Two languages, two scripts, three combinations: A (personal?) prayer-book in Syriac and Old Uyghur from Turfan (U 338)

by Adam Bremer-McCollum in

1      Introduction

In a recent issue of the New Yorker, Claire Friedman offers a humorous one-page guide to reading, or at least preparing to read, Infinite Jest:[1]

1.  Buy hardcover copy of “Infinite Jest”at brick-and-mortar bookstore. Touch paper and feel connected to hundreds of years of printed language. …

2.   Walk home and experience heft of text as bag handles dig into palm.

Embrace heaviness as evidence of import. Thumb nose at pedestrians who aren’t carrying impressive, heavy books.

The manuscript I would like to introduce to you has far fewer pages and far less heft than Infinite Jest, but its materiality is no less palpable, and thinking about it as a material object brings up questions that can help us think more realistically both about this and other texts.

The object in question is a very small booklet with prayers (partly from a known Syriac liturgical tradition) in two languages and scripts and a colophon. We first have a longer Syriac piece written in Syriac script, and then several lines in the Turkic language of Old Uyghur in Uyghur script, and then it is repeated, but not identically, again in Uyghur, but now in Syriac script, together with two lines from the Syriac Bible. Finally, there is a colophon, written in Uyghur script and language.

Among the issues that this booklet raises when taken seriously as a material and a textual object are:

•   how are language and script related?

•   what was its purpose?

•   for whom was it written?

•   is it really liturgical?

It is also probably not a text that we would expect to be copied again as we would a literary text: it is seemingly a one-off production. As the colophon seems to make clear, though, it is not entirely a private or personal text.

This document almost requires us to think about it materially, including the page layouts, the varieties of script and language combinations and their alternation, and its very small size. If we ignore any of these factors, whether in thinking about the text or in presenting it in an edition or translation, we are no longer dealing with it as it is, and we run the risk of misrepresenting it and making it something it is not. It is a very particular kind of document, but the issues it gives us a space to think about can perhaps help us interrogate other manuscripts and texts of a different character with similar questions.


1.1               Discoveries at Turfan, etc.

Along with a great many more, the manuscript was produced, used, and eventually re-discovered in the Turfan region of Central Asia. The story of the early twentieth-century expeditions (Prussian/German, French, Russian, Japanese) in Central Asia has been told before, as, for example, by Peter Hopkirk in his Foreign Devils on the Silk Road. Names such as Sven Hedin, Aurel Stein, Paul Pelliot, and Albert von le Coq will be well-known to anyone familiar with the story. The last individual, von le Coq, together with F.W.K. Müller, V.V. Radlov, Annemarie von Gabain, and others, put Turcology on a firm philological footing, something of particular relevance for our manuscript, which is partly in a Turkic language.

Among the results of these expeditions was the “acquisition” of a huge collection of manuscripts in a dizzying array of languages and scripts from around an area called Turfan. This group of texts consists of around 40,000 fragments kept in Berlin in 22 languages and 20 scripts, all the result of a series of Prussian expeditions during the years 1902-1914.[2] As with other Central Asian sites, the majority of manuscripts come from Buddhist and Manichaean communities, the minority being specifically Christian. These Christian Turfan materials amount to a bit over 1000 fragments in the following languages, with attested script⒮ in parentheses:

•   Syriac (Syriac)

•   Middle Persian (Pahlavi)

•   Sogdian (Syriac, Sogdian)

•   “New” Persian (Syriac)

•   Old Uyghur (Syriac, Uyghur)

These are, apparently, all from “a Christian monastic complex”[3] in Bulayïq (near Turfan).


1.2     Turkic languages, Old Turkic, and Old Uyghur

Our text is in a variety of Old Turkic known as Old Uyghur.[4] (The modern Uyghur language also belongs to the Turkic family but does not derive from Old Uyghur.) Although their dating is something of a guess, Old Uyghur texts date from the ninth to the thirteenth centuries.

These texts are written in a variety of scripts, including the one known as Uyghur script (< Sogdian script < East Syriac script; see further below). Some  of the later Christian material is written in a kind of Syriac script,[5] but Uyghur script, adapted from Sogdian script and “used by adherents of all religions among the Turks of Eastern Turkestan,” is most common.[6]

The largest part of this linguistic corpus is Buddhist (mostly Mahāyāna),[7] followed by Manichaean, and, least of all, Christian. For all three religious communities, the texts are generally translated texts, the Christian texts being translations from Syriac or an Iranian language.[8]

For the Christian material in Old Uyghur, the recent appearance of Peter Zieme’s Altuigurische Texte der Kirche des Ostens aus Zentralasien makes their study much easier than it had been before.[9] In this volume a scholar long experienced in this and related corpora has collected together both the earlier and the later texts together with a German translation. (Most of these texts are still not available in English, and certainly not in any single place like this.)

