Musings: Notebook to Essay

In her essay “The Nature and Aim of Fiction,” Flannery O’Connor gives a piece of advice that sounds almost rude: “The writer should never be ashamed of staring. There is nothing that doesn’t require his attention.” Just before it, she names the disciplines that can help a writer, “logic, mathematics, theology, and of course and particularly drawing,” and then she says what they have in common: “Anything that helps you to see, anything that makes you look.”

I have always liked that she reaches for drawing. Drawing is slow. You cannot draw a hand from a general idea of hands. You have to look at this one, with its particular knuckles and the light lying across it in a particular way, and you have to keep looking long after you think you have seen it. The drawing gets better at exactly the moment you stop assuming.

Staring is not getting the point at once. It is staying until the thing tells you what it is.

O’Connor goes further in the same essay. She says there is “a certain grain of stupidity that the writer of fiction can hardly do without, and this is the quality of having to stare, of not getting the point at once.” I find that a consoling sentence. So much of what we admire in quick minds is the habit of moving on, of summarizing, of knowing what a scene means before it has finished happening. A story needs the opposite habit. It needs someone willing to remain a little slow in front of a kitchen table or a stranger’s coat, and to wait.

When I read a draft that is not working, my own or someone else’s, the problem is very often a failure of staring. The writer has seen a room the way one sees a room on the way through it. The details are the expected ones. The feeling has been named instead of found. Nothing in the paragraph has been looked at long enough to surprise the person looking.

The remedy is not more adjectives. It is more attention, and attention is mostly patience. Look again at the thing you think you have already described. Ask what is actually there rather than what ought to be there. The first look gives you the category. The second, or the fifth, gives you the object.

O’Connor’s sentence has a moral weight I did not feel when I first read it. To stare without shame is to believe that the ordinary world deserves the effort. That belief, more than any technique, is where a story begins.

“The art of losing isn’t hard to master.” That is the first line of Elizabeth Bishop’s “One Art,” and it is one of the most quietly untrue sentences in American poetry. The poem knows it is untrue. That is what makes it so hard to forget.

Bishop began the poem in October 1975. Her first attempt was titled “How to Lose Things,” and it was closer to prose than to verse: reading glasses, a favorite pen, the small daily mislayings anyone could list. By the time she was finished there were seventeen drafts, now kept at Vassar, and the loose list had become a villanelle, one of the strictest forms we have. Bishop is reported to have said that the poem simply came to her. The drafts suggest otherwise.

Seventeen drafts to make something sound as if it had cost nothing.

That gap between the ease of the finished poem and the labor behind it is what I keep returning to. A villanelle repeats two lines over and over, and the danger of the form is that the repetitions go dead, the way a phrase does when you say it too many times. Bishop’s refrain does the opposite. Each time “the art of losing isn’t hard to master” comes back, it is a little less convincing, because the losses keep growing: keys, an hour, houses, cities, a continent, and finally a person. The form holds steady while the speaker’s composure gives way inside it.

Then comes the last line, with its two words in parentheses: “(Write it!) like disaster.” Those two words are the whole poem in miniature. The speaker has to order herself to finish the sentence. The poem that claimed loss was easy has to force itself to name the thing that is not easy at all.

I think of this when writers tell me a piece came quickly, as if speed were proof of truth. Sometimes it is. More often what looks effortless on the page is the residue of a great deal of effort that has been carefully removed. The drafts are where the honesty happens. The finished poem only has to be brave enough to keep it.

What I take from Bishop is permission to revise as long as a poem needs, and a warning not to mistake the polish for the point. The art of losing is hard to master. So is the art of making it look otherwise.

I used to think an ending was a lock. You turned the last sentence like a key, and the story was secured, its meaning safe inside. I wrote that way for years, and I taught that way, and I am not proud of how many stories I closed too firmly.

What changed my mind was not a theory. It was a pattern I noticed in the stories I kept going back to. None of them ended by explaining themselves. They ended a little before the explanation would have arrived. A character stands at a window. A conversation stops mid-gesture. A small action is performed, an ordinary one, and the page runs out. The story does not tell me what to feel. It leaves the feeling in my hands, and I carry it out of the room.

An ending is not a lock. It is a door left ajar, and the reader decides whether to go through.

This is harder to do than it sounds. The pull toward the last explanatory sentence is strong, because the writer is afraid, rightly, that the reader will miss the point. I know the fear well. I have written the final paragraph that says what the story meant and then deleted it, and then restored it, and then deleted it again. Usually the deletion is correct. If the story has done its work, the explanation is redundant. If it has not, no last paragraph will rescue it.

