A DNA Test Cannot Tell You Your Ancient Ancestors

For many years now, companies like Ancestry and 23andMe have been selling DNA tests that they claim can tell people where their ancestors came from. Their tests have become quite popular, despite the fact that they often present results in misleading ways that appeal to popular racist ideas about “blood quantum.” Now, some companies are trying to convince people that they can use DNA tests to trace a modern person’s genetic ancestry back to specific ancient cultures.

One such company calls itself “My True Ancestry.” This company does not conduct DNA tests of its own, but allows users to upload their DNA test results from other companies so that their software can automatically compare their genomes to those sequenced from ancient remains. The software then generates a pie chart showing what percentage of a person’s DNA supposedly comes from each ancient culture. This company, however, and others like it, are blatantly misrepresenting both how genetics works and what ancient populations were like.

How DNA tests to “determine ancestry” mislead people

I’ve written before about how companies that make genetic tests that claim to be able to determine a person’s ancestry frequently advertise these tests and present data in a misleading manner. I most notably addressed this issue in this article I wrote in February 2020 about whether modern Greeks are descended from the ancient Greeks.

Ancestry, 23andMe, and other companies that sell genetic testing for genealogy frequently produce advertisements that depict people with pie charts that portray them as having a certain percentage of their DNA from certain countries. If you buy a genetic test from them, they will give you your own pie chart that presents your ancestry in a similar manner.

This manner of portraying a person’s genetic ancestry using a pie chart is clearly inspired by the racist concept of “blood quantum,” which holds that ethnicity is biologically innate and a person who is descended from people of different ethnicities has a certain fraction of their “blood” that is “of” that ethnicity.

ABOVE: Screenshot from an advertisement for a DNA ancestry test from Ancestry, showing a man looking surprised to find out that “52%” of his DNA supposedly comes from “Ireland, Scotland, and Wales”

Ethnicity: a social construct, not genetically innate

The first problem with this way of presenting genetic ancestry is that it assumes that ethnicity is biological. In reality, ethnicity is a social construct. There is no gene or combination of genes that makes a person, say, Scottish. Instead, whether or not a person is Scottish is determined by things like:

• Do they currently live in Scotland, have they previously lived in Scotland, or do they have ancestors who lived in Scotland at some point?
• Do they identify in some sense with Scottish culture?
• Are they generally seen as Scottish by members of the Scottish community?

You’ll notice that these are all social factors, not genetic factors, because social factors are what really matter in determining a person’s ethnicity.

In some cases, ancestry is an important factor in determining a person’s ethnicity—but only when members of a community hold that, in order to be a member of their community, you need to be descended from members of the community. This is therefore still a social factor, because it is a community standard, not an inherent genetic quality.

ABOVE: Photograph from Wikimedia Commons of men of Scottish descent marching in kilts playing the bagpipes in Christchurch, New Zealand

Genetic markers

It is true that some genes—known as genetic markers—generally tend to be more common among people with ancestors from certain populations. These genetic markers, however, do not determine a person’s ethnicity; they merely happen to correlate with it and, indeed, most of the time, they don’t even correlate with it especially closely. There is often much greater genetic diversity within a single given population than there is between two different populations.

Having or not having a certain genetic marker does not automatically mean that someone does or does not have ancestors from a specific population. Even if a genetic marker is more common within a given population, not everyone within that population will have that particular marker. Furthermore, the same marker is nearly always present in at least some people from other populations, only with less frequency.

For instance, the ABCC11 gene variant that causes people who have it to have dry, flaky earwax and less body odor is most common among people with East Asian ancestry, but it is found in some people with European ancestry as well. Meanwhile, the LCT-13910 C/T gene variant that causes people who have it’s bodies to continue to produce the enzyme lactase, which digests the protein lactose, a sugar commonly found in milk and other dairy products, after childhood is most common among people with European ancestry, but it is found in people of other ancestries as well, only less commonly.

ABOVE: Map from Wikimedia Commons showing the frequency of lactase persistence across different native populations of Afro-Eurasia and Australia. Lactase persistence is most common among people with European ancestry, but is found in other populations as well.

Sometimes, the same genetic marker is simultaneously common in multiple populations that are not closely related to each other. For instance, the MTHFR C677T gene variant is most common in people of Mexican, Colombian, Chinese, Mongolian, South Korean, Japanese, Italian, French, and Greek ancestries. It is also less commonly found in other populations of people from all over the world, including Spain, Britain, Ireland, Poland, Tunisia, Turkey, and Iraq.

