The Old Turkic script is an alphabet of some forty signs. The signs are angular and run from right to left. They fit Turkic sounds more exactly than any script its users could have taken over unchanged. Scholars have argued about where the signs come from since Vilhelm Thomsen read them in 1893, and the question is still open. The argument runs almost entirely on the shapes of the letters. That is why it stays open.

The state of the evidence

Three fixed points bound the question. There are few others.

The Chinese chapters contradict each other. The Zhou Shu describes the customs of the Göktürks in the sixth century. It says that their written characters resemble those of the Hu.[1] The Sui Shu says the opposite: they have no writing and notch wood to make their agreements.[2] The Bei Shi contains both notices without choosing between them.[3] Guzev and Sergey Klyashtorny do not set the two notices against each other. They understand them as two writing systems side by side in the sixth century. One is the script of the Hu. The other is notched wooden sticks, which they count among the forerunners of the runiform signs.[4]

The oldest monument the Göktürk state raised is not in Turkic at all. The Bugut inscription of about 584 is cut in Sogdian script and in the Sogdian language. The script was in official use in the khaganate in the last quarter of the sixth century.[3]

The oldest known runiform text is the Çoyren inscription in Mongolia. It is dated to between 688 and 691, a century after Bugut.[5]

Everything else in the argument is inference from the form of the letters.

The Aramaic derivation

The view now generally held derives the alphabet from the Aramaic of the eastern Iranian world. The Sogdian script descends from the same stock. The shapes were then adapted to the chisel.[5] Thomsen was the first to read the inscriptions. He argued for this descent, and it has held the field since.[6] Pritsak and Róna-Tas hold to an Aramaic origin by way of an Iranian language. Livshits and Sims-Williams hold to a Sogdian model.[7] Radloff accepted Thomsen’s Aramaic origin. He also thought that the runic script of the Goths, whom the Huns met in the west, might have left traces in the Orkhon letters.[8]

Accounts within it differ on who made the alphabet, when, and out of what.

Gerard Clauson regards it as a deliberate creation. In his account, the alphabet was built on Irano-Aramaic and supplemented from Greek. It was made for Istemi’s embassies to Byzantium. Its inventor was a cultured Hephthalite or a Sogdian.[5] The objection is practical. An alphabet devised for Constantinople would need readers in Constantinople. There were none.[5] Tryjarski calls the suggestion tempting but without a scrap of historical evidence. He also reports Donner’s objection. Had a victorious khagan ordered a new script, the many historians who write of much smaller matters would have mentioned it.[9]

Annemarie von Gabain sees it as an adaptation of the Aramaic cursive used in the Arsacid chancellery. In her account, the western Göktürks made it out of their contact with Iran in the late sixth century. It was then carried east with the first wave of Manichaean missionaries.[5]

Louis Bazin treats it as a political act. In his account, the alphabet was developed in a national upsurge, and the Orkhon inscriptions are its monument.[5]

Peter B. Golden takes none of the three and leaves the making unattributed. The letters evolved out of Aramaic, with additions and adjustments, for the epigraphic purpose they served.[5] He adds a detail that the shape arguments tend to pass over. The Orkhon stones were cut by bedizçi, artisans sent from China. Their hands are in the letters as much as any ancestor alphabet is.[5]

Sergey Klyashtorny sets the beginning much earlier, and for the Sogdian script rather than the runiform one. The fitting of the Sogdian alphabet to Turkic sounds, he says, began in the fifth century, in the Gaochang period of the Ashina. That would make the Sogdian cursive older than the runiform writing.[3]

The tamga derivation

A second group of theories derives the letters not from an alphabet but from the tamgas. Tamgas were property and clan marks, cut into rock and burned into hides across the steppe. They were then reshaped into a set of sound signs. Schifner set out the derivation. Aristov, Mallitskiy and Sokolov worked on it.[7]

The derivation has a strength. It accounts for signs that look like marks in use on the steppe and like letters in use nowhere else. It also has a weakness. A mark and a letter can resemble each other for three different reasons, and the resemblance does not say which.

The mixed theory

A third theory combines both. An alphabet known through Sogdian gave the idea and part of the inventory. The tamgas gave the shapes for the rest. It accounts for more of the material than either of the others. For the same reason, it is the hardest of the three to test.

