The 10,000 Year Explosion

The subtitle is How Civilization Accelerated Human Evolution and the authors are Gregory Cochran and Henry Harpending.  I do think that such topics should receive open debate but, as with Greg Clark's book, I'm not convinced.  There is plenty on dog breeding, lactose intolerance, Genghis Khan and his children, the difficulties of settling the Andean Highlands, and just-so stories about medieval Ashkenazi Jews.  What's missing is a sense of what the hypothesis does not explain, what its limitations are, and also what exactly is being claimed beyond the particular cited examples.  The stories of "lots of recent change overall" and "current groups differ" are jammed together but of course they are very different.  Epigenetics don't receive much attention, even critically, and the lower levels of Ashkenazi social achievement before 1800 are dismissed quickly.  It's fine and indeed correct to claim they were oppressed but that opens up many doors to explain many other observed correlations.  The authors report that we have Neanderthal genes even though this seems to fly in the face of recent discoveries and more importantly the evidence that such interbreeding (if it occurred) mattered is extremely speculative.  Perhaps the authors are right but the reader is not given the tools to see why their understanding is a superior one.

Razib liked this book (see the first Amazon review) and I suppose it is a good introduction to this point of view, but overall I didn't come away feeling I obtained a superior understanding of the issues.


The mitochondrion is not the only place to seek Neanderthal ancestry. AR Templeton's paper in Science 1996 May 31;272(5266):1363a is interesting reading, suggests interbreeding is a likely explanation. But both for this question and the for rate of human evolution, detailed knowledge of the human past is necessary for certain answers, and we do not have this knowledge.

I havent read the book, but there's a subtext behind why these guys (and razib, Steve Sailer and gnxp folks) want desperately to believe in recent human evolution. They believe in large genetic differences between the races, and all used to believe in multi-regional evolution (that races evolved independently over hundreds of thousands if not millions of years). The latter view has been almost completely demolished by genetics - we now know almost all our genes come from relatively recent emigrants from Africa. So they are left with 2 hypotheses:

1. Recent human evolution caused large genetic differences among the races.
2. The Out-of-africa data is incorrect because it assumes our lack of genetic diversity is because of recent common ancestry, but really selection is widespread and all the genes we have looked at have actually been selected for and hence have low diversity.

This explains both why this gang is over-reacting to slim evidence, and also why their ideas are out of the mainstream. They may still be right of course, but usually (although not always) the scientific consensus is the right one.

What TGGP said. Here is the Hawks-Harpending paper he mentioned. I won't likely read this book, but from Tyler's description it seems full of speculation on why evolution has accelerated in the last 10,000 years. I don't know if they are on to anything, but how do get around the obvious explanation? That is, evolution accelerated because civilization allowed for much greater population sizes, and population size itself is strongly correlated to evolution.

Come on, Tyler, you can do a better job of actually explaining what the book is about than that! You're not as obtuse as you sound in this posting.

We all know that different environments select for different traits. And what bigger change in the human environment than the development of agriculture over the last 10,000 years? The simple rise in population due to agriculture increases the number of favorable mutations, which leads to their spread.

Moreover, this is a really fun, witty book.

Harpending and Cochran have a website at

The "Deleted Scenes" of sections that didn't make the final draft shows the quality of the book, since even this material that got left on the cutting room floor is so fascinating.

"Given that nearly all children live to childrearing age in advanced economies, it seems clear that evolution has either stopped or is going backwards for humans."

Okay, but a huge fraction of children don't live in advanced economies. Plus, there are sizable differences in how many children these children's parents have. All this is a form of Darwinian selection.

but there's a subtext behind why these guys (and razib, Steve Sailer and gnxp folks) want desperately to believe in recent human evolution. They believe in large genetic differences between the races, and all used to believe in multi-regional evolution (that races evolved independently over hundreds of thousands if not millions of years). The latter view has been almost completely demolished by genetics

Nice smear. You impute motives to some set of people without evidence ('desperately want'? really?). You claim that they believe in 'large genetic differences'; a more neutral assessment would be 'significant'. And they *all* believe in multiregional evolution until recently? As far as I'm aware none of them championed it.

As for the 'gang overreacting to slim evidence' I think it tells us something about you that you consider the evidence slim without reading the book.

I never believed in multi-regional evolution, and in fact mocked it. Henry is one of the people who helped develop the Out-of-Africa theory.
I argue that Neanderthal interbreeding likely occurred and was significant because A. the math says we would have easily picked up most of their favorable genes with just a few tens of matings over all time and B. people will fuck sheep, given the chance, let alone Neanderthals. Part A is obvious once you look at the population genetics, and in fact has been seen in natural populations. Part B is an experimental fact.

As for why evolution sped up over the Holocene, the past 10,000 year or so, it's a combination of strong new selective pressures and increased population size (which results in more mutations): I'm reasonably familiar with the idea, since John Hawks and I came up with it.

The book sounds crazy from Tyler's description. Well, not that crazy, I guess there have been lots of popular science books like that: lots of speculation, occasionally well-grounded, hardly ever referring to confirmed research results. Given the complexity of the issues, speculation like that is almost guaranteed to be wrong.

Hawks-Harpending is a cool paper. But it doesn't imply that there are "interesting" differences between races, though. For now, the known differences that arose during the period of accelerated are mostly "chemical": skin pigments that tune vitamin D synthesis, enzymes to digest lactose, and lots of things conferring resistance to specific diseases. Then there are the body shape differences for climate. It's entirely possible that that's it.


