December 10, 2007

Is James Watson black?

From the Times of London:

JAMES WATSON, the DNA pioneer who claimed Africans are less intelligent than whites, has been found to have 16 times more genes of black origin than the average white European.

An analysis of his genome shows that 16% of his genes are likely to have come from a black ancestor of African descent. By contrast, most people of European descent would have no more than 1%.

The study was made possible when he allowed his genome - the map of all his genes - to be published on the internet in the interests of science.

“This level is what you would expect in someone who had a great-grandparent who was African,” said Kari Stefansson of deCODE Genetics, whose company carried out the analysis. “It was very surprising to get this result for Jim.

Watson won the Nobel prize, with Francis Crick and Maurice Wilkins, after working out the structure of DNA in 1953. However, he provoked an outcry earlier this year when he suggested black people were genetically less intelligent than whites.

This weekend his critics savoured the wry twist of fate. Sir John Sulston, the Nobel laureate who helped lead the consortium that decoded the human genome, said the discovery was ironic in view of Watson’s opinions on race. “I never did agree with Watson’s remarks,” he said. “We do not understand enough about intelligence to generalise about race.”

The backlash against Watson forced him to step down as chancellor of Cold Spring Harbor Laboratory, New York state, after 39 years at the helm. He had said he was “inherently gloomy about the prospects for Africa” because “all our social policies are based on the fact that their intelligence is the same as ours - whereas all the testing says not really”.

The analysis by deCODE Genetics, an Icelandic company, also shows a further 9% of Watson’s genes are likely to have come from an ancestor of Asian descent.

Newsday
of Long Island "reports:"
"News that geneticist James Watson inherited 16 percent of his DNA from an African ancestor may provide the Nobel Prize winner with a new perspective on his ancestry. But experts Monday said the percentage of Watson's DNA possibly contributed by someone of African descent illustrates that race is a counterfeit concept, having more to do with social notions than biological ones."

And the NYT leaps in here.

It could be true that Watson is 25% nonwhite (although the graph in the Times says 27%), but it sounds unlikely to me, based on simple genealogical arithmetic that nobody else seems to have done. The only evidence I can see for this claim is that Watson has wavy or curly hair and that his father spent a year at Oberlin, the most racially liberal American college of the 19th Century. Otherwise, this claim fails most reality checks.

Watson's new autobiography, Avoid Boring People, has a fair amount of information about his ancestors, including several old photos. His mother's side of the family were recent immigrants from the British Isles:
"Mother was the only child of Lauchlin Alexander Mitchell, a Scottish-born tailor, and Elizabeth (Lizzie) Gleason, the daughter of an Irish immigrant couple (Michael Gleason and Mary Curtin) who had emigrated from Tipperary during the potato famine of the late 1840s."

So, if his mother was 100% white, as this family history suggests, then his father would have to be 50% nonwhite, which sounds extremely improbable. There's a picture of James D. Watson Sr. on p. 5, and he looks like your average white guy. (Granted, old black and white pictures can be somewhat misleading, but still ...).

Further, his father's upper middle class family history suggests that his father's side of the family sure didn't suffer from racial discrimination. If his father was 50% nonwhite, then his paternal grandparents had to average 50% nonwhite (e.g., one was 100% nonwhite, and the other 100% white). Yet, if one or both were significantly nonwhite, nobody in late 19th Century America seemed to notice! His paternal grandparents were both Episcopalians. His grandfather was a stockbroker, his grandmother was an heiress from wealthy Lake Geneva, Wisconsin. The odds that two individuals who were, on average, one-third black could have thrived in such an anti-black social environment seems absurd.

Further, Watson's father had three brothers. Did any of them displays signs of being part black? For a mulatto attempting to pass into white society, there are many fraugh passages, such as what to do with their loved ones. Today, we hear about how Race doe not exist, but for those who passed from black to white it was terribly traumatic, generally leaving behind your family and and taking on a new identity. Philip Roth's novel The Human Stain gives a strong picture of what it is like to pass from black to white. It's based on the literary critic Anatole Broyard.

