Monday, April 30, 2018

The original meaning of skin color

Averaged female face (left) and averaged male face (right). (Dupuis-Roy et al. 2009)

At puberty the skin differentiates between the sexes, including its color. Men are browner and ruddier, having more melanin and blood in their skin. By comparison, women are "the fair sex" (Edwards and Duntley 1939; Edwards and Duntley 1949; Edwards et al. 1941; Frost 1988; Frost 2010; Frost 2011; Kalla 1973; Manning et al. 2004; Mazess, 1967; van den Berghe and Frost 1986). Women also display a higher luminous contrast between their facial skin and their lips and eyes (Dupuis-Roy et al. 2009).

In most Western societies this sex difference is scarcely noticeable, being overwhelmed by the much larger differences of race and ethnicity. It has been further obscured since the 1920s by the popularity of tanning among many Western women (Segrave 2005). Nonetheless, for most of human history and prehistory it has been the main reason why skin color varies in our immediate visual environment.

The human mind tends to hardwire any mental task that comes up repeatedly. It thereby shortens response time and eliminates learning time. One such task is to identify whether a person is a man or a woman by examining certain features, such as face shape and properties of the skin, including pigmentation. So when we see the minor pigmentary differences that distinguish men and women, does an innate mechanism process that visual information? This question takes us to research on the best-known example of hardwiring.

Face recognition, gender identification, and facial color

We have an innate ability to recognize human faces. This is shown by a form of brain damage called prosopagnosia, where one may seem normal and yet be no better at recognizing a face than any other object (Farah 1996; Pascalis and Kelly 2008; Zhu et al. 2009). At the other extreme are "super-recognizers" who are as good at face recognition as prosopagnosics are bad (Russell, Duchaine, and Nakayama 2009). 

This mental mechanism is sexually differentiated to some degree. It encompasses several neural populations, some of which specialize in male faces, others in female faces, and others in both kinds indifferently (Baudouin and Brochard. 2011; Bestelmeyer et al. 2008; Jacquet and Rhodes 2008; Little et al. 2005).

To tell male and female faces apart, this mechanism seems to use facial color (Bruce and Langton 1994; Hill, Bruce, and Akamatsu 1995; Russell and Sinha 2007; Russell et al. 2006; Tarr et al. 2001; Tarr, Rossion, and Doerschner 2002). The criteria are hue (brownness and ruddiness) and luminosity (lightness of the skin versus darkness of the lip/eye area). Hue is a fast "channel" for gender identification (Dupuis-Roy et al. 2009; Nestor and Tarr 2008a; Nestor and Tarr 2008b; Tarr et al., 2001; Tarr, Rossion, and Doerschner 2002). If the observer is too far away or the lighting too dim, the brain switches to the slower but more accurate channel of luminosity (Dupuis-Roy et al. 2009).

When shown a human face, subjects can tell its gender even if the image is blurred and differs only in color (Tarr et al. 2001). Indeed, facial color seems especially crucial under conditions of poor visibility when face shape is uncertain (Yip and Sinha 2002). 

The existence of a hardwired mental mechanism may explain not only why a certain schema of facial color is unthinkingly identified as female but also why women seek to accentuate this schema in a wide range of cultures. Thus, in different parts of the world, female cosmetics have shared the same aim of increasing the contrast between facial color and lip/eye color (Russell 2003; Russell 2009; Russell 2010). In a similarly wide range of cultures, women have tried to lighten their color by avoiding the sun and wearing protective clothing (Frost 2010, pp. 120-123). Going back to earliest times, we see that lighter skin was a female norm wherever the visual arts had developed—in ancient Greece, in ancient Egypt, in ancient China and Japan, and in Mesoamerica. All of these artistic traditions systematically gave a lighter coloring to female figures than to male figures (Capart 1905, pp. 26-27; Eaverly 1999; Soustelle 1970, p. 130; Tegner 1992; Wagatsuma 1967).

A cue for sexual interest

In addition to identifying gender, facial color can arouse sexual interest, being linked to gendered notions of attractiveness. In one study, women were asked to optimize the attractiveness of facial pictures by varying the skin's darkness and ruddiness. They made the male faces darker and ruddier than the female faces (Carrito et al. 2016). In another study, women were shown pairs of facial pictures where one face was slightly darker than the other, and they had to choose the most pleasing one. When male faces were shown, the darker face was more strongly preferred by women in the first two-thirds of their menstrual cycle (high estrogen/progesterone ratio) than by women in the last third (low estrogen/progesterone ratio). There was no cyclical effect if the women were judging female faces or taking oral contraceptives (Frost 1994).

The above findings are consistent with the results of a brain-imaging study: the female subjects had a stronger neural response to pictures of "masculinized" male faces, and this response correlated with their estrogen level across the menstrual cycle (Rupp et al. 2009). In a personal communication, the lead author stated that the faces had been masculinized by making them darker and more robust in shape.

A cue for modifying emotions and behavior

Facial color can elicit other responses. In word-association tests, the lighter complexion of women brings to mind such words as innocence, purity, peace, chastity, modesty, femininity, and delicacy (Gergen 1967; Wagatsuma 1967). This sort of response likewise emerged during interviews with Japanese men: "Whiteness is a symbol of women, distinguishing them from men." "Whiteness suggests purity and moral virtue." "One's mother-image is white" (Wagatsuma 1967, pp. 417-418).

Infants too are lighter-skinned (Grande et al. 1994; Kalla 1973; Post et al. 1976). They also share other visual, auditory, and tactile cues with the adult female body: a smaller nose and chin; a higher pitch of voice; and smoother, less hairy, and more pliable skin. This is what Konrad Lorenz dubbed the Kindchenschema, which seems to have the property of reducing aggressiveness in adults and eliciting care and nurturance (Frost 2010, pp. 134-135; Grande et al. 1994; Lorenz 1971, pp. 154-164). 

