This review was commissioned by John Wilson and meant for Books and Culture. Alas, it will not be published there.

“Each of us remembers our own first time,” Matthew Kirschenbaum writes near the beginning of his literary history of word processing — but he rightly adds, “at least … those of us of a certain age.” If, like me, you grew up writing things by hand and then at some point acquired a typewriter, then yes, your first writing on a computer may well have felt like a pretty big deal.

The heart of the matter was mistakes. When typing on a typewriter, you made mistakes, and then had to decide what, if anything, to do about them; and woe be unto you if you didn’t notice a mistyped word until after you had removed the sheet of paper from the machine. If you caught it immediately after typing, or even a few lines later, then you could roll the platen back to the proper spot and use correcting material — Wite-Out and Liquid Paper were the two dominant brands, though fancy typewriters had their own built-in correction tape — to cover the offending marks and replace them with the right ones. But if you had already removed the paper, then you had to re-insert it and try, by making minute adjustments with the roller or the paper itself, to get everything set just so — but perfect success was rare. You’d often end up with the new letters or words slightly out of alignment with the rest of the page. Sometimes the results would look so bad that you’d rip the paper out of the machine in frustration and type the whole page again, but by that time you’d be tired and more likely to make further mistakes….

Moreover, if you were writing under any kind of time pressure — and I primarily used a typewriter to compose my research papers in college and graduate school, so time pressure was the norm — you were faced with a different sort of problem. Scanning a page for correctable mistakes, you were also likely to notice that you had phrased a point awkwardly, or left out an important piece of information. What to do? Fix it, or let it be? Often the answer depended on where in the paper the deficiencies appeared, because if they were to be found on, say, the second page of the paper, then any additions would force the retyping of that page but of every subsequent page — something not even to be contemplated when you were doing your final bleary-eyed 2 AM inspection of a paper that had to be turned in when you walked into your 9 AM class. You’d look at your lamentably imprecise or incomplete or just plain fuddled work and think, Ah, forget it. Good enough for government work — and fall into bed and turn out the light.

The advent of “word processing” — what an odd phrase — electronic writing, writing on a computer, whatever you call it, meant a sudden and complete end to these endless deliberations and tests of your fine motor skills. You could change anything! anywhere! right up to the point of printing the thing out — and if you had the financial wherewithal or institutional permissions that allowed you to ignore the cost of paper and ink, you could even print out a document, edit it, and then print it out again. A brave new world indeed. Thus, as the novelist Anne Rice once commented, when you’re using a word processor “There’s really no excuse for not writing the perfect book.”

But there’s the rub, isn’t there? For some few writers the advent of word processing was a pure blessing: Stanley Elkin, for instance, whose multiple sclerosis made it impossible for him to hold a pen properly or press a typewriter’s keys with sufficient force, said that the arrival of his first word-processing machine was “the most important day of my literary life.” But for most professional writers — and let’s remember that Track Changes is a literary history of word processing, not meant to cover the full range of its cultural significance — the blessing was mixed. As Rice says, now that endless revision is available to you, as a writer you have no excuse for failing to produce “the perfect book” — or rather, no excuse save the limitations of your own talent.

As a result, the many writers’ comments on word processors that Kirschenbaum cites here tend to be curiously ambivalent: it’s often difficult to tell whether they’re praising or damning the machines. So the poet Louis Simpson says that writing on a word processor “tells you your writing is not final,” which sounds like a good thing, but then he continues: “It enables you to think you are writing when you are not, when you are only making notes or the outline of a poem you may write at a later time.” Which sounds … not so good? It’s hard to tell, though if you look at Simpson’s whole essay, which appeared in the New York Times Book Review in 1988, you’ll see that he meant to warn writers against using those dangerous machines. (Simpson’s article received a quick and sharp rebuttal from William F. Buckley, Jr., an early user of and advocate for word processors.)

Similarly, the philosopher Jacques Derrida, whom Kirschenbaum quotes on the same page:

Previously, after a certain number of versions, everything came to a halt — that was enough. Not that you thought the text was perfect, but after a certain period of metamorphosis the process was interrupted. With the computer, everything is rapid and so easy; you get to thinking you can go on revising forever.

Yes, “you get to thinking” that — but it’s not true, is it? At a certain point revision is arrested by publishers’ deadlines or by the ultimate deadline, death itself. The prospect of indefinite revision is illusory.

