Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Three theories for facial paedomorphosis in human evolution and the preference for facial underdevelopment Wehr, Paul Arthur 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2005-104554.pdf [ 7.21MB ]
Metadata
JSON: 831-1.0092275.json
JSON-LD: 831-1.0092275-ld.json
RDF/XML (Pretty): 831-1.0092275-rdf.xml
RDF/JSON: 831-1.0092275-rdf.json
Turtle: 831-1.0092275-turtle.txt
N-Triples: 831-1.0092275-rdf-ntriples.txt
Original Record: 831-1.0092275-source.json
Full Text
831-1.0092275-fulltext.txt
Citation
831-1.0092275.ris

Full Text

THREE THEORIES FOR FACIAL PAEDOMORPHOSIS IN HUMAN EVOLUTION AND THE PREFERENCE FOR FACIAL UNDERDEVELOPMENT by PAUL ARTHUR WEHR B.A., McMaster University, 1994 M.A., California State University, Long Beach, 1998 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Psychology THE UNIVERSITY OF BRIITSH COLUMBIA March 2005 © Paul Arthur Wehr, 2005 Abstract Human faces are relatively underdeveloped compared to both ancestral human populations, and to other modern primates. Several researchers have noted a relationship between facial juvenilization and facial attractiveness in humans. Three theories were developed to account for what seems to be the co-evolution of underdevelopment in human morphology and mating preferences. In the baby-face overgeneralization hypothesis, facial underdevelopment stimulated altruistic behaviour and inhibited aggression. The youthful mimic hypothesis proposes that underdevelopment mimicked facial cues associated with youth and fertility. In the big brain hypothesis, facial underdevelopment was a necessary antecedent to brain expansion and human cognition. Predictions were generated from each hypothesis with regards to how the juvenilization preference would change as a function of context, mating strategy, and sex. Composite faces were manipulated to appear either more or less underdeveloped (juvenilized). Results indicated that the preference for juvenilized features was stronger in a mating context, stronger for male judges evaluating female targets, and stronger for a short-term mating strategy. Results were most supportive of the youthfulness hypothesis, but some support was also found for the big brain hypothesis. While the youthfulness hypothesis may explain the preference for facial underdevelopment, the big brain hypothesis might provide a more satisfactory explanation for the evolution of facial underdevelopment. i i T A B L E OF CONTENTS Abstract ii Table of Contents iii List of Tables iv List of Figures l v CHAPTER I Background and Research Question 1.1 Introduction 1 1.2 The Preference for Facial Juvenilization 3 1.3 Evolutionary Theories of Facial Underdevelopment 14 1.4 Summary and Plan of Action 51 CHAPTER II Experiment 1 2.1 Introduction 54 2.2 Methods 57 2.3 Results 63 2.4 Discussion 73 CHAPTER III Experiment 2 3.1 Introduction 78 3.2 Item Selection and Desirability 80 3.3 Methods 84 3.4 Results 85 3.5 Discussion 93 CHAPTER IV General Discussion 4.1 Results of the Predictions 100 4.2 Comparing the Hypotheses 105 4.3 Juvenilization or Feminization? 116 4.4 Summary and Conclusion 120 Bibliography 122 Appendix 135 in LIST OF T A B L E S TABLE 1.1 Summary of Hypotheses p. 53 TABLE 2.1 Individual Items Used in Experiment One p. 72 TABLE 3.1 Individual Items Used in Experiment Two p. 90 LIST OF FIGURES FIGURE 2.1 Examples of Stimuli p. 60 FIGURE 2.2 Mean Juvenilization Scores for Men and Women in Each Context p. 67 FIGURE 2.3 Mean Juvenilization Scores for Short-term and Long-term Mating Items Broken down Sex p. 68 FIGURE 3.1 Mean Desirability Scores of Items for Sexual Partners and Platonic Relationships p. 83 FIGURE 3.2 Mean Juvenilization Scores for Men and Women on Mating-relevant and Mating-neutral Items p. 87 iv Chapter I - Background and Research Question 1.1 Introduction Individuals of a species vary in their ability to survive and reproduce successfully. Those who mate and produce viable offspring will pass relatively more copies of their genes into the next generation compared to those who leave few or no offspring. Members of sexually reproducing species should exercise choosiness when selecting a mating partner to the extent that (a) individuals vary in their capacity to produce viable offspring, and (b) the physiological and material resources required for sexual reproduction are costly. Not only will they maximize their own fitness by linking their reproductive success to a high quality partner, but they will maximize the fitness of their offspring who stand to inherit at least a portion of the other parent's fitness enhancing virtues (Andersson, 1994). An individual's genotypic quality, however, cannot be assessed directly but must instead be inferred from his or her phenotype. Physical attractiveness can be conceptualized as an evolved, adaptive preference for phenotypes that, on average, reliably and honestly signal high fitness, mate value, or genotypic quality in the target (Barber, 1995; Berry, 2000; Jones, 1996; Symons, 1995). A number of factors involved in facial attractiveness have been linked to differences in fitness. Bilateral facial symmetry, for instance, is negatively related to developmental volatility, and positively related to semen quality in men, and, accordingly, is positively related to both ratings of male physical attractiveness and to number of sexual opportunities (Gangestad & Thornhill, 1997; Gangestad, Thornhill, & Yeo, 1994; Soler et. al., 2003; Thornhill & Gangestad, 1994). Facial attractiveness also predicts longevity. When male and female faces from a 1920 high school yearbook were 1 rated on attractiveness and perceived health, researchers found a significant relationship between attractiveness in early adulthood and age at death whereby more attractive individuals were discovered to have lived longer lives. Furthermore, attractiveness was a better predictor of life-span than was perceived health at the time the photo was taken. Although this was true for faces of both sexes, the relationship was weaker, but still significant, for female faces (Henderson & Anglin, 2003). Mathematically average features also tend to be rated as highly attractive in human faces. Mental retardation and disease are often coupled with a distinctive appearance, and even relatively minor learning disabilities are often complemented with unique facial anomalies (Zebrowitz, 1997). Composites created using graphic morphing software not only reflect the mathematical average of the faces used in their construction, but they are also rated as more attractive (Langlois & Roggman, 1990; Langlois, Roggman, & Musselman, 1994; Wehr, MacDonald, Lindner, & Yeung, 2001), and caricatures of faces that exaggerate deviations from the average tend to be rated low on attractiveness, whereas anti-caricatures tend to be rated more favourably (Rhodes & Tremewan, 1996; Rhodes, Sumich & Byatt, 1999). While symmetry and averageness have been the primary focus of facial attractiveness research, a number of studies have demonstrated an effect of facial juvenilization whereby decreasing the apparent growth in a face enhances its physical appeal (e.g., Cunningham, 1986; Jones, 1995; Wehr, MacDonald, Lindner, & Yeung, 2001). What makes this discovery particularly intriguing is the fact that the human face has evolved by means of under- (or truncated) development. This stands in stark contrast 2 to the predominant trend in human ontogeny which is over- (or extended) development (Brace, 1962; Parker & McKinney, 1999; Weidenreich, 1947). Chapter one has two objectives. First, to review and discuss the psychological and phylogenetic evidence demonstrating both the preference for facial juvenilization, and the ontogenetic history of facial underdevelopment in human evolution. The second objective is to propose a number of evolutionary theories on the potential adaptiveness of facial underdevelopment, each offering a different explanation for the attractiveness of juvenilized features in adult faces. In the Baby-face Over generalization hypothesis, facial underdevelopment is proposed to have evolved in response to overgeneralization of infant-directed behaviour to adults with infant-like facial features. In the Youthful Mimicry hypothesis, human mate preferences are proposed to have exerted selection pressure on morphological development, and the inclination to favour younger mating partners led to progressively less facial growth in subsequent human populations. In the Big Brain hypothesis, facial underdevelopment is proposed to have been one of several necessary antecedents to the evolution of human cognition and intelligence. 1.2 The Preference for Facial Juvenilization 1.2.1 Definition of Terms Several terms used throughout this manuscript share similar but distinct meanings. Paedomorphosis refers to the phylogenetic processes that lead to the evolution of facial underdevelopment in humans. A paedomorphic feature, then, is one that has evolved through changes in the timing and rate of development and is species specific. Juvenilization is the act or process by which experimenters manipulate the level of ostensible development in the face. Underdevelopment will most often refer to 3 individuals who, compared to others in the population, possess a relatively more juvenile facial appearance. 1.2.2 Facial Juvenilization and Attractiveness In general, juvenilized features, particularly those located in the face's centre, are rated as highly attractive. In reviewing evidence for this effect, three types of studies are relevant: correlational forays that divide the face into piecemeal and measure each feature proportion with exactness; experimental designs that manipulate the apparent level of facial development; and studies of "baby-faced" adults whose facial proportions are reminiscent of those observed in infants. While the first two types of studies have been able to demonstrate a preference for juvenilized facial features, the discussion will begin with the third class of investigation, those that have focused on babyish individuals, which has been relatively less successful. In one of the first studies of its kind, male and female facial schematics were created using a police Identi-kit and then manipulated to appear either more or less babyish by altering the dimensions of the eyes, nose and ears, and the vertical placement of features on the face. Schematics with larger eyes, shorter ears and noses, and lower placement were considered relatively more babyish than those with smaller eyes, longer features and higher placement. In general, increasing the babyishness of schematic faces did not enhance their physical appeal. Both high and low vertical placement of features were rated as less attractive compared to the original schematic, and adjusting the length of the nose and ears had no effect on attractiveness. However, there was evidence that judges preferred a babyish eye-size. Although no more attractive than the original eye-4 size, faces with larger eyes were rated as both more attractive and more desirable as a dating partner than faces with smaller eyes (Zebrowitz-McArthur & Apatow, 1983-1984). The failure of this pioneering effort to uncover a substantial juvenilization effect on attractiveness might be (in part) an artefact of its stimulus set, which consisted of only two male and two female schematics. Subsequent studies on the perceptions of baby-faced individuals have included larger samples and utilized photos of actual people in place of drawings. However, results of these investigations still show little support for the attractiveness of baby faces. When, for example, researchers asked male and female judges to rate 20 black and white photos of men on a variety of dimensions including babyishness and attractiveness, they found a moderately weak relationship whereby male faces that were rated high on babyishness were also rated as more attractive (Berry & Zebrowitz-McArthur, 1985). Two additional studies, one utilizing a large sample of 245 colour photos of men and women (Berry, 1991), and the other presenting both static and dynamic faces of male and female targets (Zebrowitz-McArthur & Montepare, 1989) found little to no relationship between babyishness and facial attractiveness. It might be tempting to conclude from the baby-face literature that facial juvenilization is actually w^attractive in adult faces. However, it is important to note that infancy is restricted to only the first few years of childhood and, as such, fails to adequately represent the full range of human juvenile morphology. Facial juvenilization is characterized by relative underdevelopment in an adult face whereas perceptions of babyishness reflect the degree to which an individual looks like an infant. It is actually counter-intuitive to expect perceptions of babyishness, or perceptions of children in general, to be rated as sexually attractive since, by definition, they are not yet sexually 5 mature. Finally, when considering the inter-relationships of three variables, it is not unexpected for two variables to correlate with the third but not with each other. Juvenilized features may be rated as both more babyish and more attractive despite the fact that a babyish perception is not sexually attractive. Thus, the absence of a relationship between perceptions of babyishness and those of physical attractiveness does not preclude the existence of a relationship between facial underdevelopment and attractiveness. A better investigation would study the features directly and independently of their ability to remind judges of sexually immature individuals. A series of correlational studies by Michael Cunningham has explored the facial morphologies associated with male and female attractiveness in both foreign and domestic samples. In one study, Caucasian male undergraduates rated black and white photos of both college women and a sample of international pageant winners on facial attractiveness. Ratings were subsequently correlated with 24 facial-metric distances measured in each photo using a micrometer. Analysis of the correlations revealed a positive relationship between rated attractiveness and several juvenile features including larger eye-size, wider eye-separations, smaller noses, and shorter chins. Two female secondary sex characteristics were also associated with attractiveness: wider cheekbones and narrower cheeks. Not only were women with these features rated as more attractive, but they were also rated higher on other attributes important within a mating context: they were rated as being brighter, more sociable, more fertile, as having fewer medical problems, and, as an indication of the woman's desirability to men in general, more likely to have an affair (Cunningham, 1986). 6 Similar results were found for undergraduate male faces when attractiveness ratings were provided by college-age women: men possessing the right mixture of juvenile features (larger eye-size and smaller noses) and secondary sex characteristics (prominent jaws and long chins) were rated higher on attractiveness (Cunningham, Barbee & Pike, 1990). More importantly, however, was the fact that cross-cultural correlations were consistent with North American samples. Asian and Hispanic students who had recently arrived in the United States, in combination with Taiwanese students living in their home country, rated the attractiveness of Asian, African, and European women from various countries. Although Asians were relatively less enthusiastic with regards to sexually mature features in the periphery, all the samples found juvenile features in the centre of the face attractive (Cunningham et al., 1995). By measuring the amount of exposure participants had to western movies and magazines prior to their participation in the study, the experimenters were able to rule out cultural contagion as an explanation for the consistency of attractiveness preferences since there was no difference between high- and low-exposure groups. Results from Cunningham's correlational studies provide support for both juvenilization and maturity in facial attractiveness. The authors proposed a multiple fitness model of facial attractiveness wherein juvenile and mature features act conjointly to signal both youth and sexual maturity of the target (Cunningham, 1986). It should be noted however, that (a) attractive juvenile features outnumbered attractive mature features, (b) the secondary sex characteristics associated with attractiveness are highly inter-correlated, and (c) correlational studies by themselves cannot establish cause and effect relationships. 7 The best evidence for the affect of facial juvenilization on attractiveness comes from laboratory experiments. Jones (1995) measured the impact of juvenilization on facial attractiveness in schematics of male and female faces using cardiodal strains. Positive strains produce a mature appearance by expanding the face outward near the base, and contracting it near the peak. A juvenile appearance is produced using a negative strain that expands the peak and contracts the base. For female faces rated low on attractiveness beforehand, the application of juvenile strains led to a substantial increase in attractiveness ratings. Ratings remained steady for other female faces, and for low to medium attractive male faces. Only in highly attractive male faces did juvenilization decrease attractiveness. While in many cases, juvenilization failed to enhance a face's appeal, neither did it generally impair it. Conversely, mature cardiodal strains always decreased a target face's allure regardless of its sex or initial attractiveness (Jones, 1995). Several studies have found large eyes (a juvenile feature) to be highly attractive. Of the babyish facial features studied in Zebrowitz-McArthur and Apatow (1983-1984), large eyes was the only one that was rated as physically attractive and desirable in a dating partner of either sex. Another study manipulating eye-size in colour photos of women found that not only do adults prefer large eyes in female faces, but five-month old babies expressed preference for them by gazing at them longer (Geldart, Maurer, & Carney, 1999). Some qualification may be necessary when considering the attractiveness of big eyes, however. One research group, for example, found that large eyes were rated as more attractive when combined with a manipulation increasing lip-fullness, but not in combination with a wider eye-separation (Paunonen, Ewan, Earthy, Lefave, & Goldberg, 8 1999) suggesting that the appeal of large eyes may depend on what other juvenile features are present in the face. Furthermore, the results from one cross-cultural investigation suggests that the influence of big eyes on attractiveness is weaker in male faces when compared to female faces. Participants from Brazil, Russia, the United States and members of the Ache, an indigenous group of hunter-gatherers from Paraguay, ranked male and female faces from both western and non-western populations on attractiveness. Female attractiveness rankings were positively correlated with eye-size in the Brazilian, American (U.S.), and Ache samples, but correlations for male faces and for female faces ranked by Russians, although positive in direction, did not reach significance (Jones & Hill, 1993). This lack of significance, however, may have resulted from the studies curious method of measuring eye-size, which was calculated for each face as a ratio of the eye-width to face-height, thereby mixing lengths associated with both a horizontal and vertical axis in the process. When eye-size is measured using ratios derived from a consistent axis, as they were in Cunningham's correlational studies, the correlations between eye-size and attractiveness were stronger and statistically significant for both male and female faces (Cunningham, 1986; Cunningham, Barbee & Pike, 1990), and this was true for American, Hispanic, and Taiwanese participants (Cunningham et. al., 1995). The attractiveness of juvenilization in female faces is further demonstrated in a tendency to "evolve" a juvenilized female face in a study using genetic algorithms programmed to imitate the process of natural selection. Faces were "crossbred" to create new generations of faces, with faces rated higher in attractiveness being crossbred more often than lower-rated faces. After many such generations, the evolved female face had a 9 larger forehead, a smaller, lower positioned face, fuller lips, and a narrower mouth (all juvenile features) compared to a composite of ordinary women's faces (Johnston & Franklin, 1993). Composite faces constructed using graphic morphing software provide useful stimuli for use in attractiveness studies because they offer attractive baselines free of idiosyncratic confounds, and they estimate the population mean from which they are drawn. Composites of male and female faces of Asian and European heritage were manipulated in Wehr, MacDonald, Lindner and Yeung (2001) to either increase or decrease juvenilization. Four facial features (eye-size, nose-size, lip-height, and eye-separation) were altered using Adobe Photoshop to construct five versions of each feature (two juvenile, two mature, and the original composite). Each version was compared to each of its alternatives in a forced choice format in which participants selected the face they thought was the most attractive. A significant and large linear effect was detected such that increased juvenilization resulted in greater facial attractiveness. There were two other significant findings that qualified the relationship between facial juvenilization and attractiveness. There was a significant interaction between the sex of the judge and target such that male judges ranking male targets preferred a slightly less, but still significantly, juvenilized appearance compared to other target-judge combinations. A significant main effect for the judge's race demonstrated an Asian preference for slightly less juvenilization, but again their preference was still significantly juvenilized overall. In general, it appears that juvenilized adult faces are physically attractive. Although research on perceptions of baby-faced adults does not agree with this conclusion, several reasons were offered to explain why this might be the case. In 10 particular, it was noted that perceptions associated with juveniles are unlikely to correlate positively with sexual attraction even if juvenile proportions in an adult face do. When researchers studied feature morphology directly, they noted that juvenile proportions were physically attractive in adult faces, particularly for features in the centre of the face. Furthermore, experiments that manipulated the degree of juvenilization in adult faces, both for schematic drawings and for photographs, while controlling for other aspects of the face found an increase in attractiveness with greater juvenilization. 1.2.3 Facial Paedomorphosis in Human Ontogeny Heterochrony describes phylogenetic changes resulting specifically from modifications to the timing or rate of development. Paedomorphosis is the heterochronic process whereby adult forms are "juvenilized" so that descendent populations are underdeveloped relative to ancestral populations. Underdevelopment occurs when the developmental trajectory is truncated, either through late onset or early offset in developmental timing or by retarding the rate of development. Conversely, peramorphosis "adultifies" descendent populations so that they are relatively overdeveloped (or augmented) compared to the ancestral population. Overdevelopment occurs by extending the developmental trajectory, either through early onset or late offset in developmental timing or by accelerating the rate of development (Gould, 1977; McKinney & McNamara, 1991). Heterochrony can be global in the sense that it can affect the entire phylogeny of a species or, more likely, it can be specific to individual features or involve a mosaic of different modifications. Although heterochrony normally describes differences between ancestral and descendant populations, it is also useful for describing individual differences within a population (Parker & McKinney, 1999). 11 Vigorous debate has raged over which heterochronic process exemplifies human ontogeny. Bolk (1926) first, and Gould (1977) and Montagu (1989) later, argued in favour of paedomorphosis, offering two main arguments to support their position. First, the superficial resemblance of juvenile chimpanzees to adult humans suggested to these scientists that human development was relatively truncated (paedomorphic) when compared to other primates. Second, the rate at which humans transition from one life-history stage to the next is relatively slower compared to other primates, past or present. This was interpreted as an example of retarded development - a paedomorphic process. However, humans do not develop slower, but rather at the same rate as other primates (Shea, 1988). Thus, the slower transition from one life-history stage to the next actually results in greater development (peramorphosis). In fact, the attributes that best characterize humans as a species - bipedal locomotion, extended periods of learning, and big brains - are each the result of overdevelopment (McKinney & McNamara, 1991; McNamara, 1997). Bipedal locomotion is possible due to extended development of leg and foot bones. The sutures responsible for fusing bone plates in the human skull close later, allowing the braincase to continue its expansion to make room for the largest of the primate brains. Compared to other primates, humans gestate longer, spend a longer period of time in the developmentally rapid stages of childhood and adolescence, and live longer adult lives. It has been suggested that shorter periods of learning would be inadequate to fully develop the information processing skills made possible by a larger, more complexly organized brain. Thus, extending each life-history stage may have ultimately paved the way for possibly our most human quality, higher cognitive functioning. (McKinney & McNamara, 1991; McNamara, 1997). 12 No species is entirely under- or overdeveloped, however, but is instead composed of a "mosaic" of heterochronic forces (McNamara, 1997). Whereas the human body and brain reflect peramorphosis, dental and facial growth generally reflect underdevelopment and paedomorphosis (Shea, 1989; McNamara, 1997, Coss & Schowengerdt, 1998). Although human facial development follows the same general pattern as our closest relatives (chimpanzees, gorillas, and orang-utans), human facial features begin comparatively more juvenile and change less over their developmental period. Compared to most other primates, humans have larger eyes, a lower positioned face, smaller jaws, and lack a muzzle (Shea, 1989). This is not to say that human faces are exclusively paedomorphic. Some juvenile traits do not result from paedomorphosis. Reduced dentition, for example, partially results from developmental retardation, but also reflects bone resorption (not a heterochronic process) in the jaw (Schultz, 1950). On the other hand, paedomorphosis is responsible for some non-juvenile traits. The human chin, absent in other primates, results from differential retardation of the upper and lower portions of the jaw. Thus, although the result of paedomorphosis acting on the jaw, the chin is not, strictly speaking, a paedomorphic feature since it is not retained from a juvenile form. The prominence of the human nose serves as another example, resulting from retardation of the jaw. As the dental arch decreases, the lips are drawn back, creating the illusion that the nose has actually extended (Glanville, 1969). While human faces may not be exclusively paedomorphic, the human face is, nevertheless, primarily the result of paedomorphosis and has become increasingly less developed over evolutionary time. 13 While the debate over this and that feature's true heterochronic origin may spark interest in phylogenetic circles, the more pressing question to psychological research is why humans should have evolved underdeveloped faces when overdevelopment seems to have shaped so much of human morphology? Because the current focus in heterochronic research has been to undermine the admittedly false notion that paedomorphosis was a phylogenetically central factor in human evolution, the selection pressures involved in facial juvenilization have been largely neglected or dismissed as trivial (McNamara, 1997). This disregard may have been over reactive given that the forces in question were powerful enough to dissociate the developmental trajectory of the face from the rest of the body. Best intentions notwithstanding, facial paedomorphosis may have played an integral role in human evolution. 1.3 Evolutionary Theories of Facial Underdevelopment The preceding discussion establishes two important points. First, human facial ontogeny is marked primarily by paedomorphosis leading to a relatively underdeveloped facial structure. Second, juvenilization is relatively more attractive in adult faces compared to adultification. This suggests, at least superficially, that human morphology and the psychological architecture governing mate preferences may have co-evolved, at least with regards to the face and standards of facial attractiveness. This raises some important questions. Was facial paedomorphosis and the preference for facial underdevelopment adaptive in the ancestral environment? If so, what were the selective forces that favoured facial paedomorphosis and the preference for facial underdevelopment? Did facial paedomorphosis and the preference for facial underdevelopment co-evolve (were they designed by the same or related selection forces 14 simultaneously) or did they evolve separately, either by one leading to the other or by completely independent processes? Three possible theories are proposed to account for facial paedomorphosis in human evolution, each offering its own explanation for the attractiveness of underdevelopment. First, the Baby-face Over generalization hypothesis proposes that individuals with underdeveloped faces elicited altruistic behaviour and inhibited aggression, thereby increasing their fitness. In this theory, paedomorphosis is driven by perceptions associated with facial underdevelopment, which is found aesthetically appealing due to its association with an infantile appearance. Second, facial underdevelopment, according to the Youthful Mimicry hypothesis, was a result of natural selection favouring morphological features that enhanced youthfulness cues in the face, and explains the attractiveness of facial underdeveloped in terms of the preference for youthful mating partners who tend to be more fertile and possess a higher residual reproductive capacity. Again, in this hypothesis, paedomorphosis is proposed to be driven by perceptions associated with facial underdevelopment. Third, in the Big Brain hypothesis, facial underdevelopment is explained as a necessary antecedent to the evolution of larger brain size and greater cognitive complexity and, as such, represents a generally honest and reliable cue to fitness in the form of intellectual development. Instead of preceding paedomorphosis, this alternative suggests that the preference for underdevelopment followed the heterochronic forces shaping the face's structure. Subsequent to full development, each theory will be used to generate a set of predictions with regards to the preference for facial juvenilization. Is the preference for juvenilization: (1) Stronger inside or outside of a mating context? (2) Stronger for short-15 term or long-term mating strategies? and (3) Stronger for men judging female faces or for women judging male faces? 