An ilteber ruled his own people and answered to somebody else. The rank is the steppe answer to a problem every conquering power meets: what to do with a king you have beaten but do not intend to replace. The answer was to leave him where he was, give him a title in the victor’s system, and put a tax officer at his elbow.

What the rank did

The Xin Tang Shu describes the arrangement in one sentence. Tong Yabgu, having made himself master of the western states, conferred on all their rulers the title xielifa, and set a tudun over each to supervise them and collect the taxes.[1] Two offices, and the division between them is the whole design: the local king keeps his throne and his people, and a man of the khan’s keeps the accounts. Nothing is said about garrisons, and nothing needed to be. A king who has been left his kingdom has more to lose by revolt than a governor would.

The title is found in use, not only in a description of it. A Sogdian contract of 639 from the bazaar of Gaochang is dated by the year of the divine and great Ilteber-king of Gaochang, which means the Chinese king of Turfan bore the rank under the western Göktürk and dated his documents by it.[2] A title used to date a contract is a title that a notary and his customers both recognised.

In the Caucasus

The earliest dated attestation outside the Chinese records stands in Armenian. In the winter of 681 the prince of the Huns north of Derbent came down into Albania with generals and ensigns, halberdiers, archers and armoured cavalry; the History of the Caucasian Albanians calls him the general and great prince of the Huns, Alp Ilitʿuer.[3] The following winter an Albanian bishop was sent to his court, and the account of that embassy is the fullest picture any source gives of a people under Khazar overlordship.[4]

Alp is an epithet — brave — and the rest is the title. The identification was made in 1903, and it was made by setting three things side by side: the Armenian forms, the bare litʿber that Łewond gives for a commander out of the Khazar horde, and the official style of the Bulgar ruler in Ibn Faḍlān’s report, where the initial syllable had been struck out by copyists who mistook it for the Arabic article.[5] The English translator of the Armenian history follows that reading and prints the manuscript variants beneath it: Alpi Litʿuir, Alpiali Tuer, Alpʿilitʿwer, Alpʿ hoł tʿuel, Ilu-Tʿvel, and more — no two witnesses agree, which is what one expects of a foreign title in a script that had no use for it.[6]

The author of that identification added a caution that is usually left off when the identification is quoted. Having established that the Armenian ilutver and litʿber and the Arabic form are one word, he wrote that he left the decision whether this word has anything to do with Old Turkic äl-täbär to the Turkologists.[7] The chain from the Caucasus to the Volga he regarded as proved. The last link, to the steppe title, he left to others.

This work takes that last link as holding. The rank does in the Caucasus exactly what it does in the Tarim — a king left on his own throne under a khan who is not of his people — and a title that behaves the same way in both places, under powers that were in contact, is one title. The caution was the caution of a man who would not decide a question outside his field, and it should be read as that rather than as doubt about the identification.

The disputed equation

That last link is still argued, and the argument runs through the Chinese spelling.

On one side stands the equation 頡利發 xielifa = ilteber, which is the reading most handbooks print. Friedrich Hirth and F. W. K. Müller proposed that the Chinese kʿi-li-fa should be read il-di-bar; Wilhelm Barthold took sse-li-fa to be elteber; Faruk Sümer reports the equation as settled and analyses the word as il, realm, with a suffix; Ahmet Taşağıl gives ilteber as the Göktürk equivalent of xielifa; Kürşat Yıldırım equates the Rouran szu-li-fa with it in the same way.[8] On this reading the rank the Göktürk gave their subject kings and the rank the Caucasian Huns gave their prince are one rank, and the Chinese, Armenian, Arabic and Sogdian witnesses all name it.

On the other side the objection is phonetic and it is old. Paul Pelliot rejected the Hirth–Müller reading outright and held that kʿi-li-fa cannot yield ilteber.[8] The case has been pressed further: that the sounds simply do not match, that hsie- is a sibilant and the remainder gives -l- and -b-, and that the Chinese transcription therefore renders a Turkic silbäg rather than ilteber — a word that would belong instead with Zilgibis, the name East Roman writers record among the Caucasian Huns in 521, and with Selifān among the Khazars of Dagestan in 916 and 943.[9] Peter B. Golden has separately resisted joining Selifān to the Chinese sse-li-fa and calls the content of the title obscure, with Caucasian or Iranian connections not excluded.[10]

The dispute matters for what may be said about the rank. If the equation holds, the Chinese notices of xielifa among the Rouran, the Göktürk and the Uyghurs are all notices of ilteber, and the rank has a documented history of three centuries. If it does not, the Chinese notices belong to a different title, and what is left of ilteber is the Sogdian contract, the Armenian history and the Arabic report — enough to establish the rank, not enough to trace it.

In this work

This work uses Ilteber for the rank and records the Chinese equation as disputed rather than as established. Where an article says that a ruler bore the title, it says so on the strength of a source that names him, not on the strength of the equation. The article on the title of khan carries the ranks together, and Beg carries the rank that is native to the tribe rather than granted from above — the two are easy to confuse, and the difference is the direction the authority comes from.

References

  1. ^ Xin Tangshu (New Book of Tang), juan 215
  2. ^ de la Vaissière 2005a
  3. ^ Movsēs Dasxurancʿi, juan 2
  4. ^ Movsēs Dasxurancʿi, juan 2
  5. ^ Marquart 1903, p. 114
  6. ^ Dowsett 1961, p. 150
  7. ^ Marquart 1903, p. 115
  8. ^ a b Gedikli 2022, p. 444
  9. ^ Gedikli 2022, p. 448
  10. ^ Gedikli 2022, p. 442

Sources

Literature

  • Étienne de la Vaissière, Sogdian Traders: A History (2005)
  • Édouard Chavannes, Documents sur les Tou-Kiue (Turcs) occidentaux (1903)
  • Josef Marquart, Osteuropäische und ostasiatische Streifzüge (1903)
  • Charles J. F. Dowsett (Übers.), The History of the Caucasian Albanians by Movsēs Dasxuranci (1961)
  • Yusuf Gedikli, Kuzey Kafkasya Hunlarındaki Zilgibis'in, Dağıstan Hazarlarındaki Selifân'ın, Juan Juan, Göktürk, Eski Uygurlardaki Hsie-li-fa Sanlarının Çözümleri, Köken ve Anlamları (2022)
  • Zehra Odabaşı, Türk Devletlerinde Kullanılan İdari Ünvanlar / Used Administrative Titles in Turkish States (2019)

Related

What links here4

Places
Gaochang
Institutions and Titles
Beg · Khan (Khagan)
Sources
Movsēs Dasxurancʿi

Cite this article

Yavuz, Emre-E. “Ilteber.” Turkic Atlas, revised 13 September 2026. https://turkicatlas.business-5d4.workers.dev/institutions/ilteber/

Last revised: 13 September 2026