And the homunculus sits there controlling behavior. There are some things outside its purview—seizures blow the homunculus’s fuses, requiring it to reboot the system and check for damaged files. Same with alcohol, Alzheimer’s disease, a severed spinal cord, hypoglycemic shock.
There are domains where the homunculus and that brain biology stuff have worked out a détente—for example, biology is usually automatically regulating your respiration, unless you must take a deep breath before singing an aria, in which case the homunculus briefly overrides the automatic pilot.
But other than that, the homunculus makes decisions. Sure, it takes careful note of all the inputs and information from the brain, checks your hormone levels, skims the neurobiology journals, takes it all under advisement, and then, after reflecting and deliberating, decides what you do. A homunculus in your brain, but not of it, operating independently of the material rules of the universe that constitute modern science.
That’s what mitigated free will is about. I see incredibly smart people recoil from this and attempt to argue against the extremity of this picture rather than accept its basic validity: “You’re setting up a straw homunculus, suggesting that I think that other than the likes of seizures or brain injuries, we are making all our decisions freely. No, no, my free will is much softer and lurks around the edges of biology, like when I freely decide which socks to wear.” But the frequency or significance with which free will exerts itself doesn’t matter. Even if 99.99 percent of your actions are biologically determined (in the broadest sense of this book), and it is only once a decade that you claim to have chosen out of “free will” to floss your teeth from left to right instead of the reverse, you’ve tacitly invoked a homunculus operating outside the rules of science.
This is how most people accommodate the supposed coexistence of free will and biological influences on behavior.* For them, nearly all discussions come down to figuring what our putative homunculus should and shouldn’t be expected to be capable of. To get a feel for that, let’s look at some of these debates.
Age, Maturity of Groups, Maturity of Individuals
In 2005’s Roper v. Simmons decision, the Supreme Court ruled that you can’t execute someone for a crime committed before the age of eighteen. The appropriate reasoning was straight out of chapters 6 and 7: the brain, especially the frontal cortex, is not yet at adult levels of emotional regulation and impulse control. In other words, adolescents, with their adolescent brains, aren’t as culpable as adults. The reasoning was the same as why the pig was executed but not her piglets.
In the years since, there have been related rulings. In 2010’s Graham v. Florida and 2012’s Miller v. Alabama, the Court emphasized that juvenile offenders have the highest potential for reform (because of their still-developing brains) and thus banned life sentences without parole for them.
These decisions have prompted a number of debates:
Just because adolescents are, on the average, less neurobiologically and behaviorally mature than adults doesn’t rule out the possibility of some individual adolescents being as mature, thus being appropriately held to adult standards of culpability. Related to that is the obvious absurdity of implying that something neurobiologically magical happens on the morning of someone’s eighteenth birthday, endowing them with adult levels of self-control. The usual response to these points is that, yes, these are true, but the law often relies on group-level attributes with arbitrary age boundaries (e.g., the age at which someone can vote, drink, or drive). There is this willingness because you can’t test every teenager each year, month, hour, to determine whether they are mature enough yet to, say, vote. But it’s worth doing so when it comes to a teenage murderer.
In another contrarian view the issue isn’t whether a seventeen-year-old is as mature as an adult but whether they are mature enough. Sandra Day O’Connor, in dissenting from the Roper decision, wrote, “The fact that juveniles are generally less culpable for their misconduct than adults does not necessarily mean that a 17-year-old murderer cannot be sufficiently culpable to merit the death penalty” (her emphasis). Another dissenter, the late Antonin Scalia, wrote that it is “absurd to think that one must be mature enough to drive carefully, to drink responsibly, or to vote intelligently, in order to be mature enough to understand that murdering another human being is profoundly wrong.”11
Amid these differing opinions everyone, including O’Connor and Scalia, agrees that there exist age-related boundaries on free will—everyone’s homunculus was once too young to have its adult powers.12 Maybe it wasn’t tall enough yet to reach all the control dials; maybe it was distracted from its job by fretting about that gross pimple on its forehead. And that needs to be considered during legal judgments. Just as with piglets and pigs, it’s just an issue of when a homunculus is old enough.
The Nature and Magnitude of Brain Damage
Essentially everyone working with a model of mitigated free will accepts that if there is enough brain damage, responsibility for a criminal act goes out the window. Even Stephen Morse of the University of Pennsylvania, a strident critic of neuroscience in the courtroom (much more later), concedes, “Suppose we could show that the higher deliberative centers in the brain seem to be disabled in these cases. If these are people who cannot control episodes of gross irrationality, we’ve learned something that might be relevant to the legal ascription of responsibility.”13 In this view, mitigating biological factors are relevant if the capacity for reasoning has been grossly impaired.
Thus, if someone had their entire frontal cortex destroyed, you probably shouldn’t hold them responsible for their actions, because their rationality is grossly impaired when deciding their own courses of action.14 But the issue then becomes where a line is drawn on a continuum—what if 99 percent of the frontal cortex is destroyed? What if 98 percent? This is of great practical importance, since a large percentage of death row inmates have a history of damage to the frontal cortex, particularly of the most disabling type, namely early in life.