What does “Christian Uyghur literature” look like? Earlier texts include a text on the Magi, various prayer and confessional texts, a text on St. George, another on Paul and Thekla, a text on God and the devil, among others, and there are later texts (from Xaraxoto), such as a homily on almsgiving, a hagiographic fragment, and a poem from the White Pagoda 白 塔 (Bái tǎ), which seems to be in heptasyllabic meter, a pattern known in both Syriac and in Turkic verse.[10]


2          The manuscript U 338

Images of the manuscript are available at the BBAW’s Turfan site.[11] Here is the basic metadata:

•   shelfmark: U 338 (olim T II B 41, no. 1)

•   5 sheets // 10 folios // 20 sides

•   measures 6.3-6.4 cm x 7.0-7.3 cm

That is, this is a very small manuscript. Zieme calls it a booklet (“Büchlein”),[12] and it is indeed that, but a diminutive suffix is only barely enough to convey how very small it is! To put its size in practical perspective, a credit card is about 8 x 5 cm.

Side 5, the only page with both scripts

Side 5, the only page with both scripts

There are two distinct scribes:

•  scribe X penned the Syriac text in Syriac script

•  scribe Y wrote the Uyghur text in Uyghur script and Syriac script[13]

One of these, scribe Y, names himself in the colophon, but fails to mention the other at all, and there is no indication of their relationship. Were they teacher and student, senior monk and junior monk, or something else entirely? His own name is Bäküz ≈ Bacchus, a name common in Syriac, not least thanks to the popularity of the saintly duo of Sergius and Bacchus.[14]

The text was found folded in such a way that the original order of the pages was obscured, and the pages/sides were thus numbered in the wrong order. The original order of the pages, divined by Zieme, and followed by Dickens, is: 

1.   side 8, line 1-side 15, line 4: Syriac prayer in Syriac script

2.    side 16, line 1-side 19, line 4: Old Uyghur prayer in Uyghur script

3.    side 20, line 1-side 5, line 1:[15] Old Uyghur prayer in Syriac script + two lines in Syriac[16]

4.    side 5, line 2-side 7, line 6: Old Uyghur colophon in Uyghur script

That is, the new sides 8-20 are the (unnumbered) old sides 1-13, and the new sides 1-7 are the old 14-20. The new numbers, written when the object was found, are clearly visible in the photos.

An aside: This manuscript is hardly the only one to have suffered such a fate. The Syriac manuscript, Saint Mark’s Monastery, Jerusalem, 180[17] serves as another—probably contemporaneous! —example of a manuscript with present disordering. Both that manuscript and U 338 thus become a kind of choose-your-own- adventure manuscript!

This reconstructed order agrees with some physical data[18] of the book itself, but it also presents a coherent literary structure, with

1.   a leading Syriac prayer text, Syriac being a liturgical and prestige language[19] for Christians, Uyghur-speaking and otherwise, in Turfan;

2.    followed by a “vernacular” prayer text that partly matches the Syriac prayer but also includes a Syriac intrusion, which is nevertheless still written in Uyghur script;

3.    and finally the colophon comes at the end (where we expect to find it).[20]

Dickens (307) suggests that this may be a kind of school text —this does not obviate its being a personal spiritual text, too[21] —with the better scribe writing first (the Syriac prayer), and the less experienced scribe writing afterward (everything else). I am not convinced that this is so. Neither the handwriting nor the orthog- raphy of either scribe seems particularly practiced. If our manuscript is indeed not all that professional, it stands in the company of a number of other allographic texts.[22]

Thinking further about the binding, we find that it is such that the book can easily be opened to any page and read cyclically. As Dickens puts it, “the type of thread binding used enables the booklet to be cycled through and left open at any page.”[23]  This seems to mean that the “covers”(current sides 6 and 7) do not really matter as such, and that the book can be opened to any page-spread, not unlike some cloth books for babies and toddlers.

The only relatively solid anchor for a date, or at least a terminus post quem, is the presence of a Mongolian word to designate Christians. The fact that the text uses this word most likely situates its writing in or after the thirteenth century.

