The Art of Revision: Black Box

Peter Ho Davies

In The Art of Revision: The Last Word, forthcoming on November 2 from Graywolf Press, Peter Ho Davies delivers a long-overdue and deeply satisfying exploration of a frequently discussed yet widely misunderstood and underestimated aspect of every writer’s working life. The author of three novels, including A Lie Someone Told You About Yourself (Mariner Books, 2021), and two story collections, begins his careful study of revision by addressing its invisibility, with the majority of readers and writers seeing only the final draft. Davies uses examples from his own books as well as from the work of other writers, including Raymond Carver, Carmen Maria Machado, and Flannery O’Connor, to illustrate his points. By turns deeply personal and intellectually rigorous, The Art of Revision: The Last Word is a moving appeal to writers on the importance of reflecting not only on their process, but also on their lives. The following is an excerpt from the first chapter.

The Art of Revision: The Last Word will be published by Graywolf Press on November 2, 2021. 

As a teacher of writing, I’ve often been struck by the sense that revision is an overlooked, under-addressed, even invisible aspect of our work: the “elephant in the workshop,” if you will. It’s notionally what most of our discussion in fiction classes, or constructive feedback from any trusted reader, points toward. Ideally the value—the lessons, the suggestions, the encouragement and energy—of such feedback extends into revision of the stories shared. But while a typical manuscript critique will raise issues about a draft and discuss possible remedies, the revisions resulting from those discussions are themselves likely to remain invisible to most of us. Even students I work with in an MFA program may only share a couple drafts of a story over two years, and then not always with the same readers. The upshot is that for all the work we share, we don’t get to see very much of each others’ revision process. 

This isn’t just a function of creative writing teaching. Revision has historically been hidden away from the view of all but a select few—editors mostly. As readers we don’t, for instance, typically have access to multiple drafts of published work. With a few exceptions—early drafts of The Great Gatsby, say, or Lady Chatterley’s Lover, brought to light largely through scholarly interest—most of us never get to trace the genesis of any work, other than our own, which seems strange and a little problematic given that in most every other aspect of our craft writers are encouraged to learn from other writers by reading and studying their work. In that fashion, we might well learn dialogue skills from one writer, plotting from another, techniques for how to move in and out of flashback, etc, etc. And to be sure we often deploy these skills on later drafts in the revision process. But what if revision isn’t just an accumulation of other skills, but, as I hope to suggest, a skill in itself, a technique of its own, a state of mind even? In that case, the paucity of its examples deny us access to an essential tool.

We might, in passing, compare this state of affairs to the practices in other arts. Painters, say, leave us sketches, studies, cartoons (often prized as works of art in their own right). Many gallery shows contain series of works, variations on a theme, which differ only slightly from one another and might (loosely) be imagined as a set of drafts. And compared to the black box of literary revision, a rehearsal process—in the dramatic arts or music—is an open book. 

There are to be sure a handful of notable instances of published drafts that we might draw on. The textbook pairing—fitting given his own avowal that “it’s instructive, and heartening both, to look at the early drafts of great writers”—is a couple of Raymond Carver stories: “The Bath” and “A Small Good Thing.” And yet the utility of those famous examples is complicated by what we now know: that these are less instances of revision by Carver, than of far-reaching cuts and changes by his editor Gordon Lish, which Carver himself felt deeply equivocal about. Fascinating and instructive as the competing versions of these stories are, the revision process can once again be obscured if it’s too tightly conflated with the editorial process.

Over the years I’ve collected several other instances in which two drafts of the same story have been published. The Irish writer Frank O’Connor published multiple versions of the same story during his long career. The other O’Connor, Flannery, revisited one of her first published stories “The Geranium," a troubling engagement with race, at the very end of her career in a stringent reckoning entitled “Judgment Day.” More recently several of Wells Tower’s stories in his collection Everything Ravaged, Everything Burned illustrate the sometimes radical changes between the version of a story originally published in a literary journal and the version collected in a book. The tell-tale phrase “some of these stories previously appeared in different form in the following journals” can be found in the acknowledgments of many short story collections, but Tower is unique in that different versions of his story “Retreat” appear in two issues of the same journal, McSweeney’s. Lately, in an encouraging development, a new magazine Draft: The Journal of Process, has begun publishing fascinating early versions of stories, including Carmen Maria Machado’s “The Husband Stitch.” 