1.3.1 Hypothesis 1: Over-generalizing Infantile Perceptions Facial underdevelopment may have evolved to exploit a facial recognition system originally designed for infant care whereby parental investment and inhibition of aggression were overgeneralized to adults with a relatively more juvenile facial appearance (Coss, 1991; Zebrowitz-McArthur & Apatow, 1983-1984). A juvenile recognition and response mechanism would have dramatically increased offspring survivorship in the ancestral environment by motivating and directing parental effort towards those juveniles (infants) with the greatest need (Alley, 1983; Wilson, 1975). One eliciting stimulus for this mechanism may have been infant morphology. Uniquely juvenile features stimulate parental behaviour and inhibit aggression in a wide variety of species (Eibl-Eibesfeldt, 1975; Harper, 1971; Tinbergen, 1964). Thus, the greater an individual's physical resemblance to an infant, the more likely he or she requires parental care, and the more likely parental effort will be mobilized and aggression inhibited. One important consequence of this design feature is that responses might be generalized to any individual possessing sufficiently infantile qualities including unrelated infants (Wilson, 1975), juveniles of other species (Lorenz, 1971), as well as older juveniles and "baby-faced" adults (Zebrowitz, 1997). Other adaptive mechanisms, such as parental solicitude (Daly & Wilson, 1980), may have evolved subsequently to bias investment towards children who are the most likely to contribute to parental fitness. Nevertheless, the emotional and perceptual responses associated with infants in general may have afforded some fitness enhancing benefits to individuals who merely resembled 16 } an infant to some degree. If true, the perceptions associated with facially underdeveloped adults should correspond to those of infants, and there should be evidence that these individuals do indeed receive greater investment from other adults and are aggressed against less often than their more mature-faced peers. Adaptive Responses to Infant Faces Infantile facial features and cranium shape can easily be distinguished from those belonging to human adults or even older children. Infants have a larger cranium-to-body-size ratio, and the shape of the cranium is more spherical in form. As the individual matures, the cranium grows unevenly and at a slower rate relative to the rest of the body, becoming less spherical and more oblong in shape. The area dedicated to the central features of the face most commonly employed for recognition and identification (eyes, nose, and mouth) composes a smaller fraction of the overall infant head-size, and is located in a lower vertical position. With the relatively more rapid development of the jaw and the advent of dentition, growth in the lower face begins to outpace that of the rest of the face, and features settle into a more central position. The features themselves also change with development. Infant eyes are nearly full grown to begin with and compose a relatively larger proportion of the face. This "wide-eyed" appearance is gradually lost as the rest of the face outpaces the eyes in growth. Compared to adults, infant noses are smaller, more concave and lack a bridge, and infant mouths are both narrower in width and, due to fat deposits in the lips, taller in height. Finally, some characteristically infantile peripheral features include chubbier cheeks, higher and finer eyebrows, and a smaller receding chin (Enlow, 1982). 17 The perceptual, affective, and behavioural responses aroused by infantile facial features and cranium shape seem to facilitate parental investment. When comparing head shapes reflecting different stages of development, the perceptions associated with the most infantile cranium profiles are consistent with an individual who requires considerable care (Abbie, 1947). In one study, infantile heads were perceived as more dependant, less alert, less intelligent, physically weaker, more helpless, and more vulnerable than outlines depicting head shapes of older children or adults (Berry & Zebrowitz-McArthur, 1986). The willingness to invest parental effort may partially depend on the degree to which an individual's appearance resembles that of a prototypical infant. Infantile head shapes are judged as both cuter and more lovable than the outlines of older children, and the more "baby-like" an infant's appearance, the more interest parents are likely to display and the more positive their affective response. Mothers tend to smile more at photos of their infants, and fathers tend to interact more with toddlers who possess archetypal features (Hildebrandt & Fitzgerald, 1983). People tend to show the same generally positive responses to infants even when they are not related to them. Relative to interactions with other adults, people tend to smile, gesture, and vocalize more when interacting with an infant, and they are also more likely to approach and maintain a closer proximity to an infant than they are to another adult (Zebrowitz, 1997). An adult's willingness to incur costs while actively defending an infant from violence may also depend on the degree to which he or she resembles a prototypically babyish appearance. A babyish cranium shape is more likely to elicit protection from adults compared to that of older children. A series of studies presenting profile and 18 portrait drawings traced from radiographs of children asked participants to identify which they would feel most inclined to defend from a physical attack. In two experiments, these head shapes were manipulated using cardiodal strains to increase or decrease their apparent cranio-facial development. In a third experiment, a series of head shapes representing actual growth were used. Results were consistent across all three experiments: the more infantile the outline, the more it was judged by college students to merit protection (Alley, 1983). Characteristics of infant faces and cranium shape may also play an important role in defusing stress induced aggression associated with child care. Although the manner in which people typically interact with infants is generally characterized by non-aggression, some highly aversive infant behaviour may provoke a violent response. To effectively persuade parents to approach an infant and relieve the source of its distress, crying must be sufficiently unpleasant to motivate its cessation (Bowlby, 1969). Instead of eliciting parental care, however, sometimes crying can act to trigger an abusive episode (Frodi, et al., 1978), and even non-abusive parents experience physiological changes consistent with increased agitation and express feelings of irritation and annoyance when exposed to a crying baby (Frodi, Lamb, Leavitt, & Donovan, 1978). Prematurely born infants, who lack a true babyish appearance, are at greater risk of violence and neglect. The eyes of premature babies are not as wide, and their craniums lack the characteristic roundness of full-term infants. And, according to one estimate, the risk for battering increases as much as five-fold for children who are born prematurely (Zebrowitz, 1997). The results of one study comparing the physiological responses of adults to pre- and full-term babies suggests that the physical appearance of 19 premature infants fail to restrain hostile reactions to crying. Parents who watched a videotape of a pre-term infant who was quiet for two minutes, then crying for two minutes, then quiet again experienced greater increases in skin conductance and diastolic blood pressure, and indicated less interest in interacting with the infant compared to parents who viewed a full-term infant (Frodi, Lamb, Leavitt, & Donovan, 1978). Although some of the parents' reactions to the prematurely born infant could be accounted for by the distinctiveness of its cry, the pre-term cry was most aversive when it was coupled with video of the pre-term infant. Apparently, some ofthe negative effects of pre-term cries could be ameliorated by the physical attributes of a full-term infant. Another study found similar results using line drawings of infant faces at various stages of development ranging from nine weeks premature to full-term. College students were less interested in caring for, or getting close to, pre-term babies, and perceived them to be less cute, less lovable, less fun, and more irritating than full-term babies. Moreover, participants' reactions became increasingly more negative as the term decreased, and the physical appearance of the child became less prototypically baby-like (Maier, Holmes, Slaymaker, & Reich, 1984). Responses to Infant Faces are Overgeneralized to Baby-faced Adults Infant features provide important cues that are highly functional within the context of the parent/offspring relationship. They signal to adults the developmental status of an infant, facilitate parental effort in the form of nurturance and protection by increasing an adult's willingness to invest in the infant and inhibit aggressive reactions to aversive stimuli associated with infant care. Furthermore, responses are generally manifested towards infants regardless of his or her relatedness to the adult. Facial 20 underdevelopment may have been selected for if these reactions to infantile stimuli were generalized to include adult relationships in ways that enhanced fitness. For example, paedomorphic adults may have continued to elicit parental investment past childhood and into early adulthood, or perhaps they were more successful in soliciting help and protection from tribe members, or experienced other forms of altruism more frequently. Perhaps tribe members were more willing to engage in social exchange with them, or they experienced less aggression from rivals when competing for mates, or received less severe sanctions for violating social norms. If true, we would expect the perceptual, affective, and behavioural responses elicited by infant faces to generalize to adults with a relatively underdeveloped facial morphology. Adults with less developed faces tend to possess many of the same features associated with an infantile appearance - a rounder, less angular face shape, larger and rounder eyes, a smaller nose, thinner and higher eyebrows, and lower vertical positioning of the central facial features which, in turn, make the forehead appear larger and the chin to appear shorter (Zebrowitz, 1997). Adults with an underdeveloped appearance are often perceived to exhibit child-like qualities. In one study, schematics of male and female faces were first manipulated to either increase or decrease their babyish appearance by altering the eye-size, nose and ear-length, and the vertical placement of features on the face. Participants then rated the faces on their perceived strength, dominance, intelligence, warmth, and honesty. The higher the face was rated on babyishness, the higher it was rated on submissiveness, warmth, and honesty, and the lower it was rated on strength and shrewdness. No relationship was found with regard to intelligence, however. Furthermore, predictions regarding baby-faced adult behaviour 21 were consistent with those of personality perceptions. The more babyish the target, the less likely they were thought to be able to perform acts of strength (lift heavy boxes), cruelty (give the cold shoulder), or dishonesty (cheat on an exam), and the more likely they were thought to believe a tall tale (Zebrowitz-McArthur & Apatow, 1983-1984). In a subsequent study, researchers asked male and female judges to rate 20 black and white photos of men on a variety of dimensions and found facial babyishness to predict ratings of perceived naivete, honesty, kindness, and warmth, but not responsibility (Berry & Zebrowitz-McArther, 1985). These original studies focused on perceptions associated with photos and drawings of young adult faces with a babyish appearance. This oyergeneralization effect is not restricted to static faces, however, and physical appearance seems more important than some other socially relevant information, such as an individual's voice. In a study comparing portraits and film clips of individuals varying in facial babyishness with and without voice information, perceptions of interpersonal warmth and physical weakness were predicted by the degree of facial babyishness in the target's face. Although these perceptions were weaker when the target's voice was inconsistent with his or her appearance, a babyish face coupled with a mature voice for example, the target's voice had no effect on perception unless it was the only information available (Zebrowitz-McArthur & Montepare, 1989). Thus, the childlike perceptions associated with baby-faced adults are generalized to the moving, talking faces that would have been encountered in the ancestral environment. The effect of facial babyishness on interpersonal perception is not restricted to youthful individuals either, but seems to endure across one's life span. University 22 undergraduates rated portraits of infants, preschoolers, grade-schoolers, young adults, and older adults on facial babyishness along with a number of personality scales. The data were consistent for targets in every age group. Individuals with babyish faces were rated higher on interpersonal warmth, honesty, naivete, physical weakness, and lower on social autonomy compared to their more facially developed peers (Zebrowitz & Montepare, 1992). Finally, the effect of facial babyishness on perceptions of personality is not limited to Western samples. In a cross-cultural investigation , the judgements of babyish Caucasian faces provided by a Korean sample were nearly identical to those provided previously by North American participants (Zebrowitz-McArhtur & Berry, 1987). Thus, the fitness enhancing benefits associated with facial underdevelopment, whatever they may have been, as they relate to interpersonal perception could have been realized across the entire life span of the individual. And yet, for the overgeneralization hypothesis to be true, perceptions of baby-faced adults must translate into action, with more infantile adults receiving greater investment and less aggression than their more mature-faced peers. Juveniles of other mammals and some birds also have a babyish appearance and also elicit nurturing feelings and behaviour from humans (Lorenz, 1971), but is the same true of babyish human adults? At the very least, the results of one study indicate that facially underdeveloped men and women receive more helping behaviour. Researchers who "lost" fictional resumes in shopping centres and outdoor markets in both the eastern United States and the African nation of Kenya included a picture of the applicant with the correct postage and a note indicating that the letter needed to be mailed that day. Eight applicants were used in all, four men and four women, two of whom were black and two 23 white. Each picture was manipulated to appear either more juvenile or mature by altering the size of the eyes and lips in the applicant's face. An undergraduate sample confirmed that juvenilized faces were indeed perceived to possess more infant-like qualities in that they were rated as more submissive, weak, naive, compassionate, and honest looking. Helping behaviour was defined as mailing the resume on the applicant's behalf. Resumes depicting a juvenilized appearance were mailed more often for both female and the Caucasian male applicants but not for the African male applicant (Keating, Randall, Kendrick & Gutshall, 2003). That the juvenilized African male applicant failed to receive greater help may reflect the perception of Black men, both in the United States and in some African countries, as dangerous and aggressive individuals, starkly contrasting the behaviour typical exhibited by infants. Although this is the only study to demonstrate this effect thus far, it does suggest that babyish looking adults with underdeveloped faces may have received greater investment, at least measured by helping behaviour, which may have increased their fitness in the ancestral environment. Alternatively, fitness might have been enhanced if facially underdeveloped individuals in the ancestral environment encountered less aggression, or if aggressive acts were weaker in severity. Being the target of aggression can incur severe costs in the form of injury or death that dramatically reduce fitness. Violating important social norms governing the group and cheating in social exchanges are two actions that could have provoked an aggressive reaction in the ancestral environment since they inflict costs on the perpetrator's contemporaries. While these actions can elicit aggression in the modern environment as well, they may also find resolution in a criminal or civil court. Results from two separate studies suggest that baby-faced adults are more likely to receive 24 verdicts in their favour and are more likely to avoid punishment in legal cases, at least for allegations of premeditated contraventions. In the first study, an undergraduate sample read fictitious criminal case summaries accompanied by a photo of either a baby-faced or mature-faced male defendant controlling for facial expression and attractiveness. Cases varied in the seriousness of their crime (unreported income in the form of tips versus sale of a potentially hazardous product), and on the intent of the defendant (deliberate action versus negligence). The likelihood of a guilty judgement corresponded to the facial appearance of the defendant. Babyish looking men, who were perceived as less competent than their more mature-faced peers, were more likely judged guilty of a negligent infraction, and less likely judged guilty of a deliberate transgression (Berry & Zebrowits-McArthur, 1988). In a separate condition of the same experiment, judges were asked to recommend sentences for the defendants who, in this case, admitted their guilt. Results indicated a contrast effect where harsher sentences were recommended for babyish men who had deliberately broken the law and for mature-faced men who behaved negligently, but lighter sentences for deliberate criminal acts by mature faced men and negligent deeds by baby-faced men (Berry & Zebrowits-McArthur, 1988). Thus, facially underdeveloped adults in the ancestral environment may have been able to deliberately subvert socially proscribed conventions without paying a heavy penalty as long as they maintained their innocence. Furthermore, given that baby-faced adults are perceived as honest and, therefore, less likely to commit crimes intentionally, they may find themselves facing criminal charges less often altogether. Conversely, the likelihood of being charged with negligence might be greater since baby-faced adults are 25 perceived as less competent and more likely to make mistakes. Nevertheless, intentional violations are more likely to be pursued, punished, and (generally) carry much stiffer penalties. Thus, the costs of violating social norms in the ancestral environment may have been much lower for facially underdeveloped individuals whose actions would more likely be attributed to an accident than to villainy. Civil court cases between litigants often reflect instances of cheating in social exchanges rather than violations to socially prescribed rules. How do baby-faced defendants fare in civil court decisions? In an investigation of over 500 actual small claims cases from the Boston area, the babyishness and attractiveness of both the plaintiff and defendant were rated by a pair of observers. In cases where the defendant denied responsibility, the probability of losing the case depended on both the intent and appearance of the defendant. For cases involving accusations of wilful misconduct, the likelihood of an unfavourable ruling decreased as the rated babyishness of the defendant increased, with 92 percent of the most mature-faced defendants losing their cases and only 45 percent of the most baby-faced defendants losing theirs. Contrary to the criminal case findings of the previous study, babyish and mature faced defendants were equally likely to lose their case when they were charged with negligence. Thus, facially underdeveloped individuals may have incurred fewer costs when accused of cheating on social exchanges in the ancestral environment. Moreover, in cases where the defendant admitted responsibility but disputed the amount of the claim, the amount of the award depended on the degree of babyishness of both the plaintiff and the defendant, with the largest stipends awarded to baby-faced plaintiffs who were suing mature-faced defendants (Zebrowitz & McDonald, 1991). This tendency could reflect a more general 26 tendency for people to protect facially underdeveloped individuals, perceived as relatively helpless, from the unscrupulous designs of deceitful persons who may try to exploit them. Summary & Predictions If the benefits to fitness arising from these reactions to facial underdevelopment were sufficient to influence the course of evolution, they probably did so outside the realm of mate selection. While infant faces are physically attractive to adults (Alley, 1981), and while adults prefer to look at pictures of babies over those of other adults (Zebrowitz, 1997), they are not sexually attractive. Babyish profiles are rated as less sexy compared to more mature cephalic outlines (Berry & Zebrowitz-McArthur, 1986), and babyish adults are not more often preferred as dating partners, although they are not generally discriminated against either (Berry, 1991; Zebrowitz-McArthur & Apatow, 1983-1984). Thus, facial underdevelopment should only be found attractive when the context is non-sexual and, consequently, the preference for underdevelopment should be independent of mating strategy since both long- and short-term strategies should be equally irrelevant. Given that the adaptive advantage proposed by the overgeneralization hypothesis entails supplementary investment and reduced aggression, it is unclear as to whether men or women would benefit more from facial underdevelopment. It would be a mistake to conclude that women gain the greater benefit based solely on existing differences in facial development since present day morphological disparities fail to inform us as to the rate of phylogenetic change that occurred within each sex, and no attempt yet has been made by evolutionary biologists to compare the individual ontogenies of male and female 27 facial development outside of the phylogenetically older influence of sexual dimorphism. Both sexes could have benefited from supplementary parental investment during childhood, and it is difficult to argue either way as to who would have benefited more from food sharing, protection, and generally pro-social behaviour. However, these same perceptions might also limit facially juvenile looking men in their ability to rise to the top of social hierarchies and ultimately affect their ability to attract healthy, young mating partners. If so, women would have experienced a greater net benefit from facial paedomorphosis. On the other hand, most violent acts are directed towards men, so facially underdeveloped men may have benefited more than women from reduced aggression. Thus, there is no predictable sex difference in the adaptiveness of facial paedomorphosis. With regards to the preference for facial juvenilization, however, we would expect a sex difference. Women generally show more interest in infants and childcare compared to men, and women are more likely to smile and initiate contact with an unfamiliar infant (Schleidt, Schiefenhovel, Stanjek & Krell, 1980). Thus, if underdeveloped faces are attractive because they resemble, to some degree, an infant-like appearance, women should demonstrate the stronger penchant for juvenilized adult faces. In summary, perceptions of facially underdeveloped adults are similar to those of children and, particularly, infants. Babyish adults are perceived as more honest, physically weaker, less competent, and warmer than their mature-faced peers. These perceptions tend to elicit behaviour resembling parental effort that might have benefited facially underdeveloped individuals in the ancestral environment. Babyish adults receive more helping behaviour, and they are punished less often and less severely for violating social norms and contracts. Hence, human facial underdevelopment may have evolved 28 because juvenile facial features increased the likelihood of eliciting investment-like behaviour from conspecifics in the ancestral environment and inhibited aggression, thereby affording fitness enhancing benefits to those individuals whose trajectories for facial growth were developmentally truncated. The context in which this occurred was not one of sexual attraction, however. Therefore, it is hypothesized that the preference for facial juvenilization should be greater outside of a mating context and independent of mating strategy. Furthermore, although no predictable sex difference in the adaptiveness of facial underdevelopment could be made, women were hypothesized to prefer facial juvenilization in a target face more so than men. 1.3.2 Hypothesis 2: Youthful Mimicry Facial underdevelopment may have evolved as a by-product of a human mate preference favouring more youthful partners for sexual reproduction. After the age of twenty, both fertility (the likelihood that a reproductive episode w i l l result in viable offspring) and residual reproductive capacity (the probable total number of future offspring that an individual can produce) begin to decline, and those offspring that are produced in later life are at greater risk for physical and mental abnormalities. Human ancestors who preferred younger (sexually mature) partners in the ancestral environment would have left a greater number of viable progeny compared to those who preferred older partners or who did not express any age preference. This inclination for youthful partners could, in turn, have acted as a selection pressure on human ancestral morphology to retain youthful features longer. Consequently, facially underdeveloped partners might have been more desirable in the ancestral mating market because they mimicked a 29 youthful appearance, despite the fact that fertility and residual reproductive capacity probably remained constant. Fertility and Residual Reproductive Capacity Decrease With Age While the ability to produce and raise healthy, viable offspring to sexual maturity declines with age in both sexes, the descent is much more pronounced in women. The ability to produce offspring peaks near the age of twenty in women, and then steadily declines until reaching zero at the conclusion of menopause. While the average length of time it takes a woman in her early twenties to successfully conceive is four-to-five months, this period increases to seven-to-ten months by her early thirties and reaches ten-to-twelve months by her late thirties. Lower rates of fertilization and implantation in older woman are largely, though not exclusively, due to age-related changes occurring in female reproductive physiology. Egg quality degrades as the ovum ages, becoming less fertile and more irregular in shape, and the uterus becomes less capable of sustaining and nurturing the zygote (Schneider, 1978). Following conception, the probability of foetal loss is much higher in older women, with spontaneous abortions occurring in about half of all pregnancies in women over the age of 30 (Harlap, Shiono & Ramcharan, 1980; Risch, Weiss, Clarke & Miller, 1988). At around fifty years of age, following menopause, the menstrual cycle ends entirely and the ovaries cease to release eggs, permanently ending a woman's ability to produce offspring and ultimately limiting her reproductive career to about 35 years on average. Comparatively, the decline in male fertility associated with advancing age is modest and does not reach zero until death. Although monogamous men end their reproductive lives in concert with their menopausal wives, healthy men under normal 30 circumstances are physically capable of siring children at any age. Nevertheless, the conception rate does decline as a result of age-related changes in male physiology. The tissue mass of the testes softens and shrinks over time (Stearns et. al., 1974; Tillenger, 1957) and sperm quality decreases with age as the mutation rate associated with sperm production increases and chromosomal abnormalities become more frequent (Modell & Kuliev, 1990; Taysi, 1988). Although overall sperm cell production does not necessarily decrease, the concentration of normal spermatozoa declines with age and sperm motility diminishes, making fertilization of the ovum less likely to occur (MacLeod & Gold, 1953; Schneider, 1978). What is important to note, however, is that the decline in conception rate associated with paternal age is more gradual and less dramatic. Given that fertility is partly predictable through age, natural selection should have favoured individuals in the ancestral environment who preferred youthful partners for sexual relationships. Moreover, given that female fertility is more adversely affected by advancing age, the preference for youth should have been relatively stronger in ancestral men. The same conclusion can be derived when considering residual reproductive capacity. The expected number of future offspring, based on the average