At the end of The Matrix trilogy, Neo and Agent Smith are engaged in one final, interminable scene of surreal combat, a surrogate competition for an eternal battle between humans and machines. “It’s pointless to keep fighting,” Agent Smith declares to Neo. “Why do you persist?”
“Because I choose to,” Neo replies, just before the computer-generated Smith meets his demise in a cinematic celebration of human free will’s superiority to the programming that enslaves machines. Machines are mindless. The brain is a decider.
All very inspiring, except that the brain itself is a machine, a network of cells that computes its choices based on the sum of sensory inputs and their interactions with neural anatomy. “Free will” is not the defining feature of humanness, modern neuroscience implies, but is rather an illusion that endures only because biochemical complexity conceals the mechanisms of decision making.
Yet belief in free will persists as stubbornly as Neo’s resistance to electronic tyranny. Whether supposedly free choice is actually a Matrix-like mirage remains one of the great questions of human philosophical history. For centuries that question was assessed mostly with thought —uninformed by actual neurobiological knowledge. Nowadays, though, the inner workings of the brain are revealing themselves to modern methods of neuroinquiry, and free will seems merely to emerge from electrochemical networks of neuronal interactions. But like tourists exploring a strange city without a GPS map, scientists don’t know how all the neural neighborhoods are connected and occasionally encounter surprising enclaves—such as a place in the brain called the lateral habenula.
“There’s lots of new research showing that an overactive habenula has behavioral effects,” says neuropharmacologist Martine Mirrione of Brookhaven National Laboratory in Upton, N.Y.Questioning consciousness
To most people, who have never heard of the habenula, free will’s existence seems obvious, because they can make up their own mind whether to believe in it or not. Consciousness of choosing seems to imply the ability to choose. But the 19th century English historian Henry Thomas Buckle ridiculed such logic, pointing out that consciousness is often fallible. Some people profess to have consciousness of the presence of ghosts, for example. “If this boasted faculty deceives us in some things, what security have we that it will not deceive us in others?” Buckle asked.
Knowing everything about a man’s character, history and all external circumstances would in fact allow someone to accurately predict what he would do, Buckle averred. That example was hypothetical, he acknowledged. “We never can know the whole of any man’s antecedents,” he wrote. “But it is certain that the nearer we approach to a complete knowledge of the antecedents, the more likely we shall be to predict the consequent.”
Today, science’s knowledge is not nearly complete, but it’s a lot closer than in Buckle’s day. As evidence flows in from probes of animal brains and scans of living humans, the neural antecedents of the brain’s decisions are becoming more clearly visible. “Perhaps,” write neuroscientists Alireza Soltani and Xiao-Jing Wang, “we are entering a new period of consilience between the science of the brain and the science of the mind.”
Death to dualism
Such consilience would certify the death of Cartesian dualism, the mind-body distinction articulated by the French philosopher René Descartes in the 17th century. In modern neuroscience, that division dissolves—the mind is simply a reflection of different states of the brain. And brain states dictate the behaviors that masquerade as free choices.
Brains are, after all, the product of evolution. To survive and perpetuate their species, animals need food, water and sex. So brains are programmed to produce behavior that serves those ends—or seek substitutes that stimulate the same neural systems. Free will is not free to ignore these imperatives, although it isn’t always obvious how they all add up and tip the scales in favor of go or stop, do or don’t. Somehow, the brain sorts out the interplay between desire and caution, pleasure and pain, curiosity and fear. And the neural systems established by evolution for survival direct all the other decisions that animals (including people) routinely make—fight or flee, explore or hide, red or white, left or right.
Neurobiologists like to describe the sum of the brain’s many motivations with the concept of reward. In real life, the common currency for measuring reward is money (and consequently the study of the brain’s choice-making is sometimes called neuroeconomics). In the brain, that currency seems to be the molecular messenger known as dopamine.
Neurons producing dopamine are powerful forces in directing the brain’s decisions. Certain dopamine neurons in the midbrain are particularly active in driving the brain to seek rewards. But they’re not tuned simply to pleasure. Those dopamine neurons become electrically excited and release molecular messages simply in anticipation of pleasure. If the expected reward does not then materialize, those dopamine neurons take a rest. On the other hand, when an unexpected reward arrives, they fire signals vigorously. Apparently these dopamine neurons encode errors in predictions about potential rewards, so as to improve future decisions on what courses of action to pursue. In other words, dopamine neurons underlie learning how to behave based on pleasurable experiences.