But however ambivalent writers might be about the powers of the word processor, they are almost unanimous in insisting that they take full advantage of those powers. As Hannah Sullivan writes in her book The Work of Revision, which I reviewed in these pages, John Milton, centuries ago, claimed that his “celestial patroness … inspires easy my unpremeditated verse,” but writers today will tell you how much they revise until you’re sick of hearing about it. This habit predates the invention of the word processor, but has since become universal. Writers today do not aspire, as Italian Renaissance courtiers did, to the virtue called sprezzatura: a cultivated nonchalance, doing the enormously difficult as though it were easy as pie. Just the opposite: they want us to understand that their technological equipment does not make their work easier but far, far harder. And in many ways it does.

Matthew Kirschenbaum worked on Track Changes for quite some time: pieces of the book started appearing in print, or in public pixels, at least five years ago. Some of the key stories in the book have therefore been circulating in public, and the most widely-discussed of them have focused on a single question: What was the first book to be written on a word processor? This turns out to be a very difficult question to answer, not least because of the ambiguities inherent in the terms “written” and “word processor.” For instance, when John Hersey was working on his novel My Petition for More Space, he wrote a complete draft by hand and then edited it on a mainframe computer at Yale University (where he then taught). Unless I have missed something, Kirschenbaum does not say how the handwritten text got into digital form, but I assume someone entered the data for Hersey, who wanted to do things this way largely because he was interested in his book’s typesetting and the program called the Yale Editor or just E gave him some control over that process. So in a strict sense Hersey did not write the book on the machine; nor was the machine a “word processor” as such.

But in any case, Hersey, who used the Yale Editor in 1973, wouldn’t have beaten Kirschenbaum’s candidate for First Word-Processed Literary Book: Len Deighton’s Bomber, a World War II thriller published in 1970. Deighton, an English novelist who had already published several very successful thrillers, most famously The IPCRESS File in 1962, had the wherewithal to drop $10,000 — well over $50,000 in today’s money — on IBM’s Frankensteinian hybrid of a Selectric typewriter and a tape-based computing machine, the MT/ST. IBM had designed this machine for heavy office use, never imagining that any individual writer would purchase one, so minimizing the size hadn’t been a focus of the design: as a result, Deighton could only get the thing into his flat by having a window removed, which allowed it to be swung into his study by a crane.

Moreover, Deighton rarely typed on the machine himself: that task was left to his secretary, Ellenor Handley, who also took care to print sections of the book told from different points of view on appropriately color-coded paper. (This enabled Deighton to see almost at a glance whether some perspectives were over-represented in his story.) So even if Bomber is indeed the first word-processed book, the unique circumstances of its composition set it well apart from what we now think of as the digital writing life. Therefore, Kirschenbaum also wonders “who was the first author to sit down in front of a digital computer’s keyboard and compose a published work of fiction or poetry directly on the screen.”

Quite possibly it was Jerry Pournelle, or maybe it was David Gerrold or even Michael Crichton or Richard Condon; or someone else entirely whom I have overlooked. It probably happened in the year 1977 or 1978 at the latest, and it was almost certainly a popular (as opposed to highbrow) author.

After he completed Track Changes, Kirschenbaum learned that Gay Courter’s 1981 bestselling novel The Midwife was written completely on an IBM System 6 word processor that she bought when it first appeared on the market in 1977 — thus confirming his suspicion that mass-market authors were quicker to embrace this technology than self-consciously “literary” ones, and reminding us of what he says repeatedly in the book: that his account is a kind of first report from a field that we’ll continue to learn more about.

In any case, the who-was-first questions are not as interesting or as valuable as Kirschenbaum’s meticulous record of how various writers — Anne Rice, Stephen King, John Updike, David Foster Wallace — made, or did not quite make, the transition from handwritten or typewritten drafts to a full reliance on the personal computer as the site for literary writing. Wallace, for instance, always wrote in longhand and transcribed his drafts to the computer at some relatively late stage in the process. Also, when he had significantly altered a passage, he deleted earlier versions from his hard drive so he would not be tempted to revert to them.

