My Review of Nicholas Wade's Book, A Troublesome Inheritance: Genes, Race, and Human History.

Hmmm: Lumpers, splitters and the dilemma of continua. Seems to me that questions of race fall well outside the areas of expertise of economists, psychologists, and perhaps even the odd emeritus biologist.

Reminds me of "The Mismeasure of Man". I could probably make a pretty good guess at which camp is actually indulging in motivated reasoning.

It should be pretty obvious by now that race is a messed up concept.

I agree that gene differences probably explain little in "race" behavior differences (if such exist), and you did great there.
One quibble:
That race is not predictive of alleles seemed overstated - no qualifiers like mostly or even overwhelmingly. "it is not scientifically possible to classify people into a “race” ... to reliably predict other, less visible, genetic features. " Not that it predicts much, but I can do better than coin flipping about guessing alleles at a few genes by just looking at folks (AGT and CYP2D6 are famous). That is to say, looking helped doctors. Asking about your ancestry is even better than looking. Measuring the damn alleles is far better than any of that. Note that we still check if we have race by treatment interactions in the arms of clinical trials, where the difference between drug A and drug B differs between "races". I think that should continue to be looked for. (Yes, it could even be a culture correlate that is found, rather than a genetically caused difference.) We suspect genetic differences may explain differences in esophageal cancer incidence.
I thought rule of thumb was that about 90% of gene differences between people, race is not predictive. So you were right mostly, but maybe only mostly. I don't think your argument is much harmed by that.

Rork, people who come from the same, similar, or overlapping/interacting gene pools will be more likely to have similar alleles. That's it. That is sometimes mistaken for race, and our categories of "races" sometimes stand in for that geographical effect, depending on the relevant population history. But that is a very different thing than distinct (or even indistinct) races compartmentalizing alleles.

Greg, per #3: Is there a picture to show that someplace? Something like a couple or three diagrams perhaps? ("Here's how alleles would sort out in biological races... Here's how alleles _actually_ sort out in humans".It seems to be point that might be communicated visually in a way that is more immediate. Wish I could draw those pictures myself, but I am hoping their existence somewhere will help my comprehension.

Ugh. Pardon my poor proofreading.

This issue disturbs me greatly. There are really good reasons to divide up people, but race is not a good one except in really specific medical contexts. Harping on something like race is to cling to things recent and shallow in terms of impact on human difference.

The far more relevant differences involve society and culture, our personal short-term heritage in response to it, and a personal mental/emotional shape with respect to the role one has in society.

Culture itself is well addressed in the article. Unexplained IQ changes are challenging and personally I have no problem with the variation in IQ and similar in terms of the spread of information that allows better mastery of what IQ tests cover (which tends to be culturally biased). The examples of hunter-gatherer cultures that are less violent, or inconsistently violent are good examples that undermine Wade's views.

The real differences in cultures and what are thought of as "races" likely involves a combination of the history of the previous couple of centuries of experience of the people, combined with what these cultures tend to value in comparison to the things that IQ tests tend to measure. Some of these valued things in IQ tests will be there because of historical accident and won't say much about general humanity without a cultural and historical comparison. Other differences will involve cultural experiences and the science of epigenetics as it relates to emotional loading of information into a persons implicit reactions as their life proceeds, or even before it began in possible transgenerational epignentic influences.

The path that people are able to take in the cultures that they have to choose from is the final piece of significant difference. What your life prepared you for will determine your skills and aptitudes, and the world you live in will not always mesh with those well.

I think that the real differences between people are represented by major mental illnesses, but in a more stereotyped form. ADHD, autism, TS, schizophrenia, depression, anxiety, and similar are likely extreme ends of general forms of human behavior given the demographics. People are likely to be far better defined by where they fall on major cognitive system differences reveled by conditions that push individuals down particular paths.

Matt: It is basic genetics so, frankly, it is hard to find good on line internet sources because it isn't new and shiny. There are some great pictures that come from the history and geography of human genes project:

http://www.amazon.com/The-History-Geography-Human-Genes/dp/0691087504

There, allele after allele is examined and mapped, and each map for each allele shows a different story and clearly shows the close link between geography and allele variation AND the independence of this relationship across alleles.

That book was published before everything went on the internet but there are graphics out there. This paper may be helpful:
http://gutengroup.mcb.arizona.edu/Publications/Ramachandran2010.pdf

That particular book was

"This issue disturbs me greatly."