Finally, returning to the fact that there are two scripts and two languages in the text, we should note that this manuscript is not the only locus of Syriac-Uyghur interactions. Others include a Syriac-language Psalter written in Uyghur script[24] and another prayer-text in Uyghur script with alternating Syriac and Uyghur on the back (U 323).

3 The text, and a translation

Here is the complete text of the manuscript U 338, with language and script for each section clearly indicated. It is important to note, in terms of the manuscript as object, that these are codicological units only by script-language combination, that is, there are no rubrics or anything similar to mark them. I give the whole in page- by-page and line-by-line transliteration, transcription, and English translation.

Syriac prayer in Syriac script


08.1  [m]ryʾ ʾlhʾ ḥyltnʾ

08.2   d[ylk] hy ʿdtk qd

08.3  yštʾ wbḥšh rbʾ

08.4   dmšyḥʾ ʾzdbnt{ʾ}

09.1  ʿnʾ dmrʿytk mn

09.2   ṭybwtʾ drwḥʾ d

09.3  qwdšʾ hw dʾ

09.4   yt⒲hw br kynʾ dʾl

10.1  {dʾl}hwtk mšbḥtʾ

10.2   mty<h>byn drgʾ dsym

10.3  {t}<yd>ʾ dkhnwtʾ šr

10.4   yrtʾ bmrḥmnwtk

11.1   mry ʾšwytk

11.2     lbṣyrwth [dk]ynn

11.3    dnhwʾ hd̈mʾ ydyʿʾ

11.4     bgwšmʾ rbʾ

12.1  dʿ[d]tk wnšmš

12.2   ʿwd[r]nʾ rwḥnʾ lnṗš{ʾ}

12.3   tʾ dmhymnʾ ʾnt

12.4   [hk]yl mrn wʾlhn

13.1    šmlʾ ʿmn ṭybwt[k]

13.2      wʾšpʿ bʾydyn

13.3     mwhbtʾ wrḥmyk

13.4      wḥnnk nhw[w]n ʿlyn

14.1  wʿl] ʿmk [h]nʾ dkbt

14.2   lk ※ hbln mry

14.3   b[ḥnnk] d[kl]n ʾkḥ[d]

15.1    šwytyt kl[h]w[n]

15.2 ywmtʾ dḥyyn [nš]pr

15.3 lʾlhwtk bʿ[bd̈]ʾ

15.4 ṭbʾ dmnyḥ[yn]

Transcription and translation

Screen Shot 2019-05-10 at 2.32.49 PM.png


Old Uyghur prayer in Uyghur script


16.1  ymʾ pw pyz

16.2   ynk ʾtwyzw

16.3  mwz ny tyryk

16.4   ʾsʾn twtdʾcy

17.1  ʾwyzwdwmwz

17.2   ny qwtqʾrtʾ

17.3  cy ʾʾlqw ʾʾdm

17.4   ʾwqwsy ylnk

18.1  wqlʾryq sw

18.2 ywrkʾyw ʾyry

18.3  nckʾyw ʾwtlwq

18.4 yʾlyncyq yln

19.1 kwqlʾryq sw

19.2 ywrqʾyw ʾyry

19.3 nckʾyw ʾwtlwq

19.4 qʾnyq kwynkwlyn

 Transcription and translation

Screen Shot 2019-05-10 at 2.36.40 PM.png


Old Uyghur prayer + two lines in Syriac, in Syriac script

Side 3 and the Syriac "intrusion"

Side 3 and the Syriac "intrusion"


