Deciphering cuneiform tablets for the GKAB project

It often takes a long time to make full sense of the inscription on a cuneiform TT  tablet TT . Between 2007 and 2012 a team at the University of Cambridge worked on a research project to edit, interpret and contextualise the tablets from Nabu's temple in Kalhu, along with three other cuneiform "libraries". Here we take one tablet from that findspot TT  as a case study, following it from excavation in 1955 to online publication in 2010.

From site to museum

Image 1: Mallowan's PGP  team found tablets in the Ezida PGP , in a room directly opposite the god Nabu's sanctuary, as shown on this isometric plan (1). © BSAI/BISI. View large image.

In 1955 Max Mallowan's PGP  team began systematic excavation of Nabu's PGP  temple Ezida PGP , after preliminary explorations by Layard PGP  over a century earlier in 1848 (2). Across the inner courtyard from Nabu's own shrine, immediately opposite it, they found a room with a particularly wide doorway (Image 1). It had housed the temple's collection of scholarly tablets and must have been in use right until the end of empire in 612 BC. After the temple was abandoned, the walls eventually collapsed in on themselves, preserving — for the moment — the tablets TT  inside. However, the archaeologists soon discovered that later occupants of Kalhu had used the room as a quarry for bricks. As Mallowan later recalled,

Since most of the tablets were of sun-dried clay their disturbance had done them no good. When the Hellenistic TT  occupants of the site dug into these rooms, mostly in order to tear up the pavements for burnt bricks required in the construction of their graves, they cast aside in dumps these documents which were of no interest to them. Often the clay texts must have been exposed to rain before they were eventually shovelled back as the pits were refilled to make good the ground surface which had to be relevelled for building (3).

Image 2: A badly damaged fragment of tablet with a "hand-raising" (šuillakku) prayer to the goddess Tašmetu PGP  (online translation: CTN 4: 166). British Museum, ND 4405/24. View large image (246 KB).

This meant that the surviving tablets were very badly damaged and weathered (Image 2). It would take years — decades — to make sense of them.

The team's usual practice had so far been to decipher newly excavated tablets as fast as possible. But that was not feasible for the tablets from Ezida. There were simply too many and they all needed careful conservation TT  and cleaning. Following normal procedure, the tablets were taken to the Iraq Museum TT  in Baghdad PGP , where over the next few decades they were studied by several Assyriologists TT  associated with the British School of Archaeology in Iraq TT . The Iraqi Director-General of Antiquities TT , Dr Faisal El-Wailly PGP , also unusually allowed some tablets to be loaned to the British Museum TT  for long-term study (4).

From clay to paper

Image 3: The hand-copy TT  of ND 4405/24 was made by Donald Wiseman PGP  and published as CTN 4: 166 on plate 98 of that volume (5). © BSAI/BISI. View large image.

From the beginnings of academic Assyriology in the 19th century until the advent of high-quality digital photography and online publication in recent years, the standard way to disseminate cuneiform texts was to publish line drawings (or "hand-copies" TT ) of them, accompanied by a catalogue that briefly listed their contents (Image 3). This was the most efficient way to make often difficult texts available as widely as possible, in the hope that other experts would then be able to study them in more depth. Gradually the hand copies of the Ezida texts accumulated as successive scholars worked on the tablets, especially Donald Wiseman PGP  in the 1960s and Jeremy Black PGP  in the 1980s. The BSAI also encouraged the sharing of copies before they were published, so that colleagues could incorporate Nimrud manuscripts into studies of particular ancient works or genres (e.g., pieces of the ominous calendar Iqqur īpuš CTN 4: 50–53 (6)).

In 1985-86 Iraqi excavators led by Muzahim Mahmoud Hussein PGP  discovered a few more tablets in Ezida that the British archaeologists had overlooked (7). These needed to be incorporated into the publication too. But the Gulf War TT  of 1991, and the international sanctions TT  that followed, made Iraq virtually inaccessible to international researchers throughout the 1990s. Ideally there was still work to do, even though the tablets in Baghdad were out of bounds. However, in the mid-90s, forty years after their original discovery, Wiseman and Black decided that it was time to publish in any case.

