Another possibility is that heroes are more strongly affected by the “field of force” that promotes compassion. Perhaps the sight or sound, or even the idea, of someone suffering affects them more strongly than it affects the average person. This seems a critical insight, but unfortunately, neither Milgram’s nor Batson’s studies give us much information about why or how this might be. Although in some ways Milgram’s and Batson’s goals were polar opposites (with Milgram interested in forces like obedience that override compassion, and Batson interested in when compassion overrides other forces like self-interest), in other ways their goals were very similar. Both men were trained in social psychology, a discipline that has historically focused on how situations and external events affect people on average rather than focusing on variation among people. Social psychologists ask how events like orders from an authority affect people—on average. How are people influenced by the belief that a stranger is suffering—on average? The advantage of focusing on external events and situations like these is that they can be tweaked. An experimenter can seat a suffering stranger right next to a study volunteer, or choose to make the stranger audible only over an intercom, or to not make him audible at all. Tweak! When these tweaks are coupled with tight control of all the other features of the study—the testing room, the shocks, the experimenter’s instructions—you’ve got a true experiment. And what a true experiment gives you is the most satisfying kind of scientific power—the power to say that the thing you tweaked caused the thing you measured. Mr. Wallace’s audible cries caused more people to stop shocking him. Difficulty of escape caused more volunteers to take the shocks on Elaine’s behalf.
But the disadvantage of focusing on external events is that you don’t get much information about individual variability. The factors responsible for variability often can’t be tweaked. They include every biological and environmental force that has ever affected a research volunteer up until the moment he or she arrives in the lab; tweaking factors like these is usually impossible or unethical, or both. What kind of parenting Milgram’s volunteers received as children, their IQs, and their personalities could very well have played a role in how they responded. But scientists can’t tweak these things. They can’t remove babies from their homes and foist them on new parents to study parenting, or inflict IQ-altering brain damage, or give people personality-altering drugs. Experiments like these would be as horrific as Eichmann’s atrocities.
So we have to do the best we can without tweaking these fundamental features. We observe and measure parenting and intelligence and personality, try to control for extraneous variables, then statistically map possible causes onto possible effects, all with the knowledge that we may still be missing something. For instance, if an aggressive adult had harsh and punitive parents, perhaps the harsh parenting caused his aggression. Or perhaps not. People also share genes in common with their parents. So another possibility—one of many—is that observed correspondences between harsh parents and aggressive offspring result from common genetic factors causing them all to act out.
The Milgram volunteers who continued shocking Mr. Wallace all the way down the switchboard intentionally and knowingly caused him harm, which is the definition of aggression. Let’s say that the volunteers who tended to do this were raised by harsher than average parents. Even if this were true, we couldn’t say that harsh parenting caused the volunteers to act more aggressively because there are too many other possible alternatives. It comes back to the old trope about correlation not implying causation. Keep this in mind whenever you hear about developmental studies linking some behavior in parents—harsh discipline or breastfeeding or using complex language or anything else—to some outcome in children. Just because one event precedes another does not mean that it caused the other. Many of these studies aren’t designed to tease apart the roles of genetic and environmental factors, so they can’t clearly establish cause and effect.
Luckily, there are ways of getting around problems like this. One is to take advantage of natural experiments. Natural experiments result when a variable is tweaked by someone, or something, other than a scientist. They are rarely natural and are not true experiments, but they are invaluable nonetheless. One well-known example is adoption studies. It’s clearly unethical for scientists to take babies from their biological parents and give them to unrelated adults to raise, but it’s fine—admirable even—for adoption agencies to do the same thing. And what’s not unethical is for scientists to study children who have been adopted to untangle the effects of genes and parenting. Adoptions “naturally” disentangle genes and parenting by ensuring that one set of parents contributes genes and an unrelated set contributes only parenting. So scientists can study adopted children to learn about how genes and parenting contribute to nearly any outcome in children.
Another way to disentangle genetic and environmental effects is through studies of twins. Because identical twins share 100 percent of their genes whereas fraternal twins share on average only 50 percent (just as any other biological siblings do), the contributions of genes and the environment can be teased apart by studying similarities and differences between identical and fraternal twins. An even more powerful approach is to combine these methods and study identical and fraternal twins raised by their biological or adopted parents. As a result of such studies we know, for example, that identical twins are very similar to one another across multiple physical and psychological indices, even when they are raised in different households. In fact, when raised apart they are sometimes more similar to one another—in terms of their IQs, say—than fraternal twins raised in the same household. A study like this provides compelling evidence for genetic contributions to intelligence.
The results of twin and adoption studies show that some outcomes are almost entirely inherited. To no one’s surprise, for example, adopted children’s eye color corresponds more strongly to that of their biological parents than their adopted parents. The heritability of human eye color—meaning how much of the variation in eye color results from inherited factors rather than environmental factors—is about 98 percent. Environmental factors contribute almost nothing. This is how researchers could determine with certainty the eye color (blue) of the English king Richard III, who died more than 500 years ago, using DNA samples extracted from his bones. The heritability of other physical features also tends to be high. Height is about 80 percent heritable, meaning that most variation in height results from genes. The other 20 percent largely reflects the effect of nutrition or illness. At least, this is how it works in typical modern, prosperous societies.