I realize you feel hurt by the derision your little post has elicited in the comments, but this is a watershed moment for you. You can continue to take the safe, popular route, but you will sacrifice the respect of the small number of people whose opinion will, in the long run, matter.

There are new, often regional versions of neurotransmitter receptors and transporters. There are new regional versions of genes involved in the hair cells of the inner ear, key to hearing. There are new regional forms of genes that affect axon growth, synapse formation, formation of the layers of the cerebral cortex, and overall brain growth.

The fundamental assumptions of contemporary intellectual discourse are

A. Humans evolved by the process of Darwinian selection.

B. Then, as some point way in the past, humans stopped evolving. The Darwinian process was over and done with. Nothing to see, folks, just move along now.

Cochran and Harpending's book reviews the recent genome study evidence demolishing the second pillar of the reigning conventional wisdom, and provide a theoretical model for understanding why it is false.

That's a major step forward in the human sciences.

In response, Tyler complains, "What's missing is a sense of what the hypothesis does not explain..."

I suspect there were similar responses in 1859:, "Yes, Mr. Darwin, but what does your theory of natural selection _not_ explain?"

Well, we don't know yet what the Cochran-Harpending theory does not explain, but, like Darwin's theory, it just might explain a hell of a lot.

Which, I suspect, explains a lot about Tyler's aversion to thinking hard about it.

For now, the known differences that arose during the period of accelerated are mostly "chemical": skin pigments that tune vitamin D synthesis, enzymes to digest lactose, and lots of things conferring resistance to specific diseases. Then there are the body shape differences for climate. It's entirely possible that that's it.

Your list is pathetically inadequate. I haven't enough familiarity with the history of the literature to say exactly when such an enumeration would have passed muster, but 'sometime in the 1950s' would be a reasonable guess (and even then we could have added to your list of morphological differences). A search of the modern literature would provide you with a hundred significant variations, but why do that when it challenges your preconceived notions?
As for the differences being mostly 'chemical', um yeah, what do you think drives the processes of bodily growth, metabolism, and everything else. Of course the differences are chemical (or as you would have it "chemical").
And if the differences don't seem interesting, I suppose that's in the eye of the beholder. But in this case I'm thinking that 'interesting' is just a code word for 'not relevant to public policy', in which case I can assure you that you are indulging in pure wishful thinking.

Apparently, Sailor* can't help it that not everyone is as enthusiastic about this book (or any book that emphasizes racial differences) as he is. The ONLY reason why someone might not share his opinion on issues concerning races is because they are just being political correct.

*I know it's spelled as Sailer but I like it when people misspell it.


The Neanderthal thesis does NOT fly in the face of your mitochondrial DNA link. Mitochondrial DNA is passed on in a direct maternal line, while other (non-Y chromosome) DNA goes to both male and female children. With a small amount of interbreeding we wouldn't expect a random sample of mitochondrial DNA to be Neanderthal, UNLESS Neanderthal mitochondria gave a competitive advantage. However, any Neanderthal genes that DID provide a competitive advantage would sweep through the population: they (and immediately adjacent genes) would become common, while other Neanderthal DNA (including mitochondria) would remain vanishingly rare in the human gene pool.

"The calculation is an adaptation of the hypergeometric distribution, originally developed by Rudyard Kipling."

Dr. Cochran,

Would you mind explicating this for those of us in the back of the class?


" 'The calculation is an adaptation of the hypergeometric distribution, originally developed by Rudyard Kipling.'

Dr. Cochran,

Would you mind explicating this for those of us in the back of the class? "

I think Cochran was mocking the "just so story" comment in the review.


""People, go get upset over whatever but I already know this area a bit."

Prove it."

Seconded. It seems you are familiar with some of the conclusions, but don't have a firm grip on the math and biology, as illustrated by the non sequitur Wired link.

The thing about the Ashkenazi bit is that it doesn't provide any evidence of actual physical evolution (for intelligence) taking place on a limited time scale. It can all be cultural selection -- Judaism is a highly literary religion (people of the book), over time people with high verbal intelligence stick in the religion and others don't. Perhaps this issue is dealt with in the book; I know people sometimes have arguments for conflating the two forms of evolution.

The authors report that we have Neanderthal genes even though this seems to fly in the face of recent discoveries and more importantly the evidence that such interbreeding (if it occurred) mattered is extremely speculative.

That's only true if you know nothing about population genetics, and don't bother to read the population genetics in the book you've read.

Obviously for now we don't have great evidence that human - Neanderthal mating happened, or that it was common. And the authors say so.

But the key point is that it would only take a handful of matings to introduce advantageous forms of a gene from Neanderthals into humans. Each new allele favored by selection has a probability of fixing equal to 2*s, where s is the selective advantage. (Basically, how many more offspring to people produce who have it vs. do not have it.)

Much of what the Neanderthals gave us could have been useless, and so not subject to selection, and so not having the above fixation probability. But if it was useful, it would fit into the above. Moreover, it would be true for every instance where Neanderthals contributed their forms of genes into our genepool.

As far as I recall from the book, they do some back-of-the-envelope calculations to show that it wouldn't take very many matings for at least a few of their forms of a gene to stand a fighting chance of fixing in humans.

Here's an analogy that an Open Borders enthusiast will get: it wouldn't take many "cultural interbreeding" actions -- where people observe what the other group does -- for a good bit of culture to spread from one group to another. So if some type of Mexican food tastes great, we don't need common Mexican - Anglo encounters for it to take hold in America. Just a handful of exposures, and it'll stand a good chance at becoming a popular fast food, like tacos (or whatever).