I don't see any more of a such a troubled passage of the Watsons

This reminds me of this great article I wrote in 2001 about a population geneticist doing a pioneering racial admixture study, who noticed that one of his subjects was determined to be 22% black. So, he looked into it more and discovered it was him! This came as a big surprise to him and all his relatives. I wrote it up and it was a wonderful human interest story. The only problem was that it wasn't true. As a reader pointed out to me, 22% means that, say, one grandparent was 7/8ths black, which somebody would have likely noticed. Later, the population geneticist took a look at his DNA again with better methodology and found he had been way, way off originally.

I hate being wrong ...

Racial admixture analyses are reasonably good for groups, but for individuals, at present, they can throw off some funny results. For example, one commercial firm often reports that Jewish customers are a little bit American Indian. Brent Staples, a black editorial writer for the NYT, took a racial admixture test and was told he was 18% Asian, which is another unlikely finding.

I don't doubt that the paternal side of Watson's family tree, which in the case of Watson's paternal grandmother goes back to a Thomas Dewey who landed in Boston in 1633, could include some blacks and American Indians. Yet, simple arithmetic shows that the chance of him being 25% nonwhite is vanishingly small.

My published articles are archived at iSteve.com -- Steve Sailer

47 comments:

Roger Thatchery said...

Watson's enemies see this as the coup de grace. They are trying to bury Watson in African skin. It works on so many levels, you see.

Watson's enemies are every bit as confident in their myth as the men who tried Galileo. Watson is a fool for not fighting these enemies to his last breath. Maybe (!) there is a very liberal woman in his life or a general climate of matriarchy in his inner circle to which he feels he must conform. After all, the cardinal rule of science, politics and culture in modern day Western society is that female expectations must be met. That is where the high brow James Watson and the low brow Don Imus meet.

Anonymous said...

might be a problem with genotype calling from the 454 sequencing data

Anonymous said...

here's one of the major skin color variants:
http://jimwatsonsequence.cshl.edu/cgi-perl/jw_details?name=rs16891982

watson is a heterozygote

but european heterozygotes are very rare:
http://www.ncbi.nlm.nih.gov/SNP/snp_ref.cgi?rs=16891982#Diversity

Anonymous said...

So what?

It still does not convince me that a country ruled by blacks is worth lining in. I know, because I fled such a country (guess which one!). Neither would I volunarily live in a black ghetto or study at a black university, because the outcome can be predicted. That's the bottom line. Of course you get exceptions but they are then just that.

robert said...

An undiscovered adulterous insemination would make the percentage a possibility.

Evil Neocon said...

Watson's enemies don't understand his argument. Watson argued that on average, Africans had lower IQs. Which of course does not discount the odd short-bus inhabitant of say, NE Asian race or the super-genius from Africa.

Truly ignorant statistically. Do Vern Troyer or Shaq represent white and black men respectively?

If they want to argue the stats (they don't because it seems the conclusive evidence is against them) argue the stats.

ghetto watson said...

"I hate being wrong ..."

In other words, Sailer can't get over his bizarre obsession with Mark Shriver. How many years are you going to carry this grudge, Steve? Agreed that Shriver isn't the brightest of bulbs, but your own reasoning about admixture tests is even worse.

Now, Decode's test results may be "wrong", but you simply don't know what you are talking about.

There is no reason to make the a priori assumption that "27% non-white admixture" needs to be from a recent ancestor, or through only one family line. It could be that *both* Watson's father and mother had some degree of non-white genetics - possibly derived from ancestors much more far removed than "greatgrandparents" - and that this ancestry became more prominent in Watson due to recombination/independent assortment of alleles.

Decode's analysis isn't telling us either what the error bars are for these estimates, nor what parental populations were used for the gene frequencies to estimate said "admixture."

In fact, given this data and the reaction to it, one wonders whether these tests using hundreds of thousands of markers are any more accurate and precise then those using far fewer.

By the way, "Native American" in Jews does not mean recent descent from Geronimo, so you can drop that song and dance as well. More likely, it may be the signature of some ancient Asian ancestry found in common between Native Americans and some Old World Eurasian groups.

