More Recent Comments

Sunday, November 30, 2008

Greg Laden on "Race" (Again)

 
Greg Laden is one of those people who want to abolish use of the word "race" to define genetic subpopulations. He claims that there is no such thing in humans (or in other species). But even if it were biologically correct we shouldn't use the word because is has been misused by non-scientists.

Here's an example of his reasoning from The Scientific, Political, Social, and Pedagogical Context for the claim that "Race does not exist.".

The race concept has been very successful in its many nefarious applications, but this is not what I wish to speak about here. Rather, I want to acknowledge that a concept that divides humans into a particular set of groups in a useful way might be, well, useful and not such a bad thing. The fact that medical researchers use race to divide subjects, and find differences between races, and that these differences are important to know about, is important, even if it does not validate the races. What it means is that an unworkable race concept works sometimes, even if the races themselves don't exist. Nonetheless, it is reasonable to acknowledge that even though races don't really exist and many, if not most, applications of the race concept are obnoxious, it may be that its use is not entirely inappropriate all the time.

I will argue, however, that the down side of the use of race requires its abolition among scientists. Since race is usually not a biologically useful concept for humans (or many other species), and is never a truly valid concept, it is difficult to justify its use given the negative political and social consequences it carries.
Greg and I have been over this ground before. It think it's silly to pretend that races don't exist. That's carrying the anti-racist agenda too far.


31 comments :

Kambiz said...

I also think it is silly to deny that races don't exist. There's this persistent and ignorant mantra among anthropolgists, i.e. anthropolgists who don't understand population genetics and the advances made in this discipline. It is really a shame.

Anonymous said...

Okay, so now I am curious. And at the risk of beating the ground where a dead horse used to be, what are the issues in population genetics which attack Greg's claim?

How are the races separated into distinct, discrete units? I'm only guessing by your last name, Kambiz, so if my assumption in incorrect let me know and accept any necessary apology. I guess that you are Persian. I am German. People tell me that as far as "race" goes you and I are basically members of the same race, even though our skin tones are different. I have been told that Persians and Europeans are closer to each other in race than Persians are to Arabs.

Are Slavs racially distinct from other European whites? Are Firt Nations/Inuit the same or of different race as Asians? What is the racial distinction that separates Hmong and Vietnamese, or are they not discrete races but merely variations?

I am merely curious as to how population genetics disproves the ignorant mantra carried by anthropologists.

Anonymous said...

It is absolutely silly to pretend that races don't exist. But it is not at all silly to recognize that the classic race concept ... the race concept as it is almost always used by pretty much everyone ... is a piss poor way of scientifically describing populations in a number of (but not all) species, including humans.

Given that this race concept is out dated and scientifically bankrupt, I feel that the insistence that it be used is itself kinda funny.

At best, and mostly harmlessly, the term "race" stands as a shorthand for the juxtaposition of actual population variations/groupings and the peculiar sampling we end up with when we combine circumstance with data collection.

The idea that the race concept is outdated and has a number of negative side effects is not a meme, not a mantra, and it is not even a thoughtless PC argument. t is just moving forward.

(I have a feeling we may be going over this a few more times.....)

Ian said...

If races "exist" as something more than a social construct, then there must be some independent way to define them. What data exists that supports the assertion that human variation is better characterised as discontinuous variation and not continuous variation? Sure, you can find differences between African Americans and white Americans, but you'd expect that difference regardless of whether you were looking at a continuous variation or not, since their ancestors came from distant points along the continuum.

It's a simple question - is there data that suggests that human variation can be better explained by a discontinuous model than by a continuous model. Is believe in "race" based on data, or is it a faith-based approach?

Anonymous said...

Greg Laden is a neo-Lysenkoist who, had he be given a power, would happily subjugate science to his socio-political agenda.

He also is not a biologist and does not understand that, given any of the available definitions of *biological* race and the history and geographic spread of H.sapiens, races simply cannot NOT exist!

He also is unaware of the fact that geographically distant human populations have pairwise FST well above 0.25 - a value that "indicates very great genetic differentiation" (Wright) and that has been used by zoologists as a rough criterion for identifying subspecies.

DK

Anonymous said...

LOL!

Greg Laden here:
It is absolutely silly to pretend that races don't exist

Greg Laden in his own blog:
First, the parts we agree with: There is no such thing as race (biologically), race is a social construct used as a political and economic tool, 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
http://scienceblogs.com/gregladen/2008/10/the_science_museum_of_minnesot.php

Make up your mind, would you?

Lippard said...

Races may not be "natural kinds" (species probably aren't, either). But fuzzy boundaries don't make categories useless.

Personalized medicine looks to me like a worthwhile project that will have positive benefits.

Devin said...

