I think the concept of race can be a useful, if flawed, tool for anthropological and social/political discussions. Certainly we can find occasions where the perceptions of race have figured into the decisions and actions of of groups of people. Whether it would be beneficial to replace the word “race” with another term, I am not sure. In the interests of communicating ideas regarding the treatment and perception of groups, the word has already (for better or worse) established itself in the language.
The problem occurs when we try to apply those perceptions to biological reality. The evidence (as copiously cited by Collounsbury) indicates that we do not have any way to genetically determine any group that we would call a race (unless we break “race” down into so many tiny populations that the word becomes meaningless).
peace kept claiming that we could get to race based on a percentage of common genetic traits. When he initially claimed that we could identify race by an examination of HLA-typing in blood, several of us looked at his citations and discovered that within any given “race,” a substantial percentage of the population would actually share more characterstics with people from other “races” and that no set of characteristics could be shown to appear always and exclusively within any one group. I asked how we were supposed to identify anyone’s race by traits that occur in multiple groups and he replied that we were just supposed to look at the averages. I dunno. If I look at a characteristic “belonging” to Group A and I find that characteristic in a member of Group B, does that mean that the person on Group B is “really” an A? He never did explain how that worked except to claim that it was obvious.
There are diseases and genetic characteristics that are more prevalent among certain populations. This is simply another side to the superficial characteristics with which we are already familiar. People who intrabreed over a long period will share similar characteristics. It is possible that if a population was separated from humanity for a sufficient length of time, a “race” could emerge. The historical (genetic) record, however, indicates that interbreeding has remained sufficiently high throughout the human population that that event has not occurred.
Colors:
Up until roughly the Renaissance, there was no orange in Europe. No European language had a word to identify that color. Certainly, there were oak trees that changed color in the autumn and sunsets and other phenomena that we would describe as “orange.” However, prior to the introduction of the Indian fruit by Persian merchants at the end of the middle ages, Europeans never found a need to identify that color. There was red with a lot of yellow in it and yellow with a lot of red in it (even gold was described as “yellow” or, occasionally, “red”).
Had we drawn a color bar extending from the deepest red to the brightest yellow on a piece of paper, handed a Medieval scholar a pencil and asked him to draw a line separating each color, he would most likely have drawn a line halfway between the two ends, right in the middle of our “orange.” If we gave the same task to a second grader in an American school, today, the child would probably draw two lines, separating red from orange and orange from yellow.
If our culture continues to expand its uses of color (in advertising, pedagogy, or screen design, etc.) it is possible that future generations will recognize as distinct, colors that we now place in the red-orange or yellow-orange bands of the spectrum.
Race appears in the same fashion between anthropology and biology. We can “see” the “distinct” races at the middle of any band, but any line we draw to identify the separations between those races fails on the biological level.
The obvious reason for this is that genetics, unlike light, is not a two-dimensional spectrum. We can plot light according to wavelengths, but genetics will always have so many overlapping and intermixed markers that no line is truly possible.
Light and color make a decent analogy, but all analogies are imperfect.