THE IMPORTANCE OF THE PSALTER AT TURFAN

Mark DICKENS1
University of Alberta, Canada

The East Syriac Psalter

The Psalter has always played a major role in Christian liturgy, especially in
monastic contexts. This is as true of the Church of the East as it is of other
Christian traditions. Indeed, a glance through Macleanřs English translation of
East Syrian Daily Offices clearly shows how central the recitation of the Psalms
was in the daily worship of the church.2 Equally as revealing is the fact that the
printed edition of the Ḥudra, the cycle of services appointed to be said in the
Church of the East throughout the year, includes the entire Psalter at the back of
each of the three large volumes it comprises.3 Therefore, it is not surprising that
we have abundant evidence of the use of the Psalter at the Christian monastery
of Shüipang near Bulayïq where, beginning in 1905, so many Christian manu-

1
This research is based on the authorřs participation in The Christian Library of Turfan re-
search project, funded by the Arts and Humanities Research Council (AHRC), UK and headed
up by Dr. Erica Hunter, who, along with the present author, will produce a catalogue of all
known Syriac fragments from Turfan. Thanks to the following individuals who provided help-
ful information or answered questions by email during the preparation of this article: Sebastian
Brock, Erica Hunter, Desmond Durkin-Meisterernst, Sergey Minov, Simone Raschmann,
Salam Rassi, Christiane Reck, Steven Ring, David Royel (Mar Awa), Aho Shemunkasho,
Nicholas Sims-Williams, Werner Sundermann, David Taylor, Herrie van Rooy and Peter
Zieme. Thanks to the Staatsbibliothek zu Berlin - Preussischer Kulturbesitz and the Berlin-
Brandenburgische Akademie der Wissenschaften for access to and permission to reproduce
images of the relevant fragments. All images are copyright Depositum der BERLIN-
BRANDENBURGISCHEN AKADEMIE DER WISSENSCHAFTEN in der
STAATSBIBLIOTHEK ZU BERLIN – Preussischer Kulturbesitz, Orientabteilung. Low
resolution images of the various signature numbers are available on the following websites:
SyrHT on the International Dunhuang Project website (http://idp.bl.uk/, enter the signature
number in the search box); n, So or U on the Digitales Turfan-Archiv website
(http://www.bbaw.de/forschung/turfanforschung/dta/index.html, follow the relevant link in the
list). MIK images (from the Museum für Asiatische Kunst) are not currently available to the
public.
2
MACLEAN, 1894, passim, esp. 260Ŕ263. See also MATEOS, 1959, 450Ŕ456.
3
DARMO, 1960Ŕ1962.
358 Mark Dickens

script fragments were discovered by the Second and Third Prussian Turfan Ex-
peditions (1904Ŕ1907).4
The Turfan Psalter finds are important for at least two reasons. First, no other
Christian text found at Turfan was rendered in more languages and scripts than
the Psalter5 and second, the Syriac Psalter fragments from Turfan are amongst
the earliest extant anywhere and can therefore be helpful in tracing the devel-
opment of the Syriac Peshiṭta text of the Psalms.6 Before looking at the Psalters
found at Turfan, an overview of the Psalter in the East Syriac tradition is in or-
der.7 The Psalter itself is usually referred to as either 1) ‫( ܟܮܒܐ ܕܕܘܝܕ‬kthābhā d-
dhawidh), Ŗthe Book of Davidŗ (or elliptically, just ‫ܕܘܝܕ‬, ŖDavidŗ) or 2) ‫ܟܮܒܐ‬
‫( ܕܣܙܣܘ̈ܪܐ‬kthābhā d-mazmurē), Ŗthe Book of Psalms.ŗ 8 The term ‫ܣܙܣܘܪܐ‬
(mazmurā, pl. mazmurē) is used most commonly to designate each Psalm in the
East Syriac Psalter.9
Although other Syriac translations of the Bible have been used in the West
Syriac tradition, the primary translation used in the East Syriac tradition has
always been the Peshiṭta. The Psalm numbering in the Peshiṭta differs from that
of both the Hebrew Masoretic text (followed by most modern translations of the
Bible) and the Septuagint; from Psa. 115 to Psa. 147, the Peshiṭta numbering is
one behind the Hebrew numbering. 10 Furthermore, there are five additional
Psalms (Psa. 151Ŕ155) attested in some manuscripts of the Peshiṭta Psalter, but
these have not yet been encountered amongst the Turfan fragments.
The Psalter is divided up in different ways in the East Syriac and West Syriac
traditions (Church of the East and Syrian Orthodox Church, respectively). The
former divides the text into 20 major sections, each called a ‫( ܗܘܠܡ ܐ‬hulālā, pl.
hulālē), Ŗpraising, chant, hymn,ŗ from the same root as Hebrew hallelujah. The

4
On Bulayïq, see SIMS-WILLIAMS, 1990. On the Turfan expeditions, see BERLIN-BRANDENBURG
ACADEMY OF SCIENCES AND HUMANITIES, 2007.
5
The only other text which comes close is the Legend of St. George, found in Syriac (MARÓTH,
1991), Sogdian in Syriac script (HANSEN, 1941) and Uyghur Turkic (BANG, 1926).
6
Critical edition published by the Peshiṭta Institute, Leiden in WALTER, 1980.
7
On the Syriac Bible in general, including the Psalter, see BROCK, 2006. ŖEast Syriacŗ is used
throughout this article as an adjective referring to the Church of the East, in preference to the
problematic adjective ŖNestorian.ŗ
8
Transliteration of Syriac follows the classical language, including the spirantised sounds bh, gh,
dh, kh, ph and th, all of which are underlined in the transliterations.
9
Other Syriac terms for ŖPsalmŗ include ‫( ܙܘܣܪܐ‬zumārā) and ‫( ܙܣܝܪܬܐ‬zmirthā), all from the root
‫( ܙܣܪ‬zmar), Ŗto sing, chant or play on a stringed instrument.ŗ
10
Although in truth it is slightly more complicated than this: Psa. 114 and 115 in the Hebrew text
are joined together as Psa. 114 in the Peshiṭta, whereas Psa 147:1Ŕ11 and 147:12Ŕ20 in the
Hebrew are Psa. 146 and 147 in the Peshiṭta. For a helpful chart setting out the differences, see
BROCK, 2006, 138. See also MATEOS, 1959, 447. Note also that versification of individual
Psalms in the Peshiṭta frequently differs from that found in English Bibles, with many Psalms
starting at verse 2 or even 3, rather than verse 1.
The Importance of the Psalter at Turfan 359

Odes or Canticles found at the end of the Psalter (on which, see below) form
hulālā 21.11 Each hulālā is further divided into 2Ŕ4 subsections, each known as
a ‫( ܣܪܣܝܮܐ‬marmithā, pl. marmayāthā). There are 57 marmayāthā in the whole
Psalter and each marmithā contains 1Ŕ6 Psalms, except for Psa. 118 (Psa. 119
in the Hebrew text), which is divided into two marmayāthā. It is thought that
the division into marmayāthā precedes the division into hulālē.
There are several additional elements found in the East Syriac Psalter which are
relevant to the Turfan finds: 1) Psalm headings or titles; 2) Psalm farcings or
canons; 3) Prayers in the Psalter text; and 4) Odes or Canticles from other Old
Testament books.
Psalm headings or titles: The Psalms in both the Hebrew Masoretic text and
the Septuagint have headings, understood to have been added at some point af-
ter the composition of each Psalm to indicate the author and sometimes the cir-
cumstances in which it was written. The Psalms in the Peshiṭta text originally
did not include headings, but these were added in the 5th century or later. They
are different from those in the Hebrew or Septuagint text and moreover they
differ between the East Syriac and West Syriac traditions. The East Syriac
Psalm headings are derived from the commentaries of Theodore of Mopsuestia
on the Psalms and primarily give information on the historical setting for each,
although many also indicate a prophetic dimension to the Psalm in question,
stating that it was written about some future time in Jewish history, such as the
time of the Maccabees. Example from Psa. 1: ‫ܙܦܝܐ‬ ̈ ‫ܬܫܥܝܮܐ ܘܣܪܬܝܧܘܬܐ ܕܥܢ‬
‫ܫܧܝ̈ܪܐ‬, ŖAn account and exhortation about the virtuous ways.ŗ12
Psalm farcings or canons: The Syriac term ‫( ܩܧܘܦܐ‬qānonā, pl. qānonē),
Ŗcanon,ŗ has many uses, but in this case refers to sentences which are usually
inserted after the first verse of each Psalm or between the first and second half
of the first verse. These are attributed to Patriarch Mar Aba I (540Ŕ552), a con-
vert from Zoroastrianism who travelled to Nisibis, Edessa, Antioch, Egypt and
Constantinople before becoming Catholicos (Patriarch) in 540. Example from
Psa. 1: ‫ܛܘܒ ܠ ܐܝܧܐ ܕܛܥܨ ܦܝܪܟ ܘܐܬܗܓܝ ܒܧܤܘܩܟ ܣܪܝܐ ܒܡܡܝܐ ܘܒܐܝܤܤ ܐ‬,
ŖBlessed is he who has borne your yoke and has meditated on your law, O Lord,
by night and by day.ŗ13
Prayers in the Psalter text: The East Syriac Psalter also includes prayers in-
serted before each new marmithā or hulālā. Example from Psa. 1: ‫ܐܫܘܐ ܠܨ‬
‫ ܘܦܗܘܐ ܨܒܝܧܨ‬.‫ܕܫܧܪܝܨ ܠܤܪܘܬܟ‬ ‫ܣܪܢ ܘܐܠܗܢ ܕܦܮܕܒܪ ܒܕܘܒ̈ܪܐ ܕܣܝܮܪܘܬܐ‬
‫ ܐܒܐ ܘܒܪܐ ܘܪܘܚܐ‬.‫ܣܪܐ ܕܟܡܝ‬ ‫ ܘܒܗ ܦܮܗܓܐ ܒܡܡܝܐ ܘܒܐܝܤܤ ܐ‬.‫ܒܧܤܘܩܟ‬
11
See MACLEAN, 1894, 259.
12
Syriac text from DARMO, 1960, 241. The Syriac text of all the headings (without translations)
can be found in BLOEMENDAAL, 1960.
13
Syriac text from DARMO, 1960, 241. Translations of all the farcings/canons can be found in
MACLEAN, 1894, 236-248 and MATEOS, 1959, 315Ŕ338.
360 Mark Dickens