Infants are light-skinned in other primates. This is particularly so with langurs, baboons, and macaques, whose skin is pink in newborns and almost black in adults. The distinct infant coloration not only helps parents find wayward offspring but also elicits caregiving and defensive reactions. As it disappears with age, infants no longer attract the same interest and are less often sought out and held by adult females (Alley 1980; Alley 2014; Blaffer-Hrdy 2000, pp. 446-448; Booth 1962; Jay 1962). 

In humans, this infant coloration is striking in dark-skinned peoples. In Kenya, newborn children are often called mzungu ("European" in Swahili), and a new mother may tell her neighbors to come and see her mzungu (Walentowitz 2008). Among the Tuareg, children are said to be born "white" because of the freshness and moisture of the womb (Walentowitz 2008). The cause is often thought to be a previous spiritual life:

There is a rather widespread concept in Black Africa, according to which human beings, before "coming" into this world, dwell in heaven, where they are white. For, heaven itself is white and all the beings dwelling there are also white. Therefore the whiter a child is at birth, the more splendid it is. In other words, at that particular moment in a person's life, special importance is attached to the whiteness of his colour, which is endowed with exceptional qualities. (Zahan 1974, p. 385)

Another Africanist makes the same point: "black is thus the color of maturity [...] White on the other hand is a sign of the before-life and the after-life: the African newborn is light-skinned and the color of mourning is white kaolin" (Maertens 1978, p. 41).

Evolution of women's lighter skin

The above suggests that lighter coloration, as a social signal, went through four stages of evolution:

1. Initially, newborn primates were light-skinned because they had no need for pigment in the womb.

2. Adults recognized light skin as a mark of infancy. Selection then favored hardwiring of certain behavioral and emotional responses, particularly by females and to a lesser extent by all members of the local group. This mechanism could nonetheless be overridden by strange males that invade the group and kill the young (Alley 1980).

3. The same selection pressure caused infants to remain lighter-colored until they no longer had to be cared for. This was particularly so in those species where care of offspring was greater and lasted longer.

4. In humans, slower maturation, higher paternal investment, and longer-lasting pair bonds increased the risk of male neglect and aggression, thus creating a similar selection pressure and causing women to mimic key features of the Kindchenschema.

This evolutionary path was described by the ethologist Russell Guthrie:

I believe the sexual differences in skin color resulted from female whiteness being selected for because it is opposite the threat coloration, although the selection pressures may have been rather mild. Light skin seems to be more paedomorphic, since individuals of all races tend to darken with age. Even in the gorilla, the most heavily pigmented of the hominoids, the young are born with very little pigment. [...] Thus, a lighter colored individual may present a less threatening, more juvenile image. (Guthrie 1970)

From this perspective, women acquired a lighter color to modify rather than arouse sexual interest. This hypothesis is supported by a two-part study where men were first shown pictures of women and asked to rate their attractiveness. Lighter-skinned women were not rated more attractive than darker-skinned women. In the second part, eye movements were tracked, and it was found that lighter-skinned women were viewed for a longer time than darker-skinned women. The longer duration may indicate a slower rise and fall in sexual interest (Garza et al. 2016). 

By altering the trajectory of sexual interest, women's lighter skin may modify male behavior by dampening strong emotions, like aggression, and inducing feelings of care. This is perhaps a clue to why many women embraced the tanned look during the 20th century, in defiance of older norms of femininity. The new look enabled them to exploit an erotic sensibility that had earlier been stigmatized. In Victorian-era novels the "dark lady" is an "impetuous," "ardent," and "passionate" object of short-lived romances (Carpenter 1936, p. 254). Similarly, in French novels of the same period "[t]he love incarnated by brown women appears as the conceptual equivalent of a devouring femininity, thus making them similar to the mythical figure of Lilith" (Atzenhoffer, 2011, p. 6).This motif goes back at least to the Middle Ages in various European cultures and highlights an alternate form of eroticism:

[...] dark girls [...] are inevitably imagined as sexually more available than their fairer sisters, with whom they are implicitly or explicitly contrasted. In addition, the change of a girl's complexion, such as being burned by the sun, is to be understood as symbolic of her having crossed a sexual threshold without the benefit of marriage. (Vasvari 1999)

Identifying the brain regions that process facial color

There is a large body of research on the processing of facial color in the human mind, particularly on the brain regions involved. Thorstenson (2018) has reviewed the literature:

[...] there is considerable evidence suggesting that color is not merely an accessory of faces, but is rather a complex and crucial feature in facial processing. While classic  work  on  neural  processing  has  suggested  a  primary  cortical  area  responsible for face processing (FFA; Kanwisher, McDermott, & Chun, 1997) and a primary cortical area responsible for color processing (V4; McKeefry & Zeki, 1997), more recent work has revealed several areas in the temporal lobes specialized for face processing (Moeller, Freiwald, & Tsao, 2008; Tsao, Moeller, & Freiwald, 2008). Further, recent work has revealed consistent patterns of connected face and color selective cortical areas (Lafer-Sousa & Conway, 2013), possibly reflecting a shared overlap of visual processing between faces and color (Nakajima, Minami, Tanabe, Sadato, & Nakauchi, 2014; Stephen & Perrett, 2015). Additionally, the N170 component,  which  reflects  the  neural  processing  of  faces  in  event-related  potential  (ERP) studies, has been shown to respond to facial color information (Nakajima, Minami, & Nakauchi, 2012), but not to non-faces (Botzel & Grusser, 1989).

Thorstenson (2018) also reviews the possibilities for social signaling. Facial reddening is associated with anger and other intense emotions. Facial color can indicate certain disease states. Finally, there is a rise and fall in facial ruddiness and darkness over the menstrual cycle, with female facial color being lightest at ovulation.