The resultant book, Literary Texts from the Temple of Nabû (Cuneiform Texts from Nimrud, 4), appeared in 1996 (8). After a brief introduction it presents a systematic catalogue of the 259 tablets that Mallowan's team found in Ezida plus a list of the 14 discovered by Mahmoud's. There then follow hand copies of almost all of the 259 British-excavated tablets, plus a few photographs at the end for illustrative purposes. This allowed large numbers of Assyriologists access to the contents of tablets that to this day remain practically inaccessible to most, in the Iraq Museum. In the two decades since the publication of CTN 4, around half the texts in it have been re-edited or re-used somehow.

From paper to pixels

Image 4: The GKAB project used Oracc's ATF standard to transliterate texts for publication, as shown here by Dr Greta Van Buylaere's file for CTN 4: 166. View large image.

Specialist Assyriological publications are typically hard to find and hard to read. Most books are printed in batches of only a few hundred, while many journals are not available online. Both books and journals can be prohibitively expensive. Even if researchers or students have access to a well-stocked academic library, they often find that Assyriologists write for the small community of cuneiform-literate experts and not for interested outsiders.

Online publication of cuneiform texts is rapidly helping to solve the first problem, if not always addressing the second one. In 2007 the UK Arts and Humanities Research Council funded a research team at the University of Cambridge to produce an online edition of the tablets from Ezida, along with the contents of three other cuneiform "libraries", and to analyse them historically. The project was called The Geography of Knowledge in Assyria and Babylonia, 800–200 BC and used Oracc for online publication.

Let us look at the first process to start with, following a tablet that the British excavators numbered ND 4405/24. The index to Wiseman and Black's book (9) tells us that ND 4405/24 (we will come back to that number) appears in the volume as CTN 4: 166. The tablet was copied by Donald Wiseman and is still on long-term loan to the British Museum (Image 2, Image 3). Even though only the top right-hand corner now survives Wiseman was able to identify the inscription on it as a šuillakku-prayer to the goddess Tašmetu PGP , Nabu's divine consort, who was also worshipped in Kalhu's Ezida temple. When project member Dr Greta Van Buylaere chose to work on it in 2010, along with other prayers like it, it had not yet been formally studied.

The first stage was to make an alphabetic transliteration of the cuneiform signs on the tablet, by reading Wiseman's copy and by reading a digital photo of the tablet itself, made during a study visit to the British Museum (Image 4). The initial challenge, as always, was to identify individual words, as cuneiform script does not typically leave spaces between them. Tašmetu's name was partially visible on the first line, while a form of the Akkadian TT  word šurbû, "very great", was clear at the beginning of the second. This is a common epithet TT  of gods and goddesses, which occurs a dozen more times in the GKAB corpus.

The presence of this word encouraged Van Buylaere to look for other adjectives typically associated with deities. It turned out that the first five lines of the prayer were full of them. The fact that she was working on other prayers in the GKAB corpus at the same time, and was therefore familiar with the sort of language to expect in this genre, was particularly helpful.

However, because the tablet was not complete, some words had been partially broken or damaged. In later lines it wasn't possible to identify every word, or even every sign. Following standard Assyriological conventions, Van Buylaere marked missing sequences with square brackets [...], readable signs which didn't fit into words with CAPITAL LETTERS, and unreadable signs as x. In the process she also discovered a few discrepancies between Wiseman's copy and the original inscription, marking the new readings with an asterisk *.

Image 5: Oracc's lemmatisation program draws on large pre-existing glossaries of Akkadian and Sumerian and allows editors to add new words too, as seen here in Dr Greta Van Buylaere's file for CTN 4: 166. View large image.

The next stage in the process was to lemmatise and translate the text (Image 5). These are two facets of the same procedure. Lemmatisation consists of linking each word in the text with the headword of a dictionary entry, and deciding which grammatical form it takes. This is particularly important in a language like Akkadian in cuneiform script, as variant spellings on the one hand, and different genders, cases and tenses on the other, can dramatically change the appearance of a word. In an online corpus, much lemmatisation can be done automatically by the computer, based on words and spellings that are already in the system. However, it also needs to be checked by a human being.