20.1 [{}] ymʾ pw

20.2  pyzyng ʾt

20.3 wyswm wzny

20.4  tyryg ʾsn

01.1  twttʾcy

01.2   ʾwyzwtwmwz

01.3  ny qwtqʾrtʾcy

01.4   hnmz mšyhʾ

02.1  tngry nyng

02.2   sʾbygy tʾb

02.3  lʾky kwngly

02.4   ʾwyzʾ

03.1  ʿmmʾ pryqʾ

03.2   gnšʾ šbyhʾ

03.3  typ ʾt

03.4   ʾwyzʾ ʾsyl

03.5  myš gww ʾ

03.6  wyzʾ gwlʾl

03.7  myš ʾlky

03.8  šlyk ʾr{y}kʾ

03.9  kwtlʾr kʾ

Transcription and translation 

Screen Shot 2019-05-10 at 2.43.19 PM.png


Old Uyghur colophon in Uyghur script


05.1   ʾwd yyl rʾm

05.2  ʾʾy ʾwyc ʾwtwz

05.3   qʾ mn pʾkwz

06.1  pytytym pwyʾ

06.2   ny ʾtm ywnʾn

06.3   qʾ tʾkswn

06.4  ʾʾmyn ʾʾmyn

06.5  typ mwny tʾk

06.6  pyys qʾtʾ pytyp

06.7  ʾβdʾ ʾwqwl qyz qʾ

06.8  ʾwqytyp ʾsydmʾky

06.9  pwlzwn ʾmyn

Transcription and translation

Screen Shot 2019-05-10 at 2.45.31 PM.png



4      Writing systems

Thinking in terms of a matrix of script and language for this little manuscript brings us to the practice of allography. At least when defined broadly as the writing of a language in something other than a previously used or the usual script, allography is more common than might be assumed at first.[38] In the strictest sense, only those languages that use a writing system created ad hoc, like Armenian, or languages that use a heavily evolved or reworked writing system, such as Gǝʿǝz, can be called non-allographic.[39] Coptic, and even more so,[40] Old Nubian, employ allographic writing systems, as is English. Also, the fact that Coptic and Old Nubian modify the alphabet —in these cases by addition —does not obviate their being known as examples of allography. These additions, each from another writing system that had been used (Demotic and Meroitic, respectively), is proof itself that one system has been replaced by another. There are many other examples of more obvious cases of allography, such as Arabic written in Hebrew (Judeo-Arabic) or Syriac script (Garšūnī par excellence).[41] Finally, transliteration itself[42] —but presumably not transcription, which involves so much interpretation and additions to what is there in the “mere” script itself? —is also an example of allography, as is the shift in the copying of manuscripts from one main script type to another more or less eventually closely related: we may think, for example, of uncial to minuscule in Greek manuscripts, and erkat’agir and bolorgir in Armenian manuscripts.

As for Syriac script and Uyghur script, which are combined here, they are in a roundabout way two re-workings of the same thing, an earlier Aramaic script, in the case of Uyghur script, via Sogdian. We saw that the fourth combination possible of script and language in our text, that is, Syriac language written in Uyghur script, does not occur here, but it does elsewhere: there is, as already noted, a Syriac-language psalter in Uyghur script also from Turfan.[43] In other cases we can find the same language, but a different script-type, as in Estrangela and Serto, respectively for the main text and colophon, both in Syriac, in SMMJ 129, which is dated to the year 806.

To judge from the writing in U 338, at least, it was not a practiced and trained scribe for either script or either language. This is apparent both in the general aesthetic effect of either scribe’s handiwork, and in some irregularities in spelling, both for the Uyghur-script[44] and the Syriac-script parts —for the latter in both Syriac[45] and Old Uyghur.

Sides 16 and 20 are the rough line-by-line equivalents of the Uyghur-script and Syriac-script versions of the same part of the prayer, the Syriac side (20) taking a bit more space than the Uyghur, and a side-by-side comparison helps highlight their particular practices:

16.1 ym’ pw pyz 20.1 [{}] ym’ pw

16.2 ynk ʾtwyzw 20.2 pyzyng ʾt

16.3 mwz ny tyryk 20.3 wyswnm wzny

16.4 ʾsʾn … . 20.4 tyryg ʾsn

Sides 16 and 20, with the same parts in Uyghur script and Syriac script

Sides 16 and 20, with the same parts in Uyghur script and Syriac script

Sides 16 and 20, with the same parts in Uyghur script and Syriac script

Sides 16 and 20, with the same parts in Uyghur script and Syriac script

In these four lines we see written Uyghur k for Syriac g, Uyghur z for Syriac s, different spacing for suffixes, and an explicit ālaφ in Uyghur that is missing in Syriac.

5      Thinking about the object

Here are some immediate impressions from looking at this little book in its original order:

1.   sloppy Syriac at the beginning, and looking distinctly not like a “literary” manuscript, including its size

2.    hole in sides 11-12

3.    much of the writing on side 14 effaced or otherwise unclear

4.    at side 16 we first have Uyghur script and language; there is no rubric or any kind of break to mark the shift in script or text

5.    at side 20 we return to Syriac script, but still Uyghur language; there is an erased word, which seems to be in Uyghur script, not Syriac; again there is nothing to mark the change in script or text

6.    at side 3 we have the Syriac-language “intrusion”, but no change in script

7.    side 5 is the only one with two scripts: the first line is in Uyghur in Syriac script, but the page’s remaining three lines, the beginning of the colophon, are in Uyghur script

8.    side 6 and 7, which contain the remainder of the colophon, are, respectively, the back and front covers

These impressions, tied to the investigations above, yield the following musings—subjective, no doubt! —about this and similar —in all sorts of ways! —textual objects.