What I try to ask now, when a draft is nearly done, is a plainer question: what is the last true thing? Not the last clever thing, or the most resonant one, but the last thing that is simply so, and that the reader can stand on. It is often an object, or a gesture, or a bit of weather. It is rarely a conclusion.

When I teach this, students sometimes hear it as a license for vagueness, and it is the opposite. An open ending takes more precision than a closed one, because the whole story has to be built to make the silence mean something. The door can only be left ajar if the room behind it has been furnished with care.

I still write the explanatory paragraph sometimes. I just keep it in a separate file now, where it can do no harm, and where I can look at it and be reassured that I know what the story is about. Then I close the file, and let the reader find out for themselves.

When I sit down with a manuscript, the first thing I have to resist is my own opinion. It is not that I don’t have one. It is that my opinion, at the start, is almost always premature. I have not yet understood the book. I have only read it once, and a book read once is a book half-read.

So I do not begin with judgment. I begin with questions. Not the questions I want to answer, but the questions the manuscript is already asking, whether the writer knows it or not. Where is this going? What is it trying to become? What is it afraid of?

The editor’s question is not the editor’s answer. It is the writer’s intention, asked back to them.

A good question does not impose. It opens. If I tell a writer “this paragraph is weak,” I have given them a verdict, and a verdict is a wall. But if I ask “what is this paragraph doing that the one before it has not already done?” I have given them a door. They can walk through it themselves, and what they find on the other side will be their own, not mine.

This is the hardest thing to teach young editors, and the hardest thing to remember myself. The urge to fix is strong. The urge to rewrite a sentence the way I would write it is almost physical. But the moment I rewrite a writer’s sentence in my own voice, I have stopped editing and started co-authoring, and the writer did not ask for a co-author. They asked for a reader who would help them see their own book more clearly.

So I ask, and I wait, and I ask again. The best editing I have ever done felt less like work and more like listening—to the book, and to the writer, and to the gap between what the writer meant and what the page actually says. That gap is where the work is. The question is the tool that finds it.

I read by return. The books I trust are not the ones I finished but the ones I went back to, and what was different between the first reading and the second was rarely the book.

The first time through, I am listening for the plot—where it goes, what happens next. The second time, the plot is already mine, so I can hear what the story is quietly saying: the sentence I hurried past, the image doing small work in a corner, the recurrence I was too busy to notice. Plot is what a story gives a stranger; the second reading is what it keeps for a friend.

This is close to what I mean when I say my own work returns to the same faces and rooms. A short story read once is a person met once; a story reread is a person met again, and each meeting changes what you saw the time before. Nothing is lost in the return. It is carried forward and heard differently.

I tell my students this, and they do not quite believe me, the way I did not at their age. But a book you have finished is not finished with you. The ones that matter arrange to be read again. They leave a room still lit, a door still open—and you will find it only if you come back.

Alice Munro can move thirty years in a single sentence, and you will not notice it happening until you are already there. A story that begins with a woman at a kitchen table in 1955 will, by its third page, have arrived at the same woman in a hospital room in 1985, and the leap will feel less like a cut than like a breath—the way memory itself moves, not in sequence but in association, the past surfacing inside the present the way a face surfaces in a window.

I return to Munro when I have forgotten what the short story can do. She is the proof that a story does not need a plot to hold you; it needs a mind, moving. Her stories are not built on events so much as on the way a life keeps returning to the same few events, worrying them, turning them over, finding something new in them each time.

The short story is not a small novel. It is a different instrument, and Munro plays it the way it was meant to be played.

What I have learned from her, and what I try to teach, is that time in a story is not a straight line to be followed but a material to be shaped. A writer can spend three pages on a single afternoon and then cross twenty years in a paragraph, and the reader will accept both—will barely notice the second—if the compression is earned. The trick is not to summarize the years but to let them pass through a single detail: a coat that no longer fits, a name that has changed, a room that is the same but emptier.

Munro taught me that the short story’s power is not in what it includes but in what it lets the reader feel pass by. The years she skips are not missing. They are present in the way the present moment is heavy with them. That is what I want my own stories to do—to make the reader feel the weight of a whole life in the space of a single scene.

I read my poems aloud before I trust them. Not to perform them—to hear them. There is a difference, and it took me years to learn it.

The eye forgives too much. On the page, a line can look right and still be dead—the rhythm off by a syllable, the ending landing where it should not. But the ear does not forgive. Read the poem aloud, slowly, and every false note announces itself. The tongue stumbles where the line is wrong. The breath runs out where the sentence has gone on too long. The voice drops where the poem has lost its music.

