Posted: Sun 07 Sep 2008 03:16 Post subject: Can DNA Tell What “Race” You Are?
Can DNA Tell What “Race” You Are?
Molecular anthropologists are often asked if DNA markers can tell what “race” you are. The short answer is “no.” Mitochondrial DNA and Y haplogroups can tell from which continent your matrilineal and patrilineal ancestors came. And if you live in the Americas, autosomal mapping can tell what fraction of your ancestors came from Africa as slaves, what fraction came from Europe as colonists, and what fraction were Native Americans. But no DNA can tell your “race.”
Oddly, there is no contradiction between saying, “DNA cannot tell what race you are,” and saying, “DNA can tell where your ancestors came from.” To see why, consider one by one, mitochondrial DNA markers, Y chromosome markers, and autosomal markers.
Mitchondrial markers (mtDNA haplogroups) were discovered first and have been well studied. Everyone carries them. They descend through the maternal line, passing from mother to daughter. Men get theirs from their mothers but cannot pass them on. It reminds you of the clan system of the Native American tribes of the U.S. southeast, so mtDNA haplogroups are often called “clans.” The map shows prehistoric migrations of mothers who, over tens of thousands of years, carried the different mtDNA markers around the globe. Here are three examples: If your mtDNA is “H” then you descend from mammoth hunters who moved into Europe about 45 millennia ago. If your mtDNA is “B,” then you either descend from Native Americans or from their ancestors who remained in Asia about 22 millennia ago. If your mtDNA is L3, then you descend from east Africans.
Why does this not tell your “race”? Because ten generations ago (around the year 1800) you had a thousand ancestors, only one of whom (your mother’s, mother’s, mother’s …. mother) had that particular mtDNA. Your other 999 ancestors might all have had totally different ancestry. For example, my own mtDNA is “A” (Native American) and yet I am northern European in appearance and my family has no tradition of having any Native American ancestry at all. Hence, my mtDNA may be Native American, but my “race” in most people’s eyes is not.
Y chromsome markers (Y haplogroups) are more recent but are also well understood. Only males carry them. They descend through the paternal line, passing from father to son. It reminds you of the way surnames work in the Western world, so Y haplogroups are often used by genealogists pursuing surname ancestry. The map shows prehistoric migrations of fathers who, over the past tens of thousands of years, carried the different Y markers around the globe. If your Y is “R1b” then your ancestors arrived in Europe before the ice ages. If your Y is “Q,” then you either descend from Native Americans or from their ancestors who remained in Asia. If your Y is ExE3b, then you descend from west Africans.
Why does this not tell your “race”? Again, it is because only one man, out of your thousand ancestors from 1800, carried your Y haplogroup. The other 999 might have had totally different ancestry. For example, my own Y is R1* (from Cameroon in west Africa). And yet, as already mentioned, I have a northern European appearance and no credible family tradition of sub-Saharan ancestry. Hence, the “race” that most people see in me matches neither my mtDNA nor my Y.
Autosomal markers are different from mtDNA or Y haplogroups in that they measure the average continent-of-ancestry admixture in your entire DNA. The technique of autosomal mapping is more recent that either mtDNA or Y, and it is constantly being refined. It was originally devised for New World inhabitants, to measure what fraction of their ancestry came from each of the three demographic sources: Africa, Europe, and Native America. Autosomal mapping comes closer to identifying “race” than the other two. With a couple of exceptions, if your DNA admixture is overwhelmingly of European origin you are likely to be seen as White. And if it is of mostly African origin you will probably be considered Black.
The exceptions are because most of the population of the Americas is mixed to some extent. Virtually all African Americans have some European DNA admixture, about one-third of White Americans show detectable African admixture, and all Latin Americans are blended in proportions matching each region’s colonial population.
It often happens that an American who looks White and is seen as White actually has significant African DNA, and one who looks Black and is seen as Black has more than half European DNA. But most importantly, Hispanics are usually seen (and see themselves) as a third “race,” no matter what their DNA admixture shows.
In conclusion, neither mtDNA nor Y DNA can give any hint as to your “racial” membership in U.S. society. Autosomal DNA, if very lopsided to one continent or another, can suggest what you look like. But this does not work for those African Americans who choose to self-identify as Black despite having mostly European DNA, nor for White Americans, most of whom are unaware of having African DNA, nor for Hispanics, whose “racial” self-identity is determined by their culture, not by their genes.
Wasn't the second image old? I mean didn't scientists make a new map?
