Changing My Mind

On sources, revision, and order

Lydia Davis
Author photo
Photo: Theo Cote

When I was starting out as a writer, I thought, not knowing any better, that I would be a writer of traditional short stories, though I didn’t put it quite that way. As my writing developed, I began departing from this form, further and further as the years have gone by, but I have revisited it now and then because it is a very solid and trustworthy form. One example of this return to a more traditional form is a story called “The Walk,” written about twelve years ago.

“The Walk” is set in Oxford, England, during and after a literary conference on translation—this would be typical, very acceptable subject matter for a New Yorker short story, for instance. The main characters are a translator, modeled on myself, and a critic, who is a composite of a couple of people I know. Much of the action is taken from real events at the time of a real translation conference in Oxford, and some elements in it are fictional. (One of the turning points in my development as a writer was the realization that I could, with great satisfaction, write fictional stories that were accounts of actual events, only thinly disguised.)

The main action in this story is simply a walk taken by the two principal characters. The central drama—not highly dramatic—is the narrator’s perception that there is a resemblance between the walk that she takes with the critic and a passage in her translation of Proust’s Swann’s Way. At another, more conventionally dramatic moment in the story, the narrator nearly sets off a fire alarm in the building where she is staying. In reality, the main character in the actual situation—I myself—did set off the fire alarm. But to recount this episode as it actually happened, with all the students evacuated onto the lawn, some in their bathrobes with wet hair, etc., myself apologizing profusely, would have taken the story in the wrong direction; as it is, this is a highly intellectual, even rarefied story.

As is almost always true of any story, this one was born of not just one thing but several. First, every now and then, when the opportunity presents itself, I like to reproduce the traditional short story, though usually with some less traditional variations. I am, in a way, mimicking even the traditional voice in which such a story is told, entering the persona of a certain kind of typical narrator:

A translator and a critic happened to be together in the great university town of Oxford, having been invited to take part in a conference on translation. The conference occupied all of one Saturday, and that evening, they had dinner alone together, though not entirely by choice. Everyone else who had participated in the conference or attended it had departed, even the organizers. Only they had chosen to stay a second night in the rooms provided for them in the college in which the conference had taken place, a down-at-heels building with stained carpets in the hallways, a smell of mildew in the guest rooms, and creaking iron bedsteads.

(Originally I had not named the town anywhere in the story, because I prefer not to put place names on places, but then, considering that among other things the narrator of the story is looking for the home of the famous editor of the Oxford English Dictionary, this began to seem unnecessarily coy.)

Second, I was moved by the physical beauty of Oxford as I had experienced it, especially at evening—the beauty of the buildings, with their varied architecture, in the evening light—and therefore wanted to describe the place. Third, I had for some years been very interested in the story of the creation of the Oxford English Dictionary, which took place right there. In a wonderful book called Caught in the Web of Words, Elisabeth Murray, the granddaughter of the editor, James Murray, recounts how he worked; I was interested in the creation of the dictionary not only because of a general interest in philology, reference books, and people obsessed with language, but also because of the human story, the fact that the man had involved his many children in the project, working in a little house in their backyard—and not only his children but various correspondents from around the world, including incarcerated criminals, who would send him words and quotations containing the words. Fourth, I was moved to frustration by one critic’s remarks about my translation of Swann’s Way, and it was mildly enjoyable to give expression to some of my reactions in the guise of fiction. Fifth, I was pleased and amused by the actual event of the walk, which paralleled or alluded to the walk in Swann’s Way itself, and I wanted to reproduce that in the story. This coincidence was probably the starting point of the story, what sparked it.

I was going to say that the desire was there from the beginning to include an element that would not have been allowed into a completely traditional short story, and that was the quoting of an extensive passage from Proust’s novel in not just one but two translated versions which to a casual reader would appear to be almost identical. But, really, I did not plan that from the beginning; it happened as I was writing.

What makes a piece of writing come into being? I have gradually, over the years, come to see the close parallel between the impulse to translate and the impulse to write something original: just as I want to capture something outside myself in a piece of original writing, when I translate something I also want to capture it, in this case to reproduce it in English.