‫ܕܩܘܕܫܐ ܠܥܡܤܝܨ‬, ŖMake us worthy, O our Lord and our God, to be led in the
ways of virtue that are pleasing to your lordship, and may our desire be in your
law, that we would meditate on it by night and by day, O Lord of all, eternal
Father, Son and Holy Spirit.ŗ14
Odes or Canticles: In addition to the 150 Psalms, East Syriac Psalters typically
include several Odes or Canticles, in Syriac ‫( ܬܫܒܘܚܮܐ‬teshboḥtā, pl.
teshbḥāthā), consisting of the following Psalm-like passages from elsewhere in
the Old Testament, normally placed at the end of the Psalter and considered to
form hulālā 21. In the East Syriac (and Maronite) tradition, the normal order is:
1) Ex. 15:1Ŕ21 (the First Song of Moses); 2) Deut. 32:1Ŕ43 (the Second and
Third Songs of Moses, sometimes divided into verses 1Ŕ20 and 21Ŕ43); 3) Isa.
42:10Ŕ13; 45:8 (the Song of Isaiah). The oldest known East Syriac manuscripts
of the Psalter (e.g. 8a1, an 8th century complete Bible) have them in this order,
but later manuscripts (e.g. 12t4, a 12th century manuscript containing the Psalms
and Odes) put the Song of Isaiah second, a practice evident in SyrHT 96, the
only Turfan Psalter fragment with more than one Ode extant.15

Role of the Psalter in the Church of the East

The Psalter is integral to the liturgy of the Church of the East, with Psalms be-
ing used at numerous places in the service, whether on festival days (including
Sundays, saintsř days, feast days and fasts) or ferial days (ordinary weekdays).
In non-monastic communities, each day traditionally consisted of four daily
services: 1) ‫ܬܫܤܬܮܐ ܕܪܣܬܐ‬, teshmeshtā dh-ramshā, ŖEvening Service, Ves-
persŗ; 2) ‫ܩܘܒܥܐ‬, subhā„ā, ŖComplineŗ; 3) ‫ܨܠܘܬܐ ܕܠܡܝܐ‬, ṣlothā dh-lelyā,
ŖNight Prayer, Nocturnsŗ; and 4) ‫ܨܠܘܬܐ ܕܨܦܪܐ‬, ṣlothā dh-ṣaphrā, ŖMorning
Prayer, Matins.ŗ However, the practice of Compline has gradually declined over
the centuries, so that today it is virtually non-existent.16
Two marmayāthā (usually totalling six or seven Psalms) are recited near the
beginning of ferial evening services (Vespers).17 After the Old Testament read-
ing (called ‫ܩܪܝܧܐ‬, qeryānā), several verses from a Psalm (called ‫ܫܘܪܝܐ‬, shurāyā,
pl. shurāyē), introduce an anthem or hymn. Many anthems also typically begin
with a verse from the Psalms. Prayers are interspersed with more Psalms, fol-

14
Syriac text from DARMO, 1960, 241. Translations of all the prayers can be found in MACLEAN,
1894, 86Ŕ95.
15
For a list of Psalter manuscripts used in the preparation of the Leiden critical edition, see
WALTER, 1980, viiŔxxix. A critical edition of the Odes has also been published by the Peshiṭta
Institute, Leiden in SCHNEIDER, 1972.
16
On the hours of prayer in the Church of the East, see BADGER, 1852, 16Ŕ18.
17
The following description of the role of the Psalter in various services is based on MACLEAN,
1894. See also the descriptions in BADGER, 1852, 21-22; TAFT, 1986, 225Ŕ237.
The Importance of the Psalter at Turfan 361

lowed by a second shurāyā and anthem. Other verses from the Psalms are scat-
tered throughout the remainder of the service, often used to introduce specific
liturgical components. This is especially true of the Martyrsř Anthems, where
each clause is introduced with a Psalmic quotation.18 Festival evening services
differ slightly, with only one marmithā near the beginning, three shurāyē and no
Martyrsř Anthem. Again, various prayers and Psalms are interspersed through-
out other parts of the service.19
The entire Psalter is recited twice through each week during the ferial night
service (Nocturns): hulālē 1Ŕ7 (Psa. 1Ŕ58) on Monday, hulālē 8Ŕ14 (Psa. 59Ŕ
101) on Tuesday and hulālē 15Ŕ21 (Psa. 102Ŕ150, plus the Odes) on Wednes-
day, the same pattern then repeating for Thursday, Friday and Saturday.20 Addi-
tionally, when the Eucharist is celebrated during this service, an anthem called
the ‫( ܩܡܮܐ‬qālthā) is sung, which includes 1Ŕ3 Psalms. Later on in the night ser-
vice, the ‫( ܫܘܒܛܐ‬shubhḥ ), a Psalm of praise, is recited and on Wednesday
nights, the ‫( ܣܘܬܒܐ‬mawtbhā), an anthem sung while sitting, features verses
from the Psalms at the beginning of each clause, similar to the Martyrsř An-
thems.21 Again, festival night services are different. On feast days, the entire
Psalter is recited, divided up into three groups of hulālē: 1Ŕ11 (Psa. 1Ŕ81), 12Ŕ
17 (Psa. 82Ŕ119:88) and 18Ŕ21 (Psa. 119:89Ŕ150, plus the Odes). On Sundays,
either hulālē 5Ŕ11 (Psa. 37Ŕ81) or 12Ŕ18 (Psa. 82Ŕ131) are recited at the begin-
ning of the service, followed later on in the service by either hulālē 12Ŕ14 (Psa.
82Ŕ101) or 19Ŕ21 (Psa. 132Ŕ150, plus the Odes), respectively. As with the
ferial night service, the Eucharist is accompanied by the qālthā, with different
Psalms appointed to be sung for different Sundays in the church year.22
Psalms interspersed with prayers begin both the ferial and festival morning
services (Matins) and, as with the ferial evening service, each clause of the
Martyrsř Anthem is introduced with a quotation from the Psalms. 23 The Psalms
also play an important role in the Compline service (occurring between Ves-
pers and Nocturns); again, interspersed Psalms and prayers begin the service
and verses from the Psalms introduce the Anthems of the Departed recited at
this service.24 Finally, the whole Psalter is read through twice weekly during the

18
For an example of a ferial evening service, see MACLEAN, 1894, 1Ŕ23.
19
For an example of a festival evening service, see MACLEAN, 1894, 68Ŕ84.
20
MACLEAN, 1894, 86. Following Maclean, Psalm numbers here follow the Hebrew text, not the
Peshiṭta.
21
For an example of a ferial night service, see MACLEAN, 1894, 85Ŕ86, 95Ŕ102. For the mawtbhā,
see MACLEAN, 1894, 130Ŕ150.
22
For an example of a festival night service, see MACLEAN, 1894, 151Ŕ163.
23
For an example of a ferial morning service, see MACLEAN, 1894, 103Ŕ109 (followed by the
morning Martyrsř Anthems: 109-130). For the festival morning service, see MACLEAN, 1894,
164Ŕ184.
24
For an example of a Compline service, see MACLEAN, 1894, 185Ŕ190 (followed by the An-
362 Mark Dickens

Great Fast of Lent and once daily during the Rogation of the Ninevites. Psalms
are also recited at the beginning of baptismal, Eucharist and burial services.25
As noted above, this pattern of three or four daily services (Vespers, Nocturns,
and Matins, along with Compline at an earlier date) has been fairly typical in
the Church of the East throughout most of its existence. In monastic communi-
ties, however (as at Turfan), there have traditionally been seven daily services.
This practice is traced back by ŘAbdisho bar Berikha (d. 1318), metropolitan of
Nisibis and Armenia under Patriarch Yahballaha III, to ŖJesus Christ, our good
and merciful God and Lordŗ and Ŗthe Catholic Fathersŗ who Ŗappointed the
same for monks and anchorites, and their successors ordained that each of these
seven services should consist of three hoolâlé.ŗ26 Thus, theoretically at least, the
whole Psalter (21 hulālē) was recited each day in monasteries of the Church of
the East. The above summary should suffice to demonstrate that the liturgy of
the Church of the East, particularly in monastic communities, was and is uni-
maginable without the Psalter.
The use of the Psalms in a monastic community is further illustrated in the Book
of Governors (Historia Monastica) by Thomas of Marga (ca. 840). The author
tells us how the Psalms were sung in services (sometimes so Ŗsweetlyŗ that the
preacher was unable to concentrate on the sermon he was to preach),27 recited
during funerals,28 chanted by schoolboys who were educated in monastic com-
plexes (sometimes to the ire of the ascetics trying to meditate in their cells),29
sung by monks while travelling in the countryside,30 recited ceaselessly (along
with hymns) by ascetics in solitude,31 taught to new converts,32 and even sung
as an accompaniment to the performing of miracles.33 Thomas also tells us of
the monastic reforms instituted by Babai of Gebhilta (early 8th cent.), including
the standardization of the tunes used in singing songs, prayers, Psalms, antipho-
nal responses and anthems.34 Most, if not all, of these practices, including the
musical reforms, were likely present in the monastic community at Turfan.

thems of the Departed: 190Ŕ204).

