Track Changes: A Literary History of Word Processing
by Matthew G. Kirschenbaum

Reviewed by Matt Margini

Reading Track Changes: A Literary History of Word Processing, Matthew Kirschenbaum’s deft and detailed foray into the history of word processing, reminded me inevitably of Halt and Catch Fire, AMC’s criminally underrated drama about tech entrepreneurs in the 1980s. More specifically, Kirschenbaum’s book transported me back to a great scene at the end of the first season, when Joe MacMillan—one of the protagonists, a Don Draper wannabe with Steve Jobsian pretentions—wanders through a warren of hotel rooms at a PC convention and stumbles upon something truly, horrifyingly new. Like a statue of the Virgin Mary, the machine is surrounded by candles and pilgrims. At first it looks like any other computer, if maybe a little cuter: a gray box encasing a cathode-ray screen. But then it speaks: “Hello, my name is Macintosh.” And Joe just stands there, dumbstruck, as though he’s seen a vision of his own death.  

The scene is a little corny, but it allegorizes a master plot that courses through the history of technology: the inexorable defeat of jagged freedom at the hands of hegemonic polish. Despite his slim-cut power suits and corporate airs, Joe is a vaquero in personal computing’s Wild West; the machine he and his nimble, hungry team have been building, with blood and borrowed money, is a device for enthusiasts who know how to delve into the innards of the beast. They’ve built a car for people who know their way around the engine. But the Macintosh is something different: a black box in the truest sense, rendering its own inner workings invisible beneath a layer of ersatz friendliness. It is experienced as accessible precisely because it is inaccessible; anyone can use it, but only because it is obdurately, beautifully opaque. The show lingers on the Macintosh’s debut not just because it flummoxes Joe, but because it portends what consumer computer interfaces will become in their current form—in other words, the friendly and all-powerful opacity with which they greet us now. When was the last time you knew what your computer was actually doing, under the hood? When was the last time you stopped to contemplate where and how the electrons move? 

Such questions lurk in the background of Track Changes, a study which, like Halt and Catch Fire, can feel downright elegiac in the way it evokes a world before the polished hegemony of Google Docs and Microsoft Word—a world in which computers were exasperating, brutish, demanding in their quirks, yet also somehow liberating in the way they granted power to those who could tame them. The title bills it as a “history” of word processing, and Kirschenbaum provides valuable insights into where word processing came from both materially and socially; he has a whole chapter, for instance, about how the technology emerged from a gendered history of white collar labor, as a tool for secretaries—the first “word processors”—who were compartmentalized in separate rooms and entrusted with increasingly regimented, computer-like tasks. But Kirschenbaum’s larger goal is to contemplate the effects of word processors on literary production: what happened to writers and their writing—materially, formally, even psychologically—after computers became available as tools of the trade. 

Why does it matter that Stephen King used a Wang System 5 word processor for more than 15 years; that Amy Tan (like Ralph Ellison and Michael Chabon) was an early adopter of the Osborne 1, the first truly portable personal computer; that George R. R. Martin uses, to this day, the ancient software WordStar; or that Jonathan Franzen still prefers the company of a chunky early 2000s Dell laptop? In some sense, these details matter for the same reason we cannot help but fetishize Austen’s writing desk, Dickens’s cat-paw letter-opener, or Darwin’s humble chamber pot: they humanize empyreal geniuses, bringing them back into the world of everyday materiality from which they seem to have departed. But word processors do not interest Kirschenbaum merely because (like these other fetish objects) they were attached to the writing lives of interesting people. They interest him because they were both easy and hard for these interesting people to work with, creating new sparks of frisson between man and machine that inspired writers just as much as they improved, or complicated, their workflow. 

One point of particular interest for Kirschenbaum is the way in which the arduousness of working with the earliest incarnations of the digital word processor made visible the miraculousness—the sheer ontological strangeness—of what they actually do. First of all, Kirschenbaum’s excursions into the early years of word processing show us how, although all forms of writing allow the writer to play God, as it were—granting the author the ability to create worlds out of words—the phenomenological experience of this cosmogonic capability is wholly different to the writer typing away at her computer. Presenting text suspended in a void of black early word processors were especially capable of turning the act of writing into an eerie facsimile of the Word, of God’s creative labor. In word processing, “text blinks on and off, winking in and out of existence with comparative ease.” Haruki Murakami’s 1Q84 thematizes this idea by depicting a writer whose computer brings literal “whole new worlds into being” in the form of parallel universes; similarly, Stephen King’s story “The Word Processor” depicts a King-like author using a word processor—King’s own Wang—to delete his wife and conjure a more desirable family. Virginia Heffernan describes the effect in her own paean to the medium: “You rocket out into the unknown, into profound solitude, and every word of yours becomes the kind of outer-space skywriting that opens ‘Star Wars.’”  

