Kenrick Keefe:Age preferences in mates studies suggest two things.First,experiments can help us understand the proximate mechanisms underlying 4.0 evolved cognitive heuristics.Second,processing of age, gender,and attractiveness seems to be made at a pre- conscious level. 出 3.0 Experimental research that independently varied ver- bal information about chronological age and apparent youthfulness of potential partners would be useful.Such research could address whether cues associated with 出 2.0 apparent youthfulness in a female may be more important than cognitive information about her age.In line with Symons's(1989)arguments in this regard,and with the 1.0 (Equal sex ratio) above discussion of interactions,we suspect that physical appearance(which provides some indication of reproduc- tive condition)is probably more important than chrono- 0.5 logical age.There are a number of competing predictions that might be made about the particular attraction mecha- nism that works overall,or which mechanism applies in which social situations.Symons's(1989)reasoning sug- gests that a possible sexual attraction mechanism in males 10s 20s 30s 40 might take the form:"Look for a woman who is sexually mature,but has not yet borne offspring."Given the evidence that people quickly process information about DECADE OF AGE AT MARRIAGE age and attractiveness,such a cognitive mechanism Figure 12. Sex ratios in marriage across decades(calculated as seems plausible.This line of reasoning suggests other number of females/males getting married at a particular age empirical questions.Are people's judgments of a wom- range).Note:The Y axis is arranged so that a 2:1 male ratio and a an's attractiveness influenced by whether she has borne 2:1 female ratio are equidistant from the line representing an equal sex ratio. offspring?If so,what are the proximate cues that allow such differentiations?How good are men at distinguish- ing women who are sexually mature from those who are doubt relates to lower remarriage rates among older not,and how are different proximate cues weighted in women. making such decisions?How good are men at distinguish- Table 2 shows that discrepancies in age ratios are not ing women who are postmenopausal,and again,what are limited to our samples but are found in diverse geograph- the proximate cues? ic locations.The table also demonstrates how the unbal- anced marriage ratio for older females is mirrored by an 9.6.Implications of gender differences early marriage ratio unfavorable to younger males. in age preference The data from Table 2 are taken from Table 24 of the United Nations Demographic Yearbook(1989).Examina- We noted earlier that the differential patterns of male and tion of similar statistics compiled there for other nations female preferences demonstrated a failure of matching for reveals that the age-linked gender difference in marriage the two sexes.The graphs depicting male and female ratio appears worldwide.It is interesting to note,how- marriage and preference patterns seem to suggest a ever,that when UN demographers computed marriage related logical inconsistency.How can it be that females' rates for each sex(in Table 25,following the one from marriage ages match their unvarying preferences across the life span,whereas male's marriage ages match their Table 2.Number of men/women marrying at ages below 20. varying age preferences?The answer is that the ratios of and above 50,in diverse geographic regions(men are listed each sex marrying change across the decades.Among first in each coluinn). teenagers who get married,females outnumber males by a considerable margin.The members come closer during the 20s,and then begin to reverse themselves after the Age of marriage 30s.Figure 12 depicts the changing age ratios for the Seattle marriage sample. Region. Under 20 Over 50 49/1227 A similar pattern appeared across all the marriage Mauritius (Africa) (.04) 559/188 (.75) samples.For example,on Poro between 1913 and 1939. Cuba (N.America) 4685/20,957 (.18) 3962/1584 (.71) 384 teenage males married,compared to 928 teenage Singapore (Asia) 457/4,304 (.10) 240/62 (.80) females.On the other hand,there were 66 marriages Malta (Europe) 93/484 (.16) 66/21 (.76) involving men over 40,but only 30 marriages for women Fiji (Oceania) 538/1952 （.22) 123/46 (.73) over 40.Thus,these data help explain the commonly noted discrepancy in remarriage rates in older men and Note:Data are from the year 1975,as reported in the 1987 women;they also suggest that the phenomenon is not Demographic Yearook(United Nations 1989).We did not look simply a function of higher mortality in males.Other at the numbers before choosing a location,but simply included researchers have noted that divorce and remarriage pat- places designed to be as geographically separate as possible, terns in older males fit with a fitness-maximizing strategy with the only limitation being that marriages for 1975 were (Lockard Adams 1981;Mackey 1980).This tendency no reported by the United Nations. 90 BEHAVIORAL AND BRAIN SCIENCES (1992)15:1
