Athens

Atul Gawande's Advice on Writing: 22 Rewrites, 30 Hours a Month, and Three Hearts

- Moritz Wallawitsch

Atul Gawande is a surgeon, public health researcher, and staff writer at The New Yorker. He wrote Being Mortal, The Checklist Manifesto, Complications, and Better. He did not start as a writer. He taught himself every part of it: research, structure, sentences, stories. While operating on patients.

His advice comes from David Perell's How I Write podcast, a Boston Globe profile, and the full YouTube transcript of his conversation with Perell.

22 Rewrites

Gawande did not come to writing naturally. He came to medicine naturally. Writing was something he taught himself from scratch. In 1996, his friend Jake Weisberg started Slate magazine - "Netscape Navigator days" - and could not get paid journalists to write for an internet shot in the dark. So he asked friends. Gawande's first piece was a diary from surgical residency, essentially a blog before the word existed.

Weisberg was his first editor. He would hand pieces back and say, "This is what you're doing well. This is what you're not doing well. Do more of the first, do less of the second." Then: "Here, rewrite it." Gawande had never rewritten anything in college. "Who rewrites a thing?"

His first New Yorker piece required 22 back-and-forths with his editor Henry Finder. Five complete rewrites. Nine months for 4,000 words. Every round filled him with anxiety. But every round taught him something that made it significantly better, so he could not object.

Now he needs three or four revisions per piece. He went from dreading revision to loving it. "The first draft is always painful, but the revisions, I know that will make it better."

30 Hours a Month

I set a goal of 30 hours a month, like an average of an hour a day. That felt doable.

Gawande writes between surgical cases. On weekends stolen from family. In 30-minute increments, eight of them minimum on writing days. He does not have a writer's schedule. He has a surgeon's schedule with writing wedged into every gap. The lesson: you do not need perfect conditions. You need consistency.

The Well Dries Up

Gawande maintains surgery, policy work, and writing simultaneously. He believes abandoning medicine would kill his writing. "The well will dry up. What will I have to say if I'm not out there?"

This is the opposite of the "quit your job to write" advice. Gawande's writing is fueled by doing the work he writes about. Without patients, without operating rooms, without the confusion of real medicine, the material vanishes. Experience feeds the page.

Two Types of Ideas

Gawande maintains a list of 400+ potential article angles in Apple Notes. Most are incomplete. He identifies two types:

  • A cool story without larger meaning. Incomplete.
  • An abstract concept without a narrative vehicle. Incomplete.

His piece "The Itch" started as a medical case report about a woman who had an itch so severe on her forehead that she scratched through her skull and injured her own brain. It went on his list as a cool story. Only later, when he connected it to ideas about sensation and how the brain works, did it become a publishable piece. One of his goals for that essay: "I wanted to make you itchy." That is the moment a cool story meets a larger frame.

Another example: a piece about hospital handwashing. Two million people picking up infections because someone does not wash their hands. The abstract concept was clear. The vehicle was following the woman at his hospital whose entire job was infection control. "She's like, I didn't go to medical school to be the handwashing police." Then Purell came out, changing her life. Now the story had turns.

Three Hearts of Every Piece

Every piece of writing needs three hearts beating at once:

  • An intellectual heart. The idea, the argument, the thing you learned.
  • A narrative heart. The story structure that carries the reader forward.
  • An emotional heart. Why anyone should care.

If any heart stops, the piece dies. His book Being Mortal found its emotional heart when his book club dinner delivered universal feedback: cut the nursing home history by half, give us more of your father. Gawande had resisted. He thought the family story was too indulgent. "People don't want to hear that. Let me teach you something." Of course it was the opposite.

The father's story became the spine of the book. Gawande's father, a surgeon with a brain tumor, was asked what minimum quality of life he would find acceptable. One patient had answered: "If I can watch football on television and eat chocolate ice cream, that would be good enough for me." Gawande's father said: "No way is that good enough for me." And they were off. His father's priorities evolved as his capacities declined - from operating, to connecting with people through the Rotary Club, to something smaller still. But at each stage, there was something that made life worth living. That is the emotional heart of the book.

