Abstract
Skin color is one of the first features that we notice in another person and, therefore, it plays a significant role in the mate selection process as well as in the assessments of attractiveness of others. However, almost all modern research showing a preference for lighter skin tone (particularly in women) was conducted within populations of relatively light skin color. The current study was conducted among the Yali people, who are dark-skinned and native to the isolated highlands of West Papua. We found that for both males (n = 53) and females (n = 53) preferred skin tone was either average or slightly lighter than the average. At the same time, we found that the male preference for lighter skin tone in females was correlated with contact with other cultures. We discuss our results in the context of social and biological theories explaining skin tone preferences.
Introduction
Within almost all human populations, skin color varies as a function of sex and age (Samson, Fink, & Matts, 2010). In terms of age, infants of both sexes are very light-skinned (Garn, Selby, & Crawford, 1956). In addition, in most regions women tend to have a slightly lighter skin color than men (of the same genetic background; Jablonski & Chaplin, 2000; van den Berghe & Frost, 1986). The sex difference in skin color emerges at puberty and seems to be hormonally mediated. Under the influence of estrogens, women retain a lighter skin complexion, whereas in boys secretion of androgens drives the deposition of epidermal melanin and stimulates increased production of hemoglobin (van den Berghe & Frost, 1986). Lighter skin color in women is also associated with the necessity of absorbing more UV radiation, needed for the synthesis of provitamin D3, which is in high demand during pregnancy and lactation (Jablonski & Chaplin, 2000). In addition, lighter skin color might serve as an indicator of fertility, as skin color darkens with age, and changes slightly during the menstrual cycle. It is the lightest during ovulation (Garn et al., 1956) and becomes slightly darker during pregnancy and while using oral contraception (Wong & Ellis, 1984).
All aforementioned facts could influence male and female skin-color preferences (Frost, 1988, 1989, 1994, 2006). Attractiveness of skin color was investigated both in the context of ethnic and monoethnic population comparisons. Generally, pale-skinned women were considered more attractive than darker skinned women in a wide range of cultures including North America (Feinman & Gill, 1978), Europe (Pawlowski & Szymanczyk, 2008), and Asia (B. J. Dixson, A. F. Dixson, Li, & Anderson, 2007; Rondilla & Spickard, 2007; Yamaguchi & Aoki, 2008). Also, a study investigating the self-preference of dark-skinned women generally confirmed the aforementioned results. Women tended to like their natural or a slightly lighter skin tone when asked to assess their own faces (Bond & Cash, 1992; Hill, 2002—African Americans; Sahay & Piran, 1997—South Asian Canadians). Some of these results were discussed in the context of racial prejudice (e.g., Feinman & Gill, 1978), but other explanation might be that men, more than women, prefer lighter skin tone in women.
Despite that, some studies conducted in Western countries (Fink, Grammer, & Thornhill, 2001; Smith, Cornelissen, & Tovée, 2007; Swami, Furnham, & Joshi, 2008) showed that tanned skin was considered to be more attractive. Authors presenting such results suggested that it might be due to the fact that in the modern, Western culture darker skin tone may indicate healthy life style as well as social status.
In conclusion, several mechanisms play different parts in shaping the preferences toward skin tone. While it is possible that some of these mechanisms act in opposing ways, the ultimate result is an effect of their joint action. Despite a vast body of research conducted on skin tone preferences, many issues concerned with the subject of skin tone attractiveness still remain unclear and require further investigation. In our study we focused on two main issues: (a) Whether the preference for lighter skin color in women rather than in men is universal or more likely to appear in regions with lower UV radiation, and (b) Are the described preferences under the influence of sociocultural norms and trends?
