A new, ambitious project led by Enrique Jiménez started in April 2018 at Ludwig-Maximilians-Universität München: the goal of the Electronic Babylonian Literature (eBL) Project is to significantly contribute to the reconstruction of Babylonian literature of the first millennium BCE. Babylonian literature is a key part of the world’s cultural heritage: the project aims to preserve and understand its legacy, in our modern-day world but also for future generations. We are very grateful to Enrique Jimenez for having sent to us this engaging and comprehensive article, which tells us much not only about the Electronic Babylonian Literature Project and the nature of the thousands of fragments it deals with, but also on “where we come from” and how the pioneers of Assyriology dealt with the thorny problem of the fragmentariness of Mesopotamian Literature.
⁂
[… who s]aw the Deep, […] the country,
[who] knew […], […] all […]
[… who] saw the Deep, […] the country,
[who] knew […], […] all […]
He who saw the Deep, the foundation of the country,
who knew the proper ways, was wise in all matters!
Gilgamesh, who saw the Deep, the foundation of the country,
who knew the proper ways, was wise in all matters!
The first quote represents the beginning of the Epic of Gilgamesh as known from the 19th century onwards (Jeremias 1891: 14). The second shows the text fully restored, in the form it achieved over one hundred years later, in 2007 (see George 2007). Throughout the twentieth century, therefore, only the fragmentary version of the prologue of the Epic was known: generations of readers, when first confronted with the foremost classic of ancient Mesopotamian literature, experienced the frustration of reading a fragmentary text, of being allowed merely a latticed glimpse into the world of the Babylonians. This frustration is every cuneiform scholar’s bread and butter, often to be consumed “when one struggles with a fragmentary text in the Students’ Room of the British Museum and suspects with more or less reason that unidentified pieces are lying in drawers just a few meters away” (Borger 1991: 41-42). It is precisely against this frustration that the Electronic Babylonian Literature Project has declared war.
The Electronic Babylonian Literature (eBL) Project started in April 2018 at Ludwig Maximilian University of Munich thanks to the generous support of a Sofja Kovalevskaja Award from the Alexander von Humboldt Fundation. The goal of the project is to bring Babylonian literature to the point of what can currently be reconstructed. Moreover, it aims to make accessible a large mass of transliterations of fragments of cuneiform tablets and a tool to allow scholars to search it quickly, thus providing a lasting solution to the abiding problem of the fragmentariness of Mesopotamian Literature.
1. The Fragmentarium
The backbone of the project is the Fragmentarium, a corpus that one day will contain transliterations of all cuneiform fragments in the world. The Fragmentarium is conceived as a pool where scholars can fish for new pieces of their texts. The Fragmentarium shares its motto with the Statue of Liberty: “Give me your tired, your poor, your huddled masses” – of fragments! Indeed, thousands of hapless fragments, which in their current form are fundamentally unusable, live in the Fragmentarium and await the day when they will be identified and made meaningful.
Figure 1: A page from George Smith’s notebooks
(© British Library Add MS 30413 f. 28a)
The compilation of a comprehensive Fragmentarium is an old dream in Assyriology, which many scholars have tried to realize since the dawn of the discipline. The first Assyriologist to transliterate all the fragments he got his hands on was George Smith, who left over thirty notebooks teeming with transliterations (see fig. 1). Smith already planned to let other scholars use his preliminary transliterations: indeed, he wrote in the last notebook, shortly before his death, “I intended to work it out but desire now that my antiquities and notes may be thrown open to all students. I have done my duty thoroughly” (British Library Add MS 30425 f. 28a). Smith’s desires were never realized, and that was also the luck of the collections of transliterations of many illustrious scholars after Smith, such as J. N. Strassmaier, F. W. Geers, and W. G. Lambert. These scholars’s admirable fragmentaria, thousands of pages filled with copies and transliterations of tablets and fragments, enjoyed very limited circulation during their lives and after their deaths, and thus they never constituted a final, universal Fragmentarium. The eBL project has several advantages over these scholars that leads us to believe that we will be more successful: first, the eBL Fragmentarium is a collective enterprise, in which the entire eBL team participates, as well as other scholars past and present. In addition to the Nachlässe of several giants of the field, kindly made available by their academic executors and processed into the Fragmentarium by L. Sáenz and E. Gogokhia, several contemporary Assyriologists have generously ceded their own collections of transliterations for the purpose of compiling a Fragmentarium, among them M. J. Geller, U. S. Koch, and W. R. Mayer.