Though providing a good review of the literature, Thorstenson should have mentioned three studies on female skin pigmentation over the menstrual cycle. McGuiness (1961) and Snell and Turner (1966) observed that facial skin darkens near the end of the cycle, particularly the peri-ocular skin of brunettes. Edwards and Duntley (1949) found that the buttocks visibly redden over the cycle, being lightest on the 13th day and darkest on the 25th day.


Today, skin color is seen through the lens of ethnic and racial conflict, yet this is not the sole meaning it has had for humans. For most of history and prehistory it was seen through a sexual lens, as a mark of masculinity or femininity.

This older meaning has received much less interest, even from academics. It is perhaps no coincidence that interest has come disproportionately from non-Western scholars like Hiroshi Wagatsuma, Kenichi Aoki, Mikiko Ashikari, and Aloke Kalla. In contrast, Western scholars, and Americans in particular, generally view the psychological meaning of skin color as a legacy of slavery.


Alley, T. R. (1980). Infantile colouration as an elicitor of caretaking behaviour in Old World primates. Primates 21(3): 416-429.   

Alley, T.R. (2014[1986]). An ecological analysis of the protection of primate infants. In V. McCabe and G.J. Balzano (eds). Event Cognition: An Ecological Perspective. Routledge. 

Aoki, K. (2002). Sexual selection as a cause of human skin colour variation: Darwin's hypothesis revisited. Annals of Human Biology 29(6): 589-608.   

Ashikari, M. (2005). Cultivating Japanese whiteness. The 'whitening' cosmetics boom and the Japanese identity. Journal of Material Culture 10(1):73-91.    

Atzenhoffer, R. (2011). Les hommes préfèrent les blondes. Les lectrices aussi. Effet de psychologie, horizons idéologiques et valeurs morales des héroïnes dans l'œuvre romanesque de H. Courths-Mahler, Colloque national (CNRIUT), Villeneuve d'Ascq, 8-10 juin 2009,   

Baudouin, J.-Y., and R. Brochard. (2011). Gender-based prototype formation in face recognition. Journal of Experimental Psychology: Learning, Memory, and Cognition 37(4): 888-898.    

Bestelmeyer, P.E.G., B.C. Jones, L.M. DeBruine, A.C. Little, D.I. Perrett, A. Schneider, L.L.M. Welling, C.A. Conway. (2008). Sex-contingent face aftereffects depend on perceptual category rather than structural encoding. Cognition 107(1): 353-365.    

Blaffer-Hrdy, S. (2000). Mother Nature: Maternal Instincts and How They Shape the Human Species. New York: Ballantine Books.

Booth, C. (1962). Some observations on behavior of Cercopithecus monkeys. Annals of the New York Academy of Sciences 102: 477-487.   

Bruce, V., and S. Langton. (1994). The use of pigmentation and shading information in recognising the sex and identities of faces. Perception 23(7): 803-822.   

Capart, J. (1905). Primitive Art in Egypt. London: H. Grevel.

Carpenter, F. I. (1936). Puritans preferred blondes. The heroines of Melville and Hawthorne. New England Quarterly 9(2): 253-272.   

Carrito, M.L., I.M.B. dos Santos, C.E. Lefevre, R.D. Whitehead, C.F. da Silva, and D.I. Perrett. (2016). The role of sexually dimorphic skin colour and shape in attractiveness of male faces. Evolution and Human Behavior 37(2): 125-33.    

Dupuis-Roy, N., I. Fortin, D. Fiset, and F. Gosselin. (2009). Uncovering gender discrimination cues in a realistic setting. Journal of Vision 9(2): 10, 1-8.    

Eaverly, M.A. (1999). Color and gender in ancient painting: A pan-Mediterranean approach. In N.L. Wicker and B. Arnold (eds). From the Ground Up: Beyond Gender Theory in Archaeology. Proceedings of the Fifth Gender and Archaeology Conference, University of Wisconsin-Milwaukee (pp. 5-10). Oxford (England): British Archaeological Reports.

Edwards, E.A., and S.Q. Duntley. (1939). The pigments and color of living human skin. American Journal of Anatomy 65(1): 1-33.    

Edwards, E.A., and S.Q. Duntley. (1949). Cutaneous vascular changes in women in reference to the menstrual cycle and ovariectomy. American Journal of Obstetrics & Gynecology 57(3): 501-509.    

Edwards, E.A., J.B. Hamilton, S.Q. Duntley, and G. Hubert. (1941). Cutaneous vascular and pigmentary changes in castrate and eunuchoid men. Endocrinology 28(1): 119-128.    

Farah, M. J. (1996). Is face recognition 'special'? Evidence from neuropsychology. Behavioural
Brain Research 76(1-2): 181-189.    

Frost, P. (1988). Human skin color: A possible relationship between its sexual dimorphism and its social perception. Perspectives in Biology and Medicine 32(1): 38-58.    

Frost, P. (2010). Femmes claires, hommes foncés. Les racines oubliées du colorisme. Quebec City: Les Presses de l'Université Laval, 202 p.

Frost, P. (2011). Hue and luminosity of human skin: a visual cue for gender recognition and other mental tasks. Human Ethology Bulletin 26(2): 25-34.    
Garza, R., R.R. Heredia, and A.B. Cieslicka. (2016). Male and female perception of physical attractiveness. An eye movement study. Evolutionary Psychology 14(1): 1-16.    

Gergen, K.J. (1967). The significance of skin color in human relations. Daedalus 96: 390-406.  

Grande, R., E. Gutierrez, E. Latorre, and F. Arguelles. (1994). Physiological variations in the pigmentation of newborn infants. Human Biology 66(3): 495-507.
Guthrie, R.D. (1970). Evolution of human threat display organs. in T. Dobzhansky, M.K. Hecht, and W.C. Steere (eds). Evolutionary Biology 4: 257-302. New York: Appleton-Century Crofts.

Hill, H., V. Bruce, and S. Akamatsu. (1995). Perceiving the sex and race of faces: The role of shape and colour. Proceedings of the Royal Society B: Biological Sciences 261(1362): 367-373.    