A caveat is that the heritability of some traits, like height, may fluctuate depending on the environment. For example, when food is scarce, the heritability of height decreases. This is because genes encode people’s maximum potential height—the height they can achieve with adequate health and nutrition during childhood. Food scarcity prevents people from reaching their potential, and the greater the deprivation, the greater the difference between their genetic potential and their actual adult height. In a malnourished population, those children who get 70 percent of the calories they need end up even shorter than the children who get 90 percent of what they need. As a result, widespread environmental factors like food availability account for much more than 20 percent of the differences in malnourished children’s heights. And as the proportion of the variability accounted for by the environment goes up, the proportion accounted for by genes inevitably goes down.
When children are getting enough food, however, more of it will make no difference. Children who consume 100 percent of the calories they need—enough to compensate for the calories they expend through activity and growth—will not be shorter than children who consume 110 percent of the calories they need. Once you have enough food to reach your maximum potential height, getting more food has no further effects. This is why public health efforts aimed at improving a population’s well-being tend to focus on reducing poverty rather than increasing weal
th: even small reductions in poverty can improve overall outcomes in a way that increases in wealth never will.
Even in prosperous populations, the heritability of other physical traits is somewhat lower than it is for height. For body weight, heritability hovers around 50 percent. This makes sense, in part because body weight has no maximum potential value, nor is it fixed by adulthood. So your parents’ choices and other environmental factors that shape your diet and lifestyle can more strongly shape your body composition than your height. But that 50 percent genetic contribution should not be ignored—body shape is not infinitely modifiable. No biological offspring of stocky parents will ever be rail-thin, even if the child is adopted and raised by paleo-vegan Pilates devotees. No diet could turn Kim Kardashian into Kendall Jenner, her lanky half-sister. Kendall carries the genes of tall, lanky Caitlyn Jenner, whereas Kim’s father was the short, stocky Robert Kardashian Sr. Biology is not destiny, but it does place limits on destiny.
This is true for essentially every complex human trait, from physical traits like body shape or facial appearance to psychological traits like aggressiveness or extraversion. According to the famed behavioral geneticist Eric Turkheimer, the first law of behavioral genetics is that all human behavioral traits are heritable. Like body composition, most psychological traits—our mental composition, you might say—are about 50 percent heritable. A massive study reported in the journal Nature Genetics showed that, across fifty years of studies of hundreds of thousands of pairs of twins, genes account for, on average, 47 percent of variance in cognitive traits like intelligence and memory and 46 percent of variance in psychiatric traits, including aggression. Parenting and other environmental factors undoubtedly shape outcomes as well, but genetic factors are at least as influential—and often more influential. This helps explain why twins adopted into separate families and later reunited find themselves tickled by the similarities they discover—from hair color to preferred hairstyle to preferred hobbies—despite their separate upbringings.
In Milgram’s era, most psychologists would have considered delving into the heritability of aggression, or any other personality variable in humans, a fool’s errand. From the early twentieth century and extending well into the 1960s, the tenets of a school of thought called behaviorism dominated psychology. Behaviorists like John Watson and B. F. Skinner viewed observable variations in animal and human behavior as primarily a result of their learning histories. If an organism—a pigeon, a rat, a monkey—had been previously rewarded (or “positively reinforced,” as the behaviorists termed it) for pushing a button, it would come to push the button more. If it had been punished for pushing the button, it would push it less. Two animals in adjacent cages that pushed their buttons different numbers of times must have experienced different prior outcomes for doing so. The behaviorists’ views were very influential—Skinner (yet another of the Harvard Psychology Department’s famous faculty) is today considered the single most influential psychologist of the last century.
Skinner’s experiments were beautifully designed and their results compelling. The ingenious “Skinner boxes” that he created to test his predictions were preserved in a lovingly curated exhibition in the basement of William James Hall that I used to pass on my way to my classes there. I marveled at the ingenuity of the little boxes festooned with elaborate arrays of tiny wires and pulleys and buttons and drawers. Of course, Skinner’s experiments used little metal boxes because all of his participants were rats and pigeons. In truth, the scope of his research was very narrow. He measured only simple behaviors that could be tested in one of his boxes, like lever-pulling and button-pecking; then, like other behaviorists, he extrapolated wildly from his findings, arguing that all variability in all animals’ behaviors—from rat aggression to human language and love—was best understood as resulting from learning histories. Skinner famously mused in his novel Walden Two, “What is love except another name for the use of positive reinforcement? Or vice versa.”