reproductive performance of similar members of the same species, also declines with age (Daly & Wilson, 1983). The average thirty-year-old woman, for instance, has the same total reproductive capacity as the average twenty-year-old woman but, because she has already experienced ten more years of reproduction to date, the average thirty-year-old will necessarily have fewer future offspring. From a reproductive standpoint then, individuals in early adulthood make "better" choices as mating partners because they have greater residual reproductive capacity. Because female reproduction takes place within a 31 restricted window of opportunity while male reproduction does not necessarily, residual reproductive capacity declines relatively faster in women. Thus, from a reproductive standpoint, pressure to choose relatively young mating partners is much stronger for men. In addition to fertility and residual reproductive capacity (factors relating to the production of offspring), a mate preference favouring youthful partners would have enhanced the quality of offspring. For children to function as effective vehicles of gene transmission, they would have to be healthy and free of genetic abnormalities. Errors in DNA replication that occur during sperm cell production in older men have been linked to several congenital disorders including dwarfism, haemophilia, and muscular dystrophy (Jones et al., 1975; Penrose, 1955). While the risk of birth defects increases with the age of both parents, syndromes related to maternal age tend to be both more frequent in their occurrence and more devastating in their consequences. Maternal age-related syndromes typically result from possession of an extra chromosome, the most common of which is Down's syndrome. Marked by brachycephalicism (shortened posterior/anterior axis of the skull), mental and physical retardation, accelerated aging, immunological complications, and early onset of senile dementia (Schneider, 1978), the risk of Down's is (1 in 45) for women over the age of 45 compared to 1 case in every 2000 pregnancies for twenty-year-old mothers (Collman & Stoller, 1962). Despite the fact that maternal and paternal ages are highly correlated, the higher risk of Down's syndrome is primarily associated with advancing maternal age. When the age of the mother is held constant, there is no difference between men of different age groups in the likelihood of Down's, whereas the risk of Down's continues to increase with maternal age when paternal age is controlled (Cohen & Lilienfeld, 1970; Penrose, 1933). Also strongly related to advanced 32 maternal age are Edward and Patau syndromes, both of which involve extreme mental and physical retardation and only a marginal chance of survival past the first year of infancy (Schneider, 1978). Thus, the cost to offspring quality should have been steep enough in the ancestral environment to drive selection to favour youthfulness in mating partners and should have been particularly strong in shaping male mating preferences. Human Mate Preferences Favour Youthfulness Ancestral humans who preferred youthful partners generally would have succeeded in producing healthy viable offspring to a greater extent than those who lacked this mate preference—the latter experiencing greater difficulty conceiving, being less likely to carry the foetus to full term, and being more likely to produce offspring with serious health issues. Thus, we would expect modern human populations to demonstrate a preference for relatively youthful partners in sexual relationships. Two lines of evidence confirm this prediction: perceptions of female fertility accurately reflect decreases in fertility associated with the aging process, and the amount men must pay in cultures that require a bride price decreases as the age of the bride increases. As age-related cues in female faces increase, people perceive them as less fertile and rate them as less desirable for mating. Furnham, Mistry, and McClellannd (2004) measured people's perceptions of female faces at three different reproductively viable ages. Using a digital morphing technique, they were able to estimate what the same female faces would look like in their early twenties, early thirties, and late thirties. Male and female judges then rated one picture of each woman on a number of mating relevant dimensions. Overall, there was a linear decline in the desirability of the faces as age increased. The young face was rated as more physically attractive and sexier than the 33 older faces, and was perceived as more fertile and healthier as well. Male judges rated the twenty-something faces as the most desirable for both short-term opportunistic and long-term committed relationships, and female judges rated the younger faces as the most desirable to men in general. Thus, as the age of the faces increased, attractiveness ratings decreased, as did perceptions of fertility and health. Furthermore, the decline was more pronounced between the early and late-thirties compared to that observed between the early twenties and early thirties (Furnham, Mistry, & McClellannd, 2004). Given the smaller disparity in the age difference between early and late thirties, this suggests that the decline in attractiveness and desirability of female faces accelerates after the thirtieth year (on average) of a woman's life. Furthermore, perceptions of declining fertility seem to be at the root of female desirability as marriage partners in cultures where a bride price must be paid before the wedding. In some human societies, the mating system requires the groom or his kin to transfer money or goods to the bride's family in exchange for the woman's hand in marriage (Daly & Wilson, 1983). In practical terms, the groom is purchasing the exclusive right to link his reproductive success to that of the bride's. The more desirable she is as a reproductive partner, the more expensive the bride price he must pay. Younger brides with greater residual reproductive capacity demand higher bride prices than older brides (Buss, 1994). In cases where the woman proves unfaithful, or the union fails to result in children, the husband may attempt to return the woman to her family and demand a refund (Daly & Wilson, 1983). The amount reimbursed, however, is not the original sum, but reflects her reproductive potential at the time she is returned. The older the woman is, the smaller the refund (Buss, 1994). 34 Thus far, the focus of the discussion has been restricted to the effect of youthfulness on female fertility and desirability within sexual relationships, establishing a strong male preference for younger, reproductively mature women. While male fertility also decreases across the life span, the negative effect of age on male desirability is less dramatic since male reproductive capacity declines at a slower rate and abnormalities associated with paternal age have less severe consequences for offspring fitness. But there are additional reasons why male desirability may not dramatically decrease with age. First, older men tend to have accumulated greater status and wealth, which could enhance their desirability to the extent that women generally covet them. Second, older men possess greater emotional stability, are less violent, and are more conscientious than younger men. Women typically prefer men with these temperamental traits because they impose fewer costs to female fitness in the form of abuse, selfishness, and infidelity. Third, older men have longer track records from which women can evaluate both the men's ability and willingness to act as providers (Buss, 1989; Buss, 1994). That is, older men have established a history from which women can more accurately gage both his ability to acquire resources and his willingness to share them with her. For these reasons, women are unlikely to have inherited a strong preference for youthfulness. This is not to suggest that they should prefer elderly or even radically older men, but rather that— unlike ancestral men—there was probably not a strong selection pressure favouring ancestral women to prefer youthful partners. Studies investigating both male and female desirability as a function of age have confirmed that the preference for youthful partners is relatively stronger in men than in women. Kenrick and Keefe (1992) studied personals advertisements in "singles" 35 newspapers from Arizona in which ages of both the advertiser and the partner sought were indicated. Relative to their own age, men in their twenties were interested in women who ranged in age from five years younger to five years older than themselves. The minimum and maximum ages for female partners increased with the man's age, but at a much slower rate until, by their fifties, men were seeking women who were 5 to 15 years their junior. These results are consistent with a male mate preference for young, fertile women. As men get older, fertile women become increasingly younger by comparison and, consequently, so does the male standard of an ideal hypothetical mating partner. Conversely, women generally preferred older men. This standard varied little across age groups, with the minimum age ranging from 0 to 5 years younger, and the maximum ranging from 8 to 11 years older (Kenrick & Keefe, 1992). Differences between male and female mate preferences cannot be attributed to either disparities in earning power, or to unique aspects of American culture. The same general pattern of results was found when comparing men and women of high and low socio-economic status. Although high SES individuals tended to advertise for relatively younger partners, men wanted younger partners than women at every level of SES (Kenrick & Keefe, 1992). Thus, even after economic parity has been achieved, men should still exhibit a stronger preference for youthfulness than women. Similar results were also obtained from German and Dutch samples as well as for personal advertisements printed in India. In the latter, where ads tended to be placed by the individual's kin and not the individual themselves, the maximum age for a female partner was even lower than in western samples. It seems that female reproductive capacity is even more important to family members whose interests may be concerned primarily 36 with the production of grandchildren (Kenrick & Keefe, 1992). In another study of 37 cultures spanning every continent and consisting of a variety of mating and economic systems, men preferred women who were on average 2.5 years younger than themselves. In relatively monogamous cultures, men preferred women about one year younger than themselves, while in relatively polygamous societies—where men must wait until they can acquire the resources to attract wives—men preferred women about seven years younger than themselves. Regardless of the mating system, men in traditional societies preferred women who had just attained sexual maturity (following menarche), but who had not yet bore her first child (Buss, 1989; Buss, 1994). Actual marriage statistics are consistent with this pattern of preferences. When they analyzed all the marriages that took place in Phoenix and Seattle during a one-month period in 1986, Kenrick and Keefe (1992) found that men tended to marry women who were only a year or two younger than themselves when they were in their twenties, but who married increasingly younger women as they got older until, by their fifties, they were marrying women almost 10 years younger. Women, on the other hand, tended to marry men who were 1 to 5 years older. Thus, not only do men express an interest in youthful women, but they tend to marry them as well. Marriage statistics collected from earlier last century and from other geographical areas confirmed this trend. Kenrick and Keefe (1992) analyzed a sample of 100 marriages which took place in Phoenix in 1923 and all the marriages which occurred between 1913 and 1939 on the Philippine island of Poro whose people, at that time, had limited experience with American or European cultures. The researchers discovered that men married increasingly younger women as they got older while women married slightly older men regardless of their own age 37 (Kenrick & Keefe, 1992). In fact, the differences in age between husband and wife were even more drastic in these "earlier" samples compared to those from Phoenix and Seattle during the eighties. Early last century, the age of female marriage partners decreased much more steeply in relation to their husbands' ages, with men in their sixties marrying women who were on average 15 to 20 years younger than themselves. If evolution designed men to prefer youthful women because female fertility peaks in early adulthood, why would older men still prefer women who have exceeded the age of maximum fertility? Why would men in their fifties and sixties, for example, continue to prefer women in their thirties, forties or even fifties as opposed to women in their early twenties? There are several possible reasons why older men may choose a mating partner whose age is greater than maximal fertility, but still younger than themselves. First, with the advent of cosmetics and cosmetic surgery, and low exposure to the sun and harsh living, women retain a youthful appearance longer than they otherwise would have in the ancestral environment (Buunk, Dijkstra, Kenrick, & Warntjes, 2001). Consequently, they continue to exhibit youthfulness cues past the age of maximal fertility. Second, men and women adjust their preferences to market pressures and pursue what they can obtain (Kenrick & Keefe, 1992). Elderly men who lack lavish resources cannot compete for twenty-something women and, therefore, pursue women who are older than maximal fertility but who are still younger than themselves. Third, age is only one of a multitude of considerations in evaluating partners for romantic relationships. Indeed, older men may find that they have more in common with moderately younger women, particularly when considering such women for long-term committed relationships (Kenrick & Keefe, 1992). 38 Theoretically, if market pressures were relieved, male and female preferences would reflect their ideal mate (Buunk, Dijkstra, Kenrick, & Warntjes, 2001). Men and women ranging in ages of twenty to sixty years of age were asked to indicate the minimum and maximum ages for partners they would consider in five relationship types that increased in their involvement level: Marriage, serious relationship, the act of falling in love, casual relationship, and sexual fantasy. This last "relationship" type requires no commitment from either the individual or the object of their fantasy and should reflect unconstrained mating preferences. Overall, women consistently preferred men from their own age group, ranging from a few years younger to a few years older than themselves. Men, conversely, preferred women ranging from much younger to slightly older than themselves. This was consistent with Kenrick and Keefe's (1992) findings. However, the male preference for youthful partners across the life span interacted with the type of relationship under consideration. For twenty-year-old men, the age range (19-25 years) for long-term relationships was narrower than the range for less committed relationships (17-31 years). In other words, men were more discriminating when choosing a marriage partner in whom they would invest heavily, and the women they preferred were either at or near their peak fertility. As the age of the men increased, their preference for youthful marriage partners became less restrictive as market forces came into play. Sixty-year-old men would consider higher minimum and maximum ages as well as a larger overall range of ages (43-55 years). But the age range for sexual fantasies reached a ceiling at 40 years and increased no further (Buunk, Dijkstra, Kenrick & Warntjes, 2001). While sixty-year-old men actually preferred to marry women over the age of 40, their sexual fantasies still focused on youthful women with high fertility. This suggests that the preference for 39 women past the age of maximal fertility in older men reflects market pressures in the dating game. If for some reason, market pressures were rendered irrelevant, the female age preferred by older men may decrease to track fertility with greater precision. Thus, expectations forecasted by the evolutionary view are borne out. Overall, men generally prefer youthful women for a variety of sexual relationships, and they prefer and marry women increasingly younger than themselves the older they get. Conversely, women do not show a similar preference for youthful mating partners but, instead, tend to prefer men from their own age group, or slightly older. Summary & Predictions Male and female fertility decreases across the life span. As parents get older, the probability of successful conception, a full-term pregnancy, and birth of a healthy normal offspring become increasingly unlikely. Although this is true for both men and women, the decline in fertility is more pronounced for women, and the consequences of maternal age-related abnormalities in offspring are more severe. Moreover, several aspects of male mate value correlate positively with maturity that may compensate for the decline in male fertility. Hence, natural selection has shaped male, but not female preferences to prefer youthfulness in mating partners. Therefore, if facial paedomorphosis evolved because of fitness enhancing benefits resulting from youthful mimicry, men should express a stronger preference for juvenilization in female faces than women do in male faces. Furthermore, given that this hypothesis places the origin of facial underdevelopment within mate selection, we would expect the preference for facial juvenilization to be stronger for judgments made within a mating context. For instance, 40 juvenilized features should be relatively more salient to a judge when the target is being considered for a date as opposed to employment. The best years for a woman to reproduce ranges from her early to late twenties, when the likelihood of conception is still high but the danger of maternal and child mortality is at its lowest. However, men find women most attractive before this time, during the years when fertility and ovulatory cycles first begin, but maternal and child mortality are still high. This suggests that men's preference for young women might reflect a long-term mating strategy where they attempt to monopolize a woman's full reproductive capacity (Symons, 1979). However, this does not preclude the possibility that youth is highly desirable in short-term mating partners, and, in fact, one study already reviewed, suggests that it may be relatively more important for short-term mating. When investigating the minimum age preference for partners across different levels of relationship involvement, men preferred younger partners overall for casual affairs and sexual fantasies compared to marriage or a committed relationship. The same effect was observed in women, but, predictably, the preference was much stronger in the male respondents (Buunk, Dijkstra, Kenrick & Warntjes, 2001). Thus, if the youthfulness hypothesis is correct, we would predict a stronger preference for juvenilization in a face for short-term relationships, particularly when the judge is a man and the target is a woman. 1.3.3 Hypothesis 3: An Antecedent to Human Cognition Over evolutionary time, humans have become increasingly more specialized as a cultural species. This has been possible primarily due to peramorphosis (overdevelopment) of the skull resulting in a dramatic increase in brain size that 41 ultimately provided the information-processing muscle required for complex cognitive functioning. At some point in human ontogeny, however, structural constraints may have limited peramorphosis, preventing further increases to brain volume. This obstacle may have been superseded when the developmental trajectories of the face and body (including the braincase) dissociated, with the face becoming relatively underdeveloped and the latter relatively overdeveloped. Thus, facial paedomorphosis may have been a necessary antecedent in attaining supplementary increases to brain capacity in human evolution and, by extension, higher cognitive functioning. Consequently, the extent of development observable in an individual's face may be an honest cue to his or her cognitive capacity or intellectual potential, and the allure of facial juvenilization may reflect people's preference for relatively intelligent mating partners. Heterochrony in Brain and Face Development Modern day humans possess a brain capacity that far exceeds that of any other primate in evolution. The human brain is three to four times larger than either the chimpanzee or gorilla brain (Jerison, 1973; Stephan, Fraham & Baron, 1981); and within the hominid line, brain size has increased steadily over the past three and a half million years, rising from a volume of 500cc in australopithecus afarensis to 1370cc in homo sapiens (McNamara, 1997). In no small measure, peramorphosis was responsible for this impressive increase. The appearance of a big brain in human evolution coincided with overdevelopment of the braincase and lengthening life-history stages. Obviously, a bigger braincase has a larger holding capacity, allowing for additional brain enlargement. But to fully develop the sophisticated information processing machinery that bigger brains afford, those stages in life critical for learning and psychological development also 42 required expansion (McKinney & McNamara, 1991). Structural constraints, however, are likely to have prevented parallel increases in cranial and facial growth, forcing the developmental trajectories of the two entities to dissociate, whereby the braincase became overdeveloped via peramorphosis and the face relatively under-developed through paedomorphosis (Gould, 1977). There is also good reason to suppose that natural selection may have favoured facial paedomorphosis directly. If further increases in the braincase became too costly, additional room for an expanding brain could be found by reducing the size of the face's features and lowering their position, thereby reducing the room necessary to support and maintain features, and making more efficient use ofthe skull's existing volume. Thus, just as prolonged life-history stages and an enlarged braincase were antecedents to brain expansion and cognitive prowess, so too might have facial paedomorphosis. Two lines of evidence lend support to this conjecture. The first establishes a link between facial underdevelopment and cranial capacity in the fossil record. Eruption of the first molar (and jaw development in general) is highly correlated with cranial capacity across all primate families, with later developmental onsets resulting from paedomorphosis predicting larger brain volumes. This relationship is strong enough that age at first tooth eruption is the gold standard for estimating brain size in the fossil record when incomplete skulls are recovered (McNamara, 1997). The second line of evidence comes from the simultaneous appearance of facial underdevelopment and cognitively demanding tasks in physical and cultural evolution. Higher cognitive functioning in humans reflects greater complexity in mental construction (Gibson, 2002). Humans are able to combine and recombine large sets of 43 objects or actions to create practically an infinite number of arrangements within a given framework. For instance, words can be integrated and reintegrated to create sentences that differ in meaning so long as said sentences adhere to certain language structures, such as the subject-verb-object format that English follows. "We will bring our enemies to justice" has a separate and distinct meaning from "We will bring justice to our enemies"1. Sentences, too, can be arranged and rearranged in a similar fashion to express an unlimited array of narratives, descriptions, and lines of arguments, as well as conjure up a plethora of ideas, connotations, and emotions. An analogous process is required for cooking and tool use. Different foodstuffs (potatoes, carrots, or asparagus) can be combined and recombined with various actions (peeling, chopping, mashing), tools (knife, food processor, grater), methods of cooking (baking, frying, steaming), and other ingredients to create any of nearly an infinite number of recipes. Similarly, raw materials can be worked and reworked into different patterns to create a variety of useful items. Hides and cordage can be weaved together to make clothing, or they can be intertwined with wooden shafts in a different configuration to make tents (Gibson, 2002). Although other primates are known to prepare food and use tools, they lack this ability to combine and recombine elements in original configurations to create new tools or cooking methods (Gibson, 2002). Coincidently, or perhaps not coincidently, the trend towards greater facial paedomorphosis in human evolution corresponds with the appearance of advances in both stone-age tool use (Klein, 1989; Wynn, Tierson & Palmer, 1996) and innovations in cooking (Brace, 1991). 1 Original quotation: "Whether we bring our enemies to justice or bring justice to our enemies, justice will be done." George W. Bush, Sept 20, 2001. 44 Brain Size and Cognition Implicit in the big brain hypothesis is a causal link between brain size and cognition. Strictly speaking, a larger brain does not necessarily mean greater intelligence. The average weight of an Asian elephant brain measures approximately 7500 grams, or five times heavier than a human brain. Dolphins and killer whales also have larger brains overall when compared to humans. Direct comparisons of gross brain volume between distantly related species are inappropriate because they fail to account for differences in body size (Kuhlenbeck, 1973). Somatic regions with greater sensory perception and motor agility require greater proportions of the brain for their maintenance and operation. Thus, in stark contrast to the human brain, which has a relatively large proportion dedicated to complex mental construction, a large proportion of the pachydermatous brain is designed to control its enormous, highly sensitive, and dexterous proboscis. An alternative means is to base the comparisons on ratios of brain size-to-body weight. However, use of this method still leads to erroneous conclusions. For example, the brain composes 2.1% of a human's total mass, whereas the brain of a mouse accounts for 3.2% of its total body weight. Again, the information processing capabilities of mice pale in comparison to the mental tasks capable of being performed by human cognition. Using brain-to-body size ratios to compare distantly related species has two critical shortcomings. First, they fail to distinguish between those parts of the brain responsible for basic functioning and survival and those responsible for higher cognitive functioning. And second, they do little to inform us as to the number and complexity of the brain's neuronal connections (Kuhlenbeck, 1973). For these reasons, it is better to restrict 45 interspecies comparisons to close relatives of humans—the pongids or other, extinct hominids—when studying brain size as it relates to intelligence. A larger brain size provides the additional neurons and greater neuronal complexity that would be required to boost cognitive performance. However, not all structures in the brain contribute equally to intelligence. If intelligence were the ability to combine and recombine elements into original configurations, the cognitive tasks associated with this ability would take place in the neocortex. However, certain subcortical structures are also necessary since elements often include a motor component. For example, one of the earliest known stone tools was manufactured through a technique called knapping, in which sharp flakes that could be used like a knife were sheared off a rock core by striking it at the proper angle with another stone. Knapping requires a number of cognitive abilities—including spatial perception to provide a frame of reference for the objects involved, mental rotation to orient these objects properly to find the precise angle at which to strike the core, and spatial visualization to picture the flake embedded in the core before it has been struck (Wynn, Tierson, & Palmer, 1996). Knapping also requires considerable strength and motor control to apply the stone to the core at the exact point with the necessary force and at the correct orientation to shear off a flake of the desired shape and size. The cerebellum is responsible for smoothing fine motor sequences—such as those used in knapping, speech production, and writing—and functions together with the basal ganglia during procedural learning to automate these motor procedures. Although it is evident that brain size has increased over evolutionary time within the hominid line, it is impossible to compare the relative sizes of individual structures 46 since the relevant tissue is absent from the fossil record. However, comparisons with analogous structures in our closest relatives, the pongids, indicate that enlargement was not uniform throughout the human brain, but rather concentrated in the neocortex, cerebellum, and basal ganglia, which are two to three times larger in the human brain (Jerison, 1973; Stephan, Fraham & Baron, 1981). Compared to chimpanzees and gorillas, the neocortex of humans has grown proportionately larger than the cerebellum, which has, in turn, grown proportionately larger than most other subcortical structures (Gibson, 2002). As a result, pongids lack the motor control necessary for human-like speech or hand-eye coordination. Although apes are capable of uttering an inventory of vowel sounds comparable to some human languages (Snowden, 1999), they are unable to construct unique sound sequences by combining different motor patterns in a rapid and smooth manner that would resemble human-like speech (Gibson, 2002). Similarly, although chimpanzees can be trained in the art of knapping, their blows lack the control and precision that would allow them to duplicate the results achieved by hominid knappers (Toth et. al., 1993). Thus, it appears that enlargement of the brain in human evolution was primarily the result of peramorphosis acting on those structures responsible for cognitive functioning. Intelligence in Mate Selection Intelligence can provide multiple fitness enhancing benefits to its possessor, including better problem solving, improved interpersonal and parenting