With regards to Staples, I'd like to know exactly what test and version he took, what exactly the results were, and some sort of knowledge that the "18% Asian" isn't simply measuring a degree of Native American ancestry.

Steve should stop fuming over "being wrong due to that damn Shriver" five years ago and start writing more intelligently about genetic testing.

Decode, on the other hand, had better give more information on their tests if they wish to charge $985 for them.

ghetto watson said...

16 + 9 = 25, Steve, not 27.

Your fixation on Shriver is so all consuming, your rage so overwhelming, that you can no longer do simple maths.

For shame.

is ghetto watson good at basketball? said...

"As a reader pointed out to me, 22% means that, say, one grandparent was 7/8ths black, which somebody would have likely noticed"

Not at all. If you hate "being wrong" Steve, be as skeptical of your "readers" as you are of population geneticists.

22% "admixture" *may* mean a 7/8 grandparent of that race, *or*, it can mean systemic admixture from that source in one's family tree, possibly from sources much more ancient than a "grandparent."

Steve Sailer said...

"By the way, "Native American" in Jews does not mean recent descent from Geronimo, so you can drop that song and dance as well. More likely, it may be the signature of some ancient Asian ancestry found in common between Native Americans and some Old World Eurasian groups."

Well, duh ...

That's the point. If somebody who knows that his four grandparents were all born in the ghetto in Minsk takes the DNAPrint test and is told he is 4% Native American ... because some ancestors in Eurasia at the end of the Ice Age turned West instead of going East and crossing the Bering Strait, you shouldn't go around calling him part-Native American. It's an error by the test, one that presumably will be made to happen less as the technology and expertise improves.

Similarly, if James Watson inherited through, say, his Irish Nana some genes that are used by Decode as a marker of sub-Saharan ancestry but that have been, say, in his ancestors in Ireland since they came Out of Africa many thousands of years ago, well, the newspapers shouldn't call him part-black. That's not how we use the terms in the English language.

Now, I wouldn't at all be surprised if Watson was, say, 1/32 black, 1/6th black and 1/11th something else is socially improbable.

Somebody is a little over-worked up here.

Steve Sailer said...

Ghetto Watson says:

"22% "admixture" *may* mean a 7/8 grandparent of that race, *or*, it can mean systemic admixture from that source in one's family tree, possibly from sources much more ancient than a "grandparent.""

This is Humpty-Dumptyism, using words to mean whatever you want them to. At this rate, I could start my own DNA Admixture testing service. You send me a swab, or if you like, skip the swab part (it's icky and unnecessary), and I'll tell you your ancestry:

"All your ancestors came, ultimately, from Africa!"

See, I'll always be right! Who needs to sequence anything? If you complain that you aren't getting your $985's worth, pointing out that you have orange hair and freckles and probably haven't had many African ancestors since the Ice Age, I'll just get mad at you and tell you that I'm using words to mean what I want them to mean, not what other people think they should mean.

ghettto watson said...

"That's the point. If somebody who knows that his four grandparents were all born in the ghetto in Minsk takes the DNAPrint test and is told he is 4% Native American ... because some ancestors in Eurasia at the end of the Ice Age turned West instead of going East and crossing the Bering Strait, you shouldn't go around calling him part-Native American."

First, Steve, "4% Native American" is below the level of statistical significance for the test and can be viewed as, possibly, experimental error. Second, who is saying that a Jew with even stst. sig. "Native American" genetic signature is "Native American?" I didn't do it, DNAP doesn't do it - only misinformed "experts" such as yourself do it. As you said, "duh!"

"It's an error by the test..."

No. It's called (a) statistical signficance, and/or (b) ancient ancestry.

"...one that presumably will be made to happen less as the technology and expertise improves."

Apparently then, one million markers assayed by Decode aren't enough. Possibly this is so, but one also gets the feeling that the only "accurate" genetic test will be one that meets the preconceptions of some people.

