The Avars of Europe first appear in the Byzantine sources in or around 558, when their envoys reached Constantinople.[1] Where their ruling core came from has been argued since the eighteenth century.[2] The khagans and the ruling nobility most likely came from the Rouran, the people whose khaganate the Göktürks destroyed between 552 and 555.[3][4][5] That theory comes first below. Four alternatives give the Rouran a smaller part or none: the pseudo-Avar theory drawn from Theophylact, the Hephthalite theory, the Ogur and Töles theory, and the Yueban theory. For the history of the Avars in Europe see Avars.

The evidence

The Chinese histories end the Rouran khaganate between 552 and 555 (see the Göktürk–Rouran Wars).[6][7] Neither the Bei Shi nor the Zhou Shu names a flight to the west. A survey of 2026 likewise finds no clear information on a link between the Avars and the Rouran in the Chinese sources.[8] The evidence for a migration comes from the Byzantine side.[3]

In 568 the Göktürk envoys at Constantinople said that some of the Avars still obeyed the Göktürks and that those who had fled numbered about twenty thousand.[9] In 576 Türk-shad called the Avars “our slaves who have fled their masters” and named them the Uarkhonitai.[10] The Göktürk ruler whom Menander calls Silzibul (see Istemi) is said to have declared, when he learned of the flight, that the Avars would not escape his swords and that he would attack them once the war against the Hephthalites was over.[11] Two further writers say that the Avars came as fugitives. Evagrius Scholasticus, who finished his Ecclesiastical History about 592, writes that they had fled in a body from the Göktürks, their neighbours, after ill treatment, and Theodore Syncellus wrote in 626 that the Avars under Baian had come to Europe as fugitives.[12]

The Rouran theory

Joseph de Guignes conjectured that the Avars of Baian descended from the Rouran, although Theophylact says otherwise.[2] A survey of 2026 calls him the first to link the two peoples.[13] The thesis became generally accepted, though critical voices were raised against it from the start.[2] The chronology seemed to support it, and so did the braided hair that was known of both peoples.[2] The Liang Shu describes the Rouran with braided hair, and the first Avar envoys at Constantinople wore theirs braided and tied with ribbons.[14][15]

From the Chinese side, Uchida Ginpū made the case for a common stock of the Rouran and the Avars most fully.[16] He identified the Avars of Theophylact who fled partly to Taugast and partly to the Mukri with Rouran who went to the Western Wei and to the Wuji of Manchuria, and he held that the Rouran destroyed in 555 were only one group of them.[17][18] Paul Pelliot wrote of the “Jouan-jouan ou Avar” of the Chinese records, and in a note he declined to say who the Avars of the Byzantine writers really were.[19]

Peter B. Golden gave the fullest recent argument. He starts from the statements of 568 and 576, which show the Avars of Europe as fugitives from the Göktürks.[3] The ruler of the European Avars bore the title khagan, which the Hephthalites did not use, and in Golden’s view this must have angered the Göktürks, to whom the title now belonged by right of conquest.[3] He concludes that these Avars were a Rouran remnant, or a core connected with the Rouran state, that fled the Göktürk conquest of 552 and the massacre of 555 and reached the Pontic steppe in the late 550s.[20] He does not exclude that Hephthalite elements joined them.[20] The khagans of Europe, he writes, are survivors of the Asian Avar ruling clan or close kin.[21]

The authors of a study of 2022 state in their historical background that at least the Avar khagans and their core group were descended from the Rouran, and that their genetic results support this view.[4] Batın Durmaz reaches the same result by a synthesis. In it the Avar elite is Rouran, its following largely Ogur, and the khagans, the title and the institutions come from the nobility.[5] Turkish authors who hold the identification include, in the list of a survey of 2026, Akdes Nimet Kurat, Bahaeddin Ögel, Özkan İzgi and Haşim Özel.[22] Ögel calls the ties of the European Avars with the Rouran state very probable.[23]

The theory has to explain how a Rouran remnant crossed Central Asia. The same survey asks how Rouran masses reached the Urals through Dzungaria, the Töles tribes at Tarbagatay and the lands of the Hephthalites.[24] It also counts the interval between the fall of the Rouran and the appearance of the Avar name in the Byzantine sources as a problem.[25]

Theophylact and the pseudo-Avars

Theophylact distinguishes two sets of Avars. The Avars whom the Göktürks defeated fled partly to Taugast and partly to the Mukri. The people on the Danube were Var and Chunni, Ogurs who took the Avar name for its prestige.[26] The Göktürks called the Avars of Europe Varchonites.[27]