skills, and better retention of cultural knowledge (Barkow, 1989; Buss, 1994). Intelligence may be particularly pertinent with regards to resource acquisition and status in men. As a study done by Jencks (1979) found, intelligent men advance rapidly up social hierarchies and 47 tend to earn more money in the United States. Human ancestors who preferred to mate with intelligent partners would have shared in these benefits and would have simultaneously increased the fitness of their offspring who would stand to inherit half of his or her genes for intelligence from the other parent. Both men and women place a premium on intelligence when choosing a long-term partner for marriage, and in most cultures this preference is equally strong in both sexes. Where differences do occur, women tend to show a slightly stronger preference for intelligence in their male partners (Buss, 1989). Given that facial paedomorphosis may have contributed to cognitive development through further brain expansion, facial growth may be an accurate indicator of intellectual aptitude. Kozeny (1968) found the faces of intelligent women possessed wider (larger) eyes, a smaller mouth, and a larger forehead—features consistent with underdevelopment. In an effort to replicate these results, another research group administered an "Intelligence Structure Test" to a group of female students in Germany, selecting the 14 highest and 11 lowest scoring women to act as comparison groups. Investigators measured the faces of these women and reported on four facial features that were of particular interest to them. Results indicated that the more intelligent women had significantly larger eyes and larger foreheads, consistent with a juvenile appearance. Although they did not have fuller lips, they did have narrower mouths, which would produce a smaller lip width-to-height ratio, also indicative of a juvenile appearance. On the other hand, the more intelligent women also had wider jaws at the level of the mandible and longer "lower faces". Whether this indicates that the women had more developed jaws (a mature trait) or a more rounded face (a juvenile trait) is unclear from 48 the description of the study (Kiener & Keiper, 1979). Although modest in their support, the results of these two studies seem to suggest that at least some aspects of facial underdevelopment may serve as honest signals to intelligence, at least in Caucasian female targets. Whether or not people actually attend to morphological correlates of brainpower within a mating context is not clear. At least three meta-analytic studies have investigated the relationship between physical attractiveness and intelligence with mixed results. Feingold (1992) found 22 studies and concluded that there was no relationship between physical attractiveness and cognitive ability measured by either intelligence or academic performance. Langlois, Kalakanis, Rubenstein, Larson, Hallam, and Smoot (2000) found a modest relationship between intelligence and attractiveness for adults, and a much larger effect for children. Jackson, Hunter, and Hodge (1995) found a positive relationship between competence and attractiveness that was again stronger in children than in adults. In a recent empirical investigation, attractiveness was found to be a better predictor of IQ than perceived intelligence. Targets were individuals born between 1920 and 1922 in Oakland, California, who participated in the Intergenerational Studies of Development and Aging (IGS) at the University of California, Berkeley. Only individuals for whom there was a photo and IQ score available were included in the study. Photos were selected at different stages of the individual's life—including childhood, puberty, adolescence, middle and late adulthood. Undergraduate judges rated each photo on its attractiveness, symmetry, averageness, and perceived intelligence. Attractiveness was a significant predictor of IQ at every age between childhood and middle adulthood with correlations ranging from (.16) to (.26). Furthermore, perceptions 49 of intelligence seem to be moderated by physical attraction to the individual—partial correlations between IQ and perceived intelligence controlling for attractiveness were not significant (Zebrowitz, Hall, Murphy & Rhodes, 2002). Thus, intelligent people are attractive, and attractive faces are perceived to be intelligent. Unfortunately, it is impossible to determine whether facial underdevelopment was responsible for the observed effect since it was not included as a variable in the study. However, given that the effect of attractiveness could not be explained by either symmetry, which only correlated with IQ during childhood, or averageness, which only correlated with IQ during puberty, it is entirely possible that facial juvenilization was responsible. Summary & Predictions Facial paedomorphosis may have contributed to the evolution of human cognition as a necessary requirement for further enlargement of the brain, thereby increasing neuronal density and complexity and enhancing processing capabilities. Both facial underdevelopment and attractiveness correlate with cognitive performance in modern human populations, suggesting that reduced development in the face offers an honest cue to intelligence. Given that intelligence is itself a central criterion in mate choice, particularly for long-term partners, facial cues to underdevelopment may be exploited by men and women to evaluate potential mates. Although this hypothesis does not propose that facial paedomorphosis originates within mate selection, it would nonetheless predict the preference for facial juvenilization to be relatively stronger for judgments made within a mating context. Greater intelligence should have benefited both male and female fitness in the ancestral environment and, consequently, both sexes should have valued high intelligence 50 in long-term mating partners. Both sexes rated intelligence as critically important in Buss's (1989) cross-cultural survey of mating preferences. And compared to other attributes that were considered in the study, intelligence was consistently ranked in the top third across all cultures. In most cases, there was no sex difference in the desire for an intelligent partner, but where differences did occur, it was the women who placed greater emphasis. This disparity becomes more apparent when constraints are placed on the respondents' choices so that they must decide between attributes that are necessities and those that are luxuries. Li, Kenrick, Bailey and Linsenmeier (2002) allotted men and women "mate dollars" to design their ideal mating partners. When their budget was restricted so that they could only afford those qualities considered necessities, women spent the largest proportion of their budget on intelligence, whereas men allotted a greater proportion of their budget on physical attractiveness and comparably less on intelligence (although intelligence was also highly valued). Given that women place relatively greater emphasis on cerebral aptitude in mating partners, the preference for facial underdevelopment is predicted to be greater for women judging male faces. Finally, given that the importance of intelligence as a criterion for mate choice is relatively stronger for marriage rather than for opportunistic mating, we would predict a stronger preference for facial juvenilization in long-term mating partners. 1.4 Summary and Plan of Action Morphological features associated with health, fertility, and genetic quality are often found physically attractive. Several empirical investigations have identified a preference for facial juvenilization in human faces. Together with the fact that facial paedomorphosis has been a hallmark in human ontogeny suggests, ostensibly, that facial 51 underdevelopment may have afforded fitness enhancing benefits that were favoured by natural selection . Three hypotheses were presented on the potential adaptiveness of facial paedomorphosis. The first proposed that facial underdevelopment exploited an existing facial recognition system that was originally designed for infant care to elicit ancillary altruism and to inhibit aggression. In the second, facial underdevelopment was explained as a by-product of (primarily) a male preference for youthful mating partners who are generally more fertile and have greater residual reproductive capacity. The third suggested that facial paedomorphosis was a necessary corollary to brain expansion and, in so doing, was one of multiple antecedents to human cognition. Each hypothesis generated a unique set of falsifiable predictions (summarized in Table 1.1). Both the youthful mimic and big brain hypotheses predicted a stronger preference for facial juvenilization in a mating context, whereas the baby-face overgeneralization hypothesis predicted a stronger preference when the context was platonic. In addition, men were predicted to exhibit a stronger preference for juvenilization by the youthful mimic hypothesis, whereas the baby-face overgeneralization and big brain hypotheses predicted women to show the stronger effect. Finally, the youthful mimic hypothesis predicted a stronger preference for juvenilization in short-term mating, whereas the big brain hypothesis predicted a stronger preference in long-term mating. Thus, in total, two predictions were proposed to falsify the baby-faced hypothesis, and three each to falsify the youthfulness and big-brain hypotheses. 52 Table Ll Summary of Hypotheses Hypothesis Baby-face Youthfulness Big Brain Context Non-sexual context higher Mating context higher Mating context higher Sex Difference Women judging men higher Men judging women higher Women judging men higher Mating Strategy None Short-term higher Long-term higher Two experiments were designed to test these predictions. Facial composites were constructed by way of morphing software and manipulated using Adobe Photoshop to vary the amount of facial juvenilization. Judges then ranked faces on a number of dimensions that varied in their relevance to a mating context and short- and long-term mating strategies. In experiment one, a between subjects design was employed whereby participants were randomly assigned to rank faces on either mating-relevant items or on mating-independent items. A repeated measures design was utilized in experiment two in an effort to replicate the results of experiment one, and additional mating-independent items were included to control for confounding factors that might have explained the results of experiment one. 53 Chapter 2 - Experiment One 2.1 Introduction Each hypothesis in chapter one made a set of predictions with regards to the effect of juvenilization on facial attractiveness. The baby-faced overgeneralization hypothesis predicted that the preference for facial underdevelopment would be stronger for platonic relationships outside of a mating context and would be stronger for female judges. The youthful mimicry hypothesis predicted that the preference for facial underdevelopment would be stronger for sexual relationships in a mating context, would be stronger for men judging female faces, and would be stronger when evaluating short-term mating partners. Finally, the big brain hypothesis predicted a stronger preference for facial underdevelopment in a mating context, for women judging male faces, and for long-term mating partners. To test these predictions, participants rank ordered a continuum of opposite-sex faces varying in facial development on various dimensions related to mating, arranging them from the most to least representative. Some items were indicative of a long-term mating strategy while others suggested a short-term strategy. If the preference for facial underdevelopment is specific to a mating context, then juvenilized faces should be ranked higher than chance alone would predict, since all the items reflect some aspect of mating. Conversely, if facial underdevelopment is irrelevant to mating interests, then no effect for juvenilization should be evident, and rankings should be random. Moreover, if the preference for facial underdevelopment is relatively stronger for short-term or long-term mating, then the juvenilized face should be ranked higher on items relevant to that mating 54 strategy. Finally, the third prediction in the set of three can be tested by comparing the responses provided by male and female participants. However, from previous studies, we know that juvenilized faces tend to be rated as physically attractive, and we also know from the halo effect that attractive faces tend to be perceived positively on a wide variety of dimensions, including those related to mating preferences (Dion, Berscheid, & Walster, 1972). Thus, juvenilized faces may be ranked higher on mating items simply because of their desirability, and not because they are germane to mating. Therefore, to unconfound juvenilization from the halo effect, a between-subjects design was used, whereby a separate sample of judges ranked the same faces on mating-neutral items. These items were meant to be desirable in general, but not necessarily relevant to mating interests in particular. Thus, the applicability of the items to mating interests in the mating condition created a context in which mating was at the forefront, whereas no such context was present in the mating-neutral condition. If the preference for facial underdevelopment is truly indigenous to mating, juvenilized faces should be ranked (a) higher than chance alone would predict in the mating condition, and (b) higher in the mating condition than in the mating-neutral condition. The relationship between an individual's preference for juvenilization and his or her socio-sexual orientation provides an alternative means of testing predictions related to mating strategy. Items on the Socio-sexual Orientation Inventory (SOI) measure behaviour and attitudes associated with short-term sexual relationships. Individuals scoring high on the scale are described as having an unrestricted orientation towards casual sex, while those scoring low are described as restricted in their orientation. Unrestricted individuals tend to report more sexual partners overall and more one-night 55 stands. Their attitudes towards sex outside of a committed relationship are less conservative, and they tend to fantasize more often about sex with people who are not their romantic partners (Simpson & Gangestad, 1991). If facial underdevelopment is preferred in long-term committed relationships, then judges with a restricted orientation towards casual sex should exhibit a stronger preference for juvenilization. If, however, facial underdevelopment is more indicative of short-term opportunistic mating, then judges with an unrestricted orientation should express a stronger juvenilization preference. Finally, correlations between attractiveness and other mating items may help to confirm or eliminate one or more hypotheses. The youthful mimicry hypothesis explains the preference for facial underdevelopment as a by-product of the male preference for younger partners. Facial paedomorphosis, in return, was favoured by natural selection because individuals with truncated development were perceived to be more youthful than their more developed cohorts. Hence, perceptions of youth are at the core of the preference for facial juvenilization. Therefore, we might expect a significant association between rankings of attractiveness and rankings of youthfulness provided by male, but not necessarily female, judges. Men who rank juvenilized faces high on attractiveness should also rank them high on youthfulness. The same argument does not apply to the big brain hypothesis, which explicates the preference for facial underdevelopment as a function of brain size and intelligence. While it is entirely possible that judges who rank juvenilized faces high on attractiveness will also rank them high on intelligence, it is by no means a necessary stipulation of this hypothesis. Just as an awareness of the link between symmetry and developmental stability is unnecessary to appreciate faces with 56 balanced proportions, a conscious awareness of the link between facial underdevelopment and intelligence (assuming there is one) may be equally superfluous. Human psychology is more likely designed to recognize high fitness in potential mating partners rather than to explain the origin of an attractive feature's appeal. Of course this counter-argument may also bear some relevance on the youthful mimic hypothesis, and the baby-faced overgeneralization hypothesis as well. It is possible that individuals appreciate features that mimic youth without being consciously aware of the reason. Thus, by themselves, these correlational predictions are a relatively weak test of the youthful mimic hypothesis. Nevertheless, they may provide some interesting insight. 2.2 Methods 2.2.1 Subjects Judges were drawn from the UBC Psychology Department subject pool and consisted of 30 male and 50 female undergraduates who received course credit for their participation. The sample included 34 Caucasians, 31 Asians, and 15 other individuals. Forty participants each were randomly assigned to either the mating or mating-neutral condition (15 male and 25 female each). The average age of the sample was 22.3 years. 2.2.2 Stimuli Targets were also drawn from the psychology subject pool but were recruited following completion of an earlier unrelated experiment. Thirty-two men and 32 women (16 Asian and 16 European of each sex) agreed to pose for a picture. Photos were taken using a three mega-pixel digital camera in front of a plain coloured wall under artificial lighting conditions. Targets tied their hair back away from their face, wore plain white t-shirts, and removed any jewellery they had been wearing. Female targets were asked to 57 remove any makeup and male targets were all clean shaven. Due to the poor florescent lighting at the time the photo was taken, the colour contrast and saturation were corrected using the "levels" and "variations" commands respectively available in Adobe Photoshop® 7.0 for Windows®. Standardized corrections were applied to all photos. Composite faces were created by combining 16 targets into the same image using Morph™ 2.0 for Windows®. The software uses anchor points and reference lines to link the features of two faces and generate a short movie file that depicts the gradual transformation of one face into the other. The middle frame of this movie depicts the mathematical average of the two faces. Using this method, 16 target faces were morphed into eight 2-face composites which were subsequently morphed into four 4-face composites, then two 8-face composites and finally a single 16-face composite. Four composites were made: An Asian male and female and an European male and female. These composites represent the mathematical average of the 16 targets used to construct them and estimate the population mean of their respective group. Features were manipulated using Adobe Photoshop® 7.0 for Windows® to either increase or decrease their resemblance to a juvenile appearance. Thus, in addition to the original mathematically average composite there was a juvenilized and adultified version of each feature. Three features were investigated in this experiment: eye-size, nose-size, and lip-height. To increase the juvenile appearance of these features, eye-area was increased by 10%, nose-width was decreased by 10%, and lip-height was increased by 10%. Manipulations in the opposite direction were used to compose the adultified versions of these features. Each feature was manipulated both separately and in 58 conjunction with the other two. Thus four sets were included in the experiment: eyes, nose, lips, and all three features combined (see Figure 2.1 for examples). 2.2.3 Procedure The judges' task was to rank order the adultified, average, and juvenilized features on ten personality traits, ranking the version most representative of the item first, and least representative last. Items in the mating condition suggested a mating context while those in the mating-neutral condition were extraneous to mating preferences. Trials were presented to participants via computer using a programme written in Visual Basic™ 4.0 for Windows®. Each trial presented an adultified, average, and juvenilized version of either an Asian or European opposite-sex face in random position combined with one of ten items. Thus, delivered in random order, there were two faces crossed with four feature sets and ten items for a total of 80 trials. Items in the mating condition were chosen from the literature on human mate preferences and were selected for the purpose of creating a mating salient context. Thus, these items were not meant to represent an exhaustive list of all possible considerations typically involved in mate selection. Items were presented in adjective form and included: attractive; healthy; honest; intelligent; warm personality; youthful; submissive (for female targets) or dominant (for male targets); good parent; good spouse; and good lover. Health, intelligence and personal warmth were selected because women tend to prefer men with these traits when looking for a mate. Good health has been rated by women from cross-cultural samples as anywhere from important to indispensable in potential marriage partners (Buss, 1989), intelligence has been ranked by women as one of the top five desirable traits in potential mating partners (Buss, 1994), and aspects of 59 EXAMPLES OF STIMULI GREATER JUVENILIZATION A V E R A G E LESS JUVENILIZATION Figure 2.1 Examples of the Caucasian male, and Asian female faces. The centre image is a 16-face composite estimating math-ematically average proportions of the population. Images to the left have been manipulated to contain greater juvenilization (facial underdevelopment). Images to the right contain less juvenilization (facial over-develoemnt). personal warmth, such as kindness and sincerity, are the single most sought after quality listed by women in lonely hearts advertisements (Harrison & Saeed, 1977). Likewise, men find women who are attractive and youthful highly desirable as mating partners (Buss, 1994; Kenrick & Keefe, 1992; Symons, 1979). Intelligence and personal warmth characterized two of three factors in an analysis of personality traits associated with mating preferences (Buss & Barnes, 1986). Women often use a man's honesty to judge whether his desire for a long-term relationship is sincere and, to attract women, men often combine honesty with cues of emotional attachment and commitment (Schmitt & Buss, 1996). Evidence of a mate's honesty may also assuage a man's concern for his partner's fidelity. Although unrelated to general likeability, dominant and submissive behaviour, when considered separately from related constructs such as aggression, have been linked to male and female sexual attractiveness respectively (Keating, 1985; Sadalla, Kenrick & Vershure, 1987). Finally, good "parent", "spouse", and "lover" were chosen by the experimenter for their face validity, and to provide a direct measure of the judges' partiality for facial juvenilization in short- and long-term romantic partners. If facial underdevelopment is pertinent to long-term committed relationships, then juvenilized faces should be ranked higher as potential spouses and parents. Conversely, if juvenilization functions within short-term opportunistic mating, then juvenilized faces should make more popular lovers. Furthermore, it should be noted that men also value women with good character and women appreciate male attractiveness. Where men and women tend to differ is in the relative importance they place on such attributes. Men tend to emphasize youthfulness and attractiveness whereas women tend to emphasize 61 good character (among other things). Thus, each of the 10 traits listed above should contribute to establishing a mating context for both sexes. Five of the items in the mating-neutral condition reflected generally positive traits that are not specifically related to mating - lucky, athletic, musical, spiritual, and funny. Four other items unrelated to mating were reverse scored so that low rankings would suggest greater positivity - messy, nervous, impatient, and critical. These last items were drawn from a big five questionnaire and, thus, reflect general aspects of human personality. The tenth item, emotional, was originally included with the reverse scored items but was later left positively scored. The reason for this was that judges seemed to vary in their interpretation of the item, some seeing it as positive (emotionally astute), others negative (overly emotional). The purpose of this condition was to present a list of items that were generally positive but not necessarily desirable for mating in particular. These items might be preferred for mating by some people (a highly devout individual may consider spirituality essential, for example), but there is no evidence that any of these items are ubiquitously sought after by most people. Again, these items are not meant to be inclusive of all auspicious traits that are neutral to mating interests. Evident in Figure 2.1 is the subtlety in which the adultified, average, and juvenilized versions differ from one another, making independent ratings impossible. Rankings were chosen in lieu of the more common paired comparisons format to: (a) reduce the number of trials, and (b) to see whether simultaneous viewing would help the participant discriminate between versions leading (perhaps) to a greater effect size. A difficulty arising from this format occurs at the beginning of each trial when the previous image is replaced. Because the new image is only marginally different, it creates an 62 illusion of movement similar to that used in a series of illustrations to create animation. This problem was resolved by alternating the Asian and European faces on consecutive trials. 2.2.4 Measures Before the trials began, all participants were prompted for their sex, age, and race. Female participants were also asked to indicate if they were pregnant (none were), if they were using a pill form of contraception, and how many days it had been since the end of their last menstrual cycle. A growing number of empirical investigations have demonstrated that female mate preferences, at least with regards to standards of male facial attractiveness, can vary as a function of the menstrual cycle (e.g., Johnston, Hagel, Franklin, Fink & Grammer, 2001; Penton-Voak et al., 1999). Data on the ovulatory status of female participants was collected to determine whether a similar effect would be observed for facial juvenilization. Following the trials, all participants were asked to indicate whether they were currently in a romantic relationship and, i f so, whether it was a committed or casual relationship and then given the Socio-sexual Orientation Inventory (SOI). After all trials and measures were completed, participants were debriefed and given course credit. 2.3 Results 2.3.1 Scoring A juvenilization score was computed for each trail by adding the number of juvenile versions that were ranked higher than relatively adult versions and subtracting the number of adult versions that were ranked higher than relatively juvenile versions. For instance, if the juvenilized version was ranked first, average second, and adultified 63 third, the score would sum to three (Juvenilized > Adultified; Juvenilized > Average; Average > Adultified). Thus each score can range from +3 (all juvenile comparisons higher) to -3 (all juvenile comparisons lower). If there is no effect of juvenilization, the mean score for the sample will be zero. If there is an effect for juvenilization, the mean score will be greater than zero. 2.3.2 Trait by Feature Interaction Before testing the predictions generated by the three hypotheses, a pair of two-way repeated measures A N O V A s tested the effects of feature manipulation by item in each condition with feature entered as a four-level within-groups factor and item entered as a ten-level within-groups factor. Because the items differed in each condition (either mating-relevant or mating-neutral), it was necessary to perform separate analyses for each rather than include condition as a two level between-subjects factor. The feature factor could be represented by a single set of scores i f it can be demonstrated that the effect of juvenilization is the same for each level of feature across all ten items or, in other words, if the interaction between feature and item fails to reach statistical significance. Furthermore, the set of scores could be drawn from the multi-feature face i f the effect of juvenilization can be demonstrated to increase linearly with the number of features altered or, in other words, if the main effect for feature is significant and the strongest effect is observed in the face where all features are manipulated jointly. This would dramatically simplify subsequent analyses. There was a significant main effect for item in both the mating condition (F^ 351) = 5.02, p < .001) and the mating-neutral condition ( F ^ s i ) = 5.02, p < .001). Tukey's HSD was calculated for each condition to identify significant differences. In the mating 64 condition (HSD = .49), juvenilization scores for honesty, attractiveness, and youthfulness were significantly greater than those for submissiveness, parenting skill, and warmth. Honesty was also significantly greater than intelligence. For the mating neutral condition (HSD = .57), nervous was significantly lower than all items except funny, which in turn was significantly lower than musical and messy. There was also a significant main effect for feature in both the mating (F(3i0i)* = 25.96, p < .001) and mating-neutral conditions (FQ, in) = 11.27, p < .001). Tukey's paired comparisons were hand calculated in the mating condition because sphericity had been violated, requiring a different error term for each comparison. Three homogeneous subsets were identified wherein the means for the multi-feature face (1.19) and the nose manipulation (0.94) were included in the highest subset. The second subset contained the nose and eye manipulations (0.69), while the final subset consisted of the mean for the lip manipulation (0.05). Sphericity was not violated in the mating-neutral condition so Tukey's HSD was calculated (HSD = .30). There were only two subsets, one containing the means for the multi-feature face (0.43), the eyes (0.40), and the nose (0.22), with the other containing the mean for the lips (-0.17). The mean for the multi-feature face was the highest in both conditions, suggesting that there was a linear effect of juvenilization whereby greater juvenilization lead to higher rankings, particularly in the mating condition. However, this mean was not significantly greater than that for the nose manipulation in either condition or the eye manipulation in the mating-neutral condition. The interaction between feature and item was not significant in either the mating (F(i5,570)* = 1.73, ns) or the mating-neutral ( F ^ 1053) = 1.71, ns) conditions. Therefore, * Assumption of sphericity was violated. Degrees of freedom reflect Greenhouse-Geisser adjustment. 