Several adoptions

Tryjarski holds that the script was not invented once. It was taken over several times, at different moments, on the decision of tribal chiefs and rulers.[10] Besides the Göktürks, he counts the Kirghiz, the Karluks, the Uyghurs, the Chik and the Bulgars among those who took it up. He adds that the Avars and Iranians such as the Alans may also have done so.[11] He reports two ways of sorting the alphabets. Kyzlasov separates an Asian group from a Eurasian group. The Asian group has the Orkhon, the Yenisei and the Talas. The Eurasian group has the Don, the Kuban, the southern Yenisei, Achiktash and Isfara. The script of Nagyszentmiklós belongs to neither.[12] Róna-Tas sets out four phases. The first is the taking over of an Aramaic-type alphabet. The second is the making of new letters for Turkic. The third is the passing of the alphabet to a people with a common Turkic speech. The fourth is the rise of local variants such as the Yenisei texts.[13]

Independent growth from pictures

Viktor Guzev regards the script as an independent growth. The signs, he says, show every stage of a writing system side by side: pictures, word signs, syllable signs and letters. He holds that this would be impossible if the script had been copied from a finished alphabet.[14] At least five signs are undisputed pictures: at (horse), ay (moon), äb (tent), är (man) and ok (arrow).[15] Guzev argues against both the Aramaic and the Sogdian derivation. His argument rests on how the script is built, not on its shapes. A script copied from an alphabet would keep its inner form. The runiform signs keep word and syllable values that no alphabet has.[16] For the outer form, he finds the tamgas a better match than Sogdian letters.[17]

Guzev and Sergey Klyashtorny put the matter differently. In their view, the script was not invented in the sixth or seventh century but reformed then. Existing sign forms were sorted into a standard alphabet. Some of them received new sound values.[18] They list three ways in which the Göktürk state could meet its need for writing. It could keep a foreign script for the chancery. It could bend the Sogdian cursive to Turkic. Or it could make a script of its own out of the signs already in use. The first was the course of the First Khaganate. The second led to the Old Uyghur script.[4] They also find runiform-looking signs on a silver cup from the Issyk mound near Almaty (see the Issyk inscription). They find such signs on a trilingual stone from Dasht-i Navur in Afghanistan as well. If these prove to be the script of Iranian-speaking nomads, they would be forerunners of the Turks.[18]

An older archaic script

D. D. Vasilʼev holds that an archaic form of runiform writing existed first in the Altai and the Sayan. It is found in short texts on rock at natural places of cult. The Orkhon script is, he says, this older writing, reformed under strong Sogdian influence in the second half of the seventh century.[19] He argues against an early derivation from Sogdian in the oases of eastern Turkestan. No early runiform texts have come from there, and the oldest and most primitive ones are found in the Altai.[20]

Place and date of origin

A. M. Şçerbak places the birth of the script in southern Siberia, because most of the runiform texts are found there.[21] He holds that it arose most probably under the Second Khaganate, since no earlier runiform text is known.[22] Kasımcan Sadıkov thinks it more probable that the script took shape in Central Asia and spread from there to the Altai and the Orkhon valley. He thinks it less probable that it travelled west from the Orkhon and Yenisei region.[23]

The argument from the writing system

A fourth line of argument asks how the writing system works, not where its letters came from. The script is an alphabet and not a syllabary. The signs that stand for a consonant together with a vowel show only the quality of the consonant beside particular vowels. They appear beside separate vowel letters.[24] Its design suits the language it writes. It keeps front and back harmony apart in the consonants as well as in the rounded vowels. It marks no tone. It implies the vowel before a consonant and not after it, which no other script of its kind does.[25] Three of its signs look like what they name: the half moon ay, the arrow ok and the tent äb, ‘house’.[26] Guzev takes the opposite view of the signs. He reads the signs for ok, ök and ık, and the cluster signs such as nt, nç and rt, as word or syllable signs of the form vowel plus consonant, not as letters.[27]

The script was in common use over the whole country of the early Turks, and not outside it. This is taken to point to a Turkic creation, in the way the Tangut and the Khitan made scripts of their own, with a remote Semitic motive at most.[28] The Göktürk empires then probably wrote the Sogdian–Uyghur script at first. They had come to settled civilisation through the Sogdians. The runiform script then came back in the second khaganate, as part of the return to the Turkic way of life that the Tonyukuk inscription preaches. Its first home may lie in central south Siberia and not in Mongolia.[29]

The limits of the argument from shape

All three theories argue from form. Form will not choose between them. A letter that looks like a tamga may descend from it, may resemble it by accident, or may have been picked because the shape was already familiar to the men who would read it.

