When I first started writing for a real publication, I taught myself “reporting” with a simple self-made curriculum unfolding over six or seven articles. The first two pieces I wrote from my head, with reference to things I already knew or to books I’d read. For the third, I actually got out of the house, but didn’t yet have to play the journalist; I just wrote about taking a flying lesson in a small airplane. The fourth article required more gumption: I decided to shadow a friend of mine for a day while he did his job as a derivatives trader. I’m not sure how he got me in the door.
Real reporting involves talking to strangers. For my fifth article I did a single phone interview with someone I’d never met. That wasn’t so bad. For the seventh article, the real leap, I shadowed someone I didn’t know—my old driving instructor. I’d only met him briefly many years before. I asked to go along for a ride in his teaching car—two steering wheels, two sets of pedals. Later, we pulled over and chatted about his life and work. I had not gone to school for this. But I liked the advice I’d gotten from a journeyman reporter. He said, if you tell someone you’re a journalist they’re going to believe you. Your job is to honor their trust.
This was most of what I knew about nonfiction writing when I managed to land an assignment, on spec, to profile Douglas Hofstadter for a piece in the Atlantic’s print magazine. That felt like a big break. But I also wasn’t quite sure what I was supposed to do.
I ended up relying on a very short user’s manual I’d discovered in the The John McPhee Reader, a book of collected journalism from the New Yorker writer John McPhee. In the introduction, William L. Howarth, who edited the collection, described McPhee’s method for producing what the New Yorker called “fact pieces,” or deeply reported nonfiction. I liked the sound of the method, and I liked the products of it. So I just did my best to copy what Howarth said McPhee did. It’s basically the process I’ve used ever since. The method is not that hard to describe and it’s so useful that I think it bears broadcasting. In fact I think those two or three pages from Howarth’s introduction are a decent substitute for journalism school, at one one-hundred-thousandth the price.
In brief, McPhee’s idea is to never face a blank page. Instead, in stage one he accumulates notes; in stage two he selects them; in stage three he structures them; and in stage four he writes. By the time he is crafting sentences the structure of the piece as a whole, and of each section, even paragraph, and the logic connecting them all, is already determined, thanks to the mechanical work done in the first three stages. McPhee is on rails the whole time he writes his first draft. From there it’s all downhill and the standard thing that everybody does: revision, revision again, then refinement—a sculptor with ax, then knife, then scalpel.
Stage 1: Gathering notes
A simple but important question I had going into that first big assignment was when do you start writing? Should you do a little reporting, then write a little, then fill in the holes with more reporting? Or should you do all the reporting up front? (By “reporting” I just mean reading, doing research, calling people, going places, spending time with people in person.)
In the McPhee method, you do all the reporting up front. McPhee usually had one person at the center of each piece, so he would aim to spend a lot of time with that person. He’d go on long backpacking trips with them, or stay at their cottage for a season, or drive across the country with them. He’d immerse himself in their lives for months. And along the way he’d talk to their family, their friends, coworkers, rivals, other people in the same field—to say nothing of all the calls or visits he’d make to experts who could weigh in on this or that.
You can get a sense for what deep reporting looks like by reading one of McPhee’s articles. In “A Sense of Where You Are,” McPhee profiled Bill Bradley, then a wunderkind basketball star at Princeton. (Later, a U.S. senator.) In one set piece, Bradley is practicing jump shots in a gym somewhere.
Last summer, the floor of the Princeton gym was being resurfaced, so Bradley had to put in several practice sessions at the Lawrenceville School. His first afternoon at Lawrenceville, he began by shooting fourteen-foot jump shots from the right side. He got off to a bad start, and he kept missing them. Six in a row hit the back rim of the basket and bounced out. He stopped, looking discomfited, and seemed to be making an adjustment in his mind. Then he went up for another jump shot from the same spot and hit it cleanly. Four more shots went in without a miss, and then he paused and said, “You want to know something? That basket is about an inch and a half low.” Some weeks later, I went back to Lawrenceville with a steel tape, borrowed a stepladder, and measured the height of the basket. It was nine feet ten and seven-eighths inches above the floor, or one and one-eighth inches too low.
Another paragraph discusses Bradley’s uncanny vision on court—hence, “a sense of where you are”—as if Bradley could see “out the back of his head.”