Michael Whitby traces the distinction to a chronological error, since Theophylact placed the Göktürk victories under Maurice, and holds that his speculation about Avar origins is historically worthless.[28] Golden agrees and removes the pseudo-Avars from the analysis.[21] In his account Károly Czeglédy took the change of name for an old commonplace of the classical writers, with no connection to historical reality.[21] He held that the European Avars had full right to the names Avar and Uar-Hun, since they and the Rouran alike were of Uar-Hun origin.[21]

Walter Pohl answers that the claim of the Göktürks, that the Avars of Europe were not true Avars, was their official diplomatic position, as Menander shows, and that to dismiss Theophylact outright would be the wrong start.[2] He understands the pseudo-Avar line as political and not ethnic. On this view the Avars of Europe were a composite group that took the name Avars by decision, and the name Varchonites marks the mixture.[29] Quite likely the Rouran called themselves Avars, but they were not the only bearers of the name.[27] He doubts that the Rouran core, destroyed between 552 and 555, could have carried the migration.[30] He allows, however, that survivors of the ruling clan legitimized the title khagan for the new Avars, and holds it not unlikely that Baian could claim descent from that clan.[31]

Durmaz matches the two names of Theophylact to the two parts of the Avars. Var stands for those of Rouran origin and Chunni for Hephthalite remnants, to whom Ogurs were added on the road west. He concedes that this rests on no clear written source.[32]

The Hephthalite theory

A second theory looks to the Hephthalite realm, where the names Uar and Hun stand together. After Josef Marquart and René Grousset, Károly Czeglédy argued further that the Avars of Hungary, whom the Byzantines called Uarkhonitai, were the Uar-Hun group of the Hephthalites.[33] Hans Wilhelm Haussig placed the Avars among the Hua of the Chinese sources, together with the Hephthalites, and held that they had lived near the Caucasus for a century when they made contact with the Byzantines.[34]

The evidence for the Hephthalite connection is the occurrence of the names Uar and Var in the Hephthalite realm. Golden notes that the Hua, Uar and Var were at first subordinate to the Rouran, and that the Tongdian says the Hephthalites spoke a language different from that of the Rouran.[35] He does not take the Avars of Europe for Hephthalites, because of the title khagan, and allows only that Hephthalite elements joined them.[3] Pohl raises a chronological objection. The first Avar envoys reached Constantinople in or around 558, before the war of the Göktürks against the Hephthalites was over, so the fall of the Hephthalites cannot have set the Avars in motion.[1][36]

Durmaz joins the two theories. He holds that the Avar migration and the first war of the Hephthalites with the Göktürks fell close together, and that the Avars took in Hephthalite remnants who would not submit. He takes coins of the Kidarite type found in Poland and Hungary as a trace of this.[37] One survey of Avar history leaves open whether the ancestors of the Danubian Avars were the Rouran or a nation of the Hephthalite realm, and allows that a combination of both may cover what happened.[38]

The Ogur and Töles theories

A third theory derives the Avars from the Ogurs. Josef Marquart and C. A. Macartney traced them to the Ogurs of the Altai.[34] Pohl objects that Priscus mentions the Avars and the Ogurs as neighbours and enemies of one another.[34] The Soviet scholars Mikhail Artamonov and Lev Gumilev saw the main body of the Avars in Ogur tribes from the Altai whom the Göktürks drove out, and gave their name as Var and Hun.[39] Czeglédy, by contrast, separated the Ogurs from the Avars and identified them with the Tiele of the Chinese sources.[34]

İbrahim Kafesoğlu gives the Ogur theory in its fullest Turkish form. He argues that the Avars whom Priscus names in the fifth century and the Abar of Zacharias Rhetor cannot be the Rouran destroyed in 555.[40] He identifies Theophylact’s false Avars with the Warkhon, that is Var and Hun, who lived beside the Ogur tribes between western Turkestan and the northern Caucasus and around the Don and the Volga, and whom Menander calls Avars.[41] These were a union of Turkic tribes that had once belonged to the Rouran, took part in the founding of the Hephthalite state, and later came to the lands north of the Caspian and the Aral Sea.[42] Together with the Ogur tribes they met there, and with some Mongol and Alan elements, they formed the Turkic masses from which the founders and the ruling class of the Avar khaganate came.[42] A survey of 2026 counts his account in the third theory, the Ogur and Töles one.[43] Ögel identifies the pseudo-Avars as Turks living on the Volga, or tribes related to them.[44]