Clare Cavanagh, the translator of Czesław Miłosz and Wisława Szymborska, has written an essay about translation called “The Art of Losing: Polish Poetry and Translation,” which she closes with the following statement:

Of course translating poetry is impossible: all the best things are. But the impulse that drives one to try is not so far removed, I think, from the force that sends the lyric poet out time after time to master the world in a few lines of verse. You see a wonderful thing in front of you, and you want it. You try reading it over and over, you see if you can memorize it, or copy it out line by line. And nothing works; it’s still there. So if it doesn’t already exist in English, you turn to translation; you try remaking it in your own language, in your own words, in the vain hope of getting it once and for all, of finally making it your own. And sometimes you even feel, for a while at least, for a day or two or even a couple of weeks, that you’ve got it, it’s worked, the poem’s yours. But then you turn back to the poem itself at some point, and you have to hit your head against the wall and laugh: it’s still there.

Contrast this with a passage from Swann’s Way in which Proust describes what it is like for the young Marcel to want to capture in writing something that moves him:

Then, quite apart from all these literary preoccupations and not connected to them in any way, suddenly a roof, a glimmer of sun on a stone, the smell of the road would stop me because of a particular pleasure they gave me, and also because they seemed to be concealing, beyond what I could see, something which they were inviting me to come take and which despite my efforts I could not manage to discover. Since I felt that it could be found within them, I would stay there, motionless, looking, breathing, trying to go with my thoughts beyond the image or the smell. And if I had to catch up with my grandfather, continue on my way, I would try to find them again by closing my eyes; I would concentrate on recalling precisely the line of the roof, the shade of the stone which, without my being able to understand why, had seemed to me so full, so ready to open, to yield me the thing for which they themselves were merely a cover.

And, in fact, as I copied out this passage, I saw a connection here that I hadn’t seen when I chose to quote it, between Proust’s examples of a roof and a glimmer of sun on a stone and my being moved by the physical beauty of Oxford—because that was part of the beauty of the city for me, the way the sun as it lowered in the sky toward sunset shone with a warm, honey-colored light on the roofs of the town and the stones of the buildings and the cobblestoned streets.

Next I want to move from Proust’s more sublime sort of inspiration to a more ridiculous one. Yet, though the subject matter is less sublime, the impulse feels the same: here is material that you relish, that you want to devour somehow. This story was born of a group email I received. “Nancy Brown Will Be in Town” is about a woman returning to a community in order to prepare to move away for good.

Here is the story:

Nancy Brown Will Be in Town

Nancy Brown will be in town. She will be in town to sell her things. Nancy Brown is moving far away. She would like to sell her queen mattress.
Do we want her queen mattress? Do we want her ottoman? Do we want her bath items?
It is time to say good-bye to Nancy Brown.
We have enjoyed her friendship. We have enjoyed her tennis lessons.

Before I show you the email that inspired this story, I thought it might be interesting to quote an earlier version and explain how I revised it.

Nancy Brown

Nancy Brown will be in town. We are told that Nancy Brown will be in town. We are told that Nancy Brown wants to sell her mattress. Do we want to buy Nancy Brown’s queen mattress? Do we want her ottoman? Do we want her bath items?
If she will be in town, why is she selling her things? Oh, she is only coming back to town to sell her things. She will soon be leaving town again. She will not be coming back. It is time to say good-bye to Nancy Brown.
We have enjoyed her tennis lessons. We have enjoyed her friendship.

1. First, the title: “Nancy Brown” is OK. I like the name, but “Nancy Brown Will Be in Town” is even nicer, and I don’t mind repeating the title in the first line—something often done in poems, less often done in prose.

2. You could probably hear that the earlier version was wordier overall. Too wordy. It was 106 words (including the title) as opposed to 72 in the final version. I like repetitions, but the earlier version included repetitions that didn’t move the piece forward much. It included thinking through questions and answers that could be cut back in the first little paragraph.