For most of the time, Mexicans may be interacting culturally with one another -- speaking Spanish, watching Spanish-language TV, listening to Mexican music -- and ditto for Anglos. And a lot of what Mexicans have to offer wouldn't get far at all.

But despite that, we have a few things that are incredibly frequent in Anglo culture now: tacos, Spanish loanwords like "adios," and so on.


Actually, they do present evidence (it's in the original paper I linked above, too) that most Ashkenazi were in cognitively more demanding fields like finance, and that richer ones had a big advantage in surviving children (seen in wills and the like). But regardless, differential emigration/outmarriage would still result in the physical evolution of the ethnic group of Ashkenazi Jews, so that we would wind up with real, physical, genetic differences predisposing the people who identify as Ashkenazi Jews to greater IQ, and particularly verbal ability.

Imagine that I am breeding cattle for greater milk production. In Scenario 1, I only allow the most productive cows in each generation to breed. In Scenario 2, I drop the less productive cows of each generation on an island. In both scenarios I wind up with a population with different biological characteristics for enhanced milk production. I don't quite understand what you mean by 'physical evolution.'

Looks like Tyler is committing some crimethink here. I guess some uncomfortable truths just can't be faced straight on. Burn him!

Kentan speaks of "any book that emphasizes racial differences"

I don't see any emphasis of race differences in the book--they hardly mention them.


Never mind the above comment - I see if you hit previous you see the older comments. Sorry. I take the insult back.

I think someone may need to learn how to go from page 1 to page 2 of the comment thread.

Always interesting to see the form that hallucinations take.

Please cut the insults for our host, folks. Particularly if you are going to accuse Tyler of censorship in a comment box *directly below the link to the previous page of comments.*

Steve Sailer's point about paging the comments is a good one. Several people on have complained about theirs also, but at least their blog provider doesn't start a new page until it reaches 50 posts. A bigger problem with marginal revolution's comments is that a link (like the one Sailer has) goes directly to the **last** page of comments which is really disconcerting and irritating.

Not being a scientist nor well-read in the field discussed, I am not sure what is being contended in any of these arguments about the Cochran and Harpending book. I think I understand that the authors' assertion that evolution of humans is advancing more than the mainstream scientists believe (interesting word, believe, in a scientific context) it is or should, is annoying to the currently accepted wisdom. I think there is an implicit (perhaps explicit) assertion by the authors that there are non-PC difference among humans due to different rates of evolution, depending on factors such as whether a given set of people has urbanized. Perhaps I will learn more as this conversation goes on.

A thing that occurs to me is that long ago, cats, dogs, cattle and other non-humans collaborated with humans (and vice versa) to form social groupings for beneficial purposes, although rather one-sided against the cattle. I wonder if the evolution of these animals has had any interesting development alongside humans? I am not stating the question well, so perhaps a more informed reader can guide me in this.

Best wishes,
Ron Pavellas

Colugo says "perhaps they can tell us what they make of this article" referring to a recent Hofer et al. paper. I would point folks to John Hawks' discussion on his blog: what this article calls "large" differences in fact correspond to average Fst values of (what we think is) the neutral genome. Hawks points out that an allele with a frequency of 30% in one population and 70% in another (a large difference according to Hofer et al.) corresponds to an Fst of 16%, about the human average. So this is IMHO a paper that there was no reason to write.

What is the difference between the Coon and Wolpoff models? People _repeat_ over and over that Coon denied gene flow among regions, but he clear as a bell in the first few pages of his book that there must have been gene flow.


Ron Pavellas asks "I wonder if the evolution of these animals has had any interesting development alongside humans"

We point out a number of interesting parallels in our book. One great example is a myostatin mutation in beef cattle that causes grotesque muscle overgrowth. This would have been deadly in the wild but with veterinarians these animals can be born, and they are valuable because they have a lot more meat. There are some mutations in one of the bone morphogenetic proteins and their receptors that cause twinning in sheep. In the wild twins would never have made it but under domestication these mutants have been favored.

I don't see a picture of a myostatin mutant bull in our book but there is a picture of whippets on p. 219 with the same mutations. Whippets with one copy of the mutant are faster and win races but whippets with two copies (bully whippets) have muscle spasms and are poor racers. Same principle...


Ricardo says "As to the interbreeding with Neanderthals hypothesis, I think both sides should ask themselves what the falsifiable hypothesis is and what would it take to falsify it"

Yes, absolutely, of course. But it is non-trivial to come up with interesting and plausible hypotheses too. (These are also called speculations and just-so stories.) The reason to throw this one out is that it could explain why the northern arm of the modern human expansion led to the so-called creative explosion of the Upper Paleolithic while nothing like that ever happened in the southern arm. This is an old and interesting problem in anthropology but one that is never discussed in public because of PC run amok. FOXP2 is a good candidate for a gene stolen from archaic humans.

This is not my own favorite hypothesis about the creative explosion. I suspect that the people of the northern arm got into an ecological situation where men could work very little and otherwise parasitize women (like Northwest Coast Indians or Plains Indians after the horse) while people of the southern arm were in an ecology where men had to work to help feed the children (like Bushmen or Great Basin Shoshone.) One sees elaboration of art and technology in the former kinds of human societies, not so much in the latter.


Coon's model had some races transitioning to more modern grades sooner with others lagging, therefore living races represent different evolutionary stages. Wolpoff's model has a single interconnected but globally distributed species (since erectus) evolving in tandem on different continents.

(Anyone else read that goofy book Race by James Baker?)

Between allelic surfing and hitchhiking, it seems possible that the signal of recent positive selection can be overestimated in a rapidly migrating and expanding species.

