Thursday, March 25, 2021

Learning from Lucifer

Via Daily Nous, I came across this funny comic on "Effective Villainy", which in turn got me thinking about what we could learn from the apparent symmetry between good and evil.  After all, it might be clearer what evil calls for, in which case we could -- if we accept symmetry -- then draw interesting conclusions about what's right and good.

Consider side-constraints, as in this evil transplant scenario: a car crash victim is about to bleed out, but then their healthy organs would save five transplant patients -- unless Lucifer sweeps in and "heroically" stems the bleeding, saving the car crash victim (and ensuring the other five die).  What should_[subscript evil] Lucifer do?  It seems clear that a side-constraint against saving lives would make no sense, from the perspective of evil: the rational pursuit of evil would lead Lucifer to do whatever ultimately proves most harmful, regardless of whether he had to get his hands "dirty" helping (ick!) various individuals along the way.   Better_[sub evil] to save one than to allow five to be saved.

Or consider aggregation. Opponents of aggregation may ground their view in either axiological intuitions (e.g. in Scanlon's transmitter room case, it just seems worse for the one person to suffer from electrocution than for the billions to miss out on seeing the football game live), or in hand-waving claims that aggregation somehow fails to respect the "separateness of persons".  Parfit showed the former to be incoherent, and I've argued that the latter is baseless, but suppose you're not yet convinced. Let's learn from Lucifer!

In the transmitter room case, it's natural to assume that Lucifer would want the guy to be electrocuted, which I think reveals our intuition that this is (seemingly! but we need not endorse this seeming upon reflection) the worse outcome -- not actually a case in which what's right diverges from what's consequentially best.  So that seems like a problem for the anti-consequentialist's use of this thought experiment.

In other cases, where it's clearer that the many's interests really do (factually) outweigh the one's, it would seem weird for Lucifer to prioritize harming the one a lot over harming the many by more in aggregate (except insofar as this was motivated on prioritarian grounds, perhaps).  It would seem especially bizarre for Lucifer to embrace a strict "numbers don't count" view, and be indifferent between drowning one or drowning five in a lifeboat case. ("Killing more people isn't worse for anyone than killing one is for the one, so in what sense is it really worse at all?  Why should I bother to kill more?" - Evil Taurek. Given how poorly Evil Taurek seems to understand evil-doing, why would you consider his light-side self any less confused?)

Overall, then, it seems clear that Satan would be a consequentialist.  So, shouldn't you follow suit? (Just, you know, aiming at the opposite ends...)

Or should we instead reject the apparent symmetry, and think that genuine morality should look rather different from the simple opposite of pure immorality?  Perhaps one could think that morality is responsive to certain considerations -- respect for persons, or whatnot -- that evil is simply indifferent to (rather than being positively opposed to)?  It'd be interesting to hear the idea spelled out in greater depth.

For those who are on board with the symmetry, what other lessons might be drawn?  (I like Michael Slote's suggestion, from his 1985 Common-Sense Morality and Consequentialism, p.77, that the implausibility of the minimalist view that only the very worst option is wrong provides us with grounds to doubt maximizing accounts of the right...)

Population ethics remains tricky.  Would Lucifer prefer to immiserate existing people, or bring into existence an entirely new population of even more miserable future people (who would otherwise not exist at all)?  Not obvious...


  1. I don't think this approach clarifies things. The examples you give PRESUME that Satan is a consequentialist, and don't use any reasoning that hasn't already been well-developed for why Agent_of_Good is/should be consequentialist. I dearly love SMBC and that particular strip, but I strongly expect that the symmetry extends to the ambiguity of "good/bad".

    To test this, I'd expect a deontologist to argue that Satan wouldn't stop the bleeding, preferring the individual harm over the collective good.

    I really enjoy that comic

    1. I'd be interested to hear from deontologists -- I would actually be surprised if many had that reaction. (I'd expect more to share Brandon's view, below, that the Devil is a consequentialist and so much the worse for consequentialism!)

  2. Take the good transplant scenario. It seems that there are at least two types of evil that are relevant: (1) the evil of more deaths or fewer savings and (2) the evil of using a person as a mere means. And, in the good transplant case, we must trade one for the other: we either (a) kill the one to save the five and have the evil of using the one as a mere means to saving the five or (b) don't kill the one to save the five and have the evil of more deaths. In the evil transplant case, things are not symmetrical. Since you don't use a person as a mere means by saving their life, we only have one type of evil to contend with: the evil of more deaths. Thus, in the evil transplant case, Lucifer will clearly save the one rather than save the five (by not saving the one). For, in this case, there is just a choice between one death or five deaths. And five deaths are more evil than one death. But since the evil transplant case doesn't involve two types of evil and the good transplant case does, I don't see how we can learn anything useful about the one from the other even if we assume that apparent symmetry between good and evil.

