Further Fragments of Sogdian Manichaean Riddles?

Christiane Reck <reck@bbaw.de> is a research associate at Goettingen Academy of Sciences and Humanities at the project Union Catalogue of Oriental Manuscripts in Germany. She has published Gesegnet sei dieser Tag: Manichäische Festtagshymnen – Edition der mittelpersischen und parthischen Sonntags-, Montags- und Bemahymnen (Turnhout, 2004) and three volumes of the Catalogue of Mitteliranische Handschriften in soghdischer Schrift (Stuttgart, 2006, 2016 and 2018). Currently, she is cataloguing the Persian manuscripts in Germany.

Riddles are a popular literary and oral genre for entertainment and education of young and elderly people all over the world. Riddles should be solved to win the heart and realm of king’s daughters or to save lives in many tales. But riddles also help one to memorize mythological and religious facts and ideas. The use of riddles in the pre-Islamic period of Iranian literature is examined in the summary by G. Windfuhr.[1] One of the most well-known examples of the Middle Persian tradition of wisdom literature (andarz) is the Pahlavi text “Mādayān ī Ĵōšt ī Friyān.”[2] The culture of riddles continued in Islamic times as a unique literary form by famous authors such as Nūr ad-Dīn Jāmī, for example and remain the subject of scholarly analysis and discussion.

At the beginning of the 20th century, four German expeditions to Chinese Turkestan (present-day Xinjiang Autonomous Region in China) took place. They brought thousands of manuscript fragments to Berlin. Among these findings from the Turfan oasis are some fragments which contain Sogdian Christian riddles, published by W. Sundermann.[3] The Manichaean literature of Central Asia also includes amusing didactic genres. In 1945, W. B. Henning published a Sogdian text[4] (So 10100h) containing questions of the King Xusrō,[5] who asks the water sprite (γntrw) about many diverse issues. These questions are mostly comparisons such as, “What is higher than the sky? What is lower than the earth?” and so on. The Manichaean affinity results from the manuscript itself, the book format, and the handwriting which belong to a so-called multiple text manuscript, which contains tales with obvious Manichaean contents. Unfortunately this fragment preserves only the questions without a single answer. The Sogdian texts very often seem to be enigmatic. B. Gharib’s dictionary was a great step forward unravelling some of the mysteries. I dedicate this work to her as a token of my gratitude.

The Berlin Turfan collection stores three further fragments with texts which can be interpreted as kinds of riddles. They remained unpublished all over the time because of their enigmatic character. All the three texts are written on the reverse side of Chinese scrolls, which confirms the Manichaean affiliation of them. All three fragments were found in Toyoq during the second German expedition to the Turfan oasis in 1904-5. The findings of the expedition are marked by the letter T followed by the number of the expedition (I to IV), the finding place (T for Toyoq, for example) and the number of the bundles for transportation. Two of the fragments (Ch/So 14743 and Ch/So 20150) were collected in the same bundle as marked by finding sigle  T II T 1435.

There was another fragment among the Turfan texts of which now only a transliteration in Hebrew letters exists in the Lentz estate in the Hamburg University. The recent shelf number of this lost fragment is *So 21005.[6] The text on one side, labeled as verso in the Catalogue but recto in the edition, contains letter formulas.[7] On the other side, several words have survived. Lentz categorized them as “Schreibübungen” (writing exercises). However, they have been mentioned tentatively in the catalogue Reck VOHD 18, 1 as a riddle, with the answer ’’ph  āp  “water,” which cannot be approved anymore. This fragment has not been edited here.

Among the Old Uyghur texts from Turfan only few fragments with riddles are known up to the present day. Two examples from the Berlin Turfan collection should be mentioned here:

  1. Ch/U 6905 verso:[8] From the altogether preserved eight text lines only the lines /v/4-8/ are to be determined as a riddle with certainty. The clue for this identification is to be found in the concluding question, bo nägü sav ol (“What does it mean?”).
  2. Mainz 842 + U 4990 verso:[9] The question in this riddle is “Which living being do we see?,” the empty space after this question preserved on these joined fragments clearly indicates that the text is finished with this question.[10] Those texts and some proverbs or sentences are similarly enigmatic like ours.[11]
  3. Ch/So 14743 verso:[12]