Does the HapMap project have anything to say about races? I'm too busy to look through the data myself...

Peter said...

As far as I can see, Greg's point is simply that unless the actual genetic structure of the human population bears some resemblance to racial distinctions as commonly understood, then using the "race" terminology is at best misleading and at worst harmful.

Personally, I can't really see the value in any categorisation that would lump all "Africans" into one category.

The various races, as commonly understood, do of course describe some features of human genetic variation: primarily skin colour, shape of eyes and shape or nose. However, unless that proves to correlate with other more significant phenotypes, isn't it about time to conclude that's all they describe, draw a line and move on to a more useful system?

DiscoveredJoys said...

It seems to me that Greg compares the social concept of "race" with the taxonomic use of "race", and asserts his own values for each.

He has a visceral reaction to the social concept of "race" which may be understandable, but he allows it to colour his feelings about the taxonomic use of "race". He dismisses the taxonomic use of race as of little use and its loss therefore a small price to pay for feeling good about himself. This may be good politics, but it is poor science.

I was going to bang on about demes and clines, and classification forcing fuzzy nature into a digital data structure, but I suspect that the debate isn't about facts, only feelings.

Anonymous said...

DK sez:
"...geographically distant human populations have pairwise FST well above 0.25 - a value that "indicates very great genetic differentiation" (Wright) and that has been used by zoologists as a rough criterion for identifying subspecies. "

Please provide a cite for that. It's contrary to what I have seen in the past, which indicated that humans do not have nearly enough genetic diversity to be considered subspecies. If you have more recent information, I would very much like a reference for it, please.

Larry Moran said...

Peter says,

As far as I can see, Greg's point is simply that unless the actual genetic structure of the human population bears some resemblance to racial distinctions as commonly understood, then using the "race" terminology is at best misleading and at worst harmful.

That's pretty much how I see it as well. Greg is saying that racism is bad and the general public uses the word race in many contexts that are socially unacceptable.

As a result of this abuse by the general public, the scientific term "race" is no longer valid and races don't actually exist.

It's sort of like saying that the general public misuses the word "evolution" therefore evolution doesn't exist.

The logic escapes me.

Anonymous said...

Wow, is this the best we can do? A half dozen remarks suggesting that I'm shaping the scientific argument to suit my politics (an unprovable and absurd assertion) and one single 'fact' that tells us that we see a genetic distance in known-to-vary alleles that is modestly impressive, and thus races exist? Srsly?

The collective argument ... that this is about feelings or politics and thus we can skip the scientific discourse ... is almost worthy of pity.

Come on, folks, you can surely do better than that.

Anonymous said...

Dr. Moran, I think it might be helpful if you explained what you mean by "the scientific term 'race'," or to put it another way, what the scientifically valid meaning of the term is. Are you speaking of a concatenation of appearance-linked traits (skin color, eye, nose and lip configuration), or of something else? If the former, what scientific usefulness would you say it has?

Larry Moran said...

Greg Laden asks,

Wow, is this the best we can do?

What are you expecting? Every reasonable biologist knows that our species can be subdivided into distinct groups that have been genetically isolated for thousands of years.

We also know that the isolation is breaking down in modern times. This is particularly obvious in racially diverse nations like Great Britain, Mexico, Canada, Australia, and Brazil (among others).

We all know that the deepest division is between Africans and everyone else. The entire Out-of-Africa hypothesis is based on those genetic differences. Along comes Greg Laden declaring that these differences between groups such as Africans and Asians don't exist. They are all figments of the imaginations of racist bigots.

Most people are speechless in the face of such blatant nonsense.

Anonymous said...

To Luna_the_cat:
"...geographically distant human populations have pairwise FST well above 0.25" Please provide a cite for that

I was wrong providing the numbers. I was referring to Nature,451(7181):998-1003, Fig.2a - but I now realize that it lists FSTs for genetic distance, NOT allele frequencies. Not the same thing. Damn, I should have paid attention. Looking at a number of papers now, it is clear that the correct number for allele variation is not much more than 0.15 at exremes; the global number is somewhere around 0.10-0.12.

See however, this paper that makes it clear that FST is too simple-minded a measure when it comes to addressing human diversity: Human Biology, 2003, 75(4):449–471.

Anonymous said...

Thank you, DK, I will read that Human biology paper if I can get hold of it; the Nature paper was definitely more on the lines of what I remembered.

General question: What is the problem with using "ethnicity" to refer to haplotype groups, rather than "race"? Especially given that "race" IS a loaded and misused word.

Anonymous said...

Where have I said that differences between people don't exist?

The Africa/Non African distinction is a good place to look. For the most part, the "african" groups are not representative of "Africa" as a continent, so the starting point is invalid. The biggest differences are in range of variation with Africa being the most variable. Everyone else is a subset of African for a very large majority of variant alleles. That does not make Africans more different or an African-Non African divide somehow deeper. As if "African" was a population.

