Maestripieri et al. appear to be unaware of important advances in understanding the role of attractiveness in mate preferences and relationship outcomes. In the context of relationships, evidence does not support the claim that a partner's physical attractiveness is more valued by men than women. Specifically, the meta-analysis of 95 articles by Eastwick et al. (Reference Eastwick, Luchies, Finkel and Hunt2014a) showed that physical attractiveness predicted romantic attraction very similarly for men (r=0.43) and women (r=0.40); see also the critique by Meltzer et al. (Reference Meltzer, McNulty, Jackson and Karney2014) and the rejoinder by Eastwick et al. (Reference Eastwick, Neff, Finkel, Luchies and Hunt2014b). For example, Selterman et al. (Reference Selterman, Chagnon and Mackinnon2015) found that participants in a speed-dating study were more interested in potential partners who were attractive and good potential earners, with neither preference moderated by gender. However, the greater valuing of attractiveness by men than women is robust in assessments of preferences for ideal hypothetical partners (e.g., Feingold Reference Feingold1992b), a context far less relevant to fitness outcomes than actual relationships.

Also weak is evidence that attractiveness is an honest signal of health and reproductive success. Reviews have found attractiveness and related facial characteristics weakly or inconsistently linked to actual health (Henderson et al. Reference Henderson, Holzleitner, Talamas and Perrett2016; Weeden & Sabini Reference Weeden and Sabini2005). In addition, Weeden and Sabini (Reference Weeden and Sabini2005) found women's health was related not to facial attractiveness, but to waist-to-hip ratio and weight. Yet, most researchers cited by Maestripieri et al. operationalized attractiveness by head or head-and-shoulders photos varying in facial attractiveness.

Even more critical to the sexual motives argument of Maestripieri et al. would be evidence that attractiveness serves as an honest signal of fertility. This prediction is best examined in populations with low contraceptive use. A study in rural Senegal, where there is little access to modern birth control, found that facial attractiveness negatively predicted age-specific reproduction in both sexes (Silva et al. Reference Silva, Lummaa, Muller, Raymond and Alvergne2012). These investigators also found a negative relation of facial attractiveness to lifetime reproductive success among men who graduated from the West Point Military Academy in 1950. Earlier studies relating attractiveness to reproductive outcomes have produced inconsistent findings (see Silva et al. Reference Silva, Lummaa, Muller, Raymond and Alvergne2012). For example, in a large U.S. sample, attractiveness showed only a weak positive relation to reproductive success in women and men, once marriage was taken into account (Jokela Reference Jokela2009). In general, evidence that fitness gains would follow from attractiveness preferences is weak and inconsistent.

The authors also neglected the relation between attractiveness and trustworthiness, thereby underestimating trustworthiness's impact. Considerable evidence shows that attractive people are trusted more than unattractive people (e.g., Kaisler & Leder Reference Kaisler and Leder2016; Palmer & Peterson Reference Palmer and Peterson2016). However, trustworthiness is not merely one of several positive traits associated with perceived attractiveness, but is a central trait revealed directly from physical appearance. In addition to attractiveness, faces are informative about two other evolutionarily adaptive attributes – namely, trustworthiness and dominance (Oosterhof & Todorov Reference Oosterhof and Todorov2008). Research has thus shown that three core dimensions describe the information directly available from faces: trustworthiness, dominance, and youthful attractiveness (Sutherland et al. Reference Sutherland, Oldmeadow, Santos, Towler, Michael and Young2013).

