Do You Have Free Will?

Writing a review is an exercise in free will. Not only can I tell you what I want about the book and whether I liked it or not, but I also get to choose how to begin. If I decide to start with a personal anecdote, that’s what you will get. And I have the ability—the freedom—to start in other ways instead. These facts may seem too obvious to mention. But they are denied by Robert Sapolsky, a professor of biology at Stanford whose new book, Determined, argues, “We have no free will at all.”

The challenge to the cherished notion of free will comes from what philosophers call “causal determinism.” This is the idea that everything that happens is the product of prior causes, stretching back into a past that was not up to us. We do not originate our choices ex nihilo; instead, they are determined by our history. As Sapolsky puts it, bluntly:

The upshot, for him, is that “there can be no such thing as blame, and that punishment as retribution is indefensible.” It’s a shocking conclusion. Imagine a murder committed in cold blood by a ruthless killer, pursuing personal gain; the murder is premeditated, carefully planned, entirely in character. Now imagine that the victim is someone you love. For many of us, this scenario, even when hypothetical, provokes feelings of resentment and blame, a desire to punish the killer. For Sapolsky, none of these responses can be justified.

Determined – A Science Of Life Without Free Will

By Robert M. Sapolsky

In making his case, Sapolsky distances himself from a recently fashionable critique of free will inspired by the neuroscientist Benjamin Libet, whose brain-imaging studies suggest to some that our “decisions” are epiphenomenal, a superficial side effect of the real decisions made by the unconscious brain. Sapolsky is not convinced. He believes that our intentions make a difference in the world. What concerns him is instead a question he sets in italics, and asks more than once: “Where did that intent come from in the first place?” What Sapolsky argues, in hundreds of pages of neurobiology, genetics, Darwinian selection, chaos theory, and quantum mechanics—all explained with diagrams and effervescent prose—is that, however unpredictable our actions are in practice, our intentions are caused by factors that were caused in turn by conditions that existed before we were even born. Ergo, free will is a myth.

That is Sapolsky’s argument in a nutshell. And if the last step—from determinism to the total absence of free will—went by quickly in my telling, it goes by very quickly in his book. Before he gets to the evidence for determinism, Sapolsky spends about a page on what it would mean to say that we are free: “Show me a neuron (or brain) whose generation of a behavior is independent of the sum of its biological past,” he writes, “and for the purposes of this book, you’ve demonstrated free will.”

Now, it’s no surprise that, if you define free will as a violation of determinism, the truth of determinism is not compatible with free will. What’s frustrating is that Sapolsky knows that the majority of philosophers—he himself estimates 90 percent—do not accept that definition. They are what are called “compatibilists,” thinkers who defend the existence of free will not by denying that events outside of our control determine our actions, but by giving philosophical accounts of freedom, blame, and punishment that don’t require Sapolsky’s miraculous neuron: accounts that are compatible with determinism.

Perhaps the most influential compatibilist in history is the Scottish philosopher David Hume, who helped invent the science of the mind. “By liberty,” he wrote in 1748, “we can only mean a power of acting or not acting, according to the determinations of the will; that is, if we choose to remain at rest, we may; if we choose to move, we also may.” Hume called this “hypothetical liberty” because it has an “if-then” structure. If I choose to type this sentence, then I’ll press the appropriate keys and the words I’ve decided to type will appear on my computer screen. If I decided to type other words, my fingers would move differently. For Hume, that’s all it takes for me to have the freedom, or liberty, to do otherwise: My actions depend on my decisions. This is an instance of causality at work, not a violation of causal law. It involves the determination of what I do by what I intend, given a hospitable environment, not a failure of determinism. Nor does it turn on the prior history of my intentions. If Hume is right, Sapolsky’s question—“Where did that intent come from in the first place?”—is irrelevant to free will.