You may be right. The 2005 version shown above is the latest that I could find. I was able to find an earlier 2004 version, but no later one. If you can find a later one, let me know and I shall replace the one above.
Joined: 07 Oct 2007 {Posts: 248 } Location: United States
Posted: Sun 07 Sep 2008 14:16 Post subject:
If a North African underwent autosomal DNA testing, would his results show up as African, Asian, or European? I wonder if these tests could give us a clue as to how much "mixing" went on between continental boundaries, ie the populations in between these differentiated groups. Personally I want to see the results of East Africans and nomadic Sahelian ethnic groups.
Question: Are Arabs and groups of Asia minor grouped with Europeans, or Asians?
If a North African underwent autosomal DNA testing, would his results show up as African, Asian, or European? I wonder if these tests could give us a clue as to how much "mixing" went on between continental boundaries, ie the populations in between these differentiated groups. Personally I want to see the results of East Africans and nomadic Sahelian ethnic groups.
I do not know of anyone trying to use autosomal testing in that way. I doubt that it would work.
Autosomal testing works in the New World because the place was populated by 3 million Europeans and 11 million sub-saharan Africans after the natives were nearly wiped out. So you had three clearly distinct populations who were thrown together on the same day.
The Old World is not like that. Berbers, Arabs, and such are their own populations, who have been in place since Neolithic times and subject to the constant ebb and flow of migrations for tens of thousands of years. You could determine what fraction of the DNA of today's Berbers, say, came from Europe and what fraction came from sub-Sahara. I believe that several studies in our collection did just that, using mtDNA (limited for individual genealogies but useful for averaging populations). As I recall, they show slight Euro and slight sub-Saharan mtDNA (less than 20 percent) and overwhelmingly Berber DNA (their own thing).
As I recall from Arabian Knight's presentations, Arabs are similar. Sub-saharan DNA only on the maternal side, very little European DNA, but overwhelming of Arab DNA.
Grasshoppa wrote:
Question: Are Arabs and groups of Asia minor grouped with Europeans, or Asians?
It probably depends on who is doing the grouping and why. The studies we have collected treat them as distinct populations.
For that matter, come to think of it, the studies we have treat sub-Saharan Africans as dozens of genetically distinct populations also. The only exception, where sub-Saharans are all lumped into one category, is for autosomal admixture in the New World, and that is only because the goal is to examine the mixing among the New World's three demographic sources.
As I said, I am not sure that the New World's three demographic sources work for analyzing people in the Old World.
Posted: Mon 08 Sep 2008 04:42 Post subject: Re: Can DNA Tell What “Race” You Are?
fwsweet wrote:
In conclusion, neither mtDNA nor Y DNA can give any hint as to your “racial” membership in U.S. society. Autosomal DNA, if very lopsided to one continent or another, can suggest what you look like. But this does not work for those African Americans who choose to self-identify as Black despite having mostly European DNA, nor for White Americans unaware of their African DNA, nor for Hispanics, whose “racial” membership is determined by their culture, not by their genes.
The last paragraph seems to presume that White Americans who find out about some African ancestry would change their racial identification. I doubt that many would turn into Bliss Broyard, Jillian Sim, Mark Stebbins, etc. I would have said: "But this does not work for those Americans who believe they should self-identify as Black despite having mostly European DNA." Also: "...nor for Hispanics, who insist that their 'racial' membership is determined by their culture, not by their genes."
Posted: Mon 08 Sep 2008 19:56 Post subject: Re: Can DNA Tell What “Race” You Are?
Powell wrote:
The last paragraph seems to presume that White Americans who find out about some African ancestry would change their racial identification. I doubt that many would turn into Bliss Broyard, Jillian Sim, Mark Stebbins, etc.
I agree. I have changed it to be clearer.
Powell wrote:
I would have said: "But this does not work for those Americans who believe they should self-identify as Black despite having mostly European DNA."
Sorry, I don't see any improvement between your "...who believe they should self-identify as Black..." and my "...who choose to self-identify as Black..."
Powell wrote:
"...nor for Hispanics, who insist that their 'racial' membership is determined by their culture, not by their genes."
I have changed it to, "whose 'racial' self-identity is..." since "membership" sounds a bit strange. But I cannot see where "insist" adds anything, especially since Hispanics are seldom openly confronted with the issue to the point where they must "insist" on anything.