65 the effect of each feature on rankings was independent of the item under consideration. From this we can conclude that that the relationship between juvenilization and items was global in nature, at least for the features and items used in the current investigation. Given that judgements of juvenilized features appear to be consistent across items, and the effect appears to be a function of the number of juvenilization manipulations included in the stimuli, data from the multi-feature face will be used as the dependent variable for the remaining analyses since it demonstrated the strongest effect. 2.3.3 Context, Sex, and Mating Strategy Two ANOVAs tested for interactions between the context of the judgement, the sex of the judge, and the relevance of the item to short- and long-term mating strategies. A two-way between groups ANOVA tested the effects of judge's sex (male or female) and context (mating or mating-neutral) whereby each was entered as a two-level between-subjects factors. Results are plotted in Figure 2.2. There was a significant main effect for context (F(i;76) = 23.20, p < .001), whereby juvenilization scores were significantly higher in the mating context (1.19) compared to the mating-neutral context (0.43). Thus, juvenilized faces were ranked higher when judgements were consistent with a mating decision. The main effect for sex was also significant (V(ij6)= 5.01, p < .05). Juvenilization scores based on male rankings of female faces (1.05) were higher than those for female rankings of male faces (0.66), indicating that the effect of juvenilization was stronger for male judges. The interaction between sex and context was not significant (F(i>76) = 2.78, p = .10). The difference between male and female juvenilization scores did not depend on the context of the judgement. 66 Mean Juvenilization Scores for Men and Women in each context 2i • Men • Women <u 1.2H © 0 5 0.8^ 0.4H 0 Mating Neutral Context Figure 2.2 Juvenilized faces were ranked higher when judgments reflected a mating context as opposed to a context void of mating. Men tended to rank juvenile faces higher than women, regardless of the context. A two-way between-within ANOVA tested the effects of judge's sex and mating strategy (short- or long-term), whereby strategy was entered as a two-level within subjects factor. Results are plotted in Figure 2.3. The main effect for sex was again significant (F(i>38) = 6.89, p < .05). Male juvenilization scores of female faces (1.60) were higher within the mating context compared to female scores of male faces (0.93), indicating again that the preference for juvenilization was stronger in male judges. The main effect for mating strategy was also significant (F(i;38) = 10.14, .p < 01). Juvenilization scores for items relevant to short-term mating (1.39) were higher than scores for long-term mating (0.99), suggesting that juvenile features may be more desirable in partners when pursuing an opportunistic mating strategy. The interaction between sex and strategy was not significant (F^g) < 1.00, p > .50). The difference 67 between male and female juvenilization scores did not depend on the item's relevance to short- or long-term mating strategies. Mean Juvenilization Scores for Short-term and Long-term Mating Items Broken down by Sex Male Sex • ST Mating • L T Mating Female Figure 2.3 Juvenilized faces were ranked higher on items related to a short-term mating strategy as opposed to a long-term strategy. This was true for both male and female judges. 2.3.4 Moderating Variables: Race, Romantic Relationship, and Ovulation In addition to the central predictions generated in chapter one, race, romantic relationship, and probability of ovulation were investigated to determine whether any of these factors qualified the effects in the preceding section. It is feasible that the judges' preference for juvenilization may have depended on either their race or the race of the target (or both), or whether they were currently in a romantic relationship, or, if female, whether they were currently ovulating. Unfortunately, with regards to this last factor, there were only seven women (three in the mating condition and four in the mating 68 neutral condition) who reported the end of their last menses had occurred between 6 and 12 days prior to their participation in the experiment, the window of time when ovulation is most likely to occur in women with regular 28-day cycles. This was an insufficient number to analyze the effect of ovulation on the preference for facial juvenilization. A two-way between subjects ANOVA was used to test the possibility that the preference for juvenilization was qualified by the judge's current relationship status. Context was entered as a two-level between-groups factor and the type of relationship (committed or no current relationship) was also entered as a two-level between-groups factor. Participants describing their romantic relationship as casual were excluded because there were too few to include as a third level of the relationship factor. One participant in a committed relationship was randomly eliminated from the mating condition and one who had no romantic relationship was eliminated from the mating-neutral condition to meet the requirement of proportional cell sizes, leaving 68 judges in the analysis. There was no significant effect for relationship type (F(i,64) < 1-00, p > .50) nor was the interaction between relationship type and context significant (F(i_64) < 1-00, p > .50). Regardless of the judges' relationship status, juvenilized faces were ranked higher in the mating context compared to the mating-neutral context. A three-way between-within ANOVA was used to test for the effect of race whereby context and the judge's race (Asian and Caucasian only) were entered as two-level between-subjects factors, and target's race (Asian and Caucasian) was entered as a two level within-subjects factor. Only Asian and Caucasian participants were included in the analysis because it was desirable to study own vs. other race comparisons. Three female Asian participants were randomly eliminated from the mating condition to meet 69 the requirement of proportional cell size. Thus, the total number of judges used in the analysis was 62. There was no main effect for judge's race (F^sg) < 1.00, p > .50) or for target's race (F(i;5g) < 1.00, p > .50), nor for any of the interactions involving the target or judge's race. Thus, juvenilization scores for both Asians and Caucasians were higher in the mating context regardless of whether the judgements being made were inter- or intra-racial. 2.3.5 Analysis of Individual Items The youthfulness and big brain hypotheses predicted that juvenilized faces would be ranked higher than chance when the item was related to mating. Table 2.1 lists the items for the mating and mating-neutral conditions in order of descending means. The scoring method explained previously assigned positive scores to rankings favouring juvenilization and negative scores to rankings favouring adultification. Thus, if there was no preference for juvenilization, positive and negative scores would cancel out and the mean will not differ significantly from zero. Reported t-tests are one-sample tests comparing the means to zero and the associated p-values are reported after Bonferroni correction for 20 tests. Effect sizes (d) were calculated by dividing the mean for each item by the pooled standard deviation for all items within its condition and are presented in the far right column. To calculate the mean effect size in the mating neutral condition, negative effect sizes were converted to zeroes since analyses reflect tests of directional hypotheses. In the mating condition, faces with juvenilized features were judged to be more attractive, youthful, honest, healthy, intelligent, and submissive, and to be better lovers, parents, and spouses compared to adultified features. They were not judged to have 70 warmer personalities however. In the mating-neutral condition, juvenilized faces were judged to be significantly less messy, more spiritual, and to have greater musical ability than adultified faces. No other traits reached statistical significance. Thus, when considering both conditions together, juvenilized faces were ranked higher than chance alone would predict for twelve of the twenty total items. However, when the two conditions were considered separately, juvenilized features were significantly associated with nine of ten mating items but only three of ten mating-neutral items. Furthermore, the average effect size of the mating items (d = .82) was more than two and half times larger than that observed for mating-neutral items (d = .30). Thus, not only were juvenilized features ranked higher than adultified features in a mating context, but rankings of juvenilized and adultified features reflected randomness when items were unrelated to mating. Although men ranked juvenilized female faces relatively higher on mating items in general, they did not rank them higher than women on attractiveness specifically. Male (2.06) and female (1.56) juvenilization scores for the attractiveness item were not significantly different (tpg) = 1.24, p = .22). Men did, however, find facial underdevelopment relatively more desirable when considering the faces for a romantic relationship. Men ranked juvenilized faces significantly higher than women for both a long-term spouse (t@8) = 2.11, p < .05), and for a short-term lover (tpg) = 2.33, p < .05). This is consistent with the youthfulness hypothesis, as was the overall correlation between attractiveness and youthfulness (.33), which was statistically significant. However, the youthfulness hypothesis also predicted that this relationship should be 71 stronger in male judges, but there was no difference between the coefficients observed for men (.28) and women (.31). Table 2.1 Means presented in descending order for items in the mating and mating-neutral conditions reporting one-sample t-tests and effect sizes. Mating Traits X Sx t(39) d Attractive 1.75 1.26 ***8.81 1.20 Youthful 1.65 1.41 ***7 42 1.13 Honest 1.45 1.41 ***6 49 .99 Healthy 1.35 1.27 ***6 71 .92 Sub/Dom 1.05 1.38 **4.82 .72 Lover 1.03 1.42 **4.56 .71 Spouse 1.00 1.62 **3.91 .68 Intelligent 0.93 1.70 *3.44 .64 Parent 0.90 1.55 *3.67 .62 Warmth 0.75 1.55 *3.06 .51 Mean 1.19 1.46 .82 SD 0.34 0.14 Neutral Traits X Sx t(39) d Messy (-) 1.00 1.48 **4.26 .60 Spiritual 0.88 1.52 *3.64 .53 Musical 0.83 1.58 *3.30 .50 Impatient (-) 0.63 1.89 2.09 .38 Lucky 0.53 1.57 2.12 .32 Emotional 0.50 1.78 1.77 .30 Talkative 0.48 1.45 2.07 .29 Critical (-) 0.05 1.88 0.17 .03 Funny -0.18 1.62 -0.69 .00 Nervous (-) -0.45 1.78 -1.60 .00 Mean 0.43 1.66 .30 SD 0.48 0.16 Note. Significance values for one-sample t-tests comparing the mean to zero are Bonferroni corrected for 20 tests. ***/?<.001. **/?<.01. *p<.05. 72 2.3.6 Socio-sexual Orientation The Socio-sexual Orientation Inventory (SOI) measures behaviour and attitudes towards short-term sexual relationships (Simpson & Gangestad, 1991). High scores reflect less restricted behaviour and attitudes towards casual sex wherein low scores reflect greater restriction. To examine the relationship between socio-sexuality and judgements of juvenilization, scores on the SOI were correlated with juvenilization scores for the attractiveness item. This restricted the analysis to the forty individuals in the mating condition (25 women and 15 men). Because mean differences between groups can greatly affect the correlation coefficient when groups are pooled, and because substantial sex differences exist when endorsing items on the SOI, coefficients were calculated separately for men and women. Although juvenilization correlated with an unrestricted orientation in women (r = .28) and a restricted orientation in men (r = -.33), neither of these coefficients reached statistical significance, possibly due to the small sample size. Thus, these results cannot ascertain the nature of the relationship between the preference for facial underdevelopment and socio-sexual orientation. 2.4 Discussion Both the youthfulness and big brain hypotheses predicted a stronger relationship between facial underdevelopment and sexual desire, whereas the baby-face overgeneralization hypothesis explained the preference for underdeveloped features as a by-product of a non-sexual attraction to infants. Results of the experiment indicated that the preference for facial underdevelopment was indeed germane to a mating context. Juvenilized faces were ranked higher in the mating condition compared to the mating-73 neutral condition, and juvenilized faces were ranked statistically higher than adultified faces on nine of ten items involved in mate selection. Although this was true of three items in the mating-neutral condition, juvenile features were not generally ranked higher when items were unrelated or only weakly related to mating preferences, and the average effect size for mating items was more than two and a half times larger than that observed for mating-neutral items. Moreover, this effect was not qualified by race, or by the relationship status of the judge - although the test involving the relationship factor was incomplete given that only two of the three relationship types could be included in the analysis. A similar problem with inadequate sample size prevented any testing of a potentially mediating effect associated with the follicular phase of the menstrual cycle. The big brain hypothesis predicted a sex difference in the desirability of facial underdevelopment whereby women would show superior potency. Conversely, the youthfulness hypothesis predicted that men would find facial underdevelopment more desirable than women, particularly within a mating context. The conclusions regarding sex differences in the data are mixed. On the one hand, juvenilization scores for attractiveness were not significantly different between men and women, suggesting that both sexes find juvenile features equally appealing. On the other hand, scores were significantly greater for men when judging the target in terms of a potential spouse or lover, suggesting that men find juvenilization more sexually alluring. However, it does not appear that this difference was driven by perceptions of youthfulness since there was no significant difference between men and women in how youthful they ranked juvenilized faces, nor was there any difference in the correlation coefficients measuring the relationship between youthfulness and desirability as a lover (rmen = .31; rw o m en = -28) 74 or between youthfulness and desirability as a spouse (rmen = -12; r w o men = .16). For both men and women, the higher they ranked juvenilized faces on youthfulness, the higher they ranked them as desirable lovers and, to a lesser degree, spouses. The baby-face overgeneralization hypothesis did not predict a difference in the juvenilization preference based on mating strategy since the appeal of facial underdevelopment occurs outside of a mating context. The youthfulness and big brain hypotheses predicted juvenilization would be more desirable in short- and long-term mating partners respectively. Consistent with the youthfulness hypothesis, juvenilized faces were ranked relatively higher for items describing a short-term mating strategy. Unfortunately, it was not possible to provide a further, unequivocal test of this prediction using the socio-sexual orientation data due to an inadequate sample size, particularly for male judges. However, if facial underdevelopment is vital for a short-term mating strategy, unrestricted orientations as measured by the SOI should be associated with a stronger preference for facial juvenilization. The general trend of the coefficients indicated that the correlation for female judges was consistent with this prediction, but the correlation for male judges was not. This issue will be revisited in study two, which utilizes a stronger sample size. Facial underdevelopment is sexually attractive to both sexes, but perhaps more so for men judging potential female partners, and seems to be most desirable when the strategy reflects short-term mating. Overall, the results of experiment one are most consistent with the youthfulness hypothesis while providing only partial support for the big brain hypothesis, and conflicting with the baby-face overgeneralization hypothesis entirely. However, depending on which elements of the data on sex differences one 75 focused on, support could be garnered for predictions proposed by either the youthfulness or big brain hypotheses. Men ranked juvenilized features higher for potential romantic partners, but both sexes found juvenilization equally attractive in general. Thus, the results do not clearly favour the prediction of one hypothesis over its alternative. Furthermore, although consistent with the youthfulness hypothesis, the stronger preference observed for facial underdevelopment in short-term mating may have been an artefact of small sample size. The analysis focusing on mating strategy included only half the total sample - those judges who participated in the mating condition - and included only 15 male judges. It is very possible that the sample was unrepresentative, and that the juvenilization preference would be stronger in long-term mating if a larger sample were utilized. Sample size also appears to have been responsible for a number of non-significant results involving race, relationship, ovulatory status, and socio-sexual orientation. Race was demonstrated to qualify the preference for juvenilization in Wehr et. al. (2001), but the interaction was a good deal smaller than the main effect observed for juvenilization. The sample size used in the current experiment may have lacked the power to detect it. Another potential flaw in the design of experiment one entailed the items chosen for the mating-neutral condition, which may have lacked the same level of salience, or desirability, that items in the mating condition possessed. For instance, the frequency in which people normally judge someone's luck, spirituality, or musical talent probably pales in comparison to how often they judge someone's health, intelligence or attractiveness as a mating partner. Alternatively, perhaps health, and intelligence are more desirable dimensions than luck or musical ability. Either way, the amount of 76 vigilance a judge was likely to commit to each judgement may have differed between the two conditions. Choosing a high quality mating partner, after all, is a far more interesting task, and carries far more serious consequences for one's future. To resolve these issues, a second experiment was designed to replicate the results using a larger sample and slightly different methodology. 77 Chapter 3 - Experiment Two 3.1 Introduction The goal of experiment two was to replicate the results of the foregoing experiment using a slightly different methodology and utilizing a larger sample size. To this end, a within subjects design was employed whereby items varied in their relevance to mating as opposed to providing a definitive mating context. Additionally, a paired comparisons format was used instead of rankings. As was the case in experiment one, the objective of the second study was to test the predictions generated from the baby-face, youthful mimic, and big brain hypotheses. Namely, to evaluate the preference for facial underdevelopment as a function of context, sex of the judge, and mating strategy. Five item sets varying in their relevance to mating replaced the mating and mating-neutral contexts from experiment one. Two of the sets were applicable to short-and long-term mating strategies respectively. Two additional sets of items were included to control for the halo effect, one composed of items from experiment one and the other consisting of the big five personality factors. A fifth set of baby-face items was drawn from Zebrowitz- McArthur and Apatow (1983-1984) and included to further investigate the baby-face overgeneralization hypothesis. If the preference for underdeveloped faces is a by-product of perceptions associated with infant faces, then juvenilization should be selected more frequently for items that are characteristic of infants than for control items, which do not reflect infantile qualities. In addition to paired comparisons, item correlations with attractiveness offer an alternative means in which to test the relevance of facial underdevelopment to mating and sexual attraction. If juvenilization is associated with mating interests, the coefficients 78 between attractiveness and other mating-relevant items should be both statistically significant and higher than those observed between attractiveness and control items. Furthermore, if the relationships between attractiveness and the other mating items are mediated by perceptions of youth or, in other words, the preference for facial underdevelopment is a product of the youthful mimic hypothesis, correlations between attractiveness and other mating items should decrease after the covariation due to youthfulness has been partialed out. This is assuming, of course, that judges are aware of the reason why they find juvenilized features attractive. Hence, these partial correlations do not provide a strong test of the youthful mimicry hypothesis itself, but only serve to supplement the analysis in determining whether the preference for youth is deliberate. Before continuing with the principal analysis of the experiment, the issue of item desirability will be addressed. A separate sample of individuals rated each item for its appeal in four types of relationships: two related to mating (sexual encounter and committed relationship), and two unrelated to mating (a co-worker and a close friend). If the preference for underdevelopment within a mating context is due to either the salience or desirability of the items themselves, then mating items should be rated equally advantageous for both romantic relationships and close personal non-sexual relationships. Furthermore, they should be rated as more desirable than platonic items overall. If facial underdevelopment is truly related to a context of mating, then the desirability of mating items should be specific to sexual relationships, and should be rated as no more important than non-mating items for platonic relationships. A second objective of this last task was to verify that short- and long-term mating items were truly indicative of short- and long-term mating strategies. 79 3.2 Item Selection and Desirability 3.2.1 Item Selection Senior undergraduate and graduate students from the psychology department at UBC rated a list of 32 items on a five-point scale indicating how desirable each quality was in a person for four different relationship types: a short-term partner, a long-term partner, a co-worker, and a close intimate friend (see appendix for a copy of the questionnaire). Half of the male and female participants rated the items for romantic relationships while the other half rated platonic relationships. The scale ranged from one (unimportant or undesirable) to three (nice but not necessary) to five (critically important). Item ratings for the romantic relationships were used to select the items that would later be used as the short- and long-term items in the experiment. The top three items reported by both men and women for a short-term partner were sexy, attractive, and a skilled lover. Warmth was also in the top five for both sexes while health was important to women, and honesty and youthfulness were important to men. Despite high rankings for warmth and honesty, it was decided to place these items in the long-term mating set where they were rated as even more important, leaving the short-term mating set to include attractive, sexy, healthy, skilled lover, and youthfulness. Joining warmth and honesty in the long-term mating set were intelligence and quality as a spouse, which were ranked in the top five by both sexes. Parenting skill was also included based on its face validity despite the somewhat surprising fact that it was not rated as very important by either men or women. Also surprising was the fact that submissiveness/dominance was rated in the bottom half of items by both sexes for both 80 partner types. At least at a conscious level, submissive women and dominant men don't provide good options as mating partners for our sample. In addition to the short- and long-term mating items, three additional sets were included. Five items were recycled from experiment one to control for the halo effect. These included lucky, athletic, musical, spiritual, and funny. There was concern that these five items alone would be inadequate to control for the halo effect since they were fewer in number (half as many used in experiment one) and were very specific in scope. To remedy this, the big five factors of extraversion, agreeableness, conscientiousness, emotional stability, and openness were used to compose a second set of control items. The big five factors are more general in scope, defining the parameters of human personality. The final set was to include baby-face items. Some items associated with a baby's face were already in use as mating items, including honesty and warmth (Zebrowitz-McArthur & Apatow, 1983-1984). To prevent confounding with the mating items, baby-face items were chosen that did not carry a strong connotation of romantic attraction. Naivete, physical strength (reverse scored), sociability, and "baby-face" were selected from research on perceptions of babyish individuals (Zebrowitz-McArthur & Apatow, 1983-1984; Berry & Zebrowitz-McArthur, 1985). Because submissiveness was rated so low on our item desirability questionnaire and, since it is a quality associated with both infants and baby-faced adults, it was also included in the baby-face item set. 3.2.2 Item Desirability It is possible that the observed effects in experiment one were due primarily to differences in the desirability of the items, whereby mating items were consistently coveted by judges to a greater degree than mating-neutral items. If true, then mating 81 items should be rated as relatively more desirable for a variety of relationship types, both sexual and platonic. A three-way between-within ANOVA was used to analyze the item desirability data collected from the separate questionnaire to test the veracity of this possibility. Item set (short- and long-term mating, halo, big five, and baby-face) was entered as a five-level within-subjects factor, sex of the respondent (male and female) was entered as a two-level between-subjects factor, and relationship type (sexual and platonic) was entered as a two-level between-subjects factor. Desirability score was the dependent variable. Scores for opportunistic and committed romantic relationships were averaged, as were scores for co-worker and close friend, to create the two levels of the relationship factor. Both the main effects for item set (F(3jH9) = 179.06 , p < .001) and relationship type (F(i;44) = 29.33, p < .001) were significant. However, these effects were qualified by a significant interaction ( F ^ ^ ) = 62.23*, p < .001) depicted in Figure 3.1. Neither the main effect for sex nor any of the remaining interaction effects were significant. To break down the significant interaction, simple main effects compared sexual and platonic relationships at each level of item set. The simple main effect for short-term mating items was significant (F(i;n6) = 10.84, p < .001) indicating that short-term mating items were rated as more desirable for sexual relationships compared to platonic relationships. None of the other simple main effects were significant (all F(i; m) < 1 00, p > .50). Long-term mating, halo, big five, and baby-face items were equally desirable for both sexual and platonic relationships. In conclusion, item desirability in and of itself could not account for the effect of juvenilization observed for short-term mating items in experiment one since these items were relatively undesirable when the context lacked a * Assumption of sphericity was violated. Greenhouse-Geisser corrections to df are reported. 82 sexual reference. On the other hand, item desirability may explain the results observed for long-term mating items since the five mating-neutral items were rated as less desirable for both sexual and platonic relationships. Mean Desirability Scores of Items for Sexual Partners and Platonic Relationships Qi U O CZ3 ST Mating L T Mating Halo Item Set Big 5 Baby-face Figure 3.1 Big-five items are an appropriate control for the halo effect. The mean desirability of the big-five personality dimensions is equal to that observed for long-term mating items and short-term mating items in sexual relationships. As far as experiment two is concerned, the results from the item-desirability analysis recommend using the big five personality item set to control for the halo effect since the mean desirability for this set (3.81) was not significantly different than that observed for either the short-term (3.07) or long-term (3.91) mating sets. Given that the means for the original mating-neutral items from experiment one and the baby-face set were both rated lower in desirability, any results involving these items will necessarily be confounded with the alternative explanation of item-valence. Therefore, only data 83 involving the mating items and their appropriate control group, the big five personality items, will be included in the remaining analyses. 3.3 Methods 3.3.1 Subjects In addition to utilizing a more powerful statistical design, a much larger sample was recruited for experiment two to address issues of power that plagued some of the analyses in experiment one. Participants were again drawn from the departmental subject pool and consisted of 61 male and 167 female undergraduates who were given course credit for their participation. The sample included 62 Caucasians, 139 Asians, and 27 other individuals. The average age of the sample was 20 years. 