Careful lemmatisation results in accurate translation, as Van Buylaere could be confident that she knew the meaning and function of almost every word in the text. For the more damaged passages, her translations had to be more cautious. These show up in the online edition in italics.

Finally, project leader Dr Eleanor Robson proofread the transliteration, lemmatisation and translation, checking for typos and for reasonable consistency with the rest of the GKAB corpus. She then uploaded the file and integrated its lemmatisation with the project's glossary. That enables users of the online project to search for any English or Akkadian word in the text, regardless of spelling, and to see where else it appears in the corpus.

From text to context

Editing cuneiform manuscripts of a single ancient composition or genre is a useful way to gain an overview of a work that may only survive in fragments at any one location. It also enables historians to track changes in that work across time and place. However, as well as these diachronic TT  studies it is also useful to study individual tablets in their contexts of use, production and abandonment. By looking at archival and archaeological assemblages TT , historians can begin to see why particular individuals and groups were interested in particular compositions, and how those compositions fitted into their composers' and copyists' wider intellectual, social and political world.

The GKAB project began that analytic process for an academic readership (e.g., (10), (11), (12)). This website, which is a spin-off from the GKAB work, is designed to reach out beyond specialists to anyone who is interested in the ancient world.

What more can be said, then, about the tablet ND 4405/24, and its prayer to Tašmetu? The excavation number — prefaced by ND for Nimrud — can be linked back to notes made in the field about exactly where the object was discovered. It turns out that ND 4405/24 was one of 114 pieces of tablet found in a pit that had been dug in antiquity in or near to the temple's tablet room (13). Eighty of those pieces were registered together under the main number 4405, with our tablet assigned the secondary number 24. Unfortunately this doesn't tell us anything at all about where the tablet was originally used or kept for, as we have seen, in this case the tablets were dug up and dumped again a few hundred years after the temple had been abandoned.

Many of the scholarly tablets found in Nabu's temple had colophons TT  on the back of them, naming the composer or copyist of the text. This information can allow historians to glimpse the range of works a particular individual produced or owned, and/or to situate them within a particular family or professional group. Again, we draw a blank in this particular case, because the surviving surface of the back of our tablet is blank. And we do not yet know enough about spelling habits or handwriting styles to be able to identify the author from those features alone.

That leaves us with the content of the text itself. It is one of several šuʾila (Sumerian) or šuillakku (Akkadian) prayers to deities found in Ezida:

ND 4324 = CTN 4: 165 [not edited]

A bilingual prayer to Nabu, badly adapted from a widely circulating prayer to the Babylonian PGP  god Marduk PGP , which ends in a prayer for the late seventh-century king Sin-šarru-iškun PGP  (14)

ND 4405 = CTN 4: 166

Our fragmentary prayer to Tašmetu, in Akkadian, which has no known duplicates

ND 5487 = CTN 4: 167

An Akkadian prayer to the wise god Ea PGP , which was widely known and used in purification rituals TT  (15)

ND 5493 = CTN 4: 168

A compilation of seven prayers to the goddesses Ištar PGP , Nanaya PGP , Nisaba PGP , and Tašmetu, most of which are also known from other Assyrian cities (16)

ND 4405/29 = CTN 4: 169 [not edited]

An Akkadian prayer to the Babylonian god Marduk, also known from other cities

ND 4405/72 = CTN 4: 179

A fragmentary catalogue of Akkadian šuillakku-prayers to gods including Papsukkal PGP , Nergal PGP  and the Arrow constellation PGP , none of which survive in the collection itself.

None of these texts, apart from the first, has any distinguishing features that allow it to be dated. There is no particular reason to suppose that they were written as a set. Yet together they form a coherent group of prayers to deities that mattered to the literate scholars of the Assyrian royal court: Nabu and Tašmetu most obviously, but also divine figures closely associated with them. Marduk was Nabu's father, while the three goddesses were often thought to have much in common with Tašmetu. So even in these anonymous, enigmatic fragments we can see a group ideal of deity taking shape. Some of it was shared with the wider literate community of early first-millennium Assyria and Babylonia, but some — such as our hymn to Tašmetu — currently seems to be unique to the scholars of Kalhu.

