[Editor's Note: "Why Men Rape" by Randy Thornhill and Craig T. Palmer (January/February), has stirred passionate reactions--both among our readers and, via the mass media, across the nation. Letters from our readers ran the gamut: from huzzahs of approval to darts dripping with scorn, from harrowing first-person accounts by rape survivors to scientific dissections of Thornhill and Palmer's methodology. We have space only to print a sampling of the letters we received, but we thank all our correspondents for taking the time to express their views.]

On behalf of the sections of Women in Science, Psychology, and Anthropology of the New York Academy of Sciences, we are writing to object to the premise of the article by Randy Thornhill and Craig T. Palmer ["Why Men Rape," January/February 2000]. Although decades of critical research have provided at least partial biochemical explanations for what were previously thought to be purely behavioral phenomena, such as a genetic propensity for alcoholism, it has proved much more difficult to provide evidence that human behavior is subject to Darwinian natural selection. Indeed, so-called social Darwinism is now considered illegitimate by most social scientists because there is scant evidence for it.

Even from an evolutionary perspective, Messrs. Thornhill and Palmer's theory is flawed. For a rapist's genetic material to be passed on, thereby making rape a behavior that is selected for, rapes must result in pregnancies. But contrary to popular belief, the pregnancy rate from rape is quite low. And the assertion that "only" 22 percent of rape victims are subjected to gratuitous physical violence during the rape--as if the act itself were not violent--completely ignores the emotional impact of rape on the victim. In the aftermath of a rape, the victim often suffers post-traumatic stress disorder and other psychological damage that has long-term consequences for her ability to function as an effective mother for any ensuing offspring.

Messrs. Thornhill and Palmer's Darwinian perspective on human behavior is even more deficient. The claim that men's reason, respect for others, sense of responsibility and social morals can be overridden by a primal urge to pass on genetic material is indefensible, given the evidence for the complexities of human sexuality and its variability both across and within cultures.

The authors' assertion that rape takes place in other species as well as among humans is likewise a spurious line of reasoning. Unlike the mating practices of other members of the animal kingdom, human sexual behavior is informed by culturally induced desires and rationales. The fact that, as the authors state, "in many cultures rape is treated as a crime against the victim's husband," or that raping the women among one's enemies is an enduring tactic of war, provides strong evidence that rape is an act of violence used to assert power.

Rape is also a common means of control, domination and humiliation among male prison inmates, an activity without any procreative advantage or function whatsoever. Many of Messrs. Thornhill and Palmer's arguments seem suspiciously like intellectualized versions of archaic thoughts about rape that the feminist movement has worked so hard to do away with.

¶In spite of Randy Thornhill and Craig Palmer's insistence on distinguishing sexual assault involving penetration, or rape, from all other forms of sexual assault and from all other forms of violence against women, it is far from clear that they are entirely separate patterns of behavior. Men who rape in domestic settings commonly perpetrate nonpenetrative sexual assault and nonsexual assault as well. How can nonpenetrative sexual assault possibly be about reproduction? There is a clear linkage between those three forms of violence that is virtually ignored by the authors: the power difference between the perpetrator of the assault and the survivor.

Messrs. Thornhill and Palmer argue that rape is about sex, not power, because men are able to maintain an erection long enough to achieve penetration. Sexual arousal is obviously very complicated and is not necessarily related to reproductive potential in any clear way. Perhaps, instead, men who rape are sexually aroused by their power and domination over the women (or men, or children) whom they are assaulting.

The authors also argue that rape evolved as a reproductive technique to be employed when men were unable to attract partners through other methods. But that does not explain rapes perpetrated by men who are considered "attractive" according to the criteria put forth in the article: men who are physically fit, wealthy, apparently good potential fathers. It does not explain rapes perpetrated by men who have partners, or who have reproduced. It does not explain rapes perpetrated against women who are their partners, or with whom they have had consensual sex in the past. It does not explain rapes perpetrated by men against other men, or rapes by women. Messrs. Thornhill and Palmer dismiss rapes perpetrated against women who are not of childbearing age, indicating that such men are simply attracted to "inappropriate" women (i.e., women who are unable to participate in reproduction). Do they have any evidence to support this assumption?

