I’ve mentioned before Robert Sapolsky’s recent book Determined: A Science of Life Without Free Will, a 528-page behemoth that at times is a bit of a slog and at other times an inspiration. (See here, here, here, and here for previous posts about it.) I found his argument against libertarian free will convincing, but of course I already believed that there is no good argument for libertarian (“you-could-have-done-otherwise”) free will (LFW), so I was on his side from the outset. I’m a hard determinist, and that’s based on seeing that the laws of physics obtain everywhere. But people are still maintaining not just that we can confect some form of free will despite the truth of determinism (these people are called “compatibilists”), but that we have real libertarian free will. They are wrong.
The video below, arguing for LFW, came in an email from Quillette touting their most popular articles of 2024. But this was a short (4.5-minute) video, not an article, and I don’t think the video was one of the top items. Perhaps the note referred to a Quillette article by Stuart Doyle (below) on which the video is based, but that article was published in 2023.
At any rate, listen to the video first, and then, if you want to see what I consider an unconvincing argument against free will (though it does make some fair criticisms of Determined), click on the headline below to read Doyle’s argument that we have “not disproven free will.”
The narrator of the video isn’t named, but she pretty much parrots what’s in Doyle’s essay, emphasizing an argument for free will that Doyle considers dispositive, but to me seems irrelevant.
You may notice some problems with the “rebuttal” described in the video. For example, it seems irrelevant to argue that “just because a neuron doesn’t have free will doesn’t mean that the bearer of a collection of neurons (a person) doesn’t have free will.” This is an argument that the emergent property of LFW can still appear even if neurons themselves behave according to physical law (a large argument in Sapolsky’s book). Also, if quantum physics is truly and fundamentally unpredictable (and we don’t know this for sure), that itself, says the narrator, poses a problem for free will, because it means that, at any given moment, a quantum event may change your behavior.
There are two problems with the quantum-indeterminacy argument. First, nobody ever maintained that quantum events like the movement of an electron can result from one’s volition (“will”), so unpredictability at a given moment does not prove volition. Further, we don’t even know (and many of us doubt) that a quantum event can change human behavior or decisions on a macro level. Some people have calculated that it can’t. So the whole issue of quantum unpredictability is irrelevant to the main problem: whether, at a given moment, you can, through your own agency, have behaved or decided differently.
This brings up the problem of predictability. The narrator’s (and Doyle’s) argument is that if you cannot predict someone’s behavior or decision—even with perfect knowledge of everything—then we have free will. As I just said, quantum physics may cause such fundamental unpredictability, but doesn’t support the notion that we have LFW Yet the video and Doyle suggest there is another form of fundamental unpredictability that can cause a lack of predictability despite perfect physical knowledge: computational undecidability. Both the narrator and Doyle accuse Sapolsky of complete ignorance of this concept, which, they say, constitutes “a major flaw in Sapolsky’s argument.” The narrator says that if human behavior is fundamentally unpredictable, then it supports the idea that free will exists. The premise of this criticism is, of course, is that if you can’t predict human behavior and decisions, even with perfect physical knowledge, then you can’t say that we lack free will. But these arguments using predictability are flimsy arguments against determinism, and, in fact, we’ll never have the perfect knowledge we need to predict behavior.
The problem is that quantum mechanics can in principle wreck perfect predictability of behavior, but that possibility doesn’t support free will. So does “computational undecidability”, another thing that impedes prediction, leave room for free will? I don’t think so (see below).
The essay by Stuart Doyle on which this video is based can be accessed by clicking the link below, or you can find it (archived here). Doyle is a graduate student in psychology at the University of Kansas.
Let me start by saying that Doyle’s essay, while it makes its points clearly and strongly, seems almost mean, as if Doyle takes great joy in telling us how stupid Sapolsky is. And this is coming from someone (me) who’s been accused of the same thing. (I plead not guilty, at least for my published work.). But for a scholar publishing a rebuttal on a major site, it seems to me uncharitable to say stuff like this:
Sapolsky’s conclusions about morality and politics stand on nothing beyond his personal tastes. His book was marketed with such authoritative headlines as “Stanford scientist, after decades of study, concludes: We don’t have free will.” In contrast to the hype, Determined is ultimately a collection of partial arguments, conjoined incoherently. And Robert Sapolsky is to blame.
Sapolsky is to blame? Well, yes, of course he is, he’s the author, but the concept of blaming someone for writing a book they don’t like, and and accusing them of incoherence (I disagree) is not civil discourse. But let’s move on.