References

1. Mallowan, M.E.L., 1966. Nimrud and Its Remains, vols. I-II, London: Collins, pp. I, 233 fig. 195. (Find in text ^)
2. Oates, D., 1957. "Ezida: The Temple of Nabu", Iraq 19, pp. 26 (PDF available via JSTOR [http://www.jstor.org/stable/4199614] for subscribers). (Find in text ^)
3. Mallowan, M.E.L., 1966. Nimrud and Its Remains, vols. I-II, London: Collins, pp. I 271. (Find in text ^)
4. Wiseman, D. and J.A. Black, 1996. Literary texts from the temple of Nabû (Cuneiform Texts from Nimrud 4), London: British School of Archaeology in Iraq (free PDF from BISI), pp. v. (Find in text ^)
5. Wiseman, D. and J.A. Black, 1996. Literary texts from the temple of Nabû (Cuneiform Texts from Nimrud 4), London: British School of Archaeology in Iraq (free PDF from BISI), p. 41. (Find in text ^)
6. Labat, R., 1965. Un calendrier babylonien des travaux des signes et des mois (séries iqqur îpuš) (Bibliothèque de l'École des Hautes Études 321), Paris: Librarie Honoré Champion. (Find in text ^)
7. Hussein, M.M. and J.A. Black, 1985–86. "Recent work in the Nabu Temple, Nimrud", Sumer 44: 135–55. (Find in text ^)
8. Wiseman, D. and J.A. Black, 1996. Literary texts from the temple of Nabû (Cuneiform Texts from Nimrud 4), London: British School of Archaeology in Iraq (free PDF from BISI). (Find in text ^)
9. Wiseman, D. and J.A. Black, 1996. Literary texts from the temple of Nabû (Cuneiform Texts from Nimrud 4), London: British School of Archaeology in Iraq (free PDF from BISI), pp. 25, 41. (Find in text ^)
10. Robson, E., 2011. "The production and dissemination of scholarly knowledge", in K. Radner and E. Robson (eds.), The Oxford Handbook of Cuneiform Culture, Oxford: Oxford University Press, 557–76. (Find in text ^)
11. Robson, E., 2013. "Reading the libraries of Assyria and Babylonia", in J. König, K. Oikonomopoulos, and G. Woolf (eds.), Ancient Libraries. Cambridge: Cambridge University Press, 38–56. (Find in text ^)
12. Robson, E., 2014. "Tracing networks of cuneiform scholarship with Oracc, GKAB and Google Earth", in M. Rutz and M. Kersel (eds.), Archaeologies of Text: Archaeology, Technology and Ethics (Joukowsky Institute Publications, 4), Oxford: Oxbow Books, 142–63. (Find in text ^)
13. Wiseman, D. and J.A. Black, 1996. Literary texts from the temple of Nabû (Cuneiform Texts from Nimrud 4), London: British School of Archaeology in Iraq (free PDF from BISI), p. 49. (Find in text ^)
14. Cooper, J.S., 1970. "A Sumerian šu-íl-la from Nimrud with a prayer for Sin-šar-iškun", Iraq 32: 51-67 (PDF available via JSTOR for subscribers). (Find in text ^)
15. Foster, B.R., 2005. Before the Muses: An Anthology of Akkadian Literature, Bethesda, MD: CDL Press, pp. III.36b. (Find in text ^)
16. Lambert, W.G., 1999-2000. "Literary Texts from Nimrud", Archiv für Orientforschung 46-7: 149-155 (PDF available from JSTOR for subscribers), pp. 152-155. (Find in text ^)

Eleanor Robson

Eleanor Robson, 'Deciphering cuneiform tablets for the GKAB project', Nimrud: Materialities of Assyrian Knowledge Production, The Nimrud Project at Oracc.org, 2019 [http://oracc.museum.upenn.edu/nimrud/modernnimrud/atthemuseum/decipheringtablets/]