Inside the khaganate: a civil war in which Kabar rebels were killed or driven into exile (9th century). Across its frontiers: Itil's minaret pulled down and its muezzins killed in reprisal (922); Byzantine towns attacked and "both man and woman" killed (940s). In its afterlife: Khazar studies suppressed in the USSR (1951), an antisemitic national myth (Gumilev), the "not real Jews" slur drawn from Koestler (1976), and the "Khazarian mafia" theory used to justify a war (2022).
CONNECTIONS · 737–969 · RELIGION · From Rabbinic Jewish (Talmudic and Geonic) → Khazar (Khazar khaganate)

A Khazar king chooses Judaism (c. 750) — and the myth that followed

Between a Caliphate that had just burned his capital and an empire that called its allies spiritual sons, a Khazar ruler took the religion of the Jews in his towns. The court converted; the steppe did not. A thousand years later the story was turned into a weapon against Jews who had never been Khazars.

In the eighth or ninth century the ruler of the Khazar steppe empire adopted Judaism, a faith without an emperor. His people did not follow. A millennium later the story became an antisemitic slur.

A page of brown parchment covered in lines of square Hebrew script, with a short line of different, runic-looking characters near the lower left.
The Kievan Letter, a Hebrew letter of recommendation written in Kiev in the tenth century for Yaʿaqov bar Ḥanukkah and signed by members of a Jewish community whose names are partly Turkic. Identified by Norman Golb in 1962 among the Cairo Genizah papers, it is the strongest evidence that Khazar converts practised Judaism outside the royal court.
Unknown scribe for the Jewish community of Kiev. Kievan Letter (Cairo Genizah, T-S 12.122), 10th century. Cambridge University Library. Public domain via Wikimedia Commons. · Public domain

Before: a steppe empire between two monotheisms

In the middle of the eighth century the lower Volga, the Don and the steppe north of the Caucasus belonged to the Khazars. They had come out of the Western Türk empire, whose western wing they inherited in the decades after 630, and by 650 a Khazar khagan ruled from the Caspian to the Crimea 23. The khaganate was a federation more than a nation. Peter Golden, whose 1980 Khazar Studies remains the philological foundation of the field, describes a Turkic ruling clan presiding over Bulghars, Alans, Magyars, Burtas, Finno-Ugric forest peoples, Slavic tributaries on the Dnieper and the Oka, and the Iranian-speaking towns of the Caucasus foothills 23. Archaeologists call the material world of this federation the Saltovo-Mayaki culture: fortified hill towns of white limestone on the Don and the Donets, semi-dugout houses, wheel-made grey pottery, and cemeteries in which nomad and farmer were buried side by side 1524. Svetlana Pletnyova, whose 1976 synthesis is still the standard Russian introduction, describes a society in slow transition from seasonal nomadism to settled farming, with winter camps hardening into villages along the rivers 24.

The khaganate lived on its position. It stood across the routes that linked the Caliphate's markets in Iraq and Khwarazm, the Byzantine ports of the Crimea, and the forest zone that sold furs, honey, wax and slaves. Its capital Itil, on the Volga delta, was a double city: the khagan's island palace on one bank, a merchants' town with bathhouses, markets and mosques on the other 9. The Khazars taxed what passed. A tenth part of goods moving along the river went to the ruler, according to the Arabic geographers, and the tolls financed an army that could meet the Caliphate in the field 29. The site of Itil itself has never been securely located. Excavations at Samosdelka in the Volga delta since 2003 were announced as its discovery, a claim most archaeologists no longer accept, and the capital of the khaganate remains known only from the texts 15. That absence is the first sign of a pattern this record will meet again: the Khazars are vivid in the writings of their neighbours and faint in their own ground.

It had met the Caliphate for a century. The Arab–Khazar wars ran in two long rounds, from 642 to 652 and from 722 to 737. At Balanjar in 652 the Arab commander ʿAbd al-Raḥmān ibn Rabīʿa was killed and his army destroyed; at Marj Ardabīl in December 730 a Khazar army under Barjik annihilated the army of al-Jarrāḥ ibn ʿAbdallāh and killed its general 223. In the years between, Khazar armies raided deep into Armenia and Azerbaijan, and Byzantium courted them as allies: in 627 a Khazar force joined the emperor Heraclius at the siege of Tiflis against the Sasanian Persians 23. In 737 the Umayyad prince Marwān ibn Muḥammad answered with the largest expedition the Caliphate ever sent north. He crossed the Caucasus, reached the Volga, and forced the khagan to sue for peace and accept Islam. The Arabic historians report the khagan's conversion and the two teachers sent to instruct him; none reports that it lasted 12. Within a generation Marwān had left for the civil war that ended the Umayyad dynasty, and the frontier settled at Derbent.

The religion of the khagan

What did the Khazars worship before any of the monotheisms reached them? The answer has to be pieced together from hostile outsiders and from comparison with other Türk peoples. The ruling clan worshipped Tengri, the sky god of the Türk inscriptions, and shared the steppe belief that the khagan's fortune, his qut, was a gift of heaven that held the realm together 2223. Boris Zhivkov, whose 2010 Bulgarian monograph appeared in English from Brill in 2015, reads the Khazar court as a sacral kingship of the old steppe type, in which the khagan was less a governor than a living talisman of the state 22. In 682 Bishop Israel of Caucasian Albania, sent to convert the Huns of the eastern Caucasus, a people closely tied to the Khazars, recorded their worship of a giant sky god and of sacred trees, and his mission ended with the trees cut down 1.

