You get excited about an idea. You sit down. You start typing. The thing you produce has the right shape but the wrong weight — it's not grounded in you yet, it's a thought-balloon, and three weeks later you reread it and don't recognize whose voice that was.1 Karlsson noticed this looking back at his own pieces: the ones he loved had something the others didn't. He'd been carrying them for a year. He'd read books about them, had conversations about them, lived around them long enough that they stopped being ideas and became part of how he saw. The 5-step method is a process for letting writing accumulate that way — not a workflow for producing essays faster, but a workflow for not writing the ones that aren't ready yet.
Karlsson lays out five steps, but they're not stages in a production line. Each one is a different relationship to the material. The point isn't to move efficiently through them; it's to know which one you're in.
Step 1 — Exploring. You read, talk, journal, notice. No piece yet. You're not even trying to write — you're trying to live with the topic long enough that it starts speaking to you. Karlsson gives this a year, sometimes more. The signal that Step 1 is done is the arrival of an animating question — a question with personal stakes, something that, if answered, would change how he is in his life. "How could this help me become a better father." Not "what is the truth about parenting" but "what do I, specifically, need to understand here so I can show up for my kids."1 When that question lands, the material is ready.
Step 2 — Vantage Point. Now the animating question turns on like a magnet, and all the accumulated notes start lining up toward it. Karlsson describes this as suddenly seeing the landscape — he can tell where he needs to go, which stories will help, which arguments belong, which images carry weight. The material starts to shape itself, not because he's organizing it but because the question is pulling it.1
Step 3 — Clusters and Sequencing. Out of the lining-up, groups form. Three or four ideas turn out to belong together as a unit. Karlsson thinks of essays as "a series of centers" — geometrical objects with energy fields between them. He lies on the sofa and pictures the clusters as having colors and feelings. He moves them around in his head, asks where the energy releases when they snap into the right order. That's the right order. Now they snapped into place.1
Step 4 — Writing the Core. Once the sequencing snaps, type fast. Karlsson sometimes lies on the sofa with voice transcription and talks the essay out three or four times until it crystallizes. The goal at this stage is not to write well — it's to capture the body-felt energy gradient before it dissipates. If he can laugh and cry while writing it, that's what reaches the reader. He'll edit later, but he wants the first draft to be a coherent whole, not a patchwork of separate fragments stitched together.1
Step 5 — Editing. Edit through it. And — this is the move that pulls the method together — write the beginning and the ending last. When Karlsson tries to write an opening first, he gets pretentious, tense, overworked. Two or three hours go into a paragraph that doesn't hold. But once the piece exists, the opening is obvious — usually a throwaway line from somewhere in the middle that, when moved to the top, makes the whole thing cohere. In Looking for Alice, the opening turned out to be a throwaway line about Gertrude Stein answering "no, I just like Alice" when asked if she was a lesbian. He didn't write that line as an opening. He moved it there once he understood what the piece was actually about.1
The method gets its power from what each step won't do.
Step 1 refuses to write. You can take notes, but you can't draft. The whole point of the year is to let the thing become yours, and writing-too-early collapses that into a thought you've already had.
Step 2 refuses to start composing. You're seeing the landscape — not walking through it yet. The magnetic field has to do its work without you forcing the shape.
Step 3 refuses to be linear. Essays happen linearly in the final form, but the clusters in your head are not linear, they're geometric. Karlsson actually moves them spatially. The ordering is felt before it's typed.
Step 4 refuses to edit. You're not writing for elegance; you're writing for transmission of the body-state. Edit after you've captured it. If you stop to fix a sentence, the energy gradient breaks.
Step 5 refuses to write the opening from scratch. The opening reveals itself once the piece is whole. Force it earlier and you produce the pretentious paragraph that won't hold.
