AN UNPUBLISHED COPTIC SAHIDIC PSALTER CODEX AT THE UNIVERSITY MUSEUM IN PHILADELPHIA:  A PRELIMINARY REPORT

 
by Robert A. Kraft (University of Pennsylvania) [updated from 15ja2010]
 
[originally = pp. 81-89 in ARMENIAN AND BIBLICAL STUDIES,
ed Michael E. Stone (Jerusalem: St. James Press, 1976)]

 
The Egyptology section of the University of Pennsylvania Museum in Philadelphia is in possession of portions of a Coptic/Sahidic paper codex of Psalms 1-49, including parts of the original cover materials.\1/  A summary description is provided here:
\1/ Appreciation is due to Dr. David O'Connor, curator of the Egyptology section, and to the staff of the University Museum for permission to work with this material and for assistance in various connections.  Students who have assisted particularly with this Coptic/Sahidic manuscript include Janet Timbie, Tom Woehrle, Patricia Crown, Lynne Cohen, Ann Bloomdahl, and Saundra Sterling.  See also n. 4 below. Further updated information on the collection is available online.

Inventory numbers: 

This material is listed as University Museum E16261, plus fragments inventoried separately under the numbers E16305 (items "i", "k", "l") and E16528.
 

Preserved portions:

There are five relatively large fragments of cover material (leather and linen, with paper filler; and some evidence of Arabic writing on the paper materials);\2/ and large sections of 108 folios, in three different types of writing.
\2/ It has not yet been established precisely how the cover materials connect with the preserved folios, but the pages and cover matter were all stored together and presumably were acquired together by the Museum.  Nevertheless, the possibility remains open that these covers never were part of the Psalm codex. 

Condition:

Water damage affects approximately half of each page; pages were not in their original sequence when damage occurred; between five and seven folios are missing, prior to last page preserved (=p. "227", containing Ps 49.1-3a); in most instances, water damaged portions have broken away from the better preserved portions of the page; there are frequent minor lacunae in damaged portions; special care is necessary to separate materials fused by the water damage. [[page 82]]

Original sizes of pages:

The original page size was approximately 18.5 by 13.5 cm, for hands nos. 1 and 3, and 17.5 by 12 cm for hand no. 2.
 

Description of writing:

Hand no. 1 is a sloppy, irregular Coptic uncial in black ink with minimal punctuation and about 18 lines to the page, with a single column per page.  Three folios containing this writing are preserved, all from the opening section of the codex (pages 1/2, 3/4, 7/8).

Hand no. 2 is a neat, regular, upright Coptic uncial in dark brown ink with some red decoration and careful attention to the strophic structure of the material (use of marginal letters, punctuation, identation); it has about 17 lines per page, in a single column. Only two folios containing this writing are preserved, from near the beginning of the codex (pages 14/15, 16/17).

Hand no. 3 is a neat, regular, slightly slanting Coptic uncial in moderately dark brown ink with red decoration and some attention to thought units (punctuation, use of marginal letters); it is normally 16 lines per page, in a single column with an average of 12-13 (10-16 range) letters per line.  Large parts of 103 folios in this hand have been identified (page 18 = Ps 7 onward).  Page numbers appear at the upper outside margin, signature numbers at the upper inside margin on the first and last pages of each signature.  Titles to new psalms are numbered and set off by broken lines from the rest of the material.  Large, decorated initial letters are used at the start of each psalm; there is haphazard use of enlarged letters set in the left margin, often (but not necessarily) when a new sentence or clause happens to begin on a new line; red ink is used to decorate marginal letters, to highlight punctuation and abbreviations, and to set off page/signature numbers and titles of new psalms.  