Tyler's link on Neanderthal mitochonria DNA blows out of the water the idea he knows what he's talking about.

He's still a preening pansy. =)

It seems every few weeks there are documentaries on my TV that mention the skeletons of Neanderthal-Cro Magnon crossbreeds found first in Portugal (the most recent, 23-25,000 ya), but also now in east-Central Europe from 30-40,000 years ago, around the time modern humans first entered Europe. It looks like any hypothesis there was _no_ interbreeding has already been falsified.

It does remain possible that these crossbreeds did not contribute anything to our current gene pool though, eg if they were never accepted into modern-human groups. I think this is unlikely, but not impossible.

henry harpending at Jan 23, 2009 12:50:40 AM:

FOXP2 is a good candidate for a gene stolen from archaic humans.

For anyone else intrigued by this aside, here's Nicholas Wade's 10/18/07 NYT piece on the recovery of the Neanderthal FOXP2 sequence:
Neanderthals May Have Had Gene for Speech.

Wade is also the author of Before The Dawn (2006), a survey of human doings from "Out of Africa" to the invention of writing. That book should be a good companion to "The 10,000 Year Explosion," which the comments (if not Tyler's review) make me look forward to reading.

The 2007 article by Johannes Krause et al. describing the Paabo lab's work on FOXP2 is at the Current Biology website. The language-enabling form of FOXP2 is shared by modern humans and, they show, by Neanderthals, but not by chimpanzees and other living primates.

In contrast to Harpending's remark, Krause et al. suggest that the language-enabling sequence was already fixed hundreds of thousands of years ago, when the ancestors of Neanderthals and homo sapiens diverged. If correct, FOXP2 wouldn't be a candidate for a gene to have swept through modern human populations as a result of inter-species mating.

"The PC-ness of an idea has nothing to do with whether it is correct. Most racial ideas that are now not-PC were well-accepted only a few generations ago."
You slip from "PC" to "well-accepted." Were environmentalist ideas subject to active persecution when they were not yet well-accepted?

Compare two views: creationism (whether under the banner of ID or not), and the belief in average genetically-based IQ differences between self-identified races (e.g. ). No one is afraid of of creationists doing research to support their claims, since it is well-established that they are false, and when the Templeton Foundation offered grants to do such research, there were no creationist takers. Scientists can write clear and well-supported rebuttals of all the creationist claims. The only concern is with attempts to force this false view into science classrooms. This is a view that is not well-accepted, but there is no PC thought police attempting to prevent it from getting its fair shot.

In contrast, people are terrified of research into group differences in behavioral genetics, harassing researchers, using explicit double standards in reviewing publications, refusing to fund experiments, etc, etc. These tactics of PC are used because the evidence for the despised view is so much stronger than is generally admitted, and counterarguments so much weaker, that an actual discussion on the merits would favor belief in group differences. Instead, we get ad hominem attacks and other attempts to change the subject.

Creationists have a long history of retreat, as new evidence invariably weakens their case. In contrast, Jensen et al regularly make unpopular predictions that turn out to be right, but that their opponents cannot explain in a coherent framework. New evidence, e.g. MRI, twin studies, transracial adoption studies, DNA, etc, keeps supporting them.
Most of the relevant experts (when surveyed anonymously) think that Jensen is right (, but also hold a low opinion of him for making the truth public.

In these circumstances, PC does have something to do with our assessment of correctness: it tells us that we are getting a biased picture of the evidence, that experiments or data that would have strengthened one side are not being done or published, and that deferral to public interpretations of the literature (where one view is rewarded and the other punished for non-scientific reasons) is unwarranted (as opposed to the examination of the literature itself, correcting for publication bias, etc).

Ron is interested in his lactose intolerance and his carrier state for alpha-thalassemia. The frequency of tolerance declines a lot as one goes south in Europe, down to a gene frequency of 25% or so. If I had to guess I would say something like a lot of Greek ancestry. Of course I already knew that :-).


Randall Parker: I don't doubt that nature selection has been important since the emergence of anatomically modern Homo sapiens and that it increased since the adoption of agriculture. The question is by how much. Just maybe it's way less than Harpending, Cochran, and Hawks think.

Utilitarian: Isn't it possible that transracial adoption studies are detecting differential prenatal developmental programming rather than genetic differences? You don't know otherwise until we have transuterine studies.

I think Cochran is right that there was probably limited human-neadertal hybridization; it requires less special pleading than the purist Out of Africa models. However, all factions in the neandertal and modern human origins & evolution debates have been indulging in rampant speculation for a long time. There is a lot of inference about Neandertal cognitive abilities based on the presence or absence of certain archaeological features, and, as usual, way too much is being assumed. Same with anatomically modern humans, various stone age "revolutions" etc. (See: Mithen, Binford, Dibble, Richard Klein, Bruce Lahn...) They're good yarns, but that's all most of these models are.

One of the tedious parts of this debate is all the "well poisoning." Individuals may disagree without either being scared of being politically incorrect (although they might be) or a racist (although they might be). What's really sad is that this tone makes it into the scientific papers on the subject, where it has no place.


Could additional heterozygosity from the hybrid European-Middle Eastern nature of the Ashkenazi population counterbalance/mask diminished heterozygosity from genetic drift in the measures in question?

Henry: Regarding earlier comments, the idea that "race" is an arbitrary concept entailing the arbitrary groupings within a continuous population is better illustrated by Novembre et al. 2008 in Nature. They find that genetic differences are consistent with a diffusion model, not distinct groups.

Greg: The point of the Risch et al. (2003) reference was to point out that the rhetoric on these claims, from both sides, makes it sound like these conclusions are not debated.

