An Evolutionary Perspective on Physical Attractiveness

Evolutionary Anthropology 97

DOUG JONES

Everyday experience suggests that physical attractiveness is important in per- to others, and that mate choices based sonal-and especially sexual-relationships. This impression is confirmed by a on these preferences have evolution- large body of social psychological research.’ z2 Cross-cultural surveys and ethno- ary consequences. However, he did graphic accounts show that concern with the attractiveness of potential mates is not take up the question of where pref- also common in non-Westernsocietiesand in tribal and peasant cultures3 However, erences come from in the first place - social psychologists and anthropologists have often had a hard time explaining why e.g., why peahens prefer peacocks attractiveness should count for so much, or why some features rather than others with large and showy trains. In recent should seem particularly attractive. The theoretical difficulties in accounting for work on sexual selection and mate physical attraction are brought out in a Brazilian saying, “Beleza n2o p6e na mesa” choice,6,7 this question occupies a cen- (“Good looks don’t put anything on the table”), which points to the absence of any tral place. This work asks how natural evident practical advantage to choosing an attractive mate. Faced with these diffi- selection might affect mating prefer- culties, a growing number of researchers in biology, psychology, and anthropology ences and why one set of mating pref- have turned to the modern theory of sexual selection, which has been highly suc- erences might lead to greater cessful in explaining nonhuman animals’ attractions to traits of no direct ecological reproductive success than another. utility. In this article, I survey recent efforts to apply the theory of sexual selection to From this perspective, mate choice is human physical attraction. seen as being guided by adaptations having the function of assessing the “matevalue”of potential mates, where Sexual selection occurs when some train of the peacock. He argued that mate value is the expected reproduc- individuals have characteristics that mate choice provides one avenue for tive success from mating with a given lead them to succeed in mating and the operation of sexual selection: individual relative to mating ran- fertilization at the expense of others of Traits without value in the struggle for domly or relative to some other base- their sex. Darwin, in On the Origin of existence might nonetheless persist in line reproductive success. Species4 and The Descent of Man and a population if they were attractive to This adaptationist approach to Selection in Relation to Sex,5 intro- members of the other sex and resulted mate preferences has several limita- duced the concept of sexual selection in more or better matings for carriers tions. Mate preferences are likely to be and argued that it had been important of these traits. The theory of sexual se- affected not only by adaptations for in shaping traits of no obvious value in lection via mate choice met with a cool assessing mate value, but by nonadag the struggle for existence, like the ant- reception in Darwin’s time, and played tive sensory and cognitive biases. Fur- lers of the stag and the extravagant little role in the Modern Synthesis of thermore, mate value may have an the 1930s and 1940s that united Dar- arbitrary component: Owing to the se- winian evolutionary theory and the lective benefits of having attractive new science of genetics. But the period offspring, it may be adaptive for indi- Doug Jones has carried out research from the 1970s to the present has seen viduals to adopt even idiosyncratic lo- comparing standards of physical an explosion of theoretical and em- cal mate preferences, provided these attractiveness in the United States, Brazil, Paraguay, and Russia. He is the author of pirical investigation that has strongly are both heritable and shared by the Sexual Selection, Physical Attractiveness vindicated Darwin’s interest in the rest of the population. Hence, the pre- and Facial Neoteny: Cross-Cultural topic. The modern literature on sexual dictive power of an adaptationist ap- Evidence and Implications, published in Current Anthropology, and Physical selection is reviewed at book length by proach may be limited. Nevertheless, Attractiveness and the Theory of Sexual Anderson6 and, in a more popular this approach is one of the most pow- Selection: Results from Five Populations, published by the University of Michigan vein, by Cr~nin.~ erful theoretical tools at our disposal, Museum of Anthropology. His future The theory of sexual selection as it and has considerable explanatory research will focus on the life-history has developed since the 1970s differs power even when organisms behave consequences of being attractive or unattractive. from Darwin’s theory in at least one maladaptively or in response to idi- important respect. Darwin was par- osyncratic local selection pressures. ticularly concerned to show that or- In this review of recent empirical Key words: Sexual selection, mate value, neoteny, fluctuating asymmetry, social learning ganisms prefer some potential mates and theoretical work applying the 98 Evolutionary Anthropology ARTICLES

TABLE 1. Adaptive Problem Relevant Physical Cues Possible Evolutionary Consequences Assessing the fecundity of a potential mate. Waist-to-hip ratio Fat deposition in female hips, buttocks, and (Among humans, this is a particular problem thighs, exaggerating low waist-to-hip ratio for males, given high age-related variability Facial propoflions Facial neoteny (especially in females, and in aduit female fecundity.) especially in populations where a large proportion of females survive past menopause) Skin color Reduced pigmentation (especially in females, and especially in climates where ecological costs of reduced pigmentation are low) Assessing the general health and parasite load Fluctuating asymmetry Structures specialized to advertise symmetry of a potential mate. (Health and parasite Anemia Structures specialized to advertise absence of load are likely to affect a mate‘s fecundity anemia (e.g. everted lips) and provisioning ability, and may correlate with genetic load.) Assessing mate value in the face of variation Average features Stabilizing selection preserving local or regional between populations in morphology, variants clothing, and adornment, selection Features favored by other Divergent selection exaggerating local or pressures, and correlates of mate value. population members regional variants modern theory of sexual selection to sessing the fecundity of potential the expected reproductive success human physical attraction, I will ad- mates. from a short-term mateship with a dress several questions: What major Two distinctive features of human partner of a given age, divided by the adaptive problems do humans face in reproduction are likely to be especially expected reproductive success from a assessing the more value of a potential important to understanding the rele- short-term mateship with a random mate? What physical cues reliably af- vance of fecundty to attractiveness in adult of the other sex. This figure ford information relevant to solving our species. First, in contrast to many shows that there is much more age-re- these problems? and What are the evo- primate species in which ovulation is lated variance in female than in male lutionary consequences of mate signaled by visual, olfactory, or behav- fecundity. In fact, given an age struc- choice? Three sections of this paper ioral cues, human females do not ad- ture realistically associated with a life take up adaptive problems, including vertise ovulation. Cues to female expectancy at birth of 35 years and evaluating the fecundity and the over- fecundity in our species track long- zero population growth, age-related all health and parasite load of a poten- term variations rather than short- variance in fecundity is almost 10 tial mate and allowing for local term, cyclical ones. Second, human times greater in adult females than variation in morphological cues and females, in contrast to most other pri- adult males. Consequently, there will selection pressures. A final section mate females (as well as human be much stronger selection on males considers the possible consequences males), experience major changes in than females to choose partners on the of mate preferences for the evolution fecundity before and during meno- basis of age. of human morphology. Adaptive prob- pause. Female fecundlty declines sig- It is also possible to use agelfecun- lems, relevant physical cues, and pos- nificantly after age 35 and falls to zero dity curves, in conjunction with sible evolutionary consequences are between ages 45 and 50, while a sub- agelmortality curves, to construct summarized in Table 1. stantial fraction of women commonly curves of “long-term mate value”’* live well past this age even under pre- that measure expected reproductive FECUNDITY AND AGE modern demographic regimes8z9 success in a mateship lasting until the The mate value of potential mates Figure 1 shows how marital fertility death of one partner relative to repro- will depend on their fecundity (the changes with age for females and ductive success from mating ran- likelihood that mating will result in males in natural fertility (noncontra- domly. Long-term mate value depends conception and the birth of offspring); cepting) populations, based on data on the age of both partners: a 30-year- the expected value of any parental care from Henry,Io Howell,6 and Goldman old woman will have a higher long- they will provide; and the genetic con- and Montgomery.’’ Although absolute term mate value for a 50-year-old man tribution they will make to the off- levels of fertility vary considerably than for a 20-year-old man because the spring. Fecundity may be particularly across populations, the shape of the 50 year-old is more likely to die or be- relevant to physical attractiveness: be- age versus fertility curve is sufficiently come infertile before his partner’s fe- cause past reproduction is an ex- stable that the curve in Figure 1 gives cundity declines significantly.Figure 2 tremely imperfect guide to present a good approximation of relative fe- shows curves for long-term matevalue fecundity, organisms may be espe- cundity in any natural fertility popula- given mortality rates realistically asso- cially dependent on physical cues tion. In Figure 1, fertility is expressed ciated with a life expectancy at birth rather than reproductive history in as- in terms of “short-term mate value,” or of 35 years. ARTIELES Evolutionary Anthropology 99

mented in the small number of studies that have made this a focus of research.13J4Thus, a plausible case can I\ be made on both theoretical and empirical grounds that human beings have universal adaptive emotional responses to signs of age, and that these P le5t responses operate especially strongly in males' responses to females. Many visible traits afford informa- i tion about female age and are thus good candidates for criteria of attractiveness. Furthermore, many traits that afford information about age-related changes in female fecundity are affected by levels of sex hormones that correlate with ovarian function, and zs.bO\*Q' Q\*m O\*O\bO\b+ N N m m 3 * v) IA 3 \o P P 00 00 Q' v) thusalsoprovideinformationonnon- 4 ?! 5 4 4 $ 5 4 5 9 5 & cn age-related differences in female fecundity. I will consider three such female age (years) traits and summarize recent research suggesting that they may indeed be universals of female attractiveness. B 2.5 - Waist to Hip Ratio A large body of research demon- -- 2 strates the importance of the waist-to- hip ratio (WHR), the ratio of the circumference of the waist at its nar- -$I 1.5 -- rowest and the circumference of the (d> hips at their widest, as a correlate of 49 female fecundity.15 Until puberty B 1.- WHRs for both boys and girls are about 0.9. At puberty, expansion of girls' pelves results in a decreasing 0.5 -- WHR. This expansion is accentuated by sex-specific fat deposition in the hips and buttocks, resulting in a WHR

11111II1 of 0.7 to 0.8 in females by age 20. From 03 I 100 Evolutionary Anthropology ARTICLES

A 2.5 cans, and Indonesians have similar 20-24 preferences for low female WHRs.16 He argued that although preferences 2 regarding absolute circumferences of waist and hips vary across populations and over time (which might be d2 1.5 predicted on evolutionary grounds, a> given variation in selective advantages 8 and disadvantages of fatness), prefer- 41 ence for a low ratio of waist to hips is likely to be relatively invariant because low WHR correlates with high 0.5 fecundity across most environments.

Facial Neoteny 0 Facial proportions also provide cues to age-related and non-age related changes in fecundity. In moving from youth to maturity, adults normally experience some reduction in the apparent size of the eyes and upper face relative to the overall height 2.5 T of the face, partly because of contin- 20-24 ued growth of the lower face and _____ partly because of sagging tissue age of 30-34 around the orbital area. Aging also normally results in increases in the 1.5 40-44 relative sizes of nose and ears as the cartilage in these organs outgrows the bone in the craniofacial skeleton, as well as reduction in the fullness of the lips as connective tissue atrophies." My own research on standards of physical attractiveness across cultures has focused on these changes in facial proportions and aesthetic responses to them.'*JBJ9 To date, this research has been carried out in the United States, Brazil, and Russia, and among the Ache Indians of Paraguay and the age of male (years) Hiwi Indians of Venezuela. (Kim Hill of the University of New Mexico as- Figure 2. A (top): Age and long-term mate value of females 0s a function of age of female and age of her partner. B (bottom): Age and long-term mate value of males as a function of age of sisted with data collection among the male and age of his partner. Ache and collected all data for the Hiwi.) The data collected include facial photographs of both sexes among magazine from 1955 to 1990 and win- lines of evidence that men judge Americans, Brazilians, and Ache, and women with WHRs at the low end of ners of the Miss America contest from ratings of the attractiveness of oppo- the normal range as being especially 1923 to 1987. Both centerfold models site-sex facial photographs of each of attractive. In one study, males in the and contest winners showed substan- these three populations by members United States were given a series of tial decreases in body mass index (a of all five populations in the study. drawings of female figures arranged in measure of body weight corrected for Measurements of the positions of a rectangular grid. Along one axis, height) over time, presumably demon- facial landmarks on the photographs these figures differed in body weight, strating changing attitudes toward fat collected confirmed that with increas- and along the other axis in WHR. and leanness. But while both waist ing age eyes grow smaller, noses larger, Within each weight class, males chose measurements and hip measurements and lips less full in relation to face females with the lowest WHRs as the declined over time, the ratio of the two height. Using multiple regression, it most attractive. Singh also has pre- was virtually constant - and low - at proved possible to construct equa- sented data on the body weight and 0.7. Singh also demonstrated that tions predicting age as a function of WHR of centerfold models in Playboy European Americans, African Ameri- eye width, nose height, and lip height ARTICLES Evolutionary Anthropology 101 among the Ache, the sample with the were judged to be especially attractive cieties, although the practice of tan- widest range of ages. These age-pre- by both English and Japanese raters, ning may have altered that preference dictor equations then made it possible as were caricatures based on Japanese in this century. Attraction to light skin to construct an index of “neoteny”for men’s ratings of Japanese women‘s and hair in the West is commonly at- each individual - the difference be- faces. In both cases, caricatures had tributed to social stratification in skin tween actual age and age as predicted large eyes and gracile In an- color resulting from colonialism,2sbut on the basis of facial proportions. other series of studies, measurements this argument does not stand up to his- (Thus, in this context, “neoteny”refers of photographic facial proportions torical scrutiny. Within Europe, at- to features shown to distinguish ado- have shown that women rated attrac- traction to light skin long antedates lescents and young adults from old tive by Asian, Hispanic, black, and Northern European political and cul- adults, not babies from adults.) This white males in the United States and tural dominance. In ancient Rome, index turned out to be a consistent by rural Taiwanese males have large light skin was a criterion of feminine predictor of female attractiveness rat- eyes and small noses and chin^.