Review of SMM Exhibit on Race and Racism

"There is no such thing as race (biologically) "
That is a hypothesis that is possibly falsifiable through experimental verification.
How about the following experiment.
Take a blood samples from three hundred individuals, one hundred from China, one hundred from Nigeria and one hundred from Sweden (all the individuals being typical examples of the socially constructed races of 'oriental', 'black' and 'white', respectively).
Now carry out a biological test on these blood samples, SNP analysis. What are the chances that in a blind analysis you would be able to use the results to accurately place 100% of these individuals into the correct 'race'?
If the answer approached 100% probability would that not falsify your thesis?

Sigmund -- why bother with blood samples? Just take photos instead.

Sigmund -- why bother with blood samples? Just take photos instead.

I think Greg is mistakenly mixing up two points here.
First, is there a biological difference between 'races'?
I would suggest there is a empirically detectable difference - for instance as revealed by SNP analysis.
Second, is there a difference which would allow you to predict abilities of an individual on the basis of his or her race?
This assumption, I think most of us agree, underlies the racism seen in societies at present, yet has zero basis in any biological results yet revealed. There is so much variation within races for almost every ability that it is impossible to use 'race' itself as a defining factor for ability.

Sigmund, why would you not sample more broadly across each of your races? Are Swedes somehow more white than Greeks and Spaniards? Are Nigerians more black than Eritrians and Namibians? Are the Chinese more asian than the Japanese and the Nepalese?

Stephanie Z, thats not a bad point.
Why not, indeed, sample broadly from the populations to make up the 100 of each race.
Would we still be able to predict the answers to a similar degree of accuracy?

"If the answer approached 100% probability would that not falsify your thesis?"

Sigmund, if human genetic variation is clinical (which I believe is unquestioned) then how would this test falsify the hypothesis "There is no such thing as race (biologically)"?

Feel free to correct me if I'm wrong, but... hasn't this work you all are talking about already been done? By Lewontin and later by Barbujani et al? I was under the impression there is mroe variation within populations than between populations, and that this is an important component of the argument against biological race.

Hmmm... comments like this

even efforts to use race in a "positive" way such as in medicine or forensics are doomed to failure because of the lack of biological validity of the concept, and so on and so forth.

seem incorrect to me. I believe there are two discussions to be had on the idea that a patient's race influences treatment.

Even assuming that it is correct that there are no meaningful biological differences between races, and therefore that it would be fruitless to pursue the idea that therapeutics in general would be different depending upon race, there still exists a very significant difference in incidence of disease along racial lines.

Perhaps that isn't what you meant at all in saying that, so I'll merely point out that to say that race oughtn't be a consideration in medicine is to completely ignore that the beta-globin diseases constitute the most common class of genetic diseases in the world, and that they primarily affect equatorial, largely black regions.

Ignoring that fact could have grave consequences, not only (especially) for the patient, but also for the doctor's malpractice insurance carrier.

Hmmm... comments like this

even efforts to use race in a "positive" way such as in medicine or forensics are doomed to failure because of the lack of biological validity of the concept, and so on and so forth.

seem incorrect to me. I believe there are two discussions to be had on the idea that a patient's race influences treatment.

Even assuming that it is correct that there are no meaningful biological differences between races, and therefore that it would be fruitless to pursue the idea that therapeutics in general would be different depending upon race, there still exists a very significant difference in incidence of disease along racial lines.

Perhaps that isn't what you meant at all in saying that, so I'll merely point out that to say that race oughtn't be a consideration in medicine is to completely ignore that the beta-globin diseases constitute the most common class of genetic diseases in the world, and that they primarily affect equatorial, largely black regions.

Ignoring that fact could have grave consequences, not only (especially) for the patient, but also for the doctor's malpractice insurance carrier.

Oh, good. Seems my wireless is working now.

Oh and BTW, Stephanie Z, it's spelled Eritrean.

Thanks, Temaharay. I should really not try to type early in the morning. Everything looks wrong, so it's hard to tell what actually is.

So, Sigmund, since we should be sampling broadly, how about we simplify the process and do it in a city that already has a diverse population, like New York?

Do we even need the whole city Stephanie?
Why not a single building - such as the United nations Building? I think we are pretty much guaranteed to get the best representation on the entire planet in that particular location. Who could object to that?

Sigmund, I would give an awful lot to see you ask the diplomats for DNA samples. :) But let's assume they would give them to you. How would you then assign these people to your racial groups?

My racial groups?
It wasn't me who came up with the idea!
I still think we are mixing up two separate matters in this topic. Obviously there is a social aspect to the term 'race' in terms of assigning individuals to one group or another on the basis of physical differences. These common differences are due to ancestry in widely separated geographic locations. The fact that such historical separation means there are lots of genetic markers that can distinguish to a high degree of accuracy these sorts of questions of ancestry should not be ignored, nor, however, should it be used as an excuse for discrimination.

help me! Help me!!! I'm at Gustavus Adolphus trying to blog the Nobel conference semi-live the the network is not letting me out! I can only communicate through comments!!!

Sigmund, you were the one to suggest the experiment. I wanted one that was better designed.

I'm not talking about discrimination here. I'm asking what the concept of race buys us, what it tells us that something else doesn't tell us better. I'm not arguing that isolated populations won't have some genetic similarities, but that isn't how we talk about race. How far do we have to travel down the line from all humanity to individual genetic makeup before we get to groupings that have any kind of meaningful predictive power for the individual? And if we can't say anything meaningful before we get far beyond what anyone would call race, why shouldn't we ignore it?

"I'm asking what the concept of race buys us, what it tells us that something else doesn't tell us better. "
Clearly individual testing (such as genome sequencing or SNP analysis) will reveal a lot more than something as crude as ones race. However does that mean that we cannot get any statistically meaningful data even from this crude measure?
For instance, should people of Irish origin be given the same health advice as native Aboriginal Australians regarding how they deal with the effect of tropical sunshine?
There are probably many such examples of how one's ancestry may affect ones health differently than other peoples of different ancestry.

Sigmund, I would argue that in these days of ozone depletion, yes, they should get very similar advice. Okay, I realize that was perhaps not the best example you could have come up with, but seriously, thinking like that has been known to lead to dark-skinned people not being screened for skin cancer, which is a bad thing. They're not immune, just somewhat less likely to get it.

You can tell someone's ancestry from their genes, right?

People are more closely related to some people than others, eg. your average ethnic Khoisan is more closely related to another Khoisan than to an ethnic Inuit.

These different ethnic groups have different distributions of various alleles that mark them out. There are differences because, eg. their ancestors adapted to different parts of the world. That's true for their different skin pigmentations, for example.

So how can you say, "There is no such thing as race (biologically)"? There's certainly differential genetic distributions between populations according to commonness of ancestry.

It's true that many people are more or less mixed-race, and that there's no bright dividing lines separating ancestral groups. Obviously, the Khoisan don't have entirely different genomes as a group from the Inuit. But there are systematic genetic differences as a result of their divergent evolution over thousands of years.

Why don't we define race like that and accept that it exists?

I am engaged in other activitreis at this moment, but I have time to give a small and possibly effective response to some of the comments above.

I am thinking of six traits that vary among humans around the world. Take traits 3, 4, 5 and 6 from a group of people in china, a group of people (natives) in Brazil, a group of people in Austraila, and a group of people in Francde. No cheating . Everybody should be native to the land you sampe them in at least twenty or so generations back.

Traits 3,4,5 and 6 will sort out so that you can unambiguously tell these groups apart pretty much as the race model many of you are working so heartily to defend would predict. These may be traits about skin color, hair color, nose shape, etc. etc.

Find. You can have your damn races and racial categories, because the experiment worked.

But I'm asking you to go one level beyond this in thinking. I want you to also look at the traits 1 and 2 in my collection of data.

Traits 1 and two also sort out like the other traits do, but they are not physical features of the people at all. They are something else. And once you know what they are, you will suddenly, in a flash, understand the bankruptcy of your model.

Anybody want to guess what traints 1 and 2 are?

I'll give you a couple of hours.

Sigmund,

Please read the articles here:

http://raceandgenomics.ssrc.org/

The exhibit does not have a single photograph of a black man swinging, dead, from a lamp post or a tree.

http://www.withoutsanctuary.org/main.html

Mother tongue? National or ethnic identity? So what?

An anecdote: Many years ago, when I was in college in Indiana, an acquaintance of mine from Venezuela came up to me and nervously asked if she could speak to me privately. I didn't know what to expect, and when we finally found a private space, I was astonished to find that she wanted to talk to about race. She'd been in the US about three months, and was completely confused about race in America, but she knew enough to know it was a taboo subject here.

You see, it all started when someone referred to "that black guy over there," and she looked, and she couldn't see any black people. None. And yet her interlocuter insisted there was a black man standing right there. Then someone else referred to a third person as Hispanic, and, despite being Hispanic herself, she could not figure out how anyone could distinguish people racially on the basis of language.

We wound up discussing the histories of the United States and Venezuela in terms of colonization, immigration, slavery, etc., and how the two countries wound up with completely different sets of racial categories. And yet when we looked at groups of diverse people, we each saw exactly what we were conditioned to see. (For example, in Venezuela, "black"/"negro" refers to what we might call "maroons," escaped slaves who fled to the jungle and still lead essentially African lives. Whereas what Americans call "black," would in Venezuela be considered "mestizo," or essentially normal, unmarked Venezuelans in all their shades of brown and biege.) I know intellectually that race is culturally determined and socially constructed, and yet when I look at a large, diverse group of people, I can't not see those people as representatives of the racial types that characterize North Americans.

Another example: A century ago, any American could look a picture of various northern European people, and pick out the Irish. And they'd probably do better than chance. How many people could do that today, now that "Irish" is no longer considered a separate race?

If I were putting together a museum exhibit on race, I'd like to see a large photo showing a diverse group of humans, and then put up a series of overlays showing how they'd be classed racially by, say, a North American, a South American, a West African, an East Asian, a European, etc. I suspect there'd be a surprising amount of divergence.

A skeletal biologist explains why biological assessments of "race" are problematic.

In short, "race" is an unhappy mix of ancestry (biological), ethnicity (cultural), and self-identification (personal). And even ancestry is not straightforward; most techniques for identifying race or ancestry from skeletons (e.g., Fordisc) were developed using huge skeletal collections amassed decades ago in which remains are catalogued by how people self-identified--not by an external, objective assessment of ancestry, and without allowances for ancestries that include multiple populations. For example, a person of Caucasian, African, and Caribbean ancestry might self-identify as "black" in a census. Taking that person's skeleton as representative of African ancestry would be a mistake. But this is the kind of data on which statistical determinations of "race" are built.

It seems that most of the comments broadly supporting the blog-post do so by arguing that common ideas about race are unsubstantiated.

