American Scientist

A Natural History of Rape: Biological Bases of Sexual Coercion. Randy Thornhill and Craig T. Palmer. xvi + 243 pp. The MIT Press, 1999. $28.95.

Why Sex Matters: A Darwinian Look at Human Behavior. Bobbi S. Low. xviii + 412 pp. Princeton University Press, 1999. $29.95.

The enterprise of trying to understand the evolved bases of human social behavior using a Darwinian framework is a worthy one indeed. The authors of both of these books state that an evolutionary paradigm may, in helping us to better understand ourselves, help us to better the human condition in which we toil. It has taken nearly 30 years for biological explanations for the social behavior of humans to gain widespread acceptance in the behavioral sciences, but today Darwinian science stands steady to help us better ourselves.

It is therefore all the more disappointing that Randy Thornhill and Craig T. Palmer, seeking to use principles of the nascent field of evolutionary psychology, have written a book about rape that is more of an ideological rant than an empirical, well-reasoned analysis. Because of its controversial subject matter, A Natural History of Rape has already received more attention than it deserves, but to summarize: The authors argue that rape is best understood in an evolutionary framework, as a behavioral adaptation molded by sexual selection. Rape, according to the authors, is a means of enhancing male reproductive success, a mating strategy used by some men as a heinous but viable alternative to the usual pattern of courtship. This of course has enraged those who consider rape to be first and foremost a violent act against women, rather than an attempt at reproduction. Thornhill and Palmer attempt to argue that rape occurs in contexts consistent with a Darwinian explanation, and they tell us over and over that their goal is to put an end to the behavior through a better understanding of it.

It is perfectly reasonable to hypothesize that rape, murder and any number of other behaviors have biological foundations. But let's consider Thornhill and Palmer's argument on its Darwinian merits, or lack of them. First and foremost, the authors make their argument nearly free of empirical evidence. They make sweeping species-wide statements about male-female mating preferences (for example, "Humans are mildly polygynous") taken straight from first-generation 1972 sociobiological predictions. They may not have noticed, but the discipline has moved far beyond these caricatures, aided by long-term field studies of nonhuman primates and other animals in which the predictions can be tested.

Thornhill and Palmer seem to loathe primary data and the scientific literature in general, preferring to cite recent pop books that survey the literature. At one point the authors cite a 1979 book eight times in two pages as their authority on human mating, even though that book (Evolution of Human Sexuality by Donald Symons) was an early think piece rather than a thorough review of actual behavior. The first data that bear on their argument do not appear in the book until page 100, and the latter half of the book is mainly an impassioned and rambling bashing of the social sciences.

I read chapter after chapter hoping to find evidence that bore on the key question, Is rape an effective mating strategy? Finally, at the end of chapter four, the authors briefly address that central question by analyzing the results of what they consider the best study conducted of the reproductive success of rapists. The answer? Rapists probably succeeded in impregnating their victims in only about 2 percent of cases. Moreover, in the study they cite, most rape-related pregnancies resulted in abortions or miscarriages; only 38 percent of the pregnancies led to a birth. The odds that a rapist will successfully father a child from a single attack are therefore less than 1 in 100. Weigh this gain against the potential cost to an accused rapist's lifetime reproductive success (as a result of imprisonment, ostracism, injury or death) and you have a negative sum gain for a rapist's effort. In other words, Thornhill and Palmer unintentionally but convincingly show that rape is certainly not a mating adaptation.

To make the case that rape is an evolved adaptation, the authors must argue that it is a cultural universal. Yet the number of pages in the book devoted to cross-cultural evidence of rape (two and a half, by my count) is less than the space given to accounts of "rape" in insects (four pages). This is, I presume, because Thornhill's real expertise is on the mating behavior of flies. Is there something wrong with this picture?

A Natural History of Rape is, I am afraid, the worst that evolutionary psychology has to offer. It proceeds by employing cardboard accounts of human behavior rather than by portraying behavior itself. More important, it illustrates a fundamental problem that evolutionary psychologists have yet to come to terms with. Whether they are seeking evolved cognitive mechanisms for incest avoidance, racism or rape, evolutionary psychologists have relied heavily on questionnaire studies—that is, they have taken what people say they do as solid evidence about what they do. Every behavioral scientist knows that people don't necessarily do what they say they do and that actual behavior matters more than words. A society may condemn rape in its moral code but tacitly allow it in practice. If a traditional culture espouses a love of its daughters, but meanwhile daughters are neglected in favor of sons, we should suggest that biological factors are at work. These questions can be addressed with behavioral and genetic data to either prove or disprove that the behavior is associated with enhanced reproductive success.

Thornhill and Palmer apparently scorn this approach and consider an adaptation to be merely any trait that seems well designed, a criterion that is subjective to the point of absurdity. They happily ignore behavioral evidence that a more careful evolutionary biologist would bring to bear—What is the reproductive payoff for rapists?—in favor of unsupported premises and provocative allegations.

Three decades of observational data on a wide range of animal species have allowed biologists to test hypotheses about evolutionary bases for myriad social behaviors. Do dominant males really father more infants? This question can be answered with long-term data on behavior and reproductive success, and the results (sometimes it's good to be dominant and sometimes it isn't) tend to complicate rather than simplify the picture. This is a sign of healthy, progressive science. Darwinian anthropologists who study human societies try to learn how and why cultural values and actual behavior are at odds; therein may lie some evidence of deeper drives. Thornhill and Palmer's book and the fairly positive reaction it has received from their colleagues show that many evolutionary psychologists are mired in ideology, in which allegations, stereotypes and suggestions serve as substitutes for data. A Natural History of Rape, focusing as it does on what the authors claim about rape rather than on much-needed data about the behavior and its context, does nothing to advance and much to hurt both the cause of rape prevention and that of evolutionary psychology.

Whereas Thornhill and Palmer fall flat in their attempt to show why sex matters, in Why Sex Matters Bobbi Low takes a far more expansive and better-documented look at the broad biocultural bases of human social behavior. Low has written a very thorough review of the current state of the art of human behavioral biology. Although she breaks no new ground, this is a book that could well be assigned in a course on human behavior or read by a nonscientist interested in knowing why all these biologists have turned their attention to the study of people. Low begins with a strong theoretical overview of the field and carefully explains why it is important to seek Darwinian explanations for human social behavior. The evidence for a role for evolution in human behavior is developed from Darwinian first principles and also from information about the reproductive ecology of a range of animal species. Low then moves on to examine the range of issues that human evolutionary ecologists have concerned themselves with over the past decade: explaining mate choice, the connection between status and reproductive success, warfare, and other aspects of the human condition that seem intuitively strange except in the light of biocultural explanation. Low considers the nuances and some of the contradictions of biocultural traits, which lends authority to her critiques. Why Sex Matters has something of a textbook feel, but it is deftly written and far more interesting for someone reading about human behavioral biology for the first time than a more standard textbook would be. It would make an excellent resource for students in undergraduate courses in biocultural issues in anthropology or biology.

Low slows down in her survey of human behavior long enough to devote several chapters to detailed examination of two areas: Reviewing her own research on sex, fertility and class in preindustrial Europe, Low takes a provocative look at how and why people in different social and economic settings have varied strategies for getting their genes and their resources into the next generation. Then in several authoritatively written chapters she examines biological and cultural aspects of warfare. Low weaves the synergy of culture and biology into her explanations, and the resulting book is a fine analysis of how the different levels of explanation form a complex tapestry.