"Similarly, if James Watson inherited through, say, his Irish Nana some genes that are used by Decode as a marker of sub-Saharan ancestry but that have been, say, in his ancestors in Ireland since they came Out of Africa many thousands of years ago..."

These companies use markers that differentiate between groups, not markers in common since "out of Africa." You're really dense here Steve.

"...well, the newspapers shouldn't call him part-black. That's not how we use the terms in the English language."

Let Decode tell us the methodology they are using so we can judge for ourselves, rather than listening to your pontification on the subject.

Mary Pat said...

If you want, I could figure out some probabilities for you. I'd need more info, though.

These "findings" are reported as 22% asian, etc., but what I want to know is what that means -- 22% of the genes looked at predominantly show up in asian people? I would need to know what they mean by that.

ghetto Watson said...

"This is Humpty-Dumptyism, using words to mean whatever you want them to"

No, Steve, you are a priori assuming genetic data are wrong just because you do not like the results. Don't misunderstand me - they may in fact be wrong - but if so, that'll be discerned from the evidence, not because Steve Sailer thinks it is "socially unacceptable" that "person X" has a particular genetic profile.

Sailer's hypothetical "out of Africa" genotyping test is really stupid - that's the kind of pablum one would expect from race deniers.

No, Steve, population geneticists (which includes the folks at Decode) differentiate population groups with markers that differ in gene frequencies between said groups. If they are just detecting markers common in everyone from "out of Africa", then all white folks would be getting results like Watson's.

We need to wait and see what Decodeme data look like for a range of tested individuals, but there are certainly large differences in results from people taking other "admixture" tests.

If such tests were just looking at common deep ancestry, results would be essentially the same. Yes, we are 'all' "from Africa", and, yet, Steve, there are wide ranges in African specific gene signatures.

So, what, is someone with less African than James Watson part Martian then? How come they don't share the same "deep common out of Africa" ancestry as does Watson?

Are they from Atlantis perhaps? Namor the submariner?

Who is twisting language and science now, Steve?

ghetto watson said...

"pointing out that you have orange hair and freckles and probably haven't had many African ancestors since the Ice Age"

I see. Having "orange hair and freckles" means one cannot have many "African ancestors since the Ice Age." The Sailer method of racial classification.

Here's a better genetic test than Decode or DNAPrint. Just have Sailer and wife look at pictures of people and decide which mulattoes look "Greek" or like a "Portuguese Count." That's real science.

What the hell do Decode know about population science anyway?

Guys like Sailer want to deconstruct population genetics like the most base race deniers when data do not fit their preconceptions, and then they accuse others of "getting all worked up."

No, Steve - Watson's data may indeed be wrong, but let us see some EVIDENCE of it, not your childish "calculations" about his "grandparents" "proving" the data "wrong."

ghetto watson said...

Especially when he was young, Malcolm X had reddish hair and freckles. I guess then he didn't have any African ancestors since the Ice Age, right, Steve?

Maybe, like DuBois, he was a "portuguese count."

ghetto watson said...

By the way, I agree that the newspaper coverage of Watson's data is misleading and politically motivated.

And, no, no one of intelligence will say that "Watson is black."

On the other hand, you cannot airily dismiss the significance of these genetic data by just saying that the test is "wrong."

In essence, Sailer is suggesting that only if the test had shown Watson 1/32 or less African would it have been "right."

In other words, first we assume what we think the results should be, and accept the results only if they meet these preconceptions.

I think Steve should stop writing on science, a subject he obviously does not understand.

EVIDENCE, Steve, not your personal opinions.

george said...

"Maybe (!) there is a very liberal woman in his life or a general climate of matriarchy in his inner circle to which he feels he must conform. After all, the cardinal rule of science, politics and culture in modern day Western society is that female expectations must be met. That is where the high brow James Watson and the low brow Don Imus meet."

What are you babbling about? Did some woman tick you off today or are you always bitter?

Jeff Burton said...

Jewish customers are a little bit American Indian

Mormons will love that.

Anonymous said...

