Hari names his favorite writer in passing — George Simenon, the French novelist of The Maigret mysteries and dozens of other titles. A complete genius. If anyone watching, if you have not — have you read him? Oh my god. He's — I was going to send you some of his books.1 Then he quotes the Simenon discipline that governs his own editing: In his editing process, I go through and strike out everything that sounds like writing.
For Hari, clarity is absolutely the most important thing. Lyricism is allowed — Hari has lyrical passages in his books — but only when I really feel I've earned them with the reader. He goes lyrical at moments of great emotional truth for the people I write about. Not at moments when he's showing off. Not at moments when he hasn't given the reader enough material to feel the truth that lyricism would otherwise be ornamenting.
The discipline is earn before you sing. If the chapter has done the work — built the character, traced the scene, made the reader feel something they didn't feel at the start — then lyricism is the right register. If the chapter hasn't done that work, lyricism is decoration. Decoration corrupts.
Hari is naming a style-restraint discipline in the same family as several other vault disciplines but with a specific operational locus. The discipline says: prose that calls attention to itself is failing the work; prose that disappears into the story is succeeding. Simenon's strike out everything that sounds like writing is the operational rule. Hari adopts it with one exception: at moments of great emotional truth, lyricism is what the moment deserves.
The mechanism has two parts.
Part 1: Default to invisible prose. Most of the manuscript's prose should be doing its work without the reader noticing. The reader is in the story, in the argument, in the people. The prose has gotten out of the way. The Simenon strike-out test catches sentences that have intruded into the reader's attention — sentences whose word-choice or rhythm has made the reader briefly notice the writer rather than the work.
Part 2: Lyricism at earned moments. The moments when Hari permits himself lyrical prose are exactly the moments where the lyrical register matches the emotional truth on the page. The 91-year-old Vietnamese woman is dying. The drug dealer is telling his conception story. The depressed teenager is confronting his real losses. These moments deserve prose that registers the depth. The lyricism is appropriate at these moments because it's load-bearing — the prose's musicality is doing emotional work that flatter prose couldn't do.
The writer who runs lyricism throughout — even competently — produces a book that reads as one tone. Every paragraph is at the same emotional pitch. The reader's responsiveness fatigues. By page 200 the reader has been at maximum-emotional-register for 200 pages, and they've stopped registering the register because everything is at the same intensity.
The writer who runs Hari's discipline — flat clarity by default, lyricism at earned moments — produces a book that modulates. The flat passages give the reader room to think. The lyrical passages arrive at the moments when emotion should land. The contrast is what makes both registers work. Without the flat default, the lyrical moments have no foundation to stand on. They just sound like more of the same.
This is also why writers Hari admires — Simenon, but also George Orwell, Joan Didion in her later work, Tony Judt — share the same architecture. Their prose is mostly transparent, with carefully chosen moments of registered intensity. The discipline is durable across generations of strong nonfiction work.
This page deepens True Intelligence Is Making Complicated Things Comprehensible (Hari) — the grandmother-discipline is what flat clarity serves. The two pages combined describe Hari's full style position: write so the grandmother can read every sentence, and reserve the lyrical moments for when the material genuinely earns them.
It connects to Restraint — Pull Your Punches (Orlean) — Orlean's pick-one-and-hit-it-hard principle is the same discipline at sentence scale. Orlean uses one adjective where a less-disciplined writer uses three; Hari uses flat prose for nine pages and one lyrical paragraph in the tenth. Different scales of the same operational discipline.
It sits in productive tension with Voice as Literary Texture (Deen) — Deen's voice-modulation framework explicitly endorses mode-switching as a craft mechanism. Hari's earned-lyricism is a specific mode-switch within Deen's broader system. The vault should hold both: Deen's general framework, Hari's specific instantiation.
Most of Lost Connections runs in flat-clarity register. Hari explains studies. Hari describes interviews. Hari builds an argument step by step. The prose is doing its work invisibly. Then, at the end of Chapter Two, the Vietnam apple scene closes with a moment of lyrical pressure: I had been killed by a poisoned apple. I was Snow White. I was Alan Turing. I was Eve.1
The four short sentences are operating at a different register than the prose around them. They are lyrical — rhythmic, image-rich, allusion-loaded. They are also earned — Hari has just spent fifteen pages building toward the moment, has built the reader's identification with him on the hospital floor, has loaded the question of what does a poisoned apple symbolize. By the time the sentences arrive, the reader is prepared for the register-shift. The lyricism lands because the architecture has been built to receive it.
