Is there such a biological thing as a racial group?

So in this pit thread, a position came up for debate that I’ve seen a lot on the dope. Diogenese, as usual, put it in the most blunt and absolute terms:

My question is… what precisely does he mean by that? What do other dopers who make similar but less absolutely stated claims mean, and are those claims true?
My thoughts on the topic:
-It’s certainly true (and I don’t think anyone would argue) that it is 100% not possible to absolutely and meaningfully divide every single human being on earth into one of a fixed number of racial groups. That’s one (of many) reasons why white separatists who propose a “no black people can live here” law are being idiots.
-It’s also certainly true that things are a lot more complicated than just “oh, all those dark skinned people who look kind of African to me are ‘black’, I know them when I see them, and they all share common genes”
-That said, I do think the general consensus view is a bit facile…

So let’s say we want to discuss 4 racial groups in the US: White, Black, Asian, and Latino.

So, we take 100000 random guessers and 100000 random targets. And we show each of the guessers a picture of each of the targets and say “to the best of your ability, sort this person into one of those 4 categories, or (E) no idea, or (F) other.” (Presumably we give them some incentive to do their honest best job).

I claim the following characteristics will be true of this experiment:
(1) there will be a large number of people among the targets who are identified as the same one of ABCD by a huge statistical majority of the guessers (although some will be mostly E, some will be mostly F, and some will be just about every cominbation of the above you can imagine). We will call people who are generally guessed as being in the same group like this “consensusers”.
(2) If a “consensuser” is closely related by blood to other targets, it is more likely than chance (although not certain) that those other targets will also be consensusers, and of the same type
(3) There will be a strong (although again, not perfect) correlation between someone being a consensuser, and that person self-identifying as a particular racial group (obviously there is some trickiness here as “Asian” has subgroups, etc).
(4) If you take other measurable biological traits (ie, muscle type distribution, height, or susceptibility to sickle cell anemia) and test all the targets for them, and then graph those vs. consensusers, you will notice at least some correlations.
So, (a) does anyone question my claims, and (b) does anyone have any comment on their bearing to the “there is no such thing biologically as race” question?

(To avoid doing the kinda douchey thing where someone posts a controversial OP and refuses to actually take a stand on the issue, my opinion is that if racial groupings were truly “biologically meaningless”, then (4) would certainly not be true. If you can predict anything biological/physiological based on the race, even with a very low degree of confidence, as long as it’s still statistically significant, then it’s disingenuous to say that there is no biological meaning at all. But that’s really ALL I’m saying, please don’t think this is part 1 and part 2 is going to demonstrate that blacks are inferior, or anything of that sort.)

Actually, beyond the most superficial (and often inaccurate) “observations,” you won’t.

If you limit your groups to people who have immigrated to, (or been imported to), the U.S., you can somewhat get away with that sort of claim, because the actual process of deciding who could enter, (or be brought to), North America puts limits on the genetic pool from which those people have come. So, for instance, we have a lot of claims about “black” people that actually only refer to people from a limited range of African nations from about Senegal to about Congo. People from Africa whose ancestors lived in what is now Kenya, Ethiopia, South Africa, and other regions outside that range do not show up in lists of very good sprinters, they tend to lack any susceptibility to Sickle Cell Anemia, (while a number of white people are susceptible). Similarly, many claims have been made about Asians, or Chinese, or even “Mongoloids” that are based on the fact that the overwhelming majority of Chinese immigrants to the U.S. have been Han and characteristics of non-Han populations are ignored when discussing what “Asians” are like.
When all the populations of the various purported “races” are observed, the differences far outweigh the similarities while similarities between separate groups in other “races” become more noticeable.

People who need to divide the world into three to six “races” want to ignore those discrepancies while making grand claims about what the “races” are like (even while disagreeing among themselves as to which and how many races exist).
People who prefer to use the more scientific analyses regarding the several hundred separate genetic populations in the world avoid those gross and distorted generalizations.

I’m not making any specific claims about how many races there are, or that in fact there is any such number. But let’s assume for a second that you are exactly correct and people from Congo have a higher susceptibility to sickle cell, while people from Ethiopia do not.