Me too. As far as I can see, "race" doesn't describe or clarify anything useful that can't be talked about more appropriately in specific (perhaps medical) terms. So why put so much effort into defining and defending an unproductive term, especially one so larded with hateful overtones, except to maintain an outdated world view that happily labels 'the other'?

Thank you Greg. The Cavalli-Svorza is now in my Amazon cart (a used copy - the new one is pricey!). I'll read the Ramachandran first, though, for practical reasons.
By the way, the reviews on the Cavalli-Sforza are peppered with submissions by folks who could benefit from reading your review of Wade.

@4: The argument sounds too close to race cannot predict alleles cause race doesn't exist or is flawed as conceived in the old days. It's true folks tend to say we want to know your ethnicity (location of your ancestors mostly) rather than your race in medical genetics. When looking at a person trying to figure out their genes, maybe race is the wrong word for the classification one is trying to achieve. If the point is that the old use of the term race is goofy, I agree. I've worked with folks from far west (Ivory Coast) and east Africa (Madagascar) who may be less related to each other than I am (Germany) to one of them. Lumping the three of us into races does seem like horribly oversimplifying and being capricious in how that's done as well.
I'd rather talk about biology.

Very nice review Greg

It has been said (and I've been trying to find out by who, so if you are reading this tell me!) that if we wanted to redo the race concept with our currently available information (instead of the crappy information that led to "Caucasian, African, Asian, etc) and we figured we'd want about 12 races (more is unwieldy, fewer is unrealistic) we could do so but nine of them would have to be African.

I'm actually a bit disappointed that they deleted that long comment full of irrational conclusions by the guy defending Wade's work. I've been practicing responding to those arguments and by the time I finished, it was gone. Oh well.

Joshua, I didn't see that comment! Oh well.

It was some person named JayMan.

It was full of appeals to authorities with badly reasoned opinions and characterizations, references to articles that did not actually support either their point, or contradict points you were making, selective use of only some relevant scientific concepts (with the exclusion of others that did not support their case), conflations of systems that will have have very different selection as if they are the same, no awareness for the idea of neutral mutations accumulating in populations but having no fitness effects, conflation of genetics and heredity, and much much more.

It's fascinating the way these people use the same forms of creationism while pretending to be scientific.

I am a complete outsider to this topic, so please stay friendly, I am just naive.

Greg, for me as outsider you convincingly showed that races do not exist. The transitions are smooth and any classification (into races) would thus be arbitrary. (The EOFs you showed in the above picture look like the ones you get for any spatially correlated random field.)

However, you do acknowledge that there are differences, but that they depend on geographical distance and are separate for every allele. If we compare this to the colors of the rainbow, one could say that the names for the colors are arbitrary and a cultural construct, but the differences in frequency of the light are real.

At the end of the review, you seem to argue that all the differences we find are cultural. Thus there you seem to argue that there are no differences, rather than that there are no races. Isn't that a too big leap?

Victor,

The groupings, not the differences, are cultural.

On color, different cultures lable colors differently, and labels are not assigned to single frequencies but to various ranges of frequencies.

As it happens, we now have very strong, practical working models of color, but there are multiple models for various purposes and none of them are absolutely perfect. FWIW, they are based on physics, the biology of the eye, psychology, and the properties and processes of pigments and light generation.

There is nothing strong or practical about race. It's like arguing abou how many angels can fit on the head of a pin.

Victor, your question in some ways complicates any possible answer but in interesting ways. For example, when we look at a rainbow we are seeing a continuum of color frequencies, but what we think we see is a discrete number of bands of distinct colors. So right there our brain is creating a category (because of the way our color sensors and neural systems work) that does not exist in the physical universe we are looking at. So in some way sit is even a better analogy to race than one might think. When we stand on a street corner in an American City we may see people we classify as "hispanic" and "black" and "middle eastern" and "east asian" and "white" but we are not only glossing over overlap and gradation that is right in front of us, but we are also seeing people who have been sampled from the real world by historical circumstances in a way that makes them categorically distinct.

Also, I'm not sure if you got this point: If every allele is acting independently (and most are) then when we see Version 6 of Gene 222, we can predict the following: The person we are looking at has Version 6 of Gene 222. There is a good chance that we will find this allele among this person's family within about second cousin. After that the probability drops depending on how widespread the allele is.If it is a very common allele, then everyone in the world may have it.If it a very rare allele the neighbor down the street may have it. But, all of this is about Version 6 of Gene2 22.

None of this lets us predict that Version 9 of Gene 8977. Nothing.

So when we see a rainbow and note that one band is blue,we can say for sure that one band of the rainbow is blue. That's all.