Supposing the claim of the article is true, I don't see what this has to do with the veracity or importance of his claims about the prospects of Africa.

Seamus said...

So when DNAPrint reports that, genetically speaking, I'm 14% American Indian, this actually means I might be part Jewish rather than part Amerind? Damn! No affirmative action bennies for me, after all. (Since I was adopted, in the days of closed adoptions, I have no clue as to the ethnic background of my natural parents--except what I can tell by looking in the mirror, which indicates a substantial European component.)

Anonymous said...

All this "proves" is that otherwise intelligent people have great difficulty of thinking straight when it comes to issues of race, because they allow their burning desire to maintain social respectability to overwhelm elementary distinctions. Even if Watson were 100% black African, it would not invalidate anything he said. I'm no genius or expert on IQ, but even I understand that when discussing group differences in intelligence, you are dealing with generalizations, probabilities, overlapping bell curves, etc. etc. The upper-class twits jumping on this factoid seem to be saying, "look! Watson said blacks aren't as smart, and he's part black! He's calling himself an idiot ROTFLOL!"

Uh, no. If this is all the dogmatic egalitarians have left, they're closer to doom than I thought.

SKT said...

The technology to determine someone's ancestry from their DNA does not exist. Period.

Every few months we see a story like this where the whitest white guy turns out to have ancestors from Burkina Faso or something, or some Mongolian goatherd is allegedly part English based on DNA studies. It's all 100% bogus.

RKU said...

Well, offhand, Watson certainly doesn't *look* 16% Sub-Saharan African...or even 16% South Italian.

For comparison, I'd guess that Cornel West is something like 25% African. And Watson doesn't look much like Cornel West.

It hardly seems very likely that all the main genes coding for visible appearance would happen to go one way while all those other "marker" genes would happen to go the other.

So even leaving aside the strong genealogical evidence, count me a little skeptical about this particular claim...

rob said...

A hundred years from now, he'll be being celebrated as black scientist.

David said...

This is a black.

This is a White.

In the light of those (entirely characteristic) photos, it's instructive that race-deniers criticize the alleged "one-drop rule" out of one side of their mouths, and out of the other side now aggressively uphold it (in a slightly different guise) as a cherished principle.

I believe the testers have a strong political motivation to fake their results in this case. After all, "confirming" Dr. Watson to be a hypocritical nut would take a great deal of heat (actual and especially potential) off their enterprise.

I'm not charging fraud, since this is a public forum. I'm merely saying that there is a strong motivation for fraud. A very strong motivation. It should be looked into.

Apparently the testers feel that no such investigation will happen or that, if it does, they will be well-protected by well-placed and grateful friends.

L'Emmerdeur said...

My question to the authors of the above-referenced articles: since he has so much black in him, does this mean his views are invalid?

*SNAP*

Note to self-important racist journalists hiding behind filthy pseudo-liberal brothel curtains: average IQ doth not to IQ distribution and fat tails speak. Retards.

none of the above said...

This story is all about irony, not about science. 99+% of the readers have no idea what the test is measuring, what the results might mean, etc., but it lends itself to Leno-style jokes.

Are there good pictures of people with similar amounts of African ancestry? I would assume that 1/6 African ancestry would give you some identifiable features most of the time (a medium-brown father marrying a pale white woman produces kids who, pretty much always, are identifiably black, if rather light skinned). But I don't have any experience with seeing pictures and ancestry fractions together.

Some confounding variables here:

a. Passing for white requires not getting a lot of the African genes that determine hair texture and skin/eye color. So if Watson has some ancestors who were passing (one or both of their parents were socially black, they were socially white), those ancestors probably didn't give him a lot of the genes linked to obviously looking black.

b. The whole skin color hierarchy among American blacks historically means that light skinned blacks often married; that assortive mating could accelerate the arrival of kids with significant African ancestry, but most of it not affecting appearance.

c. It wouldn't be a huge shock if people who were socially white but had a lot of African ancestry wound up assortively mating, since they're likely to live in the same regions, have somewhat similar appearance and culture, etc. (I don't know if this applies to Watson's relatives, though.)