For argument's sake, fine, let's accept that. But that still leaves the reality that there are ancestral population groups diffentiated as to the distribution of genes among them.

Obviously many common ideas about race are scientifically invalid. But that's got nothing to do with the reality of human population differences.

Ethylene. Nope. That is not the proper null model for genetic variation in any population. The null model is continuity. Departures from continuity can be/need to be explained, and race is one possible departure that cuold happen but usually does not happen. Except in people's minds, where it happens all the time.

Is any one prepared to answer my question about variables 1 and 2? What are variables 1 and 2? Anybody?

Ethylene asked, "Why don't we define race like that and accept that it exists?"

Because it conflicts with our Marxist world view.

> What are variables 1 and 2? Anybody?

Well, how about IQ? But that undercuts your argument.

By naming the variables 1 and 2 in this particular study of geography, I've given you a blindingly obvious hint.

I'll check back later on that.

Bob, here's a chemistry question for you: I notice that the stuff around me ... like water, maple syrup, gasoline as examples of liquids, or clay or oat meal or gravel as gannular solids etc. pretty much comes as a continuum, or really a set of different continua. For example, there is water on one end and kayro syrup on the other. So I therefore assume that the basic building blocks of matter are continuous, not atomistic. This periodic table of the elements is not at all obvious. Therefore the non-atomistic, continuous nature of matter which I see all around me and that even fits with our linguistic conventions is correct, right?

This periodic table that has been worked out by chemists. Bullshit, right? I mean, I can SEE that it is wrong.

I'm gonna guess that at least one of traits 1 & 2 is skin or hair colour. Or maybe even traits 1 & 2 are both skin and hair colour.

As a clue: your Australian sample would have to be aboriginal, since none of the rest of us can lay claim to the required 20 generations. And I do know that their skin & hair variation is genetically totally different from African, despite the fact that they are "black". You get natural blondes in some groups, for instance.

Greg, one problem, although it's probably irrelevant to your thought experiment. Can you get people with that little regional variation in their family tree in Australia and Brazil? I'm pretty sure you can't in France. Too much territorial churn.

OK, look:

Trait 3: skin color
Trait 4: hair color
Trait 5: Hair kinkiness
Trait 6: Nose length

Samples of whatever we consider indigenous from the Dordogne, France, Arnhemland Australia, the central Amazon Basin, Red Lake Indian Reservation, Minnesota and the countryside outside of Beijing.

What do you get? Four or five races, very little or zero overlap, everybody in each group looks way more like each other than they look like members of the other group, etc. etc... a nice bit of evidence for the existence of race.

Now, let's look at traits 1 and 2. I guarantee that when you look at either 1 or 2 or the two in combination, you will ALSO get five different groups with no overlap, more similarity within than between groups, etc. etc. Just like the racial traits.

The two traits are:

Latitude and longitude.

This works.

Now, take a different sample. Anchor your sample, say, in Oso. Head south to Cape Town. Take a sample every 200 km. I absolutely promise you that for every one of these traits you will find a grade, a cline, a gradual transition. You will find no evidence of race. None.

People often confuse race with variation, or once they 'get' what I just said, complain that all they need to do is to replace variation with 'race' and they can still have their very limited and quite incorrect models of racial variation. But they would be wrong.

Anybody want to guess why they would be wrong? (There are at least two major reasons).

Because ..... it gives us too much latitude? Is that it?

Confusing variation with categorization is like confusing how much gas you have in your tank with whether or not you have (some unspecified amount of) gas in your tank. It is simply incorrect and this incorrectness can lead to all sorts of problems. To make a longitudinal story short.

Well, how about IQ? But that undercuts your argument.

Do we have to play this idiotic game again?

We're talking about two different things here, populations and races. Dogs are a race of wolves. The packs of Valley A are a different population than the packs of Valley B.

Where H. sapiens is concerned we have had at least two races, perhaps more. Currently our species consists of one race, though we may have others. The Australian Aborigine has been proposed as a distinct race, having skeletal differences. Groups such as the Khoi and the San (who have differences between them as much as they have differences with other groups), pygmies, and the Somali/Ethiopics/Nilotics who are beautifully adapted to their harsh environment. In 20 or thirty generations the Eskimo may evolve into a new race. (All Inuit are Eskimos, but not all Eskimos are Inuit.)

So biologically we could have races. But not culturally/sociologically.

I do not find the idea that gradual transitions negate the concept of race very attractive
since this would also mean that there are no races of dogs since there are also crossing between different races and you can find gradual transition between dog races.
To me, if you do not find gradual transitions it is a species rather than a race.

Chondrus. You are not getting it. If you do not find gradual transitions between AND there are different groups, then perhaps you have races.

Within dogs, breeds may be seen as races, and they exist because of reproductive isolation between groups. When the reproductive isolation is not maintained the groups break down and in a very short number of generations you get mutts or wolves. Which are not breeds/races.

Alan, don't worry, people will eventually be by to say that "population = race" just as people have already bee assuming that "my sampling of humanity = race" without quite realizing it.

As you say, (I think) the groups you mention are not good examples of races because they are for each citable trait parts of a continuum.

I wish Every American with a race concept could spend a month or two traveling around Africa with their favorite "field guide to the races of humanities" to show how poorly this works.

Given that racial boundaries are illusory ... do not exist .... and racial character sets are made up of not racial traits but rather continua, I would just as soon regard Australian aborigial groups as more of the same but more extreme in some ways. If we want to test the Australian separateness and uniqueness, that falls apart the moment you go to New Caledonia. So then you have to add in the New Caledonians and ramp up the variation in this group a little. But then you've got to add in the Newguineans. Then the melaniseans, then the micronesians, etc. etc. etc. and eventually you'll see that boundary disappear as well.

Not to mention the fact that there is huge variation within australia to begin with, so some of the australian groups would have been excluded from the initial "australian race" until later, after its variation is ramped up enough by adding otherwise very similar groups.

You can slice and dice humanity a lot of ways but into races just does not work if you use actual data and actual biological concepts, don't cheat, and so on. Yet people cannot get their minds away from this concept. It astonishing.

I am not American and I have actually been travelling in Africa. What I felt there was that you have races in Africa, people from parts of Africa differ radically.
In Europe, North Africa, and to a certain extent the Americas and Asia we only have different hairstyles.

I think I understand. Greg is arguing, in effect, that the existence of mongrels falsifies purebred dog theory, which seems neither accurate nor useful.

I'm still not sure whether Greg is arguing that there never was any biological basis for race even in population groups that were effectively isolated for hundreds of generations or that races did exist but have now blended.

Robert, can you point to any groups that were effectively isolated for hundreds of generations?

Robert, when talking about human genetic variation there is no such thing as a "mongrel" nor a "purebred". Who are purebred races and who are the mongrels?

"I'm still not sure whether Greg is arguing that there never was any biological basis for race even in population groups that were effectively isolated for hundreds of generations or that races did exist but have now blended."
He lost me when he suggested latitude and longitude are biological traits.

Sigmund, what he said is that if you sample in widely varied, isolated places, excluding people who migrated to and from these places, you will find distinct differences. However, using this data to suggest that humanity is made up of distinct groups is invalid because it ignores the graded differences that occur between your sampling sites.

It's like saying that humanity is divided into blondes, redheads and black-haired people because you (a hypothetical you, not you) only sampled blondes, redheads and black-haired people. You're either saying there is no gradation of color, which is patently false, or you're saying the gradation doesn't count, which is excluding huge amounts of data in an attempt to bolster your theory.

Sigmund: I said lat and long were NOT biological traits.

And yes, race is a theoretical, but uncommon, possibility.

OK Greg, perhaps I read it a little quickly but you must admit its very unusual to describe things like a geographic location as a 'trait'.
Anyway I think the whole problem with discussing 'race' as a concept is that it gets people talking at cross purposes.
People here are obviously arguing about different definitions of the word race.
Clearly all of humanity is very closely related. Clearly the data shows that most variation occurs within populations rather than between populations. I think we can all agree on those points.
The disagreements arise when we raise the question of how to regard the 10-15% of variation that seems to be population specific. At least some of this variation is not unimportant: it has important clinical consequences (for instance in tissue compatibility matching during transplant surgery).
This 'race' specific variation was also the basis for the choice of individuals sampled for the human hapmap project. To suggest it is a mistake to imply the differences are as superficial as hair or skin color is not the same as saying 'races' don't significantly overlap genetically or that one is better than the other in some way.
How population geneticists view the term 'race' or 'deme' is very different to how a racist does (or even how many 'non-racists' in society do - including, unfortunately, many sociologists). Is there some consensus we can arrive at? The traditional societal definition of race is clearly wrong (basically skin deep and hierarchical - someone who had one 'black' grandparent and who inherited their genes for skin color would be described as black even though they may have 75% 'non black' genes). Yet at the same time we have the point that ancestry has important clinical implications (not 100% obviously but statistically relevant in situations such as transplant HLA matching). Should we simply do away with the term race and stick to deme (which is more accurate and less likely to be understood by racists or sociologists on a mission)?

People here are obviously arguing about different definitions of the word race.

I'd have to agree. I suspect this is yet another situation where the failure to operationally define a nebulous construct leads to circular discourse. Or, as succinctly expressed in Cool Hand Luke, "What we have here; is a failure; to communicate."

If we use morphological (e.g., observable physical features), geographical (e.g., where you live), psychological (e.g., personality), and sociological (e.g., socioeconomic status) definitions of race as independent variables, and alleles as the dependent variable, no doubt there would be differing rates of success with using alleles as an appropriate discriminative validity index.

Ok look: Trait 1: latitude, Trait 2: longitude, Trait 3: skin color,Trait 4: hair color, Trait 5: Hair kinkiness,
Trait 6: Nose length

If I think about this thought experiment in terms of sampling (let's say a 100 pair) of identical twins from each defined racial group in the context of the above variables, I can see how race should be viewed as a nebulous construct.

People here are obviously arguing about different definitions of the word race.

I'd have to agree. I suspect this is yet another situation where the failure to operationally define a nebulous construct leads to circular discourse. Or, as succinctly expressed in Cool Hand Luke, "What we have here; is a failure; to communicate."

If we use morphological (e.g., observable physical features), geographical (e.g., where you live), psychological (e.g., personality), and sociological (e.g., socioeconomic status) definitions of race as independent variables, and alleles as the dependent variable, no doubt there would be differing rates of success with using alleles as an appropriate discriminative validity index.

Ok look: Trait 1: latitude, Trait 2: longitude, Trait 3: skin color,Trait 4: hair color, Trait 5: Hair kinkiness,
Trait 6: Nose length

If I think about this thought experiment in terms of sampling (let's say a 100 pair) of identical twins from each defined racial group in the context of the above variables, I can see how race should be viewed as a nebulous construct.