^^^^^ attractiveness and women commonly ings given by males from all five popu- Neotenous facial features may be lightened their skins with cosmetics. lations. In other words, across five The conquest and subordination of populations of raters, including two light-skinned and fair-haired North- relatively non-Westernized groups, fe- em barbarians did not erase this pref- males were rated as more attractive to erence; instead, blond hair, destined to the extent that their faces presented in ...across five be made into wigs for Roman ma- exaggerated or “supernormal” form populations of raters, trons, became a Northern export. the proportions distinguishing young including two relatively However, fair skin and hair were not women from old women. Neoteny was criteria of male attractiveness in an- neither positively nor negatively cor- non-Westernized cient Rome, at least in heterosexual related with female ratings of male at- groups, females were contexts.26 tractiveness. Other findings also rated as more attractive Preference for females with lighter- suggest that neoteny is a component than-average skin is not limited to of female, but not male facial attrac- to the extent that their Western societies. Van den Berghe and tiveness. Female models from the cov- faces presented in Frost2’ show that in a standard cross- ers of Glamour and Cosmopolitan cultural sample, 47 of 5 1 societies for magazines were found to have exaggerated or which preferences in female skin color strongly neotenous facial propor- “supernormal”form the are mentioned preferred light skin. tions-large eyes, small noses, and full proportions For the remaining four societies, lips-relative to a sample of American sources were ambiguous rather than female undergraduates, and age pre- distinguishing young showing a clear preference for dark dictor equations assigned them very women from old women. shn. The same sample showed a con- young ages. Male models were not sistent sex difference in skin-color found to be consistently neotenous preferences. Thirty societies explicitly relative to male undergraduates. l9 preferred lighter skin in females than Other researchers have reported in males; only three explicitly pre- attractive partly as a byproduct of similar results. In one study, men were ferred the reverse. These preferences male attraction to facial markers of asked to select the most attractive of were found even in groups like the youth, at least when these are com- several faces on a computer screen. A Marquesans, Tswana, and Rundi, in bined with adult body proportions. computer graphics program then pro- which Western cultural influence is But there probably are additional se- duced a new “generation” of faces by limited and standards of beauty are in lective advantages at work. Sym~ns~~ taking the face selected and making many respects at odds with Western notes that female facial proportions small random variations in it, thus standards. Indeed, the same prefer- can provide information about fecun- simulating the cumulative trial-and- ences were found even in tribal popu- dity beyond the information they pro- error process of natural selection. The vide about age. Pregnancy results in lations like the Yanomamo, where faces resdting from multiple genera- increased bone growth, which often class differences in exposure to sun- tions of this procedure were strongly includes lengthening of the face as light probably are virtually nonex- neotenous, with larger eyes in relation well as coarsening of facial features. istent. This last finding argues against to face height, more gracile jaws, Testosterone produced by male fe- the theory that light skin is attractive smaller noses, and fuller lips.’O In an- tuses may result in masculinization of merely because it acts as a status other study, researchers produced maternal facial features. marker in societies where upper-class caricatures of attractive faces by using females spend more time indoors. It is computer graphics software to exag- important to note that the preference gerate the differences between photo- Skin Color that van den Berghe and Frost found graphs of attractive and average faces. Light skin (and sometimes light is for skin at the light end of the I-,al Caricatures based on English men’s hair) is traditionally a component of range of variation, not necessarily for ratings of Caucasian English faces female attractiveness in Western so- “white” (i.e., pink) skin. Members of 102 Evolutionary Anthropology ARTICLES culturally isolated dark-skinned popu- search must be conducted in a wider specific features like neotenous facial lations are commonly reported as say- range of populations to confirm proportions and light skin. As noted, ing that they find the skin of whether these fecundity cues are in- ~ responses to skin color have been ar- Europeans too light to be attractive. deed universal or nearly universal cri- gued to reflect widespread symbolic Skin color preferences are close teria of cross-cultural female associations of lightness and dark- enough to a cultural universal to de- attractiveness. Also, more work needs ness. Some classical ethologists3*and mand a general explanation. Some re- to be done to refine our charac- some feminists33.34 have argued that searchers have suggested that terizations of how people respond to female youth and neotenous facial fea- near-universal symbolic associations fecundity cues in the other sex. Does tures are attractive to males more be- of light and dark may play a role.28Van the response to cues to potential mate’s cause of their association with female den Berghe and Frost, however, argue age vary with the age of the rater, as subordination and male dominance for the importance of female skin suggested by models of long-term than because of their association with color as a cue to fecundity. Infants are mate value presented in Figure 2? Do fecundity. There is evidence that con- light-slunned relative to related adults. different markers interact with one an- siderations of male-female power rela- Both males and females grow darker other so that, for example, the relative tionships play a role in mate choice through childhood, but at puberty with regard to some physical charac- girls’ skin color grows lighter while teristics. At least in England and the boys’ skin continues to darken. Young United States, partnerships in which women are relatively light-skinned, females are taller than their partners but grow darker with increasing age. Young women are are strongly stigmatized and are far Thus, changes in female skin color, at less likely to occur than would be ex- least from late childhood to old age, relatively ligh t-skinned, pected on the basis of chance assort- track age-related changes in fecundity. but grow darker with ment.35 This probably reflects what This link between slun color and fe- increasing age. Thus, James calls the “male superior norm,” cundity results from an endocrinologi- which dictates that women not be cal link between estrogen production changes in female skin physically or socially dominant over and melanin production. As a result of color, at least from late their mates.36However male attrac- this link, female skin color changes tion to female youth is unlikely to re- childhood to old age, flect male insecurity alone: Kenrickl4 slightly during the menstrual cycle, be- shown that adolescent males in ing lightest at ovulation. Women who track age-related has the United States generally report that use progesterone-based birth control changes in fecundity. they are attracted to females of their pills sometimes experience darkened own age or older, which is more con- complexions. Women’s skin also dark- sistent with the fecundity cue theory ens during pregnancy, sometimes per- than with the dominance and insecu- manently. Thus, the widespread male attractiveness of different facial fea- rity theory. Jankowiak3’ notes that eth- preference for light-skinned mates, tures varies depending on the propor- nographies do not suggest that old like the preference for low waist-to-hip tions of the rest of the body? men involved with much younger ratios and neotenous facial propor- Finally, more research needs to be women generally find them especially tions, may be an adaptation for choos- done to test and extend the adaptation- easy to dominate. ing a more fecund mate.29,30 ist rationale suggested for these traits. Let me summarize the argument up to this point. Given the substantial The fecundity cue hypothesis is theo- HEALTH INDICATORS retically appealing because it ties to- changes in adult female fecundity as- Although fecundity cues are likely sociated with aging and the absence of gether several different lines of evidence. It seems to explain why hu- to be especially important in men’s any well-developed signals of monthly evaluations of women, the past or cur- man males, in contrast to the males of variation in fecundity, human males rent health status of a potential mate many other species, seem to be more must have been subject to strong selec- is likely to be important to both sexes. concerned than are females with the tion to respond reliably to cues associ- An organism that has been subject