Abstract
The current research tests the hypothesis that women have an evolved mate value calibration adaptation that functions to raise or lower their standards in a long-term mate according to their own mate value. A woman's physical attractiveness is a cardinal component of women's mate value. We correlated observer-assessed physical attractiveness (face, body, and overall) with expressed preferences for four clusters of mate characteristics (N = 214): (1) hypothesized good-gene indicators (e.g., masculinity, sexiness); (2) hypothesized good investment indicators (e.g., potential income); (3) good parenting indicators (e.g., desire for home and children), and (4) good partner indicators (e.g., being a loving partner). Results supported the hypothesis that high mate value women, as indexed by observer-judged physical attractiveness, expressed elevated standards for all four clusters of mate characteristics. Discussion focuses on potential design features of the hypothesized mate-value calibration adaptation, and suggests an important modification of the trade-off model of women's mating. A minority of women—notably those low in mate value who are able to escape male mate guarding and the manifold costs of an exposed infidelity—will pursue a mixed mating strategy, obtaining investment from one man and good genes from an extra-pair copulation partner (as the trade-off model predicts). Since the vast majority of women secure genes and direct benefits from the same man, however, most women will attempt to secure the best combination of all desired qualities from the same man.
Introduction
Humans possess a menu of mating strategies that includes long-term mating, short-term opportunistic copulations, extra-pair copulations, and serial mating (Buss, 1994/2003; 2007; Buss and Schmitt, 1993; Gangestad and Simpson, 2000; Greiling and Buss, 2000; Symons, 1979). Much empirical work has documented how mate preferences shift according to context. Women pursuing short-term mating compared to long-term mating, for example, increase the importance they place on a man's physical attractiveness, sex appeal, muscularity, and extravagant and immediate resource displays (Buss and Schmitt, 1993; Frederick and Haselton, 2007; Gangestad, Garver-Apgar, and Simpson, 2007; Haselton and Gangestad, 2006; Haselton and Miller, 2006; Pawlowski and Jasienska, 2005). Women pursuing long-term mating, in contrast, place greater importance on resource acquisition potential, such as “has a promising career” and “has good financial prospects” (Buss and Schmitt, 1993).
Mate preferences have been shown to shift as a function of personal and ecological contexts. In ecologies with a high prevalence of parasites, for example, both sexes increase the importance they place on physical attractiveness, presumed to be a powerful health cue (Gangestad and Buss, 1993). Mate preferences have also been shown to vary according to operational sex ratio (Stone, Shackelford, and Buss, 2007), cultural norms surrounding premarital sex (Buss, 1989), menstrual cycle (Gangestad et al., 2007; Pillsworth, Haselton, and Buss, 2004; Puts, 2005), age (Kenrick and Keefe, 1992), and mating system (e.g., legally polygynous versus presumptively monogamous) (Buss, 1989). One personal circumstance that has not explored in depth, however, is the role of mate value in affecting standards imposed on potential mates.
Gangestad and Simpson (2000) hypothesize that women make trade-offs in mating along two dimensions—good gene fitness indicators and good investment indicators. Good gene indicators are hypothesized to include masculinity, physical attractiveness, muscularity, symmetry, intelligence, and “confrontativeness” (Gangestad, Garver-Apgar, and Simpson, 2007). Good investment indicators are hypothesized to include resources and resource acquisition potential (Buss and Schmitt, 1993). Women might also choose “good dad” indicators such as warmth and kindness (La Cerra, 1994). Because most women cannot “get it all,” they are hypothesized to favor investment and perhaps “good dad” indicators in long-term mating, while choosing short-term affair partners who possess indicators of good genes (Gangestad and Simpson, 2000; Gangestad et al., 2007). As a caveat, it must be noted that nearly all stable personal characteristics show moderate heritability (Larsen and Buss, 2008), and hence qualities associated with being a “good dad” or “good provider” may also provide “good genes.”
Ideally, however, a woman would want all positive mate characteristics thus far identified in the scientific literature, and perhaps even more. We propose that women should value at least four clusters of characteristics in a long-term mate: (1) good genes indicators (Buss and Schmitt, 1993), (2) good resource acquisition indicators (Buss, 1989; Symons, 1979), (3) good parenting indicators (Buss, 1991), and (4) good partner indicators. These clusters may or may not covary—an empirical issue yet to be determined. A man good at acquiring resources that can be channeled to a woman and her children, for example, may or may not be a good dad or a good partner. On the other hand, it is possible that these clusters covary. Gangestad et al. (2007) reasonably argue that for most women there must be a tradeoff between good genes and good investment indicators, but it is possible that some characteristics simultaneously signal good genes, good investment, and other sought-after indicators. A prime example is the trait of “intelligence,” which is known to be highly heritable (De Fries et al., 2000) as well as being positively correlated with health, longevity, income, and socio-economic status (Jencks, 1979; Gottfredson and Deary, 2004). Some traits, in short, may simultaneously signal good genes, good health, good investment ability, and possibly good parenting proclivities.
Men who have all four clusters of desirable characteristics, of course, are rare, highly sought after, and hence difficult for most women to attract and retain. Gangestad and his colleagues (2000, 2007) argue that men with good-genes indicators are likely to pursue a short-term rather than a long-term mating strategy, and hence be reluctant to commit to one woman. Consequently, according to this argument, women have evolved adaptations for choosing a high-investing man as a long-term partner, while securing good genes through extra-pair copulations. In short, Gangestad and colleagues propose that women have evolved a mixed mating strategy, with the proper evolved function of female extra-pair copulations (EPCs) being securing access to superior genes. Versions of these arguments, albeit without using the concept of “tradeoffs,” also have been previously advanced by Symons (1979, p. 207), Smith (1984), and Buss (1994, pp. 90–91).
Hypotheses about human adaptations must include, implicitly or explicitly, the assumption that the relevant ancestral conditions for such adaptations to evolve were statistically recurrent over deep time. Based on all available ethnographic data, the most reproductively successful men historically and cross-culturally were those who married young, secured multiple wives, and opportunistically engaged in sex with other men's wives when the risks were low (Symons, 1979; personal communication, October 17, 2007). Men who were successful at having extra-pair copulations were precisely the same men who were most successful at acquiring wives—those who were high in status, good hunters, successful warriors, headmen, and successful intrasexual competitors. In the human ancestral environment, in short, it is highly unlikely that there existed a pool of desirable males who remained single and pursued an exclusively short-term mating strategy of “love ‘em and leave ‘em” (Symons, 1979; personal communication, October 17, 2007). These conditions contrast starkly with those of the modern environment, where on large college campuses and in large cities such pools of unmated men exist. Thus, if women have evolved a mixed mating strategy, adaptations to secure investment from a husband and good genes from an EPC partner, the EPC partner historically was likely to have been a successful polygynously mated man.
As a general rule, humans assortatively mate on overall mate value (Buss, 1994/2003; Buss and Barnes, 1986). The “6s” mate with other “6s,” while the “9s” can attract other “9s.” Theoretically, some women—those highest in mate value—should be able to attract and retain men who have both good gene indicators and good investment indicators.
Physical attractiveness is a cardinal component of women's mate value (Buss, 1989; Buss and Schmitt, 1993; Symons, 1979, 1995). Given limited mating budgets, men view physical attractiveness as a “necessity” rather than a “luxury” in mating (Li, Bailey, Kenrick, and Linsemeier, 2002). Consequently, a woman's mate value, as indicated by her physical attractiveness, will influence the quality of the man she is able to attract as a long-term mate. This leads to the central hypothesis of this article: Women high in mate value, compared to those lower in mate value, should impose higher standards for a suite of desirable characteristics, including hypothesized good-gene indicators, good investment indicators, good parenting indicators, and good partner indicators. The hypothesis, in short, proposes an adaptation in women that adjusts their standards up or down according to their own mate value. This hypothesis has existed in various forms for many years (e.g., Buss, 1994; Buss and Barnes, 1986; Buss and Schmitt, 1993; Little, Burt, Penton-Voak, and Perrett, 2001; Thornhill and Gangestad, in press). Buss and Schmitt (1993), for example, argued that “… expressed mate preferences may be calibrated up or down depending on one's mate value” (p. 230). Little et al. (2001) proposed and found that women who rated themselves high on physical attractiveness will express stronger preferences for men who display two cues hypothesized to indicate heritable fitness—masculinity and symmetry. Nonetheless, the broader hypothesis that high mate value women want it all has never been comprehensively tested.
Several studies find circumstantial support for this hypothesis. Two studies found that women who were younger and higher on self-rated attractiveness specified a longer list of traits that they sought or required in a potential mate than did women lower in these key markers of mate value (Pawlowski and Dunbar, 1999; Waynforth and Dunbar, 1995). Similar results have been found in Japan (Oda, 2001) and Brazil (Campos, Otta, and Siqueira, 2002). Another study found that women higher on self-rated physical attractiveness expressed a greater attraction for more masculine faces than did women lower on self-rated attractiveness (Little, Penton-Voak, Burt, and Perrett, 2001). Women with a lower waist-to-hip ratio (WHR)—a well-documented indicator of female fertility (Singh, 1993)—express a stronger attraction to more masculinized male faces than do women with a higher WHR (Penton-Voak, et al., 2003). Another study found that women with a relatively low WHR had stronger preferences for mates with resources, although contrary to the hypothesis, women with a higher WHR expressed stronger preferences for male attractiveness (Pawlowski and Jasienska, 2008). And attractive and feminine women show stronger preferences for masculinized male voices than do less attractive and less feminine women (Little, et al., 2001; Feinberg, et al., 2006). The studies on masculinity are based on the premise that the trait of masculinity is a good-genes health indicator. The rationale is that testosterone, which produces masculine features, compromises the immune system. Consequently, during adolescence when facial features and voice take their adult form, only those males who are extremely healthy can “afford” to produce high levels of testosterone.
Whereas the previous studies have focused on a delimited number of traits, notably masculine appearance and vocal quality, the current study sought to examine a large pool of potential traits. We sought to examine four clusters of expressed preference traits: (1) traits that have been hypothesized in the scientific literature to be indicators of good genes: masculinity, physical attractiveness, sex appeal, physical fitness, and intelligence (Gangestad et al., 2007; Miller, 2000); good investment ability indicators: potential income, good earning capacity, education, ambition and industriousness, favorable social status, and somewhat older age (all of these are known to be directly or indirectly linked with resource acquisition—see Buss, 1994/2003); (3) traits hypothesized by the current authors to be good parenting indicators: desire for home and children, fondness of children, emotional stability and maturity, and kind and understanding (the latter pair—kind and understanding—has also been hypothesized to be indicators of good investment proclivities—see Gangestad et al., 2007); and (4) traits hypothesized to be good partner indicators: being a loving partner, devoted to you, and loyal (perhaps best conceptualized as signs of emotional commitment). Emotional stability and maturity, hypothesized to be a good parenting indicator, could also serve as a good partner indicator. Whereas most previous studies have relied on self-rated attractiveness, the current study used a composite of impartial male and female interviewers who interviewed the participants for 40 minutes, and subsequently provided ratings facial attractiveness, bodily attractiveness, and overall attractiveness. Thus, judgments of attractiveness and expressed mate preferences came from entirely separate and independent data sources.
Materials and Methods
Participants
Participants were 214 individuals, 107 men and 107 women, who had been married less than one year at the time of testing. Participants were obtained from the public record of marriage licenses issued within a large county in the Midwest United States. All couples who had been married within the designated time period were contacted by letter and invited to participate in this study. The mean age of the male sample was 25.46 years (SD = 6.55 years). The mean age of the female sample was 24.78 years (SD = 6.24 years). Couples were given $25 per couple in return for their participation.
Procedure
Participants engaged in three separate episodes of assessment as part of a large set of studies. First, they received through the mail a battery of instruments to be completed at home in their spare time. Second, participants came to a laboratory testing session approximately one week after receiving the first battery. During this testing session, spouses were separated to preserve independence and to prevent contamination from contact or discussion. It was during this second testing session that participants completed four instruments relevant to the present report: Factors in Choosing a Mate, Preferences Concerning Potential Mates, Family and Marital Preference Questionnaire, and Goals Wanted in a Partner. Third, couples were interviewed toward the end of the second testing session using a standard set of interview questions to provide information about the couple's relationship and to give the interviewers an opportunity to observe participants so that they could provide independent assessments of each participant's physical attractiveness (face, body, and overall attractiveness). Confidentiality of all responses was assured. Not even the participant's spouse could obtain responses of their partner without written permission from his or her partner.
Mate Preferences Instruments
The “Factors in Choosing a Mate” originally was developed by sociologists in the 1930s (see Hill, 1945). This instrument seeks biographical information, age preferences in a partner, and has participants rate 18 mate characteristics on a scale that ranges from “0” (irrelevant or unimportant) to “3” (indispensable). The “Preferences Concerning Potential Mates Questionnaire,” based on a factor analysis of a much larger instrument (see Buss and Barnes, 1986), contains 13 mate characteristics that participants rank from “1” (most important) to “13” (least important). The “Family and Marital Preference Questionnaire” (FMPQ) was originally developed by Gough (1973), and expanded by Buss and Barnes (1986). It contains 83 mate characteristics, which participants range on a five-point scale ranging from “+2” (very desirable) through “0” (inconsequential) to “−2” (very undesirable). Only one study to our knowledge has examined the test-retest reliability of mate preferences (Shackelford, Schmitt, and Buss, 2005). That study, using the Factors in Choosing a Mate instrument, found that most mate preferences showed moderate temporal stability over a three-year time span, averaging +.51 for men and +.54 for women, which is considered impressive for single-item measures (Nunnally, 1978).
Judgments of Physical Attractiveness
Participants were interviewed by one male and one female interviewer, drawn from a rotating team of eight interviewers. Immediately following the interview, each of the two interviewers independently evaluated each participant on three seven-point scales: unattractive face 1 2 3 4 5 6 7 attractive face; unattractive body 1 2 3 4 5 6 7 attractive body; overall unattractive 1 2 3 4 5 6 7 overall attractive. The interviewer's ratings were highly correlated, averaging +.60 for the ratings of the male participants and +.66 for the female participants (by way of contrast, the average correlations between any two judges on ratings of physical attractiveness is typically around +.50—see Berscheid and Walster, 1974). Consequently, the ratings made by the male and female interviewers were composited with unit weighting to provide more reliable indices of facial, bodily, and overall attractiveness.
Results
Table 1 shows the correlations between the three measures of physical attractiveness derived from the interviewers and the expressed mate preferences of the female participants. It is organized into the four clusters of a priori chosen hypothesized indicators: Traits hypothesized to be good genes indicators in the published literature such as masculinity, physical attractiveness, and intelligence; traits known to be good investment ability indicators, such as financial resources; traits hypothesized in the literature to be good parenting indicators (e.g., kind and understanding) as well as those of high and obvious face-validity (e.g., “raising children well” as a high goal priority); and good partner indicators of high and obvious face-validity (e.g., “devoted to you”).
Correlations Between Women's Attractiveness and Mate Characteristics Desired
p < .05, two-tailed
p < .01, two-tailed. This table includes only hypothesized indicators specified in the text; complete statistical analyses may be obtained from the authors.
Hypothesized good genes indicators
Physically attractive women expressed significantly stronger preferences for five of the six hypothesized good genes indicators—mates who are more masculine, physically attractive, good looking, sex appeal, and physically fit. The sole exception was intelligence, which was not valued more by physically attractive than less attractive women. Men's overall physical attractiveness, in contrast, was significantly correlated with only one hypothesized good genes indicator—physically fit (r = +.24, p < .05).
Good investment ability indicators
Women's overall physical attractiveness was positively correlated with all six of the good investment abilities, in four cases significantly so—income potential, good earning capacity, college graduate, and older then self. In contrast, men's overall physical attractiveness was correlated significantly only with a preference for a mate who is a college graduate (r = +.21, p < .05).
Hypothesized good parenting indicators
Women's overall physical attractiveness was positively correlate with all six of the good parenting indicators, in five cases significantly so—desire for home and children, fond of children, likes children, and has raising children well as a goal priority. The correlations for kind and understanding were positive, but not significantly so. Men's overall physical attractiveness, in contrast, correlated positively and significantly with only one good parenting indicator—raising children well as a goal priority (r = +.19, p < .05).
Hypothesized good partner indicators
Women's physical attractiveness was positively correlated with higher standards expressed for all three good partner indicators. Only “being a loving partner” as a goal priority of the potential partner proved to be statistically significant. In contrast, men's physical attractiveness was not positively and significantly correlated with any of the good partner indicators (e.g., where women's overall physical attractiveness correlated with desire for the characteristic “being a loving partner” with an r = .24, p < .05, men's attractiveness correlated with this desire with an r = .05, ns).
Discussion
The results point to a singular conclusion: Attractive women express higher standards for nearly all hypothesized indicators of good genes, good investment abilities, good parenting abilities, and good partner traits. Physical attractiveness in men, with very few exceptions, is not linked with higher standards for these four clusters of long-term mate characteristics. These results support the central hypothesis of this article—that women high in the cardinal indicator of female mate value raise their standards for the traits that are hypothesized to have been historically most critical to female reproductive success. These results support the existence of the hypothesized mate value calibration adaptation.
The current results converge well with nearly all previous studies about what attractive women want. Whereas Campos, Otta, and Siqueira (2002), Pawlowski and Dunbar (1999), Oda (2001), and Waynforth and Dunbar (1995) found that self-rated attractive women listed a larger number of sought-after traits, the current study found that observer-judged attractiveness in women is linked with elevated preferences across all four key clusters of traits. Whereas previous studies have found that female attractiveness in women was linked with stronger preferences for hypothesized good-genes indicators of masculine faces and voices (Little et al., 2001; Feinberg et al., 2006; Penton-Voak, 2003), the current study found that observer-judged physical attractiveness in women was linked with a stronger preference for masculinity, as well as other hypothesized good-genes indicators such as physical attractiveness, sex appeal, and physical fitness. There exists one discrepancy. Whereas the current study found that attractive women expressed stronger preferences for all four clusters of characteristics, including resources and physical attractiveness, Pawlowski and Jasienska (2008) found that attractive women (as gauged by low WHR) expressed stronger preferences only for resources; however, contrary to the current study and contrary to their own hypothesis, they found that less attractive women (those with a higher WHR) expressed a stronger preference for male attractiveness. Only future studies can resolve these apparently discrepant findings.
One other puzzle remains, centering on the trait of intelligence. Intelligence has been hypothesized to be one of the cardinal indicators of good genes (Gangestad et al., 2007; Miller, 2000). The current study, however, did not find that attractive women express a stronger preference than less attractive women for intelligence in a mate. Furthermore, Gangestad et al. (2007) failed to find the hypothesized female shift in valuing intelligence more around ovulation. Although Gangestad and his colleagues state that readers should not reject intelligence as a good-genes indicator based on their single study, the current study may raise an additional doubt or add to the puzzle. If intelligence is indeed a powerful indicator of good genes, as it should be on theoretical and empirical grounds, then attractive women should value it more according to the hypothesis articulated in the current article (which they do not) and women should value it more around ovulation (especially in short-term mates) according to the trade-off model (which they do not).
Future research could profitably be directed to resolving puzzle of why women's preferences for intelligence do not shift in the predicted directions according to a woman's ovulation status or mate value.
Before discussing the potential implications of these results, several limitations of the study should be noted. First, the sample consisted of already-married individuals. The results should be replicated with different populations, particularly unmated individuals, to establish their generalizability. Second, the sample contained participants from a single culture; cross-cultural replication is needed. Third, it remains to be established that attractive women's stronger mate preferences for these key qualities in fact result in successfully securing mates possessing these qualities. There exists some evidence on this issue—women's physical attractiveness is the best known predictor of the occupational status of the man she marries and the best known predictor of hypergamy, or marrying up in socio-economic status (Elder, 1969; Taylor and Glenn, 1976; Udry and Eckland, 1984). To our knowledge, however, there exist no empirical studies that establish behavioral evidence that women high in mate value actually marry men who have hypothesized good-genes indicators such as masculinity and sex appeal, hypothesized good parenting abilities, or hypothesized good partner qualities. Fourth, it must be acknowledged that women's physical attractiveness is only one component, albeit an exceptionally important one, of women's overall mate value (Buss, 1994/2003). Fifth, future research might also add potential “protection indicators” to the suite of qualities that high mate value women potentially desire.
With these limitations in mind, what are the potential implications of these findings? The most straightforward implication is that not all women must “trade off” when selecting a long-term mate. Women high in mate value need not sacrifice good genes in order to secure good investment ability or other indicators of direct benefits. Women high in mate value, as indicated by the fundamental trait of physical attractiveness, raise their standards for all four clusters of indicators. Conversely, those women lower in mate value relax their standards for all of the key clusters of mate traits. They relax the strength of their standards not just for hypothesized good-genes indicators, but also for indicators of investment ability, parenting, and partnering. Whereas high mate value women want it all, women lower in mate value adaptively reduce their standards. These individual differences support the hypothesis that women have an evolved self-assessment mechanism that calibrates their standards to their mate value (see Buss and Schmitt, 1993, p. 230). This calibration adaptation presumably allows women to target their mating efforts toward men who are in the mate-value range that they can reasonably expect to attract and retain, and to selectively accept mating overtures from men within the mate value range they can attract and retain.
These findings raise a potential challenge to a strict version of the trade-off model proposed by Gangestad and Simpson (2000). In a world without constraints, a woman would certainly obtain the best genes, the best investment, the best parent for her children, and the best partner. And as Gangestad and colleagues correctly argue, most women cannot get all these qualities from the same man (although the degree to which these qualities are intercorrelated, and were intercorrelated in ancestral environments, remains an open question). It does not necessarily follow, though, that the trade-off that women do make will be obtaining the best possible investment from one man, while cuckolding him with a man with superior genes. Based on the extant studies of genetic cuckoldry, only a small minority of women actually sire children with men other than their regular mates—perhaps 10 - 12% (see Baker and Bellis, 1995; Cerda-Flores, Barton, Marty-Gonzalez, Rivas, and Chakraborty, 1999). Assuming that these figures approximate those of ancestral environments, this means that the majority of women, perhaps 88 - 90%, in fact obtain genes, investment, parenting, and partnership from the same man.
Consider a hypothetical numerical example in which the best genetic benefits equals a value of 5 and the best direct benefits (including investment, parenting, and partnership) equals a value of 5. A woman with a mate value of 10 is able to command it all—a perfect 5 on both clusters. Now consider women lower in mate value, say an 8. She could go for a man who maximally provides direct benefits (5), but who has only a genetic quality of 3, while cuckolding him with a man who is a 5 on genetic quality (as the tradeoff model predicts). Alternatively, she could lower her standards a bit on both, and obtain a man who is a 4 on genetic quality and a 4 on direct benefits. The same logic applies to women of lower mate value, although it is obvious that women of lower mate value, in principle, have more to gain by genetic cuckoldry than do women of higher mate value.
As Thornhill and Gangestad (in press) acknowledge, men's mate guarding adaptations, designed specifically to prevent cuckoldry, can function to prevent a woman from securing genes from a genetically superior man. And women who risk an EPC also risk losing their partner's resources, risk spousal violence, risk damage to their social reputations, and risk abandonment as a consequence of infidelity (Buss, 2000; Daly and Wilson, 1988). Combined with the finding that the vast majority of women do obtain both genes and direct benefits from the same man, and only a minority of women obtain genes from one man and direct benefits from another, these facts suggest a modification of the model proposed by Gangestad and Thornhill—one that receives circumstantial support from the current empirical study.
The implications of this modification are threefold: (1) women high in mate value can get it all from the same man, and in fact raise their standards for nearly all good genes indicators, good investment indicators, good parenting indicators, and good partner indicators; (2) most women cannot get it all, and so lower their standards across all four sets of indicators, since the odds are high that they will obtain both genetic and direct benefits from the same man; these women secure the best combination of characteristics they can obtain from the same man; (3) a minority of women, perhaps those in the low mate value range and who can manage successfully to escape the multiple costs inflicted by heavy mate guarding, a violently jealous mate, and the reputation damage caused by an infidelity exposed, pursue a dual mating strategy in which they successfully obtain direct benefits from one man while committing EPCs to obtain better genes from another man. Further research is needed to identify the qualities that distinguish which women pursue the mixed mating strategy of investment/genes tradeoff as original version of the trade-off model predicts, which women will make other sorts of trade-offs (e.g., sacrificing good parenting indicators for good resource indicators), and which women try to get the best combination of all four clusters of key qualities from the same man.
Footnotes
Acknowledgements
We would like to thank members of the BussLab and David Buss's Evolutionary Psychology Seminar for contributing to the ideas contained in this paper—Ashley Brock, Jaime Confer, Judith Easton, Diana Fleishman, Cari Goetz, Ewa Kacewicz, David Lewis, Carin Perilloux, and Yla Tausczik. Special thanks to Don Symons for specific feedback and wise reflections on these ideas.