Background:

This Coptic/Sahidic codex is part of a large collection of hitherto uncatalogued papyri and related materials (paper, leather) in varying states of preservation and in various languages (Hieroglyphics, Hieratic, Demotic, Coptic, Greek, Arabic, Pahlevi, Hebrew, and one piece each in Syriac and Latin) acquired by the University Museum near the beginning of the 20th century.\3/  As nearly as can be determined at present, the bulk of this material was purchased on behalf of the Museum in 1910 by [[p. 83]] Professor W. Max Müller from or through a Professor Moritz in Cairo.  Some of the materials seem to have been tentatively identified already when Müller obtained them from Moritz, and Müller contributed notes and observations of his own, but it was not until 1932 that the present inventory numbers were assigned to the materials by Battiscombe Gunn.  Almost no systematic work has been done on the collection over the years; indeed, many of the pieces were not even in a condition to be examined until the present writer began humidifying and flattening them in 1966.  In the 1970s, a concerted effort began with a team of students under the writer's direction to complete the job of inventorying and describing the materials.\4/
\3/ Information about the acquisition and subsequent history of the collection has been gathered by Patricia Crown for publication in the near future.  The material summarized here is from Ms Crown's report. Subsequently John R. Abercrombie published an introductory article (1978) and prepared a fuller report (ca 1980).

\4/ There is now (2010) a provisional complete inventory. In 1976, rudimentary catalogues were underway for the leather materials in Arabic and Coptic (by Richard Beal and Robert Childs).  Subsequently, a catalogue of much of the Arabic appeared posthumously by Giorgio DellaVida (1981). An inventory of the Pahlevi materials is available (by Paul Morris -- see now J. de Menasce, who obtained photos of several Pahlavi pieces for his 1957 publication in Corpus Inscriptionum Iranicarum 3.4-5, subsequently edited by Dieter Weber (1992)), and a catalogue of the Hebrew MS holdings of the University Museum and the University Library (Cairo Geniza material) was undertaken (by Francesca Rochberg and Joseph Cahn).  We were hoping to complete an inventory of the paper materials in Arabic and Coptic by Spring 1974, but it has been delayed.  The bulk of the collection, consisting of Greek, Coptic, and Arabic papyri, has required a longer period of time to organize.  The papyri in Demotic, Hieratic, and Hieroglyphic are fewer in number and were next on the schedule to be inventoried in 1974, although access to them is restricted.  Most of the groundwork for the inventories has been done by Patricia Crown, Lynne Cohen, and Ann Bloomdahl.  

The Coptic Codex:

When we rediscovered the Coptic/Sahidic Psalter in Fall of 1972, it was housed rather carelessly in a cardboard box measuring 25 by 19.5 by 4 cm. Along with the inventory number on the face of the box were the words "Coptic Book of Psalms (etc. ?). Paper."  It is not clear who was responsible originally for that tentative identification, but since the 1932 inventory card bears the same information, the identification is at least that old.  The damaged codex had been handled frequently enough subsequent to its having been damaged by water that various parts of pages were broken off from the larger lump (the water had fused parts of many pages together), and the sequence of pages that had obtained when the codex became damaged by water was no longer preserved intact.  One of the first tasks would be the delicate job of physically separating the remaining adhering pages from each other.  There was real concern that a successful separation of the water-fused portions might not even be possible.
 

Separating the Pages

In the Fall of 1972, one of my graduate students, Janet Timbie, assisted in describing the state of the MS and recording the sequence of pages as they were found at that time.  Xerox copies were made [[p. 84]] of all surface areas that could safely be placed on a flat surfaced Xerox machine, and each available page was identified.  Much of the initial identification was done by another graduate student, Thomas Woehrle (Temple University).  After consultation with Susan Nash of the Museum conservation department and Wilhelm Spawn of the American Philosophical Library in Philadelphia, I proceeded cautiously with the task of separating the materials that still adhered to each other.  Some pages and sections responded well to gentle prying apart with the help of a thin, flat metal probe -- a method used by Wilhelm Spawn when he examined the material on 19 December, 1972.  Some portions were so tightly fused, however, that attempts at dry separation proved unsatisfactory -- the individual pages sometimes would crack or would split so that one writing surface would adhere to one part of the mass needing separation, while the opposite surface of the same folio would adhere to the other part of the mass, leaving no writing visible at the place where the separation had been made.