(b) and (c) in particular could give you multiple socially white great grandparents who didn't leave you looking black, but left you with substantial African heritage.

The biggest question this raises, to me, is whether anyone has looked at these ancestry measures correlated with IQ, especially when taking skin color into account. If the black/white IQ difference is all about culture, discrimination, stereotypes, whatever, then you'd expect African ancestry to have little correlation with IQ, once you'd controlled for skin color/appearance. Socially white people with lots of African ancestry would do as well as socially white people with little African ancestry. On the other hand, if the difference is mainly about some physical differences driven by genes, then skin color ought to have little correlation with IQ, once proportion of African ancestry is controlled for.

Rob said...

further 9% of Watson’s genes are likely to have come from an ancestor of Asian descent.

If he's old stock anglo-american, he may well be ~9% native american, which often shows up as Asian, at least in some DNA tests.

OTOH, are they really trying to make the argument "Watson could not possibly be right on race, intelligence, and economic consequences, because he's just a dim black"

bob said...

The biggest question this raises, to me, is whether anyone has looked at these ancestry measures correlated with IQ, especially when taking skin color into account. If the black/white IQ difference is all about culture, discrimination, stereotypes, whatever, then you'd expect African ancestry to have little correlation with IQ, once you'd controlled for skin color/appearance. Socially white people with lots of African ancestry would do as well as socially white people with little African ancestry. On the other hand, if the difference is mainly about some physical differences driven by genes, then skin color ought to have little correlation with IQ, once proportion of African ancestry is controlled for.

no one has done that. the main problem is that african ancestry predicted from photographs is highly correlated with genetically measured african ancestry (reported by mark shriver in a talk). however, in the general social survey, they did measure skin color one year (along with IQ). richard lynn wrote a paper about the data. there's a significant but modest correlation between skin color and IQ. thus, you can't distinguish a genetic explanation from a "racism" explanation. however, other covariants, like income, can be ruled out using this data. a more detailed study, which will NEVER EVER be performed, would be needed to tease apart the question.

Les said...

Does this mean Watson is no longer an impartial scientist?

Anonymous said...

Are there good pictures of people with similar amounts of African ancestry?

Joakim Noah is 25% African, and it's quite obvious from his appearance.

Peter
Iron Rails & Iron Weights

Svigor said...

JAMES WATSON, the DNA pioneer who claimed Africans are less intelligent than whites, has been found to have 16 times more genes of black origin than the average white European.

I wonder if the Times would be kind enough to point out a few of these "genes of black origin." Funny, I've had liberals telling me for years that there are no such things, but here's a journalist saying the opposite when it suits his purpose (i.e., "we all be mixed yo!").

This weekend his critics savoured the wry twist of fate.

What's so wry about it? They've discovered it's a "black man" saying blacks are less intelligent due to genes - sounds like their bubble's been burst, not Watson's. Or maybe they just savor crow, I dunno.

TGGP said...

It said the results were consistent with one great-grandparent being African. Assuming eight-grandparents that would make Watson 12.5% African, which is not the same thing as 25%. If you still think the times were too racist for nobody to have noticed, you could divide that 12.5% by eight and say all of his great-grandparents were about 1.6% african, which is surely low enough to allow for passing.

Truth said...

They would have to test his father to make sure that he is Watson's BIOLOGICAL father. I once read that 1/5 of all children born to married couples in America are not the father's!

mrs. anonymous said...

If this is all the dogmatic egalitarians have left, they're closer to doom than I thought.

You must not know anyone in public education. You cannot even begin to imagine.

Luke said...

Since we are all out of Africa, I propose that we urge all our children to apply for AA standing when applying to college. Let the burden of proof be on the college, not the family.

Also, ghetto watson: your arguments are so cool, so diplomatic, and so intelligent, they carry all before them. It's not fair.

Anonymous said...

They would have to test his father to make sure that he is Watson's BIOLOGICAL father.

They'd have to dig him up.

Peter
Iron Rails & Iron Weights

Martin said...

