Feel Free

Determined:
A Science of Life without Free Will
by robert m. sapolsky
penguin, 528 pages, $35

Many people have been amazed by the capabilities of ChatGPT and the rapid advances in artificial intelligence. But something even more remarkable has now appeared: a book, described by its publishers as “plumb[ing] the depths of the science and philosophy of decision-making,” that claims to have been written entirely by a machine. Strange as it may sound, the book’s author—one “Robert M. Sapolsky”—clearly implies that he (or it) is a machine made of biological materials that, though possessing consciousness, sensations, and even emotions, utterly lacks free will. Indeed, the thesis of the book, Determined: A Science of Life without Free Will, is that free will is a “myth” and that all the thoughts and actions of human beings, just like those of Sapolsky, are completely determined in a mechanistic way by physical causes that are now well understood through neuroscience. And Sapolsky has not only produced an acclaimed book, but for decades has been doing cutting-edge research in neuroendocrinology, reaching the highest levels of attainment in that field, and currently holds the position of John A. and Cynthia Fry Gunn Professor at Stanford University.

Sapolsky’s book has several real merits. It is highly informative, especially on the topic of neuroscience, and is written in a lively colloquial style, by turns expository, humorous, sarcastic, and passionately moralistic. And it gives clear and forceful expression to an increasingly widespread point of view on a fundamental issue. The book suffers, however, from a grave defect: It tries to refute the reality of free will without a clear enough idea of what free will is, or at least what it traditionally was understood to be. Its text makes no mention of Aristotle, Aquinas, or the other great philosophers who shaped two millennia of thought on the subject. Instead, here is all Sapolsky has to say by way of definition:

People define free will differently. Many focus on agency, whether a person can control their actions, act with intent. Other definitions concern whether, when a behavior occurs, the person knows there are alternatives available. Others are less concerned with what you do than with vetoing what you don’t want to do.

Sapolsky feels no obligation to be more exact. For him, all aspects of human thinking and willing are entirely determined by the push and pull of physical causes, as he believes we can prove from neuroscience and physics. Yet to evaluate Sapolsky’s conclusions, we need to be precise about those aspects of human thinking and willing which the great philosophers brought to the fore. And the word Sapolsky does not mention in that paragraph is one that they would have emphasized: namely, reason.

According to Aquinas, following Aristotle, what distinguishes the human will and makes it “free” is precisely that it is rational. By contrast, Aquinas writes, other animals use no such faculty: “The sheep, seeing the wolf, judges it a thing to be shunned, from a natural and not a free judgment, because it judges, not from reason, but from natural instinct.” Of course, many subrational factors are involved in human choices. Aquinas mentions “habits and passions” as well as “temperament or disposition due to any impression whatever produced by corporeal causes.” But all of these, however refractory they may be, are ultimately, in human beings, “subject to the judgment of reason.” Whereas a nonrational animal instinctively judges a tasty fruit to be good to eat, our reason may tell us, after scientific investigation, that it contains a deadly chemical. And though we, like other animals, may have a natural urge to harm those who have harmed us, our reason may inform us of the obligation expressed by the general precept “love your enemies.”

A crucial point is that human reason is open to general abstract truths, both speculative and moral. It could not be open to such truths, which lie above it, if it were under the complete control of factors below it, such as physical and chemical forces, sensations, emotions, and the like. We must be free in this way, or we could not be rational. The great twentieth-century mathematician Hermann Weyl noted that even in the realm of mathematics, “[there must be] freedom in the theoretical act of affirmation and negation. When I reason that 2 + 2 = 4, the judgment is not forced upon me by blind natural causality.” Of course, a machine that blindly follows some algorithm might do correct calculations, such as 2 + 2 = 4, but it does not recognize mathematical truth; rather, its programmer did. The machine could just as well be programmed to output 2 + 2 = 17 and be none the wiser. A machine cannot grasp the meaning of abstract propositions, let alone judge their truth.

According to one line of thinking, which goes back at least to Aristotle, the power of abstract thought transcends the capacities of matter and therefore involves something more than the activity of a bodily organ. And that is why the human intellect and will, being rational, have traditionally been said to be immaterial “spiritual” powers. The Catechism of the Catholic Church, for example, puts it this way: “The human person, with his openness to truth and beauty, his sense of moral goodness, his freedom and the voice of his conscience . . . discerns signs of his spiritual soul, [which is] ‘irreducible to the merely material.’”

Sapolsky would reject all this, of course. I say would, because there is no sign in his book that he is aware of this entire philosophical tradition. In any event, he seems to think that a serious philosophical analysis of human thought and action is superfluous, since neuroscience can settle the question of free will on its own. 