The Yueban and other theories

Yu Taishan criticised the identification of the Rouran with the Avars in a study of the 1980s that found little notice in the West. He argued that remnants of the Rouran could hardly have escaped to the west, where the Göktürks and other former enemies blocked the way, and he identified the Avars with the Yueban.[45] In his view Yueban is another transcription of Avar or Apar. The Yueban lived west of the Rouran, were defeated by them about 450, and some of them moved west, where Priscus took note of them.[34] Pohl answers that this does not explain why the Göktürks were so angry about the flight of the Avars a hundred years later.[34]

Haussig also suggested that Avar was a Sogdian designation for aliens such as the nomad peoples, and so had no ethnic meaning. Pohl objects that it cannot explain why the Sogdian envoys of the Göktürk khagans argued so forcefully that the Avars of Europe were not real Avars.[34] Other theories add Finno-Ugric and Hungarian elements to the Avars. Gyula László derived part of them from the region of the Kama and the White Idil, and Dezső Csallány and Gyula Györffy allowed Finno-Ugric and Hungarian elements among them.[46] Arnulf Kollautz saw a Turkic and Mongol mixture.[42] Pohl does not accept any single identification. He regards it as a widespread misunderstanding to equate the Avars of Europe with one people of Central Asia, since neither the Rouran, the Tiele and the Yueban nor Hephthalites or Ogurs can simply be set equal to the Avars encamped near the Caucasus in 558.[34]

Name, title and language

The name Avar is older than the migration. Priscus names Avars about 463 as the people whose pressure set the Sabirs in motion, and a Syriac compilation of about 568 lists the Abar beyond the Caspian Gates.[47] Golden identifies the Avars of Priscus with the Rouran.[48] Kafesoğlu draws the opposite conclusion from the same notices, since they lie before the end of the Rouran.[40] Golden holds that Rouran was a Chinese name and that the people called themselves Abar.[49] Pohl finds it quite likely that the Rouran called themselves Avars, though the name was not theirs alone.[27] Etymologies of the name rest on guesswork as long as the language is not known.[50] Mongolic abarga, “worm, snake”, has been linked to the Chinese variant Ruanruan, and a Turkic avar, “rebellious, disobedient”, has also been proposed.[51]

The title khagan bears directly on the question. Shelun took it about 400, the Rouran had it from the Xianbei, and the rulers of the Avars bore it in Europe.[52][53][54][55] Pelliot held in 1915 that the Göktürks inherited the title from the Rouran, whom he calls Avars.[19] For the origin of the word itself see the origin of the title Khan and Khan (Khagan). Kafesoğlu counts the Avar offices tudun, yugurus, tarkhan and boyar among the Turkic ones.[42] The name of the first khagan, Bayan, shows how the argument runs. Pelliot took it for the Mongolic form of the name, and Kafesoğlu for a Turkic word from bay, “rich”.[56][42] A survey of the linguistic evidence finds the name used among Bulghar Turkic and Mongolic speakers alike, and of Turkic origin in Mongolic and not the reverse.[57]

The main objection to the Rouran theory is linguistic.[57] If the Rouran spoke a Mongolic language and the Avars a Bulghar Turkic one, continuity would require a language shift somewhere on the road, and no Mongolic substratum has been shown in the language of the European Avars.[57] Golden cautions that little is known of the language of the European Avars and that language shifts happen in migrations. He adds that the language of the Rouran survives only in personal names and titles, and that some recent scholarship classes it as non-Altaic.[58] Two suggestions would make a language shift unnecessary. Juha Janhunen proposes that the Rouran spoke a form of Oghuric Turkic, and András Róna-Tas that the name Yujiulü renders Ugur.[59] Golden finds the second suggestion attractive philologically but holds that it needs more historical evidence than exists.[60]

Archaeology and genetics

The archaeology gives no clear answer. The early Avar finds show no homogeneous culture: components of eastern steppe origin stand beside others with parallels in the Caucasus, the south Russian steppes and western Asia, and beside local European traditions.[57] On the Asian side, a few finds that Mongolian archaeologists attribute to the Rouran are typologically comparable with Avar antiquities, but no integrated Rouran tradition has been identified.[61] Kafesoğlu rejects the statement that the skeletons of the Avar period were predominantly Mongoloid. He cites finds of Germanic, Slavic, Iranian and Finno-Ugric types and a notable share of the Turkic type.[62]

The genetic studies draw different conclusions from their data. A survey of the genetic data in 2020 found it neutral on the question and showed only that at least some of the Avars of Europe were of east Asian ancestry.[57] Studies of the maternal and paternal lines came first. According to the survey of 2026, Veronika Csáky and colleagues found East Eurasian mitochondrial lines in 64 per cent of 25 individuals from one cemetery in a preprint of 2018, and in 69.5 per cent of Avar samples of the eighth century in a study of 2020.[63] Endre Neparáczki and colleagues found ten East Asian Y-chromosome lines, nine of them N-M178 or its branches, and the survey reports that they inferred Mongolic speech for the Avars from the occurrence of that line among the Buryats.[64] Durmaz objects that N-M178 also occurs among the Xiongnu, the Göktürks and the Tuvans, and the same survey notes that the line occurs among Mongolic speakers too.[64] A haplogroup says nothing about the language of its carriers.