I asked Bradley to go with me to the office of Dr. Henry Abrams, a Princeton ophthalmologist, who had agreed to measure Bradley’s total field. Bradley rested his chin in the middle of a device called a perimeter, and Dr. Abrams began asking when he could see a small white dot as it was slowly brought around from behind him, from above, from below, and from either side. To make sure that Bradley wasn’t, in effect, throwing hope passes, Dr. Abrams checked each point three times before plotting it on a chart. … When he finished plotting Bradley’s circles, the one for each eye was larger than the printed model and, in fact, ran completely outside it. With both eyes open and looking straight ahead, Bradley sees a hundred and ninety-five degrees on the horizontal and about seventy degrees straight down, or about fifteen and five degrees more, respectively, than what is officially considered perfection.
That is what you call access. Fundamentally, McPhee’s writing is good because it is filled with good facts. Good facts are rare, so you must give yourself enough time to acquire them. You have to get in close—build rapport with people, and watch them in their element. McPhee is not sitting across a table with his subjects, conducting an interview. He’s participating.
How long are you supposed to do that, before you come home and move on to Stage 2? When do you stop calling the next friend or expert? McPhee says, you stop when you start hearing the same stories for the third time.
*
The product of all this is notes. You write notes to yourself on reporter pads or notebooks you buy at the pharmacy; you highlight sentences in books, and type those up; you transcribe tape [1] and throw that into the file. Importantly, you don’t do much organizing. All I do is append my notes to a single giant Markdown file under a heading that says which book or article or interview or paper notebook they came from. A gem of a quote, a gem of a thought that will make a fine sentence, is intentionally cast into the same pile that includes arcana you’re probably going to discard.
Already in this stage you are deciding what’s important—literally, what’s noteworthy. I love that about writing. The beauty of having a piece of writing you’re working on is that it changes the way you think. You can’t help but relate whatever you’re doing to the work in progress. I wrote about this once, in More people should write:
When I have a piece of writing in mind, what I have, in fact, is a mental bucket: an attractor for and generator of thought. It’s like a thematic gravity well, a magnet for what would otherwise be a mess of iron filings. I’ll read books differently and listen differently in conversations. In particular I’ll remember everything better; everything will mean more to me. That’s because everything I perceive will unconsciously engage on its way in with the substance of my preoccupation. A preoccupation, in that sense, is a hell of a useful thing for a mind.
This is why it’s so useful to work on an article for a long time. If you’re reporting on something for six months, even if the really concentrated part, the key visit, is only a week or two of that, you have time for notes to accumulate. To give a sense of scale, for a profile I wrote of two Google programmers my notes file had 190,000 words; for an article about COVID and the immune system, the file had 109,000 words. (Granted, most of these words were not notes to self, but rather tape, excerpts from books and articles, and so on. Still: 190,000 words is something like four hundred pages.)
Stages 2 and 3: Bucketing, Structuring
When you get home and decide you’re done reporting, you must take stock of what you have. McPhee typed up all his handwritten notes and put them into a binder. As he read the entries, new ideas would suggest themselves, or there’d be some new research to do, and so he’d type up whatever new notes resulted. The binder would grow a bit, then settle. (At this stage McPhee would sometimes take a stab at the “lead,” that crucial first section, a page or two long, that sets up everything else. But that’s all the writing he’d do from his head, before turning his attention back to the binder of notes.)
Having re-absorbed in a few sittings everything he’d thought and written down about his topic, various large-scale structural ideas would occur to him. He’d have a sense for sections, scenes, and set pieces. As he went, he’d come up with little acronymic codes for these, and write those down on note cards. (For me, this process of deciding what the scenes and sections should be is both top-down and bottom-up: top-down, because sometimes a section you already want to include determines how you perceive the notes that relate to it; and bottom-up, because sometimes you might not realize a section needs to exist until you see it emerging from the notes.)
Having note cards standing for each structural unit makes it easy to consider different arrangements. Do you have to know about X before you can discuss Y? Yes, but you really want Y to come early, because it’s maybe the most fun section; and A and Y have to be next to each other—there’s a great juxtaposition there—keeping in mind that you sort of want to interleave scenes and exposition, so C and A need to be buffered by B… In other words you have a little constraint satisfaction problem. McPhee would solve it by moving his section-cards around on a table.
Do you see how clerical the process is? That is by design. Writing is extremely hard work, easy to procrastinate; the genius of the McPhee method is that it distributes your thinking and decision-making over time so that it rarely feels hard. I can tell you from experience that while “writing a first draft” is intimidating, “reading through all your notes” or moving note cards around on a table, contemplating structure, is not. In fact these tasks are kind of delightful.