By far the most popular neuroscience argument against free will comes from some famous experiments done by Benjamin Libet in the 1980s. Libet’s experimental subjects were asked simply to move a finger at a time of their choosing. Libet found that an electrical signal in their brains, called the “readiness potential,” started to build up about half a second before they were consciously aware of having made a decision. This fact was interpreted to mean that the brain had made the decision to move the finger while the subject was still unconscious of it, meaning that the decision could not have been free, since a free choice presumably must be conscious.

The implications of Libet-type experiments have been hotly debated by neuroscientists and philosophers for forty years. The anti–free will interpretation, however, has been largely vitiated by more recent experimental and theoretical developments, as Sapolsky admits. First, recent research strongly suggests that the “readiness potential” is not a sign that the brain has already come to a decision. Second, some Libet-type experiments have been performed that involve consequential choices—that is, choices that involve a reason to act, rather than random movement—and found no readiness potential preceding the awareness of choosing. And third, experiments have shown that a readiness potential is not strongly predictive of whether the subject will act. To quote Sapolsky: “I think all that can be concluded is that in some fairly artificial circumstances, certain measures of brain function are moderately predictive of a subsequent behavior. Free will, I believe, survives [Libet-type experiments].”

Sapolsky, however, does not base his anti–free will case just on the neural activity that occurs in the few seconds before a choice is made. Rather, he points to all the complex, interwoven chains of physical causation that led up to the choice over the course of time. “Here’s the challenge to a free-willer,” he writes:

Find me the neuron that [started the choice], the neuron that [was activated] for no reason, where no neuron spoke to it just before. Then show me that this neuron’s actions were not influenced by whether the man was tired, hungry, stressed, or in pain at the time. That nothing about this neuron’s function was altered by the sights, sounds, smells, and so on, experienced by the man in the previous minutes, nor by the levels of any hormones marinating his brain the previous hours to days, nor whether he had experienced a life-changing event in recent months or years. And show me that this neuron’s free-willed functioning wasn’t affected by the man’s genes, or by the life-long changes in regulation of those genes caused by experiences during his childhood, nor by levels of hormones he was exposed to as a fetus, when that brain was being constructed. Nor by the centuries of history and ecology that shaped the invention of the culture in which he was raised. [Emphases added]

Much of the first part of the book is taken up with detailed explanations of the mechanisms by which such factors as stress, pain, hormone levels, experiences, regulation of genes, and so forth, interact with each other in complicated ways and affect the functioning of neurons, the growth and development of the brain, and behavior.

All of which is so fascinating that one might well overlook two strange things about Sapolsky’s “challenge” to “free-willers.” First, Sapolsky challenges them to show that human beings’ free choices are not in any way “influenced by,” “altered by,” or “affected by” various physical factors. But such influences are perfectly consistent with free will, and no “free-willer” has ever doubted their reality. Many such influences are implicitly recognized by Aquinas, for example, among the “temperaments or dispositions . . . produced by corporeal causes.” Second, Sapolsky has the burden of proof backwards. One need not know exactly how free will works to have rational grounds for thinking one has it, any more than one needs to know exactly how vision works to believe that one is able to see. Rather, it is Sapolsky who has set out to prove something, namely that human thought and action are not merely influenced by physical factors but entirely “determined” by them, and to do this he has the burden of showing that no other causes are at work.

Sapolsky realizes this, and therefore several times simply asserts, apparently under the impression that he has demonstrated it, that the interwoven fabric of physical causes is “seamless.” He tells of experiments that show that judges who are hungry are “less likely” to grant paroles; that androgens in one’s fetal circulation make one “more likely” later in life to show spontaneous and reactive aggression; that stressful environments lead to increased secretions of glucocorticoids, which activate certain genes in neurons in the amygdala, making those cells “more excitable”; and many other examples of physical factors producing tendencies, correlating with behaviors, and so forth. But it is obvious that tendencies and correlations are a very far cry from a “seamless” deterministic causal account. If one is seeking a rigorous argument for physical determinism, neuroscience evidently cannot give it, at least to judge from this book. But perhaps physics can, and it is to physics (about which he candidly admits to knowing very little) that Sapolsky next turns.

For a long time, it seemed that physics did provide such an argument. In the “classical physics” that reigned from the time of Newton until the early twentieth century, all the putatively fundamental laws had equations that were deterministic—that is, in which the past uniquely and exactly determined the future. In the 1920s, however, quantum mechanics showed that the laws of physics are not deterministic. Any past state of the universe allows many possible future states, and the laws of physics determine only their relative probabilities. That revolutionary discovery eliminated the argument against free will based on the nature of physical law.

Notwithstanding this, many have argued that physics yields a kind of quasi-determinism that still outlaws free will. The point is that the effects of quantum indeterminacy tend to cancel out when systems containing many atoms are involved, just as good and bad luck tend to cancel out in a casino when many dice are rolled or cards dealt. And since even neurons contain a vast number of atoms, it is argued that quantum indeterminacy is irrelevant to how the brain functions. Sapolsky repeats this well-known argument, and it is indeed a formidable one—the only anti–free will argument in his book, I would say, that has any real force.