to attractiveness of potential mates.31 ated with age-related and other developmental stress in the past is noncyclical variations in female fecun- This hypothesis makes concrete pre- likely to be more vulnerable to further dity. A low waist-to-hip ratio, dictions about the relationship be- stress; an organism in bad health to- neotenous facial proportions, and tween age and sexual attractiveness day is more likely to be sick or to die lighter-than-averageskin all have been that are supported by current re- tomorrow. A female who chooses a consistently associated with high fe- search. Further, it suggests specific healthy mate may increase her male fecundity throughout the course traits that are likely to be components chances of having a parent around to of human evolution. Research to date of female attractiveness. However, al- assist with child care and, if vulner- suggests that these traits are indeed ternative hypotheses that are not re- ability to stress and parasitism are widely attractive to human males. lated to fecundity have been suggested partly heritable, she may also ensure Although results to date are promis- to explain both male attraction to that her offspring inherit genes for re- ing, they are also preliminaq. More re- young adult females and attractions to sistance to stress and parasites. AFUKLES Evolutionary Anthropology 103

~ ~ ~~~~

Thus, adaptive mate choice in- cal literature on fluctuating asymme- gests that a woman’s orgasms are volves, in part, choice on the basis of try.) adapted not so much to increase the cues to good and poor health. To some Recent studies have shown that overall probability of conception as to extent, such cues will vary with par- fluctuating asymmetry plays a role in ensure that some of her sexual part- ticular stressors. Social learning is sexual selection. Th~rnhill~~has ners rather than others father her likely to playa role in adapting individ- shown that female scorpionflies prefer children.49Thornhill et al.48findthat ual standards of attractiveness to such more symmetrical males and even pre- the women most likely to report or- ecologically variable indicators of fer the odors of more symmetrical gasm before or after male ejaculation health status. But other traits provide males (suggesting that symmetry and are those whose partners have the low- reliable indications of past stress and odor provide correlated cues about est levels of fluctuating asymmetry. current parasite loads across a wide health status). M0llefi3 has demon- One possible explanation for this re- range of stressors. Two such broad- strated that female barn swallows pre- sult is that males with high levels of spectrum health cues are fluctuating fer males with tails that have fluctuating asymmetry were histori- asymmetry and anemia. Both traits artificially been made more symmetri- cally poor bets as fathers because they are good candidates for cross-cultural cal. He has also shown that tail sym- had experienced greater developmen- universal criteria of low attractive- metry is negatively correlated with tal stress or carried heavier genetic ness. parasite loads.44Swaddle45 has found loads and that, as a result, female sex- that zebra finches prefer symmetry ual response displays an adaptive sen- Fluctuating Asymmetry even in artificial traits (colored leg sitivity to fluctuating asymmetry or This term refers to small random bands). This research suggests that qualities correlated with it. deviations from symmetry.38Fluctuat- many of the exaggerated ornaments These studies suggest that fluctuating asymmetry is different from direc- produced by natural selection, which ing asymmetry is important in human tional asymmetry, which involves have attracted the attention of re- sexual selection, but do not show that species-typical biologically adaptive searchers from the time of Darwin, are symmetry directly contributes to at- differences in the size and shape of the adapted at least in part to advertise tractiveness. Several studies have opposite sides of the body. Fluctuating symmetry. Precisely because large and shown that images of faces made arti- asymmetry, by contrast, is commonly showy adornments are so subject to ficially more symmetrical are usually a result of developmental stress. Stress developmental disruption, individuals less attractive than the original im- introduces “noise”into the process of who display large and showy yet sym- age~.~~,~~However, as noted earlier, development, making it more difficult metrical adornments provide poten- symmetry can take several forms, infor growing organs and body parts to tial mates with especially convincing cluding fluctuating and directional follow the developmental paths laid evidence of their freedom from stress asymmetry. The finding with regard to down by natural selection. Because and parasite loads. the symmetry-enhanced photographs random departures from developmen- Research on humans also suggests suggests that a certain amount of di- tal canalization are not llkely to be of a role for symmetry in sexual selec- rectional asymmetry actually adds to identical direction and magnitude on tion. Gangestad et al.46show that facial attractiveness, so that when fa- opposite sides of the body, develop- measures of overall bodily symmetry cial images are made more symmetri- mental stress is likely to produce fluc- are positively correlated with ratings cal the reduced attractiveness tuating asymmetry. The same of physical attractiveness based on fa- associated with lower directional reasoning suggests that genetic load cial photographs. Although raters can- asymmetry outweighs any gain associ- resulting from mutational load or ho- not assess overall symmetry from ated with lower fluctuating asymme- mozygosity is also likely to increase facial photographs, an underlying re- try. Studies based on actual facial fluctuating asymmetry, sistance to stress may influence both measurements have yielded mixed re- A large body of research suggests facial symmetry and attractiveness. sults. Grammer and Thornhill have that developmental stress and genetic Thornhill and Gangestad 47 have dem- shown a strong relationship between load are indeed associated with in- onstrated that measurements of the photographic facial symmetry and creased fluctuating asymmetry. body symmetry of males are corre- ratings of attracti~eness.~~My own Among nonhuman organisms, fluctu- lated with age at first intercourse and data do not show any consistent sig- ating asymmetry is associated with number of sexual partners. Even more nificant relation between facial fluctu- nutritional deprivation, exposure to remarkable are the findings of Thorn- ating asymmetry and ratings of facial pollutants, and increased homozygos- hill et ale on symmetry and female attractiveness when symmetry is as- ity due to inbreeding.39 Among hu- sexual response. These authors’ find- sessed either on the basis of photo- mans, correlates of fluctuating ings build on recent research suggest- graphic facial landmarks or on the asymmetry include premature birth, ing that female orgasm simultaneous basis of direct facial measurements.I2 maternal diabetes, mental retarda- with or subsequent to male ejaculation (And even when subjects report tion, and psychosis.40 Paleopatholo- draws sperm from the vagina into the greater attraction to more symmetri- gists often use fluctuating asymmetry uterus, thus increasing the probability cal faces, they may, of course, be rein skeletal remains to assess popula- of conception. Because female orgasm sponding not to symmetry, but to tion stress. (But see Smith et al4I for a is far from an inevitable occurrence other facial characteristics correlated critical review of the paleopathologi- during intercourse, this research sug- with it.) 104 Evolutionary Anthropology ARTICLES