When a genetic researcher identifies a person as having this genetic marker, they have no way of immediately knowing whether the person inherited the gene from a Mexican, Colombian, Chinese, Mongolian, South Korean, Japanese, Italian, French, or Greek ancestor, or from another ancestor from a different population who just happened to have it.

ABOVE: Map from this article published in the Indian Journal of Biochemistry showing the frequency of the MTHFR C677T allele in different countries all over the globe

More complexities involving genomes and reference populations

Things get more complicated, though, because a person’s genome is not necessarily reflective of their complete ancestry. Each person normally only has forty-six chromosomes. Assuming that you are like most humans, you inherited exactly twenty-three chromosomes from each of your parents and each of your parents inherited exactly twenty-three chromosomes from each of their parents—but you didn’t necessarily inherit a quarter of your DNA from each grandparent, since the chromosomes that are passed from your parents to you are random.

This means that, in all likelihood, you’ve inherited more DNA from some ancestors than others. Furthermore, because every person has more ancestors than chromosomes, especially as you go further back in time, there are almost certainly many ancestors from whom you’ve inherited no genetic material whatsoever.

Additionally, it’s important to remember that, when someone takes a DNA test through a company like Ancestry or 23andMe, the company doesn’t look at the person’s entire genome. Instead, they only look at the genetic markers, which actually make up an extremely tiny fraction of a percentage of your overall genome.

Genetic researchers have sorted which genetic markers are more common in certain populations based on reference groups of people with known, documented ancestors from those populations. This is what companies use to tell customers where they think their ancestors most likely came from. Which genetic markers the companies associate with a certain ethnic group, however, depends to a large extent on who is included in the reference population for that group. This is why it is important that reference populations are very large and that they accurately represent the internal diversity within each population.

ABOVE: Diagram from Wikimedia Commons showing the double helix structure of DNA

Very limited samples for ancient DNA

Genealogical DNA tests can be useful for answering certain kinds of questions under some circumstances. If you have no idea where your ancestors came from, a DNA test can give you a pretty good general idea. A DNA test, however, should not be seen as the be-all, end-all of genealogy. And, unfortunately, when people start trying to claim that they can use genetics to determine which ancient cultures you have ancestors from, things get extremely silly very quickly.

The reference populations that genetics companies like Ancestry and 23andMe rely on are generally composed of very large groups of individuals who were all alive at the time their DNA was tested, meaning they could answer questions about their personal background, and whose ancestries are well documented.

When it comes to ancient reference populations, though, we have almost exactly the opposite story. The DNA sequences that make up ancient reference populations are all extracted from the remains of ancient individuals, but only a very small percentage of ancient people’s remains have survived to the present day, most of the remains that have survived are in very poor or fragmentary condition, and only an extremely minute fraction of the ancient people whose remains have survived to the present day’s genomes have been genetically analyzed. Even the genomes that have been analyzed are usually fragmentary or incomplete.

This means that the genetic reference populations we have for ancient peoples are, by their very nature, extremely tiny and incomplete. For some ancient peoples, the entire reference population may be composed of only two or three individuals whose genomes have only been partly sequenced. For some ancient peoples, there may not be any reference population whatsoever.

ABOVE: Photograph from Forbes of the skeleton of an ancient Greek athlete on display in Taranto, Italy

Not knowing the background of ancient individuals used for DNA analysis

There is also the huge problem that all of the individuals who are included in these reference populations are long dead and, in most cases, we have little to no reliable information about who they were, what their backgrounds were, or where their ancestors originally came from.

Immigration, travel, invasions, colonization, and population change were all extremely common phenomena in the ancient world. Thus, when an archaeologist excavates human remains from the ground at a certain location, in most cases, they have no way of knowing whether the person to whom those remains belonged was born in the place where their body was found or whether their ancestors came from that place.

They essentially can only guess the person’s background and cultural affiliation based on the archaeological context in which they found the remains. In some cases, the context may reveal a lot of useful information, but, in some cases, it may reveal very little information.

So, if a DNA company tells you you’re 30% Phoenician or whatever, that doesn’t mean that 30% of your ancestors were Phoenician or that you have 30% “Phoenician genes.” Instead, it means that 30% of the genetic markers they identified within your larger genome have been identified as tending to be generally more common among the very small number of poorly-preserved skeletons that archaeologists have found in contexts that make them think those skeletons most likely belonged to Phoenician people and that have had their genomes sequenced.