"There is no such thing as race (biologically)"

is like saying

"There is no such thing as gender (biologically)"

it's true in ways that are of interest to specialist (who want to break apart and refine the definitions) or trivially true, but in all the ways that matter it's false.

Thus, it's certainly not true that "efforts to use race in a "positive" way such as in medicine or forensics are doomed to failure because of the lack of biological validity of the concept".

Briefly, in these fields and others, race is 'free' information that allows one to refine one's priors. You might argue that race is so pernicious that its use should be abandoned despite the advantages it offers, but it's just silly to claim that it's informationally void.

Suggestions for replacing race in medical practice are often impractical or nearly useless, such as the suggestion that genotyping should replace self identification. It's been demonstrated many times that genotype will pick out the same categories as self-identification, at least in the US. Which is not to say that we couldn't eventually supplement racial classification with genotype data.

Greg's suggestion that latitude and longitude -- continuous genetic variation -- defeats that clusters observed from other sampling schemes is likewise off. There are discontinuities even at a global level (doi:10.1371/journal.pgen.0010070). When you consider the immigrant populations of the new world, the lat/long suggestion obviously breaks down (doi:10.1086/427888).

More importantly, however, even if human genetic diversity were truly clinal (gennome wide), it wouldn't eliminate the fact that different populations have adapted to different local environments, producing phenotypic differences of biological and social importance. Whatever you call those differences (we tend to use the word race), they aren't eliminated by arguing that racial categories should be eliminated.

There is no such thing as race (biologically)"

is like saying

"There is no such thing as gender (biologically)"

These are absolutely utterly different statements. There is not much more significant of a difference in organismic biology than between a male and female. No, saying one is not like saying the other at all.

Briefly, in these fields and others, race is 'free' information that allows one to refine one's priors.

If you take the premise that race exists as a biologically valid concept, but that is a pretty serious error of fact.

Sig: Again, race and variation are not the same thing. Matter varies in a way that would allow me to argue that at the most basic level it is continuous and not atomistic. But I would be wrong. Similarly (but of reversed polarity) humans place each other into squared off categories that don't exist.

Greg you are arguing against a strawman on this one.
The commonly accepted idea of 'race' as a completely seperate groups of individuals is wrong.
We both agree on this point.
All I'm saying is that demes, while not completely distinct at each and every loci (human populations are overlapping venn diagrams rather than separated groups) can be distinguished by genetic means such as SNP analysis.
In population terms genetic differences will essentially show up in a stochastic manner - its only the combination of many loci that provide the biological data that allows us to assign individuals to any particular deme.
It is also the case that in many modern societies such as the US, the population mixture has resulted in previously separate demes becoming much more mixed due to intermarriage etc. In the US the 'races' are still popularly seen as distinct in a way that population geneticists would not agree with.
Its really the sociological definition of race that you are arguing against rather than the demes of modern population genetics.

Sigmund, I'm confused. You started this comment thread by trying to design an experiment that would falsify Greg's claim based on the U.S. American sociological definition of race. Now you claim that this isn't really the definition of race that we should be talking about--we should be using a definition that gives "race" a better chance of being vaguely biological. Somewhere in between, I've completely lost track.

What is your point?

Make the demes small enough and I'll go for the argument but almost all variations that people claim to link to race still won't link to the demes, and the demes will be so small that you might as well ignore them.

This is what a century of research has shown us, Sig.

"or every one of these traits you will find a grade, a cline, a gradual transition"
differences without distinction...or s/th like that.

Greg, please remember that races, sub-species, can and do interbreed. florida panthers will interbreed with common cougars. Australian Aboriginal traits show up strongest in Australia. As you get further away they show up less and less. Were the Aborigines a separate species in the classical sense, this would not happen.

Now consider the Khoikhoi and the Bushmen*. The Khoikhoi have bred with the Bushmen to some degree, yet there remains a distinct Bushman population. And both have traits distinct from their Bantu/White neighbors. None of the four can be called pure sub-species, but purity and races is not a workable combination these days. When brought into close proximity two races of a single species will mix, and that is what we see in South Africa In time you will see a descendant population that combines the Khoikhoi, Bushmen, and Bantu/White.

*According to the Wikipedia entry on the San---as academics like to kill them---the word 'san' is Khoikhoi for 'not people'. When dealing with Bushmen of different tribes they prefer the term 'Bushmen'.

The sun appears to circle the earth so we should have two races that separate our observation, lets call them sunrise and sunset. The problem is our observation has nothing to do with reality because we are categorizing the races based on inaccurate and incomplete data.

Separation by race leads to the same problem as separation by species, if you collect all the information possible to collect, you would see that all life is only slightly different.

You can see the first light in the morning and say that is the race sunrise but, can you tell me when that race of sun turns into the sunset race?

For humans race is an invalid line of separation, all it does is point out a little variation that in reality doesn't matter. The medical argument is invalid because there are illnesses that are specific to loosely related people no matter their skin color.

Greg: For traits 1 and 2 (I showed up late, so already saw your answer), I was going to go for 1) Distance between the pair being considered 2) ground level UV. Both still geographical rather than biological.

General: Race in humans is a social construct, not a biological one. You realize this very quickly if you pay any attention to what 'The Races' have been averred to be over the last 200 years and how their changes have had nothing to do with biology. 'semitic' was a race, one of about a dozen 'caucasian' races. Mediterranean was a race, as was northern European, Slavic, and several others. The genocide in Rwanda was racial between groups outsiders would have been hard put to distinguish.

You also realize it as soon as you pay more than trivial attention to the biology. Take your constellation of traits 3-90. You'll rapidly find that nearest neighbors in trait space for some of them are well-separated geographically, pairing up Finns with Kenyans, for instance. (Distance running-related traits in this case.)

A little more concrete example is my former boss. He's from southern India. According to how the US classifies race -- as a political decision -- he's Caucasian. If you go by street-level reaction, he's African.

Greg: For traits 1 and 2 (I showed up late, so already saw your answer), I was going to go for 1) Distance between the pair being considered 2) ground level UV. Both still geographical rather than biological.

General: Race in humans is a social construct, not a biological one. You realize this very quickly if you pay any attention to what 'The Races' have been averred to be over the last 200 years and how their changes have had nothing to do with biology. 'semitic' was a race, one of about a dozen 'caucasian' races. Mediterranean was a race, as was northern European, Slavic, and several others. The genocide in Rwanda was racial between groups outsiders would have been hard put to distinguish.

You also realize it as soon as you pay more than trivial attention to the biology. Take your constellation of traits 3-90. You'll rapidly find that nearest neighbors in trait space for some of them are well-separated geographically, pairing up Finns with Kenyans, for instance. (Distance running-related traits in this case.)

A little more concrete example is my former boss. He's from southern India. According to how the US classifies race -- as a political decision -- he's Caucasian. If you go by street-level reaction, he's African.

"The Khoikhoi have bred with the Bushmen to some degree, yet there remains a distinct Bushman population."

That simply is not true. There is not a "bushman population" and the term "KhoiKhoi" means "bushman who keeps cattle" ... so a person can move into and out of that category with the purchase or sale of a cow.

"And both have traits distinct from their Bantu/White neighbors."

White, yes ... the whites came from the other side of the planet. But traits that characterize a Central African or East African Bantu person (physical traits) compared to traits that characterize a classic San person are found in, for instance, Xosha, whoa are in the "bantu" group you are talking about. And so on.

"but purity and races is not a workable combination these days"

You seem to be suggesting that these days this does not work but in the past it did. This is incorrect. We have piles of research on the physical anthropology of humans over thousands of years (see my post, above) and this has never worked.

It occasionally works. Every now and then you get a group of people with a lot of physical identity within and separation without and virtually no gene flow. But that does not characterize the whole species, and it is temporary when it happens.

Regarding the words San and Bushman. The Wikipedia is not correct. If you are in South Africa, you do not use the word Bushman because it means subhuman, effectively, so you use San. If you are in Botswana, you do not use the word San because the most politically active group of these folks say "Call us Bushman." So it is complicated!

The last time I had a conversation with a bushman/san, it was a guy originally from where "bushman" is the proper term who moved to where "san" is the proper term, and he kept switching back and forth. There are more important issues!

There is no such thing as race (biologically)

From American Heritage Dictionary:
5. Biology
a) An interbreeding, usually geographically isolated population of organisms differing from other populations of the same species in the frequency of hereditary traits. A race that has been given formal taxonomic recognition is known as a subspecies.

Are you saying that geographically isolated interbreeding human populations that differ in allele frequency don't exist and never existed?
LOL.

There is no such thing as race (biologically)

From American Heritage Dictionary:
RACE
5. Biology
a) An interbreeding, usually geographically isolated population of organisms differing from other populations of the same species in the frequency of hereditary traits. A race that has been given formal taxonomic recognition is known as a subspecies.

Are you saying that geographically isolated interbreeding human populations that differ in frequencies of alleles don't exist and never existed? LOL.

DK: Most people ... no, all people it seems ... who push the race concept in human are unprepared to accept the possibility that, say, five percent of the species looks genetically like a "race" while the rest don't ... while the rest can not be put into biologically valid racial categories without making shit up.

The race concept is always put forth as a set of categories into which our species can be divided such that the dictionary definition you provide would apply. You are asking, however, a question that forces us to diverge from that widely held concept, which makes answering it in a straight forward way simply impossible (were I to remain honest).

I can think of almost no example of human populations that fit the definition you describe in any meaningful way, but the possibility is not utterly outrageous. Just non-existent for all practical purposes.

To Greg:
I can think of almost no example of human populations that fit the definition you describe in any meaningful way

Let me try to help you then:
1. "interbreeding" - CHECK. Different human populations interbreed just fine.
2. "usually geographically isolated" - CHECK. Africans are pretty well isolated from South Americans, etc, etc.
3. "differing ... in the frequency of hereditary traits" - CHECK. Lookup any SNP paper as of late comparing different populations or even self-defined races and without exception all of them will show unequivocally the differences.

So if you can't think of any populations that fit the above three criteria, here is the list (one of the many possible - remember that everything in biology is on a continuum and any classification is necessarily arbitrary; that applies to the concept of biological species to the same extent as subspecies/races): Africans, [native] Americans, Europeans + West Asians, East Asians, [native] Australians.

Africans are pretty well isolated from South Americans, etc, etc.

1) Are you sure? Better check on that more carefully. Yes, all the people at one end of the room can't hear the people at the other end of the room talking, but eveyone can hear the person standing next to them. That's a metaphor...

2) Africa and South America do not correspond to any known race that has ever been suggested ... so why do we care about these two geographical areas?

Regarding your point 3, see my thought experiment above. The sequentially assigned ID numbers in the same studies would also sort out as races, wouldn't they? This is not how one would test the hypothesis of the existence of races.