All of these examples are worth seeking out, but it’s noticeable that they have to be sought out. They have an air of the scholarly, the obscure, the collector’s item which seems contrary to the inevitable ubiquity of revision. Their very rarity is part of the problem. They make what should be the norm seem exceptional. 

And this invisibility of revision leads to both its misunderstanding, and like all invisible labor, its undervaluing. 

I’ve generally felt fortunate, over twenty-some years of teaching, that creative writing is a pretty “cool” subject for undergrads (it’s arguable the competition isn’t too stiff!). The students tend to want to be in the class and are usually willing to do the work. In my experience, however, revision is the exception—the perennially “uncool” aspect of creative writing that undergrads, but also a few graduate students, still resist.

When I ask students to revise their stories the results—even from talented writers—are often underwhelming. I get dutiful drafts, wherein the students half-heartedly take two or three suggestions from workshop, making just enough mechanical changes to begrudgingly acknowledge the feedback. This is revision as grade grubbing. Alternatively, I get “drafts” so wholly unrecognizable from their originals—not so much the baby thrown out with the bathwater, as baby, bathwater, rubber ducky, and all thrown out in the bathtub for convenience—that they’re essentially brand-new first drafts. 

Both these responses are less revisions per se than rejections of revision. The first somewhat arrogantly implies, My story doesn’t really need any improvement; the second rather abjectly suggests, My story is so bad, so hopelessly beyond repair, I can’t bear to look at it again, so I’ll start over. 

I don’t mean to suggest that all aspiring writers share these inexperienced attitudes to revision. Though, since all writers, myself included, swing from arrogance to abjection, often in the same morning, we might recognize something of ourselves in these callow responses. And even if you’ve overcome the resistance to revision that my undergrads are struggling with, even if you’re now a self-professed “lover of revision,” if you’ve ever plaintively asked “How many drafts is enough? How do I know when a story is done?” you likely share some of their underlying anxiety and resentment about revision. An anxiety that is as much about how to do it, as what it is.

Let’s start with what it’s not.

My undergraduates, I’ve come to realize, are struggling with a bias—ingrained and likely unconscious—against revision. 

Revision in high school or middle school can easily and narrowly seem to mean grammatical revision. Valuable as those lessons are, they tend to make revision feel like a chore, about as much fun as tidying one’s room, or “washing the dishes,” as Colson Whitehead puts it. The danger is that in this limited sense revision comes to seem rigid, rule-based, proper—the antithesis of everything “cool” about creativity. 

But there are other deeper cultural forces lined up against revision, and in favor of first drafts. 

There’s our tendency, say, to romanticize inspiration over perspiration, to consider the first draft the best draft. The popular image of certain writers might factor in here—think Jack Kerouac composing On the Road on that continuous roll of paper. Truman Capote’s famously withering response, “That’s not writing, that’s typing,” is worth recalling, and yet Kerouac’s 120-foot “scroll” (a single paragraph, single spaced; Kerouac could type a Benzedrine-assisted 100 words a minute) continues to have an iconic status in the cultural imagination. The truth is that Kerouac did in fact revise. Despite his editor Robert Giroux’s plaintive, “But Jack, how can you make corrections on a MS like that?” there’s evidence of as many as six drafts composed in the early 1950s, a truth partly obscured by Kerouac himself when he told Steve Allen in a TV interview that he wrote the book in three weeks. Yet there remains an allure of authenticity to this idea of the raw first draft. Plus, all writers—perhaps especially those, like myself, with feet of clay—are seduced by the myth of the genius for whom it all comes quickly and easily, fully sprung from the head of Zeus. 

As a sidebar, dare I—as a Brit—suggest, there may even be an “American” bias against revision. Consider the tendency to see the Constitution as a perfect document, one to be read—in certain quarters at least—narrowly and literally as if it were holy writ? It may be salutary, the next time you feel yourself resisting revision, to consider you’re aligning yourself with strict originalists like Antonin Scalia! As an antidote, it might help to recall that the framers were actually revisers—of a nation, of government—and that revision can be an act of revolution.

In short, we all—readers and writers, alike—tend to associate inspiration with first drafts, and perspiration with what comes later, a cultural bias that informs the lingering distrust for writing programs, as well as the general scanting of revision. 