The authors further argue that rape is a reproductive issue because women of childbearing age are more distressed by rape than are women of other ages. I question the ability of anyone to accurately assess emotional response--particularly someone who refers to the trauma experienced by a survivor of sexual assault as "distress." How were the responses of survivors quantified and compared? The questions asked by investigators to determine survivors' responses will obviously reflect the bias of whatever was determined to be an "appropriate" expression of emotional pain. A survivor who, subsequent to her rape, does not fear or hate all the men in her life is not necessarily experiencing less pain or trauma than a survivor who does.

Clare Jennings
McGill University
Montreal, Quebec

¶I applaud your publication for its willingness to publish Randy Thornhill and Craig Palmer's article on rape. Doubtless the article has led to criticism from some circles, but any serious scientist would approve of your decision. That is not to say the issue is closed--much more research remains to be done--but science can only progress by the open exchange of ideas, something you have done much to foster.

Andrew B. Heckert
University of New Mexico
Albuquerque, New Mexico

¶If rape does indeed confer an evolutionary advantage, the psychotherapeutic treatment proposed as a solution by Randy Thornhill and Craig Palmer will not do much to discourage the evolutionary tendency. The treatment, a kind of sex education, sounds suspiciously like the kind of solution that would be offered by the social scientists they are trying to debunk. A true Darwinian should realize that successful organisms adapt to changed environments, not simply to therapy. Thus the answer to rape, from a Darwinian perspective, lies in changing the environment to make it ill-suited to the unwanted genetic trait. In a couple of thousand years, we might be rid of it.

By relegating the proposed remedy to a "therapy program," policy makers would absolve the rapist of personal accountability for his crime. Rapists could claim that the state was remiss in not having instituted a sufficient evolutionary education program. But making government authorities responsible for rape is not likely to lead to a decline in the crime.

Frank R. Müller
George East, South Africa

Randy Thornhill and Craig Palmer assert that human rape is an evolved reproductive strategy, rather than an expression of hatred and violence against women. They base their conclusion, in part, on the reproductive behavior of scorpionflies of the genus Panorpa. But they provide no information in their article on whether, in Panorpa scorpionflies (not to mention in humans), there is any evidence (as opposed to assumptions) that rape actually leads to higher male reproductive success.

The only "evidence" that has been offered by Mr. Thornhill is in a 1980 paper in which he argues that males engaging in copulation by force have a lower risk of mortality than other males because they do not need to enter spiderwebs to capture nuptial prey items. Mr. Thornhill, however, does not report whether males that force copulation have previously attempted unsuccessfully to obtain nuptial prey from spiderwebs. If they have, those males could still be subject to high levels of predation. Furthermore, no empirical data on longevity or mortality rates under field conditions were gathered or reported to show that males that copulate by force really have longer life spans.

In their article for The Sciences, Messrs. Thornhill and Palmer state that, in scorpionflies, ". . . when females are willing, males are much more likely to achieve penetration and sperm transfer" than they are when the females are unwilling. In addition, Mr. Thornhill has published laboratory findings that thirty-two out of forty females were able to avoid forced copulations; in a subsequent paper, he acknowledges that forcing copulation is "the least successful" of the three mating strategies available to Panorpa males. Taken together, those data suggest that male scorpionflies that force copulations are not very successful. Without empirical data demonstrating that such males actually increase their overall or lifetime reproductive success as a result of their coercive behavior, Messrs. Thornhill and Palmer's argument falls apart.

The authors couch their argument in the language of evolutionary biology, but to our knowledge, no empirical studies have demonstrated that forced copulation in scorpionflies or rape in humans has any genetic basis. Messrs. Thornhill and Palmer do not mention any heritability studies that would buttress the assumption that forced copulation in scorpionflies is genetically based and subject to natural selection. On the basis of repeatability analyses, Mr. Thornhill suggested in a 1983 paper that in the scorpionfly Harpobittacus nigriceps (a species without forced copulation), the variation among males in their ability to capture nuptial prey is genetically based. But studies that apply rigorous quantitative genetic approaches to examine reproductive strategies in Panorpa do not appear to have been done.