DK, you very much want to believe not only in races, but the five or so standard races. Even to the point of making an absurd argument. Why? THAT is the interesting question here!!!!!

To Greg:
Africans are pretty well isolated from South Americans, etc, etc.

1) Are you sure? Better check on that more carefully.

Yes, absolutely. Until the slave trade, the two populations were pretty well geographically isolated. It's silly to deny the obvious. And the fact that today's Africans are easily separated from today's Native Americans based on a plethora of genetic traits tells that, genetically speaking, they are still relatiively isolated even today. Geography. Simple.

2) Africa and South America do not correspond to any known race that has ever been suggested ... so why do we care about these two geographical areas?

DK, you very much want to believe not only in races, but the five or so standard races.

I am sorry but that's just utter bull! Your comment is in response to my words that the list is " one of the many possible ... everything in biology is on a continuum and any classification is necessarily arbitrary" - from which it is clear that the I believe that the NUMBER of races we choose to identify has ABSOLUTELY NOTHING TO DO with the fact of the existence of biological races in any particular species. (Hint: How many colors are there? Do colors exist?) And the particular example that I gave simply reflects observed genetic clusters that roughly fit continental origins. Since you could not think of any, I figured one example would suffice :-)

So, if you want to get suggestive - fine. But I am not interested in this kind of discussion. Else you can go back to the definition and SHOW that no human populations on this planet fit this definition. I contend it is an impossible task for you.

DK: until the slave trade? Why does it always boil down to someone blaming whitey for bringing blackie over here--as if Africans don't and didn't have their own civilizations, their own impetus to travel, to conquer, as well as to enslave?

They were here long before slavery, if you believe the fact that the coca leaf ( native only to S. America) was found in Egyptian pyramids--pyramids built by slaves from all over Africa and the near and middle east.

One of those famous conquistadors (Velasquez?) reported the natives saying they were at war with a tribe of "black men" in the mountains of Central America, and even Colombus reported "black men" in boats,leaving Hispaniola upon his arrival.

Oh, and Olmec heads with cornrows...

http://ipoaa.com/africans_in_americas.htm

DK: I am wagering that you are also a religious absolutist. You guys always say " I bet you can never PROVE it 1000%..." or as you said "I contend it is an impossible task for you. "

Then, you always skip out of the debate, knowing full well that science--even though it challenges your long held and cherished G-d fantasies and biases--eventually discovers the answer, or a piece that leads us to the answer, whereas religion always posits that there are no absolute answers except a G-d...which no one can ever know, or see, or prove.harharhar

Greg:
http://ipoaa.com/africans_in_americas.htm

1. Woa, what a bunch of kookery!
2. Oh, the irony: soemone who tells that "race" does not exist, believes in the link that purports to prove - based on skin color and facial features - that Olmecs were Africans.
Q.E.D.
3. You lost your wager by a few million light years distance...

...American colonial history, slavery, etc...

Take out the 'American' and it reads truer.
You can apply it then to e.g. the saxon, or norman colonisation of britain, the chinese colonisation of malaysia or the arab colonisation of n africa.

Richard Dawkins in The Ancestors Tale points out that whenever biological criteria are applied to try and define race in biological terms, it ends up with the conclusion that measurable differences within such categories are greater than those between them.

DK:And the fact that today's Africans are easily separated from today's Native Americans based on a plethora of genetic traits tells that, genetically speaking, they are still relatiively isolated even today. Geography. Simple.

Here you are confusing distance with isolation.

Hint: How many colors are there? Do colors exist?

Colors are arbitrary names given to points on a spectrum. Like races. Maybe you are learning something new here, that would be good.

DK: Greg:
http://ipoaa.com/africans_in_americas.htm

1. Woa, what a bunch of kookery!
2. Oh, the irony: soemone who tells that "race" does not exist, believes in the link that purports to prove - based on skin color and facial features - that Olmecs were Africans.
Q.E.D.
3. You lost your wager by a few million light years distance...

Once again, you are mistaken. If you wish to argue with The Real Railer about this, fine, but don't tell me that I am wrong about something that I did not say!

Eddie: Yes, and here Dawkins is essentially referring to the well established biological anthropology to which I have been referring as well. But there are people who just HAVE to believe in race, like the medieval alchemists had to believe in the four humours or whatever. But it really is not appropriate or safe to ignore the science.

Hint: How many colors are there? Do colors exist?

Colors are arbitrary names given to points on a spectrum. Like races.

Precisely! But, of course, colors do have physical basis. And they do "exist". Like biological races (as per dictionary definition). Try going out and tell people "colors" don't exist and that their arbitrary classification makes no sense - everyone would rightfully think you are crazy in denying the obvious reality. Yet, exactly the same works fine for "races" - that is the power of denial stemming from people confusing biological and social concepts of "race".

DK: everyone would rightfully think you are crazy in denying the obvious reality

Absolutely. The average person with pre-conceived and well ingrained racial values and perspectives thinks I'm crazy. And if you believe airplanes fly because angels hold them aloft or that diseases sometimes go away because of prayer and so on, you might think aeronautics engineers or medical scientists are crazy.

And so on.

Of course, with color, to the extent that we agree on a definition of a name, saying that something is a particular color tells us what wavelength it reflects or emits. What does race tell us? As I asked above, what does it buy us as a concept? DK, what continua underlie the concept of race in such a way that race is a useful, biologically valid construct?

HP suggested:

If I were putting together a museum exhibit on race, I'd like to see a large photo showing a diverse group of humans, and then put up a series of overlays showing how they'd be classed racially by, say, a North American, a South American, a West African, an East Asian, a European, etc. I suspect there'd be a surprising amount of divergence.

There is something almost exactly like that in the exhibit. Good idea.

HP suggested:

If I were putting together a museum exhibit on race, I'd like to see a large photo showing a diverse group of humans, and then put up a series of overlays showing how they'd be classed racially by, say, a North American, a South American, a West African, an East Asian, a European, etc. I suspect there'd be a surprising amount of divergence.

There is something almost exactly like that in the exhibit. Good idea.

Bryan: Right. In fact, there are several different stations in the exhibit that allow people to explore race, identity, and ethnicity issues, including a test where visitors classify people, a series of photographs where people can write down impressions and also read what the person in the photograph said, a place where you can attempt to guess the race or ethnicity by voice...and so on and so forth.

DK, you could have used lots of other genetic differences to establish your "races", why did you choose skin color?

DK: you posited that "africans are easily separated from South Americans, pre "slavery", and I just kinda' said "what crap". Are 'they' so easily separated? The Olmec hypothesis is one proposed by actual scientists,like Ivan Van Sertima, and yes, you are right that some kooks have seized on it as a means to disseminate kookery. But it does mirror your own basis of 'fact' because you uphold racial categorization based in such physical traits.

What do you really say about the Olmec heads--as scientists more qualified than you have observed them. They were here several hundred years before "slavery". The only kookery here is your sort of desire to continue race based categorization of human beings.

Ivan Van Sertima et al. are the guys who proposed this model of racial feature based archeological interpretation of Africanism ( "They Came Before Colombus")--and some of them are relics of 1960's and 70's Pan African revolutionary ideology--ironically ideology very similar to yours in that they too have taken apparent physical 'traits' and run with them as the sole determinant of racial categories.

But they also go one further than you in that at the very least, they propose a theory that contradicts exclusive race categorization--they propose that Africans had an older and larger stake in human migration, just like the rest of the continentals, and thus are part of the same human family, sans race. You haven't made it of the eastern shelf yet, because these Olmec heads, and the hypothesis behind them mirror your own color and facial feature argument FOR race-based ( and then, I presume, social and class based) categories.

Van Sertima, et al. also discuss artifacts and 'evidence' of things that are local to Africa or S. America transplanted in other continents long before your own hypothesis. The colors painted on the royal tombs of both Egypt and Mecican/Central american tombs and artifacts are the same colors (real actual human painted colors); the coca leaf in Egypt ( I noticed that you didn't mention that) and yes, similar artifacts like the olmec heads--but you only noticed big lips, and flared noses--not the more subtle similarities to western African helmet styling, or the tenuous similarity of similar religious ritual practice.

That same face of the Olmecs can be found everywhere I have ever been. I still see it in Mexico; in China this face also has a long history too, but recently I have noticed that many Chinese women describe men with this face as 'zhong guo zhu' or "Chinese Pig" and they expect bad treatment from any man with such a face. In Japan--where "blackanese" babies are a new phenomenon, you can see this same face on the elders as well. Oh, yeah: and this face is all over South America too, and I have seen similar faces in the Scandihoovians.

Kookery is when racist arguments continue to be made in the 21st century. I mean, do you differentiate race when it comes to white folks? I mean, technically, your skin--just guessing, based in your colored worldview--is pink, and if not, some distinct shade of modern brownshirt. Do you tell people you're pink? Or are you, in your own egocentric view, a 'white person'?

My family roams the range of 'light pinkish beige' to 'distinctly pink' with red hues' and leaps the spectrum to ' olive toned', ' putty-taupe' and 'river rock red', with a few of our later progeny distinctly 'coffee and cream' or " buttery vanilla strawberry" and some 'cherry mocha' with some recent 'mellow citrus-tangerine'added...

But I acknowledge that I lost the part of the argument about you being the sort to skip out of a debate;-) However, the jury is still out as to whether you are religious, or an absolutist....

To Stephanie Z:
What does race tell us? As I asked above, what does it buy us as a concept? DK, what continua underlie the concept of race in such a way that race is a useful, biologically valid construct?

The fact that populations vary in frequencies of alleles already makes "race" it biologically valid. Just like with colors that vary with the wavelenth. So both races and colors are "real". Good we have established this.

Now you are asking a different question, question of utility. E.i., why bother? The answer: Since race/ethnicity correlates with having a particular combination of alleles (golly, I hope no one here is going to deny that - the literature on it is overwhelming!), it can serve as a crude proxy for genotype. This is important in e.g. medicine:
http://genomebiology.com/2002/3/7/comment/2007

Sure, when we can sequence and analyse genome of every single human being, that classification may become obsolete. Just like as we could sequence every living creature, there would be no need for taxonomy as we know it. But we can't do that. Not even close.

DK, humans are somewhat more complex than photons. If you have to resort to putting words in my mouth, you've already lost your argument. Again, which continua were you talking about?

I had hoped that when I said I'd asked the question before, you'd bother to go back and read that exchange. You might also want to read the discussion further up on population sampling.