It should be noted that indeterminacy at the atomic and subatomic level can produce effects even at macroscopic scales. This can happen, for example, when a quantum event triggers a system that is already on the edge, so to speak, analogous to the way a single pebble can trigger a massive rockslide. Are conditions anywhere in the human brain such that the effects of quantum indeterminacy might not cancel out? The eminent mathematician and physicist Sir Roger Penrose (who shared the 2020 Nobel Prize in Physics for unrelated work) has long been pushing the idea that “microtubules” in neurons may be the place where this happens, but Sapolsky notes that the great majority of physicists and neuroscientists dismiss this proposal as highly unlikely. Interestingly, however, a paper appeared this April in the respected Journal of Physical Chemistry claiming the observation of large quantum effects in microtubules. Whether or not this claim pans out, it certainly shows that the last word has not been said in this area.

One may concede this much to free will skeptics such as Sapolsky: The traditional conception of free will raises deep questions that we are nowhere close to answering, one of them being how free will can affect physical processes in the brain, as indeed it must if our free choices affect our behavior. But this is not a sufficient reason to deny free will. For there are many aspects of the human mind that are indubitably real and that raise questions just as deep and difficult to answer. One of these is consciousness, which even most nonhuman animals presumably possess. How consciousness fits in with physical processes is utterly mysterious. Many leading physicists and philosophers have expressed either doubts or denial that physical mechanisms can, even in principle, yield an explanation of consciousness—for example, the physicists Schrödinger, Wigner, Dyson, Linde, and Witten, and the philosophers Chalmers, Kripke, and Nagel, to name just a few of the most famous. And since consciousness is presumably a prerequisite for free will, we cannot hope to understand the latter without understanding the former. Sapolsky impatiently waves away the question “What is consciousness?” as “ridiculous”: “I don’t understand what consciousness is, can’t define it. I don’t understand philosophers’ writing about it.” This is an example of the strategy that philosophers of mind call “eliminative materialism”; what neural mechanisms cannot explain is eliminated from consideration as either unreal or not sufficiently definable to discuss. But consciousness is a reality that very few, even the most hard-boiled materialists, are willing to deny.

Consciousness is not the only such puzzle. Others are the unity of the self at one time and through time, and the ability of the human mind to draw a multiplicity of things into a single indivisible insight, as, for example, when we grasp the meaning of a sentence.

This brings us back to the power of human reason, with which we began. It is a fact of everyday life, which can scarcely be doubted, that our decisions can be influenced by reason. A judge can indeed be influenced by hunger, but he can also be influenced to change his mind by an abstract argument of great subtlety. A mathematician may come to some new conclusion due to an insight into non-Euclidean geometry or transfinite numbers. How do things such as those affect the firing of neurons, as they most certainly can? If neurons act in a quasi-deterministic way, what allows the brain to be “open” to truth?

Sapolsky does try to tackle the question of how we change our minds. But he gets no further than talking about how sea slugs change their behavior in response to having their tails shocked repeatedly and how rats change their behavior in response to Pavlovian training. Really? This is supposed to get us anywhere near to explaining the highest powers of the human mind?

Much of Sapolsky’s book, especially its last section, is devoted to showing that mankind will be better off in important ways if we stop believing in the myth of free will. We will be liberated (so to speak) from the whole mistake of moral responsibility, including moral praise and blame. The end of “moralizing” will bring an end to punishment and lead to a more humane society. But Sapolsky’s entire book, though he does not realize it, is an assault on human dignity and human values, because it is an assault on human reason. For if freedom is a myth, then so is human rationality.

One hundred and twenty years ago, G. K. Chesterton wrote an essay answering a materialist-determinist of his day named Robert Blatchford. As I read the final part of Sapolsky’s book, the end of Chesterton’s essay came to mind:

Yes, there is a liberty that has never been chained. . . . It is the liberty of the mind, that is to say, it is the one liberty on which Mr. Blatchford makes war. That which all the tyrants have left, he would extinguish. That which no gaoler could ever deny to a prisoner, [he] would deny. More numerous than can be counted, in all the wars and persecutions of the world, men have looked out of their little grated windows and said, “at least my thoughts are free.” “No, No,” says the face of Mr. Blatchford, suddenly appearing at the window, “your thoughts are the inevitable result of heredity and environment. Your thoughts are as material as your dungeons. Your thoughts are as mechanical as the guillotine.” So pants this strange comforter, from cell to cell.

Stephen M. Barr is professor emeritus of physics at the University of Delaware and president of the Society of Catholic Scientists.

Image by Mehmet Turgut Kirkgoz, public domain. Image cropped.