So I encounter some emergency situation where I have only a few seconds to pick some guy to send in to make emergency repairs in a situation where a susceptibility to sickle cell anemia would put the repairman at great risk (obviously an extremely contrived hypothetical). I have two repairmen at hand with the knowhow to do the job. One is “black” and one is “white”. Now, I certainly have no way of knowing off the top of my head whether the “black” one has Ethiopian or Congoese ancestry, and given the history of the slave trade and so forth it’s very possible that he doesn’t either. But unless there are other factors at play (ie, equivalently large “white” ethnic groups with sickle cell susceptibility) it’s still more likely that the black guy is susceptible than the white guy, even if there’s extra information that I could have, but don’t, that could let me make a better guess. So I’ve made a greater-than-chance guess about something biological based purely on perceived race. Again, I’m not trying to say that proves something deep and meaningful, but I think it’s a statement with SOME meaning.

Wouldn’t it just be easier to identify 20 or 30 specific genetic markers and label people by whether or not they have these markers?

Yes, there is. See discussion here.

http://boards.straightdope.com/sdmb/showthread.php?t=601006&page=5

There is no scenario that is not extremely contrived that gives any meaning to the way that some people try to assert that there are races.

In the U.S., we can make some moderate guesses based on the origins of immigration waves and for that, we can get away with using the culturally defined “races,” based on sloppy phenotyping, but if we are studying humanity as a whole, then the three to six “races” claim is worthless as a biological claim. If we are actually studying biology or genetics, we can use the genetic markers that gets us down to much smaller, (and far more accurate), biological populations. Claiming that there are biological “races” in humanity fails to provide useful information and distracts from actually learning anything.

Paradoxically, I kind of agree with that, but also object to someone claiming that race has NO meaning. It would be one thing to say it has limited meaning or something like that, but to make an absolute statement, particularly without a definition of what you’re talking about, seems overbroad to me…

For instance, let’s say someone says “Asian-Americans have an average IQ 10 points higher than white Americans, and it’s due to genetics”. There are plenty of interesting responses to that involving nature vs. nurture, relevance and accuracy and meaningfulness of IQ tests, etc. But I think a lot of people want to respond by saying “but there are no races, therefore your claim is wrong, QED” or “Asian-American isn’t a race, therefore your claim is wrong, QED”, and it’s not that simple. Even though Asian-Americans are a vaguely defined grouping of a wide variety of ethnicities with various amounts of intermixing, if IQ were relevant, and were genetically linked to particular ethnic groups, it would be quite possible for Asian Americans, despite being an ill-defined conglomeration, to have a high average genetic IQ… because individual of those ethnic groups (in the hypothetical where IQ is in fact genetically linked to ethnic groups) having high IQs would bring up the average for the whole group, making the statement rue.

http://www.pbs.org/race/001_WhatIsRace/001_00-home.htm

And unfortunately the junk science also includes the assumption that race is the most important element of intelligence, most of the evidence shows that social status and culture are more important.

  1. Who says that race “is the most important element of intelligence”? People say that groups have different statistical distributions. Quite a different claim.

  2. Variation between individuals in terms of intelligence is significantly due to variation in genes. As Steven Pinker notes:

http://www.nytimes.com/2009/01/11/magazine/11Genome-t.html?_r=1&pagewanted=2

Some recent research:

Also, Paul Thompson, UCLA neuroscientist:

http://www.technologyreview.com/biomedicine/22333/page2/

Population genetics finds clusters that correspond to those groups. No matter what genetic markers you choose: SNPs, STRs, no matter how you choose them: randomly or based on their “informativeness”, it is relatively easy to classify DNA into the correct continental origin.

Medically they have significance:

http://jcp.sagepub.com/content/44/10/1083.short

Let’s try with some other samples and see if you still agree with the conclusion.

So let’s say we want to discuss 4 national groups: Japanese, Liberian, Russian, and Papua New Guinean.

So, we take 100000 random guessers and 100000 random targets. And we show each of the guessers a picture of each of the targets and say “to the best of your ability, sort this person into one of those 4 categories, or (E) no idea, or (F) other.” (Presumably we give them some incentive to do their honest best job).

I claim the following characteristics will be true of this experiment:
(1) there will be a large number of people among the targets who are identified as the same one of ABCD by a huge statistical majority of the guessers (although some will be mostly E, some will be mostly F, and some will be just about every cominbation of the above you can imagine). We will call people who are generally guessed as being in the same group like this “consensusers”.
(2) If a “consensuser” is closely related by blood to other targets, it is more likely than chance (although not certain) that those other targets will also be consensusers, and of the same type
(3) There will be a strong (although again, not perfect) correlation between someone being a consensuser, and that person self-identifying as a particular nationality (obviously there is some trickiness here as “Russian” has subgroups, etc).
(4) If you take other measurable biological traits (ie, muscle type distribution, height, or susceptibility to sickle cell anemia) and test all the targets for them, and then graph those vs. consensusers, you will notice at least some correlations.
I assume you agree with all that.

