Sex or Feelings?

In 1992 the journal Psychological Science published a paper by David Buss, Randy Larsen, Drew Westen and Jennifer Semmelroth that became one of the best-known findings in evolutionary psychology. It began with a simple dilemma put to American undergraduates. Imagine that the person you are seriously involved with has become interested in someone else. Which would upset you more: your partner forming a deep emotional attachment to that person, or your partner enjoying passionate sexual intercourse with them?

The answers split by sex. A majority of the men chose the sexual infidelity as the more upsetting. A large majority of the women chose the emotional one. In a second study the researchers measured physiological responses, heart rate, skin conductance and the frowning muscle of the brow, while participants imagined each scenario, and found the men more aroused by the sexual one.

The explanation they offered was evolutionary. Fertilisation happens inside the woman's body, so a man can never be certain that a child is his; an ancestral man whose partner had sex with a rival risked investing years in another man's child. A woman is always certain the child is hers; her ancestral risk was that her partner's attention and resources would be diverted to someone else, and emotional attachment is the signal of that. Different risks, the argument goes, produced differently tuned jealousy.

It is an elegant argument, and the forced-choice result has replicated many times. That is the first thing to hold on to. Whatever the finding means, it is not a fluke.

The second thing is that almost every step between the result and the explanation has been contested.

The first objection is about method. In the original format people must choose one betrayal over the other. When researchers instead asked people to rate how upsetting each betrayal would be on its own scale, the difference between the sexes became much smaller, and both sexes tended to rate sexual infidelity as the more upsetting. Christine Harris, reviewing the evidence in 2003, argued that the sex difference was largely a product of the forced-choice format, and that other measures, physiological ones included, gave weak support to a sex-specific mechanism.

The second objection is about reasoning. In 1996 two groups of researchers independently proposed what is sometimes called the double-shot explanation. Suppose you believe, as many people do, that a woman who has sex with someone is probably in love with him, while a man can easily have sex without love. Then for a man, his partner's sexual infidelity implies emotional infidelity as well: two betrayals for the price of one. For a woman, her partner's emotional attachment implies that he will probably sleep with the other woman too, while his having sex implies much less. Each sex picks the infidelity that implies the other, and the forced choice splits without any difference in what people fundamentally fear.

The third objection is about culture. When a team led by Bram Buunk, with Buss among its members, put the dilemma to students in the Netherlands and Germany in 1996, the sex difference was present but smaller than in the United States, which suggests that whatever produces it is sensitive to the society people grow up in.

None of these objections has settled the matter, and the defenders have replied to each. Two meta-analyses published in 2012 read the accumulated studies differently: one found the sex difference across several kinds of measure, smaller outside the forced choice; the other found it largely confined to the forced choice. That is what a live scientific dispute looks like.

What is not live is the version that circulates. Magazines and popular psychology books regularly report that men care about sex and women care about feelings. That is not what the studies found even on the most favourable reading. In every version of this research, most people of both sexes find both kinds of infidelity deeply painful. The difference, where there is one, is in which they pick when somebody forces them to rank two catastrophes.

A forced choice is a powerful instrument, because it can reveal a preference that rating scales hide. It is also an instrument that turns a small difference into a split, because everyone has to land on one side or the other. Before trusting any finding of the form people prefer A to B, ask whether they were allowed to say both.