In terms of relationship between UV radiation and skin tone preferences, almost all up to date research showing a preference for lighter skin color in women was conducted within populations with relatively light skin color (Samson et al., 2010). It is possible that preferences for lighter skin color in women would be observed mainly in regions with low UV radiation, where natural selection associated with protection from UV radiation is not as strong and where it is more favorable toward darker skin color. However, it is worth noting that a study conducted in Cameroon, showed that men prefer women with skin tone darker than average (app. 27%) rather than average or lighter (app. 18%). But the author showed that distribution of preferences for the five skin colors did not differ from the one obtained by chance (B. J. Dixson, A. F. Dixson, Morgan, & Anderson, 2007).
In the context of the second research question, we need to highlight the fact that the mechanisms of described preferences are mainly based on evolutionary or biological science, but it is possible that the preferences are also affected by sociocultural norms and trends. Bond and Cash (1992), for instance, found a distinct idealization of lightness among female African Americans. A similar observation was presented by Rondilla and Spickard (2007) in Asian American females. In addition, some studies (e.g., Tovée et al., 2006) showed that the aesthetic preferences, for example, for body shape are malleable and can change with exposure to different environmental conditions and cultures. However, it remains unclear whether a similar mechanism is observed in case of skin tone and if mere contact with other cultures is sufficient to modify these preferences, or whether also environmental changes need to occur.
Materials and Method
Participants
The research was conducted among the Yali tribe (Papua, semi-independent Indonesian province). The Yali inhabit mountainous areas east of the Baliem Valley (3.92 S, 138.73 E from Wamena—central part of Baliem Valley). It is a region of mountainous tropical climate (Haberle, Hope, & DeFretes, 1991), with very high solar radiation (Kalma, 1972). Populations inhabiting New Guinea have dark skin coloration (but lighter than in some regions of Africa; Map of skin color distribution: Chaplin, 2004, p. 299).
The Yali are one of many indigenous ethnic groups in Papua (previously known as Irian Jaya). Even though Christianity is present in this region, the Yali have preserved their traditional lifestyle, for example, including polygamy and dress (however these undergo constant changes and although some men still wear only traditional koteka covering only their penis, hardly any women wear the traditional grass skirts anymore). They cultivate plots, hunt, and breed pigs (which are a marker of a man’s wealth and social position). More detailed descriptions of this population can be found in anthropological works such as: Boissiere, 2002, 2009; Koch, 1974; Milliken, 1994; Wilson, 1986; Zoellner, 1988, and others.
Because of their remote location and difficult access, the mountain terrains of Papua were not explored by Westerners until about 50 years ago. Since then they have been heavily missionized. However, until this day, relatively low numbers of tourists have visited the region, mainly because of difficult access to the Yali territory—the only access routes are via private or chartered aircraft or a mountain trek, which takes several days. All the study sites were located along a mountainous route (from southeast to northwest) surrounding the Baliem Valley from the East (hand-made map available upon request from the authors).
The first, southernmost village, relatively often visited by tourists (10-15 trekking groups yearly) was Piliam (number of participants n = 37; all the names are given in the local spelling). Other mountain villages, relatively less frequented by tourists (listed from the closest to the farthest from Piliam) were: Pui (n = 5), Hiklahin (n = 36), Ohomtongo (n = 7), Sali (n = 2), Fik-Fak (n = 6), Mogobie (n = 3), Penyam (n = 3), and a few other small settlements, where single families lived (n = 7). As was indicated by the participants themselves, those villages were visited by trekking groups approximately one to three times in the last 5 years (according to the estimates of the participants, on average they saw nine tourist groups, SD = 4.8 during the period of 5 years before the study; data applies to all the villages taking part in the study).
All the inhabitants of the visited villages were invited to take part in the study (excluding children and the elderly). The number of positive responses to the invitation was equal to approximately 40% to 50%. In the end, 106 participants were recruited to participate in the study—53 females and 53 males. The female participants were aged between 25 and 59 years (M = 38.4, SD = 8.7) and male participants between 19 and 50 years (M = 35.6, SD = 7.6). The age was self-estimated—majority of the participants did not know their exact age and relied on rough estimates. The participants were interviewed by a Papuan assistant (from the Dani tribe) fluent in English and local Yali dialects. All participants were reimbursed for their participation by receiving the equivalent of US$3.