Fortunately, the inks used in the MS proved not to be soluble in water, so it was possible to use a process whereby a fused mass of fragmentary pages could be soaked and submerged in distilled water, and then carefully separated by stripping off the layers one by one.  The results have been highly successful, for the most part, and although much work still remains to be done on various small fragments or portions of split folios that still adhere to the pages to which they were once adjacent, we are now in possession of about 90% of the surface of most preserved pages (when the water damaged portions are joined to the relatively undamaged portion of each page).  As each group of water damaged fragments was separated, dried, and cleaned, Xerox copies were made from which to work in identifying and reconstructing the individual pages.  Careful record was also kept of the sequence of pages in the fused portions, so that it would be possible to reconstruct with precision the order in which the pages occurred when the MS was damaged by water.  Table 1 records the results.  This information is crucial for the efficient and accurate identification of small fragments still adhering to their adjacent pages and for assistance in reading occasional "mirror writing" that had transferred from one page to its neighbor due to the vicissitudes undergone by the manuscript before it came into our hands.
 

Reconstructing the "original" codex.

When the MS was damaged by water at some undetermined point in time past, its pages were already jumbled, with [[pp. 85-86 = charts; p. 87]] some pages upside down and/or backwards in relation to others. Thus we find that 59 folios are damaged at the bottom, and 49 folios at the top.  On many of the former folios, the original page numbers are still preserved -- a fact which assisted us greatly in restoring the pages to their original sequence that existed when the binding of the codex was still intact.\5/  Table 2 presents a reconstruction of the original page and signature format of the codex,\6/ before it became disarranged and ultimately damaged by water.
\5/ There was some confusion in numbering in the original codex: the lower part of two leaves (four pages) from the center of signature 10 are still joined together, but their page numbers are not preserved.  On either side of this material come pp. 179/180 and 184/185, with numbers preserved.  This leaves only three numbers (181-182-183) for four pages (in a signature of 10 folios).  Although none of the page numbers in signature 1 is preserved (and there may have been a title page and similar material preceeding what is designated "1/2" on chart 2), there had to be an even number of pages between the folios "7/8" and "14/15".  If six pages were necessary for the intervening material (Ps 4.5b - 5.12b), and if Ps 1.1 began on "page 1", there was a numbering error somewhere prior to pp. 24/25 (our first preserved page numbers).  Possibly only four pages were required for Ps 4.5b - 5.12b, and possibly the page containing Ps 1,1 was numbered "page 2".  In that event, our reconstruction would need to be adjusted so that signature 1 contains the following folios: 2/3, 4/5, (6/7 lost), 8/9, (10/11, and 12/13 lost), 14/15, 16/17.  Perhaps a title-page folio was prefixed.

\6/ Evidence of actual signature designations is preserved on pp. 017 (alpha), 037 (broken), 056 (gamma), 076 (delta), 096 (epsilon), 116 (digamma), 117 (zeta), 152 (illegible), 153 (theta), 172 (theta), 191, 192 (iota alpha), 211 (iota broken).  Two sets of pages from the center of their respective signatures have been preserved with their inner margins still joined: 065 / 066 with 067 / 068 (signature 4) and 181 / 182 with 183a  / 183b (signature 10).  Notice also that pages from other such sets were adjacent when the materials were damaged: 45/46 with 47/48 (signature 3), 125/126 with 127/128 (signature 7), 143/144 with 145/146 (signature 8), 161/162 with 163/164 (signature 9), 200/201 with 202/203 (signature 11), but not the central pages in signatures 2, 5, 6 and (probably) 12. Signature 1 is anamolous.


Chart 1:[[p.85]]   UNIVERSITY MUSEUM E 16261 COPTIC/SAHIDIC PSALTER
 
Correlations between (1) the sequence of pages at the time the codex became damaged by water and (2) the original order of pages (before dislocation) as determined by preserved page numbers and sequence of content.  In the following lists, the initial number of a two number notation (e.g. 7/8, 226/227) always designates pages that faced in the same direction (pages facing upward when the pile of leaves lies on a horizontal plane), and the second number indicates pages facing the opposite direction (facing down).
 
