Edit to add:

I've read more about compatibilism, including the comments on the original post, and I think it's all been quite helpful in enhancing my understanding. The post and comments here were particularly illuminating: https://whyevolutionistrue.com/2021/04/30/why-do-we-need-free-will-compatibilism/.

I'm increasingly inclined to think that the terms "free will" and "moral responsibility" are not helpful words in getting at what I'm interested in.

What I'm really interested in is, similar to my 4th question in the original post, whether certain kinds of "reactive attitudes" (to borrow a term from P. Strawson), e.g. praise and blame, make any sense. And I think these and many other reactive attitudes really depend on the ability to do otherwise. But everyone agrees that determinism rules out the ability to do otherwise. So I'm inclined to think it never makes sense to hold these specific attitudes.

Take, for example, the person who cooly and calmly deliberates and then decides to kill someone. Maybe he stood to gain financially if he murdered his victim. We are reflexively inclined to say he is morally blameworthy. But I think there is a deep sense in which "his brain made him do it." The neurons in his brain fired in such a way that he found the argument "kill X for money" more persuasive than the alternative. Of course, I might still rationally adopt a range of negative attitudes toward the murderer in the same way that I might feel negatively toward a hurricane that killed someone. I might even use some of the same words--e.g., I might say that I "hate" the murderer and the hurricane. But just as it would be incoherent to "blame" the hurricane (in a certain sense), it is similarly incoherent to blame the murderer in this sense (assuming determinism is true).

In other words, the important point, at least for me, is that I (used to) feel differently about the murderer and the hurricane. There is a certain range of reactive attitudes (of which a certain flavor of blame is one) that I would never be tempted to ascribe to the hurricane. And, on reflection, the reason I wouldn't be tempted to "blame" a hurricane (or a child, or a person with a brain tumor, or an animal, or a killer robot), really does have something to do, at bottom, with the ability to do otherwise. When I reflect on my intuition, I feel like the hurricane "couldn't help it." (Of course, I might still incarcerate the murderer because the probability of her committing another murder conditional on her having murdered once is unacceptably high; or to for more general deterrence.)

But I think I've been programmed by natural selection to think the murderer could help it. And I now think that this is just a defect in my programming. This might not be a "defect" in every sense; maybe society functions better if we all go around thinking that the murderer could help it. But in true rationalist spirit, I'm interested in whether the intuitive feeling I have that there is a meaningful difference in blameworthiness between the murderer and the hurricane is well justified. We can argue about the definition of "blame" in much the same way you can argue about how to define "free will" and "moral responsibility." That's why I'm trying to point to the difference in our intuitive (moral?) judgments between the murderer and the hurricane to get at whether there is a real difference there or not.

Now, does any of this make a difference to how I feel or act? Actually, yes. Since I've started thinking this way, I'm less hard on myself. In the past, I might have looked at a wildly successful person and felt bad about the comparatively little I've managed to achieve. But now, even if I might still be instinctively inclined to feel that way, I reflect: It makes no sense to feel bad because everything I've done (and everything the wildly successful person has done) is just the product of genes and environment. Our brains made us do it. Similarly, if I feel proud about what I've achieved, I think to myself: I can't really take credit for any of this. My brain made me do it.

I'm less likely to get (or maybe just stay) angry with people in my life. If John does something that makes me mad, I genuinely stop and say to myself "John couldn't help it." I might still be frustrated in the same way I might be frustrated if a person hit me as the result of an involuntary spasm. Again, there are some negative reactive attitudes that I would feel quite justified in feeling toward any hardship imposed on me via an involuntary mechanism. But there is a certain kind of anger I (used to) feel toward competent adults (rooted, I think, in the sense that they chose their own actions) but I'm less likely to feel now.

I'm also more compassionate and empathetic. If someone wrongs me, I think to myself that it is literally true that if I was in their position, I would've done the exact same thing.

These are all positive changes, but I won't deny that there are also negative things about this worldview. For instance, certain types (though not necessarily all types) of gratitude seem not to make sense in a deterministic universe. And yes (because I might as well go here first), I feel less negatively toward Hitler than I would if I felt like Hitler made voluntary choices.

