Anatolia Didn't Fill With Turks; It Became Turkish

Ask where the Turks came from and the honest answer starts nowhere near Turkey. It starts on the Mongolian steppe, in the valley of a river called the Orkhon, where in 552 CE a chieftain named Bumin threw off his overlords and declared the first state ever to call itself Türk.¹ The oldest writing in any Turkic language is not in Anatolia and never was; it is carved into stone monuments that still stand in that same Orkhon valley, raised in the 720s and 730s for two Göktürk princes, Kül Tegin and Bilge Qaghan, and it took a Danish scholar, Vilhelm Thomsen, until 1893 to work out how to read it.² That is the answer to the first question, and it is not really in dispute. But it is the second question — how a peninsula that had spoken Greek for two thousand years, and had been one of the most densely Christian regions on earth, ended up mostly Turkish-speaking and mostly Muslim within four centuries — that is the interesting one, and the honest answer to it is not the one the first question sets you up to expect. Anatolia did not fill up with Turks. It became Turkish. Those are different claims, and the gap between them is the whole essay.

A people from the wrong end of Asia

Start with the steppe, because it matters that the Turks are not a Middle Eastern people who happen to speak an unusual language. Bumin's khaganate, and the rival and successor khaganates that followed it under his descendants, ruled a vast belt of grassland from Mongolia toward the Caspian, herding, raiding, and extracting tribute and trade from the Silk Road caravans that crossed their territory.¹ Where the Turkic peoples came from before that — whether an Altai homeland or somewhere further east — is still argued over by historical linguists, and the argument is not settled; Peter Golden's survey of the field treats the deeper origins as genuinely open.¹ What is not open to much doubt is the direction of travel afterward. Turkic-speaking groups moved west across the steppe over centuries, but the people who eventually ruled Anatolia did not walk out of the Altai and into the Aegean in one migration. They arrived in the Middle East first as someone else's soldiers.

The route runs through slavery, and it is worth saying that plainly because the word does real explanatory work here. From the ninth century the Abbasid caliphs in Baghdad began buying Turkic boys from the steppe and the Central Asian slave markets, training them as an elite guard, and garrisoning them at a purpose-built capital, Samarra, north of Baghdad.³ These ghilman — the term for the military slave-soldiers — were prized precisely because they were outsiders: horsemen and archers with no local tribal loyalties to complicate the caliph's chain of command, which made them useful right up until the moment they realized they held the caliph's chain of command in their own hands. Ibn Fadlan, sent north on a diplomatic and religious mission in the 920s, left an account of the Turkic and Volga peoples he met on that journey that is one of the earliest outsider descriptions we have of the steppe world the ghilman had come from — paraphrased here rather than quoted from a modern translation — and the account is a genuine and often-cited early witness.⁴ The point is not that Ibn Fadlan explains Anatolia; he doesn't. The point is that by the tenth century Turkic people were already known to the Arabic-writing world chiefly in two roles: exotic strangers on the far frontier, and armed men inside the palace.

Out of that same slave-soldier world came the first Turkic Muslim dynasty to rule serious territory, the Ghaznavids, founded by a former ghulam in Afghanistan. And out of the open steppe, not out of slavery, came the group that actually mattered for Anatolia: the Oghuz Turks, who had converted to Islam around the tenth century and organized themselves under the Seljuk family. The Seljuks beat the Ghaznavids decisively at Dandanaqan in 1040, and fifteen years later their leader, Tughril Beg, rode into Baghdad and had himself installed as the Abbasid caliph's protector — sultan in fact, with the caliph reduced to a religious figurehead who needed Tughril's soldiers more than Tughril needed his blessing.⁵ That is the pattern worth holding onto: prestige and power moving from the edge to the center, dynasty by dynasty, a full two centuries before any Turkic ruler set foot in Anatolia as a conqueror rather than a raider. The essay on the Arab conquest of the same broader region makes a parallel point about how slowly a ruling minority's religion becomes everyone else's; the Seljuk story is the same mechanism arriving from the opposite direction — see The Conquest Took Twenty Years; the Conversion Took Four Centuries.

Manzikert and what it actually opened

Anatolia's Turkish story conventionally begins on 26 August 1071, on a plain near Manzikert in the empire's eastern marches, where the Seljuk sultan Alp Arslan met the Byzantine emperor Romanos IV Diogenes in battle and won so completely that he took the emperor prisoner.⁶ It is a genuinely dramatic scene — an emperor of Rome led before a Turkish sultan's tent — and it has been treated ever since as the hinge on which Anatolia turned. That reputation is not wrong, but it is misleading if you take it to mean the Turks poured through a breach in the wall the day after the battle. Alp Arslan released Romanos on terms and went home; he had not come to conquer Anatolia, and Seljuk strategy at that date was mostly aimed south, at the Fatimids in Egypt, not west into Byzantine territory.⁶ What actually let Turkic warbands, Turkmen tribal groups largely outside any sultan's direct control, move into central Anatolia over the following decade was not the battle itself so much as what followed it: Romanos was overthrown and blinded by his own court within a year, and the empire tore itself apart in a string of civil wars in which rival Byzantine claimants repeatedly hired Turkmen mercenaries to fight other Byzantines.⁶ That is close to the modern consensus among historians of the period — the defeat's aftermath did more damage than the defeat.