Sequence when damaged         Original page numbers

         1/2                           037 / 038
         3/4                           039 / 040
         5/6                           041 / 042
         7/8                           043 / 044
         9/10                         049 / 050
        11/12                        045 / 046
        13/14                        047 / 048
        15/16                        069 / 070
        17/18                        068 / 067
        19/20                         66/65
        21/22                        051 / 052
        23/24                        053 / 054
        25/26                        055 / 056 
        27/28                        057 / 058
        29/30                        148/147
        31/32                        139/140
        33/34                        141/142
        35/36                        143/144
        37/38                        145/146
        39/40                        138/137
        41/42                        136/135
        43/44                        134/133
        45/46                         29/28
        47/48                         31/30
        49/50                        164/163
        51/52                        162/161
        53/54                        159/160
        55/56                        158/157
        57/58                        156/155
        59/60                        154/153
        61/62                        152/151
        63/64                        124/123
        65/66                        209/208
        67/68                        211/210
        69/70                         004 / 003
        71/72                         007 / 008
        73/74                         002 / 001
        75/76                         015 / 014
        77/78                         018 / 019
        79/80                         020 / 021
        81/82                         022 / 023
        83/84                         017 / 016
        85/86                        186/187
        87/88                        190/191
        89/90                        188/189
        91/92                         27/26
        93/94                        179/180
        95/96                        184/185
        97/98                         85/86
        99/100                        64/63
       101/102                        60/59
       103/104                        61/62
       105/106                        96/95
       107/108                        93/94
       109/110                        92/91
       111/112                       194/195
       113/114                       196/197
       115/116                        33/32
       117/118                       192/193
       119/120                        024 / 025
       121/122                       169/170
       123/124                       107/108
       125/126                       221/220
       127/128                       171/172
       129/130                       173/174
       131/132                       175/176
       133/134                        77/78
       135/136                        79/80
       137/138                       207/206
       139/140                       222/223
       141/142                       166/165
       143/144                       168/167
       145/146                       128/127
       147/148                       126/125
       149/150                       129/130
       151/152                        87/88
       153/154                        89/90
       155/156                       102/101
       157/158                       100/99
       159/160                        98/97
       161/162                       103/104
       163/164                       105/106
       165/166                       109/110
       167/168                       132/131
       169/170                       112/111
       171/172                       198/199
       173/174                       216/217
       175/176                       225/224
       177/178                       205/204
       179/180                       203/202
       181/182                       201/200
       183/184                       114/113
       185/186                       115/116
       187/188                       219/218
       189/190                       226/227
       191/192                       214/215
       193/194                       181/182
       195/196                      183a/183b
       197/198                        82/81
       199/200                       117/118
       201/202                       119/120
       203/204                       121/122
       205/206                        84/83   
       207/208                       177/178
       209/210                        76/75
       211/212                        74/73
       213/214                        72/71
       215/216                       150/149
                                    Missing:
                                       5/6
                                       9/10
                                      11/12
                                     13a/13b (?)   
                                      34/35
                                     36a/36b (?)
                                     212/213
                                     228/229
                                      etc.

Chart 2:[[p.86]]   UNIVERSITY MUSEUM E 16261 COPTIC/SAHIDIC PSALTER
 
Correlation between (1) reconstructed original order of the folios and (2) the sequence in which the dislocated pages appeared when the codex was damaged by water.  Folios in which the water damage affects the upper portion of the page are noted by underlining in column (1).  The other folios are damaged in their lower portion.
 
Original page      Sequence when
numbers            damaged
--------------------------------
Signature 1
 
     001 / 002         074/073
     003 / 004         070/069
     005 / 006            
     007 / 008         071/072
     009 / 010            
     011 / 012
     13a / 13b
     014 / 015         076/075
     016 / 017         084/083
--------------------------------
Signature 2
 
     018 / 019        077/078
     020 / 021        079/080
     022 / 023        081/082
     024 / 025        119/120
     026 / 027        092/091
     028 / 029        046/045
     030 / 031        048/047
     032 / 033        116/115
     034 / 035           
     36a / 36b
--------------------------------
Signature 3

      037 / 038         001/002
      039 / 040         003/004
      041 / 042         005/006
      043 / 044         007/008
      045 / 046         011/012
      047 / 048         013/014
      049 / 050         009/010
      051 / 052         021/022
      053 / 054         023/024
      055 / 056         025/026
--------------------------------
Signature 4