The Arabic writers describe the Khazar talisman in terms that sound alien even after the conversion. The khagan was seen in public rarely; subjects prostrated themselves at a distance; al-Iṣṭakhrī reports that a new khagan was half-strangled with a silk cord at his accession and asked how many years he wished to reign 223. Ibn Faḍlān, who crossed the steppe in 922, wrote that the khagan's reign was limited to forty years and that when he died his tomb was built beneath running water, so that no one could find it 9. These are reports from foreigners who never entered the palace, and they mix observation with hearsay. But they agree on the essential point, which Zhivkov and Golden both stress: the Khazar ruler's authority was religious before it was administrative, and any change of religion at court would touch the foundation of the state, not only the beliefs of the ruler 122. A conversion in this society could never be a private matter.

Jews on the northern shore

The religion that would reach the court was already on the steppe's edge. Jewish communities on the north coast of the Black Sea are older than the Khazars by six centuries. At Panticapaeum, the modern Kerch, Greek inscriptions of the first century CE record the freeing of slaves in the synagogue, under the supervision of the Jewish congregation, and similar texts survive from Gorgippia and Phanagoria across the strait 21. By the seventh century the Bosporan towns had passed under Khazar control, and their Jews with them.

Byzantine policy then drove more Jews north. The emperor Heraclius ordered Jews baptised in the 630s, and in 721–722 Leo III decreed the forced baptism of the Jews of the empire; the chronicler Theophanes records the decree and the resistance it provoked 221. The Schechter Letter, written by a Khazar Jew in the tenth century, remembers a migration from the south, from Armenia and Persia, of Jews who settled among the Khazars, married into them and nearly forgot their religion 16. Al-Masʿūdī would later describe another wave fleeing Romanos I in the 930s 9. Kevin Alan Brook's survey traces these movements and concludes that by the eighth century Jewish merchants, craftsmen and refugees were established in the khaganate's towns, from the Crimea to Itil 21. They were rabbinic Jews of the Talmudic world, connected to Babylonia, Byzantium and Iran. They were the channel through which the court would encounter Judaism.

What the steppe held around 740

On the eve of the transmission the khaganate's religious order had five features, each of which the later record preserves:

  • A sacral kingship in which the khagan's fortune, not his administration, was the foundation of the state 2223.
  • The worship of Tengri and of local spirits of mountain, tree and water, practised by the ruling clan and much of the steppe population 23.
  • Muslim, Christian and Jewish minorities already present in the towns: Muslim merchants from Khwarazm and Iraq, Christian Alans and Goths, and Jewish traders and refugees from Byzantium and Iran 121.
  • A diplomatic marriage system that tied the khagans to Byzantium: Justinian II married a khagan's sister around 704, and in 732 the khagan's daughter Tzitzak married the future Constantine V, giving the Byzantine throne an emperor, Leo IV, known as "the Khazar" 223.
  • A recent humiliation: the forced acceptance of Islam in 737, which the khagan could not keep without becoming the Caliph's client 1.

That last point is the hinge. A Khazar ruler who stayed pagan remained an outsider to both empires, a barbarian to be bought or converted. A ruler who became Muslim accepted a religious authority seated in Damascus, and later Baghdad, whose armies had just burned his capital. A ruler who became Christian accepted one seated in Constantinople, whose emperors addressed even their allies as spiritual sons. Golden, reviewing the explanations offered since the eighteenth century, finds that the most persistent is also the simplest: Judaism allowed the Khazar ruler to join the monotheist world without submitting to either of its empires 1. Dunlop had made the same point in 1954, noting that the Khazar court sat precisely where Byzantine and Islamic diplomacy met 2. There was a third book religion in the khaganate's own towns. Its scholars lived in Babylonia and Byzantium, and they commanded no armies.

The transmission: a king, a cave, and the religion of Abraham

The Khazar conversion is known from three kinds of source, and they do not agree. Each has to be read for what it can carry.

The Hebrew sources

The fullest account is Hebrew. Around the 950s Ḥasdai ibn Shaprūṭ, a Jewish physician and minister at the court of the Umayyad caliph ʿAbd al-Raḥmān III in Córdoba, heard from merchants and envoys that a Jewish kingdom existed in the east and wrote to its king, asking about its people, its borders, its armies and whether it knew the date of the Messiah's coming. A reply survives in the name of Joseph, king of the Khazars, in a short version first printed in Constantinople in the sixteenth century and a long version preserved in a manuscript collection in Saint Petersburg 24. Pavel Kokovtsov's 1932 Russian edition of the correspondence remains the critical text, and his commentary leaned, with caution, towards accepting most of it as genuine 4. Dunlop, comparing the two versions, judged that the reply as we have it had been edited but rested on authentic Khazar information, especially in its geography of the Volga and its list of subject peoples 2. Joseph's own description of his country names the river, its three towns, and the peoples who paid him tribute 11.

In Joseph's Reply, an ancestor named Bulan drove out sorcerers and idolaters, was visited by an angel, and chose the religion of Abraham after a debate among a Jew, a Christian and a Muslim, each of whom conceded that Judaism was older than his own faith. Bulan built a tabernacle on the model of Moses' tent, with gold and silver taken in war. Generations later a king called Obadiah renewed the kingdom, brought rabbinic scholars from abroad, rewarded them, and built synagogues and schools; under him the Khazars learned the Bible, the Mishnah and the Talmud 211. Joseph then lists Obadiah's successors, a line of kings with Hebrew names (Hezekiah, Manasseh, Hanukkah, Isaac, Zebulun, Menahem, Benjamin, Aaron) ending with himself 411. The list is the backbone of every dating argument, because counting its generations backwards from Joseph gives a reform somewhere in the early ninth century and a first conversion before it.