By the time the dust cleared, a Seljuk relative named Suleiman had set up a rival Turkish state deep inside formerly Byzantine Asia Minor, and he gave it a name that is one of the strangest and most telling details in this whole story: the Sultanate of Rum. Rum is simply the Arabic and Persian word for Rome, because to everyone in the Islamic world Byzantium was Rome and its land was Roman land, and the new Turkish rulers of Anatolia did not rename the country after themselves. They named it after the empire they were displacing.⁷ A separate Turkish dynasty, the Danishmendids, carved out a rival principality in the east around the same period, and for roughly a century Anatolia was a patchwork of competing Christian Byzantine, Turkish Seljuk, Turkish Danishmendid, and Crusader-adjacent territories rather than a single front line with Greeks on one side and Turks on the other.⁷ That patchwork is the actual texture of the conquest period, and it looked far messier, and far more mixed, than the single battle that gets all the credit for starting it.

The demographic math does not work as invasion

Here is the flip, and it is the argument the essay is actually built to make. Anatolia in 1071 held several million people. It had been thoroughly Hellenized in language for well over a thousand years, going back through the Roman and Hellenistic periods to the Greek colonial cities founded on its coasts centuries before Alexander, and layered under that Greek surface were the older languages of the peninsula's own deep past — Hittite, Luwian, Phrygian, and the rest, tongues that had themselves eventually given way to Greek in the centuries after Alexander's conquests.³ By the time of Manzikert, Anatolia was also one of the oldest and most densely Christianized landscapes anywhere: it had sent bishops to the first ecumenical councils, its countryside was thick with monasteries, and cities like Nicaea and Ephesus were not provincial outposts of Christianity but among its founding addresses.

Set against that, the number of Turkic-speaking newcomers — Seljuk soldiers, Turkmen tribal groups moving their herds and families onto the Anatolian plateau over the following two centuries — was not remotely large enough, on any reasonable estimate, to outnumber the resident population, let alone replace it. Nomadic pastoralist economies simply do not carry the population density that settled agriculture does; a tribal group that could field a formidable cavalry force might still represent a demographic trickle against millions of Greek-speaking farmers, townspeople, and monks. And yet within roughly four hundred years — by the time an Ottoman sultan, the grandson many times over of those same steppe warbands, took Constantinople in 1453 — the peninsula had become majority Turkish-speaking and majority Muslim. Something replaced Anatolia's language and religion without replacing anywhere near most of its people. That something is not invasion. It is conversion, intermarriage, and institutional collapse, running for four centuries at a pace no single battle could have set.

Nomadic pastoralist economies simply do not carry the population density that settled agriculture does; a demographic trickle can still out-argue millions, given four centuries and a broken church.

The scholar whose name is most attached to explaining how that happened is Speros Vryonis, whose 1971 study of the decline of medieval Hellenism in Asia Minor remains the foundational account, and whose central argument is institutional rather than military.⁸ Byzantine Anatolia's Christian identity had run through a dense, functioning church infrastructure — parish priests, bishoprics, monasteries, endowed institutions that baptized, married, buried, and taught generation after generation in Greek. Vryonis's case, in outline rather than in his own words, is that Turkish conquest and settlement steadily broke that infrastructure: bishoprics went vacant or were extinguished as their tax base and protection disappeared, monasteries lost endowments or were converted to other uses, and whole regions that had once been serviced by a working church were left without one. Where the church stopped functioning, its former parishioners were not simply abandoned to atheism; they were left facing a competing religious and social order — Turkish Muslim rule, with its own tax incentives, its own paths to status, and its own marriage market — with nothing pulling the other way. Claude Cahen's earlier work on pre-Ottoman Turkey traces the same institutional unraveling from the political and administrative side, and more recent scholarship, notably A. C. S. Peacock's work on the Seljuk world, has filled in how mixed and locally negotiated that transition actually was on the ground, region by region, rather than a single uniform wave.⁹

Break the mechanism into its parts and none of them require mass replacement of people. Conversion carried immediate, practical incentives: Muslims under Turkish rule paid different taxes than Christians did, could hold positions Christians could not, and married into the emerging ruling class in ways that pulled their children toward the new language and faith within a generation or two, since a mother's tongue is not always the tongue a child ends up speaking when the surrounding institutions, schools, courts, and marketplaces run in the other one. Intermarriage between Turkish settlers and the existing population, especially in frontier zones where Turkmen tribes and Greek villagers lived in close and often cooperative proximity for decades before any final political settlement, blurred lines that a simple ethnic-replacement story assumes were sharp to begin with. Prestige did the rest of the work that force did not: Turkish became the language of the court, the army, and eventually most cities, and prestige languages spread through populations that never moved an inch, the way Latin spread through Gaul or French spread through the Norman English aristocracy without any population of Gauls or Anglo-Saxons being physically replaced.