Jaquet, E., and G. Rhodes. (2008). Face aftereffects indicate dissociable, but not distinct, coding of male and female faces. Journal of Experimental Psychology: Human Perception and Performance 34(1): 101-112.    

Jay, P. C. (1962). Aspects of maternal behavior among langurs. Annals of the New York Academy of Sciences 102(2): 468-476.   

Kalla, A.K. (1973). Ageing and sex differences in human skin pigmentation. Zeitschrift für Morphologie und Anthropologie 65(1): 29-33.   

Little, A.C., L.M. DeBruine, and B.C. Jones. (2005). Sex-contingent face after-effects suggest distinct neural populations code male and female faces. Proceedings of the Royal Society of London, Series B 272(1578): 2283-2287.    

Lorenz, K. 1971. Studies in Animal and Human Behaviour, vol. 2. London: Methuen & Co.

Maertens, J-T. (1978). Le dessein sur la peau. Essai d'anthropologie des inscriptions tégumentaires, Ritologiques I, Paris: Aubier Montaigne.

Manning, J.T., P.E. Bundred, and F.M. Mather. (2004). Second to fourth digit ratio, sexual selection, and skin colour. Evolution and Human Behavior 25(1): 38-50.  

Mazess, R.B. (1967). Skin color in Bahamian Negroes. Human Biology 39(2): 145-154.    

McGuiness, B.W. (1961). Skin pigmentation and the menstrual cycle. British Medical Journal 2(5251): 563-565.   

Nestor, A., and M.J. Tarr. (2008a). The segmental structure of faces and its use in gender recognition. Journal of Vision 8(7): 7, 1-12,   

Nestor, A., and M.J. Tarr. (2008b). Gender recognition of human faces using color. Psychological Science 19(12): 1242-1246.   

Pascalis, O., and D.J. Kelly. (2008). Face processing, in M. Haith and J. Benson (eds), Encyclopedia of Infant and Early Childhood Development. (pp. 471-478), Elsevier.

Post, P.W., A.N. Krauss, S. Waldman, and Peter A.M. Auld. (1976). Skin reflectance of newborn infants from 25 to 44 weeks gestational age. Human Biology 48(3): 541-557.

Rupp, H.A., T.W. James, E.D. Ketterson, D.R. Sengelaub, E. Janssen, and J.R. Heiman. (2009). Neural activation in women in response to masculinized male faces: mediation by hormones and psychosexual factors. Evolution and Human Behavior 30(1): 1-10.   

Russell, R. (2003). Sex, beauty, and the relative luminance of facial features. Perception 32(9): 1093-1107.   

Russell, R. (2009). A sex difference in facial pigmentation and its exaggeration by cosmetics. Perception 38(8): 1211-1219.   

Russell, R. (2010). Why cosmetics work. In R. Adams, N. Ambady, K. Nakayama, and S. Shimojo (eds.) The science of social vision. New York: Oxford.

Russell, R., B. Duchaine, and K. Nakayama. (2009). Super-recognizers: People with extraordinary face recognition ability. Psychonomic Bulletin & Review 16(2): 252-257.  

Russell, R. and P. Sinha. (2007). Real-world face recognition: The importance of surface reflectance properties. Perception 36(9): 1368-1374.   

Segrave, K. (2005). Suntanning in 20th Century America. Jefferson (North Carolina): McFarland & Company.

Snell, R.S., and R. Turner. (1966). Skin pigmentation in relation to the menstrual cycle. The Journal of Investigative Dermatology 47(2): 147-155.  

Soustelle, J. (1970). The Daily Life of the Aztecs. Stanford, California: Stanford University Press.

Tarr, M.J., D. Kersten, Y. Cheng, and B. Rossion. (2001). It's Pat! Sexing faces using only red and green. Journal of Vision 1(3): 337, 337a.   

Tarr, M. J., B. Rossion, and K. Doerschner. (2002). Men are from Mars, women are from Venus: Behavioral and neural correlates of face sexing using color. Journal of Vision 2(7): 598, 598a,   

Tegner, E. (1992). Sex differences in skin pigmentation illustrated in art. The American Journal of Dermatopathology 14(3): 283-87.  

Thorstenson, C.A. (2018). The social psychophysics of human face color: Review and recommendations. Social Cognition 36(2): 247-273.   
van den Berghe, P.L., and P. Frost. (1986). Skin color preference, sexual dimorphism and sexual selection: A case of gene-culture co-evolution? Ethnic and Racial Studies 9(1): 87-113.  

Vasvari, L.O. (1999). A comparative approach to European folk poetry and the erotic wedding motif, CLCWeb. Comparative Literature and Culture 1(4)  

Wagatsuma, H. (1967). The social perception of skin color in Japan. Daedalus 96(2): 407-443.   

Walentowitz, S. (2008). Des êtres à peaufiner. Variations de la coloration et de la pigmentation du nouveau-né, in J-P. Albert, B. Andrieu, P. Blanchard, G. Boëtsch, and D. Chevé (eds.) Coloris Corpus (pp. 113-120), Paris: CNRS Éditions, 2008.

Yip, A.W., and P. Sinha. (2002). Contribution of color to face recognition. Perception 31(8): 995-1003.  

Zahan, D. (1974). White, Red and Black: Colour Symbolism in Black Africa, in A. Portmann and R. Ritsema (eds.) The Realms of Colour, Eranos 41 (1972), 365-395, Leiden: Eranos.

Zhu, Q., Y. Song, S. Hu, X. Li, M. Tian, Z. Zhen, Q. Dong, N. Kanwisher, and J. Liu. (2009). Heritability of the specific cognitive ability of face perception. Current Biology 20(2): 137-142.   