Kenrick Keefe:Age preferences in mates which the figures above were calculated)they neverthe- ecological factors,like sex ratios or an individual's posi- less based that rate on the number of marriageable tion in the local status hierarchy?Nyborg and Boeggild opposite sex individuals"in the same age group"(p.133). (1989)present the interesting suggestion that many gen- It would appear that,with the possible exception of der differences in behavior,although rooted in the evolu- people in their 20s,this practice is misguided for most age tionary past,are influenced at the proximal level by groups.The data we have presented indicate clearly that, hormones like testosterone.Although we have focused across sexes,the pool of marriageable partners is not mainly on psychological mechanisms,it might be possible constituted by those of the same age. to derive interesting hypotheses about how hormonal In a related vein,Townsend(1989)found that,as wom- variations across the lifespan are linked to variations in en's status increased,they indicated a preference mate preference.Thus,a consideration of life history for men with progressively higher incomes.Men did strategies raises questions that connect developmental not show a similar tendency;high status men still in- psychology,social psychology,psychobiology,and evolu- cluded women with relatively lower incomes in their pref- tionary biology. erences.Townsend noted that this results in a sex- differentiated narrowing of the pool of eligible partners. ACKNOWLEDGMENT With increasing status,men have an expanded pool, We thank Melanie R.Trost,Suzanne Grover,and Stephen G. whereas women have a more limited pool.There is a par- West for their help in gathering the data presented here.We allel in our findings;increasing age results in a differential also thank John Alcock,Mary Burleson,Peter R.Killeen,Steve shrinking of the pool of possible partners for women. Neuberg,John Reich,Melanie R.Trost,and several anonymous BBS referees for comments on an earlier draft of this manuscript. 10.Conclusion NOTES 1.Correspondence may be addressed to Douglas T.Kenrick, The findings reported here support the heuristic value of Department of Psychology,Arizona State University,Tempe, a consideration of reproductive value and fertility in AZ85287. making predictions about heterosexual attraction pro- 2.We are not aware of any mammalian species in which a cesses;they also reinforce the value of integrating social reversal occurs such that females are more dominant and subject psychological and evolutionary approaches to mating be- to selection by males for that reason.Ralls(1976)reviewed the havior (Buss 1989;Cunningham 1981;Kenrick Trost instances in which female mammals are larger than males,and 1989;Symons 1989).As evolutionary theorists in- concluded that there was no evidence of a reversed sexual creasingly emphasize facultative interactions,varying life selection process at work.For instance,females in these species history strategies,and psychological mechanisms,it did not tend to be especially dominant or to compete for high should become clearer that an evolutionary approach is investing males.However,such a reversal sometimes occurs in not incompatible with the approaches typically taken by other species (e.g.,some fish and insects).Such reversals occur when the male investment in the offspring is larger and they fit psychologists (Crawford Anderson 1989;Symons with the general parental investment theory discussed below. 1989). 3.We thank Prof.Guus van Heck of the University of Tilburg As we have indicated in the above discussion,our for providing us with the Dutch data,and Ute Hoffman and model raises a number of interesting empirical questions: Kirstin Schaefer of the University of Bielefeldt for providing the 1.What are the cultural and ecological factors that German data. might interact with age preferences?The questions about 4.We thank Dr.Nenita Estrera and Nieves Estrera for variations in sex ratios are particularly interesting. gathering the Poro data. 2.What accounts for individual differences in mate 5.One could argue that both groups in this study might have choices?In addition to the questions raised earlier,it been relatively high in resources,but a comparison of either of these patterns with those of any of the other studies reveals the might be of some interest to examine homosexual age same pattern of sex differences.Comparisons across the differ- preferences.Homosexual preferences raise several in- ent studies are discussed further below. terpretative problems,because it is uncertain whether to 6.One reviewer of a previous draft of this paper suggested expect homosexuals to act in line with their own biological that perhaps the norms are more complex,and that young males sex,or with the sex of the