25
See MACLEAN, 1894, 205Ŕ224; BADGER, 1852, 22.
26
BADGER, 1852, 17. See also BUDGE, 1893a, lv, cxxxvi, cxlviii. The Syriac text, with an accom-
panying Latin translation, can be found in MAI, 1838, 81 (Lat.), 245Ŕ246 (Syr.). Thanks to
Steven Ring and Salam Rassi for their assistance in locating this information.
27
BUDGE, 1893b, 120, 443, 444, 541Ŕ542, 611, 620.
28
BUDGE, 1893b, 122.
29
BUDGE, 1893b, 148, 221Ŕ222, 581.
30
BUDGE, 1893b, 338, 463, 506, 515, 516.
31
BUDGE, 1893b, 442, 455, 502Ŕ504, 569Ŕ570, 587, 641.
32
BUDGE, 1893b, 512Ŕ513, 609.
33
BUDGE, 1893b, 671.
34
BUDGE, 1893b, 290-294; BUDGE, 1893a, lviiŔlviii, cviiiŔcix. WRIGHT, 1894, 184Ŕ185 mistak-
enly identifies him with Babai of Nisibis.
The Importance of the Psalter at Turfan 363

Non-Syriac Psalters from Turfan

Although most Turfan Psalter fragments are in Syriac, there are also remnants
of Psalters in several other languages and scripts used in Iran and Central Asia
up to the Mongol era, namely the Pahlavi (Middle Persian), Sogdian, and New
Persian languages and the Uyghur script (although not the Uyghur language, as
we shall see below).35
The famous Pahlavi Psalter consists of 12 folios or part-folios, written in a
version of Pahlavi script more archaic than Book Pahlavi and containing por-
tions of Psa. 94Ŕ99, 118 (= Hebrew Psa. 119) and 121Ŕ136 (= Hebrew Psa.
122Ŕ137). Although generally a literal translation of the Peshiṭta, it occasion-
ally agrees more with the Hebrew text or the Septuagint. It includes the canons
of Mar Aba, but not the Psalm headings or prayers. The date of the text is uncer-
tain; the inclusion of the canons indicates it must come from after the mid-6th
century, but the original translation of the biblical text may have been composed
earlier than that.36 Of special interest is the fact that the form of Pahlavi script
used in the Psalter has also been found on a processional cross from Herat
which mentions a date of 507 or 517 (possibly a reference to the Bactrian era,
which began in 223 CE, giving a CE date of 730 or 740 for the cross).37
There are remnants of two Sogdian Psalters from Turfan, one with Syriac ru-
brics and another with the opening words of each Psalm in Greek. The former
consists of 15 fragments, including three that have been lost and one folio (So
12601 = MIK III 59) with the end of a prayer and the complete text of the Ni-
cene Creed, suggesting that the original manuscript contained more than just the
Psalter.38 The fragments are written in Sogdian script and contain portions of
Psa. 5Ŕ6, 19Ŕ20, 23Ŕ24, 28Ŕ30, 32, 33, 50, and 51. They are clearly translated
from the Peshiṭta and include the Psalm headings, but not the canons or prayers.
The first verse of each Psalm is given in both Syriac (red ink, in Syriac script)
and Sogdian (black ink, in Sogdian script).39
The second Sogdian Psalter from Turfan, also written in Sogdian script, is rep-
resented by only two fragments: So 12955 (MIK III 56), containing Psa. 33:1Ŕ
4, 8Ŕ10, and So 12950(2), containing Psa. 34:7Ŕ9, 14Ŕ16. The Sogdian transla-

35
See the Uyghur Psalter, under the next section.
36
On the Pahlavi Psalter, see ANDREAS, 1910; ANDREAS & BARR, 1933, ASMUSSEN, 1964;
GIGNOUX, 1968; SKJAERVØ, 1983; GIGNOUX, 2008. Carbon dating has also been used to date
the paper, but the findings have not yet been formally published.
37
On the Herat cross, see GIGNOUX, 2001. On the Bactrian era, see DE BLOIS, 2006.
38
The creed was published in MÜLLER, 1913, 84Ŕ87. An English translation and short commen-
tary can be found in GILLMAN & KLIMKEIT, 1999, 252Ŕ253. For the signature numbers of all
15 fragments, see RECK, 2008, 198.
39
On the Sogdian Psalter, see SCHWARTZ, 1967, 126-144; SCHWARTZ, 1974; SCHWARTZ, 1982,
158Ŕ189, 200Ŕ207; RECK, 2008, 192Ŕ193, 198Ŕ199.
364 Mark Dickens

tion follows the Peshiṭta in some places and the Septuagint in others. The open-
ing words of each Psalm are in Greek as a headline in the upper margin of each
fragment (the Greek writing possibly dating from the 9th century); this is clear
on So 12955, but only traces remain on So 12950(2). Since the extant fragments
do not contain the beginning of each Psalm, it is unclear if the Psalm headings,
canons or prayers were included in this Psalter. The Greek headlines suggest
that it originated in Ŗa Melkite community in a Sogdian-speaking environment,ŗ
possibly Tashkent.40
There are also two fragments, representing portions of a double folio from a
bilingual Syriac-New Persian Psalter written in Syriac script: MIK III 112
and SyrHT 153, containing Psa. 131:18Ŕ132:1; 133:1Ŕ3; 146:5Ŕ147:7 (= He-
brew Psa. 132:18Ŕ133:1; 134:1Ŕ3; 147:5Ŕ18). Again, it follows the Peshiṭta text
and, like the Sogdian Psalter, includes the Psalm headings, but not the canons or
prayers. Typically, verses are divided up into two phrases (called colons, on
which see below), with the Syriac text followed by a New Persian translation in
a modified Syriac script very similar to that used for Christian Sogdian. Signifi-
cantly, this is the earliest known translation from Syriac into New Persian.41

40
SIMS-WILLIAMS, 2004, 631. This article discusses So 12995; So 12950(2) will be dealt with in
a forthcoming article by the same author. See also the discussion of these fragments in
DřAIUTO, 2008 [2009]. On the Melkite community in Central Asia, see DAUVILLIER, 1953;
NASRALLAH, 1976; KLEIN, 1999. On connections between Byzantine Christianity and the
community at Bulayïq, see SUNDERMANN, 1994. Obviously, if this originally came from a
Melkite community, it would not have the East Syriac Psalm headings, canons or prayers.
41
MIK III 112, consisting of the upper half of both folios, was published in MÜLLER, 1915 and
discussed in BENVENISTE, 1938. SyrHT 153, the bottom half of one folio, was published in
SUNDERMANN, 1974, in which the author showed how the two fragments joined together.
The Importance of the Psalter at Turfan 365

Syriac Psalter Fragments from Turfan42

In all, there are 446 signature numbers from Turfan with Syriac only on them,
plus an additional 42 that are Syriac mixed with Sogdian, New Persian or Uy-
ghur (including bilingual texts), for a total of 488 signature numbers.43 Most
Syriac fragments are from liturgical books, although other genres are repre-
sented, including prayer booklets and hagiographies. So far, 50 fragments have
been identified as Syriac or bilingual Psalters, representing slightly over 10% of
the Syriac only and bilingual texts.44 A number of fragments are identified on a
typed hand-list of the Syriac fragments in the Turfan Collection 45 as coming