If those four sentences had appeared on page 4, before the Vietnam material had been built, they would have read as literary affectation. The reader would have noticed the writer. The prose would have intruded on its own work. Same sentences, different position, different result. Position is everything.
You're finishing a draft. You print it. You sit with two markers — one red, one blue.
You read with two passes.
Red pass: Mark every sentence that sounds like writing. The sentences that draw attention to themselves through unusual word choice, elevated diction, conspicuous rhythm, literary allusion. The red sentences are candidates for cutting or flattening.
Blue pass: Mark every moment of great emotional truth in the material. The moments where a real person is being witnessed at depth. Where a real scene is at maximum gravity. Where the argument is doing its load-bearing emotional work.
Now compare the two markings. The red sentences that land on blue moments are earned lyricism — keep them. The red sentences that land on non-blue passages are unearned lyricism — flatten them. The blue moments that have no red sentences are under-served by their current prose — they deserve more intensity than they have.
The audit takes a few hours. It dramatically improves the manuscript's modulation. Most writers haven't done it. Doing it is what differentiates the disciplined nonfiction writer from the writer-in-love-with-their-own-voice.
You'll know the discipline has slipped when readers describe the prose as overwritten or purple or trying too hard. The complaints land on writers who have run lyrical register too consistently. The same prose, modulated against flat clarity, would have produced vivid rather than overwritten. The difference is structural, not stylistic.
You'll know it's working when readers describe the same book using both plainspoken and moving. The combination is unusual. It's the signature of the modulation working. The plainspoken register did its grounding work; the lyrical moments landed because the grounding was there to launch from.
The principle has substantial support across the strongest nonfiction practitioners — Didion, Orwell, Judt, more recently Patrick Radden Keefe and Sheri Fink. The convergence across generations and traditions confirms the principle's operational robustness.
The tension: some genres explicitly invite consistent lyricism (nature writing, literary memoir, prose-poetry-essay hybrids). Macfarlane's Underland runs lyrical register throughout and the lyricism doesn't fatigue the reader because the subject is lyricism's territory. Hari's discipline is most clearly correct for investigative-argument-driven nonfiction. The vault should note: the principle is form-dependent, and writers should know which form they're in.
Hari and Macfarlane diverge productively on lyrical density. Macfarlane's whole project rests on the lyrical register; Hari's reserves it. Both writers produce strong work. The difference is what each form is for. Macfarlane wants the reader to linger — the lyrical density slows reading-pace, which serves nature-writing's contemplative purpose. Hari wants the reader to complete — the flat-default-with-earned-lyricism keeps pace efficient, which serves investigation's persuasion purpose. Neither writer is wrong. The vault should hold both as form-appropriate instantiations of different priorities.
Hari and Wang converge on the underlying discipline. Wang's manuscript copying trains voice-recognition. Hari's Simenon strike-out trains voice-suppression. Both writers are running disciplines that produce deliberate prose register. Both refuse the default-of-untrained-prose. The convergence reveals that style is always engineered — the question is which engineering serves which work.
Eastern Spirituality: Silence as the Condition for the Sound That Matters — the contemplative traditions consistently teach that silence is what makes specific sounds meaningful. The mantra works because of the silence around it. The bell rings because of the quiet before and after. The flat-default-with-earned-lyricism is the writing-instance of the same principle. The handshake reveals that modulation is a domain-general aesthetic principle — what works in mantric practice works in nonfiction prose.
Behavioral Mechanics: Emotional Pacing as Influence Architecture — the influence literature has documented that modulated emotional intensity produces stronger persuasion than consistent emotional intensity. The audience that has been at high-intensity for too long stops responding. The audience that has been at low-intensity becomes available for high-intensity moments. The handshake reveals that Hari's lyricism discipline is operationally identical to elite-persuader pacing-discipline. The writer who runs the modulation is doing the same operation as the political-communicator who builds toward earned emotional peaks.
The Sharpest Implication If Hari is right, the contemporary essay-Substack ecosystem may be producing consistent-lyrical writing at scale, which then underperforms its modulated counterparts in terms of actual persuasive effect. The writers who do best in the short-form attention market may be doing worse on the long-form influence dimension. The flat-default discipline is harder to develop and looks less impressive in any single paragraph, but the books built on it move minds more durably than the lyrically-saturated essays.
Generative Questions