The observation that every object in the universe obeys physical law does directly imply that there is no amorphous “will” that can affect the laws of physics, something that physicist Sean Carroll (a compatibilist) has emphasized. To me, this puts the onus on those who accept LFW to tell us what aspect of human volition is independent of the laws of physics.What form of nonphysical magic can change the output of our neurons? So far, nobody has done this. Thus, to a large extent, I think, one can tentatively accept determinism simply from knowing that every physical object obeys well-known laws and, as Carroll has written, “The laws underlying the physics of everyday life are completely understood.” Carroll:
All we need to account for everything we see in our everyday lives are a handful of particles — electrons, protons, and neutrons — interacting via a few forces — the nuclear forces, gravity, and electromagnetism — subject to the basic rules of quantum mechanics and general relativity. You can substitute up and down quarks for protons and neutrons if you like, but most of us don’t notice the substructure of nucleons on a daily basis. That’s a remarkably short list of ingredients, to account for all the marvelous diversity of things we see in the world.
So yes, Carroll is a determinist in a way that refutes libertarian free will, but in the link saying he’s a compatibilist, you’ll see that he says that we have a sort of free will instantiated in the emergent properties of humans acting as agents and expressing preferences. (Of course our tastes and preferences are also formed in our brain by the laws of physics.) Well, there is no real emergence that defies the laws of physics: emergence may not be predictable from lower-level phenomena, but it is consistent and derives from lower-level phenomena. Saying, as Doyle does, that “The ‘mechanism’ that produces deliberative choices is the whole person” is to say nothing that refutes determinism.
As I reread Doyle’s paper, I realized that although he does point out some contradictions in Sapolsky’s arguments, Doyle does nothing to dispel determinism. What appears to be the central contention of his essay is that there is another way that physical objects can behave unpredictably beyond quantum mechanics, and that way is computational decidability. But that supports LFW no more than does any unpredictability of quantum mechanics.
Here’s what Doyle says:
So what could give us the ability to surprise Laplace’s demon? Computational undecidability. This is a term describing a system that cannot be predicted, given complete knowledge of its present state. This fundamental unpredictability shows up in algorithmic computation, formal mathematical systems, and dynamical systems. Though an unpredictable dynamical system may evoke the concept of chaos, undecidability is a different sort of unpredictability. As described by one of the greatest living information theorists, C.H. Bennett:
For a dynamical system to be chaotic means that it exponentially amplifies ignorance of its initial condition; for it to be undecidable means that essential aspects of its long-term behavior—such as whether a trajectory ever enters a certain region—though determined, are unpredictable even from total knowledge of the initial condition.
If a system exhibits undecidability, then it is unpredictable even to Laplace’s demon, while a system that is merely chaotic is perfectly predictable to the demon. Chaos is only unpredictable because the initial conditions are not perfectly known. So it would be fair to dismiss that kind of unpredictability as mere ignorance—an epistemological issue, not an ontological reality. But the delineation between the epistemic and the ontic falls apart when we talk about what Laplace’s demon can’t know. An issue is “merely” epistemological when there is a fact of the matter, but the fact is unknowable. There actually is no fact about how an undecidable system will behave until it behaves. For a fact to exist, it must be in reference to some aspect of reality. But nothing about present reality could ground a fact about the future behavior of an undecidable system. In contrast, the exact actual state of present reality grounds facts about the future of chaotic systems. We just can’t know the exact actual state of present reality, thus unpredictability is “merely” epistemological in the case of chaos, but not in the case of undecidability.
Arguably, human behavior is undecidable, not just chaotic. And that would mean that human choice is free in exactly the way we’d want it to be; determined—by our own whole selves, with no fact of the matter of what we’ll choose before we choose it. But Sapolsky seems unaware of undecidability as a concept. He mislabels cellular automata as chaotic, rather than recognizing the truth that they exhibit undecidability. This is a major factual error on Sapololsky’s part.
First of all, from what I’ve read of computational undecidability, it is a phenomenon not of physical objects, but of philosophy combined with mathematical concepts and models. As Wikipedia says (and yes, I’ve read more than that article):
There are two distinct senses of the word “undecidable” in contemporary use. The first of these is the sense used in relation to Gödel’s theorems, that of a statement being neither provable nor refutable in a specified deductive system. The second sense is used in relation to computability theory and applies not to statements but to decision problems, which are countably infinite sets of questions each requiring a yes or no answer. Such a problem is said to be undecidable if there is no computable function that correctly answers every question in the problem set. The connection between these two is that if a decision problem is undecidable (in the recursion theoretical sense) then there is no consistent, effective formal system which proves for every question A in the problem either “the answer to A is yes” or “the answer to A is no”.