The writers in Kirschenbaum’s study were obsessed with this idea that the word processor brought with it a new kind of freedom—the freedom to create and delete word-worlds with abandon, in a void-like environment without any of the literal, material, or psychological frictions of the written page. They were obsessed with the way it could let them escape not only the materiality of writing but its temporality as well: the layers of history—revisions; words scratched out; a trashcan full of painstaking false starts—that encrust any manuscript, attesting to long periods of difficulty and anxiety. As Daniel Chandler has written, “writing with a word processor . . . obscures its own evolution.” Especially with track changes off—and track changes, Kirschenbaum reminds us, is a fairly recent feature—word processing can produce pieces of writing that appear “perfect,” suspended in the cosmic glass, stripped of the complex histories of painstaking revision and emendation from which they emerge. It can produce pieces of writing that are themselves like “black boxes,” seamless and opaque. (Edward Mendelson once called Microsoft Word “Platonic” because it tries so ardently to forge the ideal document; Kirschenbaum dwells, in a similar vein, on the symbolic resonances of “WordPerfect,” the name of Microsoft Word’s biggest early 90s rival.)

And yet, as Kirschenbaum reminds us, word processing in the early ’80s was a frictional, material technology like any of its predecessors, and perhaps no one was more aware of that than the early pioneers who could only seek transcendence by wrestling with the medium’s brutish beginnings. Indeed, the stories he tells—most of them clustering around the early ’80s, and particularly the year 1981, when word processers first became widely available to consumers—tend to be meaningful precisely because they show how the Godlike labor of creating word-worlds in a void was facilitated by the painstaking real-life labor of tinkering with the machine, memorizing the right commands, listening to the disk, bracing for catastrophe. For the word processing pioneers, the profound freedom of the new medium was yoked dialectically to experiences of difficulty and estrangement. 

Most of the anecdotes Kirschenbaum relates are about writers in the early ’80s becoming tinkerers, either because what they wanted to do—poetically, typographically, collaboratively—stretched the limits of the extant hardware and software, or simply because “the degree of hacking, improvisation, and problem solving the early systems required meant that the borderlines between innovation and convention, novelty and standard operating procedure, were frequently blurred in practice.” The anecdotes expand into lavish detail when the interface—which is to say, not just the software interface but the human-machine relationship—becomes a site of friction. Kirschenbaum spends multiple pages, for instance, reconstructing a “typical session” (circa 1973) between John Hersey, the author of Hiroshima, and an early word processing program that ran on a half-million-dollar PDP-10 mainframe at Yale University:

The second line of the book begins: “The tight-packed column of citizens four abreast stretches back along Church Street toward the corner of Elm.” Hersey made no changes from his handwritten draft except for the addition of the word “back,” which was performed with the Editor. To accomplish this, Hersey would first have had to maneuver the cursor (more Control-combinations!) to the blank space between “stretches” and “along.” He would have pressed ESC, then entered “5,” then chorded the Control and the A keys. This would have inserted five blank spaces into the document, one for each letter of the word plus one additional. Only then could he proceed to type ‘back’ into the spaces he had thus petitioned for.  

Because Kirschenbaum is scrupulous about avoiding technological determinism—a cautiousness that feels refreshing, given that books about writing and tech tend to be either luddite jeremiads or techno-utopian rhapsodies—he doesn’t quite make a unified argument about the ultimate effect word processing was to have on writing. But the meaning (or at least meaningfulness) of the medium of word processing itself becomes clearer, at least in a fragmentary and kaleidoscopic way, when he takes us back to a world of computing that is no longer recognizable: stories of first contact, even first love, between writers and their clunky, temperamental, complicated amanuenses. We no longer live in that world because word processing is no longer new, and because we live, to such a large extent—despite the Raspberry Pi; despite the cult of iFixit; despite the STEMification of education; despite encomiums to Net Neutrality and open-source everything—in the black-box future that the Macintosh portended. This may be the reason why Kirschenbaum doesn’t have much to say about Microsoft Word or Google Docs, the two black-box hegemons that so effortlessly process the critical mass of contemporary words. Because they are so seamless, so familiar, and so fundamentally opaque, they offer few opportunities for their users to contemplate them from the inside out. Still, the loss of that world makes it thrilling to inhabit, if only in fragments.

At the end of Track Changes, Kirschenbaum takes it upon himself to defamiliarize his own word processing by describing its inner workings in intimate detail: “A writer—say it is myself, say it is right now—presses a key and completes a circuit and a spark of voltage flashes through a silicon chip to open a logic gate that illuminates a pattern of pixels on a liquid crystal display screen (could the terminology be more wonderful?).” But these attempts to recapture the strangeness of digital writing only end up raising an elegiac prospect: word processing will never be new—or glorious, or difficult, or simply strange—again.