The last part of your question is important. There are THREE distinct things being talked about here.

1) Are their genetic sets with internal consistence (predictability) and boundaries? No.

2) (Related) Can we predict a person's allelic composition based on seeing a small subset of alleles (skin color, hair form, etc.)? No.

3) IS THE VARIATION IN social behaviors and capacities, and things like IQ explained BY VARIATION IN the alleles of some underlying genes.

That last point is important. You are asking if there are still differences even if races are not a good concept. The answer is that it depends on the trait. Genetic physical differences vary based on underlying genes for some but not all physical things. Skin color absent tanning yes, height no for the most part. What language we speak is highly heritable .... offspring resemble their parents ... but VARIATION IN WHAT LANGUAGE YOU SPEAK is not accounted for by VARIATION IN GENES. In between the most obviously learned aspects of hour humanity and the most obviously not learned aspect (language vs. skin color) is everything else. And, we have very good reason to believe that stuff like social capacities are learned, enculturated, not genetic. Those who want to say it is genetic have a very large burden of proof to address and have not done so.

The use of ALL CAPS above is not shouting. It is pointing out that questions about human VARIATION that don't ask about VARIATION need to be rephrased. "Is skin color genetic"? only trivially so, bad question. "Is variation in skin color accounted for by variation among genes?" Yes.

Culture affects what we think we see.

A cross-cultural study reveals how language shapes color perception
http://www.apa.org/monitor/feb05/hues.aspx

The history of color modeling (and color perception) stands as an instructive example of how modeling and categorizations work and don't work, and how people adapt to them. I still run into people who insist that the subtractive primaries are red, yellow, and blue. That's really, really old school (today we use CMYK). If people have a hard time letting go of old notions of color, imagine the intransigence that can arise with race!

Obstreperous Apfelmus, I agree with you. Especially about colors being cultural names for frequency bands or mixtures. Being Dutch and living in Germany, I notice that while the main colors are clear, red, blue, green, were you draw the line seems to be different for different cultures. Especially for mixtures such as brown, purple or Bordeaux red.

Greg, Thanks, that makes it clearer to me. Given that our strength is learning, which allows for rapid cultural "evolution", I agree with you that it seems to make sense that social behavior is learned and that claims to the contrary would need strong proof.

For more direct and physical traits, I am wondering if the situation is not a bit more complicated. It seems as if many traits are spatially correlated, due to local selective pressure or because it takes time to a gene to spread in a population. Skin color or lactose tolerance as adult, would come to mind.

These two traits are probably independent, but white skin is still correlated with lactose tolerance. The processes are independent, if we would "run evolution again", the same correlation is likely missing and sometimes reversed, but in the realization we have now there is a correlation. For the specific mutation that gave many Europeans lactose tolerance, skin color or the cultural category race would be predictive. Independence of two processes does not equal uncorrelated realizations.

"skin color or the cultural category race"

If so, one descriptor or the other, the specific one is best and there's no need to bring up race. It's a surpluflouous reference.

My theory is that the fixation on skin color is because it is visually striking. If we were taking about the texture of finger nails the matter would be moot.

:) agree. I guess we are a very visual species. It also amazes me that Barack Obama is seen as an African American, even on Wikipedia, just checked, while he is only half from African decent, but his skin is dark.

White skinned people have a range of lactose tolerance, and any given sample of white people in a european country will yield results from very high in Northern Europe and lower near the Middle East and Med. Ironically, people with the longest history of cattle keeping in West Asia and Europe have the lowest levels of tolerance.

While it is commonly said that the people of the "Black race" (Africans) tend to have the lowest adult lactose tolerance levels, the populations that have about 100% tolerance are pretty much only African populations, certain groups in Central Africa.

Among the people of the world who rely the absolutely MOST on cow milk for their food are African groups with near zero adult lactose tolerance.

Tolerance is distributed in a very whacky way in reference to the racial groups, and the association between cattle keeping and this particular allele variant is bad enough to seriously question the theory that they are linked, or at least, that they are usually linked. (I.e. the Central African 100% tolerant people are major cattle keepers).

So, like stature, this example is not a good example to support any sort of standard race concept.

Victor (24) This is called Hypodescent. Westerners (and maybe others) classify a person who is descendant from the dominant "race" and a lower "race" (by the standards of the time, so white = dominant or higher and black = sub dominant or lower) into the lower race category. Our classification of Barack Obama as "black" is an example of this, and an example of an insidious part of the whole race concept. More reason to put it aside.