DK: you posited that "africans are easily separated from South Americans, pre "slavery", and I just kinda' said "what crap". Are 'they' so easily separated? The Olmec hypothesis is one proposed by actual scientists,like Ivan Van Sertima, and yes, you are right that some kooks have seized on it as a means to disseminate kookery. But it does mirror your own basis of 'fact' because you uphold racial categorization based in such physical traits.
What do you really say about the Olmec heads--as scientists more qualified than you have observed them. They were here several hundred years before "slavery". The only kookery here is your sort of desire to continue race based categorization of human beings.
Ivan Van Sertima et al. are the guys who proposed this model of racial feature based archeological interpretation of Africanism ( "They Came Before Colombus")--and some of them are relics of 1960's and 70's Pan African revolutionary ideology--ironically ideology very similar to yours in that they too have taken apparent physical 'traits' and run with them as the sole determinant of racial categories.
But they also go one further than you in that at the very least, they propose a theory that contradicts exclusive race categorization--they propose that Africans had an older and larger stake in human migration, just like the rest of the continentals, and thus are part of the same human family, sans race. You haven't made it of the eastern shelf yet, because these Olmec heads, and the hypothesis behind them mirror your own color and facial feature argument FOR race-based ( and then, I presume, social and class based) categories.
Van Sertima, et al. also discuss artifacts and 'evidence' of things that are local to Africa or S. America transplanted in other continents long before your own hypothesis. The colors painted on the royal tombs of both Egypt and Mecican/Central american tombs and artifacts are the same colors (real actual human painted colors); the coca leaf in Egypt ( I noticed that you didn't mention that) and yes, similar artifacts like the olmec heads--but you only noticed big lips, and flared noses--not the more subtle similarities to western African helmet styling, or the tenuous similarity of similar religious ritual practice.
That same face of the Olmecs can be found everywhere I have ever been. I still see it in Mexico; in China this face also has a long history too, but recently I have noticed that many Chinese women describe men with this face as 'zhong guo zhu' or "Chinese Pig" and they expect bad treatment from any man with such a face. In Japan--where "blackanese" babies are a new phenomenon, you can see this same face on the elders as well. Oh, yeah: and this face is all over South America too, and I have seen similar faces in the Scandihoovians.
Kookery is when racist arguments continue to be made in the 21st century. I mean, do you differentiate race when it comes to white folks? I mean, technically, your skin--just guessing, based in your colored worldview--is pink, and if not, some distinct shade of modern brownshirt. Do you tell people you're pink? Or are you, in your own egocentric view, a 'white person'?
My family roams the range of 'light pinkish beige' to 'distinctly pink' with red hues' and leaps the spectrum to ' olive toned', ' putty-taupe' and 'river rock red', with a few of our later progeny distinctly 'coffee and cream' or " buttery vanilla strawberry" and some 'cherry mocha' with some recent 'mellow citrus-tangerine'added...
But I acknowledge that I lost the part of the bet about you being the sort to skip out of a debate;-)
However, the jury is still out as to whether you are religious, or an absolutist....

Since race/ethnicity correlates with having a particular combination of alleles (golly, I hope no one here is going to deny that - the literature on it is overwhelming!)

Denied.

Your first presumption (and gee, we don't agree on that) about humans being divisible into internally relatively homogeneous groups that have more variation between than within is the first racist error. Now you are making the second (for most racists) error. No, there is not a correlation. There might sometimes be an apparent correlation between traits presumed to be but not proven (or generally believed to be) genetic and categorical variables that are bogus ... so yes, you will find a pile of literature correlating IQ to Race, the former not being easily explained by genes and the latter being explained by ... well ... making shit up. I am guessing that that is what you are referring to.

Or is there some other literature you are referring to? If so, cite it.

It is possible that ethnicity/background/parantage can in some rare instances be useful in a medical setting to make a totally dumb guess about a diagnosis get a little less dumb. Then there are a few well known genetic diseases that one can consider. But really, the genetic correlation between race identifiers (either by observation or self-assessed) and medically important alleles is very very poor in all cases. All. Cases.

The article you cite is part of what I call an academic circle jerk. There is no new research here, but rather, a set of alternative interpretations of prior work, whereby a straw man is set up, shot down, and then the work set aside inappropriately. What is left is then a basic race-based model, but this race based model comes only from the presumptions of the authors, and is not a proper null model. The reason it is a circle jerk is something of another matter. Let's just say that one needs to have a couple of opinion pieces like this in the literature so one's race-based research can be justified by citing said opinion piece.

There are times when a physician would be wrong to ignore a person's genetic background. But more often than not this only slightly refines a diagnosis, and if the diagnosis is important and the treatment both meaningful and risky, not doing any available genetic test would be unthinkable.

Since race/ethnicity correlates with having a particular combination of alleles (golly, I hope no one here is going to deny that - the literature on it is overwhelming!)

Denied.

Your first presumption (and gee, we don't agree on that) about humans being divisible into internally relatively homogeneous groups that have more variation between than within is the first racist error. Now you are making the second (for most racists) error. No, there is not a correlation. There might sometimes be an apparent correlation between traits presumed to be but not proven (or generally believed to be) genetic and categorical variables that are bogus ... so yes, you will find a pile of literature correlating IQ to Race, the former not being easily explained by genes and the latter being explained by ... well ... making shit up. I am guessing that that is what you are referring to.

Or is there some other literature you are referring to? If so, cite it.

It is possible that ethnicity/background/parantage can in some rare instances be useful in a medical setting to make a totally dumb guess about a diagnosis get a little less dumb. Then there are a few well known genetic diseases that one can consider. But really, the genetic correlation between race identifiers (either by observation or self-assessed) and medically important alleles is very very poor in all cases. All. Cases.

The article you cite is part of what I call an academic circle jerk. There is no new research here, but rather, a set of alternative interpretations of prior work, whereby a straw man is set up, shot down, and then the work set aside inappropriately. What is left is then a basic race-based model, but this race based model comes only from the presumptions of the authors, and is not a proper null model. The reason it is a circle jerk is something of another matter. Let's just say that one needs to have a couple of opinion pieces like this in the literature so one's race-based research can be justified by citing said opinion piece.

There are times when a physician would be wrong to ignore a person's genetic background. But more often than not this only slightly refines a diagnosis, and if the diagnosis is important and the treatment both meaningful and risky, not doing any available genetic test would be unthinkable.

Or is there some other literature you are referring to? If so, cite it.

http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1196372&blobtype=…

And many, many others. Every single paper that looked at clustering of genetic markers vs populations. (Plus, I already gave a link to a review on the subject; not that I expect you to have actually read it ... but, you know, it has references).

You see, something that you say does not exist is >99% accurate predictor of having a particular combination of genetic markers. Not bad for a made up stuff that has nothing to do with reality, isn't it? On this alone, any normal person (that is, not insane denialist), would conclude that whatever people chose to call "race" has GOT to do have an underlying biological basis. And, I must say, it is entirely not suprising result (that is, for anyone who knows basic population genetics - which obviously excludes you). In fact, there is NO WAY it can be any other way around. Major "races" that roughly correspond to continental origins represent evolutionary lineages - and that MEANS differences in frequencies of alleles!

And here is a really good commentary that places this basic fact into proper context:
http://psycnet.apa.org/index.cfm?fa=buy.optionToBuy&id=2006-01690-008

Greg,

- Yes, all the people at one end of the room can't hear the people at the other end of the room talking, but eveyone can hear the person standing next to them.

Ring species. Do they disprove the "reality" of species?

- Do I read you right? "race is not biologially real" and "color is not physically real" are statements of the same general type, that both are true in the same sort of way and that he who denies either statement is akin to someone who believes angels hold up aeroplanes?

Some questions:

Because there's a continuous transformation between having lots of hair and having none, "baldness is not real." Agree?

Because as one travels in any particular direction from New York speech patterns change continuously, "accent is not linguistically real." Agree?

I find it surprising that a science blogger can so strongly reject the idea that fuzzy (and not therefore useless) categories can exist across continuous variation.

I may have missed one, but it seems uniform that the racists' examples are strictly in one variable. That's part of why they can't understand the meaninglessness of race applied to human biology. If you insist on only one character to be used to define race, and decide that 'color' is that character, then we could indeed set up a photometer and scan people from 'dark' to 'light' and arbitrarily set up three (but why are you taking 3 rather than 12, or 2? purely social reasons) 'races', 'dark', 'intermediate', 'light'. The 'dark' would include most (but not all) sub-Saharan Africans, most Indians, and Australian Aborigines. And right there you've got problems for claiming it's a biologically useful grouping, as the Australians are among the most removed, biologically, from the sub-Saharan Africans. Not only is there continuous variation, but the direction of variation -- on any trait -- also reverses direction as you follow a path around the earth.

Groups with fuzzy boundaries are one thing, and can be used sometimes. Groups that collect disparate things together, or require you to exponentially expand your number of groups are not. You could avoid the above problem by multiplying your number of 'dark' groups, but you'll have to add another trait to separate them from each other. And as soon as you do that, there will be reversing gradients between them as well. So you triple (or more) the groups again on the second trait by adding a third. And so on. You can continue the process if you like, but the end state is that the number of 'races' approaches the number of people in the world.

By the time you finish, even siblings will be getting classified as different 'races'. One of my sisters is dark haired, one is blonde.

Fuzzy groupings is one thing. Stupid groupings, we do indeed avoid in science.

DK: Again, in the paper you cite, the data are divided into categories based on self identificiation. This is the SAME phenomenon I pointed out way upstream with lattitude and longitude. The incorrect conclusion with the lat/long data would be that there is a quantum pattern to geography on the planet earth, the parallel conclusion is that there are valid races, both wrong.

And that same criticism applies to many many other papers.

Did I say that color is not physically real?

Baldness is, by the way, a binary state for most individuals.

Your accent example is a good one and one I know something about. Linguists know that dialect and accent sometimes form harsh boundaries, and other times grade much like genetic variation grades. Dialect/language and genetics/ethnicity/race have interesting parallels and interesting differences.

Did you now that the boundaries between many languages are arbitrarily drawn? I'll bet you thought languages are just like your concept of races.

We can talk about fuzzy categories and continuous variation all you want, but you are not allowed to change horses mid stream. You have been advocating for a standard race based model (with the standard five or six races). Well, actually, you can change horses mid stream if you like, but admit it when you do it please!

Robert: Sorry, I wrote my comment above having not seen yours. Right, well said.

DK, one more point about the paper you cite. The authors say this:

It is clear that the ability to define distinct genetic clusters depends on the number and type of markers used. Reports that document inability to define distinct clusters generally used only a modest number of markers and, hence, had little power to detect cluster. Studies with larger numbers of markers appear to show strong evidence of clustering. [citations removed for ease of reading]

What this actually says is that geographical origin of ancestors is not very good at predicting the presence of a particular genetic marker. If it were, small groups of markers would work nearly as well as large groups. This rather limits the utility of origin-based racial categories in medicine, even in those disorders where genetics play a large role.

D again. I'm not DK. The post about accent, baldness and ring species (again, do they demonstrate species aren't real?) was the first I've made this thread.