5.1  The booklet’s size, and the resources to produce it

At least with respect to size, we may compare this manuscript with another that has been studied recently. The Coptic Gospel of the Lots of Mary manuscript, probably of the 5th or 6th century, is 7.5 x 6.87 cm,[46] which according to the standards of modern book-collecting counts as a “miniature book”,[47]  and our booklet does, too. As Luijendijk points out, the parchment for a book of this size could be cut from one sheep.[48] This stands in stark contrast to the many much heftier books in various languages “that must have cost the lives of whole flocks, since the hide of a single animal could not supply more than one skin or two leaves of such size,” as William Wright laments for certain larger Gəʿəz manuscripts.[49] The Coptic Gospel of the Lots of Mary manuscript differs from our Syro-Turkic one in being much more professionally executed. Luijendijk notes that the Coptic scribe “achieved a professional-looking page design,”and “Given the book’s small size, the hand- writing  is  surprisingly  legible  and  quite  elegant,”and  finally,  “Both  paleography and  orthography  indicate  an  educated  copyist.”[50]   The  Turfan  manuscript  is  not necessarily difficult to read, but no one would dream of calling it elegant.

5.2  A biscriptal monastic context of production

The Lots of Mary manuscript, too, is like ours in being a monastic production.[51] By all accounts, both Syriac and Uyghur were at least mostly also monastic —not merely “monastic Christian” in the case of Uyghur —writing systems, and in just this way. The one named scribe, too, Bäküz, is probably a monk, to judge from his name, which is neither Syriac, nor Uyghur, but Greek.[52]

On a very related note, it has recently been highlighted that “[i]n some cases, it is quite clear that only persons with religious training had knowledge of allographic  writings”(den  Heijer  and  Schmidt,  32).[53]   One  category  of  allographic writing consists of

cases in which the script is that of a more or less foreign language endowed with a special status, mostly of a religious nature, and in which this script is borrowed for writing a language that is the natural, primary native language of the community in question

and “Syriac stands out as a forceful prestige language in this sense.”[54]  This would refer, in the case of this manuscript, to Old Uyghur written in Syriac, while the Syriac-language Psalter in Uyghur script serves as an opposite example.

5.3  An amulet?

Sometimes an object’s use as an amulet of some kind is evident in the object itself: for example, marks that indicate it was touched often or kept close to a person’s body, like the surviving folio of the P. Bodmer version of Zostrianos.[55] Can the Syriac-Uyghur booklet be called an amulet?[56] Well, if the term may refer broadly to an object, in this case, textual, that is considered to be of unique personal/individual value of some kind, such that it may even often or on occasion be kept on one’s body for particular protection or blessing, then yes, this may be an amulet. Much more than that we cannot say with precision.

With reference to the line translated above as, “May it be heard that (they are) writing like this five times and making boy⒮ and girl⒮ read (it) in the house! Amen!” (munï täg beš kata bitip ävdä ogul kïz okïtïp äšidmäki bolzun amin), Zieme remarks that if the text is indeed a kind of writing exercise, it is less likely to be an amulet,[57] but this is not necessarily true, is it? Imperatives of writing and/or reciting are not at all unknown in amulets.[58]

5.4  Liturgy, and not; but prayer?

As it turns out, the prayer part of the manuscript is from a known liturgical text, the prayer of the laying-on of hands, which is still used in the Church of the East rite.[59] The text as recorded in this copy differs only in minor points from the received text.[60] Yet, while this is a liturgical text, unlike usual liturgical manuscripts, there are no rubrics, and no instructions of any kind for officiant or congregation. If the booklet was really “primarily for private use”,[61] which, given the size, it must have been, maybe we can call this a micro-liturgy?

Dickens refers to this little book as showing the connections to “both the liturgical traditions of the Church of the East rooted in Mesopotamia and Persia and the surrounding Turkic culture.”[62] These are certainly all present in our short bilingual text, but I wonder if its scribes and any users did not see these things as one thing altogether, rather than both/and: that is, an Uyghur-and-Syriac-using community open to either language being written in either script.