targets they are interested in are subject to normative pressures to seekolder females,where- Homosexual choices could shed light on the sex ratio as older males are subject to normative pressures to seek question,however.For example,older homosexual younger females.This explanation is interesting,but it raises a males do not face the same favorable sex ratio as do older couple of questions.First,if such norms exist,why do the heterosexual males,so from a simple sex ratio perspec- normative theorists we cited not seem to be aware of them tive,one might not expect them to show as strong a (failing to even mention such variations)?Norms,after all,are shared rules or social conventions.Second,why do the same preference for relative youth. age-related changes operate across widely different cultures? 3.What are the cognitive mechanisms that underlie Normative explanations are not useful unless one can demon- the gender differences in age preference,and what other strate different cultures or subcultures that show variation in cognitive influences might moderate those preferences? behavior consistent with variations in norms. The findings presented in this paper also stimulate additional questions about differential life histories:Is the age of menopause related to the existence of surviving children and offspring?Is a woman's attractiveness di- minished more by reproductive than by chronological age?How are bodily changes associated with reproduc- tion (puberty,menarche,and menopause)related to BEHAVIORAL AND BRAIN SCIENCES (1992)15:1 91
Commentary/Kenrick Keefe:Age preferences in mates Open Peer Commentary youthful characteristics may act as signals for femininity whereas mature features may signal masculinity.Hence,part of the reason males tend to prefer women with a relatively youthful physical appearance,and vice versa,may be that these age Commentary submitted by the qualified professional readership of this preferences are directly tied to the preference for opposite sex journal will be considered for publication in a later issue as Continuing mates.Just as men's wide shoulders,women's breasts,and Commentary on this article.Integrative overviews and syntheses are other secondary sexual characteristics may function as hetero especially encouraged. sexual attractors or"sexual releasers"(Eibl-Eibesfeldt 1989 Horvath 1979),so too might age-related physical characteristics. In other words,the sex difference in the ages of preferred mates may also reflect our attraction to potential mates with sex- Perceived age,physical attractiveness and typical,or even supranormal,physical traits (Alley 1988). sex differences in preferred mates'ages KK cite the recent study of Langlois and Roggman(1990)as an example of an experimental study elucidating the"proximate Thomas R.Alley mechanisms"that might underlie age preferences.In arguing that attractive faces are only average,however,the model Department of Psychology,Clemson University,Clemson,SC 29634-1511 Electronlc mail:alley@clemson.bitnet advocated by Langlois and Roggman provides no clear role for age-related changes in appearance.In fact,their perspective Kenrick Keefe(KK)present a strong case for considering appears incompatible with K&K's.If aging effects were simply the sex difference in the ages of preferred mates as an evolution- incorporated into the averaging process,the most attractive ary result of divergent reproductive strategies.Their evolution- faces should have characteristics typical of an average age ary framework is admirably broad,leads to interesting questions person,whatever that might be.This mechanism would not that can be addressed by empirical studies,and is well sup- make youthful women more physically attractive,as the evolu- ported by the empirical data presented with it.Their fine-tuned tionary theories of KK,Symons(1979),and others suggest. sociobiological model ofage preferences is sufficiently precise to Langlois and Roggman note,however,that "other elements allow it to be tested (successfully)against alternative social may also be important in influencing judgments of attrac- exchange models,yet it is sufficiently open to environmental tiveness"(1990,p.120).Using a sociobiological perspective as influences to incorporate at least the majority of relevant find- in the target article,Alley and Cunningham(1991)have argued ings in social psychology.Finding little to dispute in their target that adaptive mate selection probably requires preferences for article,this commentary is devoted to a few points that supple- some atypical physical characteristics,including a preference by ment their account. males for youthful female facial characteristics. The Importance of percelved age.KK(sect.9.5)raise the In addition to the two studies mentioned by KK(sect. issue of the psychological mechanism underlying age prefer- 1.2.2),other studies have found evidence that female faces with ence.As they note,within