Monday, April 23, 2018

Debate over recent human evolution: pros and cons

Acceleration of recent human evolution. Age distribution of alleles under selection (Hawks et al. 2007)

A decade has passed since a research team led by John Hawks published a strange finding: human genetic evolution accelerated more than a hundred-fold some 10,000 years ago. This was when hunting and gathering began to give way to farming, which in turn brought other changes, all of which required adjustments to mind and body. All in all, new cultural and natural environments have reshaped 7% of the human genome over the last 40,000 years:

Some of the most radical new selective pressures have been associated with the transition to agriculture. For example, genes related to disease resistance are among the inferred functional classes most likely to show evidence of recent positive selection. Virulent epidemic diseases, including smallpox, malaria, yellow fever, typhus, and cholera, became important causes of mortality after the origin and spread of agriculture. Likewise, subsistence and dietary changes have led to selection on genes such as lactase. (Hawks et al. 2007)

Instead of adapting only to the natural environment, humans have adapted to cultural creations of their own making, things like prepared food, clothing, shelter, way of life, social organization, sedentary versus nomadic living, religious strictures, and so on.

This finding may come as a surprise. As a university student I learned that culture has greatly reduced the importance of natural selection in our species. Instead of adapting genetically to our environment, we adapt culturally. That was, and still is, the normative view.

Debates in the scientific literature: 2008 to 2010

So what are we to believe? Perhaps there have been other findings over the last decade, either pro or con.

In 2008, a research team led by Matthieu Foll and Oscar Gaggiotti calculated a higher estimate of recent human evolution: over 23% of the human genome. By using an FST test and data from 53 human populations, they found evidence for selection at 131 out of 560 random loci. When this methodology was repeated with other random loci, the same estimate of 23% came up.

A review paper by Joshua Akey notes, however, that these genome-wide scans are problematic in two ways. On the one hand, they miss genes that are known to have contributed to recent human evolution. On the other, these different scans disagree on the regions of the human genome that have been evolving rapidly:

Strikingly, only 722 regions (14.1%) were identified in two or more studies, 271 regions (5.3%) were identified in three or more studies, and 129 regions (2.5%) were identified in four or more studies (Fig. 1). Furthermore, the integrated map of positive selection does not include several of the most compelling genes with well-substantiated claims of positive selection, such as G6PD and DARC. (Akey 2009)

A closer look at the data suggests that recent evolution is highly localized on the human genome. If the size of the region is decreased, the probability increases of that region containing either no genes at all under selection or several under selection. Making the regions smaller makes it easier, strangely enough, to find regions with multiple genes under selection. "This paradoxical observation [...] is due to the marked difference in the average size of regions identified in single versus multiple studies (~80 kb and 300 kb, respectively)" (Akey 2009).

So estimates of recent human evolution seem to range from a low of 7% of the genome (Hawks et al. 2007) to a high of 23% (Foll and Gaggiotti 2008). Even the 7% estimate, however, has been criticized in the literature, specifically by two papers. The first one was Pickrell et al. (2009):

We find that putatively selected haplotypes tend to be shared among geographically close populations. [...]. This suggests that distinguishing true cases of selection from the tails of the neutral distribution may be more difficult than sometimes assumed, and raises the possibility that many loci identified as being under selection in genome scans of this kind may be false positives. Reports of ubiquitous strong (s = 1 - 5%) positive selection in the human genome (Hawks et al. 2007) may be considerably overstated. (Pickrell et al. 2009)

The argument here is that a genetic variant with high selective value should spread beyond its area of origin, instead of remaining bottled up there. Yet this is unlikely for two reasons. First, recent variants, by definition, have little time to spread very far. Second, and more importantly, the selective value of a genetic variant is a function of its natural and cultural environment. A variant that succeeds in one environment will be less successful in another.

The criticism made by Pickrell et al. (2009) was repeated by Hermisson (2009). If the data are controlled for geographic region, the evidence for recent human evolution virtually disappears:

[...] introduction of hierarchical structure based on five previously established geographic regions reduces the frequency of selection candidates from 23% (Foll and Gaggiotti, 2008) to no more than expected by chance (that is, comparable with the 1% significance level applied). (Hermisson 2009)

The implication is that recent human evolution is largely due to founder effects and other forms of genetic drift. Genetic drift, however, would not produce the observed signatures of natural selection, as Nicholas Wade noted in a review of this research the following year:

One of the signatures of natural selection is that it disturbs the undergrowth of mutations that are always accumulating along the genome. As a favored version of a gene becomes more common in a population, genomes will look increasingly alike in and around the gene. Because variation is brushed away, the favored gene's rise in popularity is called a sweep. Geneticists have developed several statistical methods for detecting sweeps, and hence of natural selection in action. (Wade 2010).

Moreover, this signature is much stronger in some geographic regions than in others:

A new approach to identifying selected genes has been developed by Anna Di Rienzo at the University of Chicago. Instead of looking at the genome and seeing what turns up, Dr. Di Rienzo and colleagues have started with genes that would be likely to change as people adopted different environments, modes of subsistence and diets, and then checked to see if different populations have responded accordingly.

She found particularly strong signals of selection in populations that live in polar regions, in people who live by foraging, and in people whose diets are rich in roots and tubers. [..] The fewest signals of selection were seen among people who live in the humid tropics, the ecoregion where the ancestral human population evolved. [...] there seem to be more genes under recent selection in East Asians and Europeans than in Africans, possibly because the people who left Africa were then forced to adapt to different environments. "It's a reasonable inference that non-Africans were becoming exposed to a wide variety of novel climates," says Dr. Stoneking of the Max Planck Institute. (Wade 2010)

Joshua Akey remains cautious on this point:

A specific example of the difficulties in interpreting signatures of spatially varying selection is the observation that non-African populations tend to show more evidence for recent positive selection relative to African populations (Akey et al. 2004; Storz et al. 2004; Williamson et al. 2007; but see Voight et al. 2006). While this may be due to increased selection as humans migrated out of Africa and were confronted with new environmental pressures (such as novel climates, diets, and pathogens), differences in demographic history or rates of recombination and mutation between African and non-African populations may obscure the relationship between signatures of selection across populations. (Akey 2009)