3. I also removed the implied “middleman” when I took out “We are told…”

4. Now, the last revision is the most interesting, to me: I changed—actually, reversed—the order of the last two sentences. Instead of “We have enjoyed her tennis lessons. We have enjoyed her friendship.” I now have: “We have enjoyed her friendship. We have enjoyed her tennis lessons.” The first order was more “logical” or traditional. It is often true that we think more conventionally first, as though reflexively, and after that we may think more adventurously and more inventively. The first, logical order placed the less important, more particular “tennis lessons” first; the more important, and more general, “friendship” second—as though moving outward from the particular to end on an appropriately general note. And yet the reversed order was actually more interesting to me: the more familiar, and expected, “friendship” first; the more surprising, even absurd “tennis lessons” second, so that the piece, which is, after all, somewhat absurd, ends on an absurd note.

As for the original email, it was the subject line, of course, that caught my attention: “Nancy Brown will be in town…” The sender was either aware, or unaware, of the felicity of the rhyme and the charm of it. Then, the body of the email:

For those of you who have enjoyed Nancy Brown’s tennis lessons, cardio tennis classes, or friendship, she’ll be in town at the end of the week for about a week and a half. She’d love to see or hear from you.
She is also getting rid of some items that she has in storage:
1 Queen mattress
1 Single mattress-box spring-frame
1 Surfboard Table
1 Scan Chair and ottoman
5 boxes of Kitchen and Bath items

What allowed me to see this email as a possible, slightly absurd piece of writing was probably, first, the charm and the singsong lyricism of the subject line: “Nancy Brown will be in town” and then the unexpectedness of the email, since I didn’t know Nancy Brown. I don’t know Nancy Brown, and yet I am suddenly drawn, intimately, into the world of Nancy Brown. Again, so much depends on context. Here, since I didn’t know her, I reacted—as though the email had been directed at me personally—by asking myself: why do I care if Nancy Brown is moving away, and why would I want Nancy Brown’s queen mattress? I read down the list of her possessions and I was not so interested—as a writer—in the single mattress box spring frame, but then I was struck by “ottoman.” I like the word ottoman—which reminds me of the Ottoman Empire and sounds so very elevated for a piece of furniture. So I wanted to keep that in my story. And then we come to the bath items—why would I want this stranger’s bath items?

Maybe I’m also struck by the sender’s including three things in the first sentence as though they were of equivalent value: “For those of you who have enjoyed Nancy Brown’s tennis lessons, cardio tennis lessons, or friendship”—another reason to reverse the order of the last line of my piece, so that “friendship” is not climactic or conclusive.

I also changed the woman Brown’s real first name, which I liked, to Nancy, just to give a little protection to the real woman, though her name was common enough. Nancy Brown was also the name of someone I knew as a teenager—who at some point, incidentally, also changed her own name, to one much more exotic.

In a New Yorker article
from 2013, John McPhee writes:

William Shawn [long-time editor of The New Yorker] once told me that my pieces were a little strange because they seemed to have three or four endings. That surely is a result of preoccupation with structure. In any case, it may have led to an experience I have sometimes had in the struggle for satisfaction at the end.
Look back upstream. If you have come to your planned ending and it doesn’t seem to be working, run your eye up the page and the page before that. You may see that your best ending is somewhere in there, that you were finished before you thought you were.

To go on to the somewhat similar situation in which the previous sentence or sentence fragment is the one you should end on, I will talk about the last two lines in my poem “Head, Heart.” I’m calling “Head, Heart” a poem because it was written to be a poem, unlike other poem-like pieces I have written:

Head, Heart

Heart weeps.
Head tries to help heart.
Head tells heart how it is, again:
You will lose the ones you love. They will all go. But even the earth will go, someday.
Heart feels better, then.
But the words of head do not remain long in the ears of heart.
Heart is so new to this.
I want them back, says heart.
Head is all heart has.
Help, head. Help heart.

First let me say something about how “Head, Heart” came to be a poem. When I was striving hardest, in my twenties, to learn to write well, I liked to read Kafka’s diaries to see the work and the thinking that went on behind his finished writing, and also the rough drafts or the false starts and the unfinished pieces. Similarly, while I was engaged on my translation of Madame Bovary, about five years ago, I liked to read Flaubert’s letters for the same reason.

