Cuneiform Digital Library Bulletin
ISSN 1540-8760
© Cuneiform Digital Library Initiative 

CDLI Publications
Editorial Notes
Bibliographical Tools

PDF Version of this Article
Get Acrobat Reader


A Note on Ur III Text Duplicates

Jacob L. Dahl < >
University of California, Los Angeles

duplicate, Ur III, Umma, Drehem

This material is based upon work supported by the
National Science Foundation under Grant No. 0000629.

§1. Among the tens of thousands of sealed and unsealed Ur III documents that have been published to date a couple of hundred unsealed documents have the subscript gaba-ri kišib-ba, “copy of a sealed document” (see G. Selz, ASJ 17 [1995] 251-274, for a philological discussion of gaba-ri, which following him [esp. p. 270] and others I translate conventionally “copy”). So far, the search for the expected administrative pair of a sealed document and its corresponding copy has been futile.


§2. There do exist, however, a small number of documents that are (almost) exact copies of each other (see for example M. Yoshikawa, “A New duplicate of YOS IV, No. 67 // V. Scheil, RA24/1, No.8c,” ASJ 7 [1985] 191-192). Some of these copies have been proven to be modern “fakes” made from a mold, presumably fabricated by using an original “clay envelope” to produce the copy (see M. Hilgert, “Notes and Observations on the Ur III Tablets from the Oriental Institute,” JCS 49 [1997]  45-50).


§3. For a list of Drehem duplicate records see M. Hilgert, OIP 121, pp. 40-42 (table 4.2). Whereas the two texts presented here are the first pair of a sealed tablet and its copy to have been recognized in the Umma corpus, Drehem has yielded numerous such examples. Hilgert describes pp. 31-42 the Drehem duplicates that can be distinguished semantically from Umma examples: the term gaba-ri kišib-ba is never used in Drehem duplicates (OIP 121, p. 31 with n. 100). Whereas Umma texts using the term gaba-ri kišib-ba were for the most part copies of receipts, all Drehem tablet copies documented transfers rather than receipts (OIP 121, pp. 38-39). Most Umma duplicates relate to the administration of the “estate” of the governor, and there is evidence to suggest that Umma duplicates were those documents the governor kept while posting the originals to Ur for inspection (for example, AAICAB 1, 1: Ashm. 1912-1143 [Šulgi 28], rev. 8: gaba-ri kišib e2-gal gal2-am3, “copy of a sealed tablet which is in the palace”). A few tablet containers (pisan dub-ba) contained copies of sealed tablets (for example, UTI 3, 2098, dating to Amar-Suen 9), most of which documented the affairs of the governor´┐Żs office; the person about whom the most duplicates were written, Lukala, fulfilled an essential function in the household of the governor, perhaps as chief household administrator (šabra (e2); see MVN 16, 1294, dated to Šu-Suen 3 x).


§4. All duplicate documents qualified by gaba-ri kišib-ba are unsealed. Although it is now in most cases possible to reconstruct the original document from its copy, it may be opportune to record some further observations here. The copy of a document that contained the statement ki PN1-ta kišib PN2, “from PN1, sealed document of PN2,” was recorded as ki PN1-ta gaba-ri kišib(-ba) PN2, “from PN1, copy of the sealed document”. The copy of a document that contained the statement ki PN1-ta PN2 šu ba-ti, “from PN1, PN2 received,” was recorded as ki PN1-ta PN2 šu ba-ti gaba-ri kišib-ba, “from PN1, PN2 received, copy of the sealed document” only when the receiving and the the sealing party were identical (see, for example, TCNU 460). When the sealing party and the receiving party were different persons (for example, in UCP 9/2/2, 1, dating to Amar-Suen 2 ix), the duplicate indicated that it was a “copy of the sealed document of PN3” (gaba-ri kišib PN3; for example, TLB 3, 68).


§5. In addition, gaba-ri is used a number of times as a “note” on the edge of a document, apparently with the simple meaning “copy”. A good example of this bookkeeping device is YOS 4, 79. This account, recording four distinct transactions, contained the following information on its edge: gaba-ri 3(diš)-kam, “it is the copy of three sealed tablets” (where we would expect the number “four”, three is possible, since two transactions may have been recorded on one and the same sealed document).


§6. In all likelihood, a renewed study of the entire body of Ur III administrative texts, with the goal of reconstructing their ancient archives, will provide us with numerous examples of such Umma sealed documents and their copies as the pair transliterated below, both of which are housed in the British Museum. The sealed document was published by T. Fish as MCS 8, 97 (= BM 113102, collation Orient 17, 43). A transliteration of its unsealed copy (BM 108081, unpublished) is included here, with the kind permission of its editors, M. Molina and M. Such-Gutierrez (to appear in the Nisaba series). Both texts are dated to the 3rd month of the 9th year of Amar-Suen.



MCS 8, 97, BM 113102

BM 108081







1(aš) gu2 2(u) 3(diš) ma-na

siki gir2-gul

1(aš) gu2 2(u) 3(diš) ma-na

siki gir2-gul

1 talent and 23 mana of

girgul wool,


udu-bi 4(u) 5(diš)

udu-bi 4(u) 6(diš)?

its sheep are 45?,


udu ba-ur4


(A: sheared sheep).


ki an-na-hi-li-bi-ta

ki an-na-hi-li-bi-ta

From Anahilibi.






kišib ensi2

gaba-ri kišib ensi2

(B: Copy of) the governor‘s



sealed tablet.

   (blank space)

(blank space)



iti še-kar-ra-gal2-la*


iti še-kar-ra-gal2-la


Month “barley arrived at the harbor”,


mu en dnanna* kar*-zi*-da


mu en dnanna kar-zi-da


Year: “the en(-priestess) of Nanna

of Karzida (was installed)”.





Col. 1









lugal kal-ga


The strong king,


lugal uri5ki-ma


The king of Ur,


lugal an-ub-da limmu2-ba


The king of the four corners.

Col. 2











the governor




of Umma,




is your slave.

Version: 30 June 2003