Technically, it is more parsimonious to assume that drift caused the high allele frequencies compared to selection acting on all those loci, which all happen to affect the same character. Fewer assumptions are being made in the former; but parsimony is only a guideline. It is not evidence.

Regarding your carefully worded statement, "very detailed genetic information shows no sign of any bottleneck affecting nuclear genes among the Ashkenazim": The reference to nuclear genes is intended to counter the data of Behar et al. (2006) using mitochondrial DNA.

Hammer et al. (2000) and Olshen et al. (2008) conclude that data from the Y-chromosome and high-density polymorphism sequencing that the Ashkenazi did experience a bottleneck. Now, we can argue over whether their conclusions are correct, but saying that there is "no sign" is overstating the case.

It's revealing to compare the intellectual quality of Tyler's original post to that of the comments by Tyler's critics.

Steve said "It's revealing to compare the intellectual quality....."

I thought the review was pretty interesting. Granted, a few things like the Neanderthal mtDNA and the epigenetics were wild shots from outer space, but his main point, that he just didn't learn much, is worth thinking about.

I think Greg Clark's book is wonderful, but while I was reading it I often wished I could grab his shoulders, shake him, yell at him "just write a logistic equation!". Economist talk about rents, production, utility, and such just goes right past me. I can make no sense of it at all, never could.

Maybe Cowen is in my position, mirror-imaged?


Josh said "The PC-ness of an idea has nothing to do with whether it is correct. Most racial ideas that are now not-PC were well-accepted only a few generations ago."

I think it does: the PC SOBs almost always get it wrong. Take a look at

for a story that will make your blood boil. When I lecture about this my students worry that I will collapse from apoplexy. At the beginning of WWII we gathered the best this country could come up with to save civilization, and there was David Blackwell right in the thick of it, 22 years old, right off the farm in Illinois.

But Blackwell spent WWII teaching arithmetic at swamp state teacher's college somewhere in Mississippi because the President of Princeton University decreed that a Negro could not be fellow at Princeton. How many American soldiers died because of that arse-kissing glad-hander who was not fit to carry pencils around for David Blackwell? Note that it was the administration at Princeton, not the greatest collection of mathematicians and physicists the world had ever seen, not John von Neumann nor Richard Feynmann, who said that a Negro had to leave, it was the administration.

The same spirit of PC that sent David Blackwell to the hinterlands has destroyed the US urban public school system (read about the judiciary and the Kansas City school system): IMHO true believers ought to be led right to some rathole and sent down it.


Henry: I have to respectfully disagree. One cannot use a non-scientific metric to evaluate ideas. In science, we cannot use the fact that someone was wrong 9 times to automatically dismiss their tenth idea. Using the status of an idea as being PC or not-PC, is a terrible way to evaluate ideas.

Dubious: Track records and prior plausibility can be included in any analysis based on a Bayesian framework for incorporating prior information. Publication bias can be very difficult to evaluate depending on the criterion that is biased. Few people actively perceive their own work as being low quality. As a result, failure to publish can be blamed on external factors, which may or may not be the case. The best bet is to evaluate studies individually. If it is outside or in your field, the best thing to do is listen to the opinions of several disagreeing experts, which generally is not that hard to find.

On the particular subject being discussed here, the use PC/not-PC rhetoric to condemn/defend positions has extended to the literature. It has been argued that such questions should not be asked because they are dangerous. On the other hand, "martyr" defenses have been used suggesting that such inquiries are being repressed not for valid scientific reasons (e.g., plausibility, practicality, quality of the research program proposed, or utility of potential findings), but because the questions are viewed as being inappropriate or not PC. Neither of these external, emotional approaches contributes productively to the development of our knowledge.

A good rule of thumb is that, when anyone in any field makes a completely definitive, authoritative statement, they are probably being a bit deceptive.

Josh at 10:56 AM: "One cannot use a non-scientific metric to evaluate ideas. In science, we cannot use the fact that someone was wrong 9 times to automatically dismiss their tenth idea. Using the status of an idea as being PC or not-PC, is a terrible way to evaluate ideas."

Josh at 11:57 AM: "A good rule of thumb is that, when anyone in any field makes a completely definitive, authoritative statement, they are probably being a bit deceptive."

"Why would you wish to be associated with such a freakshow?"

I don't know who any of them are except Phil Rushton, and there's nothing wrong with Phil except that he is a grouch. I also don't know what the BNP is. I find the topic of the meeting pretty interesting.

The level of insanity from the AAA seems to have gone way down in recent years too.


Pincher: Thank god, at least someone hasn't lost their sense of humor.

Instead of talking in hypotheticals, let's take a concrete example. In their paper on Ashkenazi intelligence, Cochran and Harpending, with Harbison, postulate that the disease variants of genes for the sphingolipid disorders present at surprisingly high frequencies among the Ashkenazim, also cause a positive, overdominant effect (individuals with one disease variant and one normal variant of a disease gene have a better phenotype than individuals with two copies of either the disease or normal variant) on intelligence. Critics have argued that the authors drew more from the speculation of plausibility than was warranted. This postulate, however, did put forward the explicit prediction that these variants should have effects on intelligence that are readily detectable on IQ tests. Again, critics have argued that a test of this prediction should have been performed before drawing conclusions. The paper in question argues that there is good reason to conduct such tests.

