The Chronicle Review

The Ghost (Writer) in the Machine


Matthew G. Kirschenbaum
June 26, 2016

Charles Dickens wrote with a feather quill. T.S. Eliot used a fountain pen engraved with his initials. Jack Kerouac typed On the Road on a portable Underwood.

The technology of writing — how words move from an author’s mind to the page — has improved drastically over the years. Perhaps the biggest advance came toward the end of the 20th century with the rise of the personal computer, allowing authors to easily revise without having to rewrite.

Goodbye, Wite-Out. Hello, delete key.

Helpful as the computer seemed, plenty of writers viewed this magical invention with skepticism (and some holdouts, fond of old-fashioned implements, still do). They worried that it would change writing for the worse, or that the glowing box might replace the writer entirely.

Matthew G. Kirschenbaum explores those anxieties in his new book, Track Changes: A Literary History of Word Processing (Harvard University Press), and attempts to identify the first novel written on a word processor. Along the way, Kirschenbaum, an associate professor of English at the University of Maryland at College Park, digs up anecdotes from the recent past that — now that computers are so thoroughly embedded in our lives — feel like ancient history.

You write that the appearance of the word processor around 1981 was "an event of the highest significance in the history of writing." Why was it such a big deal?

If you just start with the actual technology itself, the arrival of this metal, plastic, glass, and silicon box on the desk, intruding on an author’s inner sanctum, it was a very dramatic event. The hardware made noise. It threw off heat. You could hear the disks spinning. And then seeing those glowing letters take shape and marching across the screen. Peter Straub said that writing on a word processor felt like the future. And I think that’s typical of the response that a lot of writers had.

In the book you note that George R.R. Martin, author of the epic that became HBO’s Game of Thrones, still uses a program called WordStar from the 1980s, which is by today’s standards remarkably primitive. But obviously there’s a relationship there that’s been successful for him, and he remains attached to it.

I think in George Martin’s case, and it holds true for other authors who have stuck with particular software, particular platforms, it’s not just about superstition or thinking about a particular computer as their lucky charm or their totem. A program like WordStar is very different from what we’re used to today with Microsoft Word and pointing and clicking on the graphical interface. But it’s an extraordinarily powerful program that’s well designed for the person who’s willing to accept a steep learning curve, something I think we’re no longer willing to do with software.

There was a concern among writers that word processors might make their work less authentic or more automatic, and some may have tried to conceal the fact that they were using a computer.

From the late 1970s to the mid-1980s, those were real anxieties. John Barth comments on the way that the M.F.A. program at Johns Hopkins started receiving its first writing samples that were self-evidently prepared on a word processor and the kind of prejudice and bias that that created. There was a reluctance to accept those candidates because there was a notion that they would believe that they were already real writers because their writing looked so good, it looked so polished on the page. Some critics were actually under the impression that word processors were somehow writing the books.

There was a concern, too, that computers would lead to longer sentences or longer books and that economy of expression would be sacrificed. Did you find any evidence that that’s the case?

I really didn’t. All of those truisms that we associate with word processing and writing on a computer, not only do they turn out to be not true, but they turn out to be mutually contradictory. So on the one hand, the notion that word processing makes one more verbose and the books get thicker — that’s one common set of biases. Yet we also associate computers with a whole other set of styles where sentences become shorter.

Another fear was that word processing would be a disaster for historians and archivists. The idea being that changes would be deleted, that the messy process of writing and revision would become too clean, that the scholars of the future would have nothing to work with. Has that come to pass?

That hasn’t come to pass, but I would also say it’s too early for us to have a full sense of what the future of literary history is going to look like. You certainly have the potential to lose a lot when a writer is working on a computer, particularly if they’re working in the same file and saving it over and over again. But we also see writers who are "versioning" their manuscript, using the same sort of tools that software developers use to manage code.

You set out to discover the identity of the first novel written on a computer. It seems like that’s something we won’t ever know for sure. I’m curious, though, which book you think is the most likely candidate.

If we think about a word processor in the way that we conventionally envision, which is a typewriter attached to a TV set, probably the first person who does that in a way that we would recognize is a science-fiction writer working in the late 1970s named Jerry Pournelle. But there are earlier candidates as well: John Hersey, the novelist and journalist, used a mainframe computer at Yale in the early 1970s — not to compose a novel, but to revise one extensively.

The earliest and most fascinating candidate that I came across is a British novelist named Len Deighton, who in the 1960s made a name for himself writing espionage thrillers. Because he was very commercially successful, he was able to afford the first product IBM put on the market and called a word processor. It was called the Magnetic Tape/Selectric Typewriter. It sold for $10,000. Deighton was the only writer I knew of who had one in his possession.

Any thoughts about where writing technology is headed now?

We’re thinking self-consciously about digital writing technologies in a way that we haven’t for a couple of decades because Microsoft Word is no longer the only choice people have. There’s a lot more experimentation. The influence of new kinds of platforms like mobile phones and tablets. I do a lot of my writing by talking to my phone. And I think about how we’re becoming accustomed to having the computer as our collaborator. This manifested first in terms of spell check and grammar check, and now we have autocomplete and news articles that are constructed algorithmically.

And so, I think, we’re gradually moving toward a future where more and more writing is collaborative, where you have a human author but that person’s work is being informed keystroke by keystroke.

Tom Bartlett is a senior writer who covers science and other things. Follow him on Twitter @tebartl. This interview has been edited and condensed.