Larry Moran said...

Jud asks,

Dr. Moran, I think it might be helpful if you explained what you mean by "the scientific term 'race'," or to put it another way, what the scientifically valid meaning of the term is.

The effective unit of evolution is the "population," which is defined as a group of individuals that interbreed freely. Different populations will have more or less gene flow between them but they can easily be recognized as distinct populations because they have different, relatively stable, allele frequencies.

It's important to note that a difference in stable allele frequencies does NOT mean that in one population allele A has become fixed while in another it is allele B that has become fixed. This is a common error in these discussions.

Some people think that you can only identify a distinct population if every single individual conforms to some ideal genotype that distinguishes them from every single individual in another population. People who think like that don't understand the basic concepts of population genetics.

In the past, our species was subdivided into hundreds (thousands) of different populations. Some of those genetic barriers are breaking down in modern times but the evidence of the past is still clearly visible. That's why it is possible to tell you where your ancestors came from if you pay for a DNA analysis.

Different populations can be grouped according to their common ancestry. It's pretty clear, for example, that the people of Japan are more closely related to their Chinese cousins that to the Urhobo or Ebu tribes of Nigeria. The largest groupings are called "races" or "subspecies." That's the division that's right below "species" in the categorization of living organisms.

Real life evolution and populations are messy. There are no neat and tidy boundaries, especially at the geographical edges. It's often difficult to decide whether a group is a population, a variety, or a race.

In some cases it's difficult to decide whether a race, or subspecies, is a species or not. One of the classic examples is the red-winged blackbird where some of the races (subspecies) have been classified as species from time to time.

In the case of humans, there are three large groupings that generally qualify as races or subspecies: Asian, Caucasian, and African. Some people split the Asian group into North Asian and South Asian and there are other scientists who think that other groups are equivalent to races.

Are you speaking of a concatenation of appearance-linked traits (skin color, eye, nose and lip configuration), or of something else?

Those particular traits are just the tip of the iceberg when it comes to distinguishing between the three main races. There's lots of genetic data showing that the three main races have been genetically isolated for tens of thousand of years. Any individual whose ancestors come from one of those groups (no hybrids) can be unambiguously identified by their DNA.

If the former, what scientific usefulness would you say it has?

It helps us understand the evolution of our species. We now know, for example, that Asians and Caucasians share a more recent ancestor than either does with Africans. We now know that at many loci there's more genetic diversity among Africans than among Asians. This is further indication that Asians split off from the main African group and not the other way around.

We know that the natives of North and South America came from Asia and not from Europe or Africa. If there was no such thing as human races then we wouldn't be able to tell whether native Americans crossed the Atlantic from Africa or crossed a land bridge from Asia.

We are learning more and more every day about evolution and population genetics by examining our own species. It is by far the best studied of the large diverse species on the planet.

Anonymous said...

Luna_the_cat:
What is the problem with using "ethnicity" to refer to haplotype groups, rather than "race"? Especially given that "race" IS a loaded and misused word.

No problem. It's just that "race" is hierarchically higher than "ethnicity".
"Caucasian" is not the same as "Polish".

Anonymous said...

Dr. Moran -

Thanks for your reply. I think there is potential confusion between using the term "race" as you do (in the sense of subspecies, or groupings of populations) and the common understanding of the word, which ISTM features a near-exclusive reliance on appearance-related features that might misleadingly link, e.g., Africans and Australian aboriginal peoples.

If a term is scientifically precise and correct, or even quite useful (e.g., "species"), then of course by all means it ought to be used even if there is a danger of confusion on the part of laypersons. But is "race" such a term, on a scientific par with "evolution" or "species"? Or is it more akin to terms that are not scientifically precise or useful, that can in fact actively promote misunderstanding, e.g., "Darwinism"? (I pose this as a genuine rather than a rhetorical question.)

Regarding the description of Asians, Caucasians and Africans as "subspecies:" Sources such as the Genographic Project seem to show as many or more distinct populations within Africa as they do distinctions between populations in Africa and those of other geographic regions. (Not sure I expressed that particularly well, but see, e.g., the map at https://www3.nationalgeographic.com/genographic/atlas.html circa 10-5,000 B.C.) What sources would you recommend reading in support of an African/Asian/Caucasian subspecies divide?

Anonymous said...

Sorry, the URL in the previous post should be https://www3.nationalgeographic.com/genographic/atlas.html

Stephanie Zvan said...

Any individual whose ancestors come from one of those groups (no hybrids) can be unambiguously identified by their DNA.