Dr. Cochran, in a separate avenue of communication, informed me that they were prepared with collaborators to investigate this prediction, but that the study was quashed for PC reasons. This claim is difficult to verify. If it is true, the "quashers" are unlikely to admit that the reason was PC. Use of PC/notPC to evaluate a research program is unacceptable. Research funds, however, are limited. So, the project could have been quashed for numerous reasons including lack of interest, poor design, or perceived inapplicability of the results (for some, it is hard to determine what productive and ethical use society can make of genetic knowledge that certain individuals are predisposed to higher than average intelligence). I would be shocked and, frankly, disappointed if Drs. Cochran and Harpending felt that their research programs were uninteresting, poorly designed, or unimportant to society. Therefore, it would only be human to wonder if the "quashing" was not due to merit, but due to PC. In the absence of a smoking gun, it is hard to know.

Can we get a response from Cochran or Harpending on the 'quashing'?

"If it is true, the "quashers" are unlikely to admit that the reason was PC."
Actually, such admissions are surprisingly common, including in print.

Ben: Please support your assertion that the statement "race is not a valid biological concept" is "clearly incorrect" with further argument and evidence. At the very least, in the spirit of contrariness and rational debate, I choose to disagree.

I have encountered cases in my own experience, but giving the specifics would tend to reveal my identity and get me in trouble with those institutions. Fear of stigmatization of ethnic groups and PR backlash for the institutions were explicitly cited as reasons to quash research that was admitted to be scientifically important and well-designed.

Even more importantly, there have been attempts to fire or otherwise destroy the careers of researchers, which exerts a major chilling effect.

The extended persecution of Linda Gottfredson comes to mind:

Chris Brand was fired from his academic position for publishing a book on general intelligence, along with a partial defense of sexual relationships between gay men and male minors:

Helmuth Nyborg was suspended from his position for publishing an academic paper (!) on sex differences:

It was proper to fire Chris Brand for his bizarre defense of Gajdusek's abuse of those boys. I would have no problem if Arthur Butz were canned either. There are limits.

"Ben: Please support your assertion that the statement "race is not a valid biological concept" is "clearly incorrect" with further argument and evidence. At the very least, in the spirit of contrariness and rational debate, I choose to disagree."

Professor Steve Hsu has a number of comments on this, including the link I posted above regarding the Tang & Risch paper (Am. J. Hum. Genet. 76:268–275, 2005):


"The clustering data show very clearly that, in certain subspaces, the genetic variation within a particular population cluster is less than between clusters. That is, the genetic "distance" between two individuals within a cluster is typically much less than the distance between clusters. (Technical comment: this depends on the number of loci or markers used. As the number gets large the distance between clusters becomes much larger than the individual cluster radius. For continental clusters, if hundreds or thousands of markers are used the intercluster distance dominates the intracluster size. Further technical comment: you may have read the misleading statistic, spread by the intellectually dishonest Lewontin, that 85% percent of all human genetic variation occurs within groups and only 15% between groups. The statistic is true, but what is often falsely claimed is that this breakup of variances (larger within group than between group) prevents any meaningful genetic classification of populations. This false conclusion neglects the correlations in the genetic data that are revealed in a cluster analysis. See here for a simple example which shows that there can be dramatic group differences in phenotypes even if every version of every gene is found in two groups -- as long as the frequency or probability distributions are distinct. Sadly, understanding this point requires just enough mathematical ability that it has eluded all but a small number of experts.)"


Here is another link discussing that in the context of European, Nigerian and East Asian samples from HapMap and 100 African-American samples clustered based on data from ~600,000 genetic markers.

"2) Genetic distances between population clusters are roughly as follows: the distance between two neighboring western European populations is of order one in units of standard deviations and the distance to the Russian cluster is several times larger than that -- say, 3 or 4. From HapMap data, the distance from Russian to Chinese and Japanese clusters is about 18, and the distance of southern Europeans to the Nigerian cluster is about 19. The chance of mis-identifying a European as an African or E. Asian is exponentially small! (Table 5)"

Ben: The cited data shows that genetic divergence corresponds to geographic distance. Novembre et al. 2008 (in Nature) found the same thing across Europe. Individual's geographic origin could be accurately mapped based on DNA sequence. They also found that this pattern is consistent with a diffusion model of migration, not distinct populations. It would been quite surprising if we had not observed such geographic differences. This does not show that identifying a geographic point of origin gives us predictive power for biologically relevant phenotypes. If the point is just to categorize people, then race is as good a way to do it as any.

"Are we going to embark on a eugenics breeding program like Chris Brand? Gene therapy for infants predisposed to be stupid?"

Why on Earth not? We add iodine to salt, iron to bread, and have invested many billions in early education in order to enhance cognitive ability and close cognitive gaps between groups. We could close racial gaps in a single generation with assisted reproductive technology, and produce tremendous benefits. If enough men like Obama's father had contributed to sperm banks, and all the African-American single mothers used that sperm to conceive their children, then the test score gap could be gone or reversed by now. That would be a culture with scary 'eugenic' sensibilities but it would also be much happier, more scientifically advanced, more atheistic/socially liberal, more prosperous, and less plagued by racial inequality and discrimination. I could be happy with such a culture, just as I could be happy with one where PGD and embryo selection were used to eliminate cystic fibrosis.

But it is incredibly difficult to raise IQ in public debate for fear of drawing attention to the Unspeakable Truths of phenotypic differences and research into IQ and genetics is hampered by fear of proof of genetic differences beyond even unreasonable doubt (there will be public DNA sequences for representative samples of all groups, and info on the genetics of IQ can be compared against them). Once genotypic differences are established, the furious energy going into suppressing them and ineffective environmental approaches can be redirected towards biological enhancement to fix the problem instead of pretending that it doesn't exist.