In some studies, trustworthiness has actually trumped attractiveness in affecting positive outcomes. For example, in a peer-to-peer lending study, Duarte et al. (Reference Duarte, Siegel and Young2012) found little effect of borrowers' perceived attractiveness but a benefit of their perceived trustworthiness. A similar result emerged in an economic game study where perceived trustworthiness, but not attractiveness, predicted the amount of money sent to the partner (van't Wout & Sanfey Reference van't Wout and Sanfey2008). In other contexts, trustworthiness and attractiveness sometimes show high levels of shared variance (Ohanian Reference Ohanian1991; Sofer et al. Reference Sofer, Dotsch, Wigboldus and Todorov2015). Indeed, a meta-analysis found that attractiveness and trustworthiness are subserved by overlapping brain networks (Bzdok et al. Reference Bzdok, Langner, Caspers, Kurth, Habel, Zilles, Laird and Eickhoff2011). Finally, an investigation of the relation between facial trustworthiness and attractiveness in children and adults concluded that facial attractiveness might have heuristic value in signaling trustworthiness and thus guiding decision making (Ma et al. Reference Ma, Xu and Luo2015). In sum, by focusing on one aspect of face perception, namely, attractiveness, and contending that trustworthiness is merely stereotypically associated with attractiveness, the authors have ignored the feature that arguably has more adaptive significance and predictive validity.

Concerning cultural variation, preferences for attractive partners may be largely a modern phenomenon, not a pattern typical of early humans. Consider that most human evolution took place in simple band societies, characterized by small, dispersed living groups of cooperating families (Gintis et al. Reference Gintis, van Schaik and Boehm2015). Given that cultures have rules that restrict marriage options, people in band societies typically would have obtained mates from few possibilities. Moreover, people would not have been exposed to many exemplars of opposite-sex persons differing in the physical attributes that are correlated with attractiveness in WEIRD (Western, educated, industrialized, rich, democratic) cultures (Henrich et al. Reference Henrich, Heine and Norenzayan2010). With industrialization and urbanization, massive direct and indirect exposure to large numbers of people favors learning the nuances of attractiveness. From this perspective, research by Scott et al. (Reference Scott, Clark, Josephson, Boyette, Cuthill, Fried, Gibson, Hewlett, Jamieson, Jankowiak, Honey, Huang, Liebert, Purzcyki, Shaver, Snodgrass, Sosis, Sugiyama, Swami, Yu, Zhaoi and Penton-Voak2014) on preferences for opposite-sex faces is valuable because data were obtained from 12 populations differing greatly in economic development. The expected preferences for faces with sexually dimorphic features, that is, of men for feminine faces and of women for masculine faces, increased in urbanized and developed societies. These features signal attractiveness much more in contexts of exposure to large amounts of visual information about faces.

Fostered by massive media exposure, consensual cultural norms develop about what constitutes attractiveness. For example, norms about female attractiveness vary between black and white populations in the United States, with thin bodies considered attractive among whites, but attractiveness among blacks extending to heavier women (Chithambo & Huey Reference Chithambo and Huey2013). In general, standards about what attributes of bodies and faces are attractive show considerable cultural shaping, whereas cues to ugliness show more cross-cultural consensus (Sorokowski et al. Reference Sorokowski, Kościński, Sorokowska and Huanca2014). As Zebrowitz and Rhodes (Reference Zebrowitz and Rhodes2004) found, attractiveness was validly associated with health and intelligence only in a sample of persons judged below-average in attractiveness. If so, avoiding unattractive mates may have evolutionary roots because ugliness can be an honest signal for poor health and infertility. Preference for attractive mates, with attractiveness signaled by certain facial and bodily features, is more likely governed primarily by sociocultural processes.

On completion of this commentary, we were astonished that the overlap of citations between the target article and our commentary consists only of two articles (van't Wout & Sanfey Reference van't Wout and Sanfey2008; Zebrowitz & Rhodes Reference Zebrowitz and Rhodes2004). This omission of key evidence that counters the thesis of the target article illustrates the principle of attitudinal selective exposure to information (see meta-analysis by Hart et al. Reference Hart, Albarracin, Eagly, Brechan, Lindberg and Merrill2009): Unless accuracy is a clear goal, people tend to seek out and consume information congenial to their pre-existing preferences. Scientists, in particular, often find it difficult to overcome fixing on research findings based on their compatibility with their preferred theory.