Larry, I fail to see how this is different from the statistical artifacts one gets from sampling far enough apart on any smooth continuum. Where is the data that addresses the middle, and why is it okay to exclude it, particularly without looking at the size of this middle?

If there was no such thing as human races then we wouldn't be able to tell whether native Americans crossed the Atlantic from Africa or crossed a land bridge from Asia.

Again, I don't buy that race is required to explain this. We know that Native American populations came from Asia because there is more continuity with people in Asia than with people in Africa. Again, discrete categories aren't necessary.

I can see that haplotypes are useful in that they provide information about the path of a very tiny part of one's lineage. I still don't see what race buys us as a concept.

Anonymous said...

Dang! That should end "atlas.html"

Anonymous said...

"'race' is hierarchically higher than 'ethnicity'.
'Caucasian' is not the same as 'Polish'."

My argument, then, would center on usefulness. Is "Caucasian" actually a useful category? I would argue that there IS too much diversity inside the group we traditionally think of as "Caucasian" to be medically or genetically useful for the kind of study of ancestry and disease that is being done, whereas saying something like "Northern European" or "Turkic" or "Aryan Indian" would be genetically useful. ~Same for saying "African" or "black", as opposed to something like "North African", "West African", "Caribbean", "New Guinean/Pacific Islander" or even something as finely granular as "Dogon" or "Bantu" or "San" given how genetically grouped those are.

Because so far as I can tell, haplotypes (which are an undeniably useful grouping) do not so much follow "Caucasian" or "black"; haplotypes vary hugely within those large headings. No? And then even for scientists, "race" tends to devolve to the inaccurate popular understanding based around things like general skin tone.

Sigmund said...

There is a danger with statements such as Gregs:
"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" - as it gives the impression that there are NO biological underpinnings to the concept. The trouble is that 'race' as a shorthand for the continental origins of an individuals ancestors actually does have empirical backing when population genetics are taken into account. Of course there are alternative terms such as clines, demes or BGA (Bio Geographiical Ancestry) that are probably more useful scientifically than race - particularly when one considers the mistaken ideas of race held by many in the general public. However when even crude self designations of race can be clustered to an accuracy of over 99% using genetic testing it is politics rather than science to still maintain a 'lack of biological validity'.
And Greg, its not racist to say there is a statistical difference in allele frequencies between populations. What IS racist is to suggest that these differences inevitably make one of these populations inferior to the other.

Anonymous said...

Luna_the_cat said:
My argument, then, would center on usefulness. Is "Caucasian" actually a useful category?

Yes. As you know, biological races are also evolutionary lineages. We can tell that a common ancestor of Poles and Persians was Caucasian and not, say, Australoid.

Obviously, the usefulness of such broad categories varies depending on a question. It's useful as a descriptor of human history. Or, it's useful when you are doing a clinical trial with ~ 2,000 enrolled (already a major undertaking!) - where 106 bins will normally take you nowhere as far as statistics is concerned.

Anonymous said...

Given race is a biologically valid concept to some extent and given genetic isolation: do different races tend to have genetic predispositions to differing levels of intelligence?

Is there in any way, biologically, a relationship (however weak or strong) between racial groups and intelligence.

=P

Anonymous said...

I'm late to this conversation, but has anyone pointed out that Greg Laden's views on the classic race concept are compatible with those held by the great majority of anthropologists, at least in the US? i.e., there's no meaningful cluster of biologically-based traits at the so-called race level; there's more biologically-based variation within populations than between them; there were never pure races... The American Anthro Association launched a major initiative on the data to back up these claims, in language meant not for scientists but for the general public, at www.aaanet.org

Larry Moran said...

Barbara King asks,

I'm late to this conversation, but has anyone pointed out that Greg Laden's views on the classic race concept are compatible with those held by the great majority of anthropologists, at least in the US?

No, I didn't point it out on this particular thread.

It's a pretty sad state of affairs, isn't it? We have a situation where the majority of anthropologists don't understand basic biology and evolution.

I certainly hope that anthropologists in other countries don't think like that. I wonder what it would mean if it turned out that it was only *American* anthropologists who denied the existence of human subgroups? Do you think it might have something to do with letting politics and culture interfere with science?

That would be ironic, wouldn't it? The idea that anthropologists can't distinguish between science and culture is ... interesting.

Sigmund said...

Barbara, that aaanet.org site you mentioned is pretty shocking if you know anything about human genetics.
They make it seem like no new genetic variation has arisen in the past 70,000 years or so since the diaspora out of Africa - that all the variation you see in humans is already present within the African population. That is ridiculous. Have they never heard of the lactose tolerance mutation for a start.
I can perfectly well understand pitching an exhibition to a non scientific audience will require a different type of language style but it shouldn't ignore the central principle of modern biology (the fact that evolution is occuring within populations all the time).