The email arrived on a Tuesday, three days after Ancestry.com pushed its updated Turkish regional breakdown into the app. The subject line had one word: traitor. The body was blank. The customer service inbox in Ankara had already received four others that morning, all from Turkish users, all variations on the same theme.
The breakdown itself was not news to anyone who had been paying attention to population genetics. It said what the peer-reviewed literature had been saying for over a decade: most of the ancestry in the people who call themselves Turks today comes from ancient Anatolian, Balkan, and Near Eastern populations. The Central Asian Turkic contribution is real, but it is a minority component. Estimates cluster between roughly 5 and 15 percent. One recent model put the mean at 13.7 percent. A 2026 integrative review described the pattern with clinical restraint: “cultural Turkification was accompanied by limited demographic replacement.” Another analysis called the Central Asian signal an “Altaic echo” rather than a structure-forming element.
In plain language: the people of Anatolia changed their language and their religion. They did not, by and large, change their bodies.
What the algorithm also surfaced, because algorithms do not care about diplomatic sensitivities, was the proximity. Western Anatolian Turks sat close to Aegean Greek populations on the principal component plots. The distance between a man in Izmir and a man in Lesbos, measured in shared centimorgans, was smaller than the distance between that same man in Izmir and a Kazakh from the steppe.
The Turkish DNA Project, an online collective that had spent years curating its own database, did not dispute the numbers. It disputed the framing. In a tweet that was later deleted but screenshotted and circulated widely, the group accused AncestryDNA of prioritizing “to demonize the Turkish people and delegitimize [sic] their presence in Turkey rather than giving information about the genetic structure of the relevant population.” It called for a boycott. It labeled the Athens bureau chief of Greek City Times, who had reported on the results, a “white supremacist.”
A taxi driver in Ankara heard the number on the radio, yüzde on üç, thirteen percent, and turned to his passenger. “My grandfather never left Kayseri,” he said. “If he was a Turk, then who am I?” The passenger did not answer. He paid the fare and got out at Kızılay.
II
The question assumes that the answer to “who am I” should be found in a centrifuge. In Turkey, it has been found somewhere else for nearly a century, and the finding was not accidental.
In November 1928, a primary school teacher in Istanbul took down a wall chart of Arabic letters and replaced it with a chart of Latin ones. A boy in the third row raised his hand. “The things we wrote before,” he asked, “whose writing are they now?” The teacher did not answer. She told him to copy the letter A twenty times.
The alphabet reform was the first cut. The second came in 1932, at the First Turkish History Congress, held in the Dolmabahçe Palace. Delegates accepted what became known as the Turkish History Thesis: the proposition that the Turks were the creators of all ancient civilizations, including the Sumerians, the Hittites, the Egyptians, and the Greeks. A professor named Sadri Maksudi stood at the podium and announced, to applause, that the Sumerians and the Hittites had been Turks. The Sun-Language Theory followed, asserting that Turkish was the mother of all human speech.
These were not the ravings of fringe amateurs. They were state doctrine, taught in schools, printed in textbooks, and repeated by civil servants. Their function was not scientific. Their function was to give a new nation-state a pedigree that did not run through the Ottoman Empire or the caliphate. If the Turks had built the Hittite cities and spoken the first words of civilization, then the Republic of Turkey was not a parvenu carved out of a dead empire’s remains. It was the recovery of something original.
But the Thesis contained a structural problem that no amount of philological enthusiasm could solve. If the Turks were the original Anatolians, then the Anatolians had never really been Greeks, Armenians, or Kurds in any meaningful sense. They had been Turks who had forgotten their Turkishness. The Thesis absorbed difference by declaring it to be a temporary forgetting. And the instrument of remembering was language.
Mustafa Kemal himself had said it plainly: Türkçe konuşan Türk’tür. Whoever speaks Turkish is a Turk. The purification campaigns of the 1930s drove Arabic and Persian vocabulary out of official usage and replaced it with neologisms built from Turkish roots. The goal was a language that a citizen could learn, speak, and read without reaching for a lexicon of Ottoman obligation. In this architecture, blood was not the load-bearing wall. The sentence was.
III
The sentence has a cost, and the cost is paid by those who speak a different one.
In the visiting room of Diyarbakır Prison in the late 1980s, a Kurdish man named Mehmet Takar sat across from his mother. The rules of the institution were clear: Kurdish was forbidden. His mother spoke no Turkish. For the duration of the visit, they could not communicate. “Everything was a gun for them to use against us,” Takar later said. “Family was a gun.”
A quarter-century later, the gun has been replaced by a staffing table. Kurdish has been an elective subject in Turkish public schools since 2012. In the 2025-2026 academic year, 157 government-employed Kurdish-language teachers remained on the payroll across the entire country, 118 teaching Kurmanji and 39 teaching Zazaki. Another 847 graduates of Kurdish language and literature departments were waiting for appointments. Sixty thousand students had selected Kurdish courses, a demand that the teachers’ union estimated would require at least 400 instructors. The Ministry of Education allocated 10 of its 20,000 new teaching positions to Kurdish in 2024, down from 50 the year before.
Fatma Şahin graduated from Mardin Artuklu University in 2020 with a degree in Kurdish language and literature. She had wanted to teach since she was a girl. For six years she worked other jobs, whatever was available, while her file sat in the appointment queue. “My biggest dream is to be employed as a Kurdish language teacher,” she said, “to go into classrooms and teach Kurdish to Kurdish children.”
The Ministry’s explanation is that demand is insufficient. School principals, according to the education union Eğitim Sen, tell families who come to register for Kurdish courses that there are no teachers available. “On the one hand, the ministry does not hire teachers,” the union’s co-chair said, “and on the other, it says no one is asking for Kurdish lessons.”
The numbers do not resolve into a single villain. They resolve into a structure. A state built on the premise that the sentence makes the citizen will always find it easier to hire a geography teacher than a Kurdish one. Not because the geography teacher is more qualified, but because geography does not raise the question of whether the sentence itself is large enough to hold everyone who lives inside the borders.
IV
The Ancestry.com report landed in this structure like a stone dropped into a pond that had been still for a very long time.
The fact of Greek ancestry was old news. Anatolian Greeks had been converting, assimilating, and intermarrying with Muslim neighbors for centuries before the population exchange of 1923 made the categories official and violent. What landed was the implication that the sentence might be a costume, something put on over a body that had been there all along.
For a nationalism built on blood, this would be fatal. But Turkish nationalism was not built on blood. It was built on the alphabet chart, the language congress, the classroom, the appointment list. The Ancestry report could say, with the full authority of autosomal sequencing, that the man from Kayseri shared more alleles with a man from Crete than with a man from Bukhara. The man from Kayseri could shrug. His Turkishness did not depend on Bukhara. It depended on the fact that he dreamed in Turkish, cursed in Turkish, and filled out his tax forms in Turkish.
What the report could not do was make the Kurdish mother in the prison visiting room suddenly comprehensible to the guard who terminated her visit. What it could not do was turn the 847 waiting graduates into 847 appointments. What it could not do was make the sentence large enough.
The old woman in the village outside Kayseri, the one whose sheep ignored her, had heard about the DNA test from her grandson, who had read about it on his phone. She did not ask what a centimorgan was. She asked whether the test could tell her why the sheep came when she called in Turkish and scattered when she called in Kurdish, or the other way around, or whichever way it was. Her grandson said he didn’t think the test covered that.
She laughed. “Then it’s a useless test,” she said, and went back to the sheep.