The studies of 2021 and after drew different conclusions from similar lines. The same survey reports that Christine Keyser and colleagues found the Xiongnu lines Q1a and N1a also among Avars, Huns and Magyars and traced these peoples to the Xiongnu, and that Zoltán Maróti and colleagues and Lehti Saag and Robert Staniuk derived a Rouran origin from Y-chromosome lines of the C2b clade. Ke Wang and colleagues are said to have found most of their Avar samples in the C2b clade of the one Rouran sample.[65]

The genome-wide study of 2022 by Guido Alberto Gnecchi-Ruscone and colleagues bears most directly on the Rouran theory. Genome-wide data from the richest Avar burials of the seventh century show north-east Asian ancestry that matches the earlier populations of the Mongolian steppe, and the Rouran-period genome in particular.[66] Many individuals of the early and middle elite fit that genome as a single source of ancestry, although the models have little statistical power because the genome is of low coverage.[67] The same elite also matches individuals of the Xiongnu and especially the Xianbei periods.[68] The authors take the result as support for the view that at least the khagans and their core group descended from the Rouran.[4] Durmaz interprets the whole body of data as a mixture: the N, R and Q lines show continuity with the Xiongnu, the Huns and later Turkic-speaking peoples, and the C2b lines show a Rouran part.[69]

The Apar of the Orkhon inscriptions

The Kül Tegin inscription names the Apar among the mourners at the funeral of the first Göktürk khagans. The traditional view connects the name with the Avars of Europe.[70] Sergei G. Klyashtorny associated it with the pseudo-Avars.[70] Erhan Aydın reads par and identifies it with Iran, and Uchida had already identified the name with Persia.[70][71] Golden finds Aydın’s suggestion not unreasonable, since the list of mourners runs from east to west, and it seems unlikely to him that Avars would send mourners to the funeral of rulers who had destroyed their state.[70]

Unsettled points

The written evidence fits the Rouran theory at three points: the claim of the Göktürks to the Avars as runaway subjects, the title khagan, and the date of the first Avar envoys.[3][1][72] The genetic data from the elite burials point the same way, within the limits of a single genome of low coverage.[67] Still open are the size of the Rouran part among the Avars, the language of the Rouran, and the place of the Hephthalite and Ogur groups.[31][57] The theories do not exclude one another. A Rouran core of khagans and nobles with a following of Ogur and other groups is a form in which the Rouran, the Ogur and the pseudo-Avar theories meet.