3.3.2 Stimuli In experiment one, the nature of the relationship between juvenile features and mating items appeared to be both linear and global. The interaction between feature and item was not significant, and the main effect for feature was largest for the multi-feature face. Accordingly, the composites containing multiple feature manipulations were used exclusively in experiment two in order to generate the strongest possible effect. 3.3.3 Procedure and Measures The procedure and measures were the same as experiment one with the following exceptions: First, there were 25 judgement items instead of 10, divided into five sets: short- and long-term mating, the big five personality factors, the five mating-neutral items from experiment one, and a baby-face item set. Again, the analysis will focus on the first three sets listed of these five. Second, experiment two utilized a within-subjects design whereby judges responded to all 25 items from all 5 sets. Third, a paired 84 comparisons format was used in place of rankings. Thus, instead of presenting the adultified, average, and juvenilized versions together, the judges were presented with three individual comparisons (Adultified vs. Average, Adultified vs. Juvenilized, Average vs. Juvenilized) for each item. For each comparison, their task was to select the version that best represented the item. The total number of trials was equal to two faces (Asian and Caucasian) by three comparisons by 25 items for 150 trials. As in experiment one, judges completed the SOI during the final stage of experiment two. 3.4 Results 3.4.1 Scoring Scoring was identical to experiment one wherein the number of comparisons favouring a relatively adult face was subtracted from the number preferring a relatively juvenile face. For instance, if the juvenilized version was chosen over the average and the adultified versions and the average was chosen over the adultified, then the score would be 3 (juvenilized > adultified, juvenilized > average, average > adultified). Thus each score could again range from +3 (all juvenile comparisons higher) to -3 (all juvenile comparisons lower). Calculating the scores in this manner allowed a direct comparison of effect sizes based on rankings in experiment one and those based on the forced choice format used in experiment two. As before, the expected mean score for the sample will be zero if there is no preference for juvenilization, and greater than zero if juvenile faces are chosen more frequently. 3.4.2 Mating Salience, Judge's Sex, and Mating Strategy Two ANOVAs were designed to test the effects of mating context (or in this case, each item set's apparent relevance to mating), mating strategy, and sex of the judge. A 85 two-way between-within ANOVA tested the effects for judge's sex and the item's mating salience whereby sex (male and female) was entered as a two-level between-subjects factor, and mating salience (short-term, long-term, and big five) was entered as a three-level within-subjects factor. Results are plotted in Figure 3.2. The main effect for mating salience was significant (F(2,402) = 42.39*, p < .001) indicating that juvenilization effected perceptions associated with some item sets more than others. Tukey's paired comparisons were hand calculated because sphericity had been violated, requiring a different error term for each comparison. Two homogeneous subsets were identified wherein the mean for short-term items (1.25) was significantly greater than the means for long-term mating items (0.58) and the big five control items (0.66), which did not differ from each other. The main effect for sex was also significant (F(i> 226) = 17.26, p < .001) indicating that men tended to choose the juvenilized face (1.09) more often than women (0.57). As was the case in experiment one, the interaction between sex and mating salience was close to reaching significance, but failed to reach an alpha level of .05 (Fp, 402) - 2.73*, p = .07). Thus, the difference observed between men and women was similar across each item set: men more likely than women to choose the relatively juvenile face regardless of the item under consideration. Thus, the results indicate that the effect for facial underdevelopment was relatively stronger for mating-salient items, but only if they were associated with a short-term mating strategy, and was relatively stronger for male judges appraising female targets. * Assumption of sphericity was violated. Greenhouse-Geisser corrections to df are reported. 86 Mean Juvenilization Scores for Men and Women on Short- and Long-term Mating Items and Big Five Control Items 2n • Men • Women I.61 <u 1.2H u © M 0.8H 0.4H 0 STM L T M BIG5 Item Set Figure 3.2 Juvenilized faces were selected more frequently when items were relevant to short-term mating, and men tended to select juvenile faces more frequently than women. 3.4.3 Moderating Variables: Race, Romantic Relationship, and Ovulatory Status The potential mediating effects of race were assessed using a three-way between-within ANOVA wherein mating salience was entered as a three-level within-subjects factor, judge's race (Asian and Caucasian) was entered as a two-level between-subjects factor, and target's race (Asian and Caucasian) was entered as a two-level within-subjects factor. Again, the analysis was restricted to Asian and Caucasian judges in order to produce a fully crossed design wherein the means represented own- vs. other-race judgements. The main effect for target's race was statistically significant (F(i;i99) = 34.79, p < .001). Juvenile versions were selected more often for Asian targets (0.92) compared 87 to Caucasian targets (0.50). The main effect for judge's race was not significant (F(ij99) < 1.00, p > .50). Thus, there was no difference between Asian and Caucasian judges in how often they selected the juvenile version of a target. None of the two- or three-way interactions involving race were statistically significant, although the interaction between the judge's and target's race was close (F(i;i99) < 3.18, p = .08). The general trend reflected a tendency to select the relatively juvenile face even more frequently when the target was from one's own race. In sum, although juvenile faces were selected more frequently when the target was Asian, this was true regardless of the judge's race and was independent of the item's relevance to mating. To test the possibility that the judge's relationship was a mediating variable, a two-way between-within ANOVA was utilized wherein mating salience was entered as a three-level within-subjects factor, and relationship type (casual, committed, none) was entered as a three-level between-subjects factor. The main effect for relationship type was not significant (F(2,225) = 1.71, p = . 18). Juvenilization had the same effect on the judge regardless of whether he or she was currently single or involved in a casual or committed romantic relationship. The interaction between relationship and mating salience was also not significant (F o^o) < 1.00*, p > .50). Thus, the relative strength of the juvenilization effect in a mating context was neither contingent on being in a romantic relationship, nor on the relationship's involvement level. The effect of ovulation on the preference for facial juvenilization could not be tested in experiment one because there were only seven female participants whose responses indicated that they may be ovulating. Of the 167 women who participated in experiment two, 33 reported that the end of their last menses had ended between 6 and 12 * Assumption of sphericity was violated. Greenhouse-Geisser corrections to df are reported. 88 days prior to their participation, the window of time where ovulation is most likely to occur. To test the possibility that ovulation moderated the preference for juvenilization, a two-way between-within ANOVA was utilized wherein mating salience was entered as a three-level within-subjects factor, and ovulatory status was entered as a two-level between-subjects factor. Women whose cycle ended between six and 12 days prior were considered likely to be ovulating whereas women who were at some other point in their cycle were considered unlikely to be ovulating. Neither the main effect for ovulatory status (F(i;i65) < 1-00, p > .50), nor the interaction between mating salience and ovulatory status (F(2,288) = 1.51*, p = .23) were statistically significant, indicating that the tendency to select relatively juvenile faces was independent of the likelihood that a female judge was currently ovulating. 3.4.4 Analysis of Individual Items Scores significantly greater than zero indicate that the relatively juvenile face was selected more frequently, whereas scores close to zero suggest that judges' selections were random. Table 3.1 presents the means in descending order for each item set. T-tests are again one-sample tests comparing the means to zero and the associated p-values are reported after Bonferroni correction for 15 tests. Effect sizes were calculated by dividing the mean for each item by the pooled standard deviation for all items within its set. Juvenilization scores for nine of the ten mating items were significantly greater than zero, indicating that the relatively juvenilized face was selected more often than chance alone could explain. Surprisingly, intelligence was the only mating item where no effect of juvenilization was observed. Similarly, all big five items were significantly greater than zero, suggesting that some of the preference for juvenilization is actually an artefact of * Assumption of sphericity was violated. Greenhouse-Geisser corrections to df are reported. 89 the halo effect. Thus, juvenilized faces were chosen more often than adultified faces for a majority of the items in all three item sets. Table 3.1 Means presented in descending order for five item sets from experiment two accompanied by one-sample t-tests, effect sizes, zero andfirst order correlations with Attractiveness. First order correlations control for covariation due to youthfulness. STM Items X Sx t(227) d ro ri Attractive 1.45 1.52 * * * 14.41 .93 - -Youthful 1.36 1.51 * * *13 .60 .87 * * * 3 6 -Sexy 1.23 1.54 * * *12 .02 .79 * * * 65 * * * . 6 2 Lover 1.00 1.58 53 .64 * * * g j * * * . 55 Healthy 0.80 1.64 * * * 7 35 .51 * * * 24 * * * 3 j Mean 1.17 1.56 .75 .49 .49 LTM Items Spouse 0.69 1.60 * * *6 .54 .41 * * * 35 * * * 2 9 Honest 0.54 1.64 * * *4 .96 .32 .11 .03 Parent 0.43 1.74 .25 * * .25 ** .22 Warm 0.32 1.66 *2.95 .19 .15 .03 Intelligent 0.05 1.87 0.43 .03 .15 .15 Mean 0.41 1.70 .24 .20 .14 Big 5 Openness 0.66 1.68 * * * 5 92 .40 *.20 .09 Agreeable 0.64 1.71 * * *5 .62 .38 .17 .06 Emotionally Stable 0.58 1.66 * * * 5 27 .35 * * * 34 * * * . 30 Extraverted 0.47 1.76 * * * 4 03 .28 *.20 .13 Conscientious 0.41 1.54 ***3 QQ .25 .08 .03 Mean 0.55 1.67 .33 .20 .12 Grand Mean 0.71 1.64 .44 .28 .21 Note. Significance values for one-sample t-tests, zero order, and first order correlations are Bonferroni corrected for 15 tests. ***p<.00\. **/?<.01. * < .05. A comparison of the relative effect sizes, however, demonstrates clearly the superior effect associated with items from the short-term mating set (.75), where the average item effect size was three times larger than the that observed for long-term mating items (.24) and more than twice as large as the effect recorded for the big five control items (.33). Surprisingly, the modest effect size observed for long-term mating items was actually smaller than that observed for the control items. Thus, based on mean 90 juvenilization scores, it appears that the preference for facial underdevelopment within a mating context, over and above that which is due to a halo effect, occurs within a short-term mating strategy. Column five of Table 3.1 lists the zero order coefficients for each item correlated with attractiveness. If the preference for facial underdevelopment is truly germane to mating, then judges who selected the juvenilized face as being more attractive should also have selected the juvenilized face for other mating-relevant items, and should have done so more often than for the big five control items. All four of the remaining short-term mating items correlated significantly with attractiveness while only two of the long-term items (desirability as a spouse and skill as a parent) demonstrated a significant relationship, again suggesting that the preference for facial underdevelopment is specific to short-term mating. Judges who rated juvenilization high on attractiveness also rated them as sexier, healthier, more youthful looking, and to make better mating partners. Only three of the big-five items (openness, extraversion, and emotional stability) correlated significantly with attractiveness. Furthermore, the average coefficient for short-term items (.49) was more than twice the size as that for the other two groups. When sexy, youthful, desirability as a lover, and health were entered as predictors in a regressing analysis, they accounted for 54% of the total variation in the preference for juvenilization measured by attractiveness (F(4i 223) = 65.75, p < .001). Adding parenting skill, spousal quality, emotional stability, openness, and extraversion, to the regression equation only accounted for an additional 3% of the variation. Therefore, the correlational data seem to indicate that the preference for juvenilization is most closely associated with short-term, rather than long-term, mating. 91 Column six of Table 3.1 lists the first order coefficients for each item correlated with attractiveness controlling for the influence of youthfulness. If the preference for facial underdevelopment in short-term mating is primarily or exclusively the result of a preference for youthful sex partners and, furthermore, judges are aware of the potential relationship between morphology and youthfulness, then removing the covariation due to youthfulness should dramatically reduce the association between attractiveness and the other mating items. However, coefficients for the mating items decreased only marginally and none lost significance as a result of removing the covariation due to youthfulness. Youthfulness by itself accounted for 13% of the variation in attractiveness (r = .36) while the regression equation for all four short-term mating items together accounted for 54% of the variation. Therefore, the youthfulness hypothesis may not, by itself, entirely account for the preference for facial underdevelopment. At the very least, partial correlations suggest that the preference for youthful looking facial features is an unconscious appraisal. 3.4.5 Socio-sexual Orientation To examine the relationship between mating strategy and the preference for juvenilization, scores on the SOI were correlated with juvenilization scores corresponding with the attractiveness item. If facial underdevelopment is more desirable in short-term mating partners, those individuals with an unrestricted orientation should express a greater preference for facial juvenilization. Conversely, if facial underdevelopment is more desirable in long-term mating partners, the preference should be stronger for individuals with restricted orientations. Once again, coefficients were calculated separately for male and female judges. As was the case in experiment one, 92 neither the correlation for men (r = . 15) nor women (r = -.11) reached statistical significance. Thus, results from the SOI questionnaire again fail to establish any relationship between the preference for facial juvenilization and the pursuit of a short- or long-term mating strategy. 3.5 Discussion In experiment one, facial underdevelopment was demonstrated to show the largest effect in a mating context. Juvenilized faces were ranked higher than adultified faces when mating salient items were used to create a mating context. The effect of facial underdevelopment was also larger when mating reflected an opportunistic mating strategy. Juvenilized faces were ranked relatively higher on short-term, rather than long-term, mating items. And the largest effect was observed for male judges assessing female targets. Men ranked juvenilized features higher than women did across both mating and mating-neutral items. Overall, the results of experiment two replicated and extended these findings using a larger sample and somewhat different methodology. Similar to experiment one, the effect for facial underdevelopment in experiment two was stronger for items salient to mating. Juvenilized faces were selected over adultified faces more frequently when the item suggested a mating component. The effect for underdevelopment was larger for male judges and Asian targets. Male judges selected juvenilized faces more frequently than women, and both Asian and Caucasian judges selected juvenilized faces more frequently when the target was Asian - although they selected juvenilized faces more often than chance for Caucasian targets as well. These results cannot be explained by potential differences in the desirability of the mating and mating-neutral items since it was demonstrated by the item desirability 93 questionnaire that (a) the big five items used to control for the halo effect were at least as desirable as those related to mating, and (b) the desirability of short-term mating items was restricted exclusively to sexual relationships. In accordance with experiment one, the effect for facial underdevelopment was stronger when the mating decision reflected an opportunistic mating strategy. Juvenilized faces were selected more frequently when the item reflected short-term, rather than long-term, mating. The effect size for short-term mating was more than three times larger than the effect observed for long-term mating. Thus, the magnitude of the difference in juvenilization preferred for short- and long-term mating strategies was considerable. Furthermore, judges who preferred juvenilization were also more likely to rate juvenilized faces higher on other short-term mating items, rating them as more youthful, sexier, healthier, and making more desirable lovers. Although judges also rated juvenilized faces to make better parents, more desirable spouses, and to be more emotionally stable, the average coefficient for long-term mating and big five control items in general was less than half the size of the coefficients for short-term mating items. As was the case in experiment one, none of the analysis involving potential mediating variables reached statistical significance. Neither race, nor relationship status, nor ovulatory status had any effect on the preference for facial juvenilization in either experiment. While an inadequate sample size was suspected as a contributing factor to these null results in experiment one, the sample size in experiment two was more than ample. Furthermore, neither study was able to detect a significant interaction between the judge's sex and mating context (in experiment one) or mating salience (in experiment two). Given that the likelihood of a type I error in each case was only (p = .10) , this 94 might suggest that power to detect the interaction was lacking from the experimental design. However, experiment two employed a within-subjects design, which calculates the interaction effect using a smaller error term, thereby increasing power, and, again, the sample size utilized in experiment two should have been sufficient to detect a non-trivial sized effect. In sum, the interaction involving sex is either nonexistent, or it is not important. Contrary to the findings in experiment one, the results for long-term mating items in experiment two may best be explained by a halo effect. The effect for the big five personality dimensions and long-term mating items were comparable in magnitude, and their average correlation coefficient was identical. Part of the reason for this stems from the fact that the effect size for long-term items in experiment two was 45 points lower than it was in experiment one. Likewise, a decrease was observed for short-term items as well, whereby the average effect size fell 24 points between the first and second experiments. The relatively weaker effect for mating items in experiment two may have been a consequence of the within-subjects design. In experiment one, judges considered either mating items or mating-neutral items. Given that some mating items are relevant to platonic relationships as well, it is possible that, when these items were combined with mating-neutral items in the same task, their ability to trigger prototypical mate-selection responses was diluted. Personal warmth, for example, is desirable for both mating relationships but also friendships and professional relationships. When intelligence was combined with items more or less germane to romance, the context in which it was judged was probably one of mating, and this was reflected in the item's associated effect size (d = .64) in experiment one. However, when intelligence was pooled within a 95 motley of items in experiment two, the context forjudging intelligence may have been undefined, and the effect size (d = .03) decreased accordingly. The fact that larger declines in effect size were observed for mating items with greater generalizability to platonic relationships provides some support for this conjecture. The average decrease in effect size for warmth, intelligence, and honesty was (.53) points whereas the decline for desirability as a spouse or a lover was only (.17) points. It should also be noted that these relatively generalizable traits were concentrated in the long-term set of mating items, which could explain why the relative plunge in the average effect size was so much stronger for long-term mating. Thus, although within-subjects designs have greater power, the end result was to palliate the effect in the second experiment, particularly for items of the long-term mating strategy. Alternatively, or perhaps concordantly, the dissimilarity in effect sizes between the experiments may have been an outcome of their different methodologies. In experiment one, judges viewed all three alternatives in concert and ranked them from most-to-least indicative of the item under consideration. In experiment two, judges chose between two faces in a forced choice format, and comparisons of the three faces were spread over multiple trials and interrupted by comparisons between other items. The manipulations used to vary the amount of juvenilization were subtle and simultaneous viewing could have helped the participant discriminate between versions. It's impossible to determine whether the disparity in the effect sizes resulted from differences in design or methodology since the two were conflated in the experiments. Nevertheless, future investigations may want to consider using a between-subjects design and rankings in place of a paired-comparisons format. 96 One of the drawbacks to the small sample size in experiment one was that the seemingly non-trivial association recorded between socio-sexual orientation and the preference for facial juvenilization failed to reach statistical significance. These correlations were again not significant in experiment two. In fact, not only were the coefficients weaker in size, but the general trend of the relationship was the opposite in direction. In experiment one, it was women with unrestricted, and men with restricted, orientations who gave the highest rankings to juvenilized faces, but it was restricted women and unrestricted men in experiment two who demonstrated the stronger preference. Thus, there does not appear to be any reliable relationship between socio-sexual orientation and the preference for facial juvenilization. How this impacts the level of support found for the youthful mimic hypothesis is unclear. If facial underdevelopment is more important for short-term mating, as the youthfulness hypothesis predicts, then we might expect individuals with unrestricted orientations towards casual sex to evince a stronger preference for juvenilization. However, the fact that neither a restricted nor an unrestricted orientation seems to be associated with a juvenilization preference does not necessarily falsify the youthful mimic hypothesis. After all, youthfulness is only relatively more important for short-term mating, but is generally important for both long- and short-term strategies (Buunk, Dijkstra, Kenrick, & Warntjes, 2001; Kenrick & Keefe, 1992; Symons, 1979). A number of additional findings in the data were counterintuitive to the youthful mimic hypothesis. First, youthfulness by itself accounted for only thirteen percent of the total variation in attractiveness. Second, partial correlations between attractiveness and other mating items decreased only modestly after controlling for youthfulness, and none 97 of the coefficients lost statistical significance. Taken together, these first two points indicate that most of the covariation existing between attractiveness and mating is independent of youthfulness by itself. Third, although men found juvenilized faces more appealing as potential mating partners, they did not find juvenilization more attractive overall. There was no significant difference between male and female juvenilization scores for attractiveness (t(226) < 1-00, p > .50). Fourth, although marginally higher for men, correlations between the youthfulness item and attractiveness (r = .34), sexy (r = . 20), desirability as a spouse (r = .17), and lover (r = .38) were all statistically significant for female judges. Women who selected the juvenilized face as more youthful also selected it as being more attractive, sexier, and more desirable as a mating partner. Thus, the larger effect observed for male judges predicted by the youthful mimic hypothesis was not uniform throughout the data. Nevertheless, the overall basic findings of each experiment seem to support each other and the youthful mimic hypothesis. Perceptions associated with facial underdevelopment were stronger for mating dimensions, especially for short-term mating, and were stronger for male judges gauging female faces. Predictions generated from the big brain hypothesis proposed that juvenilization should have been selected more frequently for mating-relevant items. Although it was not necessary forjudges to consciously recognize a link between facial underdevelopment and cerebral fortitude, it seems likely that they should have at least selected the juvenilized face more frequently than the adultified face when the intelligence item was presented. They did not. The effect size for intelligence was practically zero, and there was no significant relationship between perceived intelligence and attractiveness. Judges who selected the juvenilized face as more intelligent did not 98 also select it as more attractive. Part of the explanation, no doubt, has to do with the generalizability of intelligence to platonic relationships, and the resulting dilution of the mating context that would ultimately weaken the effect for this item. Nonetheless, the decrease in effect size observed between experiment one and two for intelligence was more severe than other items in the same situation, such as warmth and honesty. Part of the reason why may have had to do with pooling the data across sex. The big brain hypothesis predicted a stronger preference for underdevelopment by female judges, who tend to place greater emphasis on intelligence in romantic relationships, but when juvenilization scores for men and women were compared on intelligence, the mean for male scores (0.75) exceeded that of female scores (-.20) (t(226) = 3.51, p < .001). Thus, there was indeed a sex difference in the data with regards to intelligence and underdevelopment, but it was men who perceived juvenilized faces as being more intelligent, not women. Since mean differences between groups can distort correlations, the coefficients involving perceptions of intelligence were recalculated for men and women separately. For male judges, the correlations between intelligence and attractiveness (r = .33), desirability as a spouse (r = .22), and as a lover (r = .24) were all significant. The only significant correlation for female judges was between intelligence and sexy (r = .17). Surprisingly, it was the men who selected juvenilized faces as appearing more intelligent who also found them more interesting sexual partners. This is consistent with previous findings that underdevelopment in female faces is related to actual intellectual performance (Kozeny, 1968; Kiener & Keiper, 1979). Unfortunately, to our knowledge, no attempt has been made to confirm a similar relationship in male faces. 99 Chapter 4 - General Discussion 4.1 Results of the Predictions Beginning sometime subsequent to the separation of the hominid and pongid lines, human facial morphology has followed (primarily) a paedomorphic ontogeny, (Brace, 1962; Parker & McKinney, 1999; Weidenreich, 1947). Psychological research has demonstrated what seems to be a parallel preference for underdevelopment in both male and female faces across a variety of cultures and utilizing a range of empirical techniques including (a) correlational studies (Cunningham, 1986; Cunningham, Barbee & Pike, 1990), (b) computer simulations of natural selection (Johnston & Franklin, 1993), and (c) digitally rendered manipulations to individual features (Jones, 1996; Wehr et. al., 2001). Results from the present investigation provide further support for the attractiveness of underdevelopment in human faces. Not only were they more attractive than adultified faces, but juvenilized faces were ranked higher, and selected more frequently, on dimensions related to mating in general - juvenilized faces were perceived to be healthier, sexier, more honest, and more youthful in appearance, and they were judged to make better parents and more desirable partners for long- and short-term sexual relationships. Species specific standards of physical attractiveness within mate selection can be conceptualized as an adaptive set of preferences for phenotypes advertising reproductive value, high fitness, or genomic robustness (Barber, 1995; Jones, 1996; Symons, 1995). Three theories were tendered to explain the apparent co-evolution linking facial paedomorphosis with the preference for facial underdevelopment in humans. Each theory was developed to generate a unique set of predictions dictating how the preference 100 for underdevelopment might vary as a function of (a) the context in which the judgement was encountered, (b) the sex of the individual making the judgment, and (c) the sort of mating strategy invoked by the judgement. In the baby-face over generalization hypothesis, facial underdevelopment was proposed to stimulate altruistic behaviour and inhibit aggression beyond childhood. According to this hypothesis, the preference for juvenilization should have reflected a non-sexual affinity similar to that reserved for infants. The central proposition of the youthful mimic hypothesis was that underdevelopment mimicked cues in the face that were normally associated with youth, fertility, and greater residual reproductive capacity. According to this hypothesis, the preference for juvenilization should have been strongest in male judges evaluating female targets as potential short-term mating partners. Finally, the big brain hypothesis surmised that facial underdevelopment evolved as a necessary antecedent to supplementary gains in brain volume, ultimately leading to human cognition. In this last case, the preference for juvenilization should have been most potent in female judges assessing male targets as potential long-term partners. With regards to the context of the judgement, the preference for facial underdevelopment was stronger when mating was salient. Compared to mating-neutral items, juvenilized faces were ranked higher, and chosen more frequently when assessing items applicable to mate selection. Additionally, the effect of a mating context on the preference for juvenilization appeared to be quite large. Effect sizes for mating salient items were two-to-three times larger than those observed for the mating-neutral items designed to control for the halo effect. This effect of context neither interacted with the sex of the judge, nor was it qualified by the judge's race, relationship or ovulatory status. 