A second Hebrew text tells a different story. In 1912 Solomon Schechter published a fragment from the Cairo Genizah, now Cambridge University Library T-S Misc. 35.38, written by a Khazar Jew of King Joseph's day, probably for Ḥasdai's circle 16. Its first page is lost. In this letter Jews had fled to the Khazars from Armenia and Persia generations earlier, married into the population, and lost most of their religion, keeping only circumcision and the Sabbath. One of them, a war leader, was made chief after a victory; his wife Serah and her father persuaded him to return to Judaism; when the neighbouring kingdoms protested, a disputation among Israelite, Greek and Arab sages was held; and books of the Law were found in a cave. The letter names the convert-king Sabriel 316. Zuckerman regards it as the older and more reliable of the two Hebrew accounts and builds his chronology on it 5. In this version the conversion is less a revelation than a recovery, and its hero is a soldier.

The Arabic and Greek witnesses

Outside observers confirm the result without describing the event. Ibn Rusta, writing in the first decade of the tenth century from older reports, says plainly: "Their supreme ruler is Jewish, and so is the īshā and those commanding officers and important men who support him. The rest of them follow a religion like the religion of the Turks" 9. Al-Iṣṭakhrī, around 950, describes a population of Muslims, Christians, Jews and idolaters in which the Jews were the smallest group and the king and his entourage were Jews 2. Al-Masʿūdī, writing around 943, dates the conversion of the Khazar king to the caliphate of Hārūn al-Rashīd (786–809) and reports that Jews fled to the Khazars from the Byzantine empire when the emperor Romanos I Lekapenos tried to force them to accept baptism 29. Each writer depends partly on the others, a point Stampfer would press, but Ibn Rusta's notice draws on a source of the later ninth century, and Ibn Faḍlān's of 922 was gathered by an envoy on the middle Volga, a few weeks' travel from Itil 79.

The Greek side gives one dated encounter. The Life of Constantine, the Slavonic biography of the missionary later called Cyril, describes a Byzantine embassy to the khagan in 860–861 in answer to a Khazar request for a learned man to debate with Jews and Muslims at court. Constantine studied Hebrew in Cherson on the way, and the biography sets out his arguments against his Jewish and Muslim opponents at length. His biographer claims a victory in the debate and two hundred baptisms; the khagan himself did not convert 5. Zuckerman reads the episode as the last moment before the conversion: a court that could still invite a Christian disputant had not yet bound itself to Judaism, and the debate in Joseph's Reply may preserve its memory 5. Historians who date the conversion earlier, Golden among them, read it instead as a court whose leadership was already Jewish 1.

The witnesses at a glance

The evidence for the conversion comes from nine principal sources in five languages. None is Khazar in origin except the Hebrew letters, and none describes the event as an eyewitness.

Source Date Language What it says
Ibn Rusta c. 903–913 Arabic Ruler, īshā and leading men Jewish; the rest follow a Turkic religion 9
Life of Constantine after 869 Old Church Slavonic Disputation at the khagan's court in 860–861; the khagan not baptised 5
Kievan Letter c. 930? Hebrew, with Turkic runes A Jewish community in Kiev with Turkic names 3
al-Masʿūdī c. 943 Arabic Conversion under Hārūn al-Rashīd; seven judges; refugees from Romanos I 9
Ibn Faḍlān 922 (reports gathered on the Volga) Arabic Jewish king; reprisal against the Itil mosque 9
al-Iṣṭakhrī c. 950 Arabic Jews the smallest group; the king and his entourage Jewish 2
Joseph's Reply c. 950s Hebrew Bulan's conversion; Obadiah's reform; list of kings 411
Schechter Letter 10th century Hebrew Sabriel; Jews from Armenia; Pesah's campaigns 16
Judah Halevi, Kuzari c. 1140 Judaeo-Arabic Literary retelling: the king's dream and the debate of the faiths 10

Who carried it

The sources name few individuals, and several names may refer to the same person. The table sets out what each tradition claims and where it comes from.

Name Date (as claimed) Role recorded Source
Bulan 8th or 9th century Expels sorcerers, chooses Judaism after a three-way debate Joseph's Reply 411
Sabriel (perhaps the same man) Before c. 900 Jewish war leader made chief; converts at his wife Serah's urging Schechter Letter 16
Obadiah Early 9th century? Brings rabbinic scholars, builds synagogues and schools Long version of Joseph's Reply 24
Constantine (Cyril) 860–861 Byzantine missionary; debates Jews and Muslims at the khagan's court Life of Constantine 5
Joseph c. 950s Last well-attested Jewish king; writes to Ḥasdai ibn Shaprūṭ Khazar Correspondence 4
Pesah 940s Khazar general who defeats the Rus leader HLGW Schechter Letter 16

Dating the conversion

No date in this story is secure. Five positions are argued in the current literature, and each rests on different evidence.