In summary, intriguing evidence LEARNING guishing features of the criminal face suggests a connection between fluctu- in particularly clear form. He was dis- Whereas some traits are relatively ating asymmetry and sexual selection appointed. “The individual faces are invariant indices of mate value, other in humans. The evidence that more villainous enough, but they are villain- indicators of mate value may vary over symmetrical faces are perceived as be- ous in different ways, and when they time and space. Microevolutionary ing more attractive is mixed. It may be are combined the individual peculiari- forces lead to local differentiation, so that fluctuating asymmetry has its ties disappear.”He found a similar irn- that an organism choosing a mate may greatest effect on attractiveness below provement in looks when he blended face a different distribution of metric the neck; Singh has presented evi- the faces of noncriminals. “All com- variation than did its ancestors. dence that breast symmetry is a composites are much better looking than Among humans, ornamentation and ponent of female attractivene~s.~~ their components, because the aver- body modification may further alter the sexual stimuli to which individuals aged portrait of many persons is free Anemia are exposed. In addition, selection of the irregularities that variously pressures are likely to change from blemish the looks of each of them.” Although it has not received the Although Galton’s experiments pro- same attention in the present context vided no support for the theory of a as fluctuating asymmetry, anemia is a distinctive criminal physiognomy, general index of health status in hu- they quite incidentally suggested that mans and other species that is likely to proportions close to the population play a role in judgments of physical The evidence that more average are particularly attractive. attractiveness. Organisms commonly symmetrical faces are Langlois and RoggmanS8used com- respond to infection by removing iron puter graphic software to test this from the blood stream, resulting in perceived as being “averageness effect.” They found that lower levels of blood hemoglobin. Re- more attractive is composite facial pictures produced by cent biomedical research suggests that mixed. It may be that blending individual faces generally anemia is an adaptive response to in- are rated as being more attractive than fection and neoplasia. Iron has an im- fluctuating asymmetry portant function in cell reproduction has its greatest effect most of the original pictures making so that by sequestering iron, the body up the composite. In addition, the may limit the reproduction of patho- on attractiveness below more faces going into the composite, gens and tumor cells. Of course the neck. the more attractive it seems. These chronically reduced blood hemoglo- findings suggest that the human brain bin levels will exact a toll on the organ- contains something like a face-averag- ism, lowering work capacity and ing device, which combines images of growth, but this price may be worth perceived faces to generate an image one generation to another, so that one paying in the face of serious infec- of a prototypical face (or, more likely, generation’s optimum may not be an- ti~n.~~Thus an anemic pallor is often a prototypical female and male face). other’s. Thus far, I have reviewed traits a sign of infection, and one might ex- New faces are then judged as attrac- that may provide consistent informa- pect it to figure in mate choice. Fur- tive or unattractive in part accordmg tion across a wide range of environ- to how closely they match the proto- thermore, if anemia is an adaptive ments about the fecundity and health response to parasitic infection, then the typical face. If facial attractiveness is of potential mates. In this section, in health costs of maintaining high hemo- partly the absence of distinguishing contrast, I will consider how learning globin levels may ensure that organisms features, this would also explain the rules might systematically modify carrying heavy parasite loads will not finding that attractive faces are harder preferences in the face of local vari- risk faking a healthy ruddiness. to remember.59However, Langlois and ation in traits and selection pressures. Females and males in a number of Roggman’s findings do not imply that primate species develop conspicu- the most average face is the most at- ously reddened vascularized struc- Koinophilia tractive face.60Although they found tures that are attractive to the other Between 1878 and 1881, Galton car- that their composite faces were more sex.55These structures may be adapted ried out a series of experiments in attractive than most of the faces that to advertise high hemoglobin levels composite photography57 By briefly went into them, a few of the original and freedom from infection. Among exposing a photographic plate to a suc- faces were rated as more attractive humans, many peoples have been re- cession of facial photographs, he was than any composites. Other studies ported to find red shn and lips attrac- able to produce a new photograph that suggest that, at least where males’ tive. It may be that everted human lips blended the characteristics of all those evaluations of females are concerned, have evolved to provide conspecifics faces-a kind of “average” face. Gal- the most attractive face is a supernor- with information about anemic ton’s original aim was to test the physi- mally neotenous version of the aver- ~tatus.~6The role of anemia in mate ognomic theory that criminals have age face.2* choice in humans and other species is distinctively villainous faces; he hoped Several different explanations have thus a promising topic for future in- that blending photographs of numer- been proposed as to why averageness vestigation. ous criminals would reveal the distin- might be a component of facial attrac- ARTICLES Evolutionary Anthropology 105

tiveness. First, faces with average pro- it is not clearwhat adaptive basis there first female to arrive must inspect portions might be easier to “read and is for cultural variation in standards of both males and then choose one as her might provide more accessible infor- attractiveness - that is whether vari- mate. There is some chance that this mation about age or other compo- ation tracks local differences in direc- female’s experiences with the two nents of mate value. My own tional selection pressures or is a males are misleading, so that she acci- experience and that of several other nonadaptive byproduct of symbolic dentally chooses the male with a lower anthropologists suggests that one may biases. mate value. Each subsequent female do a particularly bad job of estimating The influence of social cues on mate observes the choices of all preceding the ages of other people when they choice and standards of attractiveness females and bases her choice either on have unfamiliar facial proportions. is a topic that deserves considerably her own assessment or on the choices Second, average features might be more investigation. Social cues seem of the preceding females. Bikchandani adaptive features. Koeslag argues that sometimes to play a role in the mate et al. have shown that this process of “koinophilia”- attraction to average choices of nonhuman organisms.65 In sequential choice eventually results in or modal values of traits - is a wide- some species, females who see a male an “informational cascade.” At some spread and generally adaptive phe- in the company of other females are point, if a female sees enough females nomenon because deleterious subsequently more attracted to hm. clustered around one male, it makes mutations and environmental stresses In several cases experimental manipu- sense for her and for all subsequent are more likely to push organisms females simply to ignore her own ex- away from the average than toward perience and choose the more popular it.61Nonaverage proportions are cor- male. The authors also have shown related with a variety of developmen- that where individual experience is tal problem~.~*76~The assumption that It is not clear to what fallible, there is a strong possibility average features are the most adaptive extent population that the cascade will settle on the features will be strictly true of course wrong male. Thus, if the first female only for traits under stabilizing selec- differences in standards makes the wrong choice there is a tion; other mechanisms and learning of attractiveness result chance that she will influence the sec- rules will be required for traits under ond female to choose wrongly, in directional selection. But a preference from differences in which case the third female will be for average features may be the prod- culture and to what even more likely to choose wrongly, uct of a default rule leading to a pref- and so on. In other words, although it erence for the average value of a trait extent they result from may be adaptive, on average, to rely on unless there is some special factor fa- habituation to voring extreme values.24 the experiences of others, the wide- morphological spread habit of depending on social differences. cues may allow the whole population Social Cues to settle on the less adaptive alterna- Several lines of evidence, including tive as a result of the many copying the cross-cultural agreement in standards mistakes of a few. This principle ap- plies quite generally to social learning, of attractivenessIs and differential relations have shown that when females and means that it is dangerous to as- sponses of very young infants to im- are attracted to “popular” males it is sume that any one item of culture is ages of attractive and unattractive not merely because they seek the com- adaptive. Similar results, with