Did I say that color is not physically real?

Another poster wanted to know how, given the argument you'd made to him, you'd call color real and race unreal. You decided to bite a quite incredible bullet and say "Colors are arbitrary names given to points on a spectrum. Like races. Maybe you are learning something new here, that would be good."

You should go back and explain why color is a valid fuzzy category, but race isn't. Or why language is a valid fuzzy category but race isn't. Then again:

Linguists know that dialect and accent sometimes form harsh boundaries, and other times grade much like genetic variation grades. Dialect/language and genetics/ethnicity/race have interesting parallels and interesting differences.

I'm sure you know a lot about language. Of course there's a lot of fuzziness and politics about (say) whether Mandarin and Cantonese are "dialects" or whether Urdu and Hindi are separate "languages." It takes large doses of sophistry to go from there to language-is-not-real. I'm sure the people at Language Log would be delighted to hear that.

Anyhow, at least one understands something like the sense in which you mean race isn't real: the sort of sense in which language, color, baldness, accent and species are not real. At this point I want to conclude that the word 'real' for you is a term of art, and that you're not using it the way people normally do. That you then get unexpected conclusions is by no means surprising.

Either that or you just enjoy being needlessly fastidious with categories when it suits your (political?) purposes. Whatever.

D, color is a useful parallel to race, as is language, in that they are all social constructs that may or may not have a basis in underlying physical reality. The fact that one does have a single, fairly simple underlying reality doesn't prove that the others do. Each of our social constructs is only as valid as what it tells us about reality. That's all that determines how "real" they are.

D: What I said was that DK (sorry for confusing you) started out with something very different than fuzzy categories. This is the usual pattern of unwilling learning.

The comparison between colors and race is actually a reddish herring. There are similarities and differences that make this not a good analogy, and I consider it a straw man. Color categories are not arbitrary, in fact, and seem to be culture bound and related to the physiology of vision and so on.

Language is real. The boundaries beetween some languages is arbitrary. Languages tend to have self-sustaining features, and dialectic boundaries have very consistent and long lived self-sustaining features that only sometimes have parallels in gene flow.

Again, race and language are similar and different. Again DK is creating a straw man. Again the sophistry, as you put it.

Oh, and the final conclusion of most of these conversations is exactly what you see here. The racist who refuses to even consider the science makes the accusation that the scientist is being political. I'm guessing, D, that you have virtually no training whatsoever in the sciences, or you would know better!!!! ......:)

Greg: No apologies necessary, and thanks for the good word.

D, like other racists, you keep not noticing the important qualifier that Greg and I and others have used -- biologically real. Race is certainly a social reality. It just isn't one that comes out of objective analysis of biological data. See my examples already given. The collapse of race as a biological concept was also much more a matter of centuries of racists trying to find something to guarantee that they were in one race while the people they hated, feared, or deemed 'inferior', were some other race. After failing constantly, there was enough honesty to realize that it just isn't something that comes from the objective biological data.

Now you want to take language as your proxy for racist classification. Ok. How do we know that someone is speaking one language and not some other? One standard linguistic test is inter-comprehensibility. We say that English and Mandarin are different languages because native speakers of (only one of) either cannot understand native speakers of the other. Go to biology -- can and do members of one race mate successfully with members of another? Absolutely. Might be by way of rape, but they definitely do. So your analogy is already shot. Races cannot be mapped to languages; if you insist on analogizing biology to language, languages would have to be species.

So you drop to the analogy being races as dialects. What's the difference between a dialect and a language? An army. What are the differences between dialects? Who will hit you if you say they're speaking an other-named dialect. Social declarations, in other words. We had, therefore, the business in the US in the 90's of 'ebonics' education, where it was claimed that Blacks spoke a single dialect (or language, claims differed) and should all be taught that dialect/language in school, etc. Except they don't and never have. This was particularly clear as TV interviews shifted from people in different parts of the US, black and white, and the language much more clearly reflected what part of the country the speaker was from than what race. The Chicagoan (African) spoke much more like me (European, from Chicago area, mostly) than he did any African from the DC area (where I lived at the time). So if you want to preserve that black/white distinction by language, you'll have to do it by doubling the number of dialects in the US, with white Chicago sounding much more like black Chicago than either of them sound like New York (white or black). Same problem for 'race'. To preserve the social mapping of race into biology, you'll have to split things very close to each other, while lumping them together with things much farther away than they are to each other.

Linguistics in general seems heavy on splitters, so that there's a significant number who would not bat an eyelash, indeed they revel in the idea that my sisters and I are speaking different languages from each other. Such linguists would also happily agree that I speak (let's see: family, scientific, random public, sports, written formal, blog ...) 6-10 languages. It's just my ignorance that I call all of them 'English'.

That may work in linguistics, for what they're trying to understand (though I doubt it really does). It doesn't work in biology for what biologists are trying to understand.

StephanieZ -
yes. Of course it isn't the case that all of {language, color, race, baldness, accent} have to be real or none of them is. I was quite astounded to see Greg Laden seem to say something of the about race and the colors of light being equally real or unreal. Even the Charles Murrays of the world might consider that a bit extreme. Anyway, I obviously don't think he believes anything bizarre about colors not being real (as a physicist I'm glad!). I just don't think he has given a particularly cohesive account of how colors or accents are real but race isn't. Certainly the fuzziness test seems to me to fail.

Greg
- I agree it's quite important that languages have certain self-sustaining features and boundaries (which races don't), though I do wonder what you'd say about creoles.

- Anyway, the point of the accent / language / color / species analogies wasn't "these exist objectively / intersubjectively so race must too." Rather, it was "fuzziness doesn't obviously eliminate race as a category unless you'd have it eliminate other similar fuzzy categories."

- I think you're missing the point of the colors question. It is that color would seem to be a difficult case / counterexample to your account of what makes race unreal. In that context, you very definitely have to explain how your ideas don't render these things unreal after all. Whether language or color are in fact just as ontologically shallow as race (they obviously aren't, no matter how real or unreal race itself is) is neither here nor there.

Robert Grumbine
Asshole, reread my two blog comments and tell me what I said that makes me a racist. I refuse to waste time on you till you do. Prick.

I just don't think he has given a particularly cohesive account of how colors or accents are real but race isn't.

I don't think I need to make that comparison. I did not bring colors or accents into this discussion, and certainly would not have in the way they are being used by DK.

I agree it's quite important that languages have certain self-sustaining features and boundaries (which races don't), though I do wonder what you'd say about creoles.

Again, a very interesting topic. I'm sure I've written about creoles somewhere here or on my other blog. But what happens with languages is not that instructive as a model for race.

In order for me to say that Race is not a valid biological concept in humans one has to understand what is meant to say that race IS a valid biological concept in humans.

The latter entails these features:

1) the species can be divided into categories that are internally more homogeneous than externally with respect to the distribution of a large number of alleles.

2) This leads to the ability to predict the allelic composition of an individual for alleles that are invisible (on causual observation) based on the visibile evidence of other alleles. I see that you have black skin, kinky hair, and a certain kind of nose so I call you "African" or "Black" and thus assume that you are about to mug me. I see that you have dark hair, a whopping big nose, curly hair and a furry body so I know that you should not be trusted with my sister but you certainly can be trusted to know what to do with my money. I see that you have lily white skin, blondish hair, a properly proportioned nose and blue eyes so I know that you have a small penis but would be really good at running my country or enslaving my sister and probably can do math if you are a male; I see that you have a broad face, crazy hair, pretty dark skin and live in Australia, so I know that you are actually a form of Homo erectus; I see that you have yellowish skin, slanty eyes (like in the Olymics) and dark oily hair so I know that you are very very smart but cannot be trusted because you are also very very tricky. And so on.

3) There are typically about five or six races according to most models.

However, I observe instead:

1) The vast majority of alleles are uncorrelated to each other. The only correlations that exist are where founding populations have spread. There are hudnreds of examples of this, and since each can't be a different race (or is never classified as such) only a handful of these examples are actually employed in the race concept.

2) There are almost no clear boundaries ... indeed, there are almost no fuzzy boundaries ... between any group and any other group across the species. The closest thing we get to boundaries are the temporary distinctions that emerge from historical movements of people and similar events, which are known (from history) to eventually go away. Most purported races in fact do not have any of these boundaries at all, or only along very limited 'margins' of their population.

3) The vast majority of racial traits blend across boundaries, and the vast majority of predicted traits ... the traits that those engaged in racial thinking claim to be able to predict ... are not genetic or are so heavily influenced by non-genetic factors that claiming them to be genetic is absurd.

4) The vast majority of claims by specific race-based thinkers are those that would, if true, make the claiming-group superior in some way to the group about which the claim is being made. This would not invalidate the science (if there were real science) but it does explain the stupididty.

5) The fact that races do not exist and race-based correlatational and everything-is-genetic (even when there is not an iota of evidence) thinking does not obviate that allelic variation is real, and sometimes important. But allelic variation need not be (and in fact is not) organized in a race/subspecies pattern for humans.

Get. Over. It.

Greg: "D" is not "DK". But still:
Did I say that color is not physically real?

No, you did not although you should have had. You see, "green" is really not a physical reality. It's an abstract concept, part of the arbitrary classification. Logically, if one denies biological reality of "race" based on continuous variation, one HAS TO deny physical reality of "colors". No other way around. So you have only two choices: 1) admit that continuous variation is not a proper argument contra existence of "races" and come up with a better one or 2) stick to the insane argument that "green" does not exist.

This is the SAME phenomenon I pointed out way upstream with lattitude and longitude. The incorrect conclusion with the lat/long data would be that there is a quantum pattern to geography on the planet earth

You don't know it but you got it right (almost). 1) There are indeed "Southern states" and "Northern states". They do exist - as abstract concepts of physical reality (just like races and colors). And the basic physical reality behind them is latitude. 2) Broadly defined combinations of latitude and longitude give continents - and we know from genetics that populations on different continents differ in frequencies of genetic traits (do read the papers I gave links to). 3) I don't see how any sane person would come up with the idea of a quantum pattern to geography. Care to elaborate?

You have been advocating for a standard race based model (with the standard five or six races).

No, I have not. I twice quite explicitely said that the number of races we chose to define has absolutely nothing to with the reality of biological races in any species.

We can talk about fuzzy categories and continuous variation all you want, but you are not allowed to change horses mid stream.

What are you talking about? So far your only argument against the physical reality of races was continuous variation. The only one. (Calling me a racist does not count, I hope). Feel free to provide another one.

Your denial of the perfectly well established fact that race/ethnicity correlate with having a particular combination of alleles makes your position even less tenable. That is just as silly as denying evolution. (The other kind of denialists who refuse to accept basic scientific facts).

