Satoshi Kanazawa

The Scientific Fundamentalist

Why Do Some Men Beat Up Their Wives and Girlfriends?

Is evolutionary psychology “untestable” and “unfalsifiable”?

Posted Jul 18, 2010

In a previous post, I address the question of why so many battered women stay in their abusive relationships. (Answer: So that they could produce violent sons with the abusive husbands, who will grow up to kill many men.) Quite independent of the question of why battered women stay, another question is why some men beat up their wives and girlfriends in the first place.

Critics of evolutionary psychology often claim that evolutionary psychological explanations are “untestable” and “unfalsifiable.” As but one perfect example of the eminent testability and falsifiability of evolutionary psychological explanations, here are two competing explanations of domestic violence, formulated by the two deans of modern evolutionary psychology.

When Martin Daly and the late Margo Wilson first began studying domestic violence and uxoricide (the killing of one’s wife) in the early 1980s, they had competing explanations. Daly hypothesized that domestic violence and uxoricide resulted when the husband did not value his wife sufficiently and mistreated her as a result. Since a wife’s fertility and reproductive value decline with age, Daly predicted that older, less valuable wives were at a greater risk of spousal abuse and homicide than younger, more valuable wives.

Wilson, in contrast, hypothesized that domestic violence and uxoricide were a maladaptive byproduct of the husband’s inclination and tendency to guard his wife to make sure that she did not have sexual contact with other men. Because men should be more motivated to guard younger, more valuable wives, Wilson predicted that younger wives were at a greater risk of spousal abuse and homicide than older wives.

Both explanations use impeccable evolutionary psychological logic and derive from known first principles, but both predictions could not be true simultaneously. So Daly and Wilson got to work, as the great scientists that they were, and collected data on domestic violence and uxoricide in Canada and the United States. And they put their two competing hypotheses to an empirical test. Their data showed that younger wives were at a much greater risk of domestic violence and murder than older wives. In the end, Wilson’s prediction turned out to be true, and Daly’s false. Are evolutionary psychological explanations untestable and unfalsifiable?

Astute readers may be thinking right now, “But younger women are usually married to younger men. And younger men are more violent than older men, as you point out in your discussion of the age-crime curve. So younger women are at a greater risk of spousal abuse and murder, not because they are young but because their husbands are young and therefore more violent.”

Close, but no cigar. While it is difficult to separate the effects of the husband’s age and the wife’s age, because the two are indeed highly correlated, careful statistical analyses show that the wife’s age almost entirely determines the likelihood of being a victim of spousal abuse and homicide. Middle-aged husbands (ages 45-54) legally married in Canada to much younger wives (ages 15-24) are more than six times as likely to kill their wives than young husbands (ages 15-24) married to women of similar age. Among common-law marriages, middle-aged husbands married to much younger wives are more than 45 times as likely to kill their wives as young husbands. The effect of the wife’s age is so powerful that it overrides and even reverses a man’s tendency to become less violent with age. Thus, while it is true that younger men in general are much more violent and commit more murders than older men, young and old men kill different types of people. Young men kill other men (their male sexual rivals); older men kill their wives. As a result, the proportion of men among murder victims declines as the murderer’s age increases. For murderers aged 15-19, 86.3% of the victims are males; for murders aged 65-69, only 51.4% of them are males.