I agree.
But still not enough isolation to create any subspecies. And even if, in a hyper-technical sense, you could say that isolated populations existed hundred or thousands of years ago, they no longer exist now. Which leaves you with no scientific basis to claim the existence of races or subspecies.
Which is still not evidence of “pure” human races, if that’s what you’re trying to say. Any way you slice or dice different populations is going to be arbitrary because there are no populations that are truly isolated or that don’t blend seamlessly into some other population.
What is it that you don’t understand about the original answer I gave? If there is evidence that certain populations meet the criteria for subspecies (let’s use that term instead of “race” since it is the preferred scientific term), then that would falsify the claim. If you’ve got it, bring it.
But human variation is clinal in nature, meaning that the groups vary continuously as one population blends into the other. You can’t go anywhere and say: Here is a group that doesn’t look anything at all like its neighbors, and has been genetically isolated from them.
So if I could point to two races of people, and there was a clear dividing line between them such that nearly all of the people in one race had some aspect to their genes that was lacking in the other race (excluding people of mixed ancestry), you would consider your claim to have been falsified?
Arggghh. Is it time for a sticky on races, or a FAQ or something?
Here, I’m going to post a standard blurb on science and race. John Mace is not entirely correct, it’s possible to discern genetic populations, but, Brazil84, no genetic human population has any gene exclusively (or even almost exclusively). That includes genes for dark skin tone, or any other particular one you can imagine (see the last paragraph).
Bottom line: It’s possible to look at distributions of genetic variations in humans and historical settlement patterns, and (with some subjective judgement about where to draw lines) delineate scientifically defensible genetic populations of humans, with mixing at boundaries allowed. However, the genetic differences between populations aren’t very large, compared to the amount of variation within each population. This means that all you can say is that an individual from population A is more likely to have genetic variants x,y, and z than an individual from population B, but any individual from either population could have x, y and z; only x; only z; or none of them. Additionally, with modern travel, populations are rapidly mixing, so they mean even less.
On the other hand, ‘race’ is not based at all on genetics. Different cultures can define it differently, but in the U.S. it’s pretty much based on skin color, presence of eyefolds, hair color and texture, style of clothing, accent, musical/food/other tastes, and to a lesser degree some other facial structures. [Go watch Dave Chapelles ‘Racial Draft’ skit if you don’t believe this]
The interesting thing is that skin color differences have very little overlap with true genetic populations, if you look at the whole range of human genes (not just the few genes that determine skin color). In fact, genetically speaking there’s more variation within Africa than in the rest of the world combined: if you had to sort the entire human population into say, five, genetic groups, you’d have four groups in Africa, and one group for the rest of the world (Europeans, Native Australians, dark-skinned (Asian) Indians, Japanese, and Native Americans are all more similar to each other than major African groups are).
I never said you couldn’t. In fact I specifically said you could define any number of “genetic populations”-- which is precisely one of the problems. There isn’t any dividing line that makes any one particular division better than another one, and you end up chasing smaller and smaller populations until it just doesn’t make sense any more. We could define 2 “genetic populations” or 2,000-- it all depends on how finely you wanted to sift the genes. And even then, you’d still have “borderline” populations that didn’t obviously fit into one population as opposed to another.
No, I wasn’t arguing some idea “pure” races, really more along the lines of what Quercus spelled out. But my linked cite was more directed specifically to the OP. While not taking that study as definitive, it suggests an American (to accept your qualifications on the studies usefulness) embryo could theoretically have these genetic microsatellite markers analyzed (ok, whatever the heck that exactly is) and get an indication of what race
it fell in to.
Obviously, knowing who the parents are would be a heck of a lot better, but I could see a few uses to pinning down some “genetic clusters” to accepted ethnic groups. Only criminology ,pathology or anthropological based reasons spring to mind at the moment, however.
CarnalK: OK. Note, too, that I said much the same thing in my post #52. I would add, however, that there are going to be large segments of the populations, even in the US, for which such a test would not work (or at least be problematic)-- for those fetuses of mixed parentage (more and more people fall into that category), or from regions of the world that might not be so well understood genetically. We’re used to thinking of race in terms of Black, White, (East) Asian, and maybe Native American in the US, but our population in 2007 is much more diverse than that.
So, the answer to the OP would be “yes”, in most cases you probably could do some test that would determine the race of the fetus in the US, although quite a few would give ambiguous or indeterminate results. If you wanted to do such a test outside the US, you’d first have to establish what the racial parameters are (according to whatever social conventions exist there) and then do the proper genetic testing of the larger population (which in many cases probably has been done to some degree or another). But there isn’t anything that would prevent such a test being done in most countries, at least in principle.
Another point which seems to have been missed, here: One of the cases where genetic tests for “race” don’t work well, as others mentioned, is when the subject is of mixed race. But asking whether it’s possible to determine the race of a fetus can only possibly be of interest when there’s the possibility that the fetus is of mixed race: If the mother is (for example) white, then the fetus is going to be either white, or half white, half something else. And if it’s half white, then the test might still show the fetus as being white, since the test isn’t very good for mixed-race subjects.