The Three-Question Editing Checklist

For every sentence, Gawande asks three questions:

  1. "Can I make it visual?"
  2. "Am I conveying a new piece of information with each sentence?"
  3. "Is each sentence necessary?"

This is Klinkenborg's principle made surgical. Every sentence must earn its place. Every sentence must show something. Every sentence must advance understanding.

The Steadicam Technique

His New Yorker editor Henry Finder advised: "Imagine you're giving me the Steadicam. You need to show what's happening."

Gawande's first New Yorker piece was about automation in medicine - a computer program that could diagnose heart attacks from an EKG. He was writing about compelling data but was not showing what an EKG looks like. What do you see? What does a doctor see that you do not? What does the machine see that even the cardiologist misses? Finder walked him through it like a writing class.

This is also where Gawande's early weaknesses showed. His Slate-era metaphors were overrought and clashing. He learned to let the material speak for itself. A passage about teeth aging - enamel wearing away, gums pulling back, the appearance elongating - works because "the teeth could tell the story by themselves. I didn't need to say it was like a picket fence or gravestones with a headstone popped over."

Say Yes Before 40, No After 40

Before age 40, try everything. You do not know what energizes you. Gawande worked on Capitol Hill, played in a band, worked in labs, started writing for his friend's internet magazine, and walked into an operating room and was blown away. None of it made sense together until a theme emerged: seeing humanity through the lens of health.

After 40, commit to what you have discovered works. "If it gives me energy, I want to do more of it, and if it exhausts me, I do less of it." Now at 60, he has had three decades of obsession with the same subject from different angles. The books look scattered. They are not. Surgery, public health, writing - they are all illuminating the same thing.

Story Shapes: O's and W's

Finder would literally talk about the shape of each piece. "This one is an O" - meaning it starts where it ends. "This one is a W" - three peaks, advancing from left to right. Gawande learned to vary structure deliberately. In surgery, you try to do the same thing every time, like sinking free throws. In writing, you are constantly trying to do something new. "It gets harder as time goes on because you use up your tricks."

The Book Club Feedback Method

At 75% completion of a book, Gawande invites five or six writer friends to dinner. He gives them the manuscript. They discuss it over food. The clay is still a little wet - he is willing to tear it up. This is better than sending it out to everybody and hearing from nobody. The dinner provides deadline pressure, prevents over-commitment to weak sections, and reveals the emotional core versus extraneous material.

On AI and Editing

For his current book, Gawande needed 19th-century firsthand accounts of surgeons experiencing the spread of anesthesia. He deployed two or three different AI models. "Get me to 19th-century firsthand accounts of surgeons." It was Google on steroids. Three months of library research collapsed to two weeks. But the AI made up quotes. "Never trust AI quotes." The real accounts, once found, were not as neat and clean as what the AI fabricated - but they were gold.

For copy editing, AI effectively identifies "where is the fat." For cutting a 4,000-word talk down to 2,000 words, Gawande tried three different platforms. GPT gave him 500 words. He asked for 2,000. It apologized, gave him 1,000. Never hit the target. But he ended up with six versions, and every single one killed his opening anecdote. He had to concede: "It was a cute story, but it wasn't really working or necessary."

The lesson: AI is useful for research and line-level trimming. It is not useful for structural decisions. It cannot tell which anecdote matters. Only the writer knows that.

Key Takeaways

  • Revision transforms writing. Gawande went from 22 rewrites to loving the process.
  • 30 hours a month in 30-minute increments. Consistency beats perfection.
  • Keep doing the work you write about. The well dries up otherwise.
  • Every piece needs three hearts: intellectual, narrative, emotional.
  • Three questions per sentence: visual? new information? necessary?
  • Show with the Steadicam. Do not state. Show.
  • 400+ ideas on a list. Most are incomplete until story meets concept.
  • Book club feedback at 75%. Friends, food, honest critique.
  • AI finds research and fat. It cannot find your emotional core.

Gawande proves that writing is not a talent you are born with. It is a discipline you build, one 30-minute increment at a time. AI can help you trim the fat once you have done the hard work of finding the three hearts.

This post draws from Gawande's appearance on How I Write, the full YouTube transcript, and the Boston Globe profile of his writing routine. Athens is an AI writing editor that helps you apply Gawande's three-question checklist to every sentence.