some- adult faces,64suggest that standards of pany of other females or because a what different models, have been ob- attractiveness are not purely the prod- male’s earlier successes with females tained by Rogers6’ and Boyd and uct of arbitrary cultural variation. change his behavior in a way that Richerson.68 However, available data leave consid- makes him more attractive. Instead, erable room for disagreement about Another set of models suggests a dif- females seem to take popularity with how much, how, and why culture in- ferent rationale for attention to social other females in and of itself as a sign fluences standards of attractiveness. It cues in the development of standards of high mate value. However, it is not is not clear to what extent population of attractiveness. Fi~hel-6~argued that clear at present whether females ob- differences in standards of attractive- sexual selection might be self-rein- ness result from differences in culture serving the mate choices of other fe- forcing. Imagine a population in and to what extent they result from ha- males will be attracted not merely to which most females have a mild pref- bituation to morphological differ- those males but to physically similar erence for longer-than-average tails, ences. It is not clear to what extent the males. but a few have exceptionally strong influence of culture on standards of at- This dependence of attractiveness preferences. The last group of females tractiveness is direct (i.e. through in- on social cues has won the attention of is especially likely to choose males dividuals copying other individuals’ mathematical modelers. Bikchandani with long tails as mates and, assuming preferences) or indirect (e.g., through et al. 66 have considered a model in that tail length is heritable, to have off- differences in clothing or child-rear- which a number of females arrive se- spring with long tails. Given the gen- ing practices resulting in sexual im- quentially in a breeding territory and eral preference for long tails, these printing on different stimuli). Finally, must choose one of two males. The offspring are likely to have above-av- 106 Evolutionary Anthropology ARTIUES

territory, but this work suggests that it I I may be adaptive as a general rule to 3A Breasts adopt even idiosyncratic and arbitrary I local preferences, provided that preferred traits are heritable and give offspring a significant edge in attracting mates. These models suggest several possible adaptive bases for cultural influence on standards of attractiveness. Koinophilia will track local variations 1.00 2.00 3.00 in selection pressure as long as traits ideal size are under stabilizing selection. But for I I traits under directional selection, av- I I erage phenotypes will not be optimal 3B Buttocks phenotypes, and organisms may de- I pend on social cues for additional information about correlations between morphology and mate value. Run- away sexual selection may offer a further selective advantage to copying local mate preferences. How far these models can explain variation in standards of attractiveness is unclear, but I 1.00 2.00 3.00 I there does seem to be considerable ideal size cross-cultural variation in standards of attractiveness that goes beyond what would be expected merely on the basis of koinophilic preferences for lo- 33c Thighs Figure 3. Standards of attractiveness vary across cal modal features. Darwin recorded I I cultures. Mean responses of Brazilian, U.S., and B.78 some of the differences in standards 1 Russia Russian females (N=27,29,37) and males 1.89 (N=30,22.37) regarding ideal female body of attractiveness reported by mission- 0 female types. Subjects were asked “Would this woman aries and explorers of his time5 and be most attractive if her breastslbut- reports of local preferences for exag- R male tockslthighs were the same size, larger, or smaller?” 1 =smaller than silhouette is most at- gerated traits are commonly seconded tractive: 2=same sizeas silhouette is most attrac- by modern ethnographers. In their tive; 3=larger than silhouette is most attractive. cross-cultural survey of sexuality, Kruskal-Wallis one way nonporametric Anova Ford and Beach3’ noted considerable shows significant (pc.05) between-population variation in which physical features kideal size variation in ideal female dimensions for all features for both sexes of raters. are the focus of attention, as well as in their preferred sizes and shapes. My own research has documented a dif- erage mating success. In other words, positive feedback process-called ference in erotic focus between the females with an unusually strong pref- runaway sexual selection-is easy to United States and Brazil, with Ameri- erence for long tails are likely to have get started. However, they also raise can men focusing more on breasts and more grand-offspring than are other questions about whether the process Brazilian men focusing more on but- females. If female preferences are would be likely to lead to transitory tocks.’*The ideal female figure among heritable, the result will be a self-rein- rather than permanent exaggeration Americans in my sample has larger forcing spiral of increasing frequen- of traits.’O breasts and slimmer buttocks and Most population genetic models of thighs than does the Brazilian ideal cies of trait genes and preference runaway sexual selection assume that (see Fig. 3). genes. Thus, even a preference for an preferences are genetically transmit- One source of evidence for the im- arbitrary trait - one that is not a reli- ted. But Laland7I has recently devel- portance of social cues in the develop- able indicator of a mate’s fecundity, oped alternative models with ment of standards of physical health, or likely contribution to off- culturally transmitted preferences. He attractiveness comes from changes in spring survival - may be adaptive if it shows that cultural transmission of standards of attractiveness in the leads to the production of offspring preferences can result in runaway sex- wake of Western economic, political that members of the next generation ual selection and rapid morphological and cultural expansion. Explorers and find particularly attractive. Popula- evolution. Gene-culture models of sex- anthropologists commonly note an tion genetic models suggest that this ual selection are still relatively new aversion to European physiognomies ARTICLES Evolutionary Anthropology 107 among culturally isolated popula- consciousness-raisingamong political females have higher death rates than tions. Ache Indians in Paraguay find activists, but they have met with lim- Ache males up to puberty, at which both European and African-American ited success so far within the general point their death rates drop below faces unattractive, but are attracted to population. My research confirms those for males. In part, this change the more physically similar Asian- that somatic prejudlce is widespread, occurs because females start attract- Americans. Ache aesthetic responses but not universal, among both black ing food and protection from males9 seem to be a reaction to somatic dis- and white Bahians. However, I have Under a selective regime of this sort, tance rather than a reflection of found a nonlinear relationship be- in which markers of fecundity attract stereotypes about members of these tween race and ratings of attractive- male investment, sexual selection will populations. But standards of attrac- ness: Individuals with pronounced favor an exaggeration of fecundity tiveness have undergone Western- African facial features are rated as less cues such as a low waist to hip ratio, ization in many formerly isolated attractive than others, but individuals neotenous facial features, and fair populations. Wagatsuma72 docu- with intermediate characteristics are skin beyond their ecologically opti- mented this process in Japan. The in- not rated as less attractive than indi- mum levels. It will also favor sexual itial aesthetic response of Japanese dimorphism in these traits. exposed to Westerners was mostly Sexual selection may also favor the negative. One samurai72 visiting evolution of specialized structures to Washington, D.C. in 1860 wrote “The advertise health status and resistance women‘s skin was white, and they .. .culture-specific to stress. These might include com- were charming in their gala dresses plex, easily disrupted symmetrical decorated with gold and silver but or other local structures advertising resistance to their hair was red and their eyes preferences for extreme the stresses that produce fluctuating looked like dog eyes, which was quite traits, enduring, asymmetq, as well as vascularized ar- disheartening.”However “[tlhe subtle, if eas of skin, like the lips, which adver- not fully conscious, trend toward an may contribute to tise an absence of anemia. idealization of Western physical fea- evolutionary Sexual selection in favor of traits tures by the Japanese apparently be- that consistently signal fecundity and came of increasing importance by the divergence between health across a wide range of popula- twenties,” and has continued up to the populations. In the long tions will exert a species-wide direc- present da~.