The evidence that would decide the matter is a dated text in a transitional alphabet, in which the borrowing could be watched. No such text is known.

The dating

The debate usually runs two questions together, but they are separate. One asks where the letters came from. The other asks how old they are.

The best-supported view is that the script was not necessarily the Sogdians’ gift. It was in use under the First Khaganate, before the dated stones of the eighth century. Four things support it. The first is the Zhou Shu’s sentence. The second is the completeness of the alphabet when it first appears on stone. The third is the spread of runiform texts over country that never lay under one chancery. The fourth is that the dated monuments date the raising of stelae rather than the script. The article on the script sets them out.

It takes a side on the age of the letters and none on their ancestry. The two are compatible. An alphabet can be made early and still be made out of something borrowed.

References

  1. ^ Zhou Shu (Book of Zhou), juan 50
  2. ^ Sui Shu (Book of Sui), juan 84
  3. ^ a b c Golden 1992, p. 151
  4. ^ a b Guzev 1993, p. 31
  5. ^ a b c d e f g h Golden 1992, p. 152
  6. ^ Thomsen 1896
  7. ^ a b Tuguşeva 2012, p. 78–79
  8. ^ Tuguşeva 2012, p. 79–80
  9. ^ Edward 1993, p. 40–41
  10. ^ Edward 1993, p. 36, 40
  11. ^ Edward 1993, p. 36
  12. ^ Edward 1993, p. 39
  13. ^ Edward 1993, p. 39–40
  14. ^ Guzev 2000, p. 213–219
  15. ^ Guzev 2000, p. 212
  16. ^ Guzev 2000, p. 215, 218
  17. ^ Guzev 2000, p. 218
  18. ^ a b Guzev 1993, p. 32
  19. ^ Vasilʼev 1993, p. 64–66
  20. ^ Vasilʼev 1993, p. 65
  21. ^ M. 1990, p. 183
  22. ^ M. 1990, p. 184
  23. ^ Sadıkov 2000, p. 305
  24. ^ Erdal 2004, p. 38–39
  25. ^ Erdal 2004, p. 39–40
  26. ^ Erdal 2004, p. 40
  27. ^ Guzev 2000, p. 214–216
  28. ^ Erdal 2004, p. 38, 40
  29. ^ Erdal 2004, p. 40–41

Sources

Literature

  • Peter B. Golden, An Introduction to the History of the Turkic Peoples (1992)
  • Vilhelm Thomsen, Inscriptions de l'Orkhon déchiffrées (1896)
  • Étienne de la Vaissière, Sogdian Traders: A History (2005)
  • Emre-E. Yavuz, The Gokturk Empire: 2nd Edition (2025)
  • Marcel Erdal, A Grammar of Old Turkic (2004)
  • Viktor G. Guzev, Göktürk Yazısının Kendiliğinden Doğma (Otokton) Menşei Varsayımını Esaslandıran Deliller (2000)
  • Viktor G. Guzev; Sergey G. Klyaştornıy, Genel Yazı Nazariyesi Işığında Göktürk Yazısının Menşei Meselesi (Okunuşunun 100. Yıl Dönümü Dolayısıyla) (1993)
  • Tryjarski, Edward, Anonymity, Adaptation and Diffusion of the Asian and European Runic Scripts (1993)
  • D. D. Vasilʼev, Göktürk Yazı Kültürünün Asya’nın Merkezinden Doğu Avrupa’ya Yolu (1993)
  • L. Yu. Tuguşeva, Türk Runik Alfabesinin Kökenleri (2012)
  • Kasımcan Sadıkov, Özbekistan'da Bulunan Göktürk Harfli Yeni Metinler (2000)
  • Şçerbak, A. M., Türk „Runik“ Alfabesinin Yayılmasına Dair (1990)

Related

What links here11

Persons
Maniakh
Institutions and Titles
Post and Communication · Tally Sticks and the Golden Arrow
Languages
The Old Turkic script · The Sogdian language
Customs & Pastimes
Culture of the Turkic Peoples
Sources
Altai runiform inscriptions · Bugut inscription · Issyk inscription · Runiform inscriptions of the Khazar lands
Controversies
The Turkic ethnogenesis debate

Cite this article

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

Last revised: 8 November 2026