101 This first finding is consistent with the predictions generated by the youthfulness and big brain hypotheses, but runs contrary to expectations derived from the baby-face overgeneralization hypothesis. With regards to the sex of the judge, men (overall) demonstrated a stronger preference for facial underdevelopment. In general, men ranked juvenilized faces higher and selected them more frequently across short- and long-term mating items. And although men did not find facial underdevelopment more attractive per se, they did find it more desirable in potential mating partners - men ranked juvenilized faces higher, and selected them more frequently when evaluating the target as a possible lover or spouse. This suggests that men found facial underdevelopment more sexually enticing than women in targets of the opposite sex. Thus, this second finding seems to favour the youthfulness hypothesis over both the baby-face overgeneralization and big brain hypotheses. With regards to mating strategy, several results suggested that the preference for facial underdevelopment was stronger for short-term mating, whereas a number of other less compelling findings failed to detect a difference. In support of the former, the effect sizes for attributes generally indicative of short-term mating were significantly larger than those witnessed for long-term attributes - juvenilization was ranked higher, and selected more frequently, when the item reflected a short-term mating preference. Second, short-term mating items were better predictors of attractiveness compared to long-term items. The average correlation between attractiveness and short-term mating items was 2.5 times stronger than that observed between attractiveness and long-term items. Third, juvenilization was judged relatively more favourably when the target was 102 being evaluated as a potential lover rather than as a possible marriage partner, and, moreover, desirability as a lover was a better predictor of attractiveness than the target's appeal as a spouse. Conversely, there was no relationship between socio-sexual orientation and the preference for juvenilization in either experiment one or two. If facial underdevelopment was more closely associated with a short-term mating strategy, then individuals with less restricted orientations towards casual sex should have expressed a stronger attraction to facial juvenilization. Second, although the mean correlation between attractiveness and short-term mating items was larger overall, judges demonstrating a strong attraction to facial underdevelopment also tended to rate juvenilized faces high on parental ability and on desirability as a marriage partner. Thus, juvenilization is also relevant to long-term sexual relationships. Furthermore, the larger effect discovered for short-term mating items in experiment two may have resulted from a weaker mating context, which may have resulted in differential attenuation in the means of the two mating strategies. As discussed in chapter three, some long-term mating items (e.g. warmth, honesty, and intelligence) are fundamental to a variety of relationships, both sexual and platonic, and when these items were presented within a palliated mating context, as they were in experiment two, the target may not have been evaluated as a potential mating partner. Concentration of these items in the long-term set may have created the disparity favouring the effect for short-term mating. If so, we should expect to find greater attenuation of the long-term mating effect between experiments one and two. The average effect size for the short-term items (excluding "sexy" which was not used in experiment one) decreased by (.99 - .74 = ) 25 points between the two 103 experiments, whereas the average effect size for the long-term items decreased by (.69 -.24 =) 45 points. Thus, the effect for long-term items was more adversely influenced by the change in experimental design. If we consider those items that were most specific to sexual reproduction - desirable lover, desirable spouse, and parenting skill - the effect size decreased by (.67 - .43 =) 24 points. In contrast, those of intermediate specificity -attractiveness, youth, and health - decreased by (1.08 - .77 =) 31 points, and those most general - warmth, honesty, and intelligence - decreased by (.71 - .18 = ) 53 points. Thus, mitigating the mating context had a greater effect on items that were less specific to sexual reproduction, which were concentrated in the long-term mating set, and may have ultimately led to the weaker effect that was observed for long-term mating items in experiment two. An alternative approach to measuring the preference for facial underdevelopment as a function of mating strategy is to investigate the relationship between an item's appeal within each strategy and its respective effect for juvenilization. If underdevelopment is germane to a particular mating strategy, then juvenilization should have been selected more frequently for items that were rated as more desirable within that strategy. To this end, the effect sizes for all 10 mating items in experiment two were correlated with each of their desirability ratings for the two romantic relationships collected in the item desirability questionnaire. The coefficients for short- and long-term mating partners were (.47) and (-.67) respectively. Thus, the more desirable an attribute was rated for short-term mating, the more likely the sample was to select juvenilized faces over adultified faces on that item. Conversely, when the item was highly desirable in a long-term partner, the less likely the sample was to select facial juvenilization. 104 Thus, all considered, the preference for facial underdevelopment seems stronger for short-term, rather than long-term mating. Compared to scores for long-term items, those for short-term items were significantly greater in both experiments, and correlated higher with attractiveness. In addition, juvenilization was judged more favourably when the target was evaluated as a possible sex partner rather than as a partner in marriage and, finally, the more desirable an item was for short-term mating, the more likely participants were to choose juvenilization over adultification. These results seem to favour the predictions generated by the youthfulness hypothesis over those derived by the big brain hypothesis. 4.2 Comparing the Hypotheses To recap, the preference for facial underdevelopment was stronger within a mating context, stronger for men judging women, and stronger for a short-term mating strategy. Of the three theories, the data were least supportive of the baby-face overgeneralization hypothesis. Neither of the two predictions were supported. It was antithetical to this theory for the juvenilization preference to be sexual in character and to find the stronger affinity for juvenilization in male judges when it is women who are most interested in interacting with, and caring for, young children. Of the two remaining theories, the data were most supportive of the youthfulness hypothesis, for which all three predictions were supported. While there was some support for the big brain hypothesis, predictions regarding the judge's sex and mating strategy were not borne out in the data, and the evidence regarding a mating context was not unique to this theory. In the discussion that follows, the implications for each hypothesis with regards to the 105 preference for facial underdevelopment and the evolution of facial paedomorphosis will be considered more closely. 4.2.1 The Baby-face Over generalization Hypothesis Babyish looking individuals tend to be perceived as cuter, warmer, kinder, more submissive, more honest, more nai've and physically weaker than their more mature looking counterparts (Zebrowitz-McArthur & Apatow, 1983-1984; Berry & Zebrowitz-McArthur, 1985). Although facial underdevelopment in the present investigation shared many of these same perceptions - juvenilization was associated with greater warmth and honesty - the overgeneralization hypothesis cannot rationalize the human mate preference for facial underdevelopment. First, the strongest effect for juvenilization occurred with short-term mating items, an indication that the affinity for facial underdevelopment is indeed a sexual preference. Second, there is the antipodean finding that men judged facial juvenilization more favourably than did women. Third, further scrutiny of the "baby-face" item from experiment two (not included in the original analysis) indicated that it was no more associated with juvenilization than were the big-five items controlling for the halo effect. Thus, infantile perceptions were no more fundamental to the appeal of underdevelopment than the generally positive perceptions associated with physical attractiveness in general. Fourth, The baby-face item itself did not correlate with attractiveness. Thus, judges who demonstrated the largest affinity failed to associate juvenilization with an infantile appearance. Aside from the data, there are a number of additional reasons why the overgeneralization hypothesis may not be the best explanation for either the preference for facial underdevelopment, or the evolution of facial paedomorphosis. Although baby-106 faced individuals do receive some benefits as a result of their appearance, they must also endure a number of critical hindrances as well. Men with more babyish faces report less control over social interactions than mature faced men (Berry & Landry, 1997), and may have found it more difficult to move up social hierarchies or to maintain positions of status once they were achieved. In a study of West Point Academy graduates, for instance, more dominant looking alumni advanced to higher rank than their more submissive (babyish) looking classmates (Mazur, Mazur & Keating, 1984). And because an infantile appearance is more cute, and less sexy, baby-faced men in the ancestral environment may have found it more difficult to acquire mating opportunities. In the modern environment, high school boys who had a more dominant appearance reported more sexual activity than boys who appeared submissive, even though their physical development and rated attractiveness were comparable (Mazur, Halpern & Udry, 1994). Thus, while babyishness may have enhanced fitness by way of auxiliary altruism and reduced aggression, limitations to a man's ability to gain status and resources, and interference in his access to mating partners surely would have reduced fitness, at least for ancestral men. Another reason to doubt the veracity of the overgeneralization hypothesis as it relates to the evolution of facial morphology is that other species closely related to ancestral humans, and who faced some of the same ecological problems, did not evolve underdeveloped faces. Ancestors of other primate species must have faced the challenge of discriminating juvenile members from adults and allocating parental effort accordingly. In fact, other modern primates seem to possess an infant recognition system similar to that of humans. Other primate species prefer to look at infants rather than at 107 other adults, and approach often unrelated infants to touch, cuddle, carry, and groom them (Lancaster, 1972; Sackett, 1973). And immature members are often treated with forbearance in response to actions that would normally elicit aggressive behaviour. Presumably, facially underdeveloped adult members of other ancestral primates could have benefited from the overgeneralization of altruistic behaviour and reduced aggression that would have normally been reserved for infants. Despite this, however, neither chimpanzees, gorillas, nor orang-utans reflect a paedomorphic facial morphology. Thus, given the results of this study and other theoretical considerations, the baby-face overgeneralization hypothesis appears to be untenable. Perceptions of infant faces do not seem to be the driving force behind the preference for facial underdevelopment and, consequently, facial paedomorphosis probably did not evolve through exploitation of a facial recognition system that was originally designed to manage the distribution of parental effort. However, it is important to note that the results of this study do not prove that the baby-face overgeneralization effect is false. The potency of this effect has been demonstrated empirically during the course of numerous investigations (e.g., Berry & Zebrowitz-McArthur, 1985; Zebrowitz-McArthur & Apatow, 1983-1984), and, furthermore, several of the perceptions associated with facial juvenilization in the current study were not inconsistent with the overgeneralization effect. The point is that the mere existence of this effect can neither explain the evolution of facial underdevelopment in humans, nor the preference favouring it. 4.2.2 The Youthful Mimic Hypothesis Fertility and residual reproductive capacity decrease steadily in both men and women beginning in young adulthood. Because the decline in fertility is relatively 108 steeper for women, men tend to prefer youthful mating partners who can be, in some cases, substantially younger than themselves. Women, on the other hand, place less emphasis on youthfulness since maturity in men is less of a factor for reproduction (Buss, 1989; Kenrick & Keefe, 1992). Results indicated that the manipulations altering the level of apparent development in target faces successfully impacted their perceived age -not only were juvenilized faces judged to be more youthful, but the effect size associated with the youthfulness item was the second strongest in both experiments, preceded only by attractiveness. Furthermore, judges who demonstrated the strongest affinity for facial underdevelopment were also the ones most likely to associate juvenilization with a youthful appearance. Thus, across both sexes, less facial development was associated with greater youthfulness and greater attractiveness. Because the youthfulness hypothesis proposes that perceptions of age are the driving force behind the preference for facial underdevelopment, it was thought that youthfulness might account for a significant proportion of the variance in attractiveness since each item is (presumably) measuring the same dimension - juvenilized faces are attractive because they appear youthful. Furthermore, youthfulness might also account for a large proportion of the covariation between attractiveness and residual mating items. In other words, attractive faces make desirable mating partners because they bear a youthful appearance. If so, then removing the covariation that was due to youthfulness should have purged these associations. In the first place, youthfulness did correlate significantly with physical attractiveness, but no more so than either health, honesty, or desirable marriage partner, and the relationship was actually much weaker than that observed for sexy and desirable sex partner. And in the second case, partialing out 109 youthfulness did not weaken any of the relationships between attractiveness and the other mating items to any significant degree. Do these findings provide strong counter-evidence to the youthful mimic hypothesis? Not really. But, if it is mimicry of youthful features that is responsible for the attractiveness of facial underdevelopment, these findings suggest that the judge is unaware of their association, and that supplementary perceptions may cooperate with youthfulness to drive the preference for facial underdevelopment. Although the affinity for juvenilization may be present in both men and women, another stipulation of the juvenilization hypothesis was that the preference be relatively stronger in male judges since age, and perceptions thereof, are more closely linked to female reproductive capacity and, by extension, male sexual desire. However, no difference was detected between the sexes in their preference for facial underdevelopment - both men and women judged juvenilization to be equally (highly) attractive. Furthermore, the extent to which perceptions of youth predicted a target's physical attractiveness was about equal for men and women. Regardless of their gender, judges who perceived juvenilized targets as more youthful also rated them as more attractive. On the other hand, a sex difference was detected when the target was evaluated as a possible mating partner - underdeveloped female targets were judged to make more desirable lovers and marriage partners by male judges. Although (again) a significant predictor for both male and female judges, the extent to which youthfulness predicted desirability was much stronger for male judges - men who perceived juvenilized faces as more youthful expressed a stronger desire for them as potential 110 lovers and marriage partners, and were more likely to do so compared to female judges. Thus, overall, these findings seem consistent with the youthful mimic hypothesis. However, the fact that female judges also associate facial juvenilization with attractiveness and desirability suggests that youthfulness may only partially explain the preference for underdevelopment. By itself, youthfulness accounted for only 18% of the variation in attractiveness for male judges and 11% for female judges. When all four short-term mating items were entered as predictors into two regression analyses, one for men and one for women, the equation accounted for 43% of the variation in attractiveness for men and 59% for women. Thus, adding judgments of health, sex appeal, and desirability as a lover accounted for an additional 25% of the variation in attractiveness for male judges and a further 48% for female judges. Thus, the youthful mimic hypothesis cannot explain the preference for underdevelopment in female judges. Not only does it fail to provide an explanation as to why women might find juvenilization attractive, but perceptions of youthfulness constitute less than one quarter of the explainable variance provided by short-term mating items. And in fact, youthfulness comprises less than half of the explainable variance in attractiveness for male judges. Therefore, short-term mating items other than youthfulness account for more variation than youthfulness itself. Clearly, other factors contribute to the preference for facial underdevelopment in both sexes, but particularly for women judging male faces. Incidentally, when all mating items were entered as predictors in the equation, the proportion of variance that was explained increased to 58 percent for men and 60 percent for women. Thus, most of the variation in male attractiveness can be accounted for by 111 the short-term items alone but an additional 15% of the variation in female attractiveness can be accounted for in long-term mating. Overall, the data are most supportive of the youthfulness hypothesis. The preference for juvenilization was stronger within a mating context, was stronger for male judges evaluating female targets, and was stronger for a short-term mating strategy. Hence, all three predictions generated by this hypothesis were sustained. However, youthfulness by itself accounted for only a small proportion of the explainable variance in attractiveness (much more was explained by other mating-relevant entities), and failed to explain why women possess an equivalent attraction for facial underdevelopment. Nevertheless, the youthful mimic hypothesis seems to provide the best explanation for the juvenilization preference. As for underdevelopment itself, it remains unclear as to whether the youthful mimic hypothesis can account for the evolution of facial paedomorphosis. Although consistent with the possibility that natural selection favoured a developmental trajectory mimicking facial cues to youth and residual reproductive capacity, the results of this study were not sufficiently corroborative. 4.2.3 Big Brain Hypothesis Heterochronic processes in hominid evolution have been integral to the emergence of human cognition. While overdevelopment of the braincase and its contents enhanced facultative power and complexity, extension of the life-stages was necessary to fully maturate these newly acquired cerebral dexterities. Facial paedomorphosis may have been an important contributing factor in permitting further expansion to the braincase and economizing its existing volume. The data required to either sustain or discard this conjecture are presently unavailable given that human facial ontogeny has 112 been largely overlooked in phylogenetic circles. What empirical evidence there is seems to be supportive. For instance, primate dentition is negatively associated with brain size in the fossil record (McNamara, 1997), and acceleration towards facial underdevelopment seems to coincide with the appearance of more complex tool making and cooking technologies in cultural evolution (Brace, 1991; Klein, 1989; Wynn, Tierson & Palmer, 1996). Facial underdevelopment as a standard of physical attractiveness may reflect a generally adaptive preference for robust mating partners. If facial paedomorphosis was indeed an antecedent to human cognition, underdevelopment may have provided a reasonably honest cue to intellectual potential, thereby rendering relatively underdeveloped individuals desirable for mating. If so, juvenilization should have been more attractive within a context for which mating was highly salient. And, in fact, this prediction was supported in the data. It was further predicted that the preference for juvenilization should have been stronger for female judges viewing male targets, and for items pertinent to a long-term mating strategy. These ensuing predictions were not supported. However, there are two reasons the big brain hypothesis should be further investigated. First, as discussed previously, the relatively weaker effect observed for long-term mating items may have been an artefact of their greater generalizability to platonic relationships. Thus, the advantage observed for short-term mating items might better reflect general mating preferences. In which case, the results are not entirely irreconcilable with the big brain hypothesis. Intelligence, an important prerequisite for long-term mating, may not be critical for short-term mating, but could still be highly desirable. In other words, low intelligence might restrict your opportunities for long-term 113 mating, but high intelligence could still increase your access to short-term partners. Second, even if the big brain hypothesis is untenable in accounting for the juvenilization preference, it may still explain why paedomorphosis dominated human facial morphology. That is to say, facial underdevelopment could have evolved as a necessary antecedent to larger brains and higher cognitive functioning, but the preference for facial underdevelopment may have evolved in response to some other selection pressure, perhaps for reasons consistent with the youthfulness hypothesis - underdeveloped faces with exaggerated cues to youth (falsely) signal high fertility and greater residual reproductive capacity. Thus, this alternative suggests that the preference for facial underdevelopment and underdevelopment itself evolved independent of one another. Unlike youthfulness, the effect for intelligence in experiments one and two was much less impressive. Judges showing a strong affinity for facial underdevelopment did not associate juvenilization with high intellectual aptitude - the effect size for intelligence in experiment one was significantly greater than zero but smaller than the average effect size for mating items in general, and the effect size for intelligence in experiment two was, essentially, nil. However, compared to its two alternatives, this does not pose a large problem for the big brain theory. Perception is at the very core of the baby-face and youthfulness hypotheses - juvenilized faces were attractive because they were perceived as more infantile, or because they were perceived as more youthful. From the perspective of the big brain hypothesis, however, there is no reason to suppose that people's attraction to underdevelopment should be linked to a conscious appraisal of its origin, anymore than the predilection for symmetrical features requires an awareness of its link to developmental stability (Gangestad, Thornhill, & Yeo, 1994), or the attraction 114 to low waist-to-hip ratios (WHR) in female targets requires an awareness of the link to health and fertility (Singh, 1993). If facial paedomorphosis was associated with higher intellectual potential, then underdeveloped features would have provided an honest cue to observers of an individual's fitness and mate value. It should be noted that some other alternatives have been proposed to account for facial paedomorphosis. One such possibility is that the spherical shape of the head and diminutive position of the face, below and in front of the braincase, allows the cumbersome mass of the infant head to be packaged into a restricted area such as the inside of a uterus (Gould, 1977). Thus, the paedomorphosis may have evolved in response to larger brain size, but only with regards to the ability to give birth to offspring with relatively inflated craniums. However, this theory fails to explain why facial development continues on a retarded trajectory following birth. Another possibility is that childish features may have evolved to elicit greater parental investment, or possibly to enhance pair-bonding and evolution of the human family (Gould, 1977). This hypothesis follows a similar argument to that of the baby-face overgeneralization hypothesis, and suffers from the same weakness - the results of the current investigation demonstrate a strong sexual attraction to juvenilization. In conclusion, the results supported some elements of the big brain hypothesis but not others, and overall, the results were less supportive of the big brain hypothesis compared to the youthful mimic hypothesis. Nevertheless, unlike the baby-face hypothesis, this last hypothesis warrants further consideration, if not to elucidate the preference for facial underdevelopment, it is still a viable explanation for the occurrence of facial paedomorphosis within human phylogeny. Towards this end, research 115 documenting the relationship between underdeveloped features and intellectual performance in German women (Kiener & Keiper, 1979; Kozeny, 1968) should be further investigated. If underdevelopment evolved in response to demands for greater intelligence, there may still be evidence of this relationship in the population. Thus, support for the big brain hypothesis could be gained if underdevelopment was demonstrated to reliably predict individual differences in cerebral competence. 4.3 Juvenilization or Feminization? Correlational studies have established a relationship between underdeveloped features in the face and physical attractiveness (Cunningham, 1986, Cunningham, Barbee & Pike, 1990; Cunningham et. al., 1995). Other studies have manipulated the apparent level of development in the face to demonstrate a relationship between juvenilization and attractiveness (Johnston & Franklin, 1993; Jones, 1995; Wehr, et. al., 2001). This study provides yet another example of the latter. However, there is a parallel line of research investigating the attractiveness of feminization in human faces. One example is provided by Perrett and his collaborators (1998) who used graphic morphing techniques to enhance differences between a Japanese and European male and female composite face along a continuum ranging from 50% feminized to 50% masculinized. Both Japanese and European judges preferred feminized faces regardless of the gender or race of the target. Another example is provided by Rhodes, Hickford, and Jeffery (2000) who created multiple versions of an Asian and Caucasian man and woman ranging from 100% feminized to 100% masculinized. Again, Caucasian judges preferred significantly feminized faces for all four targets, with average amounts of feminization measuring 35% and 33% for female and male faces, respectively. 