ABOVE: Photograph from Wikimedia Commons of an Assyrian relief carving from the city of Nineveh dating to around 700 BCE depicting an Assyrian warship, probably built and oared by Phoenicians

Misrepresenting ancient cultures as genetically distinct and homogenous

So far, I’ve given what I think is a fairly decent overview of why any company that claims it can trace a modern person’s genetic ancestry back to a specific ancient culture is inherently making a very dubious claim. Now I am going to talk more specifically about My True Ancestry and why their particular presentation of data misrepresents not only how genetic testing works, but also the historical nature of ancient cultures.

It’s fairly obvious that My True Ancestry is deliberately marketing itself primarily to white people who fantasize about being descended from Vikings or Spartan warriors. The company’s logo includes an image of a Corinthian helmet resembling the ones worn by the Spartan characters in the movie 300. As I discuss in this article I wrote in November 2019, 300 and the Corinthian helmet portrayed in it have become extremely popular among white supremacists—despite the fact that the Corinthian helmet was actually most closely associated in antiquity with the city-state of Corinth, not Sparta, and the historical Spartans were actually known for wearing a different kind of helmet known as a pilos.

If you go to My True Ancestry’s website, the advertisement at the top of the screen features the ancient Celts, ancient Greeks, Lombards, ancient Romans, and Vikings, in that precise order. All of these are ancient peoples who lived in Europe and are traditionally thought of as “white.” The site quite noticeably does not feature any ancient cultures from outside Europe in their advertising. This is clearly deliberate.

ABOVE: Screenshot from the top of My True Ancestry’s website. Notice the Corinthian helmet in the logo and the advertisement about “Celtic Bloodlines.”

Below is an image of a sample pie chart that is used in a review of My True Ancestry on the Nebula Genomics site. It’s almost impossible to convey how mind-boggling wrong this whole thing looks to anyone who knows anything at all about pre-modern history. There is so much wrong with this pie chart that I could literally write a whole book debunking it and I still wouldn’t have covered everything that’s wrong with it.

For instance, the pie chart shows “Minoans,” “Ancient Greeks,” “Romans,” “Hellenic Romans,” “Byzantine Empire,” and “Ottoman Empire” as though these were discrete, genetically distinct, internally homogeneous ethnic groups, but they most certainly are not.

ABOVE: Screenshot of a sample pie chart from the Nebula Genomics review article about My True Ancestry, showing a person’s supposed ancient ancestry

“Minoans” as a genetic category?

The Minoan civilization was a civilization mainly centered on the Greek island of Krete that existed from around 3100 BCE until around 1100 BCE. The Minoans were a very early ancient people who mostly lived within a relatively small area, so, on some level, it makes sense to think that they probably shared a lot of genes in common—but that’s only until you realize that the Minoans traded extensively throughout the Mediterranean world, seem to have had extensive diplomatic ties to Egypt, and it is very likely that there was some intermarriage going on between Minoans and some other ancient peoples.

Furthermore, the Minoan people did not just die out after their civilization ended; their descendants continued living throughout the Greek islands well into the historical period. Many of them probably assimilated into Greek culture, but at least some of them retained their own culture and language. Most scholars believe that the undeciphered language known as Eteokretan, which is attested in inscriptions dating as late as the third century BCE, written using a form of the Greek alphabet, is, in fact, a late form of the Minoan language.

ABOVE: Photograph from Wikimedia Commons of an Eteokretan inscription from the site of Praisos on the island of Krete, dating to the late seventh or early sixth century BCE

“Ancient Greeks” as a genetic category?

“Ancient Greek” is also useless as a genetic category, because individual people in ancient Greece had many different ancestries and ancestors from many different groups. It is likely that all or nearly all ancient Greeks had ancestors from at least some of the various Pre-Greek peoples that lived in Greece before the arrival of the Proto-Greeks, including the Minoans. Thus, “Minoan” and “ancient Greek” are not even distinct genetic categories from each other as the use of a pie chart would imply.

Furthermore, the vast majority of ancient Greeks also almost certainly had significant ancestors from the Proto-Greeks, who spoke an Indo-European language ancestral to Greek and settled in Greece sometime around 2000 BCE or thereabouts. Thus, from the very beginning of their civilization, ancient Greek people were not ancestrally homogeneous; they had mixtures of different ancestries.

