18 Comments
User's avatar
Malcolm's avatar

Obviously, “white” means WASP. Not even Norwegians are white. 🤣

Expand full comment
Michael PRICE's avatar

"First, your DNA doesn’t contain segments labeled “European,” “African,” or “Chinese.” There’s no genetic switch for “whiteness.” Nearly all human genes exist in every population — we’re 99.9% the same species-wide.

What ancestry tests really measure is statistical similarity."

Well firstly we are not "99.9% the same species-wide" just because 99.9% of the genes have SOME representation in any particular ethnic group. All it means is that there are anomalies in each group that are similar to non-anomalies in other groups.

Secondd sufficient amount of statistical similarlity would prove someone was white (or another race) beyond any rational person's doubt. Take the probability a person of a particular race will have that allele divide it by the probability multiply it by the same thing for all the other alleles the person has*. The take that number and divide it by it's reciprocal plus itself. Do that for each race. Unless you have two high or more probabilities you have identified the race. If you do they you've identified what races they are a mixture of. This isn't finding the Higgs Boson it's basic statistics. I needed literally no more than high school statistics to figure this out.

Your problem is you think a statistical relationship can't be proof, when of course it can. Not proof in the Mathematical sense where by definition something must be true, but proof in the sense that you can literally bet your life on it. If the dice comes up "6" three times in a row that's not proof it's crooked. But if comes up "6" forty times out of sixty, yeah that's a crooked dice.

"You are not literally measuring quantities of Spanish and North African genes in people because genes do not belong to specific populations. They are not unique or labeled. Some genes are just more common in some populations than others. But they are very rarely, if ever, unique."

But the fact that they are more common in some populations than others very much does mean we can measure quantities of Spanish and North African genes. We can establish with certainty exceeded only in mathematics that someone has Spanish, North African or other ancestry.

"History books tell the story: North African Moors ruled parts of Iberia for centuries; Sicily and southern Italy were crossroads for Greeks, Arabs, North Africans, Normans.

As a result, many people in these regions today carry 5–20% North African or Near Eastern ancestry. So what happens now? Are they “less European”? "

Yes they are less European, at least in ancestry. Next question.

"It gets messier. Flip it around: if you test a North African with historic ties to Andalusia, they might show a moderate “Spanish match” — not because their family lived in Spain recently, but because these regions mixed so thoroughly in the past."

Yes, you've discovered races mix. Congratulations. That doesn't mean races don't exist or can't be genetically determined. It just means there are edge cases.

"Put in another way, if we made a North Dakota reference population and matched Norwegians against it there would be a strong match. That doesn’t mean those Norwegians came from North Dakota."

And now you've discovered reverse causation. Congratulations you really are learning.

"Let us look at the North instead of the South. Modern Finns and Estonians carry noticeable Siberian and Uralic ancestry — genetic traces that come from ancient migrations of peoples from the northeast. Does that make them “less European” than, say, a Swede or a Dane? Are they 50% white and 50% something else? Of course not — they’re just Finnish or Estonian."

No they are less white. The fact that you don't want to call them that because you think it must have political implications is just you putting your desires over facts.

"Push this logic to the present: imagine you build a California cluster today."

But that's the point, you can't build it. There is no statistical clustering between Californians similar to the ones between the Whites, Africans, Native Americans, Australians and other actual clusters. If you feed the genetic information of humans into a computer and told it to find clusters you wouldn't find one corresponding to California.

"If you tie rights — immigration, citizenship, office — to a DNA percentage, you create an incentive to game it. Companies would quietly stretch the clusters to match more people — like schools inflating grades to keep students and funding. Or a government could tighten the clusters to exclude whoever they want."

No they wouldn't because the maths wouldn't lie. You can identify white people, or Australian Aborigines or any other race rigourously and discriminate on that basis in an objective (if stupid) manner.

"If one company says you are 95% white and another says you are 80% which one should the government agency deciding if you may enter listen to?"

The one where the maths is actually correct.

"If people can pick any ancestry company to certify their “whiteness” then these companies will race to create more diverse reference groups so that more people will get over the required “whiteness” threshold."

Except again, race is real. The statistical relationships are knowable with certainty.

"The fundamental problem is that there is simply no objective measure to determine how “white” someone is. "

Let me rephrase this for you. "There is no objective way to determine whether someone's ancestry was in one particular group or not despite this ancestry being almost completely isolated from other groups for literally thousands of generations.". Now when I say it like that, does it seem likely? Or does it seem like the stupidest thing you've ever heard?

"So next time someone claims a DNA test can say who deserves to belong — remind them: there’s no “white gene.”"

Nobody said there was. Saying this will only make you look ignorant.

Expand full comment
Erik Engheim's avatar

I don't particularly like your tone, but I clicked like because you added to the discussion and gave good points and facts. So thank you for that. But you could of course have provided the exact same info in friendlier tone.