Since 2010: consensus among some, skepticism and hostility among others

After 2010, Google Scholar turns up only brief references to the original paper by John Hawks et al., most of them favorable or neutral in tone. If one judges by the scientific literature alone, there seems to be broad support for the notion that recent evolution has accelerated in our species. And the original estimate of 7% recent evolutionary change may err on the low side

Yet many people remain unconvinced. Last week Razib Khan reproached me: "you take the accelerationist hypothesis as a given. it's not. at least at that magnitude (i think most ppl agree holocene resulted in faster rate of change)." Indeed, most people seem to view these findings with incredulity, to put it mildly, as a journalist from Discover magazine found:

Not surprisingly, the new findings have raised hackles. Some scientists are alarmed by claims of ethnic differences in temperament and intelligence, fearing that they will inflame racial sensitivities. Other researchers point to limitations in the data. Yet even skeptics now admit that some human traits, at least, are evolving rapidly, challenging yesterday's hallowed beliefs. (McAuliffe 2009)

A decade later, the barriers to acceptance are still considerable. Chen et al. (2016) identifies four sources of opposition:

- Evolutionary psychologists, who believe that human nature took shape in the Pleistocene. According to this view, genetic influences on behavior are too complex to have changed much since then.

- Cultural determinists, who believe that "once humans invented culture, natural selection was halted because humans could overcome nature through culture."

- People who point out that we are all 99.9% genetically alike. So there is little room for genetic differences within our species.

- People who believe that genetic differences are inconsequential to human behavior.

There are counter-arguments to the above. Genetic influences on existing behaviors can evolve very fast (Harpending and Cochran 2002). And that figure of 99.9% genetic identity is an over-estimate, the best estimate being 99%. Even if we assume that this 1% difference is spread evenly across the genome, that tiny difference could significantly alter the way each and every gene works.

Nonetheless, such counter-arguments would still leave many unconvinced. And others wouldn't even listen. Some beliefs are foundational, being difficult to challenge without seeming to attack an entire worldview. In such cases, reactions can be nasty.

That's normal. Strong disagreement is the stuff of scientific debate. What's less normal is that some people will seek not to debate but to judge and punish. That fate befell a coauthor of the 2007 paper on recent human evolution. In 2015, the Southern Poverty Law Center (SPLC) prepared and published a file on Henry Harpending ... under the heading "Extremist Info." The opening words sounded no less ominous:

Henry Harpending is a controversial anthropologist at the University of Utah who studies human evolution and, in his words, "genetic diversity within and between human populations."

The file went on to state:

Harpending is most famous for his book, co-authored with frequent collaborator Gregory Cochran, The 10,000 Year Explosion: How Civilization Accelerated Human Evolution, which argues that humans are evolving at an accelerating rate, and that this began when the ancestors of modern Europeans and Asians left Africa. (SPLC 2015)

One wonders what exactly is intended by this public naming and shaming. After all, the SPLC has no legal mandate to judge and punish, although it seems to think so. Indeed, it acts like a law-enforcement agency without being constrained by the law and without being answerable to an elected body.

Henry Harpending died scarcely a year later, yet his file is still there on the SPLC website. Even in death he's still a grave threat … as is apparently anyone else who believes in the evidence for recent human evolution.


Akey, J.M. (2009). Constructing genomic maps of positive selection in humans: Where do we go from here? Genome Research 19: 711-722.

Chen, C., R.K. Moyzis, X. Lei, C. Chen, and Q. Dong. (2016). "The enculturated genome: Molecular evidence for recent divergent evolution in human neurotransmitter genes." In Joan Y. Chiao, Shu-Chen Li, Rebecca Seligman, Robert Turner (eds). The Oxford Handbook of Cultural Neuroscience. Oxford.

Cochran, G. and H. Harpending. (2010). The 10,000 Year Explosion: How Civilization Accelerated Human Evolution, New York: Basic Books.

Foll, M., and O. Gaggiotti. (2008). A Genome-Scan Method to Identify Selected Loci Appropriate for Both Dominant and Codominant Markers: A Bayesian Perspective. Genetics 180(2):977-993.

Harpending, H., and G. Cochran, (2002). In our genes, Proceedings of the National Academy of Science. USA. 99(1):10-12.

Hawks, J., E.T. Wang, G.M. Cochran, H.C. Harpending, and R.K. Moyzis. (2007). Recent acceleration of human adaptive evolution, Proceedings of the National Academy of Science USA 104:20753-20758.

Hermisson, J. (2009). Who believes in whole-genome scans for selection? Heredity 103, 283-284  

McAuliffe K. (2009). They don't make Homo sapiens like they used to. Our species-and individual races-have recently made big evolutionary changes to adjust to new pressures. Discover February 9

Pickrell, J.K., G. Coop, J. Novembre, S. Kudaravalli, J.Z. Li, D. Absher, B.S. Srinivasan, G.S. Barsh, R.M. Myers, M.W. Feldman, and J.K. Pritchard. (2009). Signals of recent positive selection in a worldwide sample of human populations. Genome Research 19(5): 826-837  

SPLC. (2015). Henry Harpending. Extremist Info

Wade, N. (2010). Adventures in very recent evolution. The New York Times, July 19

Monday, April 16, 2018


Reconstruction of a Mesolithic camp (Wikicommons, David Hawgood). Hunter-gatherers often slept in temporary shelters and were generally more exposed to the cold.

My last post generated many intelligent comments on Twitter. Here are my replies to each of them:

Alissa Mittnik - Department of Archaeogenetics, Max Planck Institute for the Science of Human History

That's why most of the aDNA studies you cite do not rely on those but use several 100Ks of polymorphic loci on the autosomes that are not functionally relevant, but whose variable frequencies across populations reflect their different histories of isolation and admixture.