I'm not interested in which set of beliefs would be instrumentally useful (or otherwise "better") for me or for society to adopt. I'm interested in what's actually true with respect to the question whether my instinctive reactions that there is a meaningful sense of (e.g.) blame that it makes sense to ascribe to a competent adult but not to something that manifestly lacks agency.

I haven't fully internalized this worldview, and I doubt I ever could. Some of our hard-wired instincts are impossible to fully overcome. I haven't fully grappled with what it would mean to live according to the principle that we are all meat robots.

But I'd still like to know: Have I missed something? I take it that at least some compatibilists want to say that the instinct we have to feel differently about the murderer and the hurricane is well justified. If you think so, please explain where I've gone wrong!

A final (and somewhat disconnected) point that may serve as an interesting focal point for disagreements between compatibilists and incompatibilists: I think on my view, I basically have to accept the Minority Report thought experiment in which we punish people before they've actually done anything wrong. Assuming we can unerringly predict whether they will do a crime or not, there's no difference between ex ante or ex post (except that if we wait for ex post, there's a crime victim; so we are probably required to act ex ante). Of course, it's a weird though experiment because it assumes perfect foreknowledge, etc. But even on this framing, would a compatibilist disagree that we can punish the criminal before the crime?

The original post is preserved below.

****

I believe in determinism and think (a la Sam Harris) that this means that my will is (at least in some important sense unfree) or (what amounts to the same thing) that I'm not morally responsible for my actions. But I take it from reading posts on LW that most people around these parts, following EY, are compatibilists about free will. I'm trying to understand compatibilism better, so I have some questions I hope someone can answer.

First, how would a compatibilist explain why the mentally insane (or hypnotized etc.) are not morally responsible? Consider two people, Smith and Jones, who are both murderers. Smith has a brain tumor such that he couldn't have done otherwise; the brain tumor "made" him do it. Jones is a "normal" person. I think that all would say that Smith isn't morally responsible for the murder. If asked to explain why, I think my explanation would invoke something like Smith's inability to do otherwise. "He had no other choice," etc. Why is Jones different? If determinism is true, it seems to me that his brain "made" him do it just the same as Smith's did. So how would a compatibilist differentiate between these cases?

Second, would a compatibilist think that a computer programmed with a chess-playing algorithm has free will or is responsible for its decisions? It evaluates a range of options and then makes a decision without external coercion. I think humans are basically just computers running algorithms programmed into us by natural selection. But I also think that the computer lacks responsibility in the sense I care about. If the computer lost a game, I don't think it would make sense for me to get angry with the computer in much the same way that it wouldn't make sense for me to get mad at my car for breaking down. Again, if asked to explain why, I think I would say something like "the computer couldn't help it." Would a compatibilist agree?

Third, and related to the second question, how about animals? Do they have free will? Is my dog in any sense "responsible" for peeing on the carpet? Does it make sense to say that the bear "freely chose" to eat the hiker? If not, what makes humans different?

Fourth, does it ever make sense to feel regret/remorse/guilt on a compatibilist view? Suppose I read a bunch of posts on LessWrong and think that preventing the robot apocalypse is the most important thing I could be working on. But I like reading novels, so I decide to get a PhD in English. I then think to myself "I really should've studied computer science instead." And so, I tell my friend that I regret choosing to study English instead of computer science. But my friend responds, "Your beliefs/desires, and ultimately your choice to study English, was determined by the state of your brain, which was determined by the laws of physics. So don't feel bad! You were forced to study English by the laws of physics."

Is my friend wrong? If so, why? At least intuitively, when I regret something, it seems to be because I feel like I could've done something differently. If someone held a gun to my head and forced me to study English, then I wouldn't regret studying English because I had no choice. But if determinism is true, it seems to me I had no more "choice" in studying English than I would have if someone was holding a gun to my head. And it seems to me that this logic suggests that regret is always illogical.

Finally, I think at least some versions of compatibilism want to distinguish between "internal" motivations and "external" motivations. Why is this difference important? If determinism is true, then it seems to me that I was just as constrained by the laws of physics (and the circumstances of my birth, etc.) to do X as if there were some external force requiring me to do X.