Several years ago Oprah Winfrey asked Tiger Woods what he would say to people who say that when they look him they see a black man. The issue was that some African Americans objected to Woods’ contention that he was multiracial, Cablinasian, which reflected the fact that he was ancestrally 1/2 Asian, 1/4 African, 1/8 European and 1/8 Native American. Woods is also a Therevada Buddhist by religion, taking after his Thai nationality (though mixed-race) mother, so one can argue he is quite Asian culturally. I know many people who frankly disagree with Winfrey’s assessment, that is, that Woods looks “like a black man.” Some Asian Americans have stated that “from the eyes down” he looks Asian, and frankly I can see that. Tiger Woods’ hair reflects his mixed ancestry, and it is neither straight nor kinky, but curly (he cuts it short enough that this doesn’t manifest normally, but I know people who played golf with him when he was a teenager and they attest to this). Some of my friends in college who were Asian American activists quite loudly would proclaim that Tiger Woods “looks Asian” when the topic came up.
Does Tiger Woods look Asian? Does he look like a “black man”? I am posting on this because of some responses below which asked how it was that mixed-race children can so favor one race. But to start out with I need to offer that perceptions of race are as much a matter of psychology and culture as they are of genetics. I bring up Tiger Woods because black Americans and Asian Americans might have radically different viewpoints about what race he looks like, showing that where you start from shapes where you end up at.
Part of the issue is that in the United States we accept hypodescent, which means that one drop of black blood or any hint of black features can imply that one is fully and totally black. Consider Jason Kidd, the point guard for the New Jersey Nets. He is a black man in the United States because his father is a black man (his mother is white). In fact, from the interviews and pictures I’ve seen of his father Jason Kidd is likely well below 1/2 black in ancestry, because his father looks to be mixed genetically. In the United States though people who have noticeable black ancestry are considered black. Does Jason Kidd have noticeable black ancestry? A few years ago I recall reading a story about a reporter who was in Russia where Jason Kidd was described as a white basketball player. The reporter was startled, but he had to admit that if you didn’t know Kidd’s father was African American you might not necessarily assume he was African American, especially with his hair cut short. Cultural priors matter, and in the United States we give great weight to black ancestry as determinative of one’s race. Sometimes this can get ridiculous. In Our Kind of People: Inside America’s Black Upper Class the author recounts a story where at his brother’s wedding a blonde white woman at his table started referring to “n*gger practices.” No one batted an eye and the author was shocked, but later it was explained that the woman was black and she had gone to Howard University. Of course, it is pretty obvious here that the woman did not look black, but her black ancestry and identification allowed her to use a word which is reserved for black people in the United States. Outside of American culture this makes less sense to people (in Latin America they have literally hundreds of racial terms which apply to every permutation of ancestry and appearance, which results in siblings being members of different races according to their looks).
Finally, there’s the issue of the priors you start out with. There are many people whose physical appearance are explicable by their background. For example, say your friend is an olive-skinned individual with somewhat curly hair and green eyes. You know that he has a black American grandparent. That explains their features, it is the black admixture! Now, say that this individual had an Arab name. Ah, that explains their features, they’re Arab! Now, say that they have a conventionally Jewish name. Ah, they have a Jewfro, it all makes sense! My point here is how you perceive someone and what they look like is contingent upon what you know about their ancestry. Perception isn’t objective, it is colored by the facts and preconceptions we bring to the table. In Me, Myself and Irene the main character explains that his sons, who obviously have black ancestry (his wife had an affair with a black dwarf), are dark-skinned because of Italian ancestry. This is a good laugh line, but there is the grain of truth that we tend to explain or rationalize the data we have with theories derived from the facts we assume. So for example if you have two siblings of mixed ancestry who favor one of their parents there is often the assertion that they look like two different races. But, I suspect that some sort of contrast affect is at work here, that though the siblings do favor one suite of features they do not truly look monoracial.
That being said, what about the genetics? Why don’t offspring look like the midpoint of their parents’ phenotypes? The short answer is that genetics is not blending of phenotypes, it is the mixing of discrete information bits, genes. The smaller the number of genes that control a feature the more that sampling variance can work so that the appearance deviates from the midpoint of the parental character and “favors” one parent. The reason is that human beings overlap quite a bit in physical appearance. As a parent you contribute exactly 50% of your offspring’s genes, one copy of a gene which you carry two copies of. If the parents overlap in genes and phenotype the sampling of the genes may occur in such a way that the offspring resembles one of the parents more simply because its genes on those loci resemble the state of that parent. To give a concrete example, say that there is a gene, A, which comes in two flavors, 1 and 2. These flavors impart a physical characteristic in an additive and independent fashion, so that a parent who is heterozygous, carries both 1 and 2, is equidistant in phenotype from parents which are homozygous for 1 or 2. If an individual who is heterozygous pairs up with someone who is homozygous, then the nature of inheritance is such that on the trait coded by A there is a 50% chance that the offspring will resemble the heterozygous parent and a 50% chance that the offspring will resemble the homozygous parent. Using a one gene model the discrete character of genetic inheritance is clear, the offspring inherits from both parents but resembles only one parent. As the number of genes increases such a stark separation of outcomes is far less likely, but the general principle is still operative. Parents who overlap in genes and characteristics may produce offspring who resemble one parent far more than another because of biased sampling of the genes in equation. The image to the left illustrates the process using a Venn Diagram. In this case the parents overlap in genes and appearance, and the offspring inherits from the the genes of parent 1 which resemble the state of the genes of parent 2. Therefore naturally the offspring resembles parent 2.
Above I stated that humans overlap in appearance quite a bit. This might surprise you, and you might not believe me. But I think to some extent this perception is a quirk of human psychology, we tend to shoehorn individuals into racial types which we conceive of as very distinct, perhaps more distinct than they truly are. The fact is that human normal variation is subject to some constraints in terms of function and extant genetic variation. To the left are photographs of the Icelandic singer Bjork and a Bushman woman. Contrary to rumor Bjork is not half-Chinese or part Inuit. No doubt her ancestry is a mixture of Irish and Norse homogenized through centuries of local breeding and population bottlenecks. Nevertheless, her physical appearance suggests to people that she of Asian provenance, and she has noted that she was mocked for being Chinese as a small child, suggesting that she is an extreme end of the Icelandic range. The point though is that some northern Europeans have a vague Asian cast to their features, and this does not necessarily have to do with some distant ancestry from the eastern regions of Eurasia, rather, it is simply part of the normal range of variation. In contrast to Bjork the Bushman woman is of normal type for her group. Though African one might accept that she has an Asian cast to her features, and in fact when Bushmen in Botswana met Vietnamese workers they assumed that they were of the same race (as opposed to the Bantu speaking African agriculturalists and herders who were their neighbors). We know from analysis of neutral markers that there is no close genetic relationship between Bushmen and East Asians, rather, there are only so many ways to chop up human genetic variation to produce a modal type, and so some forms reoccur. These are extreme cases, but I think it is important to keep in mind that there is overlap in the finite suite of characters which contribute to our perception of someone’s racial identity, and that may be the source of the tendency of many children to favor once side (there is also the reality that many Eurasian children, for example, are assumed to look Asian by Europeans and look European by Asians).