42
Unless otherwise noted, all Psalter fragments listed here have been identified by the present
author, with the help of SPRENGER, 1976. The ongoing task of identifying the Turfan Syriac
fragments will likely reveal more Psalter fragments to add to this list. Excluding fragments
from bilingual Syriac-Sogdian lectionaries (on which, see MÜLLER, 1907; MÜLLER, 1913;
WELLESZ, 1919; BURKITT, 1925, 119Ŕ125; SUNDERMANN, 1981), there are only a handful of
biblical fragments in Syriac from Turfan that are not from the Psalter. SyrHT 2 consists of two
fragments (T II B 18 & T II B 62) containing the remnants of a letter. The back side of T II B
18 contains Prov. 9:14Ŕ10:11, the only non-Psalter Old Testament text found at Turfan thus far
(the letter is published in MARÓTH, 1985). SyrHT 48 & 49 is the sole representative from Lec-
tionary ŖA,ŗ a double-folio containing the text of Rom. 1:24Ŕ2:6; 5:12Ŕ21; 7:1Ŕ7, the epistle
readings for the First Tuesday in Lent, the Second Sunday in Lent and the Second Friday in
Lent (MACLEAN, 1894, 270Ŕ271) in what looks like the same hand as Lectionary ŖB,ŗ Psalter
ŖCŗ and Psalter ŖK.ŗ SyrHT 123 is a single folio containing John 3:21Ŕ36, probably from a
New Testament, since the lectionary readings (for the fifth and sixth Sundays of Epiphany) di-
vide the text between verses 21 and 22 (MACLEAN, 1894, 267, 269; DIETTRICH, 1909, 165).
SyrHT 241, SyrHT 277, SyrHT 300, and n 327 all come from the same original folio, the
sole remnant from Lectionary ŖB,ŗ containing Luke 1:1Ŕ21, part of the gospel reading for the
First Sunday of Advent (MACLEAN, 1894, 265; DIETTRICH, 1909, 164; BURKITT, 1921Ŕ1923,
316). As noted above, this appears to be from the same hand as Lectionary ŖB,ŗ but the folio
dimensions are different. Four other fragments from this hand have not yet been identified:
SyrHT 17, SyrHT 370, n 311 and n 438.
43
These include SyrHT signature numbers (in the Staatsbibliothek zu Berlin), n, So and U sig-
nature numbers (in the Berlin-Brandenburgische Akademie der Wissenschaften) and MIK sig-
nature numbers (in the Museum für Asiatische Kunst). Not included are 30 Sogdian or New
Persian fragments in Syriac script which include Syriac words or phrases. Note that the signa-
ture numbers cover everything from small fragments with remnants of just one or two words to
a 61-folio manuscript in the Museum für Asiatische Kunst, so an accurate assessment of per-
centages is difficult.
44
Removing the bilingual texts from the mix does not significantly change the percentage of
Psalter fragments. A rough estimate at this point in deciphering the Syriac only and bilingual
fragments is that 80% are liturgical, 15% are biblical (Psalters and lectionaries) and 5% are
other genres.
45
Syrische Turfan-Fragmente der Deutschen Akademie der Wissenschaften zu Berlin. The com-
piler and date of compilation are not indicated on the list, but it was probably after World War
II, due to the frequent reference to missing fragments. Various suggestions have been made as
366 Mark Dickens

from various Psalters.46 Where fragments were readily identifiable as coming

from specific original Psalter manuscripts, they were designated Psalter ŖC,ŗ
Psalter ŖDŗ and Psalter ŖEŗ on this hand-list (and a separate one compiled by
someone else at a later date).47
The anonymous hand-list compilerřs system for identifying Psalters with capital
letters is unclear. Specifically, why are Psalter ŖAŗ and ŖBŗ not on the list? Al-
though a definitive answer is not possible, a plausible suggestion has been made
by Nicholas Sims-Williams, who has suggested that Psalter ŖAŗ and ŖBŗ may
refer to the Pahlavi Psalter and the bilingual Syriac-New Persian Psalter, since
these were both discovered and published early on in the history of the Turfan
Collection, before the hand-list was compiled (in contrast to the Sogdian Psal-
ters, which were not identified until the 1960řs and afterward).48 Based on this
supposition, other Psalters in the Turfan Collection that are not specifically la-
belled on the hand-list have been identified with the letters ŖFŗ through ŖM.ŗ
Psalter “C” is one of two Psalter booklets extant in the collection. In addition
to the nine folios still stitched together (SyrHT 72), there are also four separate
fragments which can either be joined to a folio or fit between some of the non-
adjacent folios in the booklet.49 It has the headings, canons and prayers, as well
as quire marks and marginalia (including the indication of new marmayāthā). It
is also one of the few Syriac manuscripts from Turfan where the text box is
ruled. The following Psalms are contained in the extant folios:
SyrHT 72, fol. 1-2 & SyrHT 348 = Psa. 72:18Ŕ74:1
SyrHT 72, fol. 3 = Psa. 117:23Ŕ118:4 (= Hebrew Psa. 118:23Ŕ119:4)
SyrHT 228 & SyrHT 379 = Psa. 137:7Ŕ138:8 (= Hebrew Psa. 138:7Ŕ
139:8)
SyrHT 79 = Psa. 140:1Ŕ141:2 (= Hebrew Psa. 141:1Ŕ142:2)
SyrHT 72, fol. 4–5 = Psa. 143:5Ŕ144:17 (= Hebrew Psa. 144:5Ŕ145:17)

to the compiler, including Olaf Hansen, who worked with the Christian Sogdian fragments
from the 1940řs to 1960řs, or an unidentified student of Anton Baumstark. Thanks to Werner
Sundermann and Peter Zieme for these suggestions.
46
Specifically those now numbered SyrHT 62, SyrHT 63 & 64, SyrHT 71, SyrHT 72, SyrHT
90 & 91, SyrHT 92 & 93, SyrHT 96, SyrHT 108, SyrHT 119, SyrHT 129, SyrHT 153,
SyrHT 180, and SyrHT 191.
47
Unfortunately, as with the first (older) hand-list, the name of the compiler and date of compila-
tion of this second list are unknown. Various people, including Werner Sundermann, added
notes about the contents of the fragments to both hand-lists. Thanks to Simone Raschmann for
this information.
48
This suggestion was made when the present author presented an initial version of this paper at
the ŖChristian Manuscripts from Turfanŗ workshop organized as part of the aforementioned
AHRC project in Berlin, March 27Ŕ28, 2009.
49
SyrHT 72, under its old signature number T II B 24, is discussed briefly in SUNDERMANN,
1975, 86Ŕ87.
The Importance of the Psalter at Turfan 367

SyrHT 72, fol. 6-9 = Psa. 145:9Ŕ150:4 (= Hebrew Psa. 146:9Ŕ150:4)

Of interest is the fact that the same scribe seems to have also penned extant fo-
lios from two separate lectionaries (Lectionaries ŖAŗ and ŖBŗ), as well as Psal-
ter ŖK,ŗ described below.50
Psalter “D,” represented by 12 fragments, is written in a very distinctive hand
(especially ‫ )ܠ‬and appears to come from the same scribe as Ḥudra B:51
SyrHT 129 = Psa. 72:8Ŕ73:4
SyrHT 377 = Psa. 74:21Ŕ23; 75:8Ŕ10
SyrHT 180 = Psa. 87:1Ŕ88:12
SyrHT 119 = Psa. 89:12Ŕ20, 23Ŕ33
SyrHT 224 (I) = Psa. 89:35Ŕ42, 48Ŕ53
SyrHT 64 = Psa. 89:46Ŕ47; 90:2
SyrHT 157 = Psa. 92:8Ŕ13; 93:1Ŕ5
SyrHT 239 = Psa. 94:12Ŕ15, 23Ŕ95:1
SyrHT 224 (II) = Psa. 95:4Ŕ7; 96:1Ŕ2
SyrHT 63 = Psa. 95:9Ŕ10; 96:5Ŕ8
SyrHT 308 = Psa. 103:2Ŕ4, 14Ŕ16
SyrHT 191 = Psa. 118:133Ŕ142, 146Ŕ153 (= Hebrew Psa. 119:133Ŕ
142, 146Ŕ153)
As can be seen, several of these fragments come from the same or adjacent fo-
lios. The Psalter includes the headings and canons, but not the prayers. It has
quire marks and the Psalm numbers (using Syriac letters) and each new mar-
mithā are designated in marginalia.
Psalter “E” is the second Psalter booklet in the collection. It consists of four
adjacent folios stitched together (SyrHT 71) and contains Psa. 22:8-26:3, in-
cluding the headings, canons and prayers. Again, marginalia designate Psalm
numbers using Syriac letters and a distinctive mark is found in the upper right
corner on the verso side of each folio which looks roughly like ‫( ♱ܚܐ܀‬this mark
is found on many of the Christian fragments from Turfan, whether Syriac or
Sogdian). Of interest also is that the later hand-list, under the entry for SyrHT
71, also mentions 2 fragments (now lost)52 with the old signature number T II B
66 No. 1 which apparently belonged to the same Psalter.
Psalter “F” is in another very distinctive hand with each line consisting of one
colon of the biblical text (each verse of the Psalms, in both the Hebrew and

50
As noted above, Lectionary ŖAŗ = SyrHT 48 & 49; Lectionary ŖBŗ = SyrHT 241, SyrHT 277,
SyrHT 300 and n 327.
51
Represented by the following fragments: SyrHT 13, SyrHT 159, SyrHT 248, SyrHT 310,
SyrHT 377.
52
The Turfan Collection experienced significant loss at the end of World War II due to its loca-
tion in Berlin.
368 Mark Dickens

Syriac text, consists of two or three cola) and ending in the punctuation mark :
in red ink. As a result, the last word on each line is elongated as necessary and
there are several places where words are omitted at the end of a line (especially
evident on SyrHT 92), in order to ensure that each new line begins with the
next colon (see also Psalters ŖGŗ and ŖJŗ below). Psalter ŖFŗ has the canons and
prayers, but no headings and is extant in 15 fragments:
SyrHT 14, 15 & 17 = Psa. 66:13Ŕ15; 67:5Ŕ653
SyrHT 174 = Psa. 73:25Ŕ27; 74:4Ŕ754
SyrHT 90 = Psa. 78:26Ŕ4555
SyrHT 91 = 78:46Ŕ64
SyrHT 93 = 79:9Ŕ80:12
SyrHT 92 = Psa. 84:3Ŕ85:5
SyrHT 172, SyrHT 175 & n 418 = Psa. 85:6Ŕ12; 86:1Ŕ856
SyrHT 173, SyrHT 176 & SyrHT 177 = Psa. 90:2Ŕ7, 9Ŕ1657
SyrHT 367 = currently unidentifiable58
Psalter “G” is represented by only one long narrow folio with another distinc-
tive hand (SyrHT 96), containing portions of the Odes found at the end of the
Psalter, namely Ex. 15:15Ŕ21; Isa. 42:10Ŕ13; 45:8 (part of the First Song of
Moses, followed by most of the Song of Isaiah). As noted above, it follows the
order of 12th century Psalter manuscripts, not 8th century ones. It does not have
the headings, canons or prayers. Similar to Psalters ŖFŗ and ŖJ,ŗ in order to keep
just one colon of the biblical text on each line, words which do not fit at the end
of lines are omitted or represented by only the first several letters of the word.59
Since these texts were so well-memorized, those reciting them presumably did
not need the missing words.

