Revision: uncommon before the 20th century? I’m not so sure
Craig Fehrman reviewed in the Boston Globe Hannah Sullivan’s The Work of Revision. I have not read Sullivan’s book, so I do not know which ideas in Fehrman’s review are his own and which are Sullivan. However I do not understand the argument.
It’s easy to assume that history’s greatest authors have been history’s greatest revisers. But that wasn’t always how it worked. Until about a century ago, according to various biographers and critics, literature proceeded through handwritten manuscripts that underwent mostly small-scale revisions.
Then something changed. In a new book, The Work of Revision, Hannah Sullivan, an English professor at Oxford University, argues that revision as we now understand it—where authors, before they publish anything, will spend weeks tearing it down and putting it back together again—is a creation of the 20th century. It was only under Modernist luminaries like Ezra Pound, T.S. Eliot, and Virginia Woolf that the practice came to seem truly essential to creating good literature. Those authors, Sullivan writes, were the first who “revised overtly, passionately, and at many points in the lifespan of their texts.”[…]
Ben Jonson, a Renaissance playwright, once observed of Shakespeare that “whatsoever he penned, he never blotted out a line.” Jonson was gently mocking the cult of Shakespeare, and it’s certainly possible that an old chest somewhere contains several radically revised versions of Shakespeare’s plays. But that seems unlikely.[…]
The problem, of course, is that we do have many variants of Shakespeare plays – e.g., the First Folio, and various Quartos for many plays. These are so different that many publishers have taken to publishing different extant versions of the plays – thus, the Norton Shakespeare, a standard college level text and anthology, publishes both the 1608 quarto text and the 1623 Folio King Lear; Arden is currently publishing Hamlet in three versions (1603, 1604, and 1623). And most Shakespeare anthologies I am aware of attempt to integrate various texts.
Now, if these do not reflect “several radically revised versions of Shakespeare’s plays,” what do they reflect?
(Fehrman-Sullivan’s theory seems to be that paper used to be expensive making revisions unlikely, and that typing forced writers to slow down so that they were able to revise. The theory as expressed in Fehrman’s review sounds unlikely to me, but again I have not read Sullivan’s book.)
BLT 
Apologies, the first version of this post omitted the link to Fehrman’s review.
I have heard Sullivan talk about the book and just read the piece in the Globe. Her idea, as I understand it, is that the different versions of Shakespeare’s plays aren’t radically different from one another. An ordinary theatregoer wouldn’t be sure afterwards which text of Lear he had heard performed, and the differences between the texts are mostly at the level of the word or sentence, so they can be represented on facing-page editions. The book says that modernist texts that were composed and returned to over many years, even decades, through a much slower writing process change in more structural ways (e.g. Ulysses). But the key point here (in your post) seems to be something different. In the case of Shakespeare, we don’t know whether the different versions reflect revision or corruption. Traditionally two different versions of a play were taken to reflect two relatively corrupted versions of the same lost original. The point is that no evidence survives of Shakespeare’s autograph — and probably isn’t likely to turn up now. (Though that would be very cool and valuable.)
Katy, thanks so much for your very informative comment.
I do think that there are some pretty big structural changes in different variants of King Lear. The Quarto has about 100 lines not in the Folio, and the Folio has about 300 lines not in the Quarto. For example, one of the most powerful scenes in King Lear is the “mock trial” scene, which is missing from the Folio.
One very good study of this is William Shakespeare: A Textual Companion.
It is only natural that in a working theater company that the staff would regularly make changes, both big and small, to plays as they were performed and revived and revived again.
Now, it is true that we lack autograph evidence of changes to Shakespeare (in fact, we have no manuscripts in Shakespeare’s hand at all), but it is only natural that we would have much more handwritten evidence of works published in the last century than before. So, if that is Sullivan’s main evidence, I think she has biased things.
Sullivan does have a point that the special demands of modernist literature must have made for especially slow writing. But the controversy over the various “corrected” versions of Ulysses, for example, show that whatever degree of revision was applied, the author was not scrupulous about making sure that the final printed version was correct.
So, I remain confused by her theory.
I think the process of an author, deliberately and significantly revising a work over a period of time before it sees the light of day, and a (playwright+theater company), revising a script as it evolves as a living, working text (as well as presumably making scribal errors as copies are made for individual performers), are so thoroughly different that they should not be compared.
If the argument rests on the price of paper, then a hundred years is way too short. I believe the move from rag-based to pulp-based paper significantly decreased its cost, and that was more like two hundred years ago. Any period that had an affordable daily newspaper has to have had affordable paper.
The argument about typewriting vs handwriting strikes me as ridiculous. I’m pretty sure I can type faster than I can write, and revising a handwritten manuscript is either just as easy, or easier than, revising typewritten material, depending on whether you’re willing to write on your typescript or insist on typewritten corrections.
Victoria, I take your point, but the Shakespeare example was raised by Fehrman in his review. I still think, for reasons that you also point out, that the Fehrman-Sullivan thesis is a bit absurd.