a sociobiological framework per- infantile or youthful characteristics are particularly attractive ceived age,not chronological or even biological age,is likely to (Alley 1988;Berry Zebrowitz-McArthur 1988;Fauss 1988) be the key variable.The reasons for this,however,are not well There is also more evidence than is presented by K&K to elucidated in their target article.The following considerations support the related point that aging has a detrimental effect on might be added to their suggestion that "nonconscious re- the physical attractiveness of women (reviewed in Alley 1988). sponses to proximal cues"are more likely to be involved than are Moreover,it appears that this detrimental effect of aging may consciously calculated strategies"such as inquiring about not apply in late childhood and adolescence when the peak of chronological age. reproductive fitness has not yet been reached (Alley 1988).Ifso, First,there are numerous changes in physical appearance this would confirm another (implicit)prediction of K K's that we can readily detect in principle and that would allow us to model,namely,that females'physical attractiveness to adult judge the maturational status of others.Existing research re- men should not simply decline with age,but should instead veals that we are quite sensitive to some of these changes and change with age in a manner corresponding to their perceived that they do permit us to make judgments of relative age (e.g., age relative to the age of maximum reproductive value. Alley 1983:Todd et al.1980). Second,chronological age is a poor index of biological age: most often,perceived age will provide a better indication of maturational status.Visually,for instance,we can easily detect the presence of secondary sexual characteristics marking the On the separation of reproduction attainment of sexual maturity.Visual information for assessing from mating preferences biological age may be present even when we are not able to use physical appearance to judge chronological age accurately(Alley Betty M.Bayer 1988). Department of Psychology,Wesleyan University,Middletown,CT 06459 Third,people can easily lie about their chronological age Electronic mail:bbayer@eagle.wesleyan.edu whereas their(approximate)biological age will usually be clearly Contrary to the recent move toward a view of human arrange- evident in the morphological and surface qualities of their ments as social and historical constructions(Gergen 1991;Scarr bodies.Although people do attempt to deceive others about 1985),Kenrick Keefe(KK)seek to ground male and female their age by using cosmetics,wigs,and other devices to alter mating preferences in the transhistorical and fixed laws of their physical appearance,it seems likely that their success is reproduction.Through a conceptual twinning of sex and eco- heavily dependent on products and techniques not available in nomics with reproduction and production,and extensive refer- our environment of evolutionary adaptedness. ence to cross-cultural and cross-generational data,KK ex- Fourth,it seems likely that accurate records,including mem- plain male and female mating preferences in terms of a ories,of chronological age would seldom be available in our functional exchange in the sexual market for reproductive gains. environment of evolutionary adaptedness. Their data,however,serve a dual purpose.Not only are they Gender,attractiveness and aging.Human females are more interpreted as supporting a sociobiological structure that deter- paedomorphic (i.e.,retain more infant-like physical charac- mines mating preferences,but they are also used by K&K to teristics as adults)than males (Gould 1977),making a more substantiate their claim that evolutionary theory can par- youthful appearance also more feminine. Consequently, simoniously subsume social psychology's unsystematized find- 92 BEHAVIORAL AND BRAIN SCIENCES(1992)15:1
Commentary/Kenrick Keefe:Age preferences in mates ings on mate preferences.As the tenability of either of these with"an evolved age preference mechanism,"are characterized claims hinges on reproduction,or"reproductive strategies,"it as typically "nonconscious responses"to cues,except under is doubly important to ask about the curious absence of any data particular conditions.K&K speculate that even men who have on reproduction. had vasectomies will show a preference for young,attractive, Stating from the outset that"differential reproductive success fertile females,despite their conscious nonintention to re- is,after all,at the heart of natural selection,"KK place produce.Elsewhere they characterize the exchange process of reproduction at the center of their work on mating preferences. 