6      The obligatory concluding words

Questions like these, and observations with an active eye toward a textual object as a real object that has existed at other times in human experience in as real a way as it may to us, put us in touch with a manuscript’s nodes of contact —its originators (e.g. authoritatively, practically, monetarily), users, and subjects (those who find themselves under its authority).[63]

It does not get much more material than the ink on the page, and ink and  page are bearers of both languages and writing systems, as well as any possible combinations thereof. No one takes this object seriously as it is, if they ignore the different combinations of Uyghur and Syriac scripts and languages that we find in this manuscript. This both forces —and helps! —us to remember its makers and intended audience⒮, and to think of them as nearby as we read, copy, and think through this object. However differently or not we may read the text from them, we don’t do so apart from them.

Side 7 (now the front cover), with the end of the colophon

Side 7 (now the front cover), with the end of the colophon

7      Bibliography

Ashurov, Barakatullo. 2015. “Sogdian Christian Texts: Socio-Cultural Observations.” Archiv Orientální 83 (2015): 53-70.

Siam Bhayro. 2012. ”A Judaeo-Syriac Medical Fragment from the Cairo Genizah”, Aramaic Studies 10: 153-72.

Siam Bhayro. 2014. “Remarks on the Genizah Judaeo-Syriac Fragment.” Aramaic Studies 12: 143-153.

Graeme Bird, ”Review of: Scripts Beyond Borders: A Survey of Allographic Traditions in the Euro-Mediterranean World.” Hugoye: Journal of Syriac Studies 21.2 (2018): 434–440.

Mark Dickens, “Syro-Uigurica II: Syriac Passages in U 338 from Turfan,” Hugoye 16 (2013): 301-324.

Johannes den Heijer, Andrea Schmidt and Tamara Pataridze, eds., Scripts Beyond Borders: A Survey of Allographic Traditions in the Euro-Mediterranean World (Louvain: Peeters, 2014).

Liv Ingeborg Lied. Nutshell_

AnneMarie Luijendijk. 2014. Forbidden Oracles?: The Gospel of the Lots of Mary. Studien und Texte zu Antike und Christentum 89. Tübingen: Mohr Siebeck.

Simone-Christiane Raschmann. 2016. “Uygur Scribbles on a Wooden Object.” In Lilla Russell-Smith and Ines Konczak-Nagel, eds. The Ruins of Kocho. Traces of Wooden Architecture on the Ancient Silk Road. Berlin, pages 42-48.

Peter Zieme, “Notes on a Bilingual Prayer Book from Bulayık,” in ed. Diet- mar W. Winkler and Li Tang, eds., Hidden Treasures and Intercultural Encounters: Studies on East Syriac Christianity in China and Central Asia, Orientalia–Patristica– Oecumenica 1 (Vienna, 2009), pp. 167-180.

All images of Manuscript U 338 are used with the permission of the Staatsbibliothek zu Berlin, where they reside:

Depositum der


in der

STAATSBIBLIOTHEK ZU BERLIN - Preussischer Kulturbesitz


Adam Bremer-McCollum is a Visiting Associate Professorial Specialist (Languages of Late Antiquity) in the Department of Theology at the University of Notre Dame.

[1] “How to Read ‘Infinite Jest’.”New Yorker. Nov. 5, 2018, 27. magazine/2018/11/05/how-to-read-infinite-jest.


[2] Cf. Dickens 2013: 303.


[3] Dickens, 304.


[4] Also belonging to Old Turkic are the language of some runiform inscriptions from the seventh to the tenth centuries as well as some later texts from the eleventh century under the Karakhanid Khanate, written mostly in Arabic script.


[5] Nina V. Pigulevskaya, “Fragments syriaques et syro-turcs de Hara-Hoto et de Tourfan,” Revue

de lOrient chrétien 30 (1935-1936): 3-46, and also the later part of Zieme’s book.


[6] Erdal, Grammar, p. 37. See further N. Sims-Williams, “The Sogdian Sound System and the

Origins of the Uyghur Script,” Journal Asiatique 269 (1981): 347-360; A. von le Coq, “Kurze Ein- fuhrung in die uigurische Schriftkunde,” Mitteilungen des Seminars for Orientalische Sprachen, West- asiatische Studien (1919): 93-109; György Kara, “Aramaic Scripts for Altaic Languages,” in Peter T. Daniels and William Bright, eds., The World’s Writing Systems (New York and Oxford, 1996), pp. 536-558 (esp. 539-545).


