Why writing is how you find out what you think

1 The bottom line

Write before you know what you think, because the draft is where the thinking happens. Ideas stay vague in your head, and the page shows you the holes, the surprises and your real position.

2 Key takeaways

• Start drafting from a hunch or a question, not a finished position. Your first ideas are imprecise and incomplete until you put them in words.
  • Paul Graham25:18 says half the ideas in an essay come while writing it.
  • Kevin Kelly12:50 takes Wired assignments on things he does not yet understand, so the writing is how he finds his view.
• Use the page to find the holes your head and your talk hide. A gap you cannot fill shows up as a place where you cannot continue.
  • Ezra Klein37:30 says conversation lets you glide past a gap, while writing stops you.
  • Michael Mauboussin2:40 often thinks he understands a topic, starts writing, and finds he does not.
• Keep the creator and the judge apart in the first draft. Editing too early shuts off what lies deeper in your mind.
  • Suleika Jaouad9:45 says the temptation to edit on a computer is too great, so she drafts elsewhere.
  • Michael Dean45:22 says to write the first draft for yourself and the second for your reader.
• Let the material overturn your premise. The draft often shows the story is about something other than what you planned.
  • David Grann1:23:49 lost two years of work on one version of Killers of the Flower Moon when he found other unconnected killings.
  • David Gelb17:16 saw the Jiro film change from a sushi information film into a story about a father and son.
• Test the result by squeezing it into one page or one line. If you cannot state it simply, you do not yet understand it.
  • Sam Altman1:29:30 often thinks he is clear, writes a one-page summary, and finds he did not understand it.
  • Michael Arndt13:23 distills a story into a bumper sticker, such as "fun is more important than winning" for Little Miss Sunshine.

Half the ideas that end up in an essay will be ones that you thought of while you were writing it.

Paul Graham, Hackers and Painters

2.1 Start before you know

You do not need a position to begin. You need a hunch, a question, or something that bothers you.

• Begin from a hunch or a question you cannot answer. Paul Graham25:18 says the first words you choose are usually wrong.
  • Start from a hunch, then reread and ask: is this correct, is this complete?
  • Write into a question you do not know the answer to. For Between Two Kingdoms, the question was how to move on from illness. Suleika Jaouad27:27
  • Kevin Kelly12:50 writes one sentence, sees it makes no sense, goes back for more interviews, then writes the next. If he does not write it, he will not have the idea.
  • Daniel Pink42:30 learned from a professor that sometimes you have to write to figure it out. In an ethics class he wrote an essay against what he thought he believed, and realized he agreed with it.
• Write about what brings you up short. Confusion is a good topic.
  • List the topics that surprise you or pose a problem you cannot answer. Write about them, often first in a journal, until you see a way through. Atul Gawande33:20
  • Write about something you do not understand in your own life, then say "oh, that's what that is." Rosanne Cash10:56
  • Being Mortal came from the discomfort of treating cancer patients Gawande could not fix. Earlier writing on end of life left him unsatisfied. Atul Gawande33:20
  • Tom Junod40:11 picks subjects that haunt him, "the stuff that gnaws at me", and writes toward the resolution.
• Skip the outline if it would tell you the story too early. A map can cap what you find.
  • Dean Koontz used to sell books from outlines, but his imagination grew as each book went on and the finished novel no longer matched what the publisher bought. Dean Koontz1:04
  • Even a two-page outline would tell Lee Child the story, and he would be bored with it. Lee Child7:48
  • Koontz says he went from an okay seller to a best seller after he stopped outlining. Dean Koontz1:04
• Do not wait to be an expert. Writing is how you become one. That makes starting less scary.
  • Schools say write only what you already know, which paralyzes people. Treat writing as the process through which you learn a subject. Ana Lorena Fabrega6:50
  • Michael Arndt51:40 used to wait for an idea, then write it down. Now he believes the act of writing generates the ideas.

2.2 Get it all out, and keep the judge away

The first draft is for finding material, so collect widely and judge later.