The paper:
http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.00100…
The figure from it:
http://www.plosgenetics.org/article/showImageLarge.action?uri=info%3Ado…

The paper:
http://www.plosgenetics.org/article/info%3Adoi%2F10.1371%2Fjournal.pgen…
The figure from it:
http://www.plosgenetics.org/article/showImageLarge.action?uri=info%3Ado…

The paper:
http://www.plosgenetics.org/article/info%3Adoi%2F10.1371%2Fjournal.pgen…
The figure from it:
http://www.plosgenetics.org/article/showImageLarge.action?uri=info%3Ado…

The paper:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedi…
The figure from it:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1852743&rende…

And so on. The list is almost endless by now. I only included few random in open access.

Question: Has there been research whereby the presumed traits themselves (rather than race) are the independent variable and allelic identification, the dependent variable?

If we assume certain alleles correlate with certain traits (e.g., IQ, personality, socioeconomic status, etc) rather than assessing by race, what about assessment via the traits themselves? As an example, determining whether similarities (and variations) exist in alleles of persons classified as 1,2, and 3 standard deviations above and below the mean on a particular standardized IQ test.

If it is observed that no defined categories of a particular trait (e.g., persons with IQs, 70, 80, 90, 110, 120, 130) correspond to systematic similarities and differences in their alleles, this would seem to be the most direct evidence that the assumed direct correspondence between alleles and observed IQ (and other such traits) is faulty. And therefore, any subsequent additional hypotheses (such as race) based on this (likely) untested premise are to be called into question.

Actually, the best evidence is swapping babies from one race to another to see if when they grow up they act like the adopted parents or the native parents.

Turns out they act like the adoptive parents in all of the traits usually assigned to race categories.

There is a reason why we have a developmental period that lasts decades.

LOL. I posted a large message that exposed Greg's fatal flaws of reasoning. Got a message acknowledging the post. Yet it never made it in here... There was nothing profane, offensive or off-topic in it. What's up with that, Greg? Can't stand it? Practicing oh-so-typical PC censorship, perhaps???

DK, did you have more than one URL in your post? Those are automatically held until Greg notices them. Happens to me all the time. He's much more interested in letting you expose your thinking to the world than you would guess.

Practicing oh-so-typical PC censorship, perhaps???

You have no clue, do you.

Your post was held in moderation by the software, default settings, because of the links. It should be free now.

Yeah, there were eight URLs. I did not know about this bug - errr, feature.

You have no clue, do you.

No I don't. Why should I have a clue about strange default settings of some particular blog?

My sincere apologies for suspecting you in censorship! I should have waited longer before posting such a message. My bad.

Greg wrote:
Actually, the best evidence is swapping babies from one race to another to see if when they grow up they act like the adopted parents or the native parents.

Turns out they act like the adoptive parents in all of the traits usually assigned to race categories.

Actually, care to provide some references to this claim? That's not what Minnesota transracial adoption study has found (adopted black children's IQ very significantly lagged behind)

http://en.wikipedia.org/wiki/Minnesota_Transracial_Adoption_Study

or what Korean adoption study found (adoptees' family income was found to be independent of adoptive parents' income; in stark contrast to biological children where the two are positively correlated).

http://www.nber.org/papers/w10894.pdf

DK, mostly because of what it says under Post a Comment and the fact that the message you got when you submitted the comment told you it was being held for the number of URLs it contained.

Actually, the best evidence is swapping babies from one race to another to see if when they grow up they act like the adopted parents or the native parents.Turns out they act like the adoptive parents in all of the traits usually assigned to race categories.

I've come across such studies before, but do not recall the methodological details at the moment. But generally I do know shared vs. nonshared environment assumptions are factors that make the data from such studies difficult to interpret (e.g., see Sundet,Eriksen, & Tambs, 2008). For example, to be certain that cross-race cultural differences account for the similarity in the behavior of adopted children to their adopted parent(s), one would have to: (a) assume that the (presumed) genes responsible for the traits connected with the biological parents are different from those of the adoptive parents; and/or (b) assume that the cultural environment of the biological parents is different from that of the adoptive parents (the nonshared environments assumption). A good example of how these environmental assumptions can make such studies difficult to interpret can be found on CogDaily's Parents' influence on kids' behavior: Not much

There is a reason why we have a developmental period that lasts decades.

Well, behavioral genetics refers to a concept called niche-picking, which suggests that as people get older, they tend to choose environments compatible with their genetics. This is based on the finding that fraternal twins become increasingly dissimilar on various traits with age, while identical twins become increasingly similar with age. As another example, what might explain the difference in brain activity of liberal vs. conservative minded college students?

Being adopted, I can say I do have some qualities similar to my adopted parents, and some particular qualities that I'm pretty sure I don't share (which may be genetic). However, I know nothing about my biological parents so can't arrive at a conclusion. I'd bet most (adopted or not) would say something similar. Which again, makes things complicated because one cannot tell whether this is more genetics, environment, or psychological (The consensus in psychology is that it involves a complex interaction and so we've moved away from the how much question to the how).

DK: There are about a dozen studies like the particular cherry-picked ones you cite, which all show the same result: No significant effect. In the case of the first study you cite, the people who did the study concluded "that racial group differences in IQ are due to environment only"

Subsequent to that particual study, the racist group including Jensen and others 'reinterpreted' the study but did a poor job. And, they and you fail to recognize what the study really means and how to interpret it. SO, the best you can do is a study that clearly shows that environment determines IQ and that a couple of yahoos dispute the conclusions of.

You realize that the NBER is a business journal, yes? Do you know the background behind this particular paper? (Oh, NBER is not peer reviewed, by the way.)

There are several questions you need to address. Step back and consider some of these issues that you have been ignoring or glossing over:

1) You insist on basic race based categories, but do you know what they are, what defines them? Are Swedes somehow more white than Greeks and Spaniards? Are Nigerians more black than Eritrians and Namibians? Are the Chinese more asian than the Japanese and the Nepalese? (see this)

2) You are defening this archaic, out of date concept of race so vigorously. What does this concept of race buy us? What does it tell us that something else would not tell us? (see this)

3) To expand on this, and use your example of color, to the extent that we agree on a definition of a name, saying that something is a particular color tells us what wavelength it reflects or emits. What does race tell us? As I asked above, what does it buy us as a concept? DK, what continua underlie the concept of race in such a way that race is a useful, biologically valid construct? (I got this idea here.)

4) Regarding the usual medical "answer" for this question, ... geographical origin of ancestors is not very good at predicting the presence of a particular genetic marker. If it were, small groups of markers would work nearly as well as large groups. This rather limits the utility of origin-based racial categories in medicine, even in those disorders where genetics play a large role. (See here.)

What I'm saying here, really, is that the tendency to insist on a race based model is often linked to a parallel tendency to a kind of sexist tunnel vision at the same time.

Or am I being too subtle...

DK, did you read the links that you provided or did you just look at the pretty numbers?

"Yet, it was previously known that adoption into upper-middle class White families has a positive influence on the IQ and school performance of White children."

Could you tell me which gene expresses the accumulation of wealth? Maybe the lottery gene by chance?

"As Scarr & Weinberg (1976) note, transracial adoption studies only control for family environment, not social environment."

So NOT social environment but I would guess that even family environment would be a stretch to "control".

But even at that, IQ doesn't measure the ability to learn.

There are several questions you need to address. ... Are Swedes somehow more white than Greeks and Spaniards?

No. No way I am stepping into your trap of addressing ill-defined questions. WTF "more white" can possibly mean as far as biology goes? Nothing, at best! I am also going to skip the whole IQ thing because that's just too complex and a distraction. Plus, I have already answered all the points you are making anyway.

So lets keep things to the point and simple. The issue is biological race. I have given you a dictionary definition of biological race. Completely obviously (review the thread, read references), human populations as we know them fit that definition. Just to make sure you remember:

RACE
5. Biology
a) An interbreeding, usually geographically isolated population of organisms differing from other populations of the same species in the frequency of hereditary traits. A race that has been given formal taxonomic recognition is known as a subspecies.

Well, you or anyone else may not hold American Heritage Dictionary to be an authority. Fine then. So lets look at few others. The excellent Long & Kittles (2003) paper entitled "Human Genetic Diversity and the Nonexistence of Biological Races" gives four different ones. The paper itself is not open access but Wikipedia has a quote:
http://en.wikipedia.org/wiki/Race_(classification_of_human_beings)#Summ…

Now, I like AHD's definition better than any given by these famous scientists but it's quite obvious that human populations fit any of them. Indeed, Long & Kittles go on to say that "a good deal of genetic diversity within groups is consistent with the entire spectrum represented by these concepts". Even Francis Collins, no matter how PC he is trying to be, has admitted that "it is not strictly true that race or ethnicity has no biological connection"
http://www.genome.gov/Pages/News/Documents/RaceandGeneticsCommentary.pdf
(This is, BTW, not a bad article. It makes a number of pretty obvious points but does it quite succintly. E.g. "race is an imperfect surrogate for ancestral geographic origin, which in turn is a surrogate for genetic variation across an individual's genome. Likewise, genome-wide variation correlates, albeit with far-from-perfect accuracy, with variation at specific loci associated with disease. Those variants interact with multiple environmental variables, with the ultimate outcome being health or disease. Considered in this context, it is apparent why self-identified race or ethnicity might be correlated with health status, through genetic or nongenetic surrogate relationships or a combination of the two. It is also evident that a true understanding of disease risk requires us to go well beyond these weak and imperfect proxy relationships. And if we are not satisfied with the use of imperfect surrogates in trying to understand hereditary causes, then we should not be satisfied with them as measures of environmental causation either."
I agree completely. That only thing that's not mentioned here and is rather ironic is that without the use of the concepts like "race" (or call it ethnicity or population or what have you - it's the biology that matters, not semantics), such a research cannot effectively proceed beyond the obviously imperfect racial classification of some kind without using them in the first place).

Since you insist that races don't exist, the onus is on you to show - using the logic and facts supported by references - that human populations do not conform to the traditional biological concepts of within species variation described by the biologists using concepts of races and/or subspecies.

If you decide to try to defend your POV, please DO NOT invoke the lack of absolute boundaries. Else you will be forced to conclude that "species", "trees", "mountains", "planets" and just about any other abstract concept do not exist either. You probably don't want to repeat the mistake you already made with "colors" :-)

Since you insist that races don't exist, the onus is on you to show - using the logic and facts supported by references - that human populations do not conform to the traditional biological concepts of within species variation described by the biologists using concepts of races and/or subspecies.

Question: How many times to I have to explain this to you?

Answer: Just this one last time, actually.

If I proffer the Periodic Table of the Elements, and you say "No, sorry, that is not right. Since you insist that the four humors don't exist, the onus is on you to show - using the logic and facts supported by references - matter does not conform to the traditional alchemic concepts bla bla bla." then that makes you an uneducated moron.