Sapolsky doesn’t mention Hume or the many philosophers influenced by him. (The closest we get is a reference to the 20th-century compatibilist Peter Strawson that confuses him with Galen Strawson, his incompatibilist son.) Nor does Sapolsky engage with the idea of hypothetical liberty: If I decide to do A, I’ll do it; if I decide not to, I won’t. So we have to read between the lines.

Read: To get happier, choose to read this column

To be fair to Sapolsky, he is not alone in giving short shrift to the sort of freedom that is compatible with determinism. In his 1788 Critique of Practical Reason, Immanuel Kant called this conception of liberty a “wretched subterfuge … the freedom of a turnspit.” (A turnspit was a wind-up rotisserie.) But Hume’s interpretation of free will has real force. When we talk in ordinary terms about what we are free, or able, to do, we aren’t talking about failures of determinism or violations of causal law. For all I know, it’s physically possible for me to vanish in a fluke of quantum mechanics. But if I told you “I am able to vanish,” I’d be lying. Hume’s sensible thought is this: Freedom means being able to perform an action if and when I decide to try. That’s why the fact that I would never choose to run a marathon doesn’t mean that I’m not free to run one. So long as “running” very slowly counts, I’m pretty sure I could.

When philosophers doubt that hypothetical liberty is liberty enough—and many do—they point to cases of addiction or compulsion, in which one’s decision is caused by an urge so powerful, it diminishes one’s freedom. If the alcoholic is compelled to choose another drink, he’s not free to go sober. Hume is therefore wrong, the reasoning goes: Even if the alcoholic would refrain if he decided to—he enjoys hypothetical liberty—the fact that he’d never make that decision, because of factors beyond his control, means that he lacks free will. Sapolsky would agree: The relevant question is not whether the addict’s decision is efficacious, or whether he would act differently if he decided otherwise, but why he makes the decision he does. If addiction is the cause, he is not free.

Ironically, the neuroscience in Sapolsky’s book suggests that the truth about compulsion may support Hume’s view. The prefrontal cortex (PFC) plays a central role in the regulation of behavior, our power to resist temptation. As Sapolsky writes, the efficacy of one’s PFC is sensitive to one’s upbringing and to temporary influences of all sorts. This plays out in the exercise of self-control. Sapolsky gives a low-key example: “Place a bowl of M&M’s in front of someone dieting. ‘Here, have all you want.’ They’re trying to resist. And if the person has just done something frontally demanding”—which exhausted their prefrontal cortex—“the person snacks on more candy than usual.”

What Sapolsky doesn’t note is that his dieter is unable to resist, and lacks the freedom to do so, in Hume’s sense. For Hume, I am free to perform an action if and when I would successfully perform it if I decided to try. The dieter tries to resist—they’ve decided not to snack too much—but their decision isn’t efficacious: They act against it or change their mind. What they lack is hypothetical liberty, the “power of acting or not acting, according to the determinations of the will.” In other words, we don’t need to ask about the origins of their intentions—Sapolsky’s question, “Where did that intent come from in the first place?”—in order to explain why the dieter lacks free will. They lack free will because their decision not to snack is ineffective. Hume’s analysis gets this right. To generalize from here: The inability to resist temptation characteristic of addiction, and dependent on the PFC, is not a problem for Hume’s view but an illustration of it. When the alcoholic cannot help but drink, he’ll end up drinking even if he intends to stay sober. His intention not to drink won’t be effective. That’s why he doesn’t have the freedom not to drink.

At other times, we manage to resist temptation. Sapolsky quotes the psychologist James Cantor on the neurobiology of pedophilia: “One cannot choose to not be a pedophile, but one can choose to not be a child molester.” Sapolsky finds this position absurd: If pedophilic desire is out of our control, so is the pedophilic action that results from it. But Cantor’s claim makes sense. Deciding not to be a pedophile won’t eliminate pedophilic desire. But deciding not to act on that desire might work. That depends on one’s biology and environment, including one’s level of self-control. It’s true that one’s strength of will is not, itself, under one’s direct control. But so what? If one has sufficient willpower, one can exercise it. If one doesn’t, one’s freedom is diminished—not by determinism or the fact that one’s decision has prior causes, but by a lack of hypothetical liberty.