53
These are three small fragments, glassed separately, each containing only a few identifiable
words, but plainly from the same original folio.
54
This fragment is glassed with SyrHT 172, SyrHT 173 and SyrHT 175, also from the same
Psalter.
55
SyrHT 90 & 91 and SyrHT 92 & 93 are preserved as two double folios.
56
Fragment n 418 was glassed with five other unrelated fragments, all of which are Syriac, but at
the time all were thought to be Sogdian in Syriac script and so received signature numbers be-
ginning with n (for nestorianische).
57
SyrHT 176 and SyrHT 177 are glassed separately from SyrHT 172Ŕ175.
58
Although the fragment is clearly from Psalter ŖF,ŗ the very few letters left on it make its identi-
fication impossible without an unwarranted expenditure of energy to track it down in the Psal-
ter.
59
However, whereas Psalter ŖFŗ only omits words for this reason a few times, Psalter ŖGŗ does it
on nearly every line, presumably necessitated due to the narrow width of the folio and there-
fore the writing area. Since we only have one folio from this Psalter, it is impossible to know
how general this was throughout the whole text.
The Importance of the Psalter at Turfan 369

The next two Psalters also have only one extant folio each. Psalter “H”
(SyrHT 62) contains Deut. 32:31Ŕ40 (the Third Song of Moses) and Psalter
“I” (SyrHT 108) contains Psa. 10:12Ŕ12:3. Psalter “J” has two extant frag-
ments: SyrHT 113, a double-folio containing Psa. 14:7Ŕ16:5; 18:35Ŕ50, and
SyrHT 164, containing Psa. 18:51Ŕ19:2; 19:4Ŕ7. In all three Psalters, there is
only one colon per line and all lines end with the punctuation mark ‫܀‬. In Psal-
ters “H” and “I” (especially the latter) many lines also end with the word
‫( ܗܠܡܘܝܐ‬Hallelujah) in order to fill out left-over space. Again (as with Psalters
ŖFŗ and ŖGŗ above), words in Psalter “J” are omitted or abbreviated at the end
of lines if there is not enough space for the entire colon. Both Psalter “I” and
Psalter “J” have the canons and prayers, but not the headings (there is not
enough of Psalter “H” to determine what additional items it had).
Psalter “K” is represented by five fragments:
SyrHT 120 = Psa. 35:7Ŕ17, 22Ŕ36:3
SyrHT 121 = Psa. 36:7Ŕ37:2, 7Ŕ16
SyrHT 220 = Psa. 37:20Ŕ23, 34Ŕ38
SyrHT 357 = Psa. 38:9Ŕ12; 39:2Ŕ3
SyrHT 358 = Psa. 141:5Ŕ8 (= Hebrew Psa. 142:5Ŕ8); 143:15Ŕ144:4 (=
Hebrew Psa. 144:15Ŕ145:4)
This Psalter has headings, but not canons or prayers.
Psalter “L” consists of two fragments that can be joined to form a partial folio
(SyrHT 98 and SyrHT 203), containing Psa. 118:32Ŕ49, 63Ŕ80 (= Hebrew Psa.
119:32-49, 63-80). Without the beginning of the Psalm, it is impossible to tell if
this Psalter had headings, canons or prayers. Finally, a small fragment in an at-
tractive hand (MIK III 110), containing Psa. 24:3Ŕ4; 25:3Ŕ6 (currently housed
in the Museum für Asiatische Kunst), is all that remains of Psalter “M.” Again,
there is too little information to determine if this Psalter originally contained
headings, canons or prayers.60
A final Psalter manuscript needs to be mentioned here, namely nine folios of a
small booklet in Syriac transliterated into Uyghur script, currently referred to as
the Uyghur Psalter (SyrHT 20-27 and MIK III 58).61 This text is one of only
two in the Turfan Collection identified as Syriac transliterations into Uyghur
script62 and was evidently prepared in order to enable Uyghur-speaking monks
who could not read Syriac script to recite the Syriac liturgy. Six of the extant

60
Thanks to Aho Shemunkasho for identifying this Psalter fragment during a visit to the Museum
für Asiatische Kunst, Dahlem, Berlin on 29 April, 2010.
61
This text will be published by the present author and Peter Zieme in the near future.
62
The other is So 20131, not yet deciphered. There are also two instances of Syriac liturgical
fragments which have several sentences of graffiti in Uyghur script on the blank reverse side
where the underlying text is Syriac: SyrHT 124 and SyrHT 286 & 287, both discussed in
DICKENS, 2009, 32Ŕ33.
370 Mark Dickens

folios contain Psalms arranged according to marmayāthā, but not in overall

numerical order (note SyrHT 20 below). Although the beginning of a new
marmithā is indicated in one place, there are no headings, canons or prayers.
The remaining three folios contain various teshbḥāthā used in the East Syriac
liturgy (not the biblical odes mentioned above, but rather canticles or hymns by
Syriac authors, such as Ephrem the Syrian). The Psalter portions are as follows:
SyrHT 23 = Psa. 11:6Ŕ13:3
SyrHT 26 = Psa. 13:3Ŕ14:7
SyrHT 20 = Psa. 14:7; 99:1Ŕ100:1
SyrHT 22 = Psa. 101:1Ŕ8
MIK III 58 = Psa. 86:11Ŕ87:7
SyrHT 21 = Psa. 87:7Ŕ88:15
The fact that the Syriac script was not used for this text, presumably because it
was no longer familiar to some (perhaps most?) of the Turfan monastic commu-
nity, suggests a late date in the Mongol era. Numerous Christian Uyghur texts
from Turfan written in Syriac script, some clearly datable to the Mongol era,
clearly indicate that the script was known amongst Uyghur Christians at Turfan
up to a certain point; how and when the script ceased to be well-known is un-
clear.
An in-depth study of where and how the Turfan Psalter fragments differ from
the standard text of the East Syriac Psalter, including the headings, canons and
prayers, is currently underway, but some initial observations about scribal errors
can be offered here. Following are some significant errors which are not merely
textual variants.63
‫ܥܤܮܐ‬̈ („amātha, Ŗauntsŗ) for ‫ܥܤܤ ܐ‬ ̈ („amē, Ŗnationsŗ) [Psalter “D,”
SyrHT 129]
‫( ܦܧܡܓܘܦܝܗܝ‬nephlghunay, Ŗthey will divide himŗ) for ‫ܦܧܡܛܘܦܝܗܝ‬
(nephl unay, Ŗthey will serve himŗ) [Psalter “D,” SyrHT 129]
‫( ܚܪܗ‬ḥaseh, Ŗhis lettuceŗ) for ‫( ܚܪܕܘ‬ḥasdu, Ŗthey have mockedŗ) [Psal-
ter “D,” SyrHT 224]
‫( ܕܦܛܥ‬d-na em, Ŗthat resuscitatesŗ) for ‫( ܕܦܗܡ‬d-nahem, Ŗthat roars,ŗ of
lions) [Psalter “E,” SyrHT 71]

63
Not included are the following variants (many which are relatively insignificant in nature) that
occur in the fragments, including 1) omission or addition of the inseparable prefixes ‫ܒ‬, ‫ ܕ‬and ‫;ܘ‬
2) omission or addition of the plural seyame marking; 3) words (especially pronouns and encli-
tics) joined together; 4) omission or addition of the third person singular masculine pronominal
suffix ‫ ;ܗ‬5) omission or addition of the first person singular objective pronominal suffix ‫ ;ܝـ‬6)
addition of ‫ ܝـ‬to turn Ŗsingularŗ pronominal suffixes (e.g., ‫ )ܗܘܢ‬into Ŗpluralŗ ones (e.g.,
‫ ;)ܝܗܘܢ‬7) addition of the third person plural masculine perfect ending ‫ܘ‬.
The Importance of the Psalter at Turfan 371

‫( ܐܬܚܢ‬eth , not Syriac) for ‫( ܐܕܚܢ‬edhhel, ŖI will fearŗ) [Psalter “E,”

