Note: this post started as a Foo Camp 2013 session.
A few weeks ago, Tim O’Reilly sent around a link to Who Edited Shakespeare?, which discussed the editor for the First Folio edition of Shakespeare’s plays. It included a lot of evidence that someone had done a lot of work regularizing spelling and doing other tasks that we’d now assign to a copyeditor or a proofreader, presumably more work than the Folio’s nominal editors, Heminges and Condell, were inclined to do or capable of doing.
It’s an interesting argument that prompted some thoughts about the nature of publishing. The process of editing creates the impression, the mythology, that a carefully crafted, consistent, and stable text exists for these plays, that the plays are static literary objects. We like to think that there is a “good” Shakespeare text, if only we had it: what Shakespeare actually wrote, and what was actually performed on stage. We have a mess of good quarto editions, bad quartos, the First Folio, apocryphal works, and more. Some versions of the plays are significantly longer than others; some scholars believe that we’re missing significant parts of Macbeth (Shakespeare’s shortest tragedy, for which the First Folio is the only source). Perhaps the worst case is Christopher Marlowe’s Doctor Faustus, which is known entirely through two early print editions, one roughly 50% longer than the other.
I’m skeptical about whether the search for a hypothetical authoritative version of Shakespeare’s text is meaningful. Shakespeare’s plays were, first and foremost, plays: they were performances staged before a live audience. If you’ve had any involvement with theater, you can imagine how that goes: “Act III, Scene iv dragged; let’s cut it next time. Act V, Scene i was great, but too short; let’s fill it out some.” The plays, as staged events, were infinitely flexible. In the years after Shakespeare, poor editors have certainly done a lot to mangle them, but I’m sure that Shakespeare himself, as a theater professional and partner in a theater company, was constantly messing around with the text.
The published versions of the plays were assembled without any involvement from Shakespeare — indeed, the First Folio was published seven years after his death. Scripts were the trade secrets of the theater companies, and no playwright before Ben Jonson wanted anything to do with publishers or publication. Most likely, the publishers paid off a bunch of actors to sit around a table and remember their lines. And it’s generally assumed that the bad quartos had to do with actors with poor memories, or who perhaps were paid in too much beer. It’s sometimes claimed that very good quartos might have had actual scripts behind them (as far as I know, without any hard evidence, since no original scripts exist).
It’s more likely that the difference between good and bad early editions has less to do with the actors or their inebriation, and more to do with which performances they took part in. Plays aren’t static, fixed literary objects; they change, according to the demands of the situation. And the differences between Quarto and Folio editions are just as likely to have to do with different performances than with the quality of the actors’ memories. The notion that there’s a canonical text for any Shakespeare play is a myth imposed by the institution of publishing (in which Shakespeare didn’t take part), and the mechanics of printing. A printing press, whether a 17th century press or a modern offset press, is a device that can’t react to change. It’s a device that needs a stable, settled text. We create that text, and the myth behind that text, as part of the publishing enterprise.
Now fast forward roughly 400 years to the present. The publishing industry is undergoing rapid change, and players who don’t adapt to the change won’t survive. One important (and undervalued) aspect of this change is that we no longer need the publishing myth of the static text. At O’Reilly, we publish print books and ebooks. For some time, we’ve been regularly updating our ebooks in the field: a day after a book has been “published,” the author can come back and say, “Hey, I need to improve that description in Chapter 3,” add a couple of pages, and push that out to the customers. The text is no longer static and canonical: it’s living and changing. We still have print customers, but increasingly, our printing is done via print-on-demand (POD). When the author updates his book, we automatically send new files to the POD printer, and customers who order a print book tomorrow will receive tomorrow’s book.
We’re in the process of re-inventing publishing around active, dynamic texts, not static, one-way texts that impart wisdom from an author to a reader. We believe that re-invention is crucial to our survival: our own survival as well as our competitors’. But let’s not stop here. What else can we do?
While we believe that Shakespeare wrote most of his great plays in their entirety, it’s well known that the process of writing in Shakespeare’s time was collaborative. A master author would sketch out the plot, write the key scenes, and let apprentices or other members of the theater company staff fill in the rest. There’s certainly stylistic evidence that Shakespeare’s early plays, in addition to several plays of doubtful authorship (the “Shakespeare Apocrypha”), were products of collaboration. Collaboration is equally critical to modern technical authorship. But it’s difficult with our older toolsets: emailing Word files back and forth is painful, and while SVN is pretty good at merging changes on line-oriented source code, prose isn’t really line-oriented. In our new approach to publishing, we’ve tackled the collaboration problem; we believe that we have the first publishing toolchain in which collaboration is native, in which it’s the default, rather than the exception.
I’ve argued for a few years now that the book isn’t really what’s important, it’s the conversation about the book. That applies as much to Shakespeare or to Faulkner as it does to technical books: if there were no conversation around King Lear or The Sound and the Fury, would they be important? So: can we build that discussion into the book? Can the book be something more than one-way communication (with updates)? For a technical book, that discussion would contain questions, answers, and clarifications. It might contain changes that the author makes in response to this discussion. (For a literary work, I grant that the author needs to butt out after he’s done his job.) Do we need the distinction between a draft or “early access” book and the first edition, or is a book something that evolves from the first few chapters, and is never finished? What happens when the notion of collaboration is extended from the authors and co-authors to the readers? Can we write and publish with an “open process,” similar to the way we develop open source software?
This reinvention raises many questions for the publishing enterprise: it requires new kinds of ebooks, new models for compensation, new models for intellectual property. These problems are all solvable (and indeed, we’re working on solving them). The big question is: what can we do now that we’re no longer tied to the myth of stable literary objects? What might the future hold, and how can we build it?