There was also a time when I liked to delve into Gerard Manley Hopkins’s journals and read, especially, his detailed and thorough descriptions of natural things—the movement of eddies of water in a brook, for instance. This journal writing of his was done in a compensatory sort of way during the seven years in which he did not allow himself to write poetry, having become convinced that it conflicted with his devotion to religion—he had recently converted to Catholicism and eventually entered the Jesuit order and was ordained a priest. (Similarly, the Objectivist poet George Oppen, an American of the twentieth century, stopped writing poetry—in his case for more than twenty years—because he, too, felt that it conflicted with his discipline, though his discipline was not religious belief but political activism and for a while communism.) It is fascinating to see how the extensive and detailed descriptions in Hopkins’s journals become concentrated into the compressed and economical images in his finished poems.

Here’s a well-known poem by Hopkins:

Spring and Fall
to a young child

Márgarét, áre you gríeving
Over Goldengrove unleaving?
Leáves, like the things of man, you
With your fresh thoughts care for, can you?
Ah! ás the heart grows older
It will come to such sights colder
By and by, nor spare a sigh
Though worlds of wanwood leafmeal lie;
And yet you wíll weep and know why.
Now no matter, child, the name:
Sórrow’s spríngs áre the same.
Nor mouth had, no nor mind, expressed
What heart heard of, ghost guessed:
It ís the blight man was born for,
It is Margaret you mourn for.

An explicator at an online site (David Coomler at the site “Hokku”) went through the poem in a very cogent, sensible manner; but when he came to the end, he expressed an opinion which I disagreed with, as to whether or not a couple of lines were “poetic” or not. Here are the last four lines of the poem again:

Nor mouth had, no nor mind, expressed
What heart heard of, ghost guessed:
It ís the blight man was born for,
It is Margaret you mourn for.

I had particularly always liked the first two of these lines, especially the first one:

Nor mouth had, no nor mind, expressed
What heart heard of, ghost guessed:

But the commentator, David, says that he feels these lines are left for the reader to “somewhat laboriously unravel” and that they “are not very poetic in their complexity…”

Another site suggests that the most beautiful line in the poem is: “Though worlds of wanwood leafmeal lie”—the w and l sounds, the assonance of “leafmeal,” and the alliteration of “leafmeal lie.” What we have here are simply two different ideas—not mutually exclusive—of what constitutes a beautiful line of poetry.

There are reasons why I like the more di∞cult lines: First and most obvious, there is the parallel alliteration: “Nor mouth… no nor mind” in the first line; and then the one-syllable words and alliteration of “heart heard… ghost guessed” in the second (ghost here meaning “spirit”). Then there is the compression of the whole, created in the following way: through the inversion of “Nor mouth had”; the elision of the article “the” in “Nor mouth”; the jamming-together of the words “no nor mind” without the usual comma punctuation. Besides the compression, there is the vehemence of the “no,” along with the bluntness, the almost animal reference of the word mouth to indicate the organ of expression: “Nor mouth had, no nor mind, expressed” rather than something more refined such as “Your lips [or tongue] had not expressed this thought; nay, your mind not even entertained it.”

The effect, to my ear, anyway, of this compression and bluntness was to make the lines convey more emotion, as though they had been wrested from the speaker almost by force.

I had thought that the alliteration and compression of Old English poetry might have led to the form of my poem “Head, Heart” but the more I look at the poem by Hopkins, the more I think that perhaps his work, too, had an influence—particularly his use of alliteration so extreme that it calls attention to itself, which nicely breaks another sometimes repeated rule, which is that your deployment of rhetorical techniques should not be so overt that it distracts from what you’re saying.

And in fact, the connection I see between Anglo-Saxon poetry and the verse of Gerard Manley Hopkins is not accidental, since, as I learn when I read more about Hopkins, he himself became fascinated with the older rhythmic structure of Anglo-Saxon poetry.

I would have to go back and do some studying to talk properly about the characteristics of Old English poetry—that is, poetry written in Old English, or Anglo-Saxon, between the seventh century A.D. and a few decades after the Norman Conquest of England, which occurred in 1066. (This date is one that should be remembered especially by writers in English, because it is after this date that the French language was aggressively introduced into the existing languages of England; the Norman Conquest resulted in our having the wonderful doubled vocabulary that we have in English, examples being subterranean and underground, cordial and hearty, omnipotent and allmighty, etc.) But what I can say is that it makes heavy use of alliteration within each line and also breaks each line into two parts, with a caesura—the technical term in poetry for a pause.