However, the most important reason that makes us believe that ours will be the last time anybody will have to compile a Fragmentarium is that the eBL project makes use of computers. The usefulness of computers for identifying fragments was foreseen by several fragmentarists of the past: thus, in the margin of one of his many notebooks of transliterations, E. V. Leichty declares the contents of that page to be “many right sides of omens too fragmentary to identify but might be good for computer search” (fig. 2).
Figure 2: A page from Erle Leichty’s notebooks (© Penn Museum,
University of Philadelphia, Leichty Folio NB 911)
The eBL Fragmentarium has been compiled from the beginning by means of an electronic application (see below). In order to prioritize which fragments should be transliterated into the Fragmentarium, we have created a button, the “Path of the Pioneers,” which yields a new unidentified and unpublished fragment every time that it is clicked (see video below). Months of clicking the button have shown that the mass of fragments still to be transliterated remains humblingly large.
In spite of its lacunas, the current coverage of the Fragmentarium is vast. Over 13.500 fragments have been fully transliterated as of February 2020, totaling almost 140.000 lines of text of all genres (fig. 3).
Figure 3: eBL edition of the extispicy fragment BM.34474
Figure 4: Montage of the join of the tiny fragment K.17569 to a manuscript of the Hymn to the Queen of Nippur
Even in its current incomplete state it is one of the largest digital Assyriological corpora. We calculate that our coverage of unpublished Nineveh fragments is over 75%. This large coverage allows us to identify even tiny pieces and assign them to a composition, such as the minuscule fragment K.17569, which contains merely one sign per line, and which was joined by Zs. Földi to a manuscript of the Hymn to the Queen of Nippur (fig. 4). It is also remarkably easy to reconstruct entire tablets from tiny fragments, such as the manuscript of a lamentation K.4981+, put together by T. Mitto from no fewer than nine discrete fragments (fig. 5). Although the speed at which fragments are transliterated is necessarily fast, the Fragmentarium is lucky enough to have enlisted the help of H. Stadhouders, whose constant revisions guarantee the quality the editions.
Figure 5: Reconstruction of the manuscript K.4981+ from nine discrete fragments previously transliterated in the Fragmentarium
Figure 6: Manuscript NipNB1 (IM.77087) of Advice to a Prince (Photograph by Anmar A. Fadhil, by permission of the State Board of Antiquities and Heritage and The Oriental Institute of the University of Chicago)
In addition to the Nineveh collection, in April 2019 LMU Munich began to collaborate with the British Museum, thanks to the good offices of Jon Taylor, in digitizing the entire Babylon Collection (over 20.000 tablets). Two photographers in the British Museum, Alberto Giannese and Ivor Kerslake, have photographed over 6.000 tablets to date, hundreds of which have been transliterated in the Fragmentarium by, among others, A. Hätinen and E. Jiménez. These photographs will be made publicly available when the project is launched. Moreover, the project’s liaison in the Iraq Museum, Anmar A. Fadhil, has photographed hundreds of tablets for the eBL project with the kind permission of the Iraqi State Board of Antiquities and Heritage, which will also be published online when the project’s portal goes live (fig. 6).
An old Assyriological joke states that one day Assyriologists will realize that there was only one tablet at Nineveh, but one broken into many pieces. The intensive work of all team members on the Fragmentarium has resulted in the identification of over 1.000 new fragments of Babylonian literature and scholarship and the discovery of well over 150 joins in the first few months of the project. We are, however, merely scratching the surface of the Fragmentarium: the great leap forward will take place when the algorithms currently under development start to query the Fragmentarium systematically for new pieces of a given text. The reconstruction of the Mesopotamian primordial tablet will then be closer than it has ever been.
2. The Corpus
If the Fragmentarium is the backbone of the project, its showcase is the eBL Corpus. The Corpus is conceived to contain editions of all “classics” of Babylonian literature in the first millennium BCE, from the Epic of Creation to the Poem of the Righteous Sufferer and the Story of the Flood (see a list of texts to be edited in fig. 7, below).