And when you use these test to determine the % admixture of people from mixed race backgrounds, you have to make some assumptions about what the source of that mixture is. If you test for admixture of White (European), Black, East Asian or American Indian, but the father turns out to be Pakistani or Saudi Arabian, then the accuracy of your data is going to be compromised.
So if I could point to two races of people, and there was a clear dividing line between them such that nearly all of the people in one race had some aspect to their genes that was lacking in the other race (excluding people of mixed ancestry), you would consider your claim to have been falsified?
Look at it this way…
blue eyes come from a gen, a bit of DNA. If (I really don’t know) the gen for blue eyes is known, then analyzing DNA and finding the blue eyes gen would mean the kid’s white, right?
No. Because there’s black folks out there with blue or grey eyes - you’re highly unlikely to find one in Africa, but I’ve met several in the US and one from the UK. Get a white great-great-greatparent with blue eyes on each side of the family and you can have someone who’s more than 75% “black” and still has that oh-so-white trait.
Heck, how many people would define Beyoncé Knowles as white or Shakira as black? I’ve never heard of any black bloodlines for Shakira. And yet, in their “Beautiful liar” video, half the time the images go by so fast and they’re purposely so a-likened, you need several viewings to tell which is which.
Does that mean you think the test is bogus?
It means I think it can not, ever, be totally accurate, among other reasons because as Colibri pointed out our definition of race is not based on genetics and the test is.
In the US, Halle Berry is African-American. In Spain she’s coffee-with-milk. Different cultures define races in different ways, and those ways have very little relationship with genetics.
But that doesn’t answer the question of whether the test is bogus or not.
It may surprise you to learn that many tests used on human beings are not totally accurate.
There’s a few issues at hand. Accuracy is one, precision is another, and interpretation is the third. Incorrect interpretation does not make the test bogus. The tests seem pretty accurate and precise at what they measure, but as it’s been pointed out many times they do not provide any useful metric in the scenario you presented. Since you have a mother and a fetus, and the race of the mother is known, there is no reason to believe the tests presented would allow you to determine the race of the father to any degree of certainty. They won’t tell you black, white, asian or whatever it is you are expecting without a terrible margin of error.
The tests aren’t bogus, your scenario is bogus.
Sorry for my absence from the thread. Work intervened. However, John Mace, Quercus and others have pretty much said what I would have said if I had been here.
Yes, that would be the case. It is precisely because such genetic differences cannot be identified when looking at populations on a global scale that we say that “race” has no scientific validity. There may be some genetic markers that are mainly confined to some local populations (as John Mace has described), but these are smaller in scale than the traditional broad racial categories of Caucasian, Negroid, Mongoloid, etc. And there are no such markers that are found in “nearly all” of the members of such groups, but are lacking from nearly all the members of one or more of the other groups. There are no clear dividing lines, either morphological or genetic, between “races” on a global scale.
I would add to this that human variation is not just clinal, it is discordant. That is, the patterns of variation of different morphological and genetic characters are not correlated with one another. For example, skin color, hair type, and blood group frequencies show quite different patterns of variation. This is another reason why races (or subspecies) cannot be defined in humans.
Depends on what exactly you mean by “bogus.” It may give reasonable results in certain limited areas or in special circumstances. On a wider scale, not so much. The latter perspective was what my post #2 was based on.
Maybe it would help a little, but I bet we’d have to go through one of these every three months or so anyway. 
Sorry if this is pointless repetition, but I’m still a little unclear here on where we’ve got to with a simple part of the discussion.
If I were to give a biologist plenty of samples (from foetuses), plenty of money for testing and as much time as he or she needed, could he/she come up with a reasonable correlation between some characteristics of the samples and the future US societal perceived race of the developed child/adolescent/adult or not?
Not guarantees, not foolproof predictions, but correlations? For example, if I were to give a biologist a billion dollars and say you can keep it if you predict the race (as perceived by society, rightly or wrongly) of the child, would s/he just shrug and
guess?
That doesn’t make sense. If there was a “clear dividing line”, then there wouldn’t be any people of “mixed ancestry”. But you can’t just look at any two populations* in isolation. You have to consider all the possible populations that they can mix with. And in that light, the boat has sailed long ago on the idea that there can be subspecies of humans.
There’s a logical contradiction in that a subspecies must exist in relative genetic isolation, but as soon as a population of humans is “discovered”, they are no longer in isolation. You might find some remote tribe in Papua New Guinea that exists in “relative isolation” from different populations, but other, similar tribes do not. We’re long past the stage where subspecies can exist. Given the way human populations have always interbred whenever they met, there is no reason to believe that any new population we find would be any different.
*you’re jumping the gun by calling them “races”, since it has to be established first that they are separate races (ie, subspecies).
See my post #54.
It depends on how well the adult population (from which those fetuses derive) was characterized genetically. In the US, that’s been done pretty well, so that in most cases you’d be able to get it right. If you were to plop that geneticist down somewhere where the adult population had not already been characterized, he probably couldn’t do much-- he wouldn’t know which genetic markers correspond to that population’s socially perceived races. He might come up with any number of schemes to sort the differences out, but he wouldn’t know which scheme to use.
Remember, these markers are in non-coding sections of the genes, so you’re not actually looking at genes that correspond to particular physical features.