The traditional concept of race in human is wrong, destructive, overturned, thrown out, and the hobgoblin of racists. It is not the null model.

The null model is random variation in genes overprinted with moderate selection, clinal distributions; in the case of a highly mobile species, biogeographical barriers having little impact in most cases; and frequent and sometimes rapid change.

I'm going to end this with two observations that I think are interesting.

If you take the view that races exist and then classify humans into races (which you can do because arbitrary categories can always be imposed on a continuum, even if incorrectly or uselessly) you would get the following two very interesting results:

1) Using the usual racial traits

... hair[shape (kinkiness), color]; eyes [color of iris, eye shape/epicanthic fold], Skin color [melanin intensity (blackness of skin), keratin intensity (yellowness of skin), subcutaneous blood effect (pinkness of skin)], nose lenth, nose width, nose 'hookiness'; Face and head shape [midface flatness/prognathism, brachiocephaly]; ratio of flacid to erect penis size; breast shape/size; and stature ...

... these are THE racial traits used for the traditional racial classificiations (there are a dozens others but these are the classic traits) and reapplied them to a world wide sample, you would get about 12-14 races.

Nine or ten of these races would be African, mainly south of the Sahara.

2) If you took the racial traits found only in bone, such as incisor shoveling, presence/absence of a few other toothy bits, malar shape, nasal apeture shape/size, shape of occipital condyles, occipital bunnning, various cranial vault measuremnet, sand so on and applied this to the recent world (i.e., adjusting for recent movements of peopl, etc.) you would get a particular racial map of the human species.

If you reapplied the same analysis to skeletal material that was betweeen 5 and 7 thousand years, you would also get a particular map (Remember ... you are not getting a racial map becausae it pops out of the data .. you are getting a racial map because you are using a technique that seeks racial groups in the data and puts them on the map... your null model is a racial grouping).

If you reapplied this method to matarial that was 10-15 thousand y.a., you would also get a racial map.

And you would find the following (see Brace's work and Howell's for exploration of this):

1) None of the maps, including the one for recent times, would conform to expecations exactly. The recent map would look much like you expected, except some of the preconcieved races would be combined, others split.

2) None of the maps would look like each other. In other words, the entire world changed racially every few thousand years. THAT is not a prediction of any racial model of today or the past. In fact, it violates a series of predictions (see the post on the review above as to the importance of this kind of data.)

3) If you reanalyze the same data using techniques that find groupings (rather than assume them) you would find at best one or two (maybe three) groups that have very marginal statistical significance, controlled mainly by major borders, and that are totally at odds with expectations.

For instance, the earliest australian material would look very very different from Eurasia. The most recent Australian material would look different too, but in different ways. But the middle Australian material would look Eurasian. There might be a grouping that included much of Europe and Central Asia and North America, that does not look even a tiny bit "asian" by modern standards. And so on.

In other words, when you look at the data using appropriate methods of analysis (and actually read the papers in stead of scanning for racial categories being used in papers as though that proved your point), your pre-determined unchangeable 19th century racialized view of the world will go all topsy turvy.

Take a few weeks off, go back to the data and literature, experience this phenomenon and enjoy the learning experience.

Shorter DK: I don't wanna talk about whether race is a valid predictive measure of anything. It's a word. See? Here are a couple more articles I don't understand where people use it as a word. It's a word damnit!

Now, when you're ready to concede that I can keep my races, I'll be over here in the corner, wrapped around my dictionary with my fingers in my ears.

La la la la....

You could lose an Arizona Navajo in a pile of Japanese Ainu, skeletally speaking. That should give any racist pause.

To Greg:
OK, your evasions started to go in cycles so it's getting boreing. To sum up:

1. You were given various definitions of a biological race, a concept that is of general value for most biological specie and has nothing special to do with humans or racism.

2. You were given facts and references that unequivocaly show that genetics of human populations is consistent with any of these definitions.

3. If you wanted to continue to insist that "race has no biological basis", all you had to do was to show that genetics of H.sapiens is such that the widly held concept of biological races/subspecies is inapplicable to human populations.

4. You repeatedly refuse to do so. You completely failed to stick to the subject matter. Instead, you keep harking back to the "traditional concept of races" (which I have never invoked), acusations of racism and unbelievably silly things like hair kinkinesss and ratio of flacid to erect penis size.

5. And that means you last the argument. That you did is not surprising at all given given the wealth of genetic information about structure of human populations coming in the past decade. What surprised me is that you haven't even tried. (The lame attempt with the continous variation thing was an attack on taxonomy at large and had nothing to do with human populations; it also failed). In other words, you had your chance and you blew it. Pretty pathetic and ultimately boring. I expected higher intellectual standards and honesty here!

But DK, you didn't respond to my points. At any point, really.

DK: You come on my blog and say the same tired old thing over and over again, show me a half dozen or so references that support my argument (refuting yours), and you dare to call ME boring and not up to your standards?

You have been spayed, my friend. Eviscerated. By me. And by a girl.

Please take the arguments to the more current post on this topic, here, or to Stephanie's blog.

"You could lose an Arizona Navajo in a pile of Japanese Ainu, skeletally speaking."
Yeah, but if you did, Kennewick man might rise from the dead, and claim that he was actually an Indian, so to inherit the stewardship of the American Casino Empire....

Well, you could also lose Kennewick Man on the main deck of the Starship Enterprise.

By the way, Real Railer, your post was automatically moderated. I wonder which key word took effect there. Most likely casino...

Greg, considering that 1920s lynchings by the "new" KKK occurred as far north as Duluth, Minn., yes, it's a big oversite that lynching pics weren't included.

Yes, and it isn't even that Duluth was an outposts for racism. The KKK started up north. Minnesota is the home of the Modern (as in E. 20th century) white supremacist and so on.

Truly amazing. What a passion to deny that race can be meaningfully talked about, particularly as a biological concept. And there aren't any breeds of dogs either and Ashkenazi Jews were not a closely interbreeding group of remarkably similar founder families that merged but understandably, since Disraeli had a character modelled on a Rothschild say "all is race" in a novel (quoted by Hitler). have come to loathe the notion of race. I don't know why you go on about bones and skeletons when modern biology has the means to assess the probable ancestry of an individual by allele frequencies and sections of DNA. These are bound to have defined early interbreeding groups descended from tiny founder populations which were isolated from others for long enough to become distinct clans, tribes and races which statistical studies of their DNA would have easily shown to be assignable in nearly every case to one ancestral line or another. And the mixing has not gone anywhere near far enough to make such distinctions a relic of pre-history. True, most of the uses of race and "race" are chosen for political or social or in-group-emotional reasons and are not useful for legitimate purpose in a modern liberal democracy - as opposed to some hypothetical primitive Hobbesian world where outsiders were enemies and few could speak each others' languages. But the fact that it is at least possible to make a case for the utility of race for medical practice (pending the day when the individual genotype is always available) proves that biological race is, for at least one purpose, a reality. And it only needs one such instance to invalidate your passionate assertion to the contrary.

Of course there are statistical generalisations that can be made about races or sub-races precisely because they are based in biology. West African sprinters and East African middle and long distance runners prove the point as a matter of plain fact and logic. I did hear one dopy left wing academic turned politician insist that eventually women would run as fast as men but I don't accuse you of such a degree of eccentricity: he managed to convince himself that accepting bribes for doing something he thougt was a good thing anyway was OK and he went to jail. You wouldn't deny the probability of a West African beating an East African in sprints and vice versa over distance was biologically based would you?

Also, if you were a would be employer of software writing trainees in New York and an employment agency gave you three lists of 100 names, one of which you identified as all East Asian, one as African-American, one as Jewish and one as WASP would you have any trouble with choosing to go first to the Jewish, or maybe East Asian, list? It might well be that it was cultural, most likely family, influences which were the reason behind the reason, but, would you deny that it may turn out, in 200 years time that much of the one SD difference in IQ type performances of African-Americans persists and, despite the fact that it is the social decisions to self-identify and of others to identify them by skin colour that has maintained their African-American separateness, it might be concluded that a generalisation about them of considerable practical import has a biological basis? Of course we will know within 10 years or so from genetics and neuro-science whether the perceived African-American *average* differences on a number of cognitive and behavioural variables could really have any biological basis. Rather than be so assertive without evidence - except bones! - why don't you concede that you can wait for the scientific answer and in the meantime merely point out that the concept of race is useful for for very few legitimate public or private purposes.

Julius: I certainly hope that you are a person without any specific HR related responsibilities.

I think I understand. Greg is arguing, in effect, that the existence of mongrels falsifies purebred dog theory, which seems neither accurate nor useful.

No. There are purebred dogs because if you interbreed breeds the AKA comes out and shoots you and your dogs... My point was that dog "races" can be maintained only with extreme measures.

Would it help to understand that 100 years ago zoologists divided most mammals into multiple subspecies, and the race concept congealed about that time as well. Subsequently, the zoologists have realized that there are almost no mammalian subspecies. They had it all wrong. All. Wrong.

Same with the race concept for the same reasons.

Come on, Bob, you're a scientist-type guy. Get with the program!

And vote for Obama! I know you must be thinking about it...

Can someone explain something to me: People pushing for a lack of biologically real races keep saying there is more difference within groups than bewteen. My question is so what? As far as I can tell, none of the people arguing for biological reality of race are claiming this. Does it say somewhere that there must more more intergroup variation than intragroup for any sort of classification to be useful? For example, I'm sure Ford trucks and Chevy trucks are generally more similar to each other than their respective companies mid-size cars, but that doesn't mean manufacturer isn't useful information when you need to get a new key made. I realize people and cars aren't very much alike. Thsi is just an example of how such a grouping could still be meaningful. What am I missing?

You're missing a key--or any other piece of information for which race is a distinguishing characteristic. A key from a Chevy will never fit a Ford. The variation within each group that people want to designate a race means that there is a huge amount of overlap between groups. Thus, race doesn't tell us anything useful about the individual.

Ace, the thing is this:

The argument about variation you hear is a canard. A wild goose chase. A very smart guy who knows a lot about genetics once made an absolutely correct statement that sounds like "there is more variation within than between yada yada yada..." and this was a meaningful statement and it was not irrelevant. But almost no one who has repeated the statement has a clue what they are talking about and everybody gets it wrong.

The reason that the concept of "race" from a biological perspective is of very limited use is because of the nature of humans, their genes, and the distribution of variants of their genes. For races to be real and valid categories, genetic population need to be more discrete and more differentiated than they are in humans. This fact is neither unusual or surprising.

In the mean time, the social, economic, and human rights consequences of maintaining a biological falsehood are not without cost. Therefore insisting on maintaining the biologically limited concept is itself a racist act. So don't do it. It might make you feel good, but nobody cares how you feel, honestly.