Procedure
Five independent judges were asked to rank 50 pictures with Papuan faces from the region of Baliem Valley (23 female and 27 male) from the palest to the darkest and then to choose an average. The pictures were taken from approximately 2 to 3 m in natural light. All pictures showed faces (similar to those taken for ID cards) with neutral expressions (or with a hint of a smile). Silhouette or clothing were not visible in the picture. The four male and four female faces chosen to be the average in terms of skin tone were both brightened and darkened by 25% to create two contrasting images (in Adobe Photoshop). Examples of the stimuli used in the study can be made available on request by emailing the first author.
Each participant was presented with four sets of three images of faces on a laminated photograph, with each image measuring 10 cm × 15 cm. The darkened (by 25%), average and brightened (by 25%) were shown simultaneously (it is important to mention that various methods can be used to brighten the skin color of the stimuli, as can be seen from the studies described in the introduction; we chose a method that can be easily replicated). When possible, the participants were tested individually.
Each Yali participant was asked to choose one face of the opposite sex, which in their opinion was the most attractive (as a representative of the opposite sex) for them. Prior to the main experiment, we conducted interviews with each of the participants in order to make sure that the experimental question and certain concepts were clear and easily understood. Each interview lasted relatively long, as participants were asked a series of questions relating to their biological data as well as the concept of attractiveness; the Yali answered questions regarding the concept of attractiveness in terms of what they find attractive in a person of the opposite sex; that is why we are fairly sure that each participant understood the procedure and answered to questions accurately.
During our research we controlled for other variables such as (a) the last time participant saw a tourist (b) distance from Wamena (number of days it takes the participant to walk to Wamena).
Yalimo seems to be a good place for conducting research concerned with changes in phenotype preferences under the influence of other cultures as it lies in a very remote location. Wamena is the biggest (approximately 10,000 people) village in Baliem Valley. It is a political, tourism, and trade center in this region. All tourist treks in Baliem Valley start and end in Wamena. The Yali villages are situated from two to three to about 14 days walking distance from Wamena (naturally Yali can travel that distance much faster than the “outsiders”). Also, some small-scale trading takes place between Yali villages and people who live in Wamena. Yali sell various objects, such as string bags or arrows in the city market, with salt, a highly prized commodity, often traded in return (Matthiessen, 1989), but at present Yali also buy tools available in Wamena shops. Apart from tourists, Yali also come into contact with Indonesians. In line with the Indonesian policy, increasing numbers of Indonesians (such as traders, clerks, and officials) migrate to settle down in Wamena. In conclusion, the bigger the distance from Wamena and the lesser the number of encounters with tourists, the lesser should be the influence of other cultures.
Results
Presented results describe the mean preferences for all male and all female faces. Distribution of the obtained results, both in male (χ2 = 9.8, df = 2, p = .007) and female ratings (χ2 = 14.7, df = 2, p < .001) differed from chance. Simultaneously, distributions of male and female choices did not differ between each other (χ2 = 0.9, df = 2, p = .64).
Both for male and female participants, the darkened face was the least likely choice (15.4% for female face, 18.5% for male face) in comparison to the average (40.4% for female face, 40.7% for male face) and brightened face (44.2% for female face, 40.7% for male face) (all p < .05; test between two proportions, Statistica software).
Observed Differences in Male and Female Preferences
To test the difference between preferences of males and females Analysis of Covariance (ANCOVA) was used with participants’ age and number of encountered tourists as controlled variables. We did not find any sex differences in this case; F(1, 100) = 0.8, p = .36, η2 = 0.01
Preferences Versus Contact With Other Cultures
For male participants, we observed a significant correlation between preferred skin tone (mean for four faces; from 1–bright, through 2–average, to 3–dark), and the last time they saw a tourist (r = 0.35, p = .005), and a similar significant correlation between preferred skin tone and the number of encounters with tourists (r = 0.38, p < .005). The correlation between preferred skin tone and the distance from Wamena was not significant, but on a trend level (r = 0.25, p = .069). We did not observe any correlation between the age of participants and preferred face (r = 0.05, p = .7).