Kerouac himself was probably both acceding to as well as perpetuating the myth, and before we accuse him of any self-aggrandizing burnishing of his own legend, we might reflect how for all writers—prone as we are to a squirming embarrassment, if not downright shame, at our fledging efforts—revision is in a sense an act of forgetting our earlier mortifying drafts, an act of forgetting moreover that manages to forget even itself. That’s why the examples cited above feel like such rarities. Final drafts tend to erase those that come before them and a successful revision invariably covers its own tracks; the invisibility of revision is not accident, but intent. Asked why he doesn’t keep rough drafts, Tobias Wolff answers for most of us, I think: “They embarrass me, to tell you the truth.”

If the myth of the Muse seems an ancient veil for revision, new technologies of writing—most obviously the displacement of the typewriter by the computer—have also contributed to its invisibility. 

I’m (just) old enough to have started writing on a typewriter—a member of the last generation of writers to precede the advent of the word processor. As such I have a firsthand sense of both the painstaking work of revision on a typewriter (white-out and corrector tape; scissors and paste) and the seduction of that winking cursor with its promise of infinite easy revision. 

The freedom to revise as much as one liked, the electronic speed of those changes, and the flawless (typographically speaking) results were almost miraculous to typewriter users. It’s no accident that the software of the day was called things like Word Perfect or Perfect Writer. All this might suggest that the virtual nature of text on a screen encourages revision, but there’s also an argument that the word processor’s ability to make a text look prematurely “finished”—thanks to font options and justified text—militates against revision, in a way the relative imperfection of typescript does not. Our manuscripts essentially look publishable long before they are.

The personal computer’s radical reshaping of the revision process is likely another reason why writers sometimes struggle to understand revision. Consider the basic idea of the draft, the notion that revision is measured draft by draft, and the regular conversations that arise in interviews with writers about how many drafts a story or novel has gone through. Discussion of discreet drafts made sense in the era of the typewriter. An author might type their way through a draft, mark it up, retype it, and repeat until done, each iteration yielding a new pile of pages, a distinct and easily numbered draft. But it makes a lot less sense to talk about a third draft or a thirtieth in an era of word processing, when a text is subject to almost continuous revision, infinite drafts, if you will, one saved on top of the other in the hard drive. As Philip Roth noted of his early experiences with a computer: “I’m doing so much changing as I go along that the drafts disappear, as it were, into the rewrites.” The concept of the draft, while it remains useful, has thus become more fluid and elusive. 

Interestingly, several contemporary writers persevere with the old methods, or some combination of writing “technologies”, starting in longhand on legal pads, say, before typing up later revisions, thus preserving some demarcation of drafts. Anne Tyler, profiled in the New York Times, subscribes to an even more elaborate process perhaps in an effort to make the most of all available technologies: “She writes in longhand, draft after draft, and when she has a section she’s satisfied with, types it into a computer. When she has a completed draft she prints it out and then rewrites it all in longhand again, and that version she reads out loud into a Dictaphone.” Tyler would seem to be re-seeing (literally revising) and even rehearing her work at various of these iterations, holding off that settled aura of inevitability that a draft accumulates after we’ve read it too many times in the same form. 

The fresh sensory apprehension of an old draft is one reason I read my own work aloud and encourage my students to do the same (stressing that it’s not enough to simply lip-sync it, however shy they might be about being over-heard by roommates). Nicholson Baker, via his character, the poet Paul Chowder, even playfully advises doing so in different voices: “a juicy Dorchester accent, or a Beatles Liverpool accent, or a perfectly composed Isabella Rosellini accent.”

If technology complicates our understanding of revision, science in the form of a 2012 academic paper by Carney and Banaji entitled “First is Best” may clarify it. As its abstract states: “A century of research on human and nonhuman animals has suggested that the first experience in a series of two or more is cognitively privileged.” 

In one cited experiment subjects were asked to choose between two similar-looking sticks of gum, and the researchers found that respondents picked the first they were shown 62 percent of the time, as opposed to the second which they opted for only 38 percent of the time. (I like this particular example, I should say, because gum is so humble, so down to earth. It’s less easy to rationalize or mythologize our preference because the first gum is more inspired or more authentic!) 