Two points here. First, Doyle gives not one example of a biological system in which “computational undecidability” would obtain. If there was one, why didn’t he mention it? It seems to me solely a mathematical/logical concept, and my (admittedly cursory) readings have turned up nothing in biology or physics that seems “computationally undecidable”, much less in a way that would give us free will.
Second, even if there is a fundamental and non-quantum form of unpredictability in physics and biology, that doesn’t open up the possibility of free will. That would depend on whether our “will” could, in some non-physical way, affect the behavior of molecules. If it cannot happen with quantum mechanics, then how can it happen with computational undecidability? Unless Doyle tells us how this mathematical/logical idea can somehow affect our behavior according to our “will”, he has no argument against determinism and thus has no argument for free will.
Now it’s true that belief in “physical determinism—folding into that term quantum and other unpredictable effects not affected by our volition)—is largely a conclusion from observing nature. But just because we cannot absolutely prove determinism of behavior from science, we can still increase determinism’s priors by various experiments. These include recent studies showing that you can predict, using brain scanning, binary decisions that people make before they are conscious of having made them. For example, if people are given a choice of adding or subtracting two numbers, scanning their brains shows that you can, with substantial probability (60-70%), predict whether they’ll add or subtract up to ten seconds before they are conscious of having made a choice. And this is from crude methods of measuring brain activity (e.g., fMRI). Perhaps by measuring individual neurons or groups of neurons we could predict even better. But the experiments so far imply that decisions are made before people are conscious of them, and that raises the Bayesian priors that people’s behaviors are determined by physics, not by their “will”.
And there are various other experiments showing that you can both increase or decrease people’s sense of volition. Electrical stimulation of the brain can make people think that they made a decision when in fact it’s purely the result of stimulating certain neurons. This causes people to make up stories of why they did things like raise their hand when a part of their brain is stimulated (“I decided to wave at that nurse”). But that sense of volition is bogus. This kind of post facto confabulation, which occurs very soon after you decide something or do something, is what makes us think what we have LFW. Further, there may be evolutionary reasons why we think we have libertarian free will, but I won’t get into those. Suffice it to say that I think that our feeling of having LFW is merely a very powerful illusion—an illusion that may have been installed in our brains by natural selection.
On the other hand, you can make people think that they didn’t have volition when in fact they did. A Ouija board is one example: people unconsciously move the “cursor” around to make words when they think that it’s moving independently of their will. There are other experiments like these, all showing that you can either strengthen or weaken people’s sense of volition and will using various psychological tricks. And they all go to refute the idea of libertarian free will
So yes, I think Sapolsky is right. His determinism agrees with nearly all the scientists (including compatibilists) who think that the notion of libertarian free will is bogus. To think otherwise is to believe that there is some kind of non-physical mental magic that can change the laws of physics.
One final point. Arguments about free will are not just philosophical wheel-spinning, for they play directly into an important part of society: reward and punishment—especially punishment. If the legal system truly embraced determinism of behavior, we could still have punishment, but it would be very different. We would punish to keep bad people off the streets, to give people a chance for rehabilitation (if they can be rehabilitated), and to deter others. But what we would not have is retributive punishment: punishment for having made the wrong choice.
Legal systems are grounded on the notion that we are morally responsible, but under determinism we’re not. Yes, we can be responsible for an act, but “moral” responsibility is intimately connected with libertarian free will; it’s the idea that we have the ability, at any given time, to act either morally or immorally (or make any any other alternative decision, even if it doesn’t involve morality). Yes, I know there are some who think that the justice system already implicitly accepts determinism, but they are wrong. For if it did, we wouldn’t have any form of retributive punishment, including capital punishment.
As for rewarding good behavior, well, yes, you couldn’t have done otherwise than, say, saved a drowning person. But rewarding people who do good is a spur for other people to do good. Even if the rewarded people don’t “deserve” plaudits in the sense that their accomplishments didn’t come from LFW, handing out rewards for things that society approves of is simply a good thing to do—for society.
Oh, a p.s. Because people feel so strongly that they do have libertarian free will, I have faced more opposition when touting determinism than when touting the truth of evolution. As I always say, “It’s much harder to convince a free-willer of the truth of determinism than to convince a creationist of the truth of evolution.” People feel so strongly that they have LFW that I have suffered two unpleasant consequences for touting determinism. I’ve told these stories before, but a big jazz musician nearly attacked me for implying that his solos were not truly extemporaneous, and that he could not have played a different solo, and on another occasion an old friend kicked me out of his house because he couldn’t abide the notion of determinism. No creationist has ever treated me in those ways!