~2Similar Westernization tional pressure. Selection will then of standards of attractiveness has been run, diverging sexual accentuate such traits up to the point noted for the Philippines, the Carib- selection may be an at which viability costs balance sexual bean, and many other societies (re- important agent of benefits. This point may differ from viewed in Jones).12 one population to another depending My research has examined “somatic speciation. on ecological conditions. Koinophilia prejudice” in Bahia, in the Brazilian and local preferences for exaggerated Northeast. Bahia has a racially strati- traits will have different conse- fied society, in which people at the top quences. Koinophilia will act as a con- of the local economic and political hi- viduals with pronounced European servative force in evolution, erarchy are of predominantly Euro- features. Further research is needed to preserving variants already present at pean ancestry, whereas the majority of determine whether this nonlinear re- high frequencies but making it more the population is of predominantly Af- lation between status markers and difficult for even advantageous novel- rican origin. As in much of Latin perceptions of attractiveness is a gen- ties to obtain an evolutionary foot- America, race is perceived more as a eral feature of somatically stratified hold. Koinophilia, while helping to continuous than as a categorical vari- societies or whether it is specifically maintain regional variations in mor- able, and no clear line divides black related to the Brazilian system of ra- phology,74 may on a longer time scale and white into distinct groups. Ba- cial stratification, in whch people of play a role in the reproductive isola- hians use a complex terminology to re- mixed race escape some of the stigma tion of species and in evolutionarysta- fer to racial mixture. Because that attaches to blacks via what Degler S~S.~~By contrast, culture-specific or appearance is more important than calls “the mulatto escape hatch.”73 other local preferences for extreme descent in Bahian racial classification, traits, if enduring, may contribute to full siblings may be given different ra- evolutionary divergence between EVOLUTIONARY CONSEQUENCES cial labels. A number of researchers populations. In the long run, diverging have reported somatic prejudice in I end where Darwin began, by con- sexual selection may be an important Brazil against African features.73In re- sidering the possible evolutionary agent of speciati0n.~6 cent years, Afro-Brazilian cultural as- consequences of standards of attrac- Darwin argued that sexual selection sociations have attempted to combat tiveness. Table 1 lists some of the ways made a major contribution to the evo- this prejudice by sponsoring beauty in which the preferences I have con- lution of human morphology. Many of pageants aimed at encouraging an sidered here might affect morphologi- Darwin’s speculations about sexual se- Afrocentric standard of beauty and by cal evolution. For example, Ache lection and the descent of man are al- 108 Evolutionary Anthropology ARTICLES most certainly wrong. The relative Fecundability and husband’s age. SOCBiol York: Anchor. hairlessness of the “naked ape” is 36:146-166. 35 Gillis JS, Avis WE (1980)The male-taller 12 Jones D (1996)Physical Attractiveness and norm in mate selection. Pers SOCPsychol Bull more likely a thermoregulatory adap- the Theory of Sexual Selection: Results from 6:396-40 1. tation” than a sexual display; vari- Five Populations. Ann Arbor: Museum of An- 36 James WH (1989) The norm for perceived ation in such traits as body thropology Press, University of Michigan. husband superiority: A cause of human assor- 13 Jankowiak W (1993)Sex, Death and Hier- tative marriage. Soc Biol36:271-278. circumference7*and nose shape79 archy in a Chinese City: An Anthropological Ac- 37 Jankowiak W (1995) A commentary on shows clear correlations with climatic count. New York: Columbia University Press. “Sexual selection, physical attractiveness and variables and straightforward func- 14 Kenrick D (1 994) Evolutionary social psy- facial neoteny.” Curr Anthropol36:737. chology: From sexual selection to social cog- 38 Van Valen L (1 962) A study of fluctuating tional rationales. But sexual selection nition. In Zanna M (ed) Advances In asymmetry Evolution 16:125-142. remains a plausible explanation for a Experimental Social Psychology, pp 75-1 2 1. 39 Parsons PA (1 992)Fluctuating asymmetry: number of other characteristics, in- New York: Academic Press. A biological monitor of environmental and 15 Singh D (1993)Adaptive significance of fe- genomic stress. Heredity 68:361-364. cluding everted lips, and scalp hair in male physical attractiveness: Role of waist-to- 40 Livshits G,Kobylianski E (1991)Fluctuat- its various forms, as well as sexually hip ratio. J Pers SOCPsychol 65:293-307. ing asymmetry as a possible measure of devel- dimorphic traits such as female 16 Singh D, Suwardi L (1 995) Ethnic and gen- opmental homeostasis in humans: A review. der consensus for the effect of waist-to-hip ra- Hum Biol63:441461. breasts, buttocks, and thighs and male tion on judgment of womenj: attractiveness. 41 Smith BH,Garn SM, Cole PE (1982)Prob- facial and body hair. These charac- Hum Nat 6:51-65. lems of sampling and interference in the study teristics seem comparable to social 17 Enlow DH (1990)Facial Growth. Philadel- of fluctuating dental asymmetry. Am J Phys phia: Harcourt Brace Jovanovich. Anthro 58281-289. signals found in other primate species. 18 Jones D, Hill K (1993)Criteria of physical 42 Thornhill R (1992)Fluctuating asymmetry Physical anthropologists have pro- attractiveness in five populations. Hum Nat and the mating system of the Japanese scor- duced of studies of “natural selection 4~271-296. pionfly, Panorpa japonica. Anim Behav 19 Jones D (1995) Sexual selection, physical 44:867-879. in action” for many traits of ecological attractiveness, and facial neoteny: Cross-cul- 43 Mdler AP (1992) Female swallow prefer- relevance. What is needed now are tural evidence and implications. Curr Anthro- ence for symmetrical male sexual ornaments. comparable studies of sexual selection pol 35:723- 748. Nature 357:238-240. 20 Johnston VS, Franklin M (1993)Is beauty 44 Merller AP (1 992) Parasites differentially in- in action. Studies of natural selection in the eye of the beholder? Ethol Sociobiol crease the degree of fluctuating asymmetry in in animal populations suggest that the 14:183-199. secondary sexual characters. J Evol Biol effect of sexual selection on morpho- 21 Perrett DI, May KA, Yoshikawa S (1994) 5:691-700. logical traits is often at least as power- Facial shape and judgments of female attrac- 45 Swaddle JP (1996) Reproductive success tiveness. Nature 368:239-242. and symmetry in zebra finches. Anim Behav ful as that of selection resulting from 22 Cunningham MR (1986) Measuring the 51 :203-210. differential mortality.*O Both evolu- physical in physical attractiveness: Quasi-ex- 46 Gangestad SW, Thornhill R, Ye0 RA (1994) tionary causes and consequences periments in the sociobiology of female facial Facial attractiveness, developmental stability, of beauty. J Pers Soc Psychol 5:925-935. and fluctuating asymmetry. Ethol Sociobiol standards of physical attractiveness 23 Cunningham MR, Roberts AR, Barbee AP, 15:73-85. have proven to be immensely impor- Bruen PB (1995)“Their ideas of beauty are, on 47 Thornhill R, Gangestad SW (1994)Fluctu- tant in the study of nonhuman organ- the whole, the same as ours”: Consistency and ating asymmetry and human sexual behavior. variability in the cross-cultural perception of Psychol Sci 5:297-302. isms. Research to date suggests that female physical attractiveness. J Pers SOCPsy- 48 Thornhill R, Gangestad SW, Comer R they are likely to be important in the chol68:261-279. (1995)Human female orgasm and mate fluc- study of humans as well. 24 Symons D (1995)Beauty is in the adapta- tuating asymmetry. Anim Behav 50:1601- tions of the beholder: The evolutionary psy- 1615. chology of human female sexual 49 Baker RR, Bellis MA (1993)Human sperm REFERENCES attractiveness. In Abramson PR, Pinkerton SD competition: Ejaculate manipulation by fe- (ed), Sexual Nature, Sexual Culture, Chicago: males and a function for the female orgasm. 