So do you now accept that nationality has a biological basis? After all it meets exactly the very standards that you set for race.

So are you seriously arguing that someone can be biologically “French”? Or are you beginning to see why this whole experiment is comepletely invalid as a method of determining biological bases?

Using fine scale analysis you can distinguish between countries in some cases. Check out this paper on European genetic substructure, discussed by Steve Hsu:

More from the interview:

http://www.pbs.org/race/000_About/002_04-background-01-07.htm

Let me get this straight. You are seriously arguing that somebody can be biologically french?

As long as you will accept probabilities instead of certainties, I don’t see anything ridiculous about that at all?

How about the two most obvious:

  1. The boundaries of France are not real. Never mind whether they are measurable, they aren’t even real. What we call France today did not even exist 2000 years ago.

  2. The “French genetics”, whatever that is, will move with migration, so Canadians of French descent will not be distinguishable from people born in France, while a Greek who moves t France will presumably not share any French genetics.

Can you answer some basic questions:

If someone lived in Alsace in 1932, were they biologically French? How about in 1945? How about 1845? Why does the state of being French change over time, even for the same individual, if being French is biological?

If a couple moved from Paris to Sydney in 1800, were there children French? How about their grandchildren. Because if French is a biological construct then it can’t change due to mere geography, right? A deer does not become a kangaroo if it is born in Australia. So why does the offspring of two biological Frenchman become not French when moved abroad?

Did the genes somehow shift to suit the whim of the cartographers? Did people give birth to children with different genotypes when the boundaries got redrawn?

Or is it more reasonable to assume that people always gave birth to children who shared 50% of their own genes? And hence “French” is a purely political construct, and not a biologically one?

If your point is that “Hispanic” is precisely as biologically founded as “French” then I agree 100%. If you are going to argue that “French” is determined by biology then we are going to have a lot to discuss.

Both obviously true. Clearly you can not currently test someone’s DNA and determine 100% whether that person was born inside the political boundaries of modern day France. But you CAN test someone’s DNA and determine there’s a fair likelihood that some percentage of their ancestry likely descends from people living in a particular region of Europe several hundred years back. And France is a pretty big region of Europe.

This doesn’t seem controversial or surprising to me at all as long as you throw in some disclaimers about lack of certainty, etc.

And if you go back far enough everyone’s ancestors came from the African continent. So what?

Sure, add some disclaimers, I’d suggest something like: None of this is useful.

And that’s it in a nutshell. You can draw any number of arbitrary boundaries and classify the people in those boundaries as a “population”.

When you have a species that varies clinally, like humans, you can’t draw objective boundaries. I can plop you down in Beijing or Berlin, and you’ll be able to tell me where you are by looking at the people around you. But, if you walk from Beijing to Berlin, there is no line you cross where suddenly people stop looking “Asian” and start looking “European”. And this is the same with our genes. We just don’t keep ourselves in isolated genetic populations.

So, if we insist on their being races, we’ll never be able to agree how many there are. Are there 6 or 600?

Excellent post, John Mace. Our use of race takes a spectrum and tries to break it into discrete parts. It does this in a clumsy way that lumps people who don’t have anything to do with each other together, and others who are closely related get separated.

You will find that the “races” are totally different in different countries. Give people a random stack of photos and have different people stack them into piles, and every culture is going to have a very different pile. I’ve spent a lot of time trying to convince doubtful Africans that in America, Beyonce is black. At times in apartheid South Africa, Chinese people were “Asian” but Japanese were “white.” So if race is meaningful, who gets “race” right?

Finally, even if there are generalizations that can be made, what exactly are you going to do with them? None of us have ever been in a situation where we need to determine someone’s likelihood of sickle-cell anemia just by looking at them. Even if black people have lower IQs, what exactly can you do with this information? Not invite them to interviews? Automatically track them into remedial classes? Obviously this would be absurd. Individual variation is too high for “race” as a biological category (as opposed to a social one) to be useful in planning just about anything.

So if an arbitrary, changeable grouping is not really useful, what is the point of even having it?