I have revised my article according to the criticism you have given here. We are probably not going to agree with I will acknowledge that many of my points were too simplified.

The clustering is an important point I should have covered, so thanks for brining that up. But I would argue my basic point still stands albeit weaker: These clusters can be manipulated.

While you cannot throw in arbitrarily people into a reference group as that would create new clusters, you can change how you sample. Include more people at the edges etc. Ways in which you can shift around the center of a cluster.

Obviously you cannot move that center wherever you want, but you can definitely tweak it and such tweaking will clearly be incentivized if you start politicizing ancestry tests.

Expand full comment
Michael PRICE's avatar

"I don't particularly like your tone, but I clicked like because you added to the discussion and gave good points and facts. So thank you for that. But you could of course have provided the exact same info in friendlier tone."

Dude you're lucky I didn't call you an ignorant moron who decided to ignore everything he knew about genetics, anthropology and statistic or knew less of it than the average high schooler. Think yourself lucky.

"I have revised my article according to the criticism you have given here."

Don't revise, delete. You said nothing worth preserving.

"We are probably not going to agree with I will acknowledge that many of my points were too simplified."

It's spelled "wrong" not "simplified".

"The clustering is an important point I should have covered, so thanks for brining that up. But I would argue my basic point still stands albeit weaker: "

It does not. You are simply wrong.

"These clusters can be manipulated."

No, because they are a physical fact which results form historical fact. The fact remains that the term "white race" is a valid category of ancestry. Yes there are people who are not 100% white, that doesn't mean that the term isn't valid.

"While you cannot throw in arbitrarily people into a reference group as that would create new clusters, you can change how you sample."

Which won't matter because for any significant amount of data the clusters reappear because again, they are physical facts.

" Include more people at the edges etc. "

But again, it's obvious who is at the edges. The math don't lie.

"Ways in which you can shift around the center of a cluster."

No, again, the cluster is a physical FACT, not an opinion.

"Obviously you cannot move that center wherever you want, but you can definitely tweak it and such tweaking will clearly be incentivized if you start politicizing ancestry tests."

You can try tweeking them all you want, the facts remain. I am not suggesting using the category "White" to do anything, I'm just saying that your attempt to pretend it isn't a valid category is a sad and pathetic attempt to impress people with fake science for a political agenda.. I don't even necessarily disagree with your agenda, but if you want to defend it try not saying things everyone intelligent can figure out is bullshit.

There are 3 billion nucleotides in a human genome. Of these all but 12 million will be identical to a reference sample [1]. Of the 12 million 6.3% of them are different due to race according to Lewontin, who was trying to downplay the difference. That's over 700,000 nucleotides. Do you really think that we can't detect this? Deny it's relevant to moral, political, intellectual or culture value all you want, just don't deny it's real.

[1]https://www.genome.gov/about-genomics/educational-resources/fact-sheets/human-genomic-variation#:~:text=On%20average%2C%20a%20person's%20genome,for%20the%20~0.4%25%20difference

Expand full comment
Leandra's avatar

Autosomal DNA test results are also only representing the last 7 generations of ancestors, approximately.

Expand full comment
Jake Carney's avatar

Why morphology isn’t a science and why psychology is by Economist and End times expert. Erik the Red.

Expand full comment
Erik Engheim's avatar

Aren't you adorable Jake. Following me everywhere like a little puppy dog. Can I help you with something?

Expand full comment
Wanha's avatar

Wow! Just wow! Tell me you know nothing about genetics, evolution or biology without telling me you know nothing about genetics, evolution or biology. Maybe some pharmacogenomics would do good to you. Or prepping elementary school again?

Expand full comment
Erik Engheim's avatar

By all means feel free to point out what is wrong and I can correct it. If you cannot I will just assume you don’t like the anti white-supremacy message.

Expand full comment
Michael PRICE's avatar

Well since you asked...

"First, your DNA doesn’t contain segments labeled “European,” “African,” or “Chinese.” There’s no genetic switch for “whiteness.” Nearly all human genes exist in every population — we’re 99.9% the same species-wide.

What ancestry tests really measure is statistical similarity."

Well firstly we are not "99.9% the same species-wide" just because 99.9% of the genes have SOME representation in any particular ethnic group. All it means is that there are anomalies in each group that are similar to non-anomalies in other groups.

Secondd sufficient amount of statistical similarlity would prove someone was white (or another race) beyond any rational person's doubt. Take the probability a person of a particular race will have that allele divide it by the probability multiply it by the same thing for all the other alleles the person has*. The take that number and divide it by it's reciprocal plus itself. Do that for each race. Unless you have two high or more probabilities you have identified the race. If you do they you've identified what races they are a mixture of. This isn't finding the Higgs Boson it's basic statistics. I needed literally no more than high school statistics to figure this out.