In other words, amid differing opinions about where a line should be drawn, believers in mitigated free will agree that massive amounts of brain damage overwhelm a homunculus, while it should be expected to handle at least some damage.
Responsibility at the Level of the Brain and at the Social Level
The renowned neuroscientist Michael Gazzaniga, one of the leading lights and elders of the field, has taken an extremely odd path in writing, “Free will is an illusion, but you’re still responsible for your actions.” This is expounded at length in a challenging book of his, Who’s in Charge? Free Will and the Science of the Brain. Gazzaniga fully accepts the entirely material nature of the brain but nonetheless sees room for responsibility. “Responsibility exists at a different level of organization: the social level, not in our determined brains.” I think either he is actually saying, “Free will is an illusion, but for practical reasons, we are still going to hold you responsible for your actions,” or he is hypothesizing some manner of homunculus that exists only at a social level. In response to the latter idea, the pages of this book show how our social world is ultimately as much a product of our determined, materialist brains as are our simple motor movements.*15
The Time Course of Decision Making
Another well-established fault line in a stance of mitigated free will is that our capacity for free will moves to the forefront with decisions that are slow and deliberative, whereas biological factors may push free will aside in split-second-decision situations. In other words, the homunculus is not always sitting right at the helm in the bunker; instead it occasionally wanders off to grab a snack, and if something exciting suddenly arises, those neurons may fire off commands to muscles and produce a behavior before the homunculus can rush back and hit that big red button on the control panel.
Issues of getting to the red button in time intersect with issues of the adolescent brain. A number of critics of Roper v. Simmons,
starting with O’Connor in her dissenting opinion, noted a seeming contradiction. The American Psychological Association (APA) had filed an amicus curiae brief in the case, emphasizing that adolescents (i.e., their brains) are so immature that they can’t be held to adult criminal standards with sentencing. Turns out that the same APA had filed a brief some years earlier in a different case, emphasizing that adolescents are sufficiently mature that they should be able to choose whether to have an abortion, even without parental consent.
Well, that’s a bit awkward, and it sure makes the APA and its ilk appear to be flip-flopping along ideological grounds, O’Connor charged. Laurence Steinberg, whose research on adolescent brain development was covered heavily in chapter 7 (and whose work was influential in the Roper v. Simmons decision), offers a logical resolution.16 Deciding whether to have an abortion involves logical reasoning about moral, social, and interpersonal issues, stretching out over days to weeks. In contrast, deciding whether to, say, shoot someone can involve issues of impulse control over the course of seconds. The frontal immaturity of the adolescent brain is more pertinent to split-second issues of impulse control than to slow, deliberative reasoning processes. Or in a mitigated-free-will framework, rapid-fire, impulsive behaviors can occur while the homunculus has gone to the bathroom.
Causation and Compulsion
Some proponents of mitigated free will distinguish between the concepts of “causation” and “compulsion.”17 In a way that feels a bit nebulous, the former involves every behavior having been caused by something, of course, but the latter reflects only a subset of behaviors being really, really caused by something, something that compromises rational, deliberative processes. In this view some behaviors are more deterministically biological than others.
This has been relevant to schizophrenic delusions. Suppose someone suffering from schizophrenia has auditory hallucinations, including a voice telling him to commit a crime; he does so.
Some courts have viewed this as not mitigating. If your friend suggests that you mug someone, the law expects you to resist, even if it’s an imaginary friend in your head.
But others see distinctions depending on qualities of the auditory hallucinations. In that view, if a schizophrenic individual commits a crime because a voice in his head demanded it, yes, his act was caused by that voice, but that doesn’t excuse the crime. In contrast, consider a schizophrenic individual committing a crime because thundering choruses of taunting, threatening, cajoling voices in his head, complete with baying hellhounds and choirs of trombones playing loud atonal music, command him every waking moment to do the crime. When he succumbs and does so, it is deemed more excusable, because those voices constituted a compulsion to act.*
Thus in this view even a sensible homunculus can lose it and agree to virtually anything, just to get the hellhounds and trombones to stop.
Starting a Behavior Versus Halting It
It is virtually ordained that any discussion of volition and biology eventually considers the “Libet experiment.”18 In the 1980s neuroscientist Benjamin Libet of UCSF reported something fascinating. A subject is hooked up to an EEG machine, which monitors patterns of electrical excitation in the brain. She sits quietly, looking at a clock. She has been instructed to flick her wrist whenever she feels like it and to note the time, down to the second, when she decided to do so.
Libet would identify something in the EEG data called a “readiness potential”—a signal from the motor cortex and supplementary premotor areas that a movement would soon be initiated. And consistently, readiness potentials appeared about half a second before the reported time of conscious intent to move. Interpretation: your brain “decided” to move before you were even aware of it. Thus, how can you claim to have chosen when to move, evidence of free will, if the cascade of neural signaling culminating in movement started before you consciously chose? Free will is an illusion.