The first text is preserved in the upper part of the reverse side of a Chinese scroll. The shelf number of the fragment is Ch/So 14743 (T II T 1435; 13,5 x 39,5 cm). This scroll fragment is cut in the middle lengthwise as it was often done in the case of re-use for Sogdian texts. The text itself is short but complete in three and a half lines with two interlinear words (cw ʾnt) whose position is not clear but which could belong near to the end of the third line. Possibly the writer forgot these two words and added them afterwards. The script is relatively clear and the letters s and š are easily distinguished. The letter z is written without connection to the following letter. This makes it relatively easier to read. Although the script and language of the fragment is Sogdian, several Middle Persian and Parthian words are observed. In Sogdian transliteration the letters t, k and x may also represent the sounds d, g and h. However, the reading is not certain in some cases, neither the translation nor the interpretation. It starts with a question and seems to give an answer like an aphorism. This text is somehow puzzling to understand. I discussed the transliteration and translation with Werner Sundermann many years ago and again with Nicholas Sims-Williams a few years ago, as it remained difficult to read and interpret this text. There were some different ideas about the readings of y and r, which now are more clearly distinguished. Nicholas Sims-Williams kindly informed me about his most recent readings and understanding of this tricky text.[13] Whereas the first part seems to be more or less clear, the second part remains difficult, mainly because of the difficulties with the words yʾδ (or βʾδ, rʾδ) and xʾnk (or xʾʾk, xnʾk). I preferred on this place to adopt the reading yʾδ as proposed by Nicholas Sims-Williams. Nicholas Sims-Williams read trw and trwγ “lie, falsehood,” ʾym as “this,” yʾd as “memory,” xʾʾk “earth” or xʾnk “house.” The word mrδmʾn can be a misspelling of mrδxmʾn or the variant form without any h/x (see DMSB, 230). But one could read *myδmʾn “*guest” as well (see Pashto ميلمه milma). Thus it could be together with the Middle Persian xʾnk an unexpected compound of an unattested Sogdian word together with a Middle Persian word for “guest house.”[14] Furthermore it is not clear where the passage cwn ʾnt should actually be inserted.[15] The reading and understanding of the third line is in any case difficult and can merely be a guess: “The memory is a guest house, from them are they.” In any way, all these are only attempts to understand this short sentence. Thus no final translation of this line is offered here.

/1/cw  ʾsty  z-βʾk  trw  ʾstyWhat is the tongue? The tongue is the lie.
/2/z-βʾk  ʾym  trw  ʾsty  pδ  z-yrt

cwn  ʾnt

This lie is in the heart.


/3/y/β/rʾδ  my/rδmʾn  xʾn/ʾk  trwγ… … … from them are they. Falsehood
/4/ʾsty xmʾk


is everything.

Non-Sogdian words:

Sogdian script:Reconstruction in Manichaean script:Translation:
trwdrwlie, deceit (MP), DMSB, 193
ʾymʿymthis (MP/Pa)
pdin etc. (MP/Pa)
zyrtzyrdheart (Pa)
yʾδyʾdmemory (MP ʾyʾd)
mrδmʾnmrdwhmʾnpeople (MP, Pa)
xʾnkxʾnghouse (MP)
trwγdrwγlie, deceit (Pa)
xmʾkhmʾgall (MP/Pa) DMSB, 216

1- Ch/So 20150 verso:[16]

The second text Ch/So 20150 (T II T 1435; 13,4 x 10,5 cm) preserves only the endings of seven lines and a trace of the line above. The presence of the interrogative cknʾʾcw “from what, from which, from whom, concerning which” could indicate that these are riddles.[17] But it could well be a kind of doctrinal text like the Sermon of the Soul,[18] as supplemented by Nicholas Sims-Williams. Unfortunately, only a few words are preserved in this fragment. One cannot thus find out, what is asked and what its answer is. The only meaningful passage is tmʾyk ʾʾtr “the hell’s fire,” as in So 18248I/r/13/.[19] The reading of x(wy)rty is proposed by Yutaka Yoshida. Because of the blurred letters and the lack of context the reading and interpretation is complicated.

