{"path":"research/the-post-that-states-the-problem-from-inside.md","content":"# The post that states the problem from inside\n\n**Date**: 2026-09-22 · **Type**: source reading · **Status**: analysis, nothing ruled ·\n**Source**: a public Swedish Facebook post of 2026-09-18 by Tildah Puronne, roughly 1,400 words,\nplus its ten-comment thread. Full text held privately at\n`.private/sources/fb-post-puronne-politik-som-spel-2026-09-18.md`, because reproducing the whole of\nsomeone's writing on a project website is publication rather than citation; this page quotes only\nwhat its argument needs.\n\n**Why it is worth a doc.** Almost every external source in this corpus arrives as a theory *about*\ndeliberation — Mutz on cross-cutting exposure, Kahan on curiosity, the DeepMind facilitation study.\nThis one is a person doing it, in public, on a live disagreement inside her own family, and arriving\nat most of this project's convictions without any of its vocabulary. That makes it useful in a way\nthe literature is not: it is evidence about whether these positions are *reachable* by someone who\nhas never heard of any of them.\n\n**It is not evidence the wager is true.** One self-selected person writing persuasively, who closes\nby naming two parties she has joined. It is a specimen, not a measurement.\n\n---\n\n## 1. What it does, in order\n\nThe first third is the problem: politics as a game (who wins, who loses, who is most dangerous for\nSweden), while people across the country sit with real fears. Ten of them are listed — and the list\nis **built in opposing pairs**, without ever announcing that it is. Fear of crime sits beside feeling\nsingled out for one's origin or religion; fear of religious extremism beside fear of racism and the\nfar right; fear that welfare is disappearing beside fear that the state takes up too much room in\npeople's lives. Then:\n\n> *\"Och alla verkar titta på varandra och tänka ungefär samma sak: 'HUR KAN NI INTE SE DET JAG SER?'\"*\n\n**That is the founding conviction felt from the inside rather than argued.** It is also, in the post,\nexplicitly the place she got stuck — not a conclusion. Everything after it is her answer, and the\nanswer is that the reason the other person holds their position is findable, and that finding it is\nwork worth doing.\n\n## 2. The three refusals, which are the no-verdict design stated by a layperson\n\n> *\"Inte automatiskt hålla med. Inte låtsas att alla påståenden är lika sanna. Inte mötas i någon\n> låtsasmitten där ingen får ha starka åsikter.\"*\n\nNot automatically agreeing. Not pretending all claims are equally true. **Not meeting in a fake\nmiddle where nobody is allowed strong opinions.**\n\nThe third clause is the one this corpus has had to argue at length and she states in six words. It is\nexactly the line between *mapping reasoning* and *splitting the difference* — and she keeps it\nfalsifiable a few paragraphs later: *\"Ibland har vi fel. Ibland finns det fakta som faktiskt\nmotsäger vår verklighetsuppfattning.\"* Sometimes we are wrong; sometimes there are facts that\ncontradict our picture of reality.\n\nSo the position is: understand without agreeing, refuse the false middle, and let facts still settle\nwhat they can. That is the product's own stance, reached by someone who has not seen it.\n\n## 3. The line that is *contributions are added, never overwritten*\n\n> *\"Tänk om vi kunde lägga HELA verkligheten på bordet istället för att slåss om vilken del av den\n> som får finnas?\"*\n\nWhat if we could put the *whole* of reality on the table instead of fighting over which part of it\ngets to **exist**.\n\nThe ratified principle says contributions are added and never overwritten, and that the mechanism is\nthe absence of a single current text for anyone to win. Her sentence names the same thing from the\nparticipant's side: the fight she is tired of is a fight over which parts are permitted to exist, and\nthe alternative is that all of them do.\n\n## 4. The wager, performed by hand on a real disagreement\n\nThe middle of the post is a descent. Her younger brother started working politically for Sweden\nDemocrats; she had a strong prior aversion; she decided to ask rather than stay angry — *what are you\nactually afraid will happen, what are you trying to protect, what do you see that I do not*. Then she\nran the same descent in the other direction, on what the right calls the left's naivety.\n\nWhat she reports finding at the bottom is a **shared list**: trygghet, frihet, gemenskap,\njämställdhet, människovärde, demokrati, barns framtid, rätten att få vara sig själv. And then the\nmove that makes it the wager rather than a platitude:\n\n> *\"Vi kanske ser helt olika hot mot dom sakerna. Vi kanske har olika idéer om hur dom ska skyddas.\"*\n\nShared structure, divergent weighting, with the disagreement relocated to *which threats bear on\nthose goods and how to protect them*. That is the sharpened form of the convergence wager — the form\nthe moral-foundations and cooperative-rules literature supports — arrived at by descending on one\nfamily argument.\n\n**The honest caveat**: she did this by hand, on one pair of positions, with a strong prior interest\nin the outcome. The corpus's own wager is about what happens when this is done at scale, by people\nwithout that motivation, and this is silent on that.\n\n## 5. Four things it supplies that the corpus does not already have\n\n**(a) The sayability condition, stated symmetrically.** The audience-matching rule asks for a\ndisagreement that is real, live and **sayable**. She gives the two-sided version:\n\n> *\"Jag vill att någon ska kunna säga 'jag är rädd för islamism' utan att samtalet tar slut där. Och\n> jag vill att någon annan ska kunna säga 'jag är rädd för vad som händer när muslimer görs till ett\n> kollektivt hot' utan att samtalet tar slut där heller.\"*\n\nSymmetry is what makes it a design constraint rather than a preference for one side. A surface that\nkeeps the first sayable and not the second has not implemented the condition; it has picked a team.\n\n**(b) The concept layer, inverted.** *\"Kanske pratar vi ibland om samma längtan med helt olika\nspråk.\"* The concept layer is built for one word carrying two senses. This is two vocabularies\ncarrying one referent, and the corpus has no machinery for it — the bridge-candidate proposal in\n`core-periphery-and-the-concept-layer.md` § 5b is the nearest thing, and it starts from a shared\nterm.\n\n**(c) What actually triggered the descent was a relationship.** Not curiosity in the abstract: her\nbrother. The eight conditions under which legibility pays\n([incentives-analysis.md](incentives-analysis.md)) do not include *someone you love is on the other\nside*, and for ordinary people it may be the strongest of them. Worth adding as a ninth, or worth\ndeciding it is out of scope because no platform can manufacture it.\n\n**(d) She names her own prior disengagement.** Years checked out, did not vote in the last election,\nexhausted by the mud-slinging and by being expected to pick a team and defend it. That is the\ndisengagement failure mode in first person — and her exit route was precisely the conversion the\ncorpus argues for: from diffuse ambivalence (*both sides seem to have a point*) to located\nuncertainty (*I know which premise we differ on*).\n\n## 6. The comment thread is the measurement, and it is not comfortable\n\nTen comments, mostly warm agreement. **Two of them land on the exact positions the post refuses.**\n\nOne reads, in full: *\"Sanningen finns någonstans i mitten av allting.\"* The truth is somewhere in\nthe middle of everything — which is the fake middle she explicitly rejected four paragraphs in.\nAnother proposes letting the players get on with their game while *\"vi lever i vår egen upplevelse på\nsidlinjen\"*, living in our own experience on the sidelines, which is the disengagement she just\ndescribed climbing out of.\n\n**So a carefully-written, popular statement of the position was read as centrism by one reader in\nten and as quietism by another**, in its own comment section, on the same day. That is a cheap,\nreal-world measurement of