In the course of our work on the GKAB project, it became clear that the meanings and functions of 'libraries' in cuneiform culture were not as clear cut as we had originally assumed.
First, there is the fundamental question of what we mean by the word 'library': a building, a room, a collection of books?
The archaeological evidence is clear that no whole buildings were dedicated to literate activities in first-millennium Assyria or Babylonia, or at least none have been found so far. But there were certainly rooms where tablets were stored. At Kalhu, the temple of Nabu had a dedicated tablet room, immediately opposite the god's shrine. In the āšipus' house(s) and in the Reš temple at Uruk, tablets were kept in particular rooms. But all of these rooms were almost certainly too dark to read or write in, so the tablets must have been removed from them in order to be read. And in the case of Huzirina we simply do not know where Qurdi-Nergal's family and their associates originally used their tablets, as they buried them outside the house for protection from invaders.
We find the Akkadian word gerginakku, "library" only on a few tablets from the Neo-Assyrian period (both Hurizina and Kalhu). It appears exclusively in colophons, in the phrase gerginakka lā taparrar, "Do not disperse the library!", suggesting that it refers to a collection, rather than a building. But it also highlights the fact that tablets could and did circulate, despite their owners' attempts to keep them stationary. So the assemblages that archaeologists find are only snapshots of what a tablet collection looked like on the day that it was abandoned, or that the building it was in got destroyed. Colophons from all collections also mention lēʾu, "(wooden) writing boards", reminding us that tablets were not the only medium of cuneiform scholarship.
We might also worry about how to define the counterpart of a "book" in cuneiform culture. Intuitively, archaeologists and Assryiologists have always distinguished "archives"—containing everyday writings such as letters, legal documents, business records—from the scholarly writings which constitute "libraries", even when they were kept (and found) in the same storage spaces. By contrast, the ephemeral by-products of elementary scribal education—word lists, spelling exercises and the like—tended not to be kept with more learned writings, if they were kept at all. A close examination of the findspots of tablets in the āšipus' house(s) has shown this particulalry clearly.
Cuneiform "books", then, roughly equate to learned or scholarly writings. They may (parts of) be standard works copied and re-copied over many generations, such as sections of the omen series Enūma Anu Ellil or chapters of The Epic of Gilgamesh. Equally, they include new compositions, whether commentaries on the standards or entirely new experiments. No library had any of the larger, multi-tablet works in their entirety though, which raises interesting questions about the accessibility of scholarship.
How did learned knowledge spread, and what were the barriers to its movement? People and tablets both moved long-distance, as well as within urban professional communities. Equally, social and political constraints inhibited the flow of scholarly exchange just as much as the scholars' own prohibitions did.
Eleanor Robson
Eleanor Robson, 'Rethinking the cuneiform library', The Geography of Knowledge, The GKAB Project, 2019 [http://oracc.museum.upenn.edu/cams/gkab/contexts/rethinkingthecuneiformlibrary/]