If we think of freedom as the ability to do otherwise, understood in Hume’s way—if I decide to do A, I will; if I decide otherwise, I won’t—then it is perfectly compatible with determinism: Freedom turns on how effective we are in executing our intentions, no matter what caused them.

But Sapolsky has a second argument, which is more about morality than the nature of free will. How can it be fair to blame or punish someone who acts wrongly if their doing so is a consequence of factors beyond their control? Something—some combination of genetics, upbringing, and environment—made that ruthless murderer ruthless, the sort of person who plans a murder and is able to follow through. If we can’t blame him for those past causes, how can we blame him for what he does? How can we justify the punishment we so desire?

When it comes to blame, I think the rhetorical question can be answered. Let’s grant, as Sapolsky does, that your character is fixed by facts that are not up to you. Still, when you act with indifference to the rights and needs of others, we can blame you for what you do—unless you have a good excuse. What counts as an excuse is a question of morality, not metaphysics. We excuse wrongful action when it’s the product of manipulation or coercion, when you don’t know what you’re doing (provided your ignorance is not willful or negligent), and perhaps when you’re unable to do otherwise. These excuses mitigate blame by showing that your behavior is not reflective of your moral character; whether or not you are to blame for the character you have is irrelevant. Nor is there pressure to acknowledge an additional excuse—that one’s behavior is determined by the past—because we do not need it to account for the excuses we accept in everyday life.

It’s ironic, again, that Sapolsky’s reasoning tends to support, not undermine, this view. For Sapolsky, “all that came before, with its varying flavors of uncontrollable luck, is what came to constitute you.” But if my present psychology, even if formed by forces outside my control, constitutes me, and my behavior reflects who I am—not ignorance or compulsion—then it issues from me and I should take responsibility for it. My wrongdoing is expressive of my moral character and therefore subject to moral blame. What is my excuse?

Read: Why TV is so worried about free well

Sapolsky goes on to draw a comparison between his project and the way in which medical conditions like epilepsy and schizophrenia, along with the behaviors they cause, came to be exempted from blame. We learned to say, “It’s not him. It’s his disease.” Which is progress. But these diseases interfere with the efficacy of one’s intentions, that hypothetical liberty, and with what Sapolsky elsewhere calls “the consistency of behavior that constitutes our moral character.” When it’s me, not my disease, acting with the consistency of behavior that constitutes my character—again, not out of ignorance or compulsion—I don’t see what gets me off the hook.

Punishment is something else. Sapolsky is appalled by our ruthless urge to see the guilty suffer. “If there’s no free will,” he writes, “there is no reform that can give retributive punishment even a whiff of moral good.” Hence his public-policy proposal: to replace the punitive carceral system with “quarantine,” the comfortable confinement of those who are a danger to others, until they aren’t.

But there are problems here. First, Sapolsky’s view implies that the perpetrator of a one-off crime—so long as we are sure it’s one-off—should go scot-free: If there’s no risk that they’ll reoffend, then there’s no benefit in quarantining them. Second, his view neglects the crucial role of punishment in deterrence: not physically preventing future crimes but giving us incentives not to commit them. Third, the moral challenge to retribution has nothing to do with freedom or determinism. Those who believe that punishment is of value even if nothing good will come of it simply want to see the guilty suffer. Sapolsky is one of many who recoil from this impulse. Unless it has some deterrent or preventative function, how can the suffering of the guilty make the world a better place? If you struggle to see an answer, debates about determinism won’t help.

Sapolsky has a lot to teach about the science of decision making and about empathy for the unfortunate. But Hume remains a better guide to the philosophy of free will. I suppose that verdict makes this a negative review, and I don’t feel great about that. Books require a lot of work, and authors have feelings. Still, I take comfort in the fact that, although I wrote these words of my own free will, Sapolsky doesn’t think I did—or that it’s fair to blame me for their potential ill effects.