This is, in other words, essentially the same diffusion story told in the essay on the Arab conquest of the wider Middle East — a small ruling and military minority whose language and faith spread through settled majorities over centuries via taxation, status, and intermarriage rather than the sword doing the whole job in the initial campaign. What makes Anatolia's case sharper, and worth pairing against that one, is the opposite linguistic outcome next door. Iran was conquered by Arab armies in the same broad era as the wider Islamic expansion, converted to Islam over a comparably long and gradual process — and kept its own language. Persian never gave way to Arabic the way Egyptian and Syrian speech eventually did, and the reasons scholars point to involve Persian's older prestige as an imperial language and its own robust literary tradition, which gave it institutional footing that could survive alongside a new faith.¹⁰ Anatolia took the opposite path: it kept something closer to its faith's demographic substance while losing its language almost entirely, because Greek's institutional support — the church network Vryonis describes — collapsed in a way Persian's comparable supports did not. Same mechanism, same rough timescale, and the linguistic outcome still went two different ways depending on what infrastructure was left standing to carry the old tongue.

What the genes and the grandchildren actually say

Modern population genetics has weighed in on this question over the last two decades, and it is worth reporting what it says carefully, because the finding is easy to state too precisely in either direction. Multiple genetic studies of present-day Anatolian populations have looked for a Central Asian genetic signature, the kind you would expect if the steppe migrations that brought the Turkish language had also brought enough people to swamp the existing gene pool. What they consistently find is that Central Asian ancestry is present but is a minority component of the modern Anatolian genome — the studies vary in how they define the comparison populations and in the precise proportions they report, so no number is asserted here, only that the finding across multiple independent studies points the same direction: modest, not dominant, and a live and still-developing area of research rather than a closed question.¹¹ That result lines up with the historical argument rather than contradicting it. It is the quiet, testable version of what Vryonis's institutional history implies in prose: most people living in Turkey today are, in the main, the descendants of Anatolia's older populations — Greek-speaking Christians and their Hittite, Luwian, and Phrygian ancestors further back still — who over four centuries adopted a new language, a new faith, and in time a new name for themselves. That is not a diminishment of Turkish identity. It is a more interesting fact about how identity actually spreads than the invasion story ever offered, and it deserves to be said as the finding it is rather than as a point scored against anyone.

Most people living in Turkey today are, in the main, the descendants of Anatolia's older populations, who over four centuries adopted a new language, a new faith, and a new name for themselves.

Carter Findley's history of the Turks across their whole span traces exactly this arc, from steppe confederation to world-spanning identity, and it is worth reading precisely because it refuses to compress that arc into a single migration story.¹² The long version runs: a khaganate on the Orkhon in 552, a slave-soldier corps at Samarra in the ninth century, a sultan in Baghdad by 1055, an emperor taken captive at Manzikert in 1071, a beylik under a minor frontier lord named Osman around 1299 or 1300 — one Turkish principality among several, not obviously destined to outlast the rest — and then, in 1453, that beylik's descendants breaching the walls of Constantinople and inheriting, in fact if not in blood, the city that had called itself Rome for eleven centuries. The essay on the battle that supposedly stopped the Arab advance into Europe makes a related point about how a single dramatic battle can absorb credit that belongs to a longer, less cinematic process; Constantinople, not Tours is the argument that the slow siege-and-defense of one city mattered more than the one battle everyone remembers. Manzikert is the Turkish mirror of that same trap: a real and important battle, wrongly asked to explain four centuries of language and identity change that it only began.

Rome's city, the steppe's grandchildren

None of this is a case for erasing the Turkish story into a Greek one, or for pretending the steppe migrations were incidental. They were not. Without Bumin's khaganate, without the Oghuz confederations, without Alp Arslan's army at Manzikert and the Turkmen warbands who followed the opening that battle's aftermath created, there is no Turkish Anatolia at all, and no Ottoman Constantinople in 1453. The language on the Orkhon stones really did travel, in the end, all the way to the Bosphorus, and that is a remarkable enough fact on its own that it does not need embellishing into a myth of replacement to be worth telling. What the evidence — textual, institutional, and now genetic — actually supports is a more interesting claim than either an invasion story or an erasure story: a small, prestigious, well-armed minority carried a language and a faith across a peninsula already full of people, and within four centuries that peninsula had made those things its own, largely without needing to become someone else's descendants to do it. The grandchildren of the steppe took Rome's last city. Most of the people who cheered them in, and most of the people who make up Turkey today, had never left.