• Spill everything onto the page before asking if it is useful. Sorting comes on the second pass.
  • Put in every story and character you might possibly use. On the second pass your conscious mind picks out what the subconscious was reaching for. Andrew Sean Greer3:24
  • Collect surprising or "crunchy" research moments as brief scenes, in no order. Then step back and ask where the story is. R.F. Kuang42:05
  • Kuang's current zero draft grew from reading about the Franco-Prussian war, German nationalism and classical music. R.F. Kuang42:05
  • Keep the tap running draft after draft. The first ideas are often recycled, like dirty water. David Perell46:54
• Draft in a way that makes the judge hard to summon. Creating and critiquing at once shuts down your depths.
  • On a computer you start polishing a sentence before you know what you are saying. Pick a drafting medium that discourages editing. Suleika Jaouad9:45
  • Write longhand while the idea is still a soup of possibilities, then type once it is clear. Andrew Hunter Murray4:25
  • Write by hand with no filter and no attempt to make it art. Then look for patterns. Robert Greene57:07
  • Murray says that when longhand works, he writes sentences he did not know were in his head. Andrew Hunter Murray4:25
  • When he wants to lash out, Chamath writes the thing he really wants to say in a Google doc and deletes it. Chamath Palihapitiya6:45
• Swing between playing and analyzing. Each mode does a different job.
  • Review your pages for structure and patterns. Then go to the other side and "just jam and just play." Andrew Stanton10:17
  • Write the first draft for yourself, chasing the material. Use structure and craft in the second draft to make it legible. Michael Dean45:22
  • Stanton rarely gets the story right in the first six drafts and is happy to hit it on the seventh or eighth. Andrew Stanton10:17
• Follow the sentences and let the characters lead. Control gets in the way of discovery.
  • Stay inside the work, follow one line to the next, and go deeper into the material instead of thinking about it from outside. Jayne Anne Phillips8:36
  • When a character says something that implies an unplanned action, trust the character, then work out where the story goes. Dean Koontz4:05
  • Elif Shafak14:02 does not know what a character will do five chapters on, and lets a side character take over.

2.3 Dig past the first pass

Your first words are often the polite version, so keep asking until they become honest and specific.

• When you are stuck, ask what you actually believe. Often you are running a social filter instead.
  • Writer's block often means you are saying what you are supposed to say or what will make others like you. Write the sentence you believe, then edit it. Stefan Sagmeister44:04
  • Write the expected account first, then what was really going on in your mind. Alain de Botton1:27:27
  • Sit alone and quiet, then note which words come up first. Do not censor them. Ada Limón15:51
  • De Botton's classroom version: everyone visits a park, and the real writing may have nothing to do with the park. Alain de Botton1:27:27
• Interview yourself in the margins. Questions push you past habitual language.
  • Beside your notes, write: what do you mean, do you have an example, give me three examples. Henrik Karlsson19:00
  • Return to a specific moment and keep asking what it was really like, then try to put words to it. Paul Millerd3:06
  • Write what happened, how you felt, and what happened next. Tucker Max41:40
  • Karlsson says that when the examples do not match the claim, he writes about why. Henrik Karlsson19:00
  • Max says writers who keep writing how they felt often realize their story is a different one. Tucker Max41:40
• Expect the truth to arrive late, and treat resistance as a signal. A bad patch can point to what matters.
  • People open with a business-card version of themselves. Your first pass is usually the persona, so keep digging. Brandon Stanton9:18
  • When pages turn bad around shameful material, ask why you are ashamed. The answer may be to cut it or to make it ridiculous. Jonathan Franzen8:49
  • Franzen floundered through six months of self-analysis on The Corrections before finding an idea that made the shameful material ridiculous. Jonathan Franzen8:49
  • When you feel a knot while writing, treat it as the place the writing will come from, and put a hand over it. David Whyte1:22:01

2.4 Let the draft find the holes and change your mind

A stalled sentence, a muddle or a surprise is the draft telling you what you have not yet worked out.

• Treat the places where you cannot continue as the holes in your thinking. Writing blocks you where talk lets you glide.
  • In writing, a hole you cannot fill shows up as a place where you cannot continue. Ezra Klein37:30
  • If you think you understand something and the draft proves otherwise, go back to first principles and poke at the weak spots. Michael Mauboussin2:40
  • When you lack an opening, it is because you do not yet know your position. Start writing, since writing forces the thought into a final form. Ben Thompson18:38
  • Klein says he very rarely stops talking in conversation for lack of preparation, but in writing it happens all the time. Ezra Klein37:30
  • After the first assembly of Jiro, Gelb saw what was missing and shot a second stage to fill the gaps. David Gelb18:24
• Ride through the confusion until the idea simplifies. A tangle usually means a missing distinction.
  • Start with a simple idea, add examples that break it, get more confused, and it compresses into a simpler formulation. Henrik Karlsson29:10
  • When a section feels jumbled, assume you are missing the right distinction, not the right wording. Keep chopping. Chris Dixon38:10
  • Chase the moment two ideas you know are true seem to conflict. Ask under what context each holds. Alex Hormozi5:50
  • After about 30 hours on one piece, Karlsson realised he had the whole framing wrong, and his opinion snapped around. Henrik Karlsson29:10
  • Dixon did about 50 iterations on the finance and social networking chapters. Chris Dixon38:10
  • Hormozi's value equation took multiple years to crystallize. Alex Hormozi5:50
• Argue against yourself. An idea you have only borrowed is dead.
  • View the idea from many sides and say where your first view is wrong. Robert Greene40:18
  • Greene says this is why his books have reversals. Robert Greene40:18
• Follow the material when it overturns your plan. Let themes grow out of it, not into it.
  • Arrive with no idea what the piece is about and stay open to research overturning your premise. David Grann1:23:49
  • If the bones are a Stegosaurus and not the promised Tyrannosaurus, have the nerve to reshape the story late. Andrew Stanton32:36
  • Do not plan themes. Write true sentences and the themes will emerge. Tucker Max54:58
  • Two years into Killers of the Flower Moon, Grann saw the book was about a culture of killing, not one evil man. David Grann1:23:49
  • Max got stuck when he was writing "Tucker Max's next memoir" instead of telling the truth. Tucker Max54:58

2.5 Test whether it is clear

Once the draft has done its work, squeeze it to check that you actually know what you think.