For female participants (mean for four faces) we found no significant correlations between preferred skin tone and the age of participants (r = 0.05, p = .7), the number of encounters with tourists in the last 5 years (r = −0.20, p = .14), the last time they saw a tourist (r = 0.07, p = .6), or distance from Wamena (r = 0.13, p = .32).
Discussion
We found that for both Yali males and Yali females the preferred skin tone was either the average or the slightly lighter than average. Therefore, despite the high solar radiation in this region relatively light skin tone was preferred. At the same time, we did not find the expected differences between male and female preferences. We also observed an interesting and significant correlation between the indicator of acculturation and skin tone preference, but we found such relation only in males.
Classical sexual selection theory predicts that traits which yield an advantage in intrasexual competition for mates, or which enhance an individual’s attractiveness in the eyes of the members of the opposite sex, should be favored during the course of evolution (Darwin, 1871). As previous research suggests, paler coloring can be linked to fertility levels and youth (e.g., Fink, Grammer, & Matts, 2006; Garn et al., 1956) and therefore, average and paler faces should be regarded as more attractive only in women. In our study we did not observe such results.
Generally, the fact that we found no sex differences in terms of preferences for male and female faces might support the hypothesis concerned with the relationship between paler complexion and perceived higher social status (see: Hulse, 1967). Lighter skin color may also be a culturally based sign, indicating that individuals with paler complexion can afford more free time necessary to rest and who do not have to undertake manual outdoor labor. These two mechanisms do not cancel each other out and can act together.
On the other hand, if social status was crucial in shaping preferences for lighter skin tone, then it should be more important in men’s faces, since social status is generally more significant for men’s attractiveness than for women’s (e.g., Buss & Barnes, 1986). This could be one of the reasons why our participants preferred relatively lighter skin tone in men.
There is hardly any research available on the skin tone preference within non-Western populations, but there is a considerable body of evidence showing that in many Asian nations, light colored faces are preferred (Dixson et al., 2007; Swami et al., 2008; Yamaguchi & Aoki, 2008). Data available from dark-colored populations are less conclusive (Dixson et al., 2007, Swami et al., 2008). Given this considerable lack of data, it is worth gathering information from such sources as ethnographic literature. Some of these sources indicate that men with lighter skin tone are less sexually desirable than darker skinned men (Hogbin, 1946, pp. 197-198; Macintosh, 1959; Pospisil, 1958, p. 46), for instance among Western highland groups of Papua-New-Guinea, “the light-skinned boy does experience some mild chaffing” (Macintosh, 1959). Among the Kapauku, “a man who has red hair and lighter skin is mocked [by the women], although he may be generous, wealthy, and brave. The whiter men (albinos) can scarcely find a mate at all” (Pospisil, 1958, p. 46). What is more, Frost (2010, pp. 83-87) recalls further relevant data from anthropologists studying horticultural societies, where men tend to choose darker skinned women as mates, because they are thought to be harder workers than lighter skinned women.
Where do the differences between the above mentioned accounts and presented results stem from? Firstly, it is not quite clear what color of skin people of New Guinea meant, when interviewed by the researchers in previous studies, for example, in Macintosh (1959). The reports from Pospisil (1958) also referred to albinos, which may indicate that the locals were describing attitudes toward individuals whose skin color changed due to an illness. This is why the method used in our study is more accurate, as we know which skin tone was the baseline.
What is more, even if dark-skinned people are regarded as hardworking it does not automatically influence their attractiveness ratings. Similar results were reported by Ardener (1954), with regard to the Ibo from Nigeria, who tend to choose “‘yellow-skinned’ girls for wives as they are regarded as beauties,” despite the fact that these are the “‘black’ girls who are said to be harder workers.” Therefore, darker skinned women (and probably also men) are thought to be harder workers, and lighter skinned women and men can be regarded as having higher social status or in the women’s case they can be considered as more beautiful.