Many Greek people in antiquity also had ancestors from various ethnic groups of Asia Minor and the Levant. For instance, the historian Herodotos of Halikarnassos (lived c. 484 – c. 425 BCE) describes Thales of Miletos (lived c. 625 – c. 545 BCE), who is traditionally said to have been the first Greek philosopher, in his book The Histories 1.170.3, saying that he was “τὸ ἀνέκαθεν γένος … Φοίνικος,” which means “by descent, the nation of the Phoenician.”

Herodotos himself was born in the city of Halikarnassos in Asia Minor and may have been of mixed Greek and Karian ancestry. The Souda, an encyclopedia written in the tenth century CE based on earlier sources, records that his father was named Lyxes. Various other ancient accounts state that his uncle or cousin was an epic poet named Panyassis. These are both distinctly Karian names.

ABOVE: Photograph from Wikimedia Commons of a second-century CE Roman marble portrait bust of the Greek historian Herodotos of Halikarnassos, a conventional portrayal based on a Greek bronze original of the fourth century BCE

“Romans” as a genetic category?!

Things get even worse, though. The terms “Minoan” and “ancient Greek” are both useless as genetic descriptors, but the term “Roman” is even more useless in this context because whether or not someone was considered “Roman” in antiquity depended whether they were a citizen of the Roman state and, from very early in Roman history, people of all variety of ethnic, geographic, and ancestral backgrounds could and did become Roman citizens.

Bret Devereaux, who is a Teaching Assistant Professor in the Department of History at North Carolina State University, recently wrote a post on his blog about how, from the very beginning, the Romans were a very diverse people. The Romans even conceived of themselves as a people with diverse origins in their mythology.

The most famous version of the founding myth of the Roman state comes from the Roman historian Titus Livius (lived c. 59 BCE – c. 17 CE), who explicitly says in his Ab Urbe Condita 1.8 that the earliest Romans who lived during the time of the very first king Romulus himself came from many different nations. He writes, as translated by Benjamin Oliver Foster:

“Next, lest his big City should be empty, Romulus resorted to a plan for increasing the inhabitants which had long been employed by the founders of cities, who gather about them an obscure and lowly multitude and pretend that the earth has raised up sons to them. In the place which is now enclosed, between the two groves as you go up the hill, he opened a sanctuary. Thither fled, from the surrounding peoples, a miscellaneous rabble, without distinction of bond or free, eager for new conditions; and these constituted the first advance in power towards that greatness at which Romulus aimed.”

As Devereaux discusses in his blog, the story Livy tells here is an ahistorical legend, but it reflects how the Romans thought about their own history. Moreover, archaeological evidence confirms that people of diverse nationalities from different parts of the Mediterranean world were present in the city of Rome from very early on. Thus, even if you are only thinking about the very earliest Romans in Latium, it’s still not accurate to think of them as a homogeneous group, genetically or otherwise.

ABOVE: Sixteenth-century CE Italian fresco by one of the Carracci brothers in the Palazzo Magnani in Bologna, depicting people of all nations coming to the sanctuary established by Romulus to become the first Romans

When it comes to later Roman history, trying to speak of “Romans” as an ethnic category becomes downright comical. For most of the period lasting from the first century CE until the permanent loss of the western territories in fifth century CE, the territories of the Roman Empire stretched from northern England in the far north to near the northern border of what is now Sudan in the far south and from Portugal in the far west to the banks of the Euphrates River in the far east.

The Roman state frequently granted Roman citizenship to people who had been born into allied or conquered nations. Over the centuries, the number of people of foreign birth who had become Roman grew until, finally, on 11 July 212 CE, the Roman emperor Caracalla issued the Constitutio Antoniniana, which granted full Roman citizenship to all free people living throughout all territories controlled by the Roman Empire.

This means that, after this decree, every single free person in all of southern Europe, North Africa, Asia Minor, and the Levant was a Roman—from the newly freed British woman in Eboracum, to the Greek-speaking Syrian-born philosopher in Athens, to the Coptic-speaking Egyptian potter in Memphis, to the elderly Jewish woman working as a dream interpreter in Rome. To speak of someone in this world as being “genetically Roman” would be like speaking of someone today being “genetically a United States citizen.” It would be completely nonsensical.

ABOVE: Map from this Reddit page showing the sprawling territorial extent of the Roman Empire in 125 CE during the rule of the emperor Hadrian

Other useless categories

This brings us to terms like “Hellenic Romans,” “Roman Hispania,” and “Gallo-Romans,” which seem to be My True Ancestry’s feeble attempt to address the issue of diversity within the Roman Empire. These terms, however, are positively silly as genetic categories, especially because they are not distinguishable from the other categories that are included.