Late in the interview Karlsson names what unifies the whole method: "My toolkit is a series of exercises for breaking my own mental models."1 Every move he makes — the year of incubation, the self-interview in the margins, the case studies, the cluster-sequencing on the sofa, the no-words essay-thinking experiment — is structured to prevent him from settling into a comfortable, ready-made way of seeing the topic. The animating question keeps him in confusion long enough that something true can come through. Most writing protocols try to reduce uncertainty so you can produce text. Karlsson's protocol sustains uncertainty so the text that emerges is from below the surface he started with.
The reader gets the result of that work without doing it: the writing carries the body-state of a person who has been sitting with the question for a year, who has been broken open by it. That's why Karlsson types fast at Step 4 — because the body-state is what he's trying to transmit, and the body-state lives in momentum.
Single-source: Karlsson's own retrospective look at his pieces (lines 22-32) identifies the year-long incubation as the common factor in his best work. He gives examples: Looking for Alice (the Gertrude Stein opening that emerged from a throwaway middle line, lines 106-108), the favorite books were not meant to be published piece (a small-observation entry that he didn't promise anything in but ended up covering education, cognition, and "all these different areas" — lines 244-250). The method is named retrospectively, not prescriptively — Karlsson is describing what he found, not advocating it as universal.
The method assumes you can afford a year per piece. Karlsson works full-time as a writer now, lives on a Danish island, and has structured his life around long-incubation. For writers without that structure, Steps 1-3 either compress (with quality loss) or never complete (and Step 4 becomes the thought-balloon Karlsson started by warning against).
The method is also psychologically demanding. Step 1 doesn't produce visible output. For writers paid per piece or driven by metrics, a year of "no output" is professionally untenable. Karlsson admits he sometimes tries to "force himself to have an idea" because he writes professionally now — but the forced version "is never the same thing." The method's discipline is not effort; it's restraint.
Karlsson is the fifth practitioner methodology in the vault's non-fiction-writing-craft hub, and the most interesting comparisons are against the four already filed: Wang, Pink, Deen, Orlean.
Wang and Karlsson both put years into pieces, but they're not loading the same thing. Wang spends a year cultivating voice — manuscript-copying Stendhal, Mahler, Beethoven; building an inheritance of solved structural problems he can deploy later. Karlsson spends a year living with the topic — reading books about it, having conversations, letting it become alive in him. Wang's year is craft-time; Karlsson's year is residence-time. Both produce essays nobody else could write, but for different reasons: Wang's pieces are unrepeatable because his voice is uncopiable; Karlsson's are unrepeatable because the topic has been metabolized into who he is.
Pink and Karlsson both have routines, but Pink's 500-800-word daily quota is a discipline of consistency and Karlsson's 7am-3pm-six-days-a-week window is a discipline of integration. Pink rigidly walls writing off from the rest of life (isolated office, no phone, no email — rigidity as liberation). Karlsson lets the writing leak into the rest of life: if he's stuck, he fixes the brick wall; if he's tired, he hangs with his kids; "the writing feeds into being a better dad, and being with the kids gives me ideas for the writing."1 Pink's model is the writer-as-engineer optimizing throughput. Karlsson's model is the writer-as-householder for whom the writing is one organ in a larger body.
Orlean and Karlsson share the most structurally — both run three-phase architectures with long research, long thinking, fast(er) writing. But Orlean's Phase 3 is mechanical (1000 words a day, every day, like a marathon at mile pace). Karlsson's Step 4 is the opposite: he waits for the sequencing to snap, then types fast in one burst to capture the energy. Orlean trusts the daily-quota to produce coherence over weeks; Karlsson distrusts daily-quota because it loses the body-state. The tension reveals something about what Phase 3 is for: if Phase 1-2 left you with a clearly-shaped piece (Orlean), Phase 3 is execution and daily-quota works; if Phase 1-2 left you with a body-feeling that has to be transmitted (Karlsson), Phase 3 has to ride the feeling and daily-quota would dissipate it.