The encounters of writers with their machines are enormously various and fun to read about. Kirschenbaum quotes a funny passage in which Jonathan Franzen described how his first word processor kept making distracting sounds that he could only silence by wedging a pencil in the guts of the machine. Franzen elsewhere describes using a laptop with no wireless access whose Ethernet port he glued shut so he could not get online — a problem not intrinsic to electronic writing but rather to internet-capable machines, and one that George R. R. Martin solves by writing on a computer that can’t connect to the internet, using the venerable word-processing program WordStar. Similarly, my friend Edward Mendelson continues to insist that WordPerfect for MS-DOS is the best word-processing program, and John McPhee writes using a computer program that a computer-scientist friend coded for him back in 1984. (I don’t use a word-processing program at all, but rather a programmer’s text editor.) If it ain’t broke, don’t fix it. And if it is broke, wedge a pencil in it.

Kirchenbaum believes that this transition to digital writing is “an event of the highest significance in the history of writing.” And yet he confesses, near the end of his book, that he’s not sure what that significance is. “Every impulse that I had to generalize about word processing — that it made books longer, that it made sentences shorter, that it made sentences longer, that it made authors more prolific — was seemingly countered by some equally compelling exemplar suggesting otherwise.” Some reviewers of Track Changes have wondered whether Kirschenbaum isn’t making too big a deal of the whole phenomenon. In the Guardian of London, Brian Dillon wrote, “This review is being drafted with a German fountain pen of 1960s design – but does it matter? Give me this A4 pad, my MacBook Air or a sharp stick and a stretch of wet sand, and I will give you a thousand words a day, no more and likely no different. Writing, it turns out, happens in the head after all.”

Maybe. But we can’t be sure, because we can’t rewind history and make Dillon write the review on his laptop, and then rewind it again, take him to the beach, and hand him a stick. I wrote this review on my laptop, but I sometimes write by speaking, using the Mac OS’s built-in dictation software, and I draft all of my books and long essays by hand, using a Pilot fountain pen and a Leuchtturm notebook. I cannot be certain, but I feel that each environment changes my writing, though probably in relatively subtle ways. For instance, I’m convinced that when I dictate my sentences are longer and employ more commas; and I think my word choice is more precise and less predictable when I am writing by hand, which is why I try to use that older technology whenever I have time. (Because writing by hand is slower, I have time to reconsider word choices before I get them on the page. But then I not only write more slowly, I have to transcribe the text later. If only Books and Culture and my book publishers would accept handwritten work!)

We typically think of the invention of printing as a massive consequential event, but Thomas Hobbes says in Leviathan (1650) that in comparison with the invention of literacy itself printing is perhaps “ingenious” but fundamentally “no great matter.” Which I suppose is true. This reminds us that assessing the importance of any technological change requires comparative judgment. The transition to word processing seemed like a very big deal at the time, because, as Hannah Sullivan puts it, it lowered the cost of revision to nearly zero. No longer did we have to go through the agonies I describe at the outset of this review. But I am now inclined to think that it was not nearly as important as the transition from stand-alone PCs to internet-enabled devices. The machine that holds a writer’s favored word-processing or text-editing application will now, barring interventions along the lines of Jonathan Franzen’s disabled Ethernet port, be connected to the endless stream of opinionating, bloviating, and hate-mongering that flows from our social-media services. And that seems to me an even more consequential change for the writer, or would-be writer, than the digitizing of writing was. Which is why I, as soon as I’ve emailed this review to John Wilson, will be closing this laptop and picking up my notebook and pen.


  1. What of those who otherwise would never have written at all? I only began to write after the advent of word processing software, because I could proof and edit easily! Lousy handwriting, no typing skills.

    I know at least one writer who published a book only because she was able to use voice recognition software to get her words down.

  2. Hi,
    I wrote "Understanding Computers"in 1984 for OUP on a microprocessor system with 64k of RAM and two 8" floppy disks, using the Wordstar package. Wordstar had just introduced an indexing program called Starindex, and much to my amazement, this was good enough to index the whole book.

  3. From William F. Buckley, Jr.'s "literary autobiography" Miles Gone By (2004): "[The digital version of the American Heritage Dictionary (AHD)] is the greatest clerical blessing of my lifetime, excepting only of course, Word Processing itself, the King of Kings…. Not long after I acquired the AHD, I was in a radio exchange with the senior U.S. liberal, Professor Arthur Schlesinger Jr., who in a casual survey of technology stunned me by saying that, in his judgment, 'word processing is the greatest invention in modern history.' Suddenly I was face to face with the flip side of Paradise. That means, doesn't it, that Professor Schlesinger will write more than he would do otherwise?"

Comments are closed.