Haplogroup U was once considered to be functionally irrelevant. Even if a gene seems to be noncoding "junk," it can still regulate what other genes do. The Drosophila genome has shown the functional value of noncoding genes:

There is now a wealth of evidence that some of the most important regions of the genome are found outside those that encode proteins, and noncoding regions of the genome have been shown to be subject to substantial levels of selective constraint, particularly in Drosophila. Recent work has suggested that these regions may also have been subject to the action of positive selection, with large fractions of noncoding divergence having been driven to fixation by adaptive evolution. [...] Here, we examine patterns of evolution at several classes of noncoding DNA in D. simulans and find that all noncoding DNA is subject to the action of negative selection, indicated by reduced levels of polymorphism and divergence and a skew in the frequency spectrum toward rare variants. (Haddrill et al. 2008)

According to a recent study, most of the human genome has some kind of function, even the noncoding regions. "These data enabled us to assign biochemical functions for 80% of the genome, in particular outside of the well-studied protein-coding regions" (The ENCODE Project Consortium 2012).

It is a myth to believe that noncoding DNA is mostly “junk.” In fact, human evolutionary change has largely occurred in that kind of DNA, apparently as a means to alter the development of complex structures like the brain:

[...] a systematic search for human-specific deletions compared with other primate genomes identified 510 such deletions in humans that fall almost exclusively in noncoding regions.

[…] Another evolutionary approach has been to focus on genomic loci that are well conserved throughout vertebrate evolution but are strikingly different in humans; these regions have been named "human accelerated regions (HARs)" [...]. So far, ~2,700 HARs have been identified, again most of them in noncoding regions: at least ~250 of these HARs seem to function as developmental enhancers in the brain. (Bae et al. 2015)

The same authors note that it is easier to determine the function of coding DNA; hence, the widespread perception that noncoding DNA serves no purpose:

It is relatively easy to detect and understand the functional consequences of changes in protein-coding sequences, compared to noncoding mutations. Mutations in a coding sequence often cause more severe phenotypes than mutations in a regulatory element associated with the same coding sequence. (Bae et al. 2015)

Alissa Mittnik

Turning around the hg U argument, one could make the case that the environmental conditions that farmers of Anatolian ancestry faced in northern Europe led to selective pressures which increased "hunter-gathererlike" functional variants (maybe introgressed) in their population. Which might lead us to underestimate the proportion of Anatolian farmer admixture.

By "environmental conditions" you seem to be referring only to the natural environment. There is also the cultural environment.

Recent human evolution has been primarily in response to the cultural environment. This may be seen in the hundred-fold acceleration of genetic change 10,000 years ago, when our ancestors began to shift from hunting and gathering to farming (Cochran and Harpending 2010; Hawks et al. 2007). By that time, humans had already spread from the tropics to the arctic. They were now adapting to new cultural environments of their own making, and not simply to existing natural environments.

Adaptation to farming was physiological, behavioral, and mental. I mentioned energy balance. Less energy was needed for body heat because sleeping environments were warmer, as were daytime environments in general. A farmer could choose the best time of day to go out into the fields. A hunter had much less choice. He could give up chasing his prey, and go home empty-handed, or he could continue chasing it hither and thither until he finally got it.

There were also mental adaptations, with some capacities being reduced. A hunter had to memorize huge quantities of spatiotemporal data for several purposes: tracking prey over time and space; predicting where they might go; charting the best path to get there; and remembering how to go home. Getting lost could be fatal, since a hunter could not always live off the land, especially in winter. This is why meat was stored in caches, whose locations likewise had to be remembered. All of that memory storage became obsolete when early Europeans became farmers. As the need for spatiotemporal memory decreased between the Mesolithic and the Neolithic, there was a corresponding reduction in cranial size (Henneberg 1988).

The Mesolithic-Neolithic transition led to reduction in other mental demands. There was less need to recognize odors (Majid and Kruspe 2018) and less need for monotony avoidance and sensation seeking (Zuckerman 2008). Meanwhile, there was a greater need to process reciprocal obligations with a larger number of people while interacting less, on average, with each person.

In sum, it is no trivial matter to go from hunting and gathering to farming. These are two very different ways of life with different demands on the mind and body. Much readjustment is needed to make the transition from one to the other.

All right. For the sake of argument, let’s assume that genetic change has been primarily in response to the natural environment. As Anatolian farmers advanced farther into northern Europe, they adapted to a colder climate by allocating more energy to body heat. To this end, they acquired functional variants like haplogroup U, perhaps through introgression. Natural selection then raised their incidence of haplogroup U to higher and higher levels.

But … that's … not … what … happened. Haplogroup U went into decline after farming came and is now rare in northern Europeans. So this is not even a "just so" story. This is an "ain't so" story. In reality, farmers could control their living conditions by building warmer homes, by spending more time indoors, and by planning when they went outdoors. Hunter-gatherers had less control, often having to stay out in the worst weather.

Alissa Mittnik

You also say WHG is a genetic dead end, which is definitely not true, WHG is one of the distinct ancestral source populations for modern Europeans. In fact, East Baltic HGs are genetically WHGs.

Brace et al. (2018) argue that early British farmers had about a 10% residue from native hunter-gatherers. Of course, those farmers also had admixture from WHGs on the continent. So the total residue is higher, all the more so without the genetic change that is wrongly attributed to admixture. So I stand corrected: WHGs did make a contribution to the present-day gene pool.

My basic point is that farmers replaced hunter-gatherers much more in western Europe than in northern Europe. In western Europe, hunter-gatherers had very low population densities, being small bands of nomads. In northern Europe, especially around the North Sea and the Baltic, they were able to achieve much higher population densities by exploiting marine resources. Consequently, those hunter-fisher-gatherers suffered less population replacement because they were too numerous to replace.

I disagree with your second point. East Baltic HGs seem to be closest to Scandinavian HGs. They show the same phenotype of fair skin and a variety of hair and eye colors. WHGs had a different phenotype: dark skin, dark hair, and blue eyes.

