THE LEGAL SYSTEM rests on the assumption that we are “practical reasoners,” a term of art that presumes, at bottom, the existence of free will. The idea is that we use conscious deliberation when deciding how to act—that is, in the absence of external duress, we make free decisions. This concept of the practical reasoner is intuitive but problematic.

The existence of free will in human behavior is the subject of an ancient debate. Arguments in support of free will are typically based on direct subjective experience (“I feel like I made the decision to lift my finger just now”). But evaluating free will requires some nuance beyond our immediate intuitions.

Consider a decision to move or speak. It feels as though free will leads you to stick out your tongue, or scrunch up your face, or call someone a name. But free will is not required to play any role in these acts. People with Tourette’s syndrome, for instance, suffer from involuntary movements and vocalizations. A typical Touretter may stick out his tongue, scrunch up his face, or call someone a name—all without choosing to do so.

We immediately learn two things from the Tourette’s patient. First, actions can occur in the absence of free will. Second, the Tourette’s patient has no free won’t. He cannot use free will to override or control what subconscious parts of his brain have decided to do. What the lack of free will and the lack of free won’t have in common is the lack of “free.” Tourette’s syndrome provides a case in which the underlying neural machinery does its thing, and we all agree that the person is not responsible.

This same phenomenon arises in people with a condition known as chorea, for whom actions of the hands, arms, legs, and face are involuntary, even though they certainly look voluntary: ask such a patient why she is moving her fingers up and down, and she will explain that she has no control over her hand. She cannot not do it. Similarly, some split-brain patients (who have had the two hemispheres of the brain surgically disconnected) develop alien-hand syndrome: while one hand buttons up a shirt, the other hand works to unbutton it. When one hand reaches for a pencil, the other bats it away. No matter how hard the patient tries, he cannot make his alien hand not do what it’s doing. The movements are not “his” to freely start or stop.

Unconscious acts are not limited to unintended shouts or wayward hands; they can be surprisingly sophisticated. Consider Kenneth Parks, a 23-year-old Canadian with a wife, a five-month-old daughter, and a close relationship with his in-laws (his mother-in-law described him as a “gentle giant”). Suffering from financial difficulties, marital problems, and a gambling addiction, he made plans to go see his in-laws to talk about his troubles.

In the wee hours of May 23, 1987, Kenneth arose from the couch on which he had fallen asleep, but he did not awaken. Sleepwalking, he climbed into his car and drove the 14 miles to his in-laws’ home. He broke in, stabbed his mother-in-law to death, and assaulted his father-in-law, who survived. Afterward, he drove himself to the police station. Once there, he said, “I think I have killed some people … My hands,” realizing for the first time that his own hands were severely cut.

Over the next year, Kenneth’s testimony was remarkably consistent, even in the face of attempts to lead him astray: he remembered nothing of the incident. Moreover, while all parties agreed that Kenneth had undoubtedly committed the murder, they also agreed that he had no motive. His defense attorneys argued that this was a case of killing while sleepwalking, known as homicidal somnambulism.

Although critics cried “Faker!,” sleepwalking is a verifiable phenomenon. On May 25, 1988, after lengthy consideration of electrical recordings from Kenneth’s brain, the jury concluded that his actions had indeed been involuntary, and declared him not guilty.

As with Tourette’s sufferers, split-brain patients, and those with choreic movements, Kenneth’s case illustrates that high-level behaviors can take place in the absence of free will. Like your heartbeat, breathing, blinking, and swallowing, even your mental machinery can run on autopilot. The crux of the question is whether all of your actions are fundamentally on autopilot or whether some little bit of you is “free” to choose, independent of the rules of biology.

This has always been the sticking point for philosophers and scientists alike. After all, there is no spot in the brain that is not densely interconnected with—and driven by—other brain parts. And that suggests that no part is independent and therefore “free.” In modern science, it is difficult to find the gap into which to slip free will—the uncaused causer—because there seems to be no part of the machinery that does not follow in a causal relationship from the other parts.

Free will may exist (it may simply be beyond our current science), but one thing seems clear: if free will does exist, it has little room in which to operate. It can at best be a small factor riding on top of vast neural networks shaped by genes and environment. In fact, free will may end up being so small that we eventually think about bad decision-making in the same way we think about any physical process, such as diabetes or lung disease.