SyrHT 71]
‫( ܒܒܪܬܐ‬b-bharthā, Ŗwith a daughter/eggŗ) for ‫( ܒܒܪܕܐ‬b-bhardhā, Ŗwith
hailŗ) [Psalter “F,” SyrHT 91]
‫( ܠܮܗܘܣ ܐ‬lathhumā, Ŗto the abyssŗ) for ‫( ܠܮܚܘܣ ܐ‬lath umā, Ŗto the
borderŗ) [Psalter “F,” SyrHT 91]
‫( ܚܟܤܮܐ‬hekhmthā, Ŗwisdomŗ) for ‫( ܚܤܮܐ‬hemthā, Ŗanger, wrathŗ)
[Psalter “F,” SyrHT 92]
‫( ܚܤܮܪܐ‬hamthrā, not Syriac) for ‫( ܚܤܪܐ‬hamrā, Ŗwineŗ) [Psalter “H,”
SyrHT 62]
‫( ܟܪܒܗ‬kesbeh, not Syriac) for ‫( ܟܪܧܗ‬kespeh, Ŗhis moneyŗ) [Psalter
“J,” SyrHT 113]
‫( ܒܟܪܘܟܘܘܬܗܘܢ‬bakhrykuthhon, not Syriac) for ‫( ܒܟܪܝܗܘܬܗܘܢ‬bakhry-
huthhon, Ŗin their sicknessŗ) [Psalter “K,” SyrHT 120]
‫( ܦܛܪܘܢ‬nehrun, not Syriac) for ‫( ܦܛܧܪܘܢ‬nehprun, Ŗmay they be
ashamedŗ) [Psalter “K,” SyrHT 120]
‫( ܣܡܧܘܬܐ‬malnuthā, not Syriac) for ‫( ܣܡܧܧܘܬܐ‬malphānuthā, Ŗteachingŗ)
[Psalter “K,” SyrHT 121]
Most of these scribal errors seem to indicate that the scribe was not well-
acquainted with Syriac. A number of them suggest that the scribes may have
been native Uyghur-speakers, especially those involving bilabial confusion (‫ܦ‬
/p/ and ‫ ܒ‬/b/) or dental confusion (‫ܬ‬/t/ or /θ/ and ‫ ܕ‬/d/ or /ð/). In other cases,
confusion between various velar, pharyngeal and glottal sounds (notably ‫ ܟ‬/k/
or /x/; ‫ ܓ‬/g/ or /γ/; ‫ ܚ‬/ / and ‫ ܗ‬/h/) could be due to phonological interference
from either Uyghur (these four Syriac letters can all be represented by one Uy-
ghur letter x when Syriac is transliterated into Uyghur script, as in the Uyghur
Psalter) or (in the case of ‫ܓ‬, ‫ ܚ‬and ‫ )ܗ‬Sogdian.
The error of ‫ ܚܪܗ‬in place of ‫ ܚܪܕܘ‬can be explained by the similarity in shape
between ‫ ܗ‬and ‫ܕܘ‬, but again suggests lack of knowledge of the Syriac lan-
guage.64 Other errors, including the omission of the letter ‫ ܦ‬/p/ or /f/ and the
insertion of the letters ‫ ܬ‬/t/ or /θ/ and ‫ ܟ‬/k/ or /x/, may or may not indicate pho-
nological interference. Indeed, the substitution of the legitimate Syriac words
‫ܥܤܮܐ‬̈ („amātha, Ŗauntsŗ) for ‫ܥܤܤ ܐ‬ ̈ („amē, Ŗnationsŗ) and ‫ ( ܚܟܤܮܐ‬ekhmthā,
Ŗwisdomŗ) for ‫ ( ܚܤܮܐ‬emthā, Ŗanger, wrathŗ) could theoretically be made by
native Syriac-speakers who were not concentrating enough while copying the
original manuscript, resulting in the confusion of two common words. More
study of these spelling errors, combined with palaeographic analysis of the dif-
ferent hands involved, could also shed light on the dating of these Psalters, since
64
Someone well-versed in Syriac would be unlikely to write about Ŗlettuceŗ in the context of Psa.
89 without realizing it was an error!
372 Mark Dickens

Uyghur influence probably indicates a later date, whereas Sogdian influence

likely points to an earlier date.65
References to the Psalter can also be found in other Syriac fragments from Tur-
fan, particularly the many liturgical fragments currently being identified.66 One
example which has already surfaced is SyrHT 87, containing a ‫( ܛܟܪܐ‬ṭakhsā,
Ŗriteŗ) for ‫( ܠܡܝܐ‬lelyā, Ŗnightŗ) that includes instructions to the priest in Sogdian
and Psalm incipits in Syriac.67

Other Uses of the Psalms in Central Asia

Although they are the largest corpus of Psalmic material from Central Asia, the
Psalter fragments from Turfan are not the only texts showing how the Psalms
were used by Christian communities in the region. Several other archaeological
finds are also of relevance to this topic. Perhaps the most significant of these is
an ostracon (potsherd) from Panjikent, Tajikistan, dating from the late 7th or
early 8th century, containing portions of Psalms 1 and 2 in Syriac written in
black ink. Although the text is the same as that of the Peshiṭta, scribal errors
indicate that the writer was a Sogdian-speaker and suggest that Ŗthe inscription
is a studentřs exercise written under dictation for the purpose of mastering the
Syrian [sic] script,ŗ possibly at Ŗa school attached to a monastery and built on
the model of the Syrian [sic] schools described in numerous sources.ŗ68
There are also several liturgical or prayer texts with quotations from the Psalms
discovered in various parts of Central Asia. A Syriac prayer for rain after
drought found in Qara-Khoto includes a quotation from Psa. 51:4, Davidřs
Psalm of repentance after committing adultery with Bathsheba. 69 Another
Syriac liturgical fragment from Dunhuang consists of excerpts from ‫ܟܮܒܐ ܕܩܕܡ‬
‫( ܘܕܒܮܪ‬kthābhā dha-qdhām wadh-bāthar), ŖBook of the Before and Afterŗ
which contain quotations from the Psalms (Psa. 15, 17, 21, 23, 24, 25, 28),
along with an interlinear text in Uyghur script of Buddhist origin from the
Mongol era.70
65
Thanks to Peter Zieme and Nicholas Sims-Williams for their observations about possible Uy-
ghur or Sogdian influence in these errors. On orthographic errors in the Syriac text of Psa. 1
and 2 written on an ostracon recovered from Panjikent which indicate that the scribe was a na-
tive Sogdian-speaker, see PAYKOVA, 1979, 165.
66
This is a laborious process, given the fact that, unlike the biblical texts, there is no concor-
dance to the printed editions of the Ḥudra or other liturgical texts of the Church of the East.
67
Discussed in DICKENS, 2009, 27. The following Psalm incipits have been identified by Sebas-
tian Brock: Psa. 102 (twice); 115 (Heb. 116); 129 (Heb. 130); 13; 26; 88; 3; 142 (Heb. 143);
128 (Heb. 129); 103 or 104; and 145 (Heb. 146).
68
PAYKOVA, 1979, 166.
69
PIGOULEWSKY [PIGULEVSKAYA], 1935/36, 14Ŕ18.
70
The Syriac text is published in DUAN, 2001, the Uyghur text in YAKUP, 2002. Although it was
The Importance of the Psalter at Turfan 373

An Uyghur Christian text from Turfan also contains at least an allusion to, if not
a quotation from, the Psalter. U 4910, written in Uyghur script, is a prayer con-
taining the phrase s(ä)n ärürs(ä)n [ögsüz]lärniŋ ögi kaŋsızlarnıŋ k[aŋı], ŖYou
are Mother to the motherless, Father to the fatherless.ŗ71 This seems to be an
expansion on Psa. 68:5 Ŕ ŖFather of the fatherless (or orphans) and judge (de-
fender) of the widows is God in his holy dwelling.ŗ72 Given the reputation of
the medieval Mongol-Turkic world for according higher status to women than
elsewhere in Ŗthe civilized worldŗ at that time (whether Muslim, Christian or
Chinese),73 it is interesting that a biblical passage referring to God only as male
has here been extended to include the feminine aspect of the divine.74
Another Psalmic quotation is preserved in a funerary tile from Chifeng, Inner
Mongolia, dated to 1253 CE. In the centre of the tile is a cross, with a vertical
inscription in Uyghur script in the lower left and right quadrants of the tile.75 In
the upper left and right quadrants are four Syriac words: ‫ܚܘܪ ܠܘܬܗ ܩܒܪܘ ܒܗ‬,
ŖLook to him, hope (or trust) in him,ŗ a quote from Psa. 34:6. Unlike other
Christian gravestones from Central Asia, the tile is not engraved, but the vertical
inscriptions are painted in black ink. As Pier Giorgio Borbone has noted, quota-
tions from Psa. 34:6 sometimes accompany the image of the cross in Syriac
manuscripts, and there is at least one other example of this text on a gravestone
with a cross from China, found at the Buddhist Temple of the Cross in the
Fangshan Region near Beijing.76

