This is the text of a talk given at the conference of the Society for Textual Scholarship on April 6, 1995, as part of a panel devoted to the edition of Thomas Middleton's works forthcoming from Oxford University Press.
In one sense the question in my title doesn't really need an answer. A great many people write electronically today, and practically everyone edits electronically. I do know of one editor whose work rarely involves anything beyond the technology of the 1940s--the typewriter, the microfilm reader, and the telephone; but I don't imagine there are many like him, and even his texts, when they go to his publishers, are immediately turned into electronic texts.
But those publishers, of course, go on to work over those texts extensively to create a different kind of end product: books printed on pieces of paper.
That's a way of writing and editing electronically that tries to create a publication of a sort that would have been comprehensible a hundred years ago. I don't want anyone to think I'm about to shout ``The book is dead!'' or say snide things about people who desire to create new books like this, because the book is still one of the most resonant symbols in our culture. Perhaps there will be scholars of twenty-first-century literature who won't need to pay much mind to books; but scholars of Renaissance and twentieth-century literature, like me, will always need to spend a great deal of time with books, and will always treasure them. Czeslaw Milosz, in a recent poem, imagined going into a library and finding the books unreadable, finding that the ``letters faded and disappeared from the pages''--a woeful experience. ``So a message saving the world is silenced forever?'' he laments. The powerful association of books with communication through time, with the preservation of human voices, is not something we're likely to slough off casually, and I at least wouldn't want to do that. On the merely practical level, electronic ways of preserving messages can do pretty much all of the things that books did; but that's not the most important thing. What is most important is the fact that we used books for this purpose for so long, and that to all of us they've conveyed messages from across the ages.
And the book will survive on its technological merits, too. Writing on electronic texts has an inclination to depict the book as a beautifully simple device, not a lot more complicated than the wheel; but the technology of reading today goes far beyond the simple idea of recording words on paper. As we've learned from research into the history of the book, we can't understand reading without thinking about the entire system that we've built up to support it. A text doesn't have its meaning as an isolated entity; it is what we can make of it, and that depends a great deal on the whole world of information in which its readers live. A reader of Middleton today can draw on an extensive array of bibliographies, catalogues, libraries, dictionaries, and histories. Our edition will include an essay by John Jowett on Middleton's seventeenth-century readers, and it will make it dramatically clear how different Middleton looked to such readers--indeed, how many of them had no way of even associating Middleton's name with the books at hand. Our reading of Middleton today is conditioned above all by the OED and the STC, two works that have been far more influential than any twentieth-century school of criticism. But the conventions of academic publishing discourage an acknowledgment of the relative importance of such reference works as compared to criticism.
An edition that is not just prepared electronically, but published in electronic form, makes its principal contribution by adding to the body of reference information. It is a new kind of book, but one that continues the work that books have always done: it creates new meaning, even if the words it contains are old; it transmits a message, and it adds something to the message in the process. The way an electronic edition extends this system is by adding a new kind of reader--because it has not one but two kinds of readers: people and machines. Machines read very fast but not well. They don't understand the plots, they don't understand irony or figurative speech, they don't understand anything much beyond the literal. But they're very good at the literal; they can't tell when a text is funny, but they can quickly pick out places where the word ``funny'' occurs. They're good at doing some things that we like to see done but hate to do: counting and finding. A text is different when these things can happen without great effort, just as it is different when it is included in a catalogue; what is possible to know about it changes. That a canonical writer is surrounded by bibliographies, editions, guides, and concordances is not merely a manifestation of literary authority; it is also a body of supplementary knowledge that changes what that writer's works mean.
To my mind there has been one great electronic edition created so far (I can say this, of course, because ours isn't done yet): the Perseus hypertext, published by Yale University Press, which isn't even described by its creators as a scholarly edition and which didn't involve much real textual work, just a collection of already-edited classical texts and other information, for the most part. But in putting these together with a wealth of other verbal and pictorial information the texts are changed immensely. We have a newly expanded ability to see where they stood in relation to the entire culture.
Many people nowadays who talk about the electronic edition see it in a very different way: not as a book, but as a photograph. That is, on this position, the great merit of an electronic edition is that it gives you a copy of something that might come within your field of view--in this case, a copy of all the pieces of paper that happen to record some form of the text's words. The copy is the same as the original, at least as far as looking goes, and looking is all that we imagine doing with these pictures, in this kind of edition.
Of course, this conference is one place where you can't get away with saying that a photographic facsimile is a true copy. But the biggest problem with facsimiles lies not in their very real deceptiveness, which after all does not negate their usefulness. It lies rather in their equally real faithfulness. They don't let us add anything new to the text, or rather to the picture. A book, an edition, can say something new; a picture can't. This, I think, is why the torrent of manuscript facsimiles of nineteenth- and twentieth-century English and American authors which rolled into research libraries in the last twenty years has somehow left everything else undisturbed and just where it was to begin with. It is certainly a good thing that we have these facsimiles to look at, but the works that changed the way we saw the texts of a few writers, such as Dickinson, were not facsimiles, but books and articles that presented an argument and not just photographs. Putting photographs into electronic form doesn't change this situation much, in large part because computers can't do very much with pictures. Machines are rather poor readers, as I said; but they are nearly useless at looking. They can't make connections between shapes in a picture and anything they might refer to, if anything beyond the simplest geometric figures is involved. In some cases they can display pictures more cheaply than a book can; but, as I've said, just making pictures available hasn't changed things so far.
I mentioned the problem of editions prepared electronically but with an eye only to the generation of printed books. That problem is rooted in the creation of electronic texts that reflect the typographic structure of a printed document--rather than the literary structure of the work, which is the approach we are taking. This approach is sometimes criticized because it is interpretive: there is an odd but persistent belief that an electronic text should be somehow ``objective.'' But our approach is of a piece with our editorial preferences. We don't see the avoidance of interpretation as a merit. We want to tell people about our understanding of Middleton, and one good way to do that is to edit his text. Moreover, an electronic text which describes the work's literary structure, rather than the way we want to see it printed, is also easier for the computer to read. A computer can't recognize a line of verse on its own; if you want a computer's assistance in any analysis that needs to take account of lines of verse, or stanzas, or speeches, or stage directions, you need to identify those features for the machine.
Why edit electronically, then? Because that's how you can record what you think about the work, and not just copy it, an operation which may assume that attitude but doesn't articulate it. The electronic text doesn't have to be a silent and transparent mediator; if we think about it we know that it isn't anyway. But it can say more than the printed book does.
Our edition is based on arguments about what we think the text is that will certainly not evade all dispute--in its choice of works to attribute to Middleton and in the critical claims made about the works, just as much as in the way they're edited. An electronic edition, for us, has two purposes which work together--to enable new ways of understanding the text, and to articulate what the editors think.