Iosif Lazaridis - Department of Genetics, Harvard Medical School

"Lazaridis et al. (2014) estimated Anatolian farmer admixture in East Baltic peoples at 30%."

There were no Anatolian farmers known at the time, so I doubt we estimated Anatolian farmer admixture; also model did not account for Yamnaya ancestry (also unsampled at the time). In Haak, Lazaridis et al. (2015) we estimate 17.4% LBK_EN ancestry in Lithuanians. Given that LBK_EN is ~10% WHG, this translates to ~15% Anatolian ancestry which seems about right.

So East Baltic peoples have ~15% Anatolian ancestry. That figure is considerably lower than the estimate of 52% for northwest Europeans (Skoglund et al. 2012). Such a difference in ancestry would surely produce a visible difference in the way people look.

Does it? Can you identify a Latvian in a room full of Dutch people? Let’s put aside the mathematical models, and their unstated assumptions. Does such a difference in ancestry seem plausible?

Razib Khan - (geneticist and science writer)

didn't read your whole piece in detail. 2 comments 1) u overread from SNP data on pig[mentation]. gen background matters for blondism in KITLG. my 2 sons r heterozygote (like 25% of Scandinavians) have brown hair. 2) ppl in the reich lab don't think SHG contributed ancestors to later ppl

Variation in hair color is determined mainly by alleles at MC1R, and these were the alleles that Günter et al. (2018) measured in their study of ancient DNA from Scandinavian hunter-gatherers. An SNP close to KITLG (rs12821256) plays a measurable but secondary role in hair color variation (Sulem et al. 2007). Using this and other loci would provide a finer-grained simulation of hair color in early Scandinavians, but the overall picture is already clear.

I'm sure the folks at David Reich's lab exclude natural selection from their mathematical models. When I was a university student I learned the normative view that culture has greatly reduced the importance of natural selection in our species. Instead of adapting genetically to our environment, we adapt culturally. In reality, culture has accelerated human evolution by creating human-made environments, each of which requires its own set of adaptations (Cochran and Harpending 2010; Hawks et al. 2007). Instead of adapting only to climate, wildlife, and vegetation, we have had to adapt to diet, clothing, shelter, way of life, social organization, sedentary versus nomadic living, religious strictures, and so on.

That is a very different view of things, and my impression is that most academics are still working with the old view.


Narva was a technically in the SHG group and it contributed ~10% to Corded Ware. About decreasing U, it can be both to the introduction of new mtDNA from both Anatolia and the Steppe, but also normal selection against it due to its heat/atp balance.

If the incidence of haplogroup U decreased partly because of Anatolian admixture, we should see a steeper decline when farming was first introduced and a gentler decline thereafter (as a result of natural selection). Instead, we see a steady decline throughout the Neolithic and post-Neolithic.

Hernan Cortes

did the corresponding hunter gatherer Y chromosome decrease at same rate?

As far as I know (and I'm willing to stand corrected), the decrease in the incidence of haplogroup U was the single largest genetic change associated with the transition from hunting and gathering to farming. I'm using the word "associated" liberally because this change continued long past the actual transition.


Bae, B-I., D. Jayaraman, and C.A. Walsh. (2015). Genetic changes shaping the human brain, Developmental Cell 32(4): 423-434.

Brace, S., Y. Diekmann, T.J. Booth, Z. Faltyskova, N. Rohland, S. Mallick, et al. (2018). Population replacement in early Neolithic Britain, BioRxiv, February 18.  

Cochran, G. and H. Harpending. (2010). The 10,000 Year Explosion: How Civilization Accelerated Human Evolution, New York: Basic Books.

Günther, T., H. Malmström, E.M. Svensson, A. Omrak, F. Sánchez-Quinto, G.M. Kilinç, et al. (2018). Population genomics of Mesolithic Scandinavia: Investigating early postglacial migration routes and high-latitude adaptation. PLoS Biol 16(1): e2003703.     

Haddrill, P.R., D. Bachtrog, and P. Andolfatto. (2008). Positive and Negative Selection on Noncoding DNA in Drosophila simulans, Molecular Biology and Evolution 25(9): 1825-1834  

Hawks, J., E.T. Wang, G.M. Cochran, H.C. Harpending, and R.K. Moyzis. (2007). Recent acceleration of human adaptive evolution, Proceedings of the National Academy of Science USA 104:20753-20758.   
Henneberg, M. (1988). Decrease of human skull size in the Holocene, Human Biology 60(3): 395-405.  

Lazaridis, I., N. Patterson, A. Mittnik, G. Renaud, S. Mallick, K. Kirsanow, et al. (2014). Ancient human genomes suggest three ancestral populations for present-day Europeans, Nature 513(7518): 409-413    

Majid, A., and N. Kruspe. (2018). Hunter-gatherer olfaction is special, Current Biology 28(3): R108-R110.   

Skoglund, P., H. Malmström, M. Raghavan, J. Storå, P. Hall,  E. Willerslev, M.T. Gilbert, A. Götherström, and M. Jakobsson. (2012). Origins and genetic legacy of Neolithic farmers and hunter-gatherers in Europe, Science 336:466-469.  

Sulem, P., D.F. Gudbjartsson, S.N. Stacey, A. Helgason, T. Rafnar, K.P. Magnusson, et al. (2007). Genetic determinants of hair, eye and skin pigmentation in Europeans, Nature Genetics 39(12): 1443-1452.  

The ENCODE Project Consortium (2012). An integrated encyclopedia of DNA elements in the human genome, Nature 489: 57-74   

Zuckerman, M. (2008). "Genetics of Sensation Seeking," (pp. 193- 210) in J. Benjamin, R.P. Ebstein, and R.H. Belmaker (eds) Molecular Genetics and the Human Personality, Washington D.D.: American Psychiatric Publishing Inc.