'biological fitness and reproductive value"as "reflect[ing] The failure to report data concerning reproduction (e.g., evolved adaptations that may operate below the level of con- number,survival,and care of oftspring),although not peculiar sciousness."One interpretation of K K's use of conscious and to K K's research(see Gladue 1989),is a critical omission in unconscious processes is that it is a device that enables theorists the light of their hypotheses,empirical assertions,and the- to position themselves in particular ways vis-a-tis others behav- oretical claims.Without data on reproduction,"reproduction ior to yield knowledge that others do not have(for additional must function in some symbolic way in this empirical test of examples of this in gender research,see Morawski 1985).If mating preferences. these processes of attraction and exchange are unconscious Reproduction functions as"the(marked)board"on which the then what is conscious?And who has it?For example,what is political contest between evolutionary and social theory is reproductive responsibility,and who will assume it?Or is this played(Haraway 1990,p.147).That is,reproduction when seen also calculated outside of consciousness? as a natural imperative turns mating preferences into "natu- Without information about reproduction,K&K's evolution- ral"selections,productive and reproductive ofa"natural"social ary meshwork of mating preferences and reproductive strat- order."Strategy,"in line with its etymological origins in mili- egies falters,as does their claim of its parsimonious explanatory tary cleverness,connotes for KK staying in the reproductive power over social psychology,despite the historical reliance of game.That is,seeking a mate (social exchange)with whom to both social and evolutionary discourses on the social exchange reproduce(evolutionary intent).Absent from K &K's empirical metaphor of a political economy of nature (Gergen 1990).Even account,as either subject or object of mating preferences,is the when the findings of KK are considered as mating preference reproductive behavior of having and raising babies.Whether data alone,they are not without problems.Personal ads selected distinguished as somatic or reproductive effort,or as reproduc- for their information on sex,age,and wealth are representative tive potential or behavior,KK provide no measures of any of of persons for whom this information is salient.Cross-cultural their various definitions of reproduction.If the stakes in this data without reference to the positions of men and women in game are not progeny (i.e.,reproduction),might they have those societies or to the number of times they marry tell us little more to do with the dominance of one explanatory model, about their social arrangements of marriage.A framework of evolution -over another,social? competition and dominance in the social exchange of mating Criticisms of Darwin have illuminated his projection onto preferences ignores alternative accounts that do not assume that nature of Malthusian economics and the workings of Victorian dominance is rooted in nature (e.g.,Haraway 1991),as well as society (Bleier 1984;Hubbard 1983).KK do what other those that go beyond a "minimax strategy of intimate relations proponents of evolutionary thought have done before.They Gergen 1990).What some of K K's findings do suggest, reflect back onto human behavior this early socially constructed however,is that our preferences are often more extreme than view of nature as nature being in and of itself a driving force of a our choices for marriage (see Figure 10),that social change is particular social order(Sayers 1982).They draw a fairly direct indeed a slow process (e.g.,compare Figures 3 and 4;Figure comparison between contemporary social arrangements and 11),and that males stay in the mating game longer than females past nomadic hunter-gatherer ways of living wherein women (see,for example,Figures 7 and 9 in which data for women in provided the bulk of child care and ..substantial calories to their 50s and 60s is absent). the diet"while men "hunted and provided crucial protein. In the end,we might well ask what substantive information on When they depict women as everyone's mother (e.g.,offspring reproduction would add to K K's essay.Is it possible that once of their own marriage,men's existing progeny, grand- teenage pregnancies,single,female-headed families,and the children,or men themselves)and men as predominantly pro- instability of fertility,along with the use of birth control and viders and protectors,KK recapitulate the social and histor- reproductive technologies,(see Sandelowski 1990)are taken ical sexual division of labor as biologically and functionally into account,"natural"reproductive strategies will appear as determined.This is but a partial view,however,a particular merely part of the variance rather than the main effect?Where cultural story of women as nurturers and men as providers(see do technological innovations in reproduction figure in these lanson-Smith 1980). invisible equations of evolutionary psychology?And,finally, According to K&K,women exchange attractiveness and men will these reproductive practices