Your problem is you think a statistical relationship can't be proof, when of course it can. Not proof in the Mathematical sense where by definition something must be true, but proof in the sense that you can literally bet your life on it. If the dice comes up "6" three times in a row that's not proof it's crooked. But if comes up "6" forty times out of sixty, yeah that's a crooked dice.

"You are not literally measuring quantities of Spanish and North African genes in people because genes do not belong to specific populations. They are not unique or labeled. Some genes are just more common in some populations than others. But they are very rarely, if ever, unique."

But the fact that they are more common in some populations than others very much does mean we can measure quantities of Spanish and North African genes. We can establish with certainty exceeded only in mathematics that someone has Spanish, North African or other ancestry.

"History books tell the story: North African Moors ruled parts of Iberia for centuries; Sicily and southern Italy were crossroads for Greeks, Arabs, North Africans, Normans.

As a result, many people in these regions today carry 5–20% North African or Near Eastern ancestry. So what happens now? Are they “less European”? "

Yes they are less European, at least in ancestry. Next question.

"It gets messier. Flip it around: if you test a North African with historic ties to Andalusia, they might show a moderate “Spanish match” — not because their family lived in Spain recently, but because these regions mixed so thoroughly in the past."

Yes, you've discovered races mix. Congratulations. That doesn't mean races don't exist or can't be genetically determined. It just means there are edge cases.

"Put in another way, if we made a North Dakota reference population and matched Norwegians against it there would be a strong match. That doesn’t mean those Norwegians came from North Dakota."

And now you've discovered reverse causation. Congratulations you really are learning.

"Let us look at the North instead of the South. Modern Finns and Estonians carry noticeable Siberian and Uralic ancestry — genetic traces that come from ancient migrations of peoples from the northeast. Does that make them “less European” than, say, a Swede or a Dane? Are they 50% white and 50% something else? Of course not — they’re just Finnish or Estonian."

No they are less white. The fact that you don't want to call them that because you think it must have political implications is just you putting your desires over facts.

"Push this logic to the present: imagine you build a California cluster today."

But that's the point, you can't build it. There is no statistical clustering between Californians similar to the ones between the Whites, Africans, Native Americans, Australians and other actual clusters. If you feed the genetic information of humans into a computer and told it to find clusters you wouldn't find one corresponding to California.

"If you tie rights — immigration, citizenship, office — to a DNA percentage, you create an incentive to game it. Companies would quietly stretch the clusters to match more people — like schools inflating grades to keep students and funding. Or a government could tighten the clusters to exclude whoever they want."

No they wouldn't because the maths wouldn't lie. You can identify white people, or Australian Aborigines or any other race rigourously and discriminate on that basis in an objective (if stupid) manner.

"If one company says you are 95% white and another says you are 80% which one should the government agency deciding if you may enter listen to?"

The one where the maths is actually correct.

"If people can pick any ancestry company to certify their “whiteness” then these companies will race to create more diverse reference groups so that more people will get over the required “whiteness” threshold."

Except again, race is real. The statistical relationships are knowable with certainty.

"The fundamental problem is that there is simply no objective measure to determine how “white” someone is. "

Let me rephrase this for you. "There is no objective way to determine whether someone's ancestry was in one particular group or not despite this ancestry being almost completely isolated from other groups for literally thousands of generations.". Now when I say it like that, does it seem likely? Or does it seem like the stupidest thing you've ever heard?

"So next time someone claims a DNA test can say who deserves to belong — remind them: there’s no “white gene.”"

Nobody said there was. Saying this will only make you look ignorant.

* don't do this for the things we share with all humans or all hominids, it won't change the result but it is a waste of time.

Expand full comment
Erik Engheim's avatar

Not sure why you double post, but as I mentioned, I have revised the article now to address your criticism. I acknowledge you had many valid points. But we are not going to agree fully on this.

Expand full comment
Wanha's avatar

Just ask AI to point out the key problems. It wrote it in the beginning, after all.

Expand full comment
Erik Engheim's avatar

It doesn’t see any problems so now it is on you to make your case.

Expand full comment
Wanha's avatar

Your thickness is amazing 😍

Expand full comment
Erik Engheim's avatar

Listen, come with something useful or I am just deleting everything you wrote. If you have genuine useful criticism to come with, it gets to stay. I don't have any problems with criticism. But if all you got is trolling, you are out.

Expand full comment
Wanha's avatar

Looks like your style in promoting yourself, despite clearly being as ignorant as one can be on multiple topics you write, I'm happy to withdraw myself from your Substack. Tack och adjö och samma på norska.

Expand full comment
Don Quixote's Reckless Son's avatar

There is a person on this thread who knows nothing about those subjects but it ain't Erik. I'm sorry the education system failed you.

Expand full comment