Figure 7: Frontpage of the eBL Corpus, with the list of texts currently being edited as part of the project
The texts are furnished with a normalized version and a score edition (fig. 8). The eBL editions use many previously unpublished or unedited fragments, often identified by the members of the project, and thus represent cutting-edge versions of the texts. Newly identified fragments, as well as important new readings of texts, are being published in the series of articles From the Electronic Babylonian Literature Lab (Kaskal). Moreover, the eBL score editions are based on collation of all available manuscripts.
Figure 8: Prototype of the eBL edition of Enūma eliš I
It may seem redundant to collate all manuscripts of texts that are, as a rule, very well edited, but surprises here and there have shown it to be of great profit. One example will suffice: in a passage in the fourth tablet of Enūma eliš Marduk, having defeated Tiamat, carefully examines the goddess’ body in order to enact his “ingenious scheme.” A key word in l. 136 represents Tiamat’s corpse as a “foetus” (kūbu), a description that has fascinated scholars ever since the key manuscript was published in 1887, and has even attracted the attention of scholars of religion such as Mircea Eliade. Collation of the passage undertaken for the preparation of the eBL edition by A. C. Heinrich (fig. 9) has, however, shown the reading of the signs in every single edition of the Epic of Creation since the 19th century to be erroneous: as it turns out, the text describes Tiamat as a serkūpu, a word that must have troubled the Babylonians, since their own dictionaries explain it simply as “sea” (tiāmtu). Many other similar discoveries stud the eBL editions and turn them into the most reliable, up-to-date renderings of Babylonian literature. The editions will be constantly updated, so that every reader, specialist or not, will have immediate access to the current state of scholarship.
Figure 9: eBL edition of Enūma eliš IV 135–137
To make the editions even more accessible, B. R. Foster has kindly agreed to update his English translations of the Babylonian classics and to publish them online. Moreover, A. A. Fadhil will produce Arabic translations of some of the texts and equally publish them on the project’s website.
Last but not least, many unpublished manuscripts of Babylonian Literature in the Sippar Library are being prepared for publication by Fadhil and Jiménez, and will be used for the eBL electronic editions. A conservation project begun in 2018 has enabled three stays in the Iraq Museum of the conservator C. Gütschow, who has conserved several tablets from the Library and trained IM conservators in innovative tablet conservation techniques (see here, in Arabic, and here, in English). The tablets thus conserved, at their maximum readability, are in the process of being published, with the new findings being incorporated into the eBL editions.
3. The Platform
Figure 10: Transliteration of K.12564 parsed and saved as a data tree in the JSON format
The project’s developer, J. Laasonen, is developing a powerful platform for the transliteration of cuneiform texts. The application parses ATF into a data tree in the JSON (JavaScript Object Notation) format. The text so parsed is stored in MongoDB, a document-oriented NoSQL database that uses JSON-like format to store data (fig. 10). The database is hosted in two virtual machines at the Leibniz Supercomputing Centre (LRZ) of the Bavarian Academy of Sciences and Humanities. The JSON-like scheme is easily readable by the computer, and allows the easy lemmatization of Akkadian texts, which is based on the Concise Dictionary of Akkadian, courtesy of Harrassowitz.
Moreover, the application will align the standardized, normalized lines of the texts in the eBL corpus with the manuscript lines of the texts. As is well known, the cuneiform writing system knows no set orthography, which means that every word can be written in a number of different ways, only some of which are explicable or predictable. An Assyriological quip explains the reason for this to be that the Babylonians did not possess a copy of von Soden’s Grundriß der akkadischen Grammatik in their libraries. Be that as it may, the computer should learn from the body of aligned score editions how a given word in a given text can be written. Using this information, the computer should be able to make predictions about how other words in other texts should be written, and therefore to identify even the tiniest fragments in the pool of the Fragmentarium, regardless of their orthography.
At the end of the project we will make available a search field in which scholars can enter transliterations of the texts they are working on, and algorithms will search for matching sequences and return a list of possible duplicates, arranged according to their similarity.
⁂
The eBL Corpus and its Fragmentarium are scheduled to be publicly released in the Fall of 2021. One day we will not need a Fragmentarium at all: all cuneiform fragments will be identified and joined to their manuscripts; all texts will be complete, and one will be able to read Gilgamesh and the Flood Story from cover to cover. That day is still far away, but we hope that the tools we are creating will bring it much closer to its realization.
Figure 11: The Munich division of the Electronic Babylonian Literature team
(photo by Alexa Bartelmus)