Furthermore, natural selection can only act on phenotypic variation, but Messrs. Thornhill and Palmer present no information on the existence, nature or extent of variation in the trait in question, either in animals or in humans. In Panorpa, males can switch from one mating strategy to another; it is not clear how much variation exists in the likelihood of choosing one strategy over another, or in the frequency of switching. Thus the suggestion that human rape is under genetic control is not based on any solid genetic studies of Panorpa or of humans, but rather on a willingness to interpret bits and pieces of ambiguous data selectively. It is a classic example of adaptive storytelling (in the sense of Stephen Jay Gould's famous "just-so stories").

In one of the more scientifically startling passages in their paper, Messrs. Thornhill and Palmer discuss the "notal organ" of scorpionfly males, which holds the female during forced copulation, and they conclude that the organ evolved specifically for "raping" females. The authors then write:

Human males obviously have no external organ specifically designed for rape. One must therefore look to the male psyche--to a potential mental rape organ--to discover any special-purpose adaptation of the human male to rape.

What are the authors suggesting here? Apparently that the notal organ of scorpionflies is homologous to the "male psyche" of the human rapist. That proposition is an absurdity; there is no way that the authors, or anyone else with even a modicum of training in biology and evolution, could really believe that notal organs and human brains share a common ancestry by descent through evolutionary history. If one assumes instead that Messrs. Thornhill and Palmer are speaking about analogous organs, there is no empirical evidence to suggest that notal organs and "male psyches" evolved convergently as a result of common evolutionary pressures.

There are also numerous logical problems with Messrs. Thornhill and Palmer's arguments, which we have no space to elaborate here. Suffice it to say that even their own definition of rape includes behavior (e.g., oral and anal penetration) that cannot possibly lead to reproduction, and that rigorous studies by social scientists on rape trauma do not support many of their assertions. Moreover, psychological research underscores the complexity and heterogeneity of sexually aggressive behaviors, which appear to be influenced by a multiplicity of bio-psychosocial factors that are not reducible to simplistic, singular explanations such as the biological imperative to reproduce.

Zuleyma Tang-Martinez
Mindy Mechanic
University of Missouri–Saint Louis
Saint Louis, Missouri

¶The status of women must be considered when studying rape. Only recently--and, indeed, only in some countries--have women gained equal footing with men. The expectation of most American women, that they will have a relationship based on love, did not exist for most of human history. Men owned and ruled. Arranged marriages and routine affairs, including sexual intercourse, did not depend on a woman's consent.

Although as a woman, I do not care for the conclusions of Randy Thornhill and Craig Palmer, I cannot easily cast off the arguments they present. History provides some supporting evidence, such as the practices of separating the sexes and supervising mixed company. In recent decades, and particularly with the advent of the birth-control pill, neither rape nor out-of-wedlock births have declined.

I am sure the shriek of voices will continue: many people want their theories to support the ideas of free will and equality. But one must remember that equality is an ideal, not necessarily a reality. Rape is primitive, aggressive sex. The authors' theory seems to have a fair amount of foundation.

Elise Beaulieu
Scituate, Massachusetts

Randy Thornhill and Craig T. Palmer reply: We thank all the writers for their interest in our evolutionary analysis of rape. The responses to our article illustrate common misunderstandings--such as the naturalistic fallacy, the myth of genetic determinism and the failure to distinguish between proximate and ultimate causation--that are fully explained in our book, A Natural History of Rape.

Several of the responses, however, also misrepresent our central argument, attributing to us the view that human rape is an adaptation that was favored by Darwinian selection. In fact, we conclude that the existing evidence cannot determine whether human rape is an adaptation (i.e., a behavior that was favored by direct selection for raping) or a by-product of various other sexual adaptations. Following the basic principles of the scientific method, we then suggest various testable predictions that could settle that issue in the near future. Now that the strong emotional reaction that greeted the publication of our work has started to subside, we hope that the authors of the responses, and readers in general, will start to consider what is actually in our book. To aid them we will address some of the specific misunderstandings in these responses.