"Would we get more impact from discovering the environmental factors (something we can manipulate) influencing intelligence?"
Not in the U.S. (in the developing world, micronutrients give more bang for the buck).

Dubious: Eugenics can be mocked on both ethical and scientific grounds. While the progress of domestic breeding programs are frequently cited, they also have substantial drawbacks. Selective breeding on a limited character set reduces genetic variation and tends to increase recessive alleles. For examples, take a look at the issues of pure bred dogs, pure bred horses, or the runty plants that are bred together to form modern seed stocks. There is no analogy between single nutrient interventions and selective breeding. Let's also not forget that the steps necessary for a human selective breeding program could quite possible be more detrimental to society than the results helpful, even under the rosiest scenario. It is easy to compare the starting point and the end point and be satisfied. The intervening steps would be extremely messy. Not to mention that are current understanding of the genetics underlying human traits is far too limited to run such a program efficiently.

My point on Gottfredson was that we cannot draw conclusions either way from her testimony. A book publisher withdrawing an offer is about business. Businesses make decisions for PC reasons. Their goal is to stay in business. Writing inflammatory stuff, whether correct or not, is both a good way to get publisher interest and a good way to get them to back away. My argument was not that no one had ever suffered for non-PC views, but that their persecutors would not admit to using PC as the reason. The only thing close to that standard was the editor's comments, but that case does not stand out in any field of research.

Your assertion that research funds are not limited implies that you have never been involved in the attempt to acquire research funds. Currently, about 90% of research proposals are not funded. I never argued that any honest approach toward building our knowledge should not be funded or is fundamentally useless. Gaining knowledge is always worthwhile. Funding, however, must include prioritization based on several factors, including the expected utility to society.

The assertion of an Ashkenazi advantage in verbal ability and disadvantage in visuospatial ability has always been perplexing. The argument for "g" is that there is a close correlation among different "intelligence abilities." If the correlation breaks down in one group, that indicates that these abilities are separable either due to genetic or cultural factors.

I would guess that the probability of any of your 5 postulates being "proven" in 5-10 years is close to 0 (P<0.1). We have, to date, found no evidence of large effect genetic variants affecting intelligence. I'm highly skeptical of Cochran, Hardy, and Harpending's suggestion that Ashkenazi disease alleles have a large, overdominant effect on intelligence. My prediction raises two issues. One, what is constituted as proof. Two, that the scientific method operates by disproving hypotheses, not by "proving" them. So, I would prefer to restate my predictions as a close to 0 (P<0.1) probability that the alternative hypotheses to those you have presented will be disproven.

"Your assertion that research funds are not limited implies"

Of course funds are finite and competition fierce, but given the existing priorities, including an intense interest in research to show that discrimination is responsible for group differences, refusing to test the competing hypothesis (the refutation of which would support the discrimination view) shows bad faith.

"The assertion of an Ashkenazi advantage in verbal ability and disadvantage in visuospatial ability has always been perplexing."

To you, you mean. You can read up on group factors and the hierarchical model yourself.

"So, I would prefer to restate my predictions as a close to 0 (P<0.1) probability that the alternative hypotheses to those you have presented will be disproven." Fascinating, we have a big divergence of opinion. If spending a few million on admixture studies, testing the Ashkenazi, etc, would almost certainly refute all the scientific racists of the world, why wouldn't you support doing the experiments? Then the James Watsons and Charles Murrays and Greg Cochrans of the world wouldn't have legs to stand on. It would reduce stereotype threat, weaken opposition to affirmative action, and have many other great consequences (in the world where genetic factors aren't responsible). In other words, results supporting no-difference hypotheses would have utility. You asked what the utility of evidence supporting the above hypotheses would be, but with such a low P, how is it that the expected utility of the experiments isn't positive? Very probably we get to refute inaccurate racism, with a tiny, tiny chance of confirming hereditarian views (and incidentally realizing that our models of many social phenomena had been mistaken, leading to ineffective action). With Steve I don't see how you can reasonably be so unmoved by the opinions of the experts in the field (surveyed anonymously) that you think they are almost surely wrong, while not understanding the psychological, statistical, and biological evidence motivating their views. However, since you're on record now, I'm content to wait on this until the Ashkenazi data become available, which shouldn't be too far off given the exponentially falling costs of testing. At that time one of us will look like a clown and we can update our beliefs based on the results.

My probability estimates are based on several lines of reasoning; but they are just a guess based on ad hoc Bayesian style inference. Personally, I would prefer to not make such predictions on scientific outcomes, but Dubious announced the intention to interpret a lack of repsonse on my part as either a concession or intellectual dishonesty, necessitating the following disclaimer. If I have missed replying to your point below, I apologize, but that omission does not in any way imply that I agree with, disagree with, or concede the point.