Proposed date Principal advocate Basis
c. 740 Dunlop (1954), following Halevi The Kuzari's chronology and Joseph's generations of kings 210
786–809 al-Masʿūdī Reign of Hārūn al-Rashīd 9
c. 837/8 Kovalev (2005), Olsson (2013) Coins struck in the name of Moses 68
c. 861 Zuckerman (1995) After Constantine's mission, which found the khagan not yet Jewish 5
Never Stampfer (2013) The internal sources are late or unreliable; the reliable ones are silent 7

Peter Golden's 2007 survey argues for a conversion in stages: an initial adoption in the mid-eighth century, after the humiliation of 737, and a rabbinic reform that made Judaism the court's settled religion by the early ninth 1. Constantine Zuckerman, from a close study of the Schechter Letter and the earliest quotations of Joseph's Reply, rejects the two-stage model. In his reconstruction, "the official introduction of Judaism in the Khazar state can be dated ca 861, very soon after Constantine's mission", in a single act by the commander Bulan, and "the conversion to Judaism became the occasion for a major constitutional reform of the Khazar state" 5. J. T. Olsson, writing in the Journal of the Royal Asiatic Society in 2013, reads the same events as a coup: a Jewish general seizes effective power in the crisis of the 830s, when Magyar attacks, Byzantine pressure and famine had weakened the khagan, signals it on the coinage, and completes the formal Judaization at his coronation around 861 6. The three historians disagree about the date. They agree that the conversion was an act of the court, made for reasons of state, and that its leading figure was a military man.

The case that it never happened

In 2013 Shaul Stampfer of the Hebrew University published the most radical reading, a seventy-page article in Jewish Social Studies. He argued that Joseph's Reply and the Schechter Letter are later literary creations or of doubtful origin, that the Arabic notices depend on one another, and that the most reliable contemporary witnesses say nothing. Sallām the Interpreter, sent north by the caliph al-Wāthiq in 842 in search of Alexander's wall, does not mention Jewish Khazars; nor does a letter of the Patriarch Nicholas Mystikos of around 914, written when the patriarch had every reason to note an infidel king 7. Stampfer also noted the near-total absence of Jewish material in Khazar archaeology. His conclusion was blunt: "There never was a conversion of a Khazar king or of the Khazar elite" 7. In his reading the conversion was a legend, given literary form in Hebrew and repeated by Arabic geographers who copied one another.

Most specialists have not followed him. The 837/8 dirhams, Ibn Rusta's independent notice, Ibn Faḍlān's report of 922 and the Kievan Letter, discussed below, are hard to explain away together 168. But Stampfer's challenge changed the terms of the argument. It forced historians to separate what the sources can bear from what centuries of retelling had added. What can be defended is narrow: by the early tenth century the khagan, the king and much of the court professed Judaism, and outside observers from Baghdad to Constantinople knew it. How deep it went below the court is a separate question, and the answer from every kind of evidence is: not very. The rest of this record works within that narrow claim.

What changed and what was replaced

A diarchy: the khagan withdraws, the king rules

The most visible change was constitutional. Tenth-century observers describe two rulers. The khagan, of the old charismatic clan, lived secluded and was shown to the people rarely; the bek or king, whom the Arabic writers call by titles such as khāqān bih, commanded the army, appointed governors and made war and peace 29. Ibn Faḍlān describes the king entering the khagan's presence barefoot, bowing, carrying a stick of wood that he lit before speaking, and he reports that the great khagan kept twenty-five wives, each the daughter of a neighbouring ruler, taken by consent or by force 9. The secluded khagan of the 920s still held the sacral charge of the old steppe order, even though the king who governed in his name was a Jew. Whether this division of power came before or because of the conversion is the centre of the dating debate. For Zuckerman and Olsson the conversion was the moment the military commander took the throne in all but name, leaving the khagan the ritual role of the steppe kingship 56. Golden is more cautious and sees a diarchy that may have older Türk roots, comparable to arrangements among other steppe peoples 1. Either way, the result was an unusual arrangement: a Jewish court presiding over a sacral institution inherited from the worship of the sky.

The faith on the coins

The strongest material evidence came from Gotland. In 1999 the Spillings hoard, the largest Viking-age silver hoard ever found, yielded dirhams struck in 837/8 in imitation of Abbasid coins, but with a difference. Where the Islamic formula read "Muhammad is the messenger of God", a small group of these coins reads "Moses is the messenger of God" 8. Roman Kovalev, who published them in 2005, argued that this "special issue" was a deliberate statement of identity by a Jewish ruler of the Khazars, minted for a trade that ran from the Caspian to the Baltic, and that it dates the presence of Judaism at the head of the state to 837/8 at the latest 8. The coins show the transmission's character better than any chronicle. They kept the visual language of the Caliphate, the Arabic script, the layout, the weight standard, and replaced only its prophet. Judaism here was a declaration of independence in the idiom of the power it declined to join, struck in silver that would travel to Scandinavia in the purses of Rus traders.

A small, worn silver coin with concentric bands of Arabic inscription, shown beside its museum label.
A Khazar "Moses" dirham of 837/8 from the Spillings hoard, Gotland. Struck in imitation of Abbasid silver, it replaces the Islamic formula "Muhammad is the messenger of God" with "Moses is the messenger of God" — the earliest dated material sign of Judaism at the head of the Khazar state.
W.carter. Khazar "Moses" dirham, 837/8, from the Spillings hoard. Gotland Museum, Visby. CC BY-SA 4.0 via Wikimedia Commons. · CC BY-SA 4.0

Judges for every faith

Below the court the khaganate stayed plural, and its institutions were adapted to that fact. Al-Masʿūdī describes seven judges at Itil: two for the Muslims, judging by Islamic law; two for the Khazars, judging by the Torah; two for the Christians, judging by the Gospel; and one for the Slavs, the Rus and other pagans, judging by custom and reason 29. The royal guard, the Arsiyya, were Muslims from Khwarazm, who served on condition that they would not be sent against fellow Muslims 9. The arrangement was tested in 912–913, when a Rus fleet, allowed passage down the Volga in return for half its booty, raided the Muslim shores of the Caspian. On its return the Muslim guard demanded the right to avenge their co-religionists; the king, by al-Masʿūdī's account, could only warn the Rus, and the guard fell on them and destroyed most of the fleet 29. The Jewish king ruled through a Muslim army that obeyed him only within the limits of its own faith.