References

  1. ^ a b c Pohl 2018, p. 20, 42
  2. ^ a b c d e Pohl 2018, p. 41
  3. ^ a b c d e f Golden 2013, p. 62–63
  4. ^ a b c al. 2022, e6
  5. ^ a b Durmaz 2026, p. 59
  6. ^ Zhou Shu (Book of Zhou), juan 50
  7. ^ Bei Shi (History of the Northern Dynasties), juan 98
  8. ^ Durmaz 2026, p. 31
  9. ^ Menander Protector; Blockley 1985, fr. 10.1, p. 117
  10. ^ Menander Protector; Blockley 1985, fr. 19.1, p. 175
  11. ^ Menander Protector; Blockley 1985, fr. 4.2, p. 47
  12. ^ Pohl 2018, p. 37
  13. ^ Durmaz 2026, p. 33
  14. ^ Liang Shu, juan 54
  15. ^ Pohl 2018, p. 20
  16. ^ 内田吟風 1962, p. 76,86
  17. ^ 内田吟風 1962, p. 78
  18. ^ 内田吟風 1962, p. 82-85
  19. ^ a b Pelliot 1915, p. 688–689
  20. ^ a b Golden 2013, p. 63
  21. ^ a b c d Golden 2013, p. 65
  22. ^ Durmaz 2026, p. 33–34
  23. ^ Ögel 1984, p. 128
  24. ^ Durmaz 2026, p. 37
  25. ^ Durmaz 2026, p. 37, 44
  26. ^ Theophylact Simocatta; Whitby 1986, p. 188–190
  27. ^ a b c Pohl 2018, p. 45
  28. ^ Whitby 1988, p. 315, 317
  29. ^ Pohl 2018, p. 45–46
  30. ^ Pohl 2018, p. 36, 46
  31. ^ a b Pohl 2018, p. 46
  32. ^ Durmaz 2026, p. 37–38
  33. ^ Rásonyi 1993, p. 73
  34. ^ a b c d e f g h Pohl 2018, p. 42
  35. ^ Golden 2013, p. 60
  36. ^ Whitby 1986, p. 188–189
  37. ^ Durmaz 2026, p. 38
  38. ^ Sinor 1990, p. 207
  39. ^ Rásonyi 1993, p. 79
  40. ^ a b Kafesoğlu 1997, p. 161
  41. ^ Kafesoğlu 1997, p. 161–162
  42. ^ a b c d e Kafesoğlu 1997, p. 162
  43. ^ Durmaz 2026, p. 35
  44. ^ Ögel 1984, p. 116
  45. ^ Pohl 2018, p. 41–42
  46. ^ Rásonyi 1993, p. 78–79
  47. ^ Golden 2013, p. 49–50, 59
  48. ^ Golden 2013, p. 49–50
  49. ^ Golden 2013, p. 66
  50. ^ Golden 2013, p. 59
  51. ^ Pohl 2018, p. 40–41
  52. ^ Wei Shu, juan 103
  53. ^ Golden 2013, p. 54
  54. ^ Theophylact Simocatta; Whitby 1986, p. 189
  55. ^ Menander Protector; Blockley 1985, fr. 25.1, p. 219
  56. ^ Pelliot 1915, p. 689
  57. ^ a b c d e f Savelyev 2020
  58. ^ Golden 2013, p. 51–52
  59. ^ Golden 2013, p. 52, 55
  60. ^ Golden 2013, p. 55
  61. ^ Kradin 2005, p. 150
  62. ^ Kafesoğlu 1997, p. 163
  63. ^ Durmaz 2026, p. 41, 42
  64. ^ a b Durmaz 2026, p. 42
  65. ^ Durmaz 2026, p. 43
  66. ^ al. 2022, p. 1, 7
  67. ^ a b al. 2022, p. 5
  68. ^ al. 2022, p. 7
  69. ^ Durmaz 2026, p. 44
  70. ^ a b c d Golden 2013, p. 61
  71. ^ 内田吟風 1962, p. 87
  72. ^ Durmaz 2026, p. 58

Sources

Literature

  • Peter B. Golden, Some Notes on the Avars and Rouran (2013)
  • Walter Pohl, The Avars: A Steppe Empire in Central Europe, 567–822 (2018)
  • Guido Alberto Gnecchi-Ruscone et al., Ancient Genomes Reveal Origin and Rapid Trans-Eurasian Migration of 7th Century Avar Elites (2022)
  • Alexander Savelyev; Choongwon Jeong, Early Nomads of the Eastern Steppe and Their Tentative Connections in the West (2020)
  • Michael Whitby, Mary Whitby (trans.), The History of Theophylact Simocatta (1986)
  • Michael Whitby, The Emperor Maurice and His Historian (1988)
  • R. C. Blockley, The History of Menander the Guardsman: Introductory Essay, Text, Translation, and Historiographical Notes (1985)
  • Uchida Ginpū 内田吟風, 柔然アヴァール同族論に關する諸問題 (1962)
  • Paul Pelliot, L'origine de T'ou-kiue, nom chinois des Turcs (1915)
  • Batın Durmaz, Bozkırın Muamması: Avarların Köken Paradigması ve Yeni Bir Sentez Denemesi (2026)
  • Batın Durmaz, Avar Etnogenezinin Oluşumunda Akhunların Etkisine Arkeogenetik Üzerinden Bir Bakış (2026)
  • Nikolay N. Kradin, From Tribal Confederation to Empire: The Evolution of the Rouran Society (2005)
  • Denis Sinor (ed.), The Cambridge History of Early Inner Asia (1990)
  • İbrahim Kafesoğlu, Türk Millî Kültürü (1997)
  • László Rásonyi, Tarihte Türklük (1993)
  • Bahaeddin Ögel, İslâmiyetten Önce Türk Kültür Tarihi. Orta Asya Kaynak ve Buluntularına Göre (1984)

Related

What links here3

Peoples
Avars
Institutions and Titles
Khan (Khagan)
Controversies
Archaeogenetics of the Steppe Empires

Cite this article

Yavuz, Emre-E. “The origin of the Avars.” Turkic Atlas, revised 8 November 2026. https://turkicatlas.org/historiography/avar-origin-debate/

Last revised: 8 November 2026