The poem is not finished until it has been said, not just seen.

I think this is why poems were spoken before they were written. The line break is not a typographical choice; it is a notation for the breath. A short line asks you to pause. A long line asks you to keep going, to run out of air a little, to feel the pressure of the thought before it releases. When I read a poem aloud, I am reading the poet’s instructions for how to breathe.

My students resist this at first. They want to read silently, quickly, the way they read everything else. But a poem is not information to be scanned. It is a small machine made of sound, and it only runs when you turn it on with your voice. So I make them read aloud—their own poems, and the poems of others—until the ear wakes up and starts doing the work the eye has been skipping.

The test is simple, and I use it on everything I write now, not just poems. If a sentence is hard to say, it is hard to read. If it is hard to read, it is not finished. The ear knows before the mind does. Listen to it.

We use the word epiphany now to mean almost any small realization—a moment of clarity, a sudden understanding. We have Joyce to thank for that, and we have, as usual, thinned the word out in the borrowing. What Joyce meant by it was something more specific, and I think about it whenever I sit down to write.

As a young man in Dublin, Joyce kept a notebook of what he called epiphanies—short sketches, a few lines each, of moments he had witnessed or overheard: a fragment of conversation, a gesture, a look that passed across a face. He believed these moments, gathered faithfully, were the raw material of art. Not the grand events, not the plotted turning points, but the small, sudden flashes in which, as he put it, the soul of the commonest object seemed to radiate.

The ordinary moment, seen clearly enough, gives off light.

I think of this when I teach, because students arrive believing they need a big subject—a death, a betrayal, a catastrophe—in order to have something worth writing about. And I keep telling them what Joyce kept telling himself: the material is already there, in the overheard sentence, the remembered gesture, the thing you noticed and could not say why it mattered.

The discipline is not in finding the moment. It is in trusting it. An epiphany is not manufactured; it is recognized. You cannot decide to have one. You can only learn to notice, and to write the noticing down before it fades.

I keep a notebook for the same reason Joyce did. Not for ideas, exactly—for the small things that arrive already lit: the sentence overheard on the train, the way a student’s face changed when she finally understood. These are not the finished work. They are the promise of it.

The word has been worn thin, but the thing itself has not. It is still out there, on the train and at the kitchen table, for anyone willing to look long enough.

Every writing student knows Chekhov’s gun. Show a rifle on the wall in the first act, the rule goes, and it must go off by the third. The advice has been repeated so often, and so confidently, that most people who quote it have never read the letter it comes from—and would be surprised by what Chekhov actually meant.

He did not mean that every object must be used. He meant that nothing unnecessary should be there in the first place. The gun is not a promise the writer has to keep. It is a warning the writer gives themselves. If it does not belong to the story, take it down from the wall before the reader ever sees it.

I think about this constantly when I edit. The manuscripts that arrive on my desk are rarely short of material. They are short of necessity. A scene that does the work of the scene before it. A metaphor that is beautiful and also, in the end, decorative. A detail the writer loves because it is true, not because it is needed.

Cutting is not a verdict on the writer’s talent. It is an act of mercy toward the story, which is trying to become itself and keeps getting interrupted.

The hardest cut is always the one the writer loves most. I know this from both sides of the desk. The sentence I am proudest of is very often the sentence that has to go, because it is proud of itself, and the reader can feel it.

Chekhov was not a cold writer. He had learned that holding back is a kindness to the reader. Leave the wall bare. Trust the room.

A boy who asked too many questions, and a former nun who told him he had permeable membranes. This note grew into an essay, and the whole story—Carol, her sentence, and what Ernest Hartmann called boundary thinness—is told in “Thin Places.”

I keep these two on the same shelf, though they never met and would likely have unsettled each other if they had. Walt Whitman and Emily Dickinson are commonly named the first great American poets, and I think the label holds—not because they wrote alike, but because each, independently and almost simultaneously, broke with English precedent and invented a native way of hearing. Whitman did it by knocking down the wall between the poet and the world. Here he is at the opening of “Song of Myself,” already refusing the modesty English verse expected of him:

I celebrate myself, and sing myself,
And what I assume you shall assume,
For every atom belonging to me as good belongs to you.

I loafe and invite my soul,
I lean and loafe at my ease observing a spear of summer grass. —Walt Whitman, “Song of Myself,” 1

Dickinson broke the same wall from the opposite direction—not by enlarging the self until it held everyone, but by narrowing the aperture until a single perception could not be paraphrased away. Where Whitman’s line runs long, cumulative, and breathing, hers stops, doubts itself, and starts again on a dash:

I’m Nobody! Who are you?
Are you — Nobody — too?
Then there’s a pair of us!
Don’t tell! they’d advertise — you know!