So what would Watson's critics make of this? "You can't believe what James Watson says. He's part black, and everybody knows that those people aren't so...... Oh, nevermind."

Anonymous said...

I agree it's a silly argument. Let's assume he's part black. So what? How are his arguments any less valid?

Liberals assume that race realists must be white supremacists or hate black people or whatever. If David Duke turned out to be part black or part Jewish, it would mean something.

Mary Pat said...

Let's just think about how much of your grandparents' genes you actually have.

Consider the 50% of your genes from your father. Well, he got 50% from each of his parents, so you should be an admixture of 25% of each of your paternal grandparents, right?

Not really.

It is theoretically possible (though not likely) that you could have ended up with 50% paternal grandfather and 0% paternal grandmother. All that is required is that they add up to 50%. And you can push that reasoning back as far as you like (though it gets less and less likely the farther back you go.) We already know it works like this with regards to the Y chromosome and mitochondrial DNA -- it's passed wholesale along from father to son, and mother to children, the only differences being mutations. (X & Y don't crossover, do they? I'm no biologist.) It's already known that there are correlations between genes that are physically close, as it's not like there's a perfect 50/50 shuffle for each gene.

So though you're not likely 50/0 with regards to your paternal grandparents, it's not shocking if you're pretty far from 25/25.

This is why I'm asking for probabilities and correlations by race. This "22% African" tells me nothing. If it's 22% of certain gene markers he has is associated predominantly with people from Africa, there's probably not much there there. Is this being done in a Bayesian manner?

ghetto watson said...

First, 3 points:
1. I agree with Watson's original comments on race and IQ. And, his genetic data should have no bearing on that.

2. Agreed that the newspapers' smug and irresponsible comments on this are wrongheaded and should be condemned.
3. I also agree that these results are somewhat suspicious as it does seem unlikely (not proof!) that Watson can be 16% "African"
Now:

"The technology to determine someone's ancestry from their DNA does not exist. Period."
The technology does exist, and is not only from commercially available tests but, more importantly, from population genetics laboratories such as demonstrated by the Price et al. Seldin et al., Witherspoon et al. papers, not to mention Neil Risch's voluminous work in racial clustering. You do not know what you are talking about. Period.

"Every few months we see a story like this where the whitest white guy turns out to have ancestors from Burkina Faso or something, or some Mongolian goatherd is allegedly part English based on DNA studies. It's all 100% bogus."

Citations, please. If you are talking about mtDNA or NRY studies, I'll agree that's "bogus" from an individual standpoint, but that's not what Decodeme is doing. They're looking at 1 million autosomal markers.

"For comparison, I'd guess that Cornel West is something like 25% African. And Watson doesn't look much like Cornel West."

That's what one can call the Steve Sailer racial analysis method. First you "guess" that West is "something like" 25% African - without any more evidence than what you think about his appearance. Then you say that Watson "doesn't look much like Cornel West" - as if that proves anything other than your opinion.

Once again - do not misunderstand. I'm also a bit skeptical as well that Watson can be up to 16% African, that seems at least several fold too high. It's quite possible that Decode is "off" in their algorithms and/or parental populations. As I said, the data may be wrong.

But this skepticism is properly addressed by having Decode's methodology examined, by observing what others get when tested with Decode, and, importantly, to have Watson double checked by another company using different markers and algorithms - DNAP for example.

One does not properly address the problem by stating - completely incorrectly - that ancestry cannot be determined by DNA (it certainly can), not by making "guesses" about Cornel West, and not by Steve's Shriver-obsessed rants about grandmothers and greatgrandmothers.


Now, looking at the Times article, I can't say I'm too impressed with Decode's current ancestry product. All they gave for Watson was a pie graph of European, African, and Asian percentages. Not only are there no error bars given and no estimate of statistical significance for any of these measures, but one can question the categories themselves. What do they mean by "European?" Where would a South Asian or Middle Easterner fall in this spectrum? Are Amerinds in the "Asian" group? So, only 3 ill-defined categories, no error bars, no information on what's significant and what's not.

Not good.