Hail the habenula
Sound decisions depend on more than seeking pleasure, though. It’s also important to learn what choices will turn out to be bad. And the latest research suggests that that’s a job for the habenula.
It’s an obscure structure found deep in the brain, beneath the corpus callosum near the thalamus and in front of the pineal gland (the small body identified by Descartes as the seat of the soul, the source of free will). “Virtually all kinds of vertebrates have this habenula, which suggests that it is very important for survival,” says Okihide Hikosaka of the National Eye Institute, an NIH agency in Bethesda, Md.
When a monkey is faced with a nonrewarding choice, neurons in the lateral part of the habenula fire their signals rapidly, Hikosaka and Masayuki Matsumoto reported in Nature last year. When the habenula neurons fire, dopamine neurons slow down. Apparently the habenula warns against bad choices by suppressing dopamine activity, either directly or perhaps via intermediary neurons.
“Dopamine neurons contribute to learning of actions based on good experiences,” Hikosaka says, “whereas lateral habenula neurons are probably involved in learning of actions based on bad experiences.”
Recent work in several other labs suggests that the habenula plays an especially key role in neuronal crosstalk, serving as a sort of relay station between the primitive parts of the brain, which control basic needs, and the most advanced frontal regions where thought and logic presumably moderate basic impulses. But nobody suggests that the habenula is the source of all decisions or the seat of human consciousness. It’s just one hub in a network of brain addresses where parts of the decision-making process are assembled. Neuroscientists discussing such issues chatter about the amygdala, the nucleus accumbens and the anterior cingulate cortex, the PFC, the OFC and the IPC. Such areas encode information on rewards, costs or how much to discount the value of rewards that will be delayed. Different neural neighborhoods control risky choices, safe bets and when to change a decision already made. And while the habenula communicates to many brain regions involved in decision making, various regions transmit messages to the habenula, too.
All of this is important for much more than just enlightening free-will philosophy or learning the nomenclature of brain anatomy. Habenula activity has been implicated in everything from stress and anxiety to psychiatric disorders and sleep. Besides influencing dopamine cells, for example, signals from the habenula suppress neurons that make serotonin, the brain chemical famous for its effects on mood. Mirrione and her collaborators at Brookhaven have shown a link between elevated habenula activity and symptoms of depression in rats.
Depressed people typically forgo pleasurable activities that would ordinarily elicit “go” signals from dopamine neurons. An overactive habenula, by damping dopamine, could drive depression by denying the brain the power to choose pleasure. Many popular antidepressants work by elevating the brain’s serotonin levels, perhaps countering the habenula signals that suppress serotonin production. But such antidepressants don’t always work. Direct intervention in the habenula might offer an alternative, Mirrione says. Their rat study “suggests that the habenula appears to be a novel target for therapeutic intervention in treatment-resistant depression,” she and her collaborators reported in November in Washington, D.C., at the annual meeting of the Society for Neuroscience.
Other studies hint that the habenula plays a role in nicotine withdrawal behaviors, with implications for helping people to quit smoking. Behavior underlying other drug addictions might also be disrupted by intercession in the habenula, Israeli scientists reported at the neuroscience meeting. Their study found that deep brain stimulation of the habenula influenced the desire of addicted rats to self-administer cocaine.
Practical and clinical implications aside, the habenula’s multiple powers, and the diversity of other brain regions it interacts with, all suggest that the original question about free will is ill-posed. Asking whether humans have free will is like asking which came first, chicken or egg. It’s not a meaningful question. For chickens and eggs, the issue is understanding DNA and genes and the chemistry controlling reproduction and heredity. For free will, the issue is understanding the complex circulation of molecular information that is massaged and manipulated at various stations by neural systems tuned to multiple decision-making considerations. That process is free will, even if it isn’t really free. So deciding whether the will is free turns out to be circular, although perhaps not viciously, like some of those fights in The Matrix.