[7] Peter Zieme, Fragmenta Buddhica Uigurica: Ausgewählte Schriften von Peter Zieme, ed. Simone-

Christiane Raschmann and Jens Wilkens, Studien zur Sprache, Geschichte und Kultur der Turkvölker 7 (Berlin, 2009).


[8] “Religious Uygur texts, which are the majority, are normally translations, reformulations,

expansions etc. of texts in other languages; Chinese, Indic, Iranian or Tokharian if the text is Buddhistic, Iranian if it is Manichaean, Iranian or Syriac if it is Christian” (Erdal, Grammar, p. 23 n. 43).


[9] Gorgias Eastern Christian Studies 41 (Piscataway, 2015). See also his earlier overview, “Zu den

nestorianisch-türkischen Turfantexten,” in Georg Hazai and Peter Zieme, eds., Sprache, Geschichte und Kultur der altaischen Völker: Protokollband der XII. Tagung der Permanent International Altaistic Con- ference 1969 in Berlin, Schriften zur Geschichte und Kultur des alten Orients 5 (Berlin, 1974), pp. 661-668 (+ 4 plates). See also Pigulevskaya, “Fragments syriaques et syro-turcs.”


[10] Cf. Zieme 176, n. 610.


[11] See (sides 1-7) and http://turfan. (sides 8-20).


[12] 2015: 47.


[13] Note that the Syriac Psalter in Uyghur script, on which see below, has somewhat nicer hand- writing.


[14] Zieme 2015: 47.


[15] Sides 1 and 20 are the present (not original) “covers” of the booklet.


[16] But not without mistakes. This short Syriac intrusion perhaps derives from 1 Pe 2:9.


[17] On which see my post at


[18] “The fact that the thread binding is tied off on the two folios containing sides 6-7 and sides 16-17, along with examination of the crease lines on both folios, clearly indicates that they were the outer and inner folios respectively of the booklet when it was stitched together” (Dickens, 306).


[19] Dickens, 315.


[20] Dickens, 307-308, sees a problem in the text thus now starting on a verso side, but this does not strike me as so problematic. In any case, he has no fitting solution to the problem, if it is indeed one.


[21] As also noted by Dickens, 308, n. 15.


[22] “…texts discussed or listed in this collection are of an unquestionably utilitarian character and hence should be regarded as documentary sources” (den Heijer and Schmidt, 30).

[23] 308. See the photo of side 6, for a glimpse of this kind of binding.


[24] Mark Dickens and Peter Zieme, “Syro-Uigurica I: A Syriac Psalter in Uyghur Script from Turfan,”in  J.  den  Heijer,  A.  B.  Schmidt  and  T.  Pataridze  ed.,  Scripts  Beyond  Borders.  A  Survey  of Allographic Traditions in the Euro-Mediterranean World, ed., Leuven: Peeters, 2014, pp. 291-328, 629- 633 [plates 9-26].


[25] Wr. ʾšwytk.


[26] Or the sG mawhaβtā.


[27] Wr. dkbt.


[28] Hendiadys: soyurkayu irinčkäyä.


[29] Hendiadys: otlug yalïnčïg.


[30] It is unclear how these three words in Uyghur relate to the preceding, and it may be that something is missing.


[31] Wr. ʿmmʾ, for ʿmʾ.


[32] Wr. pryqʾ. The word could thus also be briχā.


[33] Wr. gnšʾ.


[34] Dickens says the Syriac intrusion “sounds like a catch phrase used by the Christians to describe themselves” (316-317). This may be true, but without further evidence, it is not helpful. Zieme first suggested this Syriac part was inspired by Ps 72:17, but Dickens convincingly demonstrates that 1Pet 2:9 is a better fit. Zieme mentions Dickens’ proposal but is ultimately undecided about the definite source: “…sie ist offenbar nur eine Paraphrase nach türkischem Stil mit einem strengen Parallelismus” (2015: 58-59).


[35] The first month.


[36] As rightly mentioned by Dickens in a note (309, n. 18), the reference to merit-transfer, even though it indeed uses a Sanskrit loanword —and, it should be noted, one not uncommon in Uyghur

—need not imply anything exactly resembling the Buddhist practice, especially since we often find

an analogous practice in Syriac, including Church of the East, scribal traditions. For the Buddhist

practice, see George J. Tanabe, Jr., “Merit and Merit-making,” in Encyclopedia of Buddhism, vol. 2, 532-534.