Preliminary Conclusions and Outstanding Questions about the Turfan

Psalters

The prevalence and variety of Psalters from Turfan, whether written in Syriac,
Sogdian or New Persian or transliterated into Uyghur script, testifies to their
importance in the life of the Christian community there. The Syriac Psalters in

originally thought that the Uyghur text might also be of Christian origin, Yakup has shown
clearly that it is a Buddhist text and therefore not related in any way to the Syriac fragment on
which it was written.
71
ZIEME, 1974, 663. The updated transcription by Peter Zieme reproduced here is slightly differ-
ent from that in the article.
72
Translation of Psa. 68:6 in the Peshiṭta (Psa. 68:5 in English Bibles).
73
See for instance, Ibn Battutařs comment on the status of women in the Qipchaq Khanate under
Özbeg Khan (1313Ŕ1341): ŖA remarkable thing which I saw in this country was the respect
shown to women by the Turks, for they hold a more dignified position than the menŗ (GIBB,
1929, 146). On this theme in general, see also RYAN, 1998.
74
This aspect is however, suggested in other verses in the Bible, e.g. Psa. 131:2; Isa. 42:14; 46:3;
49:15; 66:13; I Thess. 2:7.
75
On the Uyghur inscription, see HAMILTON & NIU, 1994.
76
On the use of Psa. 34:6 in China and Central Asia, see BORBONE, 2006.
374 Mark Dickens

particular provide insight into the level of Syriac proficiency in the monastic
community. Not surprisingly, the scribal hands differ considerably; some were
obviously accomplished calligraphers who wrote their Psalters very carefully,
but others were less attentive in their copying and allowed various errors to
creep in. The omission of words in some Psalters due to space limitations re-
minds us that these texts were memorized by heart, so that omitted words did
not hinder recitation of the Psalter.
More work still needs to be done with the Syriac Psalter fragments from Turfan.
A careful study of scribal errors in the fragments may give us further insight
into how the Sogdian and Turkic speakers in the Turfan Christian community
pronounced Syriac, as well as how knowledge of both the Syriac script and lan-
guage changed over time in the community. Although comparison with the
critical edition of the Peshiṭta Psalter has not revealed any significant variants
from the established text of the Psalter itself, the headings, canons and prayers
still need to be studied more in-depth to determine if there are any important
differences from the standard text of these components of the Psalter, especially
as they are preserved in the printed edition of the Ḥudra.77
As noted above, the Syriac Psalters from Turfan are amongst the earliest extant
examples of this text. Although none have been dated definitively, they pre-
sumably come from sometime between the 9th and 13th centuries, the standard
date range for Turfan manuscript fragments. Given the fairly developed state of
Syriac palaeographical study, this task should not be excessively difficult. Per-
haps more accurate dating of the Syriac fragments may also help in dating the
Christian fragments in Sogdian or Uyghur, especially if common scribal hands
can be identified. If any of the Syriac Psalters can be dated before the 12 th cen-
tury, they could be helpful in clarifying textual variants at an early stage, since
there are only two early manuscripts used in the critical edition of the Peshiṭta
Psalter which reflect the East Syriac tradition: 6t1 and 8t1.78 Most of the exclu-
sively East Syriac Psalter manuscripts used in the critical edition (including the
earliest ones Ŕ 12t1, 12t3 and 12t4) may well be later than many of the Syriac
Psalters from Turfan.79
Certainly, there is more to be learned from study of the Turfan Psalters which,
along with other extant liturgical fragments, testify to a highly-organized mo-
77
Comparison with the headings preserved in BLOEMENDAAL, 1960 will also be important.
78
Respectively, a 6th century manuscript in the British Library containing the Psalms and Odes,
written in the West Syriac script, but with East Syriac headings, and an 8th century manuscript
from St. Catherineřs Monastery, Mt. Sinai, also containing the Psalms and Odes, again a West
Syriac text but with some East Syriac marginal readings.
79
All 12th century manuscripts from the British Library or the Chaldean Patriarchate, Baghdad,
containing the Psalms and Odes. Other East Syriac Psalter manuscripts are significantly later:
4 from the 13th century, 3 from the 16th century, and 3 from the 17th century. Again, see the list
in WALTER, 1980, viiŔxxix.
The Importance of the Psalter at Turfan 375

nastic community which was faithful to the beliefs and practices of the Church
of the East, despite being located thousands of kilometres to the east of the
Churchřs Mesopotamian and Persian heartland.

Bibliography

ANDREAS, F. C., 1910. ŖBruchstücke einer Pehlewi-Übersetzung der Psalmen aus der
Sassanidenzeitŗ. Sitzungsberichte der Königlich Preussischen Akademie der Wis-
senschaften, 869Ŕ872.
ANDREAS, F. C. & Kaj BARR, 1933. ŖBruchstücke einer Pehlewi-Übersetzung der
Psalmenŗ. Sitzungsberichte der Preussischen Akademie der Wissenschaften, 91Ŕ
152.
ASMUSSEN, J. P., 1964. ŖThe Pahlavi Psalm 122 in Englishŗ. Dr. J. M. Unvala Memo-
rial Volume. Bombay: Kaikhusroo M. JamaspAsa, 123Ŕ126.
BADGER, George P., 1852. The Nestorians and Their Rituals, Vol. II. London: Joseph
Masters.
BANG, W., 1926. ŖTürkische Bruchstücke einer nestorianischen Georgpassionŗ. Le
Muséon, Vol. 39, 41Ŕ75.
BENVENISTE, E., 1938. ŖSur un fragment dřun psautier syro-persanŗ. Journal Asiatique,
Vol. 230, 458Ŕ462.
BERLIN-BRANDENBURG ACADEMY OF SCIENCES AND HUMANITIES, 2007. Turfan Studies.
Berlin: Berlin-Brandenburg Academy of Sciences and Humanities.
BLOEMENDAAL, W., 1960. The Headings of the Psalms in the East Syrian Church. Lei-
den: E. J. Brill.
BORBONE, Pier Giorgio, 2006. ŖPeshitta Psalm 34:6 from Syria to Chinaŗ. Text, Trans-
lation, and Tradition: Studies on the Peshitta and its Use in the Syriac Tradition
(Monographs of the Peshitta Institute Leiden, Vol. 14), W. Th. van Peursen & R. B.
ter Haar Romeny (ed.). Leiden: Brill, 1Ŕ10.
BROCK, Sebastian P., 2006. The Bible in the Syriac Tradition (Gorgias Handbooks, Vol.
7). Piscataway, NJ: Gorgias Press.
BUDGE, Ernest A. Wallis (ed.), 1893a. The Book of Governors: The Historia Monastica
of Thomas Bishop of Margâ A. D. 840, Vol. I: The Syriac Text, Introduction, etc.
London: Kegan Paul, Trench, Trübner & Co.
BUDGE, Ernest A. Wallis, tr., 1893b. The Book of Governors: The Historia Monastica of
Thomas Bishop of Margâ A. D. 840, Vol. II: The English Translation. London: Ke-
gan Paul, Trench, Trübner & Co.
BURKITT, F. Crawford, 1921Ŕ1923. ŖThe Early Syriac Lectionary Systemŗ. Proceedings
of the British Academy, Vol. 10, 301Ŕ338.
BURKITT, F. Crawford, 1925. The Religion of the Manichees. Cambridge: Cambridge
University Press.
DřAIUTO, Francesco, 2008 [2009]. ŖUn antico inno per la Resurrezioneŗ. Rivista di Stu-
di Bizantini e Neoellenici, Vol. 45 (N.S.), 3Ŕ135.
376 Mark Dickens

DARMO, Thoma, 1960. Ktaba da-Qdam wad-Batar wad-Hudra wad-Kashkol wad-

Gazza w-Qala d-`Udrane `am Ktaba d-Mazmure, Vol. I. Trichur, Kerala: Church of
the East.
DARMO, Thoma, 1961. Ktaba da-Qdam wad-Batar wad-Hudra wad-Kashkol wad-
Gazza w-Qala d-`Udrane `am Ktaba d-Mazmure, Vol. II. Trichur, Kerala: Church
of the East.
DARMO, Thoma, 1962. Ktaba da-Qdam wad-Batar wad-Hudra wad-Kashkol wad-
Gazza w-Qala d-`Udrane `am Ktaba d-Mazmure, Vol. III. Trichur, Kerala: Church
of the East.
DAUVILLIER, Jean, 1953. ŖByzantins dřasie centrale et dřextrême-orient au moyen ageŗ.
Revue des Études Byzantines, Vol. 11, 62Ŕ87.
DE BLOIS, Francois, 2006. ŖDu nouveau sur la chronologie bactrienne post-
hellénistique : lřère de 223-224 ap. J.-C.ŗ. Comptes-rendus des séances de
l‟Académie des Inscriptions et Belles-Lettres, 991Ŕ997.
DICKENS, Mark, 2009. ŖMultilingual Christian Manuscripts from Turfanŗ. Journal of
the Canadian Society for Syriac Studies, Vol. 9, 22Ŕ42.
DIETTRICH, G., 1909. ŖBericht über neuentdeckte handschriftliche Urkunden zur Ge-
schichte des Gottesdienstes in der nestorianischen Kircheŗ. Nachrichten von der
Königlichen Gesellschaft der Wissenschaften zu Göttingen (Philologisch-
historische Klasse), 160Ŕ218.
GIBB, H. A. R., tr., 1929. Ibn Battuta: Travels in Asia and Africa, 1325–1354. London:
George Routledge & Sons.
GIGNOUX, Philippe, 1968. ŖLřauteur de la version pehlevie du psautier serait-il nesto-
rien?ŗ Mémorial Mgr Gabriel Khouri-Sarkis. 1898-1968. Louvain: Imprimerie
Orientaliste, 233Ŕ244.
GIGNOUX, Philippe, 2001. ŖUne croix de procession de Hérat inscrite en pehleviŗ. Le
Muséon, Vol. 114, 291Ŕ304.
GIGNOUX, Philippe, 2008. ŖPahlavi Psalterŗ [Electronic Version]. Encyclopaedia Irani-
ca (http://www.iranicaonline.org/articles/pahlavi-psalter).
GILLMAN, Ian & Hans-Joachim KLIMKEIT, 1999. Christians in Asia before 1500. Ann
Arbor: University of Michigan Press.
HAMILTON, James & NIU Ruji, 1994. ŖDeux inscriptions funéraires turques nestoriennes
de la Chine orientaleŗ. Journal Asiatique, Vol. 282, 147Ŕ164.
HANSEN, Olaf, 1941. ŖBerliner soghdische Texte I: Bruchstücke einer soghdischen Ver-
sion der Georgspassion (C1)ŗ. Abhandlungen der Preussischen Akademie der Wis-
senschaften. Philosophisch-Historische Klasse, No. 10, 1Ŕ38.
KLEIN, Wassilios, 1999. ŖDas Orthodoxe Katholikat von Romagyris in Zentralasienŗ.
Parole de l‟Orient, Vol. 24, 235Ŕ265.
MACLEAN, Arthur John, 1894. East Syrian Daily Offices. London: Rivington, Percival
[repr: Piscataway, NJ: Gorgias Press, 2003].
MAI, Angelo (ed.), 1838. Scriptorum veterum nova collectio e vaticanis codicibus edita
ab A.M., Vol. X. Rome: Typis Collegi Urbani.
The Importance of the Psalter at Turfan 377