*
McPhee would then go through his binder and mark individual notes as belonging to one thematic bucket or another, writing the corresponding shortcode in pencil next to it. Notes that weren’t worth keeping wouldn’t get a code. “Writing is selection,” McPhee likes to say. Which of your notes are worth including? Your writing can only be as good as your taste. As Howarth writes:
McPhee has a passion for details, for they convince readers that he deals in actualities. Added to his journalist’s reverence for facts is a novelist’s propensity for symbols. His task is to burnish objects until they become reflectors of character and theme. Instead of sermonizing on thrift or prodigality, he notes that Donald Gibbie’s teapot is plugged with fourteen wood screws, or that the light in Lt. Arthur Ashe’s closet at West Point is always burning.
Having labeled each note with an acronymic code, McPhee would literally go through his big binder with a pair of scissors, cutting out all the coded notes, sorting them into file folders. The process is recursive. That is, for each file folder he’d go through the same exercise, if a little less rigorously: re-read the notes, feel out the sub-structural beats, make notecards for each of those, arrange them into a logical sequence, then label and file the notes once more.
If you do this, what you end up with is at least two levels of structure: the major sections of the piece, and for each, the flow of ideas through that section. When you go to write, everything you want to say and all the material that supports your saying it is already laid out.
If the best New Yorker fact pieces feel rich, it is maybe because they are grown bottom-up from a huge repository of notes, sifted and arranged for impact. Compare the way one might naively think writing happens: perhaps the author sits at their keyboard with an impulse to say something, and draws on the facts, metaphors, and anecdotes that come to mind. People do write that way, but it’s not what you do if you’re after the thing that McPhee is famous for. If you’re a curious reporter diving into worlds you don’t know—oranges; nuclear science; geology; the life of a long-haul trucker—you first have to go out there with some notebooks and soak, and soak, until you’ve soaked so much that you feel like you get it—and your task then is to rebuild for the reader, in compressed form, all of what you saw and learned to get you to that point. I don’t know how to do that except by building the piece out of the notes themselves.
Interlude: Organizing notes digitally
Lest you think that McPhee is a product of a bygone era, working with paper, binders, scissors and pen, he actually switched early to computers. The software he has been using since the mid-1980s is in fact more sophisticated and tailored to his needs than that of almost any writer working today.
He described his setup in detail in a piece called “Structure.” A friend of McPhee’s, named Howard Strauss, in the Princeton I.T. department, customized a text editor called Kedit for him. One sub-program was called Structur:
Structur exploded my notes. It read the codes by which each note was given a destination or destinations (including the dustbin). It created and named as many new Kedit files as there were codes, and, of course, it preserved intact the original set. In my first I.B.M. computer, Structur took about four minutes to sift and separate fifty thousand words. My first computer cost five thousand dollars. I called it a five-thousand-dollar pair of scissors. … Some of those files created by Structur could be quite long. So each one in turn needed sorting on its own, and sometimes fell into largish parts that needed even more sorting… So Howard wrote Alpha. Alpha implodes the notes it works on. It doesn’t create anything new. It reads codes and then churns a file internally, organizing it in segments in the order in which they are meant to contribute to the writing.
Hovering over a note with Structur, McPhee could type a series of keys and the note under his cursor would get assigned to a bucket, sent instantly to a separate file; another key would mint a new bucket, or change its shortcode. Then, turning to Alpha, notes that had been filed into separate buckets could be re-integrated into something like an outline. (At least that’s how I imagine it works.)
For the longest time, when I worked my own notes using the McPhee method, I would do all the bucketing via cut-and-paste. I used a program called Textmate on my Mac. From one big Markdown file I would cut notes and paste them into sub-files according to topic or scene. (A pale imitation of Structur.) Working inside one of the sub-files, I’d arrange notes within Markdown headings. Keyboard shortcuts inside Textmate helped, but the process was still tedious. (A pale imitation of Alpha.) In about 2021 I found myself lamenting that John McPhee, then already ninety years old, had a more sophisticated setup for managing his notes than I did. I’m a programmer, for God’s sake.
I ended up landing on a solution that uses org-mode inside Emacs. I call it “McPhee mode”. When I turn it on with a keyboard shortcut, my entire file of notes is optimized for refiling into buckets. Each bucket gets assigned a shortcode, corresponding to a keybinding—as simple as the letter A or l , or more complicated if I have lots of buckets. Whenever I see a line or excerpt I want to file away, I need only hit the corresponding keys to have the note slurped into a bucket. As notes are filed they get highlighted. Each note keeps track of where it came from. Doing this within a single large file makes undo/redo work automatically. The tool has made the filing process not just tolerable but even sort of pleasant.
Your browser doesn't support video playback.
A short screencast of my “McPhee mode” Emacs refiling system in action.