[37] As acknowledged by both Zieme and Dickens, the exact meaning of the colophon’s last line is

enigmatic. My translation is only a guess at a possible meaning.


[38] Unfortunately, status-quo assumptions about the usual —more quantitatively observable —or the “proper” script can lead us to de-value certain language + script combinations. It is, of course, those who have some kind of (to them) advantageous power over the situation —e.g. politically, numerically, racially, and/or in terms of class —who, wittingly or not, generally control the nexus of language + script + normativity.


[39] What about Christian Palestinian Aramaic, a language quite distinct from Syriac, albeit written in a similar, yet still different, script? Is it “Syriac script” (cf. Syriac in Melkite script)? Or is it distinct enough in ductus, including the reversed peh?


[40] Because the ductus was/remained even more Greek than that of Coptic; cf. Browne.


[41] Each of these two scripts has been used also for Persian, and, of course, several other languages (Syro-Turkic). While other factors —such as the sometimes frequent presence of loanwords from the prestige community (i.e. Hebrew and Syriac) —characterize these language versions, in this case we are especially focusing on a particular combination of script + language.


[42] But cf. Kiraz, 67.


[43] Cf. Dickens, 304 n. 6, and 315, n. 67.


[44] Zieme 2009: 170.


[45] Also, the plural marker (the syāmē dots of Syriac) are often, but not always, missing where we expect them (e.g. 11.3 for ydyʿʾ, 13.2 bʾydyn, and elsewhere). We should mention, though, that other Aramaic dialects get by fine with no recourse to something like syāmē, and so, while the syāmē dots are conventionally obligatory in many cases, they are hardly necessary from a linguistic point of view. The scribe also has no problem with splitting Syriac words across a line-break —e.g. qd|yštʾ in 08.2-3 and dʾ|yt(w)hw in 09.3-4 —something rare in typical Syriac scribal tradition, but not unusual in Uyghur writing.


[46] Luijendijk, 42, 46, 51. She adds, “The way I usually indicate its size is by holding out the palm of my hand”(51).


[47] Luijendijk, 51


[48] Luijendijk, 56.


[49] Catalogue of the Ethiopic Manuscripts in the British Museum (1877), ix.


[50] Luijendijk, 43, 44, and 46.


[51] Luijendijk 46, 66 n. 48), and she further cites D. Frankfurter, “Ritual Expertise,”128: “…but Coptic in the era of these texts was for the most part a monastic writing system.”And she adds, “… the reading skills necessary to operate the book hint at a monastic milieu”(Luijendijk, 68).


[52] The name is, of course, pre-Christian, but its use as a monastic name is due to one of the famous pair Sergius and Bachhus.


[53] For “ethnolect”and “religiolect”, see den Heijer and Schmidt, 46, following Benjamin Hary, and see den Heijer and Schmidt, 47-49 for issues of language proficiency and education. This also has bearing on the apparent monastic context of the booklet.


[54] den Heijer and Schmidt, 56.


[55] Rodolphe Kasser and Philippe Luisier, “P. Bodmer XLIII : Un feuillet de Zostrien,” Le Muséon 120 (2007): 251-272, here esp. 251. Images of the leaf are at constellations/papyri/barcode/1072205351.


[56] Bilingual amulets are known, such as the Greek-Aramaic silver amulet, on which see R. Kotan-

sky, J. Naveh, and S. Shaked, “A Greek-Aramaic Silver Amulet from Egypt in the Ashmolean Museum,” Le Museón 105 (1992): 5-25.


[57] “Die Wendung ‘es funfmal schreibend’ kann man vielleicht als eine Schulübung ansehen, ob- wohl dies der eingangs erwähnten Möglichkeit eines Amuletts widerspräche” (Zieme 2015: 47).


[58] A line that immediately comes to mind is Ḥarbā d-Mōšē § 128 (Harari 1997: 45-46): בעית אם

יומא  בת  ביצה  על  ’כת  אלתר  תילף  דתשמע  כל,  “If  you  want  to  learn  everything  you  hear immediately, write on a freshly laid egg…”


[59] See further Dickens, 312-313. The text was first identified by Mar Awa, Bishop of California for the Assyrian Church of the East.


[60] Cf. Dickens, 315-316.


[61] Dickens, 305. Cf. here Zieme 2009, 170.


[62] 317.

[63] This sentence, however imperfectly worded it may be, was much inspired by Luijendijk’s resonating line: “These explorations of the codex as object have begun to bring us into contact with…those who handled this booklet”(56).



follow us in feedly