The result was a state whose rulers held one religion and whose army, merchants and subjects held others, which may have been part of its design. The main groups recorded in Itil around 900–950 were:

  • The Jewish court: the king, the khagan and the leading officials, according to Ibn Rusta and al-Masʿūdī 9.
  • Muslims: the Arsiyya guard, the merchant colonies, and by Ibn Faḍlān's time a congregational mosque with a minaret 9.
  • Christians: Alans, Goths of the Crimea and Greeks, served by a Byzantine metropolitanate of "Gothia" whose lists name bishoprics in Khazar territory 123.
  • Pagans: much of the steppe and forest population, including the Turkic commoners who followed "a religion like the religion of the Turks" 9.
  • Jewish incomers: refugees from Byzantine persecution, including those al-Masʿūdī places in the reign of Romanos I 29.

Archaeology's near-silence

The written sources describe a Jewish court. The ground has produced little to match it. Valery Flyorov, who has excavated Khazar fortresses since the 1980s, made the absence the subject of a book in 2011: the "cities" and "castles" of the khaganate show steppe and Alan building, steppe burial, and almost nothing that can be tied to Jewish cult 15. Burials remained pagan in form across the Saltovo-Mayaki zone; no Jewish cemetery of Khazar date has been securely identified 715. There are exceptions. A gravestone with a Hebrew epitaph and two seven-branched candelabra was found in 2005 near the Don delta, and menorah graffiti appear on a few bricks and vessels 1521. The difficulty is that each item could belong to resident Jewish merchants rather than converted Khazars. The finds that bear on the question can be listed briefly:

  • Fortresses at Sarkel, Mayaki and the Tsimlyansk sites: steppe and Alan construction in brick and limestone, with no building identified as a synagogue 1415.
  • Cemeteries across the Saltovo-Mayaki zone: catacomb tombs of the Alan population and pit graves and cremations of the steppe population, all in pagan form 1524.
  • The Perevolochny Erik gravestone, found in 2005 near the Don delta: a partly preserved Hebrew epitaph with two menorahs and a palm branch 15.
  • Menorah graffiti on a handful of bricks and vessels, of uncertain authorship 1521.
  • The dirhams of 837/8 with the Moses formula, found not in Khazaria but in a Gotland hoard 8.

Flyorov's reading, which most archaeologists share, is that Judaism remained a court religion that did not reach the burial customs of the population it ruled 15. The material record therefore supports the narrow claim and refutes the broad one: a Jewish ruling stratum, yes; a Jewish people of the steppe, no.

The Kievan Letter

One document shows converted Khazars in their own hand. In 1962 Norman Golb identified among the Cambridge Genizah papers a Hebrew letter written in Kiev, then a Khazar tributary town, on behalf of a man named Yaʿaqov bar Ḥanukkah, who had stood surety for a loan to his brother; when robbers killed the brother and took the money, the creditors held Yaʿaqov liable, and the community wrote to others asking help to pay the debt 3. The signatories include men with Turkic names alongside Hebrew ones, and the letter carries a short note in Turkic runic script, read by Omeljan Pritsak as an official's endorsement. Golb and Pritsak published it in 1982 as an original document of the Khazarian Jews of Kiev, arguing that its Turkic-named signatories were proselytes or their descendants 3. Its date, around 930 in their reading, is disputed, and so is Pritsak's reading of the runic note. But the letter remains the strongest evidence that some Khazars outside the court practised Judaism, kept Hebrew names beside Turkic ones, and belonged to a Jewish community that wrote to the wider Jewish world in Hebrew. Its runic note is often described as the only written word of the Khazar language to survive.

What the steppe kept

The conversion did not replace the khaganate's older religious order. It sat on top of it. The table sets out what changed and what stayed.

Domain Before c. 740 After the conversion
Ruler's religion Tengri and the sacral khagan Judaism at court; the khagan kept as a sacral figure 15
Legitimacy Heavenly fortune (qut) Fortune plus descent from a convert-king chosen by God 422
Law Steppe custom Plural courts: Torah, sharīʿa, Gospel, custom 9
Coinage Imported Abbasid dirhams Local imitations; a special issue naming Moses (837/8) 8
Burial Kurgan and flat-grave steppe rites Unchanged across the population 15
Diplomacy Marriages with Byzantium; war with the Caliphate Neutral monotheism between both 12

What the cost was

Most religious transmissions in this atlas carry a bill paid by the people who received them. The Khazar conversion is unusual. It was not imposed on the population, it produced no mass baptism or forced circumcision, and the khaganate's courts went on judging Muslims and Christians by their own laws. Nothing in the sources suggests that a Khazar pagan was compelled to become a Jew. The documented costs are of three kinds: a civil war that may have followed the reform, a cycle of religious reprisal across the frontiers, and a thousand-year afterlife in which the story was turned against Jews. The first two belong to the ninth and tenth centuries and are modest by the standards of this atlas. The third is not modest, and it is not over. It is also the cost most often misdescribed, because the people who pay it were never Khazars and the people who exploit it rarely know the sources. This section takes the three in order and then sets out what the record does not charge to the conversion, beginning with the war that ended Khazaria.