      057 / 058        027/028
      059 / 060        102/101
      061 / 062        103/104
      063 / 064        100/99
      065 / 066        020/019
      067 / 068        018/017
      069 / 070        015/016
      071 / 072        214/213
      073 / 074        212/211
      075 / 076        210/209
--------------------------------
Signature 5

     077 / 078         133/134
     079 / 080         135/136
     081 / 082         198/197
     083 / 084         206/205
     085 / 086         097/098
     087 / 088         151/152
     089 / 090         153/154
     091 / 092         110/109
     093 / 094         107/108
     095 / 096         106/105
--------------------------------
Signature 6

     097 / 098        160/159
     099 / 100        158/157
     101 / 102        156/155
     103 / 104        161/162
     105 / 106        163/164
     107 / 108        123/124
     109 / 110        165/166
     111 / 112        170/169
     113 / 114        184/183
     115 / 116        185/186
--------------------------------
Signature 7

     117 / 118       199/200
     119 / 120       201/202
     121 / 122       203/204
     123 / 124       064/063
     125 / 126       148/147
     127 / 128       146/145
     129 / 130       149/150
     131 / 132       168/167
     133 / 134       044/043
     135 / 136       042/041
--------------------------------
Signature 8

     137 / 138        040/039
     139 / 140        031/032
     141 / 142        033/034
     143 / 144        035/036
     145 / 146        037/038
     147 / 148        030/029
     149 / 150        218/217
     151 / 152        062/061
--------------------------------
Signature 9

     153 / 154        060/059
     155 / 156        058/057
     157 / 158        056/055
     159 / 160        053/054
     161 / 162        052/051
     163 / 164        050/049
     165 / 166        142/141
     167 / 168        144/143
     169 / 170        121/122
     171 / 172        127/128
--------------------------------
Signature 10

     173 / 174        129/130
     175 / 176        131/132
     177 / 178        207/208
     179 / 180        093/094
     181 / 182        193/194
    183a  / 183b    195/196
     184 / 185        095/096
     186 / 187        085/086
     188 / 189        089/090
     190 / 191        087/088
--------------------------------
Signature 11

     192 / 193       117/118
     194 / 195       111/112
     196 / 197       113/114
     198 / 199       171/172
     200 / 201       182/181
     202 / 203       180/179
     204 / 205       178/177
     206 / 207       138/137
     208 / 209       066/065
     210 / 211       068/067
--------------------------------
Signature 12

     212 / 213
     214 / 215      191/192
     216 / 217      173/174
     218 / 219      188/187
     220 / 221      126/125
     222 / 223      139/140
     224 / 225      176/175
     226 / 227      189/190                   
        

Portions of most of the pages from twelve signature units are preserved.  Units two through twelve are written in a consistent Coptic hand (labeled "hand 3" above), but the first signature contains two startlingly different styles of writing.  Indeed, Psalm 7 actually begins on page 016b ("hand 2"), but the material was subsequently crossed out, and begins anew on page 018 ("hand 3").  It seems probable that a codex from which the initial signature had been damaged or lost was repaired in two stages: some pages from another old and damaged codex of slightly smaller format were used for some of the missing material (at least pages 014 / 015 and 016 / 017), but the opening pages had to be supplied from a third source, perhaps written by the repairer himself ("hand 1" pages 001 / 002, 003 / 004 , 007 / 008; doubtless also 5/6, and perhaps 9/10, etc.).  Thus a composite initial signature was prefixed to the other material.\7/  [[p. 88]] 

\7/ There seems to be a trace of some sort of glue on the inner margin of folio 016 / 017.  Interestingly, many of the pages of the codex still preserve (three) holes along the inner margin through which a binding thread apparently passed.  Does this mean that the signature structure based on groups of folded double pages was no longer functional (with each signature sewn separately, then united by means of the threads and cover) when the codex was repaired for its last useful stage of life?