Generally, our results show that preference for skin color (in this case lighter skin color) can change relatively quickly, and can emerge following migration to a new cultural context. Yali men who had more frequent contact with Western and Indonesian people were more likely to prefer lighter skin tone, compared to men who rarely had any contact with other cultures and lived far away from Wamena. It suggests a cultural influence on the perception of skin tone, and similarly to body composition it seems to be of huge significance (e.g., Tovée, Swami, Furnham, & Mangalparsad, 2006). Yali have been confronted with missionaries, anthropologists, administrators, and tourists—all representing high social status, so it cannot be excluded that lighter skin might be perceived as being the most beautiful and desirable because of its association with power and wealth (hypothesis presented by social scientists, among them Bond & Cash, 1992; Hulse, 1967; Hunter, 2002).
However, if we were to hypothesize that correlation between preferred skin tone and contact with Western and Indonesian people was related to the perceived high status of the representatives of other cultures, then it would constitute an interesting issue due to the fact that such correlation was observed only among male participants. This can be explained in two ways. The first possibility is that Yali women have “always” had a preference for lighter skin tone in men (sexual selection hypothesis) and that is why these preferences have not changed under the influence of encounters with other cultures. The second explanation is that they might have adopted such preferences earlier than men, whose preferences have only began changing recently. We do not have much evidence for that, however it seems that in the Yali tribe men are more traditional and conservative—for example many young men still wear only traditional kotekas, while hardly any young woman wears the traditional skirt made of grass anymore.
Our study has several limitations. First of all, similar preferences for male and female skin tone can be related to the construction of the stimuli—which in its assumptions should express the sexual dimorphism in skin color of the studied population. Therefore, Yali participants, choosing both average female and average male skin tone in reality chose between slightly lighter female faces and slightly darker male faces. Even so, such subtle differences between the skin tones should not account for the obtained results suggesting an aversion toward darker skin tone. Moreover, due to technical limitations, we cannot guarantee that the images presented to the participants constitute ideal and accurate representations of the skin color of the people in the photographs. Regardless, we consider our stimuli to be a “step forward,” especially when comparing them to other available stimuli used in similar studies from previous years (e.g., B. J. Dixson et al., 2007; Swami et al., 2008).
Second of all, we need to mention that our sample was not entirely isolated and unaffected by the Western culture (as we wrote, Yali have been confronted with missionaries, anthropologists, administrators, and tourists). Although in comparison with other studies on perceived attractiveness of skin color, Yali still can be considered relatively unaffected.
Another limitation is that the chosen sample may not be the best selection for testing hypotheses concerned with the influence of sexual selection on preferences. This is due to the fact that a significant number of marriages are arranged. At this point it is important to mention that our main focus was on mate preferences and not mate choice. It is possible that Yali men might prefer lighter skinned women on aesthetic grounds, but this aesthetic preference would not necessarily translate into mate choice.
Finally, the explanations provided in this article were made “post factum” and therefore can only serve as an interpretation. It would be interesting to conduct a simple, qualitative study in which the Yali would be interviewed about their skin color preferences.
To summarize, the presented results demonstrate a preference for lighter skin tone for both males and females in a relatively remote, dark-skinned, indigenous population. At the same time, our study shows that these preferences might be influenced by the contact with other cultures. We can also speculate that women might be more likely to adopt Western standards of attractiveness than men (however, such conclusion requires further studies).
Footnotes
Acknowledgements
We would like to thank the Editor and anonymous reviewers for all valuable comments.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a Polish Ministry of Science and Higher Education scholarships and Polish Ministry of Science and Higher Education grant (Iuventus Plus) to P.S. and Foundation for Polish Science START scholarship to A.S.