The Kabar revolt

The first cost falls inside the khaganate. Constantine VII Porphyrogennetos, writing his manual De Administrando Imperio around 950, devotes a chapter to the Kabars, "a race of the Khazars" who rebelled against their rulers. When a civil war broke out, he writes, "their first government prevailed", some of the rebels were killed, and the rest fled to the Magyars, with whom they settled as three tribes and to whom, the emperor adds, they taught the Khazar language 12. The emperor gives no date and no cause. Zuckerman, Olsson and others place the revolt in the 860s or a little earlier and connect it to the reform, reading the Kabars as the losers in a struggle over the new constitutional and religious order 56. Golden accepts the connection as possible but unproven 1. What is certain is the outcome: a defeated faction killed or driven out, and three tribes lost to the steppe neighbour who would later found the Hungarian kingdom on the middle Danube 1223. If the reform caused the war, the conversion's first victims were Khazars who refused it.

Reprisal across the frontiers

The second cost is the logic of the conversion itself. A ruler who chose a religion because it belonged to neither empire also made himself the protector of its believers inside both, and the protection was paid for in other people's lives. Two episodes are documented:

  • 922, Itil. Ibn Faḍlān reports that when the Khazar king learned that Muslims had destroyed a synagogue in Dār al-Bābūnaj, in the caliph's territory, he ordered the minaret of Itil's mosque pulled down and its muezzins killed. The king explained: "I would not have razed the mosque, were I not afraid that every synagogue in the territory of Islam would be razed!" 9.
  • The 940s, the Crimea. The Schechter Letter reports that in the reign of Joseph, after the emperor Romanos I persecuted the Jews of Byzantium, the Khazar general Pesah "marched against the cities of Romanus in fierce anger and smote both man and woman", taking several Byzantine towns, before turning on the Rus leader HLGW whom Romanos had bribed to attack the Khazars 16.

The persecution that provoked the second episode was real. Al-Masʿūdī records the flight of Byzantine Jews to the Khazars under Romanos, and the Hebrew sources agree 29. The pattern is not peculiar to the Khazars: medieval rulers routinely answered attacks on their co-religionists abroad with attacks on foreign co-religionists at home. But the conversion inserted the khaganate into it. A Jewish state could answer a synagogue with a minaret and a forced baptism with a raid, and the people who died were muezzins in Itil and townspeople in the Crimea who had no part in either decision. The sources give no numbers, and this record will not invent them.

What the conversion did not cause

The fall of Khazaria is often narrated as the end of the Jewish kingdom, and it was. It was not a consequence of the Judaism. In 965, according to the Russian Primary Chronicle, Sviatoslav of Kiev marched against the Khazars, defeated the khagan's army in the field and took Sarkel, which he renamed Belaya Vezha 13. The fortress had been built in the late 830s with Byzantine engineers under Petronas Kamateros, at the khagan's request 12. Ibn Ḥawqal, writing a few years later, reported that the Rus had left the Volga country so devastated that not a grape or a raisin remained 2. The causes were commercial and strategic: Kiev wanted the tolls of the Volga and the Don, and the khaganate's steppe allies had been drawn away by the Pechenegs. The atlas records the destruction as the fate of the receiving culture, as it records the fall of Champa, and does not charge it to the transmission. Khazar Jewish communities survived for a time in the Crimea and the Caucasus; Brook follows their traces into the eleventh and twelfth centuries, after which they disappear from the record as a distinct group 21.

A black-and-white aerial view of a rectangular walled enclosure with excavation trenches, set in flat open steppe beside a river.
Aerial photograph of the Khazar fortress of Sarkel on the lower Don during Mikhail Artamonov's excavations of the 1930s. Built in the late 830s with Byzantine engineers, it was taken by Sviatoslav of Kiev in 965 and has lain under the Tsimlyansk reservoir since 1952.
Unknown photographer. Aerial view of Sarkel during the excavations of M. I. Artamonov, 1930s; published in Artamonov, Хазарская крепость Саркел (1956). Public domain via Wikimedia Commons. · Public domain

The afterlife: a story turned into a weapon

The heaviest cost came after the Khazars were gone. It began well. Around 1140 the Andalusi poet and philosopher Judah Halevi, shortly before he left Spain for the land of Israel, made the conversion the frame of a dialogue in Arabic, the Kuzari. In its opening an angel tells the Khazar king in a dream: "Thy way of thinking is indeed pleasing to the Creator, but not thy way of acting" 10. The king questions a philosopher, a Christian and a Muslim, finds each of them pointing back to the Jews, and summons a rabbi, whose answers make up the book. Translated into Hebrew by Judah ibn Tibbon in 1167, the Kuzari became a classic of Jewish thought, and for centuries the Khazars stood in Jewish memory as proof that a gentile king, free to choose among the monotheisms, could choose Israel 210. For a dispersed people living under Christian and Muslim rulers, it was a consoling story.

In the twentieth century the story was put to other uses. The table sets out the main stages, each of which inflicted a measurable cost on scholarship, on Jews, or on both.