A “Hellenic Roman” would presumably just be a Greek person living in the Roman Empire, but this would mean that “Hellenic Roman” is a subclass of both “ancient Greek” and “Roman”—not a distinct class of its own. Likewise, a “Gallo-Roman” would presumably just be a Gaulish person living in the Roman Empire, but, once again, this would make “Gallo-Roman” a subclass of both “Gauls” and “Romans.”

The term “Byzantine Empire” is equally as useless as a genetic category as “Roman,” since the Byzantine Empire was simply the eastern part of the Roman Empire that survived after the empire lost control of most of its territories in western Europe in the fifth century CE. This empire ruled many dozens, if not hundreds, of diverse ethnic groups, each one with a distinct culture of its own, including Greeks, Jews, Armenians, Circassians, Georgians, Aromanians, Slavs, Albanians, and countless others.

Furthermore, we once again see the problem of overlap between ostensibly distinct categories. Citizens of the Byzantine Empire were considered “Roman” and, although many people of many different ethnic backgrounds lived in this empire and held citizenship, Greek people were the dominant ethnic group for most of the empire’s history. This means that, by including “Byzantine Empire” as a genetic category, My True Ancestry is once again double-listing Greeks and Romans.

It gets worse, though, because the Ottoman Empire, in turn, conquered most of the territories ruled by the Byzantine Empire and included basically all the ethnic groups that were included in the Byzantine Empire and many others. Thus, all the diverse ethnic groups that existed in both the Byzantine Empire and Ottoman Empire are double-listed in the pie chart again.

ABOVE: Map from Wikimedia Commons showing the Ottoman Empire at its greatest territorial extent in the year 1683 CE

The “Viking” problem

The sample pie chart from my True Ancestry that I’ve shown above does not include “Vikings,” but I know from looking at numerous other sample charts and watching YouTube videos in which people show their results that My True Ancestry has several different listings for “Vikings,” including “Vikings (Danish),” “Vikings (Swedish),” and “Vikings (Norwegian).”

I can understand why My True Ancestry lists “Vikings” as an ethnic group, since it is very common for ordinary people who are not historians to use the word Vikings to refer to medieval Norse people in general. Nonetheless, it kind of annoys me that they use the word this way, because, if you want to really be accurate, “Viking” is not an ethnic group at all, but rather an occupation.

The proper name for the people who lived in Scandinavia, Denmark, and Iceland during the Middle Ages and spoke the Old Norse language is, in fact, Norse. The word Viking comes from the Old Norse word víkingr, meaning “pirate” or “raider,” and refers specifically to the Norse people who went out raiding. The vast majority of Norse people who lived during the Middle Ages were not Vikings, but peasant farmers.

ABOVE: Detail from Wikimedia Commons of a depiction of a Norse ship from the Tjängvide image stone in Sweden, dating to the eighth century CE

An interesting side note about my own Norse ancestry

On an interesting side note, my mother has been doing genealogy work for the past twenty years at least and she has traced a few of her ancestral lines back surprisingly far. As a result of this, I know that, through my mother, I actually have documented ancestors with names who are recorded to have lived in Scandinavia in the 1300s CE, which is a little after the Viking Age, but close enough to all but guarantee that I had ancestors who lived there during the Viking Age itself.

I don’t have any evidence that any of my ancestors ever went on raids, so I can’t honestly claim that I have “Viking ancestors,” but I definitely have some medieval Norse ancestors. This, of course, does not make me special or unique in any way; there are millions of other people who have medieval Norse ancestors as well. Nonetheless, I thought I would share this because it’s an example of how it is possible in some cases to trace one’s ancestry back very far using historical records.

ABOVE: Photograph my father took of me standing in front of one of the bookcases in my bedroom on 21 June 2021

Conclusion

Any company that claims it can tell you which cultures your ancient ancestors from over a thousand years ago lived in based on any kind of DNA test is using entirely untrustworthy methods of genetic and historical analysis. There is simply too much complexity involved and too little data available for anyone—even a genius scientist—to be able to track such information reliably.

If you know that you have ancestors who came from a particular place hundreds of years ago, there’s a good chance that you have at least some ancestors who were living somewhere in that general area over a thousand years ago. Thus, if you know you have ancestors who originally came from, say, Greece, there’s a pretty good chance that you have at least some ancestors who were living somewhere in the vicinity of Greece in antiquity.

That’s really the most we can really say with reliability. No one can show using DNA that your ancestors were Spartans or Roman legionaries or Celtic warriors or whatever.