/1/](k/m)[   6         ]
/2/](.)  cknʾʾ[cw] Why
/3/](.)myty  rty  CWR(H)[]… And self/person/soul/body[20]
/4/](m/s/š)[1-2]ysy  ky  ZY] … who not
/5/](.)  x(wy)rty  Ο] it is eaten(?).
/6/](.)  cknʾʾc  pyδʾr] Why
/7/](.)  tmʾykw  ʾʾtr] the hell’s fire
/8/](.)nt  ZY  tšʾnt] (they) … and escape

2- Ch/U 6419 verso:[21]

The third fragment Ch/U 6419 (T II T 1198; 16 x 13,7 cm) preserves presumably the end of a text because the end of the sheet of paper seems to be cut and the sentence is terminated by a full stop, a filled circle. We do not know how much text is missing. If the text had been written on a scroll cut in the middle, then the lines would have been nearly complete. But if the text was written on a complete scroll, nearly the half of a line is missing. The text in the lines /8/ and /9/ shows, that more than a few letters should be inserted in between these passages. This implies, that this fragment was torn from a complete scroll. The beginning of the lines is preserved in most cases. The character of the text is not clear. But the interrogatives ktʾm ʾʾδy “which person” and ky “who” indicate, that certain questions are mentioned. The word “pyrnm” occurs many times in this text. It means “before” in a temporal or in a locative sense (“formerly, before, first” or “before, in the presence of”). It can be used as a postposition: see DMSB, 164. In this text it seems to be located after several parts of the body (pʾδʾ, cšm-). In l. 9 the second word is not clearly legible. The second letter should be an n. The third letter seems to be a s as it is curved to the left. But the meaning of ʾ(ns)ʾsy is not clear. Yutaka Yoshida proposed to read the third letter as x and interpret the word ”/nx’s as “battle”. Another unknown word has only the beginning of it preserved: wkʾr[.

It would be very nice to learn what the text is about. Thus I can only wish that further research in Sogdian will allow us to identify the content of this text:

/1/(ZY) pr  xwsʾnt(yʾ)  (βw)[t… is in … and in contentment. [
/2/sr(y)ʾkycw  γ(n)tʾk  ZY  ptx/γ(.)[the chief evil and …[
/3/[n](ʾ)  xwpw  ZY  nʾ  wkʾr[not good and not …[
/4/[1-2](y)  ktʾm  ʾʾδy  xcy  ky  cʾwn  [[…] Which person is it, who [   ] of/from [
/5/pʾδʾ  pyrnm  ʾʾγʾyrt  Z(Y)[before the feet (goes) …[
/6/(ZY) ky  MN  pcmʾky  xw  (.)[and who is of noble [
/7/[     10     ]  cšmʾ  py(r)[nm[               ] before the eye [
/8/[   6   ](.)  ZY  MN  pʾδʾ  (p)[yrnm[           ] and [before] the feet [
/9/ZKwy  ʾ(nx)ʾsy  pyrnm  tysʾy  ZY[in before the battle(?) enters and [
/10/pyrnm  xw  p(t)βyw  ZY  ʾ(z)[before [       ] the respect and …[
/11/[    8    ](..)γrβ  ʾ(st)[y] Ο[             ] is (much).

Summary: There are no more clearly legible and understandable Sogdian riddle texts in addition to those which had been published before this article. All the remnants are ambiguous, or only fragments are preserved. But it may be possible that some other documents will be discovered or identified in the future which could illuminate these puzzling texts.



Ch/So 14743 verso (detail)

Photo: Berlin State Library



Ch/So 20150 verso

Photo: Berlin State Library



Ch/U 6419 verso

Photo: Berlin State Library

[1]G. Windfuhr, “Riddles,” in General Introduction to Persian Literature, ed. J. T. P. de Bruijn (London: I. B. Tauris, 2009), 312‒314.

[2]M. Macuch, “Pahlavi Literature,” in Literature of Pre-Islamic Iran, ed. R. E. Emmerick and M. Macuch (A History of Persian Literature; Companion vol. 1) (London: I. B. Tauris, 2009), 116‒196, esp. 160‒172. See also I. Colditz, “Eine vergessene zoroastrische Märtyrerin?,” in A Thousand Judgements. Festschrift for Maria Macuch, ed. A. Hintze, D. Durkin-Meisterernst and Claudius Naumann (Wiesbaden: Harrassowitz Verlag, 2019), 51‒70,  esp. 51, fn. 2.

