Tuesday, January 23, 2007

Aggressive Personalities, or, The Magic of Testosterone

Last week I talked about the role of stress hormones and the ANS in aggression. I'm particularly interested in the effects of stress on aggression because my Div III (thesis) looks at the effects of stress on another behavior, altruism. But clearly there are a lot of other factors influencing aggression. A second model of aggression is the serotonin/testosterone model.

There are many different types of aggression, from the rough and tumble play of young siblings to horrific acts of war. Sometimes it can be hard to recognize, but it's there, nonetheless - I'm excited by the recent focus on aggression in middle school girls. While there are thousands of different ways to be aggressive, researchers have found that there are two main aggressive personality types. These types are so common and so distinctive that you can easily see them in fiction as well as real life, including in the best-selling Harry Potter novels.



James Potter is the title character's father. Harry is an orphan, and at first, all he knows about James is that he was a star athlete, the Head Boy of his class, and that he went on to be an Auror (a glamorous sort of magical FBI agent). As Harry grows, he finds out that James wasn't as perfect as he first seemed. James, one of the most popular boys in the school, was fond of attacking other students verbally and with magic. In one scene, he and his friend use magic to hold outcast Severus Snape upside down so his underwear is exposed for everyone to see.

Severus Snape is one of Harry's professors. Snape went to school at the same time as James Potter, and still hates him, even though the elder Potter has been dead for over ten years. Snape uses his position of power to pick on Harry, making him serve unnecessary detentions and verbally jabbing at the young orphan's emotional wounds. Harry is not the only one Snape attacks, however. He also picks on Harry's classmates, other professors, and anyone else he takes a dislike to. He often attacks even when it would be to his benefit not to, making him unpopular with most of his students and peers.

James Potter exhibits what psychologists have termed controlled aggression. Controlled aggression is a dominance behavior, used by someone to advance in a social hierarchy or to maintain their high position. Controlled aggression is almost always provoked by some sort of percieved challenge by the victim. In animals, this might be a threat call or a low-ranking male's attempt to mate with a dominant male's female. For James Potter, it was likely Snape's refusal to acknowledge and respect his position as the most popular boy in school. While controlled aggression is not always justified, it generally is a means to an end, rather than an end itself. This type of aggression is also called instrumental aggression.

Severus Snape, on the other hand, generally exhibits impulsive aggression. Impulsive aggression is caused by strong emotions such as anger and is aimed at hurting someone - that is, the aggression serves no other purpose. Perhaps because of its emotional nature, impulsive aggression is more likely to be physical and violent, but it by no means always is (likewise, controlled aggression can be incredibly violent). Snape generally restricts his aggression to hurtful comments and abuse of power, but he is often described as just barely restraining himself. One gets the sense that if Snape was not afraid of being kicked out of Hogwarts (or constrained by a magical debt - don't ask), he would have physically attacked Harry a long time ago.

Obviously any two people, fictional or otherwise, will have other biological and environmental factors influencing their behavior. People can be both controlled and impulsive aggressors, sometimes even simultaneously. But let's assume that James Potter and Severus Snape are perfect prototypes of the controlled aggressor and the impulsive aggressor. How do they differ, physiologically?

Testosterone is one of the most well-known hormones, and it is certainly the substrate most associated with aggression. Testosterone is the primary male sex hormone, and its presence drives a lot of the changes that distinguish men from women. Some people explain the higher rates of aggression in males than females by calling it a secondary sex characteristic of testosterone, like one's voice dropping, or growing facial hair. Of course, it's not that simple. While levels of testosterone in animals are generally correlated with aggression, many studies in humans have failed to find the expected association. A study of eighteen highly aggressive young boys who had been institutionalized for their violent behavior found no abnormalities in testosterone levels when compared to healthy controls (Constantino et al, 1993). Studies of older children and adults, looking at both extreme behaviors (in institutionalized subjects) and behaviors within the norm (children rated as more aggressive by parents, teachers and peers) have failed to find a correlation (Susman et al, 1987, Olweus et al, 1988, Inoff-Germain, 1998). At the same time, a number of studies have found something. In particular, James Dabbs published a number of prison studies showing that testosterone levels in inmates were correlated with the violence of the crime that landed them in jail as well as their behavior there.

The conflicting literature left everyone a bit confused. Several theories have been proposed to explain why testosterone and aggression are only sometimes linked, including the one I'm going to talk about here: the dominance hypothesis.

The dominance hypothesis is best explained by Mazur and Booth (1998). They define dominance behavior as action with the intent of achieving or maintaining status and the power, influence, and benefits that come with status. For simplicity's sake, we can consider dominance behavior and controlled aggression to be essentially the same thing. Mazur and Booth propose that testosterone levels are only linked to dominance behavior and not other forms (ie, impulsive) aggression.

This hypothesis has gotten a lot of experimental support. A 1996 study (Schaal et al) looked at behavioral assessments of nearly two hundred boys. While there was no relationship between testosterone and aggression per se, there was a strong correlation with "leadership" behaviors but no correlation with physically aggressive behaviors. Also, when previous studies are re-analyzed and types of aggressive behavior parcelled out, a pattern begins to emerge. The Olweus study cited above found testosterone was higher in those who responded to provocation ("When a teacher criticizes me, I tend to answer back and protest") but not to simple aggressiveness ("I fight with other boys at school"). In the Inoff-Germain study, they found testosterone was linked an ability to deal with their anger and express it acceptibly and non-explosively. This did not mean they were passive. There was a signifiant negative correlation between testosterone and a failure to respond to being aggressed against.

In response to these studies, a new physiological model was made, to distinguish between controlled and impulsive aggression. This model drew upon a body of work showing that low serotonin levels led to general impulsive behavior. Researchers surmised that people who were impulsively aggressive had low serotonin levels. People who exhibited dominance behavior, on the other hand, were driven by testosterone. High serotonin levels allowed them to control their impulses and to use their aggression to establish dominance. A study in chimpanzees found that unprovoked assaults and biting was associated not with high testosterone but low serotonin, and that dominance behaviors such as mounting of other males were associated with high testosterone and high serotonin levels.

Let's go back to our two wizards. If we were able to analyze blood samples from James and Severus - or, for even better accuracy, cerebrospinal fluid - we might expect that James would have both high serotonin levels and high testosterone levels, whereas Severus would have low serotonin. The prediction of low serotonin also fits with Severus's description as often looking better and upset. While I would hesitate to diagnose Snape as clinically depressed, his unhappiness fits with the association between low serotonin levels and depression.

Obviously it is a stretch to talk about the neurochemistry of fictional characters, but hopefully these examples have helped explain one of the current models of aggression for you.

185 comments: