6 Lee et al. increase female facial masculinity.In fact,the correlation largest sample that has been used to assess how natura in and sibling pairs wa ation in which sts that the ng a all poutral genes influence male and female facial masculinity experimental and correlational research yielding mixed correlation of findings regarding whether male facial masculinity is betwee Table 2).Mascu female faces by obs 001).This gests that the heritable factors underlying male facial Second.we found that the same genetic factors masculinity temale attractiveness increased facial masculinity scores for both males and betw em nbined with the negative assoctal of moes as le facially attractive.Therefore,any genetic benefits to male decrease facial atractiveness in female offspring associated with choosing a facially masculine Accordingly males who had more masculine faces hac partne would b offset by red ice attr. sisters with less attractive faces.A sister shares th rprising their brothers'facial attractiveness (r=-02 =72) decrease the attractiveness of resulting daughters.It is urthermore,male facial masculinity was not associ possible that yet-to-be-established genetic benefits to ated with rated attractiveness (r= 01,p=84). which ons outweigh these genetic detriment daughters ao Ho any suc 00 he attractiveness. facial attractiveness found here.but also perhaps ar Discussion Katina Grammer, 2012) and healt Despite the lary n facial ual linity soores that we estimated to be due to additive that the have genetic influences(49),our other findings do not sup male and female facial shape,and that masculine male port the widely held idea that male acial masculinity is faces have (or d)a selective advant of some kinc ults are difficult with notion tha and rated attractiveness contrary to the sevy sons exnl. from won n's p eference for facially masculine men as a nation of facial sexual dimorphism.This is by far the means of obtaining genetic benefits for offspring.but our results do not preclude this type of explanation.Fo e that mascu althou Table 2.Proportions of Variance of Objective Facial b ain or populations,or by women who are ovulating.Another Influences atemative is that female choice mascu Participant group ty pe se D d tra muscu Female .48L11.61】 031.00.341 .49L.39..62 Moreover the advantage of male facial masculinity Male may stem from enhanced fitness from factors that are Overall unrelated to female choice.For example,faciall Note:The wins was left free nance to male competitors (Puts,2010).In contrast to as the male facial masculinity.female facial femininity (i.e..low masculinity）is both e and 1.0 Moreover,it is not associated with brothers 6 Lee et al. increase female facial masculinity. In fact, the correlation between opposite-sex twin and sibling pairs was of the same magnitude as that between same-sex nonidentical twin and sibling pairs, which suggests that the same genes influence male and female facial masculinity; accordingly, modeling showed a genetic correlation of 1.0, p = .02, between the sexes (see Table 2). Masculine female faces were rated as less attractive than feminine female faces by observers (r = −.17, p < .001). This sug￾gests that the heritable factors underlying male facial masculinity reduce female attractiveness. Accordingly, the correlation between brother masculinity and sister attractiveness was negative, r = −.13, p = .03; that is, sis￾ters of more facially masculine men were rated as less facially attractive. Therefore, any genetic benefits to male offspring associated with choosing a facially masculine partner would be offset by reduced attractiveness of female offspring. In contrast, and unsurprisingly, there was no association between sisters’ facial masculinity and their brothers’ facial attractiveness (r = −.02, p = .72). Furthermore, male facial masculinity was not associ￾ated with rated attractiveness (r = .01, p = .84), which calls into question the sexy-sons hypothesis, according to which male facial masculinity is preferred for heritable attractiveness. Discussion Despite the large proportion of variation in facial mascu￾linity scores that we estimated to be due to additive genetic influences (.49), our other findings do not sup￾port the widely held idea that male facial masculinity is a signal for heritable genetic benefits, for two reasons. First, there was no association between male facial masculinity and rated attractiveness, contrary to the sexy-sons expla￾nation of facial sexual dimorphism. This is by far the largest sample that has been used to assess how natural variation in objective facial masculinity affects individu￾als’ attractiveness, and the finding accords with the over￾all neutral picture that emerges from previous experimental and correlational research yielding mixed findings regarding whether male facial masculinity is attractive, unattractive, or neutral (DeBruine, Jones, Smith, & Little, 2010; Perrett et al., 1998; Rhodes, 2006; Scott et al., 2012). Second, we found that the same genetic factors increased facial masculinity scores for both males and females. Combined with the negative association between female facial masculinity and attractiveness, this suggests that the genetic factors increasing male facial masculinity decrease facial attractiveness in female relatives. Accordingly, males who had more masculine faces had sisters with less attractive faces. A sister shares the same proportion (.50) of segregating genes as a daughter, so choosing a facially masculine man as a mate will tend to decrease the attractiveness of resulting daughters. It is possible that yet-to-be-established genetic benefits to sons outweigh these genetic detriments to daughters. However, any such genetic benefits would need to out￾weigh not only the detriment of masculinity to female facial attractiveness found here, but also perhaps appar￾ent detriments to female fertility (Pfluger, Oberzaucher, Katina, Holzleitner, & Grammer, 2012) and health (Thornhill & Gangestad, 2006). The existence of facial sexual dimorphism suggests that there have been different selection pressures on male and female facial shape, and that masculine male faces have (or had) a selective advantage of some kind. Our results are difficult to reconcile with the notion that the selective advantage of masculine male faces comes from women’s preference for facially masculine men as a means of obtaining genetic benefits for offspring, but our results do not preclude this type of explanation. For example, it is possible that masculine faces, although not judged as being more attractive by raters overall, are judged as more attractive by women in certain contexts or populations, or by women who are ovulating. Another alternative is that female choice acts not on facial mascu￾linity per se but on correlated traits such as body muscu￾larity or assertive behavioral tendencies. Moreover, the advantages of male facial masculinity may stem from enhanced fitness from factors that are unrelated to female choice. For example, facially mascu￾line men might gain a survival or reproductive advantage through intrasexual competition by being more robust to physical damage or by signaling formidability and domi￾nance to male competitors (Puts, 2010). In contrast to male facial masculinity, female facial femininity (i.e., low masculinity) is both heritable and associated with attrac￾tiveness. Moreover, it is not associated with brothers’ Table 2. Proportions of Variance of Objective Facial Masculinity Estimated to Be Accounted for by A (Additive Genetic), C (Shared Environmental), and E (Residual) Influences Participant group A C E Female .48 [.11, .61] .03 [.00, .34] .49 [.39, .62] Male .46 [.20, .59] .00 [.00, .17] .54 [.41, .71] Overall .49 [.28, .57] .00 [.00, .17] .51 [.43, .61] Note: The numbers in square brackets are 95% confidence intervals. Opposite-sex twins contributed to means and variances but not to variance components (i.e., the genetic correlation between opposite￾sex twins was left free to vary in the model). The genetic correlation between opposite-sex twins was estimated in the model at .50, the same as the correlation for same-sex nonidentical twins, which implies no sex limitation in facial masculinity (i.e., a perfect genetic correlation, r = 1.0, between male and female facial masculinity). Downloaded from pss.sagepub.com by Cai Xing on January 7, 2014