There is another twist to that; It is simply true that whiteness is very white with little variation and anyone who isn't all that white gets to be black or whatever. But we do tend to not do this with white-asian people ... they are "mixed" and if they are white enough they are white. But in the US we have a long history of trying to manage the breeding and other aspects of society of black people/Africans (as well as women in general) because we tend to be assholes about this sort of thing.

"Rork, people who come from the same, similar, or overlapping/interacting gene pools will be more likely to have similar alleles. That’s it. That is sometimes mistaken for race, and our categories of “races” sometimes stand in for that geographical effect, depending on the relevant population history. But that is a very different thing than distinct (or even indistinct) races compartmentalizing alleles.
"

OK.........none of that contradicts the existence of population differences on traits. But you seem to think it does.

No, greg, calling southern and and northern Europeans different "populations" instead of different "races" will not take away the height gap that exists between them.

"Are you claiming that northern and Southern Europeans are different races distinguished by height?"

That's the silliest thing iv'e ever heard. Did you even read Wade's book?

I think that a big part of the disconnect there is that the whole point to most of the objections I have read to Wade's book, is that there are no meaningful correlations between what we call race, and what you can find in terms of patterns in genetics. Race exists as a concept, but it gets used in so many ways from Darwin to today and the usage includes a lot of things that just can't be demonstrated the way that many like Wade seem to want it to.

Yes there are changes in genes between groups. The more the groups are separated in time the more changes will accumulate. There are thousands of differences between each of us and our parents. But not all genetic changes produce effects (neutral mutations), not all expressed genetic changes produce visible effects in complex systems with many parts (dilution effects for both positive and negative mutations), and not every system is as tolerant to changes (cognitive systems versus skin color) and less tolerant systems will produce fewer acceptable mutations.

Many times that I see people pointing to genetic data showing differences used in tracking things like migration patterns, the specific genetic differences they are pointing to have no phenotypes associated with them. The collection of data that do show phenotypic differences show inheritance patterns, and only a part of inheritance is genetic. I never see anyone like wade describe the magnitude of nurture effects on inheritance.
Inheritance is not equal to genetics. We also inherit predispositions (gene expression states, epigenetics) for responding to environments in different ways based on the experiences of our mother during conception.

On top of that there is stronger and stronger evidence accumulating that the experiences of parents and grandparents also change these expression predispositions. The recording of transgenerational "nurture" information is mixed up in a lot of inheritance data and must be separated from genetic effects if any genetic argument is to be made. Not to mention that the hardware that runs the nurture programs will also involve elements of genetics making them a new variable since they would technically be both nature and nurture.
"A review on the evidence of transgenerational transmission of posttraumatic stress disorder vulnerability."
http://www.ncbi.nlm.nih.gov/pubmed/24402183
http://en.wikipedia.org/wiki/Transgenerational_epigenetics
Science and medicine usually start with the most traumatic stuff, and end up discovering systems involved in positive and negative aspects of the human condition more broadly. At this point we have no reason to dismiss the idea that whole groups of people that we see as minorities get all sorts of emotional inheritance patterns that produce effects like lower IQ (not to mention height and color may have transgenerational epigenetic effects as well. Wade and his sort are cherry-pickers with a whole set of unknown unknowns they don't think about.

A huge part of what falls under the word Race involves claims of not just things like height and color, but claims of specialness for some races over others, or inferiority of some races relative to others. People can go on about skin color and height, but those are not meaningful differences when it comes to this strange social need to sort people into boxes. The conversation always get over to IQ, a thing with so many cultural effects that you cannot separate nurture from it and neither Wade or anyone else have outlined anything remotely convincing when it comes to genetics, race, and IQ.

"strange social need to sort people into boxes"

Call it tribalism, institutionalized OCD, or whatever. It pretty much comes down to cattle management of domesticated humans. As some would have it, we're all just material serving at the pleasure of our "betters".

One thing being overlooked in this discussion is that whether aggregations of phenotypically similar societies rise to the level of being different races or whether they are merely vestiges of some primitive reluctance to trust anyone that looks different does not deny the fact that such aggregations do exist and are the source of potentially self-annihilating behavior (if anyone doubts that they exist please see this data portal: http://dna.xyvy.info ). Therefore we must explore the underlying reasons for these aggregations in an attempt to better understand them and decouple ourselves from a blind adherence to behaviors that may have once been successful adaptations but are no longer so.

Greg, This might be the reference you are looking for.

http://www.ncbi.nlm.nih.gov/pubmed/20178763