folks discussing fake moral theories on Twitter, it’s important to understand when moral philosophers are simply engaging in a different (more theoretical) sort of project than an outsider might expect, resulting in criticisms missing their mark due to irrelevance. When objecting that a theory has implications that “sound bad”, pause to consider (i) whether this reflects a problem with our contingent social circumstances rather than with the timeless claims of the ethical theory in question, and (ii) whether competing views actually have more plausible implications, or whether you’re mistakenly rewarding them for incompleteness (failing to determinately answer the disputed moral question at all).

1. This is presumably because most people would interpret them as dog-whistles, implicating some much stronger, false claim (e.g. that you should essentialize the groups in question, or that any member of one demographic is more worth saving than every member of another, or that such rules should be institutionalized in harmful stigmatizing ways).

Consider how many more true claims would be assertable in a society without bigotry, and hence without fear and monitoring of intrinsically innocuous and true (but “politically incorrect”) claims that may be used as dog-whistles for coordinating bigotry.

2. I actually don’t love this standard way of framing the issue. I think it invites misunderstandings, and we’d do better to replace it with the telic vs decision-theoretic frame that I’m developing in Beyond Right and Wrong. The basic idea: core ethical theory addresses the telic question of ultimate ends, which needs to be supplemented with an account of instrumental rationality before we get any practical advice for what fallible humans “ought” to do. Naive instrumentalism is clearly unfit for human-sized minds. So there should be no temptation to infer “People should try to φ” from the mere fact that φ-ing would be objectively preferable in some (rare) circumstances C, when φ-ing generally goes disastrously.

3. Real rule consequentialism, for example, claims more strongly that following the ideal rules makes an action right (even when this particular instance of rule-following is transparently harmful). See my objections to rule consequentialism for more detail.

4. Here are some of mine.

5. Incidentally, this is why (non-analytical) metaethical naturalism is false. (Little-known fact!) More importantly, it also explains why no sound objection to a moral theory can rely on contingent premises. Demandingness, cluelessness, self-effacingness, and abusability objections are all fallacious if (or to the extent that) they wouldn’t have the same force were the identified “problematic” feature merely possible rather than actually-realized. But these features are possible to realize on every credible moral theory. So they cannot reasonably be regarded as evidence that a theory is false.

6. Notably, not all else is equal in practice: we can save lives in poor countries much more cheaply, which—Beckstead agrees—is a strong reason to do so! But the more principled response to Torres is just that Beckstead’s claim had nothing to do with race. (Thank goodness the culture has since moved on from this idiotic style of criticism, where you somehow won debate points by hallucinating violations of wokeness and baselessly attributing them to your interlocutor. Ugh.)