Nancy Steinberg, Nancy Tooney, Constance Sutton and Florence Denmark assume that an evolutionary-biological explanation is equivalent to a genetic explanation, as opposed to an environmental or cultural one. But an evolutionary-biological approach assumes no more than that every trait of an individual, including rape, is the result of a complex developmental interaction between genetic and environmental factors; neither factor is assumed to be the more important.

The four writers also appear to assume that we downplay the trauma of rape victims when we point out that only a minority of victims undergo severe physical injuries. We also explain that the crucial question is whether rapes may have resulted in pregnancies in ancestral, as opposed to current, environments, and that even a low pregnancy rate might have been sufficient for selection to favor raping behavior under certain past circumstances. In fact, we emphasize that the trauma is often even greater in cases in which the physical injuries are less severe, as one would expect from evolutionary theory. Finally, the fact that Ms. Steinberg and her colleagues equate evolutionary explanations with "social Darwinism" only shows how fully they subscribe to the naturalistic fallacy: the erroneous belief that scientific understanding provides moral guidance.

Clare Jennings also seems to mistakenly assume that evolutionary theory predicts that all behaviors will lead to reproductive success. The evolved differences between male and female behavior have given rise to many male sexual behaviors as by-products, including the specific nonreproductive sexual behaviors she mentions. In our book we explain how the data clearly falsify a simplistic "mate-deprivation" explanation of rape; nevertheless, the data do not eliminate the possibility that one of the factors contributing to some rapes is a difficulty in obtaining consensual sex. Finally, the methods for measuring the psychological pain of rape victims were introduced by the investigators who originally studied that pain; we simply used their data to test evolutionary predictions about the psychological trauma of rape.

Frank Müller equates our evolutionary approach with the idea that rape is genetically determined, and hence that it cannot be influenced by changing environmental factors. But in fact, we stress the importance of environmental factors throughout our book, as focal points for the kinds of intervention that might prevent rape. We also emphasize that identifying such factors in no way reduces the responsibility of the rapist. We discuss such factors as the conditions in which boys are raised, and the circumstances in the environments of men and women that may lead to rape.

The first sentence in the letter by Zuleyma Tang-Martinez and Mindy Mechanic erroneously opposes ultimate causation (evolved strategy) and proximate causation (motivation) as mutually exclusive alternatives. But those two causal frameworks are complementary in biology, not antagonistic; indeed, ultimate causation explains the existence of proximate causes.

Ms. Tang-Martinez and Ms. Mechanic discuss the matter of the current reproductive success of rape in people and scorpionflies. But that is a fundamentally different issue from the question of whether or not direct selection for rape was effective in the evolutionary history of the two species. An adaptation, which comes about as a result of direct selection in the past for some particular function, may now be adaptive or maladaptive. For example, the human preference for sugar is a psychological adaptation that evolved because it was adaptive in evolutionary history, but it is often maladaptive today.

Effective selection for rape behavior in any species is demonstrated by finding a rape-specific adaptation. The notal organ of the scorpionfly appears to be a rape-specific adaptation because evidence indicates it is functionally designed for rape. The empirical search for adaptation is the standard and scientifically rigorous method in evolutionary biology for identifying effective evolutionary selection pressures. In our book we discuss several potential rape adaptations that men may possess. If such a trait is found, it would be analogous to the notal organ (same selection pressure modifying different phenotypic substrates), not homologous.

Finally, Ms. Tang-Martinez and Ms. Mechanic's claim that our book is an example of what Stephen Jay Gould has called a "just-so story" is so clearly inaccurate that it is potent testimony to the power of ideology to cloud reasoning ability. A just-so story is one that is unskeptically accepted. In contrast, our book explicitly rejects the hypothesis that human raping behavior is an adaptation, because the existing evidence does not rule out the alternative hypothesis that rape is a by-product.

Like Andrew Heckert, we thank The Sciences for having the courage to allow the scientific method to be applied to controversial topics.