116 Not only do the results of juvenilization and feminization studies converge, but so too do the features used to describe underdevelopment and feminization. This is because sexual dimorphism arises from the differential timing and rate of somatic growth that occurs during all phases of development beginning with the embryo and continuing through adulthood. In humans, male foetuses develop at a slightly faster rate (Mittwoch, 1993) and reach sexual maturity later, remaining in the developmentally rapid stage of childhood an average of 2 years longer than females (McNamara, 1995). Rapid growth and a later offset of development indicate an extended developmental trajectory (peramorphism), meaning that men are relatively overdeveloped, physically, compared to women. Thus, feminizing a face not only abates its masculinity, but it also reduces its apparent level of development. This raises an important question. What is the best way to conceptualize this type of research: juvenilization or feminization? I believe there are five reasons why juvenilization is the better approach to this construct. First, to reconcile their results with the good genes model, the researchers were required to provide separate explanations for the attractiveness of feminized male and female faces. The appeal of feminization in female faces has been explained as a function of enhancing the cues that normally advertise youth and fertility whereas various positive and negative perceptions of personality were employed to account for the same effect in male faces. For instance, masculinized male faces were judged to be more dominant, cold, dishonest, uncooperative, to lack emotional capacity, and to possess poor parenting skills. (Perrett et al., 1998; Fink & Penton-Voak, 2002). Conceptualizing the manipulation as one of juvenilization requires only a single explanation for both sexes -reduced facial development reflects a phenotype that typically increases fitness. Thus, 117 defining the construct in terms of differences in development gains the benefit of greater parsimony. Second, as stated already, one interpretation of the data on feminized male faces suggests that less masculine men might be more desirable for long-term committed relationships and parenting roles because they tend to be less aggressive and more cooperative (Perrett et al., 1998; Penton-Voak et al., 1999; Rhodes, Hickford, & Jeffery, 2000; Fink & Penton-Voak, 2002). However, the empirical evidence available in the present study fails to support this hypothesis. Although juvenilized faces were ranked higher than chance on both warmth and parenting, the effect sizes for items associated with sexual attraction were more impressive. In experiment one, the effect sizes for attractiveness and youthfulness were more than twice as large as those for parenting skill and warmth, which were on par with control items such as tidiness, spirituality, and musical ability. Even more convincing was the fact that the effect sizes for sexy and skilled lover in experiment two were, again, twice as large as those for warmth and parenting, which were on par with control items such as extraversion and conscientiousness. If the appeal of feminized male faces stems from perceptions of warmth and parenting, the effect sizes associated with these items should have been considerably larger than they were observed to be. Third, natural selection is notoriously unconcerned with courteous behaviour during pair-bonding and, in fact, lengthy expositions on male sexual proprietariness (Daly & Wilson, 1988), sexual jealousy (Buss, 2000), and sexual coercion (Thornhill, & Palmer, 2000) have suggested rather the opposite. No doubt at a conscious level women prefer men who are easier to get along with and show an interest in sharing parental 118 responsibilities, but these may not be the most important criteria women use when choosing between potential mates, and they may readily trade these positive personality traits off for something more tangible to reproductive success. Fourth, sexual dimorphism in a species is directly contingent on its mating system. In highly polygamous species, competition between males in finding mating partners is intense, and the males of these species tend to exhibit extreme sexual dimorphism. Conversely, dimorphism in relatively monogamous species tends to be minor, although some dimorphism may still occur if males differ significantly in quality (Andersson, 1994). Humans are only mildly polygamous and, consequently, only mildly dimorphic between the sexes. Therefore, enhancing secondary sexual characteristics probably has only a limited effect on facial attractiveness. Humans are highly paedomorphic regarding their faces, however. Given the relative importance of sexual selection and heterochrony in human evolution, juvenilization is the theoretically stronger conceptualization. Fifth, and perhaps most importantly, in the only experiment to directly compare the attractiveness of juvenilization and feminization, researchers concluded that juvenilization had the bigger effect. Ishi and Gyoba (2001) created 20-face composites of a Japanese man, woman, boy, and girl, and then created two arrays of adult faces ranging from 50% feminized to 50% masculinized, and 50% juvenilized to 50% adultified. Twenty-eight male and female judges ranked the three most attractive in one of two subsets, each consisting of examples from both arrays. For both the male and female target, juvenilized and feminized versions were ranked higher than adultified or masculinized versions. In addition, juvenilized versions were ranked higher than 119 feminized versions. The difference between the juvenilized and feminized versions was greater for the female target, with the male target not reaching statistical significance. However, this was probably due to the experiment's relatively small sample size and curious procedure. Although it is premature for a definitive conclusion, the results of this experiment suggest that it is juvenilization, and not feminization per se, that increases facial attractiveness. In sum, defining manipulations of juvenilization as feminization lacks plausibility, parsimony, and empirical support. The alternative is that feminization increases facial juvenilization, which has been selected for in both men and women, possibly because it mimics cues to youth and fertility, or possibly because it was as an antecedent to human cognition. This second interpretation has greater theoretical credibility, greater parsimony, and is consistent with the findings of both the current study and those investigating sexual dimorphism. 4.4 Summary and Conclusion Human evolution has been defined by facial paedomorphosis - truncation of facial development leading to an increasingly more juvenilized appearance in modern homo sapiens. Furthermore, both men and women tend to show an affinity for facial underdevelopment in the opposite sex. Three theories were proposed to account for the possible co-evolution of underdevelopment in human morphology and mating preferences. The baby-face overgeneralization hypothesis proposed that facial underdevelopment stimulated altruistic behaviour and inhibited aggression. The youthfulness hypothesis proposed that underdevelopment mimicked youthfulness cues in the face that were generally associated with fertility and greater residual reproductive 120 capacity. The big brain hypothesis proposed that facial underdevelopment was a necessary antecedent to supplementary gains in brain size, which ultimately lead to the evolution of human cognition. To compare and evaluate each of these theories, predictions were generated regarding the juvenilization preference as a function of context, mating strategy, and sex of the judge. Results indicated that the preference for juvenilized features was stronger in a mating context, for a male judge evaluating a female target, and for a short-term mating strategy. These findings were least consistent with the baby-face overgeneralization hypothesis, and most consistent with the juvenilization hypothesis. However, although it seems to explain the preference for facial juvenilization, the youthful mimic hypothesis may be insufficient to account for the evolution of facial paedomorphosis. Given that heterochrony in general has played such a significant role in the evolution of human cognition, it is possible that facial paedomorphosis may also have been an important contributing factor. Thus, although results were relatively less supportive of the big brain hypothesis with regards to the juvenilization preference, this last theory may be the most satisfactory in explaining the significance of facial paedomorphosis in human evolution. Future investigations should attempt to identify a relationship between underdeveloped facial morphology and actual intellectual performance. 121 Bibliography Abbie A. A. (1947). Headform and human evolution. Journal of Anatomy, 81, 233-258. Alley, T. R. (1981). Head shape and the perception of cuteness. Developmental Psychology, 17, 650-654. Alley, T. R. (1983). Infantile head shape as an elicitor of adult protection. Merrill-Palmer Quarterly, 29, 411-427. Andersson, M. B. (1994). Sexual selection. Princeton: Princeton University Press. Barber, N. (1995). The evolutionary psychology of physical attractiveness: Sexual selection and human morphology. Ethology and Sociology, id, 395-424. Barkow, J. (1989). Darwin, sex, and status. Toronto: University of Toronto Press. Berry, D. S. (1991). Attractive faces are not all created equal: Joint effects of facial babyishness and attractiveness on social perception. Personality and Social Psychology Bulletin, 17, 523-531. Berry, D. S. (2000). Attractiveness, attraction, and sexual selection: Evolutionary perspectives on the form and function of physical attractiveness. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 32, pp 273-342). San Diego, CA: Academic Press. Berry, D. S., & Landry, J. C. (1997). Facial maturity and daily social interaction. Journal of Personality and Social Psychology, 72, 570-580. Berry, D. S., & Zebrowitz-McArthur, L. (1985). Some components and consequences of a babyface. Journal of Personality and Social Psychology, 48, 312-323. 122 Berry, D. S., & Zebrowitz-McArthur, L. (1986). Perceiving character in faces: The impact of age-related caniofacial changes on social perception. Psychological Bulletin, 100, 3-18. Berry, D. S., & Zebrowitz-McArthur, L. (1988). What's in a face? Facial maturity and the attribution of legal responsibility. Personality and Social Psychology Bulletin, 14, 23-33. Bolk, L. (1926). Das problem der menschwerdung. Jena: Gustav Fischer. Bowlby, J. (1969). Attachment and loss (Vol. 1). New York: Basic Books. Brace, C. L. (1962). Cultural factors in the evolution of the human dentition. In M. F. A. Montagu (Ed.), Culture and the evolution of man (pp. 343-354). New York: Oxford University Press. Brace, C. L. (1991). What big teeth you had grandma! Human tooth size, past and present. In M A. Kelley & C. S. Larsen (Eds.), Advances in dental anthropology (pp. 33-57). New York: Wiley-Liss. Buss, D. M. (1989). Sex differences in human mate preferences: Evolutionary hypotheses tested in 37 cultures. Behavioural and Brain Sciences, 12, 1-49. Buss, D. M. (1994). The evolution of desire: Strategies of human mating. New York: BasicBooks. Buss, D. M. (2000). The dangerous passion: Why jealousy is as necessary as love and sex. New York: Free Press. Buss, D. M., & Barnes, M. (1986). Preferences in human mate selection. Journal of Personality and Social Psychology, 50, 559-579. Buunk, B. P., Dijkstra, P., Kenrick, D. T., & Warntjes, A. (2001). Age preferences for 123 mates as related to gender, own age, and involvement level. Evolution and Human Behavior, 22, 241-250. Cohen, B. L., & Lilienfeld, A. M. (1970). The epidemiological study of mongolism in Baltimore. Annual New York Academy of Science, 171, 320-327. Collman, R. D., & Stoller, A. (1962). A survey of mongoloid births in Victoria, Australia, 1942-1957. American Journal of Public Health, 52, 813-829. Coss, R. G. (1991). Evolutionary persistence of memory-like processes. Concepts in Neuroscience, 2, 129-168. Coss, R. G., & Schowengerdt, B. T. (1998). Evolution of the modern human face: Aesthetic and attributive judgments of a female profile warped along a continuum of paedomorphic to late archaic craniofacial structure. Ecological Psychology^ 10, 1-24. Cunningham, M. R. (1986). Measuring the physical in physical attractiveness: Quasi-experiments on the sociobiology of female facial beauty. Journal of Personality and Social Psychology, 50, 925-935. Cunningham, M. R., Barbee, A. P., & Pike, C. L. (1990). What do women want? Facialmetric assessment of multiple motives in the perception of male facial physical attractiveness. Journal of Personality and Social Psychology, 59, 61-72. Cunningham, M. R., Roberts, A. R., Barbee, A. P., Druen, P. B., & Wu, C. H. (1995). "Their ideas of beauty are, on the whole, the same as ours": Consistency and variability in the cross-cultural perception of female physical attractiveness. Journal of Personality and Social Psychology, 68, 261-279. 124 Daly, M . , & Wilson, M . (1980). Discriminative parental solicitude: a biological perspective. Journal of Marriage and the Family, 42, 277-288. Daly, M , & Wilson, M . (1983). Sex, evolution, and behavior (2nd ed.). Belmont: Wadsworth Publishing Company. Daly, M . , & Wilson, M . (1988). Homicide. New York: Aldine De Gruyter. Dion, K. , Berscheid, E., & Walster, E. (1972). What is beautiful is good. Journal of Personality and Social Psychology, 24, 285-290. Eibl-Eibesfeldt, I. (1975). Ethology: The biology of behavior (2nd ed.). New York: Holt, Rinehart & Winston. Enlow, D. H. (1982). Handbook of facial growth (2nd ed.). Philadelphia: Saunders. Feingold, A. (1992). Good looking people are not what we think. Psychological Bulletin, 11, 304-341. Fink, B., & Penton-Voak, I. (2002). Evolutionary psychology of facial attractiveness. Current Directions in Psychological Science, 11, 154-158. Frodi, A . M . , Lamb, M . E., Leavitt, L. A. , & Donovan, W. L. (1978). Fathers' and mothers' responses to infant smiles and cries. Infant Behavior and Development, 1, 187-198. Frodi, A . M . , Lamb, M . E., Leavitt, L. A. , Donovan, W. L., Neff, C , & Sherry, D. (1978). Fathers' and mothers' responses to the faces and cries of normal and premature infants. Developmental Psychology, 14, 490-498. Furnham, A. , Mistry, D., & McClelland, A. (2004). The influence of age of the face and the waist to hip ratio on judgements of female attractiveness and traits. Personality and Individual Differences, 36, 1171-1185. 125 Gangestad, S. W., & Thornhill, R. (1997). The evolutionary psychology of extrapair sex: The role of fluctuating asymmetry. Evolution and Human Behavior, 18, 69-88. Gangestad, S. W., & Thornhill, R. (2003). Facial masculinity and fluctuating asymmetry. Evolution and Human Behavior, 24, 231-241. Gangestad, S. W., Thornhill, R., & Yeo, R. A. (1994). Facial attractiveness, developmental stability, and fluctuating asymmetry. Ethology and Sociobiology, 15, 73-85. Geldart, S., Maurer, D., & Carney, K. (1999). Effects of eye size on adults' aesthetic ratings of faces and 5-month-olds' looking times. Perception, 28, 361-374. Gibson, K. R. (2002). Evolution of human intelligence: The roles of brain size and mental construction. Brain, Behavior and Evolution, 59, 10-20. Glanville, E. V . (1969). Nasal shape, prognathism and adaptation in man. American Journal of Physical Anthropology, 30, 29-37. Gould, S. J. (1977). Ontogeny and phylogeny. Cambridge: Belknap Press of Harvard University Press. Harlap, S., Shiono, P. H. , Ramcharan, S. A. (1980). A life table of spontaneous abortions and the effect of age, parity and other variables. In Porter, I. H. , Hook, E. B., (Eds.) Human embryonic and foetal death. New York: Academic Press. Harper, L. V . (1971). The young as a source of stimuli controlling caretaker behavior. Developmental Psychology, 4, 73-88. Harrison, A . A. , & Saeed, L. (1977). Let's make a deal: An analysis of revelations and stipulations in lonely hearts' advertisements. Journal of Personality and Social Psychology, 35, 257-264. 126 Henderson, J. J. A . , & Anglin, J. M . (2003). Facial attractiveness predicts longevity. Evolution and Human Behavior, 24, 351-356. Hildebrandt, K. A. , & Fitzgerald, H. E. (1983). The infant's physical attractiveness: Its effect on bonding and attachment. Infant Mental Health Journal, 4, 3-12. Ishi, H. , & Gyoba, J. (2001). Analyses of psychological aspects of attractiveness in feminized and juvenilized Japanese faces. Tohoku Psychologica Folia, 60, 29-34. Jackson, L. S., Hunter, J. E., & Hodge, C. N . (1995). Physical attractiveness and intellectual competence: A meta-analytic review. Social Psychology Quarterly, 58, 108-122. Jencks, C. (1979). Who gets ahead? The determinants of economic success in America. New York: Basic Books. Jerison, H. J. (1973). Evolution of the brain and intelligence. New York: Academic Press. Johnston, V . S., & Franklin, M . (1993). Is beauty in the eye of the beholder? Ethology and Sociobiology, 14, 183-199. Johnston, V . S., Hagel, R., Franklin, M . , Fink, B., & Grammar, K. (2001). Male facial attractiveness: Evidence for hormone-mediated adaptive design. Evolution and Human Behavior, 22, 251-267. Jones, D. (1995). Sexual selection, physical attractiveness, and facial neoteny. Current Anthropology, 36, 723-748. Jones, D. (1996). Physical attractiveness and the theory of sexual selection. Ann Arbor: Museum of Anthropology, University of Michigan. Jones, D., & Hi l l , K. (1993). Criteria of facial attractiveness in five populations. Human 127 Nature, 4, 271-296. Jones, J. L., Smith, D. W., Harvey, M . A. S., Hall, B. D., & Quan, L. (1975). Older paternal age and fresh gene mutation: Data on additional disorders. Journal of Pediatrics, 86, 84-88. Keating, C. F. (1985). Gender and the physiognomy of dominance and attractiveness. Social Psychology Quarterly, 48, 61-70. Keating, C. F., Randall, D. W., Kendrick, T., & Gutshall, K. A . (2003). Do babyfaced adults receive more help? The (cross-cultural) case of the lost resume. Journal of Nonverbal Behavior, 27, 89-109. Kenrick, D. T., & Keefe, R. C. (1992). Age preferences in mates reflect sex differences in human reproductive strategies. Behavioral and Brain Sciences, 15, 75-133. Kiener, F., & Keiper, E. (1979). Some connections between physiognomic traits and intelligence. Journal of Human Evolution, 5,805-808. Klein, R. G. (1989). The human career. Chicago: University of Chicago Press. Kozeny, E. D. (1968). Experimentelle er gebnisse uber den zasammernhang zwischen intelligenz und geschtsform in ihrer behiehung zum grazilisierungs problem. Erganzende anthropologische bemerkungen zur physiognomik der intelligenz und ihrer morphologischen entwicklungstendenz. Homo, 1, 27-48. Kuhlenbeck, H. (1973). Central nervous system of vertebrates (volume 3, part II). New York: Arnold-Backlin-Strasse. Lancaster, J. B. (1972). Play-mothering: The relations between juvenile females and young infants among free-ranging vervet monkeys. In F. E. Poirier (Ed.), Primate socialization (pp. 83-104). New York: Random House. 128 Langlois, J. H., Kalakanis, L., Rubenstein, A. J., Larson, A., Hallam, M . , & Smoot, M . (2000). Maxims or myths of beauty? A meta-analytic and theoretical review. Psychological Bulletin, 126, 390-423. Langlois, J. H., & Roggman, L. A. (1990). Attractive faces are only average. Psychological Science, 7,115-121. Langlois, J. H., Roggman, L. A., & Musselman, L. (1994). What is average and what is not average about attractive faces? Psychological Science, 5, 214-220. Li , N. P., Kenrick, D. T., Bailey, J. M . , & Linsenmeier, J. A. W. (2002). The necessities and luxuries of mate preferences: Testing the tradeoffs. Journal of Personality and Social Psychology, 82, 947-955. Lorenz, K. (1971). Studies in animal and human behavior (Vol. 2). Cambridge: Harvard University Press. MacLeod, J., & Gold, R. Z. (1953). The male factor in fertility and infertility: Semen quality in relation to age and sexual activity. Fertility and Sterility, 4, 194-207. Maier, R. A., Holmes, D. L., Slaymaker, F. L. , & Reich, J. N. (1984). The perceived attractiveness of preterm infants. Infant Behavior and Development, 7, 403-414. Mazur, A., Halpern, C , & Udry, J. R. (1994). Dominant looking male teenagers copulate earlier. Ethology and Sociobiology, 15, 87-94. Mazur, A., Mazur, J., & Keating, C. (1984). Military rank attainment of a West Point Class: Effects of cadets' physical features. American Journal of Sociology, 90, 125-150. McKinney, M . L. , & McNamara, K. J. (1991). Heterochrony: The evolution of ontogeny. New York: Plenum Press. 129 McNamara, K. J. (1995). Sexual dimorphism: The role of heterochrony. In McNamara, K. J. (Ed.) Evolutionary change and heterochrony. Chichester and New York: Wiley. McNamara, K. J. (1997). Shapes of time: The evolution of growth and development. Baltimore: Johns Hopkins University Press. Mittwoch, U. (1993). Blastocysts prepare for the race to be male. Human Reproduction, 8, 1550-1555. Modell, B., & Kuliev, A. (1990). Changing paternal age distribution and the human mutation rate in Europe. Human Genetics, 86, 198-202. Montagu, A. (1989). Growing young (2nd edition). New York: McGraw-Hill. Parker, S. T., & McKinney, M. L. (1999). Origins of intelligence: The evolution of cognitive development in monkeys, apes, and humans. Baltimore: Johns Hopkins University Press. Paunonen, S. V., Ewan, K., Earthy, J., Lefave, S., & Goldberg, H. (1999). Facial features as personality cues. Journal of Personality, 67, 555-583. Penrose, L. S. (1933). The Relative effects of paternal and maternal age in mongolism. Journal of Genetics, 27, 219-224. Penrose, L. S. (1955). Parental age and mutation. Lancet, 2, 312-313. Penton-Voak, I. S., Perrett, D. I., Castles, D. L., Kobayashi, T., Burt, D. M., Murray, L. K., & Minamisawa, R. (1999). Menstrual cycle alters face preference. Nature, 399, 741-742. Perrett, D. I., Lee, K. J., Penton-Voak, I., Rowland, D., Yoshikawa, S., Burt, D. M., 130 Henzi, S. P., Castles, D. L., & Akamatsu, S. (1998). Effects of sexual dimorphism on facial attractiveness. Nature, 394, 884-887. Rhodes, G., Hickford, C , & Jeffery, L. (2000). Sex-typicality and attractiveness: Are supermale and superfemale faces super-attractive? British Journal of Psychology, 91, 125-140. Rhodes, G., Sumich, A., & Byatt, G. (1999). Are average facial configurations attractive only because of their symmetry? Psychological Science, 10, 52-58. Rhodes, G., & Tremewan, T. (1996). Averageness, exaggeration, and facial attractiveness. Psychological Science, 7,105-110. Risch, H. A., Weiss, N. S., Clarke, A. E., & Miller, A. B. (1988). Risk factors for spontaneous abortion and its recurrence. American Journal of Epidemiology, 128, 420-430. Sackett, G. P. (1973). Monkeys reared in isolation with pictures as visual input: Evidence for an innate releasing mechanism. In T. E. McGill (Ed.), Readings in animal behavior (2nd ed., pp. 263-269). New York: Holt, Rinehardt & Winston. Sadalla, E. K., Kenrick, D. T., & Vershure, B. (1987). Dominance and heterosexual attraction. Journal of Personality and Social Psychology, 52, 730-738. Schleidt, M., Schiefenhovel, K., Stanjek, K., & Krell, R. (1980). "Caring for a baby" behavior: Reactions of passersby to a mother and baby. Man-Environment Systems, 10, 73-82. Schmitt, D. P., & Buss, D. M. (1996). Strategic self-promotion and competitor derogation: Sex and context effects on the perceived effectiveness of mate attraction tactics. Journal of Personality and Social Psychology, 70, 1185-1204. 131 Schneider, E. L. (1978). The aging reproductive system^ New York: Raven Press Schultz, A. H. (1950). The physical distinctions of man. Proceedings of the American Philological Society, 94, 428-449. Shea, B. T. (1988). Heterochrony in primates. In McKinney, M. L. (Ed.) Heterochrony in Evolution. New York: Plenum Press. 237-266. Shea, B. T. (1989). Heterochrony in human evolution: The case for neoteny reconsidered. Yearbook of Physical Anthropology, 52,69-101. Simpson, J. A., & Gangestad, S. W. (1991). Individual differences in sociosexuality: Evidence for convergent and discriminant validity. Journal of Personality and Social Psychology, 60, 870-883. Singh, D. (1993). Adaptive significance of female physical attractiveness: Role of waist-to-hip ratio. Journal of Personality and Social Psychology, 65, 293-307. Snowden, C. T. (1999). An empiricist view of language evolution and development. In B.J. King (Ed.), The origins of language: What nonhuman primates can tell us (pp. 79-114). Sante Fe: School for American Research. Soler, C, Nunez, M., Gutierrez, R., Nunez, J., Medina, P., Sancho, M., Alvarez, J., & Nunez, A. (2003). Facial attractiveness in men provides clues to semen quality. Evolution and Human Behavior, 24, 199-207. Stearns, E. L., MacDonald, J. A., Kauffman, B. J., Lucman, T. S., Winters, J. S., & Faiman, C. (1974). Declining testicular function with age: Hormonal and clinical correlates. American Journal of Medicine, 57,761-766. Stephan, H., Fraham, H., & Baron, G. (1981). New and revised data on volumes of brain structures in insectivores and primates. Folia Primatolog, 35, 1-39. 132 Symons, D. (1979). The Evolution of Human Sexuality. New York: Oxford University Press. Symons, D. (1995). Beauty is in the adaptations of the beholder: The evolutionary psychology of human female sexual attractiveness. In Abramson, P. R., & Pinkerton, S. D. (Ed.) Sexual nature, sexual culture. Chicago: University of Chicago Press. 80-118. Taysi, K. (1988). Preconceptual counselling. Obstetrics and Gynecology Clinics in North America, 15, 167-178. Thornhill, R., & Gangestad, S. W. (1994). Human fluctuating asymmetry and sexual behavior. Psychological Science, 5, 297-302. Thornhill, R., & Palmer, C. T. (2000). A natural history of rape: Biological bases of sexual coercion. Cambridge Massachusetts: MIT Press. Tillenger, K. G. (1957). Testicular morphology. Acta Endocrinologica, 30, 28-39. Tinbergen, N . (1964). Social behaviour in animals (2nd ed.). London: Chapman & Hall. Toth, N . , Schick, K. , Savage-Rumbaugh, E. S., Sevcik, S., & Rumbaugh, D. (1993). Pan the tool-maker: Investigations into the stone toll-making and tool-using capabilities of a bonobo (Pan paniscus). Journal of Archaeological Science, 20, 81-91. Wehr, P., MacDonald, K. , Lindner, R., & Yeung, G. (2001). Stabilizing and directional selection on facial paedomorphosis. Human Nature, 12, 383-402. Weidenreich, F. (1947). The trend of human evolution. Evolution, 1, 221-236. Wilson, E. O. (1975). Sociobiology: The new synthesis. Cambridge, M A : Harvard University Press. 133 Wynn, T. G., Tierson, F. D., & Palmer, C. T. (1996). Evolution of sex differences in spatial cognition. Yearbook of Physical Anthropology, 39, 11-42. Zebrowitz, L. (1997). Reading faces. Oxford: Westview Press. Zebrowitz, L., Hall, J. A., Murphy, N. A., & Rhodes, G. (2002). Looking smart and looking good: Facial cues to intelligence and their origins. Personality and Social Psychology Bulletin, 28, 238-249. Zebrowitz, L., & McDonald, S. M. (1991). The impact of litigants' baby-facedness and attractiveness on adjudications in small claims courts. Law and Human Behavior, 15, 603-623. Zebrowitz, L. A., & Montepare, J. M. (1992). Impressions of babyfaced individuals across the life span. Developmental Psychology, 28, 1143-1152. Zebrowitz-McArthur, L., & Apatow, K. (1983-1984). Impressions of baby-faced adults. Social Cognition, 2, 315-342. Zebrowitz-McArthur, L., & Berry, D. S. (1987). Cross-cultural agreement in perceptions of babyfaced adults. Journal of cross-cultural psychology, 18, 165-192. Zebrowitz-McArthur, L., & Montepare, J. M. (1989). Contributions of a babyface and a childlike voice to impressions of moving and talking faces. Journal of Nonverbal Behavior, 13, 189-203. 134 Ap p e n d i x Your Sex: S I ? (please circle one) Instructions'. Please rate the following 32 items (both sides of this page) on how important they are to you personally in the context of romantic relationships. First rate all 32 items with regards to their importance for a short-term mating partner (a brief sexual relationship or encounter). Then rate each item with regards to their importance for a long-term mating partner (an extended or permanent committed relationship). Base your ratings on some hypothetical person, not anyone you know specifically. 1 = Irrelevant or Undesirable 3 = Nice but not essential 5 = Crucial Trait Short-Term Partner Long-Term Partner Agreeable Athletic Attractive 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Baby-faced Conscientious Critical 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Dominant Emotional Emotionally Stable 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Extraverted Funny A Skilled Lover 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Healthy Honest 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 135 1 = Irrelevant or undesirable 3 = Nice but not Essential 5 = Crucial Trait Short-Term Partner Long-Term Partner Patient Intelligent Lucky 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Tidy Musical Naive 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Nervous Open A Skilled Parent 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Talkative Sexy Sociable 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Spiritual A Good Spouse Strong. 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Submissive Warm Youthful 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 136 Your Sex: S I $ (please circle one) Instructions: Please rate the following 32 items (both sides of this page) on how important they are to you personally in the context of your professional and personal relationships. First rate all 32 items with regards to their importance for a co-worker (someone you work with on a daily basis). Then rate each item with regards to their importance for a close friend (someone you could trust and confide in). Base your ratings on some hypothetical person, not anyone you know specifically. 1 = Irrelevant or Undesirable 3 = Nice but not essential 5 = Crucial Trait Co-worker Close Friend Agreeable Athletic Attractive 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Baby-faced Conscientious Critical 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Dominant Emotional Emotionally Stable 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Extraverted Funny A Skilled Lover 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Healthy Honest 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 137 1 = Irrelevant or undesirable 3 = Nice but not Essential 5 = Crucial Trait Co-worker Close Friend Patient Intelligent Lucky 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Tidy Musical Naive 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Nervous Open A Skilled Parent 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Talkative Sexy Sociable 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Spiritual A Good Spouse Strong 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Submissive Warm Youthful 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 138 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092275/manifest

Comment

Related Items