On substack recently, free will has been all the rage, both in the sense that people have been using it—as they always do—and that they’ve been writing about it. Since Sam Harris—a prominent opponent of the idea that we have free will—joined substack, quite a few people have been arguing that we don’t have free will. In this post, I’m going to argue that we have free will. To be honest, I find it to be a sort of straightforward issue.
Here’s one important feature of good philosophy: it shouldn’t deviate from common sense unless it needs to. It is a cost to David Lewis’s theory, for instance, that he says that every possible world concretely exists, that there are infinite spatiotemporally disconnected leprechauns. Similarly, it’s reasonable to believe in moral realism because it is intuitive that certain things are wrong. An objection—indeed, one of the only half-decent ones—to the self-indication assumption is that it implies some very weird things about the world.
It makes sense to believe that there’s an external world, even if one doesn’t have arguments to refute solipsism. This is because it sure seems like there’s an external world, and one should believe things that seem obvious unless given a good reason to doubt them.
But if this is true, then there’s at least a very strong case in favor of free will. As even the opponents of free will admit, free will is intuitive. It sure seems like we are free to make various choices. For example, I think I made the choice to type this sentence, rather than a different one—there was another one that I explicitly considered, but decided against. Perhaps if given overwhelming evidence, we should abandon that judgment, but in the absence of strong evidence, we should believe it.
There’s an important sense in which I’m free to move my arm but I’m not free to, say, fly into the sky. That’s because I can type the sentence if I want, but can’t fly to Denmark if I want. Now, maybe that’s not the only necessary condition, but it’s certainly a necessary condition. I don’t intend to resolve the debates between those who think determinism eliminates free will and those who don’t—my claim is just that there’s a very intuitive sense in which many of our actions are free. Currently, I lean weakly towards libertarianism because 1) I think that compatibilism makes poor sense of our ability to grasp non-natural facts; 2) I have an intuition that we’re free in a deeper sense than that described by compatibilism, though it is a freedom of sorts; 3) it’s easier to see why there’s a lot of evil in the world if you believe in God and are a libertarian. But these are very deep views that most of you disagree with, so I won’t defend them in detail—instead, I intend to argue that one should at least be a compatibilist.
In response to the intuitive case for free will, those who deny free will tend to argue that compatibilists are redefining free will. But why? Why is compatibilist free will not a type of genuine freedom (to be clear, I don’t think it’s the only kind of genuine freedom we have, but it’s still a significant kind of freedom)? Generally, philosophers think that there are two conditions that must be met for you to be free: you have to be able to do otherwise and you have to be in control of your actions. Can compatibilism accommodate those?
It seems to me that it can. When we say “does A control B,” we mean something like “is B determined by A, such that changes in B depend on A?” But in that sense, I do control my actions. If my desires were different, my actions would be different. My desires are causally responsible for my actions
You might worry that this usurps control in some important sense. If what ultimately determines our actions are the initial conditions of the universe, then it’s hard to see how we’re in control. But crucially, if we’re part of the causal sequence, it can be both us and the initial conditions of the universe that control some event. For example, suppose there’s a completely deterministic universe that gives rise to a drone controller—it moves the drone in accordance with the buttons pressed. Well then, there’s an important sense in which the drone controller controls the drone. The causal history of how the drone came to be shouldn’t matter to that.
Now, can one do otherwise in a deterministic universe? It seems to me that the answer is yes. When we ask if a person could do otherwise, I take this to be about whether they could do otherwise if they wanted to. That’s why, for example, I can’t fly but I can write a different sentence. But if one wanted to do otherwise, they would do otherwise. Our desires may be determined, but our actions depends on them, so in the counterfactual sense, it’s true that we can do otherwise if we want to.
For these reasons, I think the compatibilist kind of free will is a kind of free will. It’s not the only, or even the most robust kind of free will that there is, but contrary to the free will deniers, there’s nothing illicit about it. It’s no trick. Given how strong people’s pro-free-will intuitions are, unless we have some very strong reason to abandon free will, we should hold on to it. The mere fact—if, indeed, it is a fact—that our choices are predetermined would not be sufficient reason to usurp the notion that we are, in an important sense, in control.
Furthermore, it’s certainly not true that science has established that there is no free will. I won’t address in detail all of the scientific evidence people levy in favor of free will, but it’s most commonly the Libet experiment. Libet found that a sizeable share of the time, before people make a decision to move their finger, one can detect it in their brain. This indicates, according to some, that the choice is dictated by previous brain behavior. But here’s an equally good explanation of the data: there are neural correllates of us having an urge to take an action. Whether we follow that urge is, of course, up to us. Libet didn’t find that whether one would move their finger could be perfect predicted, just that it could be predicted some of the time. But this is what those who believe in free will would expect.
Therefore, both compatibilism and libertarianism are very live options. The least plausible view of free will is the one that denies the obvious data, the strong intuition that we have free will. It’s no surprise that the very substantial majority of philosophers reject this view, according to which we are effectively puppets, no freer than Kermit. One needs very strong arguments to overturn a very strong intuition, and merely gesturing vaguely in the direction of science does not provide such an argument.
Two general points:
1) It seems like we shouldn't treat our intuitions the same across different spaces - especially if we know where that we are likely to develop bad intuitions for, say, evolutionary reasons (think scope neglect), there may be reasons to flat-out reject them. We must understand that intuitions come about via a process of empirical stuff over the course of time in an evolutionary process and can't treat them any differently because they "feel different."
2) It seems like your approach is more frequentist than Bayesian (not saying that there is anything wrong with that, I'm merely pointing it out). This is the case, in that, the initial intuition serves as a sort of null hypothesis as opposed to perhaps just a probability (and given the fact that you have no information, you may use some indifference principle to equally distribute credences among the partition). I would say, however, that it seems like if you do use this approach towards the skeptical problem, perhaps you should use it more consistently across all of epistemology (as I don't see why the cases would be so different in method).
Under compatibilism, does ChatGPT have free will?