Year Event Cost
25 December 1951 Pravda publishes "On one erroneous conception" under the pseudonym "P. Ivanov", attacking Mikhail Artamonov for treating Khazaria as a civilizing power Khazar studies frozen in the USSR; Artamonov forced to recant in print 1418
1966–1993 Lev Gumilev recasts Khazaria as a parasitic Jewish elite exploiting the steppe and Rus An antisemitic national myth adopted by Russian nationalist circles 18
1976 Arthur Koestler's The Thirteenth Tribe argues that Ashkenazi Jews descend from Khazars A thesis built to defeat racial antisemitism becomes its argument that Jews are "not real Jews" 20
2013 Behar and colleagues test 1,774 genomes from 106 populations No sign of a Khazar origin for Ashkenazi Jews 17
2022 The "Khazarian mafia" theory circulates on Telegram to justify the invasion of Ukraine Conspiracy theory weaponised in an active war 19

Moscow, 1951

The first stage struck the historians. Mikhail Artamonov, who had excavated Sarkel in the 1930s, argued in his early work that the Khazars had been a power of the first rank, a shield for the eastern Slavs against the Caliphate. On 25 December 1951, in the last years of Stalin's campaign against "rootless cosmopolitans", Pravda published an attack on this "erroneous conception", signed by a "P. Ivanov" whom no one has ever identified 18. The article denied the Khazars any civilizing role and cast the khaganate as a predator on the early Rus. Khazar studies stopped. When Artamonov's History of the Khazars finally appeared in 1962, it carried a conclusion stating that the Khazars had left no lasting mark on the Rus 14. Sarkel itself had been under the water of the Tsimlyansk reservoir since 1952, so the fortress whose excavation had made his name could no longer be studied at all. The cost here was paid in scholarship: a decade of silence and a monograph compelled to contradict its own evidence.

The second stage came from inside Soviet scholarship. Lev Gumilev, who had worked on Artamonov's excavations, turned the Khazar Jews into the villains of a theory of ethnic history in which a foreign merchant elite had parasitized the steppe and its neighbours, an alien graft on the body of the peoples it ruled. His popular books circulated widely in the late Soviet years. Michal Racyn, in a 2023 study of the two scholars, concludes that "Gumilev's radical revision of the history of Khazar Khaganate far exceeded previous Artamonov's findings and attracted the attention of Russian nationalist groups of the post-Stalinist era mainly due to its antisemitic narrative" 18. Racyn adds that Gumilev's most radical Khazar study was censored and appeared only posthumously, in 1993 18. The medieval court's Judaism had become, in this telling, the original sin of a Jewish conspiracy against Russia.

The Thirteenth Tribe and its readers

The third stage began with a well-meant book. Arthur Koestler, a Budapest-born Jew, argued in 1976 that Ashkenazi Jews descended mainly from Khazar converts, hoping that if Jews were not a "race" racial antisemitism would lose its object 20. Dunlop had already rejected any mass migration of Khazars into Europe, and Koestler's case rested on speculation 220. Its readers drew the opposite conclusion. If Ashkenazi Jews were Khazars, the argument ran, they had no history in the land of Israel and no claim to the name "Jew". The thesis was taken up by antisemites on the far right and by polemicists against Zionism, often by people who had read neither Koestler nor the sources. Population genetics has since tested the claim directly. Doron Behar and colleagues, assembling genome-wide data from 1,774 people in 106 populations, including 261 from 15 populations of the Caucasus and the region to its north, found that Ashkenazi Jews share most ancestry with other Jewish populations and with Europe and the Middle East, and no sign of a significant genetic contribution from the Khazar region 17. The genetic result does not settle what happened at the Khazar court. It settles what did not happen to the Jews of Europe.

"Khazarian mafia"

The final stage is current. Since the 2010s the claim that a cabal of "Khazarian" false Jews controls finance and government has spread through conspiracy networks, combining Koestler's thesis with older libels. In May 2022 the Institute for Strategic Dialogue documented its circulation on Telegram as a justification for Russia's invasion of Ukraine, where Ukraine itself was cast as a Khazar project 19. The researchers noted that the scarcity of sources on the Khazars and a contested hypothesis about Ashkenazi origins had made the theme easy to fill with fantasy 19. A disputed conversion of a steppe court in the ninth century has become a slur used in a twenty-first-century war. The people at whom it is aimed have no connection with the Khazars beyond the name of the religion their king once chose.

Weighing the bill

Cost When Evidence
Civil war; Kabar rebels killed or exiled to the Magyars 9th century, probably 860s De Administrando Imperio 39 12
Itil minaret destroyed and muezzins killed in reprisal 922 Ibn Faḍlān 9
Byzantine towns attacked, "both man and woman" killed 940s Schechter Letter 16
Khazar studies suppressed in the USSR 1951–1962 Racyn; Artamonov 1418
Antisemitic "Khazar" national myth 1966–1993 onward Racyn 18
"Not real Jews" argument from Koestler's thesis 1976 onward Koestler; Behar et al. 1720
"Khazarian mafia" used to justify a war 2022 ISD 19

The transmission's own bill was small: a civil war of uncertain cause and a cycle of reprisals that killed muezzins in Itil and townspeople in the Crimea. Measured by what the Khazars themselves paid, it is among the lightest in this atlas. The record's larger cost is what others did with the story after the Khazars could no longer speak for themselves: a Soviet campaign that silenced the historians, a national myth that cast Jews as an alien parasite, and a genealogical slur that denies Jews their own history. A court religion that never reached the graves of the people it ruled became, a thousand years later, an accusation against people who had never been Khazars at all.