MARÓTH, Miklós, 1985. ŖEin Brief aus Turfanŗ. Altorientalische Forschungen, Vol. 12,
283Ŕ287.
MARÓTH, Miklñs, 1991. ŖEine unbekannte Version der Georgios-Legende aus Turfanŗ.
Altorientalische Forschungen, Vol. 18, 86Ŕ108.
MATEOS, Juan, 1959. Lelya – Sapra: Essai d‟interpretation des matines chaldeenes
(Orientalia Christiana Analecta 156). Rome: Pontificio Institutum Orientalium Stu-
diorum.
MÜLLER, F. W. K., 1907. ŖNeutestamentliche Bruchstücke in soghdischer Spracheŗ.
Sitzungsberichte der Königlich Preussischen Akademie der Wissenschaften, 260Ŕ
270.
MÜLLER, F. W. K., 1913. ŖSoghdische Texte Iŗ. Abhandlungen der Königlich Preussi-
schen Akademie der Wissenschaften. Jahrgang 1912. Philosophisch-Historische
Klasse, No. II, 1Ŕ111.
MÜLLER, F. W. K., 1915. ŖEin syrisch-neupersisches Psalmenbruchstück aus Chine-
sisch-Turkistanŗ. Festschrift Eduard Sachau, Gotthold Weil (ed.). Berlin: Verlag
von Georg Reimer, 215Ŕ222.
NASRALLAH, J., 1976. ŖLřÉglise melchite en Iraq, en Perse et dans lřAsie centraleŗ.
Proche Orient Chrétien, Vol. 26, 16-33, 319Ŕ353.
PAYKOVA, A. V., 1979. ŖThe Syrian Ostracon from Panjikantŗ. Le Muséon, Vol. 92,
159Ŕ169.
PIGOULEWSKY [PIGULEVSKAYA], N., 1935Ŕ1936. ŖFragments syriaques et syro-turcs de
Hara-hoto et de Tourfanŗ. Revue de l‟Orient Chrétien, Vol. 30, 3Ŕ46.
DUAN, Qing, 2001. "Bericht über ein neuentdecktes syrisches Dokument aus Dunhuang
/ China," in Oriens Christianus, Vol. 85, 84Ŕ93.
RECK, Christiane, 2008. ŖA Survey of the Christian Sogdian Fragments in Sogdian
Script in the Berlin Turfan Collectionŗ. Controverses des chrétiens dans l‟Iran sas-
sanide (Studia Iranica - Cahier 36), Christelle Jullien (ed.). Paris: Association pour
lřAvancement des Études Iraniennes, 191Ŕ205.
RYAN, James D., 1998. ŖChristian Wives of Mongol Khans: Tartar Queens and Mis-
sionary Expectations in Asiaŗ. Journal of the Royal Asiatic Society, Vol. 8 (Ser. 3),
411Ŕ421.
SCHNEIDER, H. (ed.), 1972. The Old Testament in Syriac according to the Peshitta ver-
sion, Part IV, fascicule 6: Canticles or Odes. Leiden: E. J. Brill.
SCHWARTZ, Martin, 1967. Studies in the Texts of the Sogdian Christians. Dissertation:
Near Eastern Languages, University of California, Berkeley.
SCHWARTZ, Martin, 1974. ŖSogdian Fragments of the Book of Psalmsŗ. Altorientalische
Forschungen, Vol. 1, 257Ŕ261.
SCHWARTZ, Martin, 1982. Studies in the Texts of the Sogdian Christians (revised ver-
sion). Dissertation: Near Eastern Languages, University of California, Berkeley,
Berkeley.
SIMS-WILLIAMS, Nicholas, 1990. ŖBulayïqŗ. Encyclopaedia Iranica, Vol. 4, 545.
378 Mark Dickens

SIMS-WILLIAMS, Nicholas, 2004. ŖA Greek-Sogdian Bilingual from Bulayïqŗ. La Per-

sia e Bisanzio (Atti dei Convegni Lincei 201). Rome: Accademia Nazionale dei
Lincei, 623Ŕ631.
SKJAERVØ, O., 1983. ŖCase in Inscriptional Middle Persian, Inscriptional Parthian and
the Pahlavi Psalterŗ. Studia Iranica, Vol. 12, 47Ŕ62, 151Ŕ181.
SPRENGER, Norbert, 1976. Konkordanz zum Syrischen Psalter (Göttinger Orientfor-
schungen, I Reihe: Syriaca, Band 10, Teil 8). Wiesbaden: Otto Harrassowitz.
SUNDERMANN, Werner, 1974. ŖEinige Bemerkungen zum Syrisch-Neupersischen Psal-
menbruchstük aus Chinesisch-Turkistanŗ. Mémorial Jean de Menasce, Philippe
Gignoux & A. Tafazzoli (ed.). Louvain: Imprimerie Orientaliste, 441Ŕ452.
SUNDERMANN, Werner, 1975. ŖNachlese zu F. W. K. Müllers „Soghdischen Texten IŖ,
2. Teilŗ. Altorientalische Forschungen, Vol. 3, 55Ŕ90.
SUNDERMANN, Werner, 1981. ŖNachlese zu F. W. K. Müllers „Soghdischen Texten IŖ,
3. Teilŗ. Altorientalische Forschungen, Vol. 8, 169Ŕ225.
SUNDERMANN, Werner, 1994. ŖByzanz und Bulayïqŗ. Iranian and Indo-European Stu-
dies: Memorial Volume of Otakar Klíma, Petr Vavroušek (ed.). Praha: Enigma
Corporation, 255Ŕ264.
TAFT, Robert, 1986. The Liturgy of the Hours in East and West: the Origins of the Di-
vine Office and its Meaning for Today. Collegeville: Liturgical Press.
WALTER, D. M. (ed.), 1980. The Old Testament in Syriac according to the Peshitta ver-
sion, Part II, fascicule 3: The Book of Psalms. Leiden: E. J. Brill.
WELLESZ, Egon, 1919. ŖMiscellanea zur orientalistischen Musikgeschichte. Die Lekti-
onszeichen in den soghdischen Textenŗ. Zeitschrift für Musikwissenschaft, Vol. I,
505Ŕ515.
WRIGHT, William, 1894. A Short History of Syriac Literature. London: A. and C. Black
[repr: Piscataway, NJ: Gorgias Press, 2001].
YAKUP, Abdurishid, 2002. ŖOn the Interlinear Uyghur Poetry in the Newly Unearthed
Nestorian Textŗ. Splitter aus der Gegend von Turfan: Festschrift für Peter Zieme
anläßlich seines 60. Geburtstags, Mehmet Ölmez & Simone-Christiane Raschmann
(ed.). Istanbul-Berlin: Şafak Matbaacılık, 409Ŕ417.
ZIEME, Peter, 1974. ŖZu den nestorianish-türkischen Turfantextenŗ. Sprache, Geschich-
te und Kultur der altaischen Völker: Protokollband der XII Tagung der Permanent
International Altaistic Conference 1969 in Berlin (Schriften zur Geschichte und
Kultur des alten Orients 5), Georg Hazai & Peter Zieme (ed.). Berlin: Akademie-
Verlag, 661Ŕ668.
The Importance of the Psalter at Turfan 379

Pahlavi Psalter folio: Ps 06 Sogdian Psalter fragment: So 12950(1)

Syriac-New Persian Psalter fragment: Uyghur script Psalter folio: SyrHT 20

SyrHT 153
380 Mark Dickens

Syriac Psalter C folio: SyrHT 72 Syriac Psalter F folio: SyrHT 92

Syriac Psalter G folio: SyrHT 96 Syriac Psalter J folio: SyrHT 113