1 Hartfield E, Sprecher S (1986)Mirror, Mir- University of Chicago Press, pp 80-120. Anim Behav 46387-909. ror: The Importance of Looks in Everyday Life. 25 Lancaster RN (1991) Skin color, race and 50 Langlois JH, Roggman LA, Musselman L Albany, NY State University of New York racism in Nicaragua. Ethnology 30:339-353. (1 994)What is averageand what is not average Press. 26 Thompson LA (1989)Romans and Blacks. about attractive faces? Psychol Sci 5:214-220. 2 Jackson LA (1992)Physical Appearance and Norman, OK University of Oklahoma Press. 51 Swaddle JP, Cuthill IC (1996) Asymmetry Gender: Sociobiological and Sociocultural Per- 27 van den Berghe PL, Frost P (1986) Skin and human facial attractiveness: Symmetry spectives. Albany, New York: State University color preference, sexual dimorphism and sex- may not always be beautiful. Proc SOCLondon of New York Press. ual selection: A case of gene culture co-evolu- B260:lll-116. 3 Buss DM Sex differences in human (1989) tion? Ethnic Racial Stud 9:87-113. 52 Grammer K,Thornhill R (1994)Human mate prefernces: Evolutionary hypotheses 28 Gergen KJ (1975)The significance of skin (Homo sapiens) facial attractiveness and sex- tested in 37 cultures. Behav Brain Sci 12~149. color in human relations. In Franklin JH (ed) ual selection: The role of symmetry and aver- 4 Darwin CR (1967)On the Origin ofSpecies. Color and Race, pp 112-128. Boston: ageness. J Comp Pyschol 108:233-242. New York Atheneum. Houghton Mifflin. 53 Singh D (n.d.) Female health, attractive- 5 Darwin CR (1981)The Descent ofMan artd 29 van den Berghe PL, Frost P (1986)Skin ness and desirability for relationships: Role of Selection in Relation to Sex. Princeton: Prince- color preference, sexual dimorphism and sex- breast asymmetry and waist-to-hip ratio. ton University Press. ual selection: A case of gene culture co-evolu- Ethol Sociobiol 16:465481,in press. 6 Andersson M (1994)Sexual Selection. tion? Ethnic Racial Stud 9:87-113. 54 Weinberg ED (1984) Iron withholding: A Princeton: Princeton University Press. 30 Frost P (1988)Human skin color: Apossi- defense against infection and neoplasia. 7 Cronin H (1991) The Ant and the Peacock: ble relationshk between its sexual dirnor- Physiol Rev 64:65-102. Altruism and Sexual Selection from Darwin to phism and its social perception. Perspect Biol 55 Dixson AF (1983)Observations on the evo- Today. New York: Cambridge University Press. Med 32:38-58. lution and behavioral significance of “sexual 8 Howell N (1979)Demography of the Dobe 31 Ford CS, Beach FB (1951)Patterns ofSex- skin” in female primates. In Rosenblatt JS, !Kung. New York: Academic Press. ual Behavior. New York Harper. Hinde RA, Beer C, Busnel MC (eds) Advances 9 Hill K, Hurtado AM (1996)AcheLife History: 32 Guthrie RD (1970) Evolution of human in the Study ofBehavior, pp 63-106.New York: The Ecology and Demography of a Foraging threat display organs. Evol Biol 1:257-302. Academic Press. People. New York: Aldine de Gruyter. 33 Gowaty PA (1992) Evolutionary biology 56 Jones D (n.d.) Blood and iron: Infection, 10 Henry L (1961)Some data on natural fer- and feminism. Hum Nat 3:217-249. anemia and sexual selection. J Theoret Biol, tility. Eugen Q (now Social Biology) 8:81-91. 34 Wolf NB (1992)The Beauty Myth: How Im- submitted for publication. 11 Goldman N, Montgomery M (1989) ages of Beauty Are Used Against Women. New 57 Galton F (1907)Inquiries into Human Fac- ARTICLES Evolutionary Anthropology 109 ulty and Its Development, p 224. London: J M 65 Pruett-Jones S (1992) Independent versus 73 Degler CN (197 1) Neither Black nor White: Dent. nonindependent mate choice: Do females Slavery and Race Relations in the United States 58 Langlois JH, Roggman LA( 1990)Attractive copy each other? Am Nat 14O:lOOO-1009. and Brazil. Madison, WI: University of Wis- faces are only average. Pyschol Sci I: 115-12 1. 66 Bikchandani S, Hirshleifer D, Welch I consin Press. 59 Light LL, Hollander S. Kayra-Stuart F (1992) A theory of fads, fashion, custom, and 74 Diamond J (1992) The Third Chimpanzee. (1981) Why attractive people are harder to re- cultural change as informational cascades. J New York Harper Collins. Political Econ JOO:992-1026. member. Pers Soc Psychol Bull 7:269-276. 75 Koeslag JH ( 1995) On the engine of specia- 60 Alley TR, Cunningham MR (1991) Aver- 67 Rogers AR ( 1988) Does biology constrain tion. J Theoret Biol 277:410409. aged faces are attractive, but very attractive culture? Am Anthropol90:819-829. faces are not average. Psychol Sci 2:123-125. 68 Boyd R, Richerson PJ (1995) Why does cul- 76 West-Eberhard MJ (1983) Sexual Selec- 61 Koeslag JH, Koeslag PD (1994) Koino- ture increase human adaptability? Ethol So- tion, social competition, and speciation. Q philia. J Theoret Biol J67:5565. ciobiol16:125-143. RevBiol58:155-183. 62 Lerner IM (1954) Genetic Homeostasis. 69 Fisher RA (1958) The Genetical Theory of 77 Wheeler PE (1992)The influenceof the loss London: Oliver and Boyd. Natural Selection. New York; Dover. of functional body hair on the water budgets 63 Deutsch CK (1987) Disproportions in psy- 70 Pomiankowski A (1987) The costs of choice of early hominids. J Hum Evol23:379-388. chiatric syndromes. In Farkas LG, Munro IR in sexual selection. J Theoret Biol 128195- 78 Ruff CB (1991) Climate and body shape in (ed), Anthropornetric Facial Proportions in 218. hominid evolution. J Hum Evol22:81-105. Medicine. Springfield, IL: Thomas, pp 131- 71 Laland KN (1994) Sexual selection with a 79 weiner JS (1954) N~~ shape and climate. 141. culturally transmitted mating preference. J Phys Anthropol 12:14, 64 Langlois JH, Roggman LA, Casey RJ,Ritter Theoret POPUl Bid 45:1-15. JM, Rieser-Danner LA, Jenkins W (1987) In- 72 Wagatsuma H (1968)Thesocial perception JA (1986) in the fant preferences for attractive faces: Rudi- of skin color in Japan. In Franklin JH (ed) .. PPm 207-208. Princeton: Princeton Uni- ments of a stereotype? Dev Psychol Color and Race,-pp 129-165. Boston: verslryrress. 231363-369. Houghton Mifflin. 0 1997 Wiley-Liss, Inc.

June 27-July 1,1997 Dublin, Belfield, Dublin 4, Ireland Society for Human Biology and the American Society of Primatologists fa:353-1-7061 125 IUAES’s Commission on Human Ecology, Son Diego, California e-mail: [email protected] and the Fifth World Academic Contact Nancy Caine (local WWW: http:llwww.ucd.iel Conference on Human Ecology arrangements), Psychology Program, cgi-bin/unitel/select/list/280 Adelaide, South Australia California State University, San Theme: Is Human Evolution a Marcos, CA 92096-0001; Closed Chapter? Contact ASHB/Sth tel: 619-752-4145;fax: 619-752-4111 August 20-27, 1997 WACHE, Department of Anantomy email: [email protected] XXV International Ethological and Histiology, University of For program information contact Conference Adelaide Medical School, Adelaide, Evan Zucker, Department of Vienna, Austria SA 5005, Australia; Psychology, Loyola University, New Contact Dr. Michael Taborsky, tel: 61 8 8303 3369; Orleans, LA 701 18; Konrad Lorenz Institut fur fax: 61 8 83034398 tel: 504-865-3255; Vergleichende Verhalternsforshung, fax: 504-834-4085 A-1 160 Wien, Savoyenstrasse lA, email: [email protected] Austria. Travel awards may be April 22-23,1998 available to those who received or Supmdaminian Modes of Evolution will receive their Ph.D. between 1992 London, UK July 2-4, 1997 and 1997. Award contact: Dr. Judy Contact The Science Promotion ASAB Summer Meeting Stamps, Section of Evolution and Section, The Royal Society, “Biological Aspects of Learning” Ecology, University of California, 6 Carlton House Terrace, London, University of St. Andrews, Scotland UK Davis, CA 956 16; UK SWlY 5AG; For talks and posters contact Dr. email: [email protected] WWW: htp:l/britac3 .britac.ac.uk/rs/ Peter Slater, School of Biological and Medical Sciences, University of St. Andrews, Bute Medical Building, St. November 19-23,1997 July 26-August 1,1998 Andrews KY 16 9TS, Scotland, UK 98th Meeting of the American The 14th International Congress of tel: 44 (0) 1334-463500; Anthropological Association Anthropological and Ethnological fax: 44 (0) 1334-4636; Washington. DC Sciences email: [email protected] Theme: Toward an Anthropology of Williamsburg, VA the 2lst Century. Contact AAA Theme: The 21st Century: The Cen- Meetings Department, 4350 N. tury of Anthropology. Abstract dead- August 17-21,1997 Fairfax Dr., Suite 640, line: October 15, 1997. Contact Ms. Arlington, VA 22203; 3rd World Congress of the International Oriana Casedi, The 14th Congress Lo- Association for Scientific Exchange on tel: 703-528-1902 (ext 2) Human Violence and Human Coexistence email: [email protected] gistics Coordinator, The College of Dublin. Ireland William and Mary, Williamsburg, VA Theme: Violence and the Future of 23187-8795; Society. Contact Jessica Bates, December 1-5, 1997 tel: 757-221-1870;fax: 757-221-1734; Congress Secretary, U College, Annual Conference of the Austmlasian e-mail: [email protected]