Container-less Content? Not in This Digital Age.
Slips, fascicles, and finally august tomes are all alternative frames for the dictionary. A project as protracted and sprawling as the OED required framing in different ways for different times with different purposes. Those volumes published in 1928 were certainly designed to reflect the achievement of Murray, his team, and their successors. However, even then different editions were on offer -- you could get half volumes with different bindings, as well as the "main" edition. But, of course, a living language is impossible to frame. English is fluid, rendering the dictionary immediately outdated. The First Supplement to the OED appeared only five years after publication. What had once looked like a complete container suddenly had appendages.
By the 1980s, it was clear a new breakthrough in managing lexicographic data was needed. With help from IBM, the Second Edition OED was revolutionized. Torn apart, rekeyed, converted into code, it was eventually printed and published to much fanfare in 1989. Again the scale is vast -- 20 volumes, 615,100 entries, those definitions illustrated with some 2,436,600 quotations over 21,370 pages, using 59,000,000 words and no less than 63 kilograms of paper.
Enabled by the digitization of the OED's production process, this was an order of magnitude bigger than before. This digitization may have liberated it from print strictures, but simply provided a new (more flexible) frame for the book, which in 1989 was reframed back into print. Since 2000, however, the OED has been available online, updated every three months.
The OED today is a database, not a book or set of books, or series of sheaves, or even mass of slips. That it is a database means it has simply been reframed by screens, code, database software, information management tools, server stacks, monitors -- but not that it's unleashed from framing altogether. In its monumental journey the OED has a constantly evolving set of distribution mechanisms, best understood as overlapping methods of framing the content. Each one articulates the OED in a subtly different way, emphasizing a certain performative aspect of framing. Each acts as a "container" for the dictionary, although the language of the container hardly helps us see the Scriptorium and the database on a continuum with each other and those handsome volumes gathering dust on library shelves.