not speak to our intimate wealth and status to consummate this reproductive bargain relations as the social construction of particular historical Women's commodities are located in their bodies and men's in moments? the public world;women's commodities are determined by "nature"and therefore held passively and without volition, men's by"culture"and therefore held actively and with volition; women's reproductive fertility is limited,men's unlimited.In this market,women's soft currency of physical attributes wanes Age preferences:The crucial studies over time while men's hard currency of wealth generally waxes. have yet to be done Even women with wealth and status are shown to be no excep- tion to nature's rule of exchange (see K K's Figure 9) Peter Borkenau Reproduction is used in this scenario to render natural both the Department of Psychology,University of Blelefeld,W-4800 Bielefeld, commodities of exchange in female and male mating and the Germany cultural drama of dominance and competition amongst males for Electronic mall:upsyf270@dbiunill.bitnet the resource of female fertility (see Gross Averill 1983 Kenrick Keefe(KK)report new and highly interesting data Morawski Steele 1991). on the cross-cultural generality of sex-specific age preferences in K&K further equate male attractiveness with dominance, mates,arguing convincingly that their findings reflect sex dif manifest in wealth and status,and female attractiveness with ferences in reproductive strategies.So far,I perfectly agree with youth,representative of fertility.Judgments of attractiveness, them,but for two reasons,I cannot share their view that these although entertained as a psychological mechanism interacting findings suggest an evolutionary explanation of mate selection BEHAVIORAL AND BRAIN SCIENCES(1992)15:1 93
Commentary/Kenrick Keefe:Age preferences in mates criteria more than social-psychological economic exchange similar than age differences of dizygotic twins and their spouses models:(a)Evolutionary and social-psychological models ex- [See also Plomin Paniels "Why Are Children in the Same plain social phenomena at different levels,and (b)reproductive Family So Different From One Another?"BBS 10(1)1987. capacity may itself be a resource that is desired by the opposite Although I share K K's view that evolutionary models ofage sex.Points (a)and (b)are discussed in turn. preferences in mates are reasonable and worthy of being tested, Evolutionary models intend to explain current biological and I believe that the crucial studies have still to be done.The behavioral phenomena in terms of how these phenomena might reported cross-cultural universals in mating patterns may also have increased the organisms'inclusive fitness in the species be explained by economic exchange models that consider the evolutionary past.Among humans,the preference ofolder males universal biological differences in the reproductive capacity of for younger females is easily explained this way,whereas the males and females.And KK also report differences among preference for mates of a similar age is less easily explained.But cultures and generations!Whereas males in their 60s married this does not at all affect the raison d'etre of social-psychological women who were 20 years younger than themselves on the models,which is to provide an account of the psychological island of Poro (Figure 8),males in their 60s who married in processes underlying current mate selection,for instance,the Phoenix in 1923 had wives 15 years younger than themselves feelings of males and females toward younger and older mem- (Figure 4),and males of the same age who married in the same bers of the opposite sex,what males and females do to attract city in 1986 had wives about 10 years younger than themselves members of the opposite sex,and how sexual arousal affects (Figure 3).Thus there are cultural and cohort effects. mating behavior.The importance of both the evolutionary and So what we need are data that distinguish more clearly the psychological perspectives has become very clear in recent between different reasonable models,evolutionary models that years concerning the phenomenon of altruism.Whereas evolu- do not deny cultural factors,and social-psychological models tionary models are useful for explaining why humans help mainly that consider biological factors.What we don't need is a debate relatives(Hamilton 1964)and people they trust (Trivers 1971). about whether evolutionary models or poor social-psychological only psychological approaches are useful to clarify whether models are superior.Obviously,many of my statements can also benefactors intend to improve their own psychological condition be found in the target article.But when they contrast the or the condition of the recipient(Batson 1987).Seealso Caporael evolutionary and social-psychological models,Kenrick Keefe et al.:"Selfishness Examined"BBS 12(4)1989.] seem to be defeating their major purpose,which is to fully Admittedly,the authors of the target article write that evolu integrate the two approaches. tionary biology and social psychology should cooperate in re search on mate selection.But at other places they argue that evolutionary models account for the data better than social- psychological models:"When age preferences from mating advertisements are examined more carefully,they yield results The May-September algorithm meets that were not predicted by social economic models....The the 20th century actuarial table results do fit well with an evolutionary model."In my view,such a debate is counterproductive.Obviously,many social psychol- Gwen J.Broude ogists completely ignore the biological roots of human behavior Department of Psychology,Vassar College,Poughkeepsie,NY 12601 and therefore suggest unreasonable models that should be Electronic mail:broude@vassar.bitnet replaced by more appropriate social-psychological ones.But Are men attracted to younger women,as the social psychologists social-psychological models will never be successfully replaced suggest?Or are male preferences complicated by the additional by evolutionary models,just as biochemical models of the preference for similarity of spouse,as Kenrick Keefe(KK) human immune system will never be successfully replaced by propose in this target article?I wish to demonstrate that the evolutionary models.Science needs many levels of analysis social psychological prediction is consistent with the data,al- My second point is that the target article distinguishes insuffi though K&K are right that the basis for the preference is rooted ciently between evidence for reproductive strategies and evi in selection pressures and not in arbitrary cultural norms dence for evolutionary models.Given that many couples who KK rest their case on findings suggesting that differences in marry consciously intend to have children,the reproductive age between men and preferred or actual mates are not consis- potential of a prospective spouse should be a highly desirable tent across the male lifespan.Does this observation in fact commodity.Moreover,because most women cannot give birth undermine the social psychological prediction or support the to children after their 40s,older males who marry and want argument that men select not only for age of mate but also for children have to look for younger females.This applies to all similarity?Let me focus initially on the finding that young men cultures and generations.Thus K K's findings are easily usually do not prefer or marry women who are younger than explained by economic exchange models that include reproduc themselves and sometimes even marry slightly older women,as tive capacity as a desired commodity.(That most people intend this is one of the trends seen by the authors as causing trouble for to reproduce may nevertheless be explained by evolutionary the social psychologists.The trend is certainly not surprising, models.)Such an exchange model may predict that the prefer- and indeed what else could we expect?One assumes that what ence among older males for younger females is substantially we are witnessing here is a floor effect;the fact that a man can set affected by their subjective intention to reproduce.To my his sights only so low when it comes to yearnings after a young knowledge,this has not yet been investigated. wife.Any lower runs into barriers of cultural norms and,more to Evolutionary models,however,assume that the preference of the point,of biology.In addition,my own cross-cultural data older males for younger females has a genetic basis and evolved (based on codes from Palfrey House,unpublished)suggest that, because genes that augmented this preference spread in the in societies where the age of consent for females is younger than population.This view is very reasonable and implies stronger is the typical marrying age for males,marriages between young predictions than an economic exchange model.Evolutionary men and slightly older women disappear and,indeed,young models predict that males'preference for younger females men tend to marry slightly younger women.Thus,on a world- should be quite unaffected by their subjective intention to wide sample of 26 subsistence economy societies in which men reproduce.Moreover,because there are individual differences commonly marry at 19 years of age or younger,wives are an in the preferred age difference of a spouse,behavior-genetic average of 2.44 years younger than their husbands.This trend is methods,such as twin and adoption studies,should be applied. discrepant with the K &K data for Western samples but more Evolutionary models predict that,within twin pairs,age dif- consistent with their own findings on more traditional cultures. ferences of monozygotic twins and their spouses should be more The cross-cultural data suggest that even the youngest men 94 BEHAVIORAL AND BRAIN SCIENCES (1992)15:1