What followed

Where this lives today

The Kuzari of Judah Halevi, a classic of Jewish philosophy Khazar titles and loanwords in Hungarian and Old East Slavic Contested claims of Khazar descent among Crimean Karaites, Kumyks and Karachay-Balkars The "Khazar" slur in modern antisemitic conspiracy theory

References

  1. Golden, Peter B. "The Conversion of the Khazars to Judaism." In Peter B. Golden, Haggai Ben-Shammai and András Róna-Tas, eds., The World of the Khazars: New Perspectives. Selected Papers from the Jerusalem 1999 International Khazar Colloquium. Handbook of Oriental Studies 8/17. Leiden: Brill, 2007, 123–162. en
  2. Dunlop, D. M. The History of the Jewish Khazars. Princeton: Princeton University Press, 1954. en
  3. Golb, Norman, and Omeljan Pritsak. Khazarian Hebrew Documents of the Tenth Century. Ithaca and London: Cornell University Press, 1982. Edition, translation and facsimile of the Kievan Letter (Cambridge University Library T-S 12.122) and the Schechter Letter (T-S Misc. 35.38). en primary
  4. Коковцов, П. К. Еврейско-хазарская переписка в X веке. Ленинград: Издательство АН СССР, 1932. Critical edition and Russian translation of the letter of Ḥasdai ibn Shaprūṭ and both versions of the Reply of King Joseph. ru primary
  5. Zuckerman, Constantine. "On the Date of the Khazars' Conversion to Judaism and the Chronology of the Kings of the Rus Oleg and Igor: A Study of the Anonymous Khazar Letter from the Genizah of Cairo." Revue des études byzantines 53 (1995): 237–270. en
  6. Olsson, J. T. "Coup d'état, Coronation and Conversion: Some Reflections on the Adoption of Judaism by the Khazar Khaganate." Journal of the Royal Asiatic Society, 3rd ser., 23, no. 4 (2013): 495–526. en
  7. Stampfer, Shaul. "Did the Khazars Convert to Judaism?" Jewish Social Studies, n.s., 19, no. 3 (2013): 1–72. en
  8. Kovalev, Roman K. "Creating Khazar Identity through Coins: The Special Issue Dirhams of 837/8." In Florin Curta, ed., East Central and Eastern Europe in the Early Middle Ages. Ann Arbor: University of Michigan Press, 2005, 220–253. en
  9. Ibn Faḍlān. Ibn Fadlan and the Land of Darkness: Arab Travellers in the Far North. Translated by Paul Lunde and Caroline Stone. London: Penguin Classics, 2012. Includes the Khazar passages of Ibn Faḍlān, Ibn Rusta and al-Masʿūdī. en primary
  10. Judah Halevi. Kitab al Khazari. Translated from the Arabic by Hartwig Hirschfeld. London: George Routledge and Sons; New York: E. P. Dutton, 1905. en primary
  11. Marcus, Jacob R. The Jew in the Medieval World: A Source Book, 315–1791. Cincinnati: Union of American Hebrew Congregations, 1938, 227–232: "The Conversion of the Khazars to Judaism" (the Reply of King Joseph). en primary
  12. Constantine Porphyrogenitus. De Administrando Imperio. Greek text edited by Gy. Moravcsik, English translation by R. J. H. Jenkins. New, revised edition. Washington, DC: Dumbarton Oaks Center for Byzantine Studies, 1967. Chapters 39 (the Kabars) and 42 (Sarkel). en primary
  13. Cross, Samuel Hazzard, and Olgerd P. Sherbowitz-Wetzor, trans. and eds. The Russian Primary Chronicle: Laurentian Text. Cambridge, MA: Mediaeval Academy of America, 1953. Entry for 6473 (965). en primary
  14. Артамонов, М. И. История хазар. Ленинград: Издательство Государственного Эрмитажа, 1962. ru
  15. Флёров, В. С. «Города» и «замки» Хазарского каганата: Археологическая реальность. Москва: Мосты культуры / Гешарим, 2011. ru
  16. Schechter, Solomon. "An Unknown Khazar Document." Jewish Quarterly Review, n.s., 3, no. 2 (1912): 181–219. Edition and translation of Cambridge University Library T-S Misc. 35.38. en primary
  17. Behar, Doron M., Mait Metspalu, Yael Baran, Naama M. Kopelman, Bayazit Yunusbayev, Ariella Gladstein, Shay Tzur, et al. "No Evidence from Genome-Wide Data of a Khazar Origin for the Ashkenazi Jews." Human Biology 85, no. 6 (2013): 859–900. en
  18. Racyn, Michal. "Khazar Khaganate in the Realms of Soviet Academia: From M. I. Artamonov to L. N. Gumilev and Russian Nationalists." JANUS.NET e-journal of International Relations 14, no. 1 (May–October 2023). en
  19. Institute for Strategic Dialogue. "An Antisemitic Conspiracy Theory Is Being Shared on Telegram to Justify Russia's Invasion of Ukraine." ISD Digital Dispatches, May 2022. en
  20. Koestler, Arthur. The Thirteenth Tribe: The Khazar Empire and Its Heritage. London: Hutchinson, 1976. en
  21. Brook, Kevin Alan. The Jews of Khazaria. 3rd ed. Lanham, MD: Rowman & Littlefield, 2018. en
  22. Zhivkov, Boris. Khazaria in the Ninth and Tenth Centuries. Translated by Daria Manova. East Central and Eastern Europe in the Middle Ages, 450–1450, vol. 30. Leiden: Brill, 2015. (Bulgarian original: Sofia, 2010.) en
  23. Golden, Peter B. Khazar Studies: An Historico-Philological Inquiry into the Origins of the Khazars. 2 vols. Budapest: Akadémiai Kiadó, 1980. en
  24. Плетнёва, С. А. Хазары. Москва: Наука, 1976. ru

Further reading

Cite this article
OsakaWire Atlas. 2026. "A Khazar king chooses Judaism (c. 750) — and the myth that followed" [Hidden Threads record]. https://osakawire.com/en/atlas/khazar_judaism_750/