AGGRESSION FUELED: Contrary to popular opinion, the male hormone testosterone is not solely responsible for aggression.
© ISTOCKPHOTO.COM/Kenneth C. Zirkel

It's commonly assumed that testosterone, that stereotypically male hormone, is intimately tied to violence. The evidence is all around us: weight lifters who overdose on anabolic steroids experience "roid rage," and castration—the removal of the source of testosterone—has been a staple of animal husbandry for centuries.

But what is the nature of that relationship? If you give a normal man a shot of testosterone, will he turn into the Incredible Hulk? And do violent men have higher levels of testosterone than their more docile peers?

"[Historically,] researchers expected an increase in testosterone levels to inevitably lead to more aggression, and this didn't reliably occur," says Frank McAndrew, a professor of psychology at Knox College in Galesburg, Ill. Indeed, the latest research about testosterone and aggression indicates that there's only a weak connection between the two. And when aggression is more narrowly defined as simple physical violence, the connection all but disappears.

"What psychologists and psychiatrists say is that testosterone has a facilitative effect on aggression," comments Melvin Konner, an anthropologist at Emory University and author of The Tangled Wing: Biological Constraints on the Human Spirit. "You don't have a push-pull, click-click relationship where you inject testosterone and get aggressiveness."

Castration experiments demonstrate that testosterone is necessary for violence, but other research has shown that testosterone is not, on its own, sufficient. In this way, testosterone is less a perpetrator and more an accomplice—one that's sometimes not too far from the scene of the crime.

For example: regardless of their gender, the most violent prisoners have higher levels of testosterone than their less violent peers. Yet scientists hypothesize that this violence is just one manifestation of the much more biologically and reproductively salient goal of dominance.

"It has been suggested that the antisocial behaviors related to high testosterone are a function of the manner by which dominance is maintained in these groups," says Robert Josephs of the University of Texas at Austin. In other words, if researchers were to study other groups of folks, say the rich and famous, they might discover that testosterone is connected not to violence, but to who drives the biggest SUV or has the nicest lawn. As Josephs put it: "Perhaps slipping a shiv into your neighbor's back might play in the penitentiary, but it probably won't earn you any status points in Grosse Pointe."

One psychologist, James Dabbs of Georgia State University in Atlanta, made a career out of conducting studies connecting testosterone to every kind of lifestyle imaginable. In his book Heroes, Rogues and Lovers, he noted that athletes, actors, blue-collar workers and con men tend to have higher levels of testosterone than clerks, intellectuals and administrators.

What Dabbs didn't address was whether this correlation was the cause or an effect of the environment these men found themselves in. Which is to say, are high-testosterone males more likely to become violent criminals, or does being a violent criminal raise a man's level of testosterone?

No one really knows the answer, but a growing body of evidence suggests that testosterone is as much the result of violence as its cause. Indeed, both winning a sporting match and beating an opponent at chess can boost testosterone levels. (On the other hand, losing a sporting match, growing old and becoming obese all reduce levels of testosterone.)

"The causal arrow goes both ways," says Peter Gray of the University of Nevada, Las Vegas, whose own work shows that marriage and fatherhood lower testosterone levels. "There's evidence in humans that, just as in animals, testosterone is responsive to male-male competition."

Changes in testosterone levels in response to challenges can be further shaped by our expectations. In one experiment that put a biological spin on the red state–blue state divide, researchers at the University of Michigan at Ann Arbor had a volunteer "accidentally" bump into and then insult men who were raised either in the North or the South. The researchers hypothesized that Southerners come from a "culture of honor" in which aggressive responses to insults are culturally appropriate, and the results of their experiment bolstered that notion: Not only were Southerners more likely than their northern counterparts to respond with aggression, but their levels of testosterone also rose as a result. The Northerners, in contrast, were much less likely to experience an increase in testosterone.