[3]W. Sundermann, “Der Schüler fragt den Lehrer: Eine Sammlung biblischer Rätsel in soghdischer Sprache,” in A Green Leaf: Papers in Honour of Professor Jes P. Asmussen, (Leiden: Brill, 1988), 173‒186, [pl.?] V-XII.

[4]W. B. Henning, “Sogdian Tales,” Bulletin of the School of Oriental and African Studies 11 (1945) 3: 480‒482 with n. 3.

[5]P. Lurje, Personal Names in Sogdian Texts, IPNB 8, II (Wien: Verlag der Österreichischen Akademie der Wissenschaften, 2010), 84, no. 65.

[6]Mitteliranische Handschriften: Berliner Turfanfragmente manichäischen Inhalts in soghdischer Schrift, described by Ch. Reck, (VOHD 18,1) (Stuttgart: Steiner Verlag, 2006,) 306, nr. *442.

[7]A. Benkato, Studies in the Sogdian Epistolary Tradition (Berliner Turfantexte; 41) (Turnhout: Brepols, 2018),  41, nr. 27.

[8]S.-Ch. Raschmann, Alttürkische Handschriften, Teil 14: Dokumente: Teil 2  (VOHD 13, 22) (Stuttgart: Steiner Verlag, 2009), 184, nr. 503.

[9]The fragments are described in the database KOHD digital: https://orient-mss.kohd.adw-goe.de.

[10]I thank Simone-Christiane Raschmann for her kind help in referring the Old Uyghur texts.

[11]P. Zieme, “Fragmente von Erzählungen, Sprichwörtern und Reimsprüchen aus der altuigurischen Zeit,” AİBÜ Sosyal Bilimler Enstitüsü Dergisi, Semih Tezcan’a Armağan 13 (2013): 473‒496.

[12]Mitteliranische Handschriften: Berliner Turfanfragmente manichäischen Inhalts in soghdischer Schrift, described by Ch. Reck, (VOHD 18,1) (Stuttgart: Steiner Verlag, 2006), 162‒163, nr. 215 (wrong measurement). The recto side contains Chinese text: Taishō Tripiṭaka nr. 473 = 21, 443c52‒444a.

[13]I thank Nicholas Sims-Williams and Yutaka Yoshida very much for all their help and Alisher Begmatov for improving my English phrasing. All remaining shortcomings are, of course, my own responsibility.

[14]There are some of such multilingual text fragments among the Turfan texts, which use words from several languages; see for example Nicholas Sims-Williams, “A Multilingual Manichaean Calendar from Turfan (U130),” Language, Society and Religion in the World of the Turks: Festschrift for Larry Clark at Seventy-Five, ed. Zsuzsanns Gulácsi (Silk Road Studies 19) (Turnhout: Brepols, 2019), 251‒266.

[15]See DMSB, 67 and 64 s.v. cwn.

[16]Mitteliranische Handschriften: Berliner Turfanfragmente manichäischen Inhalts in soghdischer Schrift, described by Ch. Reck, (VOHD 18,1) (Stuttgart: Steiner Verlag, 2006), 230, nr. 310. The recto side contains Chinese text: Taishō Tripiṭaka nr. 374 = 12, 572c19‒26.

[17]Nicholas Sims-Williams, Desmond Durkin-Meisterernst, Dictionary of Manichaean Texts 3:2, Texts from Central Asia and China (Texts in Sogdian and Bactrian) (Corpus Fontium Manichaeorum: Subsidia 7)(Turnhout: Brepols, 2012), 63‒64.

[18]W. Sundermann, Der Sermon von der Seele (Berliner Turfantexte 19) (Turnhout: Brepols, 1997), 92-93, 153, 155).

[19]W.B. Henning, “The Murder of the Magi”, JRAS (1944): 133‒144, (138 and 140), repr. W. B. Henning, Selected Papers II (Téhéran-Liège: Brill, 1977), 139‒150 (144 and 146).

[20]I thank Yutaka Yoshida for the reading of CWRH.

[21]See Mitteliranische Handschriften: Berliner Turfanfragmente manichäischen Inhalts in soghdischer Schrift, as described by Ch. Reck, (VOHD 18,1) (Stuttgart: Steiner Verlag, 2006), 268, nr. 369. The recto side contains Chinese text: Taishō Tripiṭaka nr. 262 = 9, 18c22‒19a3.