Naturally, this finding generated speculation, controversy, replications, elaborations, refutations, and nuances that are beyond me. One criticism concerned a necessary limitation of the approach. In this view, there’s free will, you freely decide when to move your wrist, and that readiness potential is a consequence of your decision. What’s the five-hundred-millisecond delay, in that case? That’s the lag time between the instant when the decision to move first occurs and when (a) attention is then focused on the clock and (b) the position of the second hand is interpreted. In other words, the supposed half-second lag is an artifact of the experimental design, not a real thing. Other criticisms concerned the ambiguity of feeling that you intend to move. Other criticisms are more arcane than I can follow.
A very different interpretation of the finding was offered, interestingly, by Libet. Yes, maybe your brain prepares to initiate a behavior before there is conscious awareness of the decision, meaning that your belief that you consciously chose to move is wrong. But in that lag time is the potential to consciously choose to veto that action. In the pithy words of V. S. Ramachandran (of mirror neuron speculation in chapter 14), we may not have free will, but we have “free won’t.”19
Predictably, this intriguing counterinterpretation has fueled more discussions, experiments, and counter-counterinterpretations. For us, surveying different disputes concerning mitigated free will, this entire debate is about the nature of a homunculus’s control panel. How many of its buttons and switches and dials that go up to eleven are involved in initiating a behavior versus halting it?
—
Thus a view of mitigated free will makes room for both biological causation of behavior and free will, and all the discussions merely concern where lines in the sand are drawn and how inviolate they are. This prepares us to consider what I think is the most important line-drawing debate.
“You Must Be So Smart” Versus “You Must Have Worked So Hard”
Stanford psychologist Carol Dweck has done groundbreaking work on the psychology of motivation. In the late 1990s she reported something important. Kids do a task, take a test, something, where they do it well. You then praise them in one of two ways—“What a great score; you must be so smart” or “What a great score, you must have worked so hard.” When you praise kids for working hard, they tend to work harder the next time, show more resilience, enjoy the process more, and become more likely to value the accomplishment for its own sake (rather than for the grade). Praise kids for being smart, and precisely the opposite occurs. When it becomes all about being smart, effort begins to seem suspect, beneath you—after all, if you’re really so smart, you shouldn’t have to work hard; you glide, you don’t sweat and grunt.20
Beautiful work that has achieved cult status among many thoughtful parents of gifted kids, who want to understand when their child’s smarts shouldn’t come into the picture.
Why do “You’re so smart” and “You work so hard” have such different effects? Because they fall on either side of one of the deepest lines drawn by believers in mitigated free will. It is the belief that one assigns aptitude and impulse to biology and effort and resisting impulse to free will.
It’s cool to see natural ability in action. The great all-around athlete who has never seen pole-vaulting before, watches it once, tries it once, and soars like a pro. Or the singer whose voice has always had a natural timbre that evokes emotions you never knew existed. Or that student in your class who obviously just gets it, two seconds into your explaining something really abstruse.
That’s impressive. But then there’s inspiring. When I was a kid, I repeatedly read a book about Wilma Rudolph. She was the fastest female runner in the world in 1960, an Olympian who became a civil rights pioneer. Definitely impressive. But consider that she was born prematurely, underweight, one of twenty-two kids in a poor Tennessee family, and—get this—at age four got polio, resulting in a leg brace and a twisted foot. Polio, she was crippled by polio. And she defied every expert’s expectations, worked and worked and worked through the pain, an
d became the fastest there was. That’s inspiring.
In many domains we can sort of grasp the materialist building blocks of natural ability. Someone has the optimal ratio of slow-twitch to fast-twitch muscle fibers, producing a natural pole-vaulter. Or has vocal cords with the perfect degree of velvety peach fuzz (I’m winging it here) to produce an extraordinary voice. Or the ideal combination of neurotransmitters, receptors, transcription factors, and so on, producing a brain that rapidly intuits abstractions. And we can also perceive the building blocks in someone who is merely okay, or lousy, at any of these.
But Rudolph-esque accomplishments seem different. You’re exhausted, demoralized, and it hurts like hell but you push on; you want to take an evening off, see a movie with a friend, but resume studying; there’s that temptation, no one’s looking, everyone else does it, but you know it’s wrong. It seems so hard, so improbable to think of those same neurotransmitters, receptors, or transcription factors when considering feats of willpower. There seems a much easier answer—you’re seeing the Calvinist work ethic of a homunculus sprinkled with the right kind of fairy dust.
Here’s a great example of this dualism. Recall Jerry Sandusky, the Penn State football coach who was a horrific serial child molester. After his conviction came an opinion piece on CNN. Writing under the provocative heading of “Do pedophiles deserve sympathy?” James Cantor of the University of Toronto reviewed the neurobiology of pedophilia. For example, it runs in families in ways suggesting genes play a role. Pedophiles have atypically high rates of brain injuries during childhood. There’s evidence of endocrine abnormalities during fetal life. Does this raise the possibility that a neurobiological die is cast, that some people are destined to be this way? Precisely. Cantor concludes, “One cannot choose to not be a pedophile.”21
Behave: The Biology of Humans at Our Best and Worst Page 60