How many pages were included in this reconstituted codex is difficult to determine.  The last preserved page (p.227) breaks off in the middle of a verse (Ps 49.3). and would not have been the final page of signature 12 if, like most of the other signatures, it contained 10 folios (20 pages) (Only signature 8 has 8 folios.) If signature 12 also had 10 folios, it would have extended to page 231.  But even that would not provide sufficient room to include the entirety of Ps 49.  Thus it may be that we have only a portion of a much larger Psalter codex, or perhaps the first volume of a multi-volume Psalter.  Further research into the habits and techniques employed in producing Coptic codices is necessary in this connection.\8/
\8/ Reference needs to be made to the following articles: H. Ibscher, "Koptische Bucheinhande aus Ägypten," Berliner Museen 49 (1928), pp. 86-90; D. Cockerell "Development of Bookbinding Methods -- Coptic Influence," Library Service 4.13 (1932/1933), p. 4, and 20 (1939/40), pp. 214-233.

Dating the writing.

Careful paleographical research on the different hands in the codex is being pursued by Janet Timbie.  Initial impressions suggest a date somewhere in the ninth to eleventh century for hands 2 and 3, with "hand no. 2" possibly somewhat earlier (early tenth century?) than "hand no. 3" (end of tenth century?).  Both represent Stegemann's "schmaler Stil."\9/  Comparative material for dating the relatively careless and sloppy "hand no. 1" is more difficult to isolate.  "Hand no. 1" leaves the impression of being later than hands 2 and 3 (which is expected if the reconstruction offered above is accepted), but at this point it is impossible to give more than a general tenth to thirteenth century estimate, with some inclination towards a twelfth century date.
\9/ V. Stegemann, Koptische Palaographie (Heidelberg: Akademie der Wissenschaften, 1936).  For recent palaeographical charts see also M. Cramer, Koptische Palaographie (Wiebaden: Harrassowitz, 1964). See now Janet Timbie, "The Dating of the Coptic/Sahidic Psalter Codex from the University Museum in Philadelphia," Le Muséon 88 (1975) 387-390.

The text represented.

Work has only begun on collating this witness with other available materials.  The initial impressions gathered by Ms. Timbie, Mr. Woerhle, and myself are that our text has few major differences form the sixth/seventh century Psalter published by Budge in 1898.\10/  Orthographic [[p. 89]] variations abound.  A tendency to include the conjunctive "and" where it is lacking in Budge's text has also been noted, but whether this is peculiar to our manuscript or not still needs to be determined.\11/  As could be expected, the titles of the Psalms also show variation between our codex and other Sahidic witnesses (which vary significantly among themselves).  The absence of any convenient collection of Sahidic textual data on the Psalms and of any standard critical edition of the Sahidic Psalms makes this aspect of work on the Pennsylvania Sahidic Psalter especially difficult and time consuming.  Any detailed report must be postponed to a further date.
\10/ E. A. W. Budge, The Earliest Known Coptic Psalter (London: Paul, 1898).  For helpful lists of Sahidic MSS of the Psalms, see A Rahlfs, "Die Berliner Handschrift des Sahidischen Psalters," Abhandlungen der kgl. Gesell. der Wiss. zu Göttingen 4,4 (Berlin, Weidmann, 1901), pp. 5-7; also H. Worrell, The Coptic Psalter in the Freer Collection, ("University of Michigan Studies: Humanistic Series," 10.1; New York: Macmillan, 1916), p. xxiii f. For further bibliography, see W. Kammerer, A Coptic Bibliography (Ann Arbor: Univ. of Michigan Press, 1950), items 843-862.  
\11/ See, for example, Ps 17.48c (18.49c) "and saves me from a man" (only Theodoret supports this reading according to Rahlfs' apparatus in his Gottingen LXX ed. of 1931 [19672]); Ps 21.2b (22.3b) "and by night" (so also Greek and Hebrew); Ps 21.30a (22.31a) "and my seed" (so also Greek but not Hebrew). [William Adler subsequently made a careful comparison of the available materials, and his information needs to be added to this report.]

Postscript (2010)

Color Images have been added (more cross referencing needed)
add ifnormation from Adler textual notes

//end, 15ja2010//