Making Your Practical Ethics True
Metapreferences & path-dependence
Tyler Seacrest asks a very good question about preference consequentialism:
Consider a society that is a quarter gay and three-quarters straight, where the straight members are all very against gay sex. Under preference consequentialism, it seems clear that gay sex should be considered unethical. However, if all members had a strong metapreference that preferences should be weighed based on who is more directly affected, then the gay members preference should be given more weight on this issue since it affects them more, and thus gay sex is ethical. I’d be curious if you think metapreferences pose any kind of challenge to preference consequentialism.
I agree with Tyler: in this scenario, gay sex is ethical.1 But how does this square with preference consequentialism? Surely all the straight people would have their preferences better satisfied if the gay people abstained.
Imagine Luigi really wants to climb to the top of Mount Everest; he sees it as his calling, his chance for growth and struggle and the ability to become someone that he’s proud of. Luigi would not climb into the Experience Machine just for a perfect simulation of climbing Everest, even if the simulation would trick him into thinking it were the real deal. Luigi wants the real thing.
But Luigi is given another offer: to step into the State Update Machine. The State Update Machine will teleport him to the top of Mount Everest, clad in winter gear, with his body, brain, personality, and memories updated just as if he had actually climbed Mount Everest. Unlike the Experience Machine, this process affects Luigi’s real body and environment—it even creates tracks in the snow and campsites where Luigi would have hiked. The end state of Luigi after the State Update Machine is identical to the end state of Luigi after actually climbing Mount Everest… except that Luigi won’t have actually climbed Everest. Does the State Update Machine satisfy Luigi’s preferences?
Well, I’ve somewhat underspecified Luigi’s preferences and why he desires to climb Everest. Maybe what he wants is the personal transformation in mind and body alone, and he concludes that the Experience Machine will give him that. But he could also just want to climb Everest, terminally. There is nothing impossible or contradictory about that. If we want to capture the whole range of possible preferences, we need to accommodate this.
And so this demonstrates that preference consequentialism is path-dependent. At time t, Luigi in the State Update Machine world is indistinguishable from Luigi in the Actually Climbed Everest world. But Luigi cares about not just that state, but how he arrived at that state—by actually climbing Everest.
Return to Tyler’s gay-straight world. It’s easy to imagine:
The straight people would really prefer to live in a world where there weren’t any gay people at all, and therefore no gay sex happened.
The straight people would not prefer to murder all the gay people so that they wouldn’t be around anymore, and then wipe the memories of everyone so they believed gay people never existed. Even though the end state would be the same, that doesn’t mean you want to do a bunch of murders.2 It’s a bad pathway to the end state.
Similarly, it’s easy to imagine:
The straight people would not prefer that the gay people change their minds or be shamed, persuaded, or coerced away from having gay sex on the basis of a preference utilitarian argument. It’s a bad pathway.
The straight people would prefer that people reason and make decisions according to a weighted model where those that are directly affected get more say in the matter. That’s the code by which they want to live. Obviously it would be ideal for them if that aligned with having no gay sex at all, but in the cases where it doesn’t, their preference that you reason according to that model is more important.
This doesn’t make preference consequentialism contradictory. Imagine a world in which obeying deontological constraints really made everyone happy and really was the best way to create the greatest good for the greatest number; but reasoning by utilitarian logic made people depressed and anxious. Utilitarianism would thus recommend being some sort of Kantian in day-to-day life. But that wouldn’t make utilitarianism false. In fact, acting like a Kantian would be right precisely because utilitarianism would be true!
When people prefer not to act according to a preference consequentialist calculus, they are ultimately expressing a preference about how they should reason and act, not a preference about what metaphysics moral facts should have. If one really desires that a different metaethics be true, then this is sadly an unsatisfiable desire—just like desiring that atoms be indivisible or 1 + 1 = 3. But, just as preferring different outcomes is what makes them right or wrong, the deeply held desire to live by a particular code of action or practical ethics can make it right to do so.
Lest anyone get mad at me for trivializing gay-bashing or something like that: I am bisexual, I have had gay sex, it was pretty cool, and people shouldn’t hate on us. Also, don’t get mad at Tyler: I brought up the gay sex example myself in my last article.
The murders would be wrong independent of what the straight people thought, because the murders would ruin the chances of the gay people at living out any of their preferences, not just the sex. But it’s still possible for the straight people to have their own preference one way or the other.
I’m not so sure to me. It appears obvious that in the situation, your balancing the straight peoples preference that nobody ever do gay sex against their preference in favour of whatever moral theory, they believe in, and there is no rule saying that the second preference has to be stronger. In any case, even if the second preference is stronger, there will be cases where something only mildly violate the second preference but is highly beneficial to the first preference. In such cases, I think it’s obvious that a preference consequentialist would want to go ahead with these things.
Wow, thanks for taking the time to write such a nice response! Earlier, it struck me as possible that preference-based consequentialism might be prone to some sort of paradox in rare cases, but I don't have that intuition regarding utilitarianism. Thus that analogy of the utilitarians all adopting deontology really helped. Thus, in a society where no one likes preference consequentialism, they are free to act like it's not true and adopt deontology, but of course a preference doesn't determine if something is true or false, so preference consequentialism could still be the true moral theory in this case.