Stage 4: Drafting
Outlined in this fashion, McPhee’s writing methods may seem excessively mechanical, almost programmatic in his sorting and retrieval of data bits. But the main purpose of this routine is at once practical and aesthetic: it runs a line of order through the chaos of his notes and files, leaving him free to write on a given parcel of work at a given time. The other sections cannot come crowding in to clutter his desk and mind; he is spared that confusion by the structure of his work, by an ordained plan that cannot come tumbling down. The strategy locks him in, gives him no easy exits from the materials at hand, which he must confront with that humorless partner, the typewriter. –Howarth, The John McPhee Reader
The main advantage of the McPhee method is that so much of the hardest work is done up front. Writing can proceed without interruption. I dread but then always enjoy the intense period where I emit a first draft. It requires gearing up for. But thanks to the McPhee method, once I start drafting I’m usually at it for just five or six nights. It takes me about five or six hours each night to write a section, plowing through the notes that have been scrupulously prearranged.
I should say that in those nightly encounters I rarely go straight from notes to full sentences. I bridge the gap with a kind of pseudowriting. In effect it’s an outline that takes the form of paragraphs, mixed in with the raw notes. I write good full sentences, broken half-sentences, non-sentences that also include todos in them. I use a lot of square brackets. Suppose I’m working on a scene set at Alta, Utah, among the “snow safety” patrollers, in an article about avalanches. At the pseudowriting stage, the first paragraph might look like this:
One night earlier this winter, the only road out of Alta, Utah, was closed down. [warning signs inside hotels]… That season, there had already been [n] feet of snow… I had flown in that morning after being told that a storm had dumped… Our lodge, like the others… [concrete buildings like bunkers; 3/4 of all buildings in avalanche path!]. [the most serious avalanche hazard in the country. highway 210 has highest AHI in the country - 1045. at 40 you already require avalanche controller] … then cut to ski patrol
Underneath this mess, all the relevant notes are on hand. Writing this way you are not staring into the abyss. You’ve got good facts; they’re in a good order. Your only job at that point is to turn them into something that’s fun to read:
One night earlier this winter, the only road out of Alta, Utah, was closed down. At ski lodges, signs warned guests to stay inside or face fines. Already that season, twenty-two feet of snow had fallen, and, the day before, a storm had dropped thirty-three inches; another foot was predicted by morning. The most dangerous time for avalanches is after a rapid snowfall, and three-quarters of the buildings in Alta are threatened by a known avalanche path. A standard measure for danger on roads, the Avalanche Hazard Index, computes risk according to the size and frequency of avalanches and the number of vehicles that are exposed to them. An A.H.I. of 10 is considered moderate; at 40, the road requires the attention of a full-time avalanche forecaster. State Highway 210, which runs down the mountain to Salt Lake City, if left unprotected, would have an A.H.I. of 1,045. Just before 5 A.M., a small group of ski patrollers gathered at a base by the resort’s main lift. …
So much of writing is managing your own emotions. The virtue of “pseudowriting” is that it helps you preserve hope for as long as possible—hope that what you will eventually put in place of those square brackets will be good. Hope keeps you coming back: it is more pleasant and low-stakes to pseudowrite than to fix actual language into the draft; and it is less daunting to return to a document where it feels like all that’s left is for you to fill in some blanks. Do that enough times and you will, in fact, end up with something you can read top to bottom.
The rest of the process is the stuff everyone already does. “All writing is rewriting.” You revise, you check your facts, you do more reporting, you restructure, you revise, you compress. But god is this easier once that first draft is in hand. The McPhee method offers a reliable procedure for getting there.
If you’re interested in the McPhee method, you should start by reading William L. Howarth’s introduction to The John McPhee Reader, then the McPhee articles themselves, and finally McPhee’s own Draft No. 4: On the Writing Process.
[1] A word about tape: McPhee “never uses tape recorders when interviewing, for they inhibit some people and are too subverbal for his purposes,” according to Howarth. “The writing process must begin with words—a scrap of talk, bits of description, odd facts and inferences—and only a pencil and notebook will answer those needs with literacy and economy.” It’s true that the brain is the best filter, because what sticks with you after a conversation is often what’s most interesting, both to you and to the eventual reader. If you pay too much attention to the tape, and not to your own impressions, you might actually give the reader a worse sense of what it’s like to be around a person.
Even so, conversations move fast. Notes help. I personally can’t listen well and take good notes at the same time. I’ve settled on a routine in which I record most conversations, but write highlights and impressions as I go and just after I finish. (I recall McPhee saying somewhere he made an exception for highly technical conversations, especially among multiple people. A lot of what I write about is technical, and so tape is my friend.)