Anonymous said...

Mary Pat says:


(X & Y don't crossover, do they? I'm no biologist.)


Not true. There is a portion of the Y that does engage in recombination, and a portion that does not.

Anonymous said...

What has constantly amazed me, and now with the media on board, frightens me, is the inability for many people to understand what a bell curve is, how it works, what it means, and that "exceptions" support the "rule," the very structure of the curve!
So it doesn't necessarily matter that he has whatever percentage african genes, since it doesn't follow that all africans suffer from egregiously low IQ scores.
Some africans have been measured at 125 in particular IQ tests. The particular ones I'm speaking of were also students who were majoring in engineering in South Africa.
There are Africans with higher than world average IQ's, but they only define ONE side of the bell curve!
There are African's with VERY low IQ's, who make up the other side.
The Africans with 80-85 IQ's make up the majority of Africa, completing the bell curve!
Why is that such a brain teaser?!
Having 15% African genes doesn't "prove" anything at all, since they work in accordance with that bell curve, along with his predominant gene's bell curve.
It's a dangling variable, that in his case, seems to not have made any difference, and at the same time does NOTHING to adjust the the IQ Bell Curve of native Africans.
The song remains the same.

peace4all said...

Is Africa a country or a continent?
Afrcans have more DNA variations than the rest of the world combined. The difference between a Yoruba and a khoisan could be more than the difference between a Wollof and a Swede. We all came from Africa, wether we like it or not. We have been in African longer than we have been Chinese, Turks or Caucasians.
The ancient Greeks used to believe their white- skinned northern neighbours were intellectually inferior to them. What is intelligence? If the average prson uses less than 10% of their brain capacity, is intelligence the person who exercises more of it?
Is it nature or nurture? Are Jews smarter than Palestinians (both Semites)? Are Japanese smarter than Vietnamese (both Asians)? Are West African smarter than South Africans? West Africans are the smartest immigrant group in the united kingdom.
Until ethnic people write their own history and do their own tests, Africans, native Americans, Bangladeshis etc would be continually lynched psychologically by racists masquerading as scientists.

Is there any genetic component to pedophilia and serial killing?75% of serial killings happen in the U.S. Any genetic component here?

For 400 yrs, all the prevailing scientific tests and data proved that Blacks slaves were physically inferior to whites. Blacks, it was believed, could not box or play basketball. Intelligence was not even considered because blacks were not considered human. Now, all of a sudden, well, they can play basketball, they can box, they can play sports... they are human after all. They are human but not as human as whites, they are not physically inferior. They are equal but not as equal as whites. They are less intelligent.

The nazi scientists were brilliant in some ways as scientists. they had their data and so-called evidence that Jews and others were sub-human. These were not laymen,
they were scientists and intellectuals. They believed that not only are blue-eyed, blonde-haired caucasians smarter, they are physically superior to all.

Anonymous said...

"but Kari Stefansson, whose company assessed Watson's heritage, says he found enough errors in the public genome to have doubts about whether the 16 percent figure will hold up. For example, he says there are places where it appears that Watson has two X chromosomes, which would make him a woman."

....and the genes to correlate with "black"...rather "african populations" you blooming morons.

MrK said...

Alls these responses are disingenuous. Watson has clear African ancestry. I used to think that it was because he was part Jewish, but apparently that isn't it.

Remember that in the tv movie Double Helix, Wantson was played by Jeff Goldblum.

Look at this 1962 photograph at the Nobel Prize website.

Notice his frizzy hair, and rather pronounced maxizilla. (This was in the early 1960s, before Afros became popular, of course, so it isn't a fashion thing.)

Officially speaking, being 1/8th black would make Watson officially black in the United States (using either the '1/8th or more' and the 'one drop rule' definitions).

It should now also be clear that race in the US is a cultural definition, which has more to do with American history, it's reaction to the end of the slave trade, the historic fear of turning into the Caribbean, and the expansion of who would be a slave.

I wish more white people would come out about their black ancestry. It would create a sea change on the psychological/cultural issue of 'race' in America.