The real takeaway from this study, though, may be that those preference percentages—so heavily weighted in favor of the first choice—are only true when respondents are asked to make a rapid decision. When given more time for deliberation, they showed an equal preference for gum #1 and gum #2. The finding recalls one of my favorite pieces of writing advice: Flaubert’s famous dictum that “Talent is long patience.” I confess when I first read it that I suspected a bad translation from the French. But the older I get the more the call to patience resonates. I might even be tempted to rephrase Flaubert’s line as “revision is long patience.” 

Or consider Dorothy Parker’s famous—and widely shared—sentiment, “I hate writing, I love having written,” as a corollary of which I might observe that we are, most of us, often indecently eager to finish whatever we’re writing.

That haste makes intuitive sense in the case of first drafts. We’re often racing our own skepticism—can I write this, is it worth anything?—in those early stages, praying to get to the end before our doubts overtake us. Building a bridge across a chasm of our own doubt. (Tellingly, Jane Smiley equates patience with faith.) There’s an all-or-nothing feeling to first drafts. If we finish them we feel we have something, however flawed, to build on. But if we stop before the end, it’s not clear we have anything at all. Is a bridge half-finished even a bridge?

But what applies to first drafts needn’t apply to later ones. 

Rather, I might suggest, by way of paraphrase, “Draft in haste, revise at leisure.” A first draft, by these lights, might be viewed as a kind of whirlwind romance; revision a mode of repentance!

Finally, though, if we’re apportioning blame here for the many misunderstandings of revision, we—writers and especially teachers of creative writing—may have to take our share. One common tenet of the workshop, after all—one I subscribe to myself—is that we try to judge a story with respect to its author’s intent. That’s a respectful and pragmatic strategy in workshop. If a writer’s goal is tragedy, then asking at the start of any discussion for more jokes seems to miss the point, dismiss the author, and likely means he or she won’t heed the suggestion (even if a case might be made that most tragedies could use a few more jokes, and that they might be all the more tragic for them). But this strategy—baked in as it is to many workshops—none the less bears examination. As I point out to my own classes, it rests on an assertion that we writers know our own intent in an early or first draft when I’d argue on the contrary that we rarely know ourselves or our work (which may amount to the same thing) quite so well at the outset. 

I blame the common advice “write what you know” in general, and Hemingway in particular, for this idea of us “knowing” our stories. All those famous, bracing lines of his:

“I decided that I would write one story about each thing that I knew about.”

“Write the truest sentence that you know.”

Or—the so-called iceberg theorem:

If a writer of prose knows enough of what he is writing about he may omit things that he knows and the reader, if the writer is writing truly enough, will have a feeling of those things as strongly as though the writer had stated them. The dignity of movement of an iceberg is due to only one-eighth of it being above water.

What comes out of these quotes is the primacy of knowing what we’re writing about, the primacy of intent—write what you already know, is the implication—which in turn implies a kind of before-and-after model of writing. Before: We know what we want to write. After: We revise what we’ve written to better match what we set out to write in the first place. Revision by these lights becomes a process of perfection, the process of perfecting an initial platonic ideal of the story. 

And there are a few problems with that.

In the first place, perfection can be an impossible, even dispiriting goal. If you’ve ever thought to yourself, despairingly, I could go on revising this forever, you’ve just taken a look down that ever-receding, infinitely regressing hall of mirrors that is perfection.

In the second place, the before-and-after/knowing-then-writing model again tends to imply that inspiration is restricted to the before, which makes the actual writing feel like about as much fun—which is to say as creative—as painting by numbers. 

In the third place, as you may have already realized, this just isn’t how we do it, this isn’t our actual experience of writing. What most of us already know is that writing is a mode of thinking. As we write, as we express our initial thoughts, new ones spring to mind, to elaborate or complicate the initial ideas. We might call these inspirations, or more humbly surprises, but either way I suspect they’re at the heart of the pleasure of the enterprise. 

“Write what you know,” does have value as advice. It speaks to the fundamental authority of authorship, the contractual notion that if we ask for a reader’s attention for thirty minutes or three hundred pages, we know something about our subject. But this speaks best to an end product, a final draft. The question we might ask is how we come to know what we know, and I suspect a guiding principle of early drafts might be better phrased as “Write to know,” and of revision, “Revise to know more,” and of a final draft, “I’ve written what I now know.”



Peter Ho Davies is the author of three novels, including A Lie Someone Told You About Yourself, and two story collections. The winner of the Anisfield-Wolf Award and the PEN/Malamud Prize, he teaches at the University of Michigan and lives in Ann Arbor.