    1. Good point! It's a flawed analogy. Still, I think there's something to the general point that constraints against direct helping wouldn't really make sense from the evil-doer's perspective, which at least raises the question of why we should think constraints against direct harming are any more sensible for the right-minded.

    2. Hi Richard

      Fun scenario! If we assume Lucifer aims at maximizing death for some reason, both saving or letting die look like poor choices to me: rationally, Lucifer will instead kill the car crash victim in a painful way that also damages the organs the other people need beyond repair. Or better yet, he will just kill everyone. So, we need to put limits to Lucifer's power, but the answer regarding his rational course of action depends in part on that.

      In any case, unless we assume that he is a consequentialist (as Unknown points out, if I understand them correctly), he may well not aim at maximizing death, or even suffering. For example, he may aim actively inflicting unjust suffering and death.

      Consider the classical example: five people dying, but killing one would make organs available and that (let's say) will save them, also preventing further suffering. If Lucifer with his limited power gets the opportunity to kill the one, and enjoy the evil pleasure of inflicting pain and death by his own hand, then that would be the rational course of action, whereas just not murdering anyone - which would result in more fatalities - would be irrational given his goals.

      So, it seems to me you get side-constraints if you like, depending on Lucifer's goals. So, in addition to limiting his power, you would need to set a goal of maximizing death, or something like that.

      That aside, above I tried to find objections that do not question the implication immoral-> irrational because I take it it is generally accepted within the consequentialism-deontology debate. But leaving that said, in my view the main difficulty with these scenarios is that they are about what would be rational for Lucifer to do. As I see it, killing one to take the organs and save five is immoral for any a moral agent - which would be a human or sufficiently similar agent -, and usually irrational for humans, but not necessarily irrational.

    3. Interesting! I was assuming that Lucifer would want to do what was morally worst (or most evil), and then invite people to use their own judgment to see what that amounts to. So I guess the question is whether the "actively inflicting" version or the consequentialist version of Lucifer seems more evil. (We can assume that he enjoys a similar amount of "evil pleasure" in his choices either way. I don't think we should be too swayed by the fact that we can more easily imagine a villain enjoying the more hands-on intervention. That just seems a psychological quirk, like how altruists typically get more satisfaction from directly helping a few than by indirectly helping many more.)

  3. I don't really have much to contribute to the argument itself, but I found this post rather humorous, because I've literally said in philosophical argument with someone that the Devil is a consequentialist. To some extent this is just a function of the fact that the way consequentialism works is that you try to get more of a certain kind of result, so anything that treats this as the main thing to do will have a consequentialist structure, regardless of what the kind of result is. The example I use in ethics courses is 'murder consequentialism'; if your aim in life is make sure that the world has the greatest possible number of murders, that's clearly a consequentialist life, even if a completely insane one. (It's why, despite the fact I've never come across a philosopher who quite liked the argument, Bentham and Mill have to make their arguments that happiness/pleasure is what we should aim at because we already are aiming at them -- it rules out crazy consequentialisms, which structure alone will never do.)

    Unlike Unknown, I don't think one gets the same result, or at least not very easily, with deontology. You can have crazy rules/obligations/laws, but in general what makes rules crazy is that they are incoherent or self-defeating or for some reason impossible actually to implement. For some instrumental meaning of 'rational' you can be a very rational crazy consequentialist, systematically aiming for consequences that are really bad. It's a bit harder to see how you could be a very rational crazy deontologist, truly applying without exception a rule or standard that's actually wrong in and of itself. This is, I think, one of the reasons Kantianism always gets its supporters: there's something very plausible in the idea that every kind of moral failure involves *some* kind of inconsistency or arbitrary exception-making. I suppose this ties into the question of whether you can have a deontology that genuinely allows good and evil to be symmetrical; most deontologies seem committed to asymmetry, because the asymmetry is explicitly part of why you need this standard or rule, not another. Kantianism is the obvious case, but it's true even of divine command theories and views in which moral life is based on rules created by society. But maybe there's a more abstract kind of deontology that allows a genuine symmetry, I don't know.

  4. I love this post!

    I wouldn't say I'm a Taurekian, but here's a bit of sympathy. Even a Taurekian Satan might still "bother" with killing more -- at least, if that means killing more people in addition to those already on the chopping block. Taurek (of course) can't say it's any worse for Al and Betty to die than for just Charlie to die, but every Taurekian, to my knowledge, thinks it's worse for both Al and Betty to die than for just Al to die. This claim isn't actually in Taurek, but Kamm says he endorsed it in conversation (Morality, Morality V1), and it's commonly attributed to him in the literature, both by opponents (Kavka's "The Numbers Should Count") and friends (Lübbe's "Taurek's No Worse Claim"). And I know of some unpublished stuff that talks about it, but I guess I shouldn't spill those beans.

    Anyway, that doesn't really undercut your post. It's still pretty funny that Taurek's Satan would be indifferent between killing one and killing 50 others. But at least that's not as Satanically ineffective as, say, killing one then thinking, "Well, can't get any worse than that!"


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)