Deen and Karlsson disagree most sharply on whether structure is taught or felt. Deen's 27-pattern taxonomy is a craft-engineering apparatus — name the patterns, practice them, deploy consciously. Karlsson's clusters-as-geometrical-objects are felt in the body, not named. Both produce structurally-coherent essays. The split tells you the structure question may be terminal — you can reach craft from either pole (engineered-patterns or body-felt-fields), but the routes don't translate. Trying to teach Karlsson's method as patterns flattens it; trying to teach Deen's method as body-feeling loses the precision.
The 5-step method is not just a writing process — it's a description of how a mind makes something its own, and you find that description echoing across several domains once you start looking.
Psychology: Integration as Prerequisite for Expression — the year-long incubation is a structural cousin of Jungian integration: the topic doesn't stay an idea but gets absorbed into how the self sees, and only then can it be transmitted. Pick up an essay written too early. You can feel the gap between the writer and the material — the prose is brittle in a particular way, and when you press on it, it breaks. Now pick up an essay written after the material has lived in the writer for a year. The same press lands on something dense; the prose doesn't crumble. Writing-too-early isn't a craft mistake. It's a developmental mistake. The reader's hand knows the difference even when their analytical mind doesn't.
Eastern-spirituality: Karmas and Samskaras — the year-long incubation is also a samskara-formation process: the topic carves grooves in the mind through repeated exposure, conversation, journaling, and lived application until it becomes a habit of perception rather than a thought you have to summon. Watch the writer who has incubated a topic for two years return to it after five. They pick up where they left off without effort. The grooves are still there; the perception still operates. Watch the writer who wrote one essay from an idea they hadn't lived try to write a second on the same topic. They have to reconstruct everything from notes, and the new essay reads as if a different person wrote it. Samskara-grade material gives you continuous work across a career; idea-grade material gives you one-off pieces with no second act.
Behavioral-mechanics: Elicitation Framework — Step 3's cluster-sequencing-as-reader-state-induction is the same mechanism Hughes uses in elicitation: control the state of the interlocutor through sequence, and the content lands differently. Karlsson is doing elicitation on the reader — sequencing his clusters to produce a specific emotional/cognitive state before the key insight arrives. The Berger Van Gogh example demonstrates this directly: you look at a painting, you read "this was the last painting Van Gogh painted before he killed himself," then you look again and see the painting completely differently. Same content. Different sequence. Different person looking back. Essay-craft and influence-engineering share the operating principle — state precedes interpretation. Karlsson summons with it; Hughes extracts with it. Same lever, opposite directions.
History: Strategic Patience and Long Incubation as Leadership Trait — Karlsson at essay scale, Bismarck at geopolitical scale, both run the same operation. Bismarck spent years observing the European balance of power before initiating any of his major moves. Eisenhower's long telegram analysis preceded his containment doctrine by months of sustained reading. Now picture Bismarck at his desk in 1860, refusing to commit to a Schleswig-Holstein move because the material hadn't formed a body-felt picture yet. You can feel the same somatic discipline Karlsson describes — knowing-when-to-wait, knowing-when-the-snap-has-arrived, refusing the premature output that the deadline-mind would force. Premature output is the universal failure mode of long-form work, whether the work is an essay or a foreign policy. Writers and statesmen breathe the same air on this one. Karlsson's restraint at Step 1 isn't an artistic temperament; it's the same operational discipline that produces strategic depth across high-stakes fields.
The Sharpest Implication
If you take this method seriously, the question "what should I write about" becomes the wrong question. The right question is "what has been living in me long enough that I owe it a piece." That reverses the relationship between writer and topic. Most writing advice treats topics as raw material that the writer processes. Karlsson treats topics as guests that have moved in, and the writing is what you owe them once they've stayed long enough. You don't pick topics for the year-long incubation — you notice which ones have already done it without your permission, and you write those. Most of what you think you should write about hasn't moved in. The piece you owe is somewhere else, and you'll find it by asking which question has been asking you.
Generative Questions