Posted: Mon 08 Sep 2008 21:22 Post subject: Re: Can DNA Tell What “Race” You Are?
fwsweet wrote:
Powell wrote:
The last paragraph seems to presume that White Americans who find out about some African ancestry would change their racial identification. I doubt that many would turn into Bliss Broyard, Jillian Sim, Mark Stebbins, etc.
I agree. I have changed it to be clearer.
I would prefer "believe" to "choose" because the first answer to the question, "Why do you call yourself black?" is usually the "I had no choice" mantra, citing the "one drop rule" if the person looks white or otherwise nonblack (Gregory Howard Williams, Adrian Piper, etc.) or the "I have to be black because I look this way" excuse (Barack Obama, Colin Powell, etc.). Indeed, only after people began pointing out that they did indeed have a choice, has there been any outrage at supposedly being deprived of the "choice" (to be "black" only - I have yet to hear much outrage at being deprived of the choice to be "white" a la Anatole Broyard) they didn't have.
Powell wrote:
I would have said: "But this does not work for those Americans who believe they should self-identify as Black despite having mostly European DNA."
Sorry, I don't see any improvement between your "...who believe they should self-identify as Black..." and my "...who choose to self-identify as Black..."
Powell wrote:
"...nor for Hispanics, who insist that their 'racial' membership is determined by their culture, not by their genes."
I have changed it to, "whose 'racial' self-identity is..." since "membership" sounds a bit strange. But I cannot see where "insist" adds anything, especially since Hispanics are seldom openly confronted with the issue to the point where they must "insist" on anything.
I would agree that Hispanics don't usually have to "insist" with individuals, but political power is exercised behind the scenes. We've had previous discussions, for example, of the tremendous efforts made by Mexican Americans to be classifed as "white" regardless of individual color, ancestry or phenotype.
Joined: 30 Mar 2005 {Posts: 1082 } Location: New Jersey
Posted: Tue 16 Sep 2008 15:16 Post subject:
Grasshoppa wrote:
If a North African underwent autosomal DNA testing, would his results show up as African, Asian, or European? I wonder if these tests could give us a clue as to how much "mixing" went on between continental boundaries, ie the populations in between these differentiated groups. Personally I want to see the results of East Africans and nomadic Sahelian ethnic groups.
Question: Are Arabs and groups of Asia minor grouped with Europeans, or Asians?
According to AncestryByDNA, the four groups distinguishable in their basic test are Europeans, sub-Saharan Africans, East Asians, and Native Americans / Central Asians. Lou Charlton explained that for their purposes, in the basic test (the Euro tests break this down much better), Southwest Asians, North Africans, and South Asians are grouped with Europeans.
North Africans -- judging from the chart on AncestryByDNA's website, which shows the results of a test done with only 7 people, apparently -- register with the following admixtures: 77.4% European, 15.0% sub-Saharan African, 5.6% East Asian, and 2% Central Asian / Native American. Again, I reiterate that the test used here was the basic AncestryByDNA test. The Euro test can distinguish various subgroupings.
The mtDNA of North Africans tends to be mainly their own, but reflects admixture from sub-Saharan Africa and Europe.
North Africans -- judging from the chart on AncestryByDNA's website, which shows the results of a test done with only 7 people, apparently -- register with the following admixtures: 77.4% European, 15.0% sub-Saharan African, 5.6% East Asian, and 2% Central Asian / Native American. Again, I reiterate that the test used here was the basic AncestryByDNA test. The Euro test can distinguish various subgroupings.
Do you know how much south asian/south asian/north african is in "european"?
North Africans -- judging from the chart on AncestryByDNA's website, which shows the results of a test done with only 7 people, apparently -- register with the following admixtures: 77.4% European, 15.0% sub-Saharan African, 5.6% East Asian, and 2% Central Asian / Native American. Again, I reiterate that the test used here was the basic AncestryByDNA test. The Euro test can distinguish various subgroupings.
Do you know how much south asian/south asian/north african is in "european"?
The DNAPrint test lumps South Asians and Middle Easterners with Europeans. Some racist purists feel that these groups should not be represented here. There new Euro DNA test breaks down the purely European percentages.
Joined: 07 Oct 2007 {Posts: 248 } Location: United States
Posted: Fri 31 Oct 2008 18:58 Post subject:
I still don't understand why autosomal tests can't tell what "race" you are. If they can accurately tell ancestry based on a continental scale, why isn't this accurate in categorizing folks as "races?"
'
FAQ
Search
Register
Log in