Lines of evidence include:
1. There is no evidence of genetic variants of large effect on intelligence is non-existent.
2. The proposed mechanism leading to an Ashkenazi advantage in intelligence is implausible on the basis of cell and developmental biology. The mechanism is non-parsimonious and completely speculative.
3. The calculation of response to selection is flawed. The analysis applied is for direct selection on a character, as when a pig farmer selectively breeds larger hogs. Correlated response to selection on fitness is the approriate analysis and, using their historical parameters, does not agree with observation.
4. Heretability of intelligence differences within populations does not indicate that differences between populations are due to genetic effects.
5. There is no comparison to the probability of seeing similar results either due to cultural effects or drift, only support for the plausibility of selection.
6. The general view of population geneticists is that there is no genetic evidence for genetic differences in racial IQs and that Ashkenazi disease allele frequencies are explicable by founder effects that did occur.
7. There is no precise definition of race on which to make these judgments. The ability to show geographic clustering is inadequate. One must specify a distance limit to define race in order for these hypothesis to have meaning.
8. I am inclined to be skeptical and insist on further proof especially when any theory is supported by going directly to the media, a conspiracy of repression is alleged, or contrary results are systematically dismissed in a non-rigorous fashion (e.g., the rebuttal of Risch et al. 2003 implies incompetence on the part of the editors and reviewers at PNAS).
9. The many alleles/very small effect hypothesis implies that some alleles must be at high frequency in certain populations but not in others, which would restrict the proportion of overall variants possible. As I am simply supporting the statistical null hypothesis, plausible arguments for divergent selection pressures and simulations to demonstrate a reasonable set of parameters to get the observed genetic and phenotypic distributions before I could consider this scenario likely.

My prior assumptions are no less rigorous than Cochran's general assertion, building on the malaria example, that recessive deleterious mutations are only seen in populations because they have a positive overdominant effect, usually due to pathogen resistance.

I never said the hypothesis of overdominant effects for Ashkenazi disease alleles was not falsifiable. I stated that Dubious asking for my guess as to the probability of "proof" of a hypothesis did not square with the scientific method. Karl Popper would have predicted that there was 0 chance that any hypothesis would be proven, as it is not possible.

Positive correlations between mental tests underlies the "g" concept. Significant differences in these correlations in one population indicates that these "intelligences" may be more separable due to genetic or cultural factors than previously thought. This observation says nothing about selection on intelligence.

One must resist the urge to twist my arguments into something they are not. Concerning research priorities, my point is that, when resources are limited, there are other factors that must be considered in the prioritization of research funding. The fact that the proposed research has not been funded is not evidence of a conspiracy to repress this line of questioning or bad faith. I could propose, alternatively, that all the "repressed" research has been denied funding due to implausibility, lack of justification, or lack of utility.

The argument that a "simple" experiment to "disprove" the scientific racists creates utility does not work in reality. First, it is unlikely that these results would change many minds. The consistent negative results of high quality studies on homeopathy and therapeutic touch have not changed the minds of there adherents. Humans just don't change their minds easily. Second, the question implies that there is evidence to support the scientific racist viewpoint (i.e., that the view is valid and is well supported). In statistical terms, the scientific racist point of view must disprove the null hypothesis that there is no difference between populations other than chance effects. In Popper's terms, the burden of proof is on the scientific racists to disprove the null hypothesis and not the other way around. Every theory that is thrown out there, no matter how falsifiable, does not necessarily merit investigation in our finite world.

AMac: One of the comments to which I was replying used the term "scientific racist" (Dubious Jan 27 1:12PM). I believe this was most popularly used to describe Chris Brand. It is not my term and I do not like it myself as it is vague and carries unproductive negative connotations. Again, I was replying to a specific question about the utility of the proposed race/IQ/Tay-Sachs experiments that specifically used that phrase. I do think humans don't like to change their minds. I do not think that this or any "group" is more intransigent than any other, which is why I've stayed in the discussion for so long.

#7: My objection to race not being precisely defined is in the context of my probabilistic prediction. We expect geographic clustering due to physical constraints (this is the diffusion model referenced in Novembre et al 2008). You are right that the correlation between race and geography is completely non-surprising in this context. This correlation does not mean that the categories of race provide useful predictive power (also definition dependent. This also does not give us non-arbitrary definitions (what distance should be used as a cut-off). There are a variety of potential approaches to this problem, but we do not know what the biological relevancy of any really are. Dubious' hypotheses need to be stated in the context of a precise and consistent definition of race to be quantifiably testable.

I agree that Cochran's hypothesis is testable. We simply disagree on the priority that testing should be given.

#1: If I only made one typo it's a miracle. CochranHardyHarpending have no evidence that such variants exist. Perhaps I should have said direct evidence. Their large effect variants are speculation.

#4: My statement about heritability is drawn completely from what heritability represents. Heritability in one comparison (e.g., within a population, which is the most tested in intelligence) is not applicable to other comparisons (e.g., differences between populations). The between populations testing could be done, but it is practically much more difficult. I take exception to the use of the following logic: IQ is heritable in one population, therefore IQ variation in this population is due to genetic factors, therefore IQ differences between populations are due to genetic factors. This logic is not valid.

And, with that, I believe I shall take my leave of this thread to do the things that I actually get paid to do.

Poor Josh. I feel for the guy. He really thinks the world will fall apart if C&H are correct on their main these -- which they obviously are. In the past, I would have debated people like Josh and would have swallowed all their innuendo, robust stupidity and leftist ideological fanaticism.

But now, I read stuff like that with a sense of pity. It's like seeing people still defending monarchism, Marx or Freud. Thank C&H for writing this book. Hopefully, we can now move on to serious debate and a new age of science where truth trumps and ideologues bite dust.

Btw, Tyler, you didn't learn much while reading this book? Wow, I must have missed some pretty good books the past few years!

The book is an introduction to an idea that goes against what is "commonly known to be true" and provides some groundwork for future research. I think you're a little harsh against a book whose intentions are such. In much the same way that the blank slate is obviously wrong, though too many believed and believe it, it seems equally obvious to me that the cessation of genetic change is wrong. So long as mutations occur, there will be evolution. That much should be obvious. And this is an idea that isn't even that recent. E.O. Wilson et al proposed cultural-genetic coevolution back in the early 80's. If I have any objection to the book, it's that they left out his work.

Comments for this post are closed