• Write a one-page summary or memo. If the case will not fit, the thinking is not done.
  • Write a one-page summary and see if you really understood the idea. Sam Altman1:29:30
  • Before building a deck, write a one-page memo on why the investment makes sense. If you cannot state the case simply, drop it. Sahil Bloom5:54
  • Write the hypothesis and expected return before a project, so you cannot later tell yourself you saw the result coming. Sam Corcos38:45
  • Bloom says he did this in private equity, where everyone else went straight to PowerPoint. Sahil Bloom5:54
  • Corcos's example: eight engineering weeks, a 20% rise in food logging, about $90,000 a month. Comparing written hypotheses showed one project worth $200,000 a month against another worth $90,000. Sam Corcos38:45
• Distill the piece into one line, and judge the rest by it. The line is found, not planned.
  • Force the whole story into a bumper sticker and ask whether each scene serves it. Michael Arndt13:23
  • Keep testing titles until one captures your whole thesis. Ali Abdaal40:37
  • Toy Story 3 distills to the wisdom of letting go. Michael Arndt13:23
  • Jason Fried wrote "once upon a time you used to pay for software" and realized Once was the product name. Jason Fried48:29
• Explain it to someone who arrives cold. Talk exposes jumps that you did not notice you made.
  • Teach a partner who does no homework. They sniff out logical leaps and undefined words. Tim Mackie and Jon Collins4:11
  • Talk ideas through out loud and refine them verbally. The Cultural Tutor1:18:08
  • Mackie and Collins' wilderness series came from about 10 hours of talk before a script they felt great about. Tim Mackie and Jon Collins4:11
  • Fareed Zakaria heard George Will speak with a clarity and economy you can't get when thinking something for the first time. He credits writing first. Fareed Zakaria37:46
• Write for yourself first. Understanding comes before an audience.
  • Writing for yourself comes first, small private groups second, public writing a distant third. Sam Altman12:35
  • Write only what you are trying to figure out. "Know your audience" soon turns into pandering. Morgan Housel49:42
  • Howard Marks wrote client memos to crystallize his own thinking, and they later became The Most Important Thing. Morgan Housel49:42

2.6 What the books say

Drafting shows you where your thinking is still empty, and Minto gives you a test to find those gaps and fill them.

• Replace a blank claim like "five changes" with the actual point. Barbara Minto, The Minto Pyramid Principle
  • A line such as "We recommend five changes" names the kind of idea coming but gives no idea, and it can hide thinking that is not finished.
  • Ask why these ideas and no others belong together, then write the answer as the summary.
• Push on a vague draft until the real point appears. Barbara Minto, The Minto Pyramid Principle
  • A writer said the company had "two organization problems" but could find no order in them. Pressed, he saw they were areas needing more delegation, there were four, and the real point was the company's inability to delegate authority.

2.7 Where writers disagree

Writers differ on how much to think before drafting, and on how soon to write about something painful.

• Some think it through first, then write fast. Others discover in the draft.
  • Forsyth walks around the block to get the facts and the point clear, then writes a couple of pages as fast as he can. Mark Forsyth46:47
  • Johnny Harris learned that scripting explanations on camera in the field kept him from wrestling with the story. Now he writes the narration after it settles. Johnny Harris9:45
  • Harris says a Hong Kong shoot in 2018, with all explanations recorded on camera, was "a total nightmare". Johnny Harris9:45
  • Ted Gioia alternates: when an idea hits while reading, he writes a few paragraphs, then goes back to reading. Ted Gioia1:12:17
• Some write at the heart of the confusion. Others wait for distance. The right timing depends on the material.
  • Shukman could not write Frozen Lake during the first onslaught of grief. He wrote it once he could inhabit the experience without being overwhelmed. Henry Shukman24:43
  • Gawande, by contrast, writes about problems that confuse him, often first in a journal, until he finds a way through. Atul Gawande33:20
  • Wordsworth wrote about Snowdon seven or eight years after the climb. Henry Shukman24:43
• Writing can also talk you into a simpler view than the evidence supports. Discovery is not the same as being right.
  • When chiseling to one best argument, the person you persuade most is yourself. Make sure the formalism does not push you into a simpler reality. Ezra Klein51:10
  • Klein says on a podcast he can hold ideas in tension and leave them unresolved, which he cannot easily do in a column. Ezra Klein51:10