Likewise, the thinking went, two children in adjacent houses with different levels of aggression must simply have received differential reinforcement for aggression along the way—one rewarded for aggressive behaviors more than the other. The rewards in question needn’t be cookies or stickers. Only the world’s worst parents would reward aggression with literal prizes. But all sorts of other inadvertent behaviors on the part of parents could reward aggression, in theory. When aggression begets attention, even in the form of yelling or criticism, it can be more rewarding than no attention. A child who is ignored most of the time except when he hits his brother—in which case he gets yelled at—might actually prefer the yelling. Or if hitting his brother gets him something else he wants—his brother out of his bedroom, a toy his brother was holding—that’s a reward too. According to Skinner, if we could perfectly control the rewards and punishments that children receive from their rearing environments, we could eliminate undesirable behaviors like aggression entirely.
But there is absolutely no evidence that this is true. Rewards clearly do influence behavior, as do punishments. But heritability studies prove without a doubt that they are not the only influences. The heritability of aggression is consistently found to be around 50 percent, and for some forms of aggression it is as high as 75 percent. If that much of the variability in children’s aggression can be predicted from genetic differences among them, genes must play a major role in promoting aggression.
What all of this means is that understanding aggression—and the compassion that can inhibit it—requires more than looking at people’s behaviors inside a laboratory, where various environmental variables can be tweaked to make people behave more or less compassionately. A complete understanding of the roots of aggression and compassion also requires looking at deeply rooted, inherited variables that affect how compassionately people behave—that produce the variation in how people respond to various tweaks. Perhaps the most infamous and compelling such variable is psychopathy.
Psychopathy (pronounced sigh-COP-a-thee) is a disorder that robs the human brain of the capacity for compassion. It is characterized by a combination of callousness, poor behavioral control, and antisocial behaviors like conning and manipulation. Psychopaths need not be violent, but they often are. Only about 1 or 2 percent of the American population could be classified as true psychopaths, but among violent criminals the number may be as high as 50 percent. Psychopaths are marked by their tendency to engage in proactive aggression—acts of violence and aggression that are deliberate and purposeful rather than hot-tempered and impulsive.
Psychopathy is also highly influenced by genes, with a heritability quotient that may be as high as 70 percent. That this surprises many people I encounter reflects a common view of human aggression that is often, whether they know it or not, colored by the long shadow of behaviorism. Most people assume that violent, callous individuals must be the outcome of highly abusive or neglectful homes. But this simply isn’t true.
Take Gary Ridgway, the middle son of Mary Rita Steinman and Thomas Newton Ridgway. Gary and his brothers were raised in McMicken Heights, Washington, just north of where I grew up in Tacoma. The family was poor, no doubt: Thomas drove trucks off and on for money, and the family was crammed into a 600-square-foot house. Mary was a bossy and dominating mother—a “strong woman,” as her oldest son Greg later recalled. She and her husband had fights that sometimes turned violent—she once broke a plate across his head during a family dinner. But Gary also remembered her as a kindly figure who did jigsaw puzzles with him when he was a small boy and helped him with his reading. There was no sign of true abuse or household dynamics outside the range of normalcy for a family in the 1960s. And Gary’s brothers grew up to lead ordinary lives.
Not so for Gary, who grew up to become the Green River Killer, the most prolific serial murderer in American history. He is now serving a life sentence for forty-nine confirmed murders, and he has claimed that he committed dozens more. His first attempt at murder took place around
1963—the same year, as it happens, that Milgram was conducting his studies on obedience.
Gary was about fourteen years old and on his way to a school dance. Walking through a wooded lot, he ran across a six-year-old boy. Almost without thinking, he pulled the boy into the bushes and, using a knife he always carried with him, stabbed the boy in the ribcage, piercing his kidney. He quickly withdrew the knife and watched blood gush from the wound. Then Gary walked away, leaving the boy to die—or live. He wasn’t particularly concerned either way, except that he hoped that if the boy lived, he wouldn’t be able to identify him. (The boy did live, but never identified Ridgway as his assailant.) Later on, Ridgway couldn’t even pinpoint why he had done it. It had felt like it just happened, much as other bad things often seemed to just “happen” for him—gleaming rows of windows shattered by rocks, birds felled by a BB gun, a cat suffocated in a picnic cooler.
As Ridgway neared adulthood, things turned much darker. Relentless sexual urges awakened within him. Combined with the callousness and delight in the power of killing he already possessed, those urges turned Ridgway into an insatiable sexual sadist who raped and murdered at least forty-nine girls and women, most of them teenage runaways and adult sex workers around the town of SeaTac in the 1980s, while I was in elementary school some thirty miles to the south.
Ridgway was an unusually depraved personality even compared to other murderers—“a lean, mean killing machine,” as he called himself. Mary Ellen O’Toole, a famed FBI profiler and expert on psychopaths, spent many hours interviewing Ridgway, and she has told me that he is one of the most extreme, predatory psychopaths she has ever encountered.
The Fear Factor Page 5