Here are just a few lines from the beginning of a poem called “The Battle of Brunanburh.” It is the earliest poem contained in the Anglo-Saxon Chronicle, and it dates from 937. It celebrates the victory of King Aethelstan over the Scots and the Norse.

                                   …. feld dænnede
secga swate               siðþan sunne up
on morgentid,          mære tungol,
glad ofer grundas,   Godes candel beorht,
eces Dryhtnes,          oth seo aethele gesceaft
sag to setle.

                                      …. The field flowed
with blood of warriors  from sun up
in the morning,            glorious star
glided over the earth    God’s bright candle,
eternal lord,                 till that noble creation
sank to its seat.

Old English rhythms and word-sounds may lie somewhere in the distant or not so distant past behind my choices in “Head, Heart,” and so may Hopkins’s rhythms and word-sounds, as the Anglo-Saxon rhythms lay somewhere behind Hopkins’s rhythms. I also see the word and line choices of “Head, Heart” as being in a way deliberately somewhat awkward; I am interested in writing that is eloquent yet awkward, or awkward yet eloquent, as in the Hopkins lines that included “Nor mouth had, no nor mind.” The awkwardness implying that the writer was too moved to be more elegant. However, I didn’t think of all this or plan it before I started the poem—or even after I had finished it. The deliberate awkwardness was a natural impulse.

Here is an earlier version of the poem:

Head, Heart

Heart is disturbed.
Head is trying to help heart “deal” with this.
Head tells heart how it is, again:
Here is life, here is death, you will lose the ones you love. They will all
go. We will all go. Even the earth will go, one day.
Heart feels better, then.
Now heart feels worse again. The words of head do not remain long in the ears of heart.
Heart is so new to this.
I want them back, says heart.
Help, head. Help heart. You are all heart has.

And here, again, is the final version, for comparison:

Head, Heart

Heart weeps.
Head tries to help heart.
Head tells heart how it is, again:
You will lose the ones you love. They will all go. But even the earth will go, someday.
Heart feels better, then.
But the words of head do not remain long in the ears of heart.
Heart is so new to this.
I want them back, says heart.
Head is all heart has.
Help, head. Help heart.

The opening line of the final version is plainer and actually more Anglo-Saxon: “weeps” as opposed to “is disturbed.” The second line is also plainer and less contemporary in its language: “Head tries to help heart”—all of these being Anglo-Saxon-derived words. As opposed to the wordier and more contemporary “Head is trying to help heart ‘deal’ with this.” Although that line is completely Anglo-Saxon too, “deal with” is a contemporary expression. There is also a shift of tense. In the earlier version, the first two lines establish the situation with the use of the verb “to be” as an auxiliary with an adjective and a present participle—“is disturbed; is trying”—then in the third line comes the “action”: “Head tells heart.” In the later version, the verbs express action from the beginning: “Heart weeps / Head tries… / Head tells…” It is also blunter, and therefore, to my ears anyway, more emotional.

The revised version has been cut throughout—more than one-fifth of the material is gone. And I have reversed the last sentences so that the stronger sentences—both commands—come last, in a single line: Instead of “Help, head. Help heart. You are all heart has,” the last lines now read: “Head is all heart has. / Help, head. Help heart.”

In working on this poem—and almost everything else I have written, no matter how brief—I have needed to reread and revise it over time, allowing it to rest between readings, until it felt “right,” until I could not see any more obvious weaknesses in it. Almost always, time has to pass before I can see clearly what needs to be done to a story or poem—time for me to change my mind.

Lydia Davis is the author of Essays One, her first collection of nonfiction. Her most recent collection of stories is Can’t and Won’t.
Originally published:
January 1, 2020


Rachel Cusk

The novelist on the “feminine non-state of non-being”
Merve Emre


Renaissance Women

A new book celebrates—and sells short—Shakespeare’s sisters
Catherine Nicholson

Fady Joudah

The poet on how the war in Gaza changed his work
Aria Aber

You Might Also Like

Why I Write

Getting ready not to be
Samuel R. Delany


Aria Aber


Sign up for The Yale Review newsletter and keep up with news, events, and more.