How dreary — to be — Somebody!
How public — like a Frog —
To tell one’s name — the livelong June —
To an admiring Bog! —Emily Dickinson, 288

Set side by side, the two techniques could hardly be more opposed. Whitman writes in long, unrhymed, incantatory lines built on anaphora and cataloguing—the famous lists of occupations, landscapes, and bodies that pour down the page until the poem feels less like a description of America than an act of ingesting it. His grammar is additive: and, and, and, one clause opening onto the next, because his premise requires that nothing be left out. Dickinson writes in the clipped meter of the hymnal—common meter and ballad meter, the very forms she heard every Sunday in an Amherst pew—but she interrupts that inherited music with slant rhyme, eccentric capitalization, and the dash that has generated a century of scholarly argument. Her grammar is subtractive: she tells the truth, as she says elsewhere, by telling it slant, withholding as much as she reveals. Whitman annexes the world line by line. Dickinson interrogates a single instant until it yields more than it wants to.

Both are among my favorite poets, and I hold them there for almost opposite reasons. I love Dickinson for her intellectuality and her enigmatic lines—the sense that every poem is a locked room with the key visible but just out of reach, and that rereading her is less a pleasure than an ongoing negotiation. I love Whitman for the opposite reason: his expansive self-presentation, his celebratory tone, the way he seems constitutionally incapable of closing a door once he has opened it. I find it a little funny that I love them equally, since to love both is to love a contradiction. And yet the contradiction resolves itself, I think, at a deeper level, because both poets are, at bottom, iconoclasts of versification—perseverant and entirely certain of their own craft, unwilling to write the poems their century expected of them. They simply arrived at that certainty from opposite directions. Whitman abolished the line to make room for everyone. Dickinson fractured the line to make room for the one true thing she meant. Both are, underneath the technique, evoking the same American spirit—independence, self-reliance, the conviction that the poem answers to no authority but the poet’s own perception.

In today’s psychological terms, I would call Dickinson an introvert and Whitman an extrovert, and I do not think the anachronism does either poet a disservice. Whitman’s is a poetics of merger—he wants to absorb the crowd and be absorbed by it, to feel the “kelson of the creation” running through every body he catalogues. Dickinson’s is a poetics of withdrawal—her speaker is most herself alone, at the window, cataloguing not the world but the mind’s report of it. And this is where the two of them stop being merely a matched pair of nineteenth-century originals and start speaking, I think, for two permanent strands of American ideology: the pull toward community and the pull toward being left alone. Whitman is the poet of the “barbaric yawp,” the crowded ferry, the “I” that dilates until it contains multitudes. Dickinson is the poet of the shut door, the letter never sent, the self that insists on its right to remain illegible. America has always wanted both things at once—the town meeting and the cabin in the woods—and it is fitting that our two foreparents of poetry embody the choice so completely that neither one alone would do.

I have a more private reason for returning to Dickinson, one that has nothing to do with prosody. I lived for five years a few blocks from her homestead in Amherst, Massachusetts, now a museum, and I passed it often enough that the house stopped being a historical marker and became, in my imagination, simply inhabited. I would look up at her bedroom window and imagine her at it, looking back down at me. Sometimes I glanced into her garden and wished, in the frankly impossible way one wishes for such things, that I could travel back and know her—sit with her, even briefly, in that garden. But I suspect, as the biographical criticism consistently suggests, that any approach of mine would have sent her hurrying back inside, her black dress billowing behind her like dark contrails, the wake of a speeding jet. That, too, seems fitting. The poet who wrote “I’m Nobody” and meant it would hardly have made an exception for an admirer two centuries early or two centuries late.

I did not always make that walk alone. Often my companion past the homestead was Kathy Scherpa, one of the kindest people I have ever known, and I find now that I cherish those unhurried walks even more than the poetry itself—which is a high bar, since I love the poetry a great deal. We would pass the same fence, the same hedge, the same second-floor window, and talk of nothing urgent, which is to say we talked of everything that mattered. I did not know then to call those afternoons some of the best moments of my life. I know it now, plainly, and I suspect Dickinson—who prized nothing so highly as an afternoon rescued from the ordinary and made luminous by attention—would have approved of the accounting, even if she would not have approved of the company.

Whitman, I suspect, would have done the opposite: seen me on the path, called out, and insisted I stay for supper. That, in the end, is the whole difference between them, and the whole reason American poetry needed both.

I have written the same doorway more times than I can count. In “Copacetic” it is the doorway of my father’s hospital room, and he is at the window in the blue bathrobe my sister gave him. In Give Them Unquiet Dreams it is a doorway on Winthrop Street in Braintree, and my grandmother is standing in it with the rosary in one hand and a lit cigarette in the other. In the Aiden stories I am still writing, it is a doorway I have not yet described but that will arrive when it wants to.

For years I thought I was repeating myself—and worried about it, in the way writers are told to worry about it. I have begun, at sixty-three, to think of this less as repetition than as return. There may be a difference. You repeat what you have exhausted. You return to what you have not yet learned to see.

The grandmother in the doorway is the same grandmother, and she has always been the same grandmother. I am not putting her there each time. She is there, and I am learning to look.

I take it as a small consolation that this is not a new theory. Emerson might have called the doorway a “porch.” Chekhov might simply have gone on writing.

I read this passage from Gilead once a year, sometimes twice. It is the letter of an old minister to his young son, and I know few sentences in modern American fiction that carry as much weight so lightly:

I’ll pray, and then I’ll sleep. —Marilynne Robinson, Gilead

Five words. The last sentence of the book. What it does is refuse the summing-up that every other closing sentence in the American canon has been reaching for since Fitzgerald’s green light. Robinson trusts the reader to have understood everything. She simply names the two things her narrator is still willing to do: pray, and sleep. In that order.

I sometimes tell students that if they can end a story with two verbs and no explanation, they are already doing something most published writers never manage. Robinson can. Chekhov could. You get there by learning to trust the reader more than you trust your own summary.

People ask me, when I say I write from mercy rather than judgment, what mercy has to do with fiction. Fair question. Mercy sounds like something you extend to the deserving. In a story you have made up yourself, no one is deserving because no one is real. You could summarize a character in a sentence and be done with them. Judgment is the easier thing.

But this is what I mean. When I sit down with a character who has done something I do not admire—the father who is drinking, the mother who is failing her children, the boy who has stolen and then lied about it—I try not to have already decided what they are. I try to sit at the kitchen table with them and let them explain themselves in the way they would explain themselves to a friend at four in the morning. Not to me. To a friend. My job, if I do it well, is to have overheard what they said and set it down as accurately as I can.

Mercy in fiction is not forgiveness. It is the willingness to let a character say what they would actually say, in the words they would actually use, without the writer standing behind them with a red pencil marking them as guilty.

This is not, for the record, moral relativism. It is the opposite. It is the belief that every person, in real life, is too complicated to be reduced—and that the writer’s job is to honor that in the invented person too. If the story is good, the reader will do the judging themselves, and their judgment will be more accurate than mine could have been. I get out of the way.

I learned this, or began to learn it, from Chekhov. I try to teach it every semester. I fail at it, in my own work, all the time.

This fragment grew into the essay All We Ever Needed: The Architecture of Mercy, a full study of the short story collection built on this idea.

A passage I keep returning to, from an essay written near the end of three decades of teaching, about the classroom as communion and teaching as the pond I bathe in each morning. It is never far from my desk. The essay is here in full: “Teaching at Fifty-Six.”

I keep coming back to this passage from Trevor’s “The News from Ireland.” It is the kind of sentence that teaches you, every time you read it, what the short story is for:

There is, in his clerical opinion, no part of the world more forsaken by God, no part where charity is more thinly given, kindness more rarely shown. —William Trevor, “The News from Ireland”

What strikes me is the construction. Trevor begins with a tucked-in clause—in his clerical opinion—that identifies the speaker without naming him, and then keeps the sentence moving forward in two parallel halves whose rhythm mimics the very judgment they describe. The sentence becomes the clergyman’s mind on the page.

We are taught to admire writers for what they say. We should also admire them for how they hold the reader in a single breath.

I taught Dickinson for years. I am still teaching her, even now, by reading her in the morning before I write. This is the stanza I came back to this week:

Tell all the truth but tell it slant —
Success in Circuit lies
Too bright for our infirm Delight
The Truth’s superb surprise —Emily Dickinson

The instruction is to a poet, but I have come to think it is also an instruction to a novelist. The slant truth is the one the reader can bear; the head-on truth is the one the reader resists. Fiction’s whole technology—point of view, scene, dialogue, metaphor, the careful arrangement of what is shown and what is withheld—is in service of telling it slant.

Dickinson knew at thirty what most of us are still working out at sixty.