Book history tells us that books are the results of processes, and that what we might think of as the 'final form' of a book might, in fact, not be.
Consider this post, for example. The activity of writing it is, of course, a process: its producer (that's me, though it's unfashionable to say so) writes text, a process which takes time. He can also delete and rewrite text. Nor is the time taken necessarily continuous: the producer can go pause for lunch, go to work, go on holiday. He can even go back, in the case of a blog post, and change things after he's saved it. It's even possible that he might get bored, abandon the work (as Coleridge supposedly did when interrupted during the writing of Kubla Khan), even die, and be replaced as producer by someone else who adds more material, and gives more time to the work. (This is what happened, very notably, in the case of the medieval French poem the Roman de la Rose. It's what didn't happen to Dickens's final novel, The Mystery of Edwin Drood: it is unfinished.) I'm not dead, for the record - but you've only got my word that I'm the same person who began writing this post. Writing is a process; it takes time. We tend to think of the 'final form' as the last possible snapshot in a series, the point after which the text does not change.
But form is not just about language. You might view the 'final form' of this post, once I've hit 'publish', on a computer screen. But your 'final form' is different from someone else's, as it's on a different screen. And what if you print the post out? Then the 'final form' takes on a very different material form, no longer pixels on a screen but ink on paper. Your Penguin paperback copy of Hard Times may contain the same words as the serialization of the novel in Household Words in the 1850s, but its form is not the same. In Household Words it was surrounded by periodical journalism; in your Penguin paperback, by critical notes. In serialized form it was packaged for reading in small chunks; in paperback form, as a whole novel. Such differences in material context might affect your reading of Hard Times, just as watching a box set of the first season of The West Wing might make you react differently from someone who watched it unfold over an entire season for an hour a week. Republication is all part of the process of making and remaking texts.
As well-adjusted literary critics, we're getting used to this, but sometimes it's easy to forget, especially when the form of a work seems more or less fixed. At the extreme, consider the reaction to early modern humanist textual critiques of biblical texts, or more recent critical studies of the Koran. Holy books can sometimes resist being read as processes (or, at least, some of their interpreters can be quite resistant to this idea). Less extremely (and here we get to the point), consider the OED.
It's easy to think of the Dictionary as something immutable - or, even, to think that historically it has been so, even though it is now under revision. This can lead to attempts to treat the OED either as if it is locked permanently in the past or as if it belongs entirely to the present. The latter approach can lead to unfair accusations. One of the most common, which is my example here, is the idea that the OED's editors have been biased towards Shakespeare because they attribute to him so many first uses for words which a little light searching can antedate.
This has been a curiously long-standing accusation. (For one of its most sustained developments, see John Willinsky's book Empire of Words: The Reign of the OED.) As soon as you start to think of the OED as a process and not an object, though, it unravels.
For starters, the OED was not published en bloc. The 'first edition' was published in alphabetical fascicles (rather like serializations, but less racy) between 1884 and 1928. Supplements to OED1 were published in 1933 and in the late 1970s/early 1980s. In 1989 the texts of OED1 and the Supplement volumes were compiled to form OED2. The current revision, OED3, is adding lots of new material and revising and improving upon the old.
It's now 2008, 124 years since the first fascicle of OED1 was published. Thinking about it like that, it's very clear that the OED is a process, and that's why it's faulty thinking to accuse OED1 text of being biased towards Shakespeare because you can find an antedating for a word on the internet. The internet (and here's a shocker) did not exist at any point between 1884 and 1928. Indeed, research resources for the history of English were pretty basic compared to today, especially for the medieval and early modern periods. The OED's lexicographers used what material was available to them at the time. Shakespeare, unlike many since-rediscovered Elizabethan authors, was well represented in critical editions, concordances, etc. Put simply, if you were looking for evidence of a word in use in early modern England, and both Shakespeare and an obscure pamphleteer had used it, you might find the Shakespeare usage and not the pamphleteer because you had a Shakespeare concordance or a text of Shakespeare to hand, and the pamphlet was in some remote library somewhere, unknown. Of course, the editors and their assistants did read lots of obscure early modern texts, but they couldn't catch as much as a full-text search through Early English Books Online can today. Seen today, this can look like a pro-Shakespeare bias, but it's a selection bias, not a genuine editorial one. To try to apply the 'bias' line to first usages, in particular, is ridiculous: the OED editors always looked for the earliest uses of words they could find, and they always included them. They did not hide earlier quotations to make Shakespeare look good; they simply did not know the earlier quotations existed. That would not be pardonable today, but they are not writing it today.
The OED is being written today, though. The difference between what could be done then and what can be done now is one of the motivators for the ongoing revision: the process of making and remaking the OED needs to be resumed. New editors have picked up where old ones left off, and they change things. The OED's form now, consulted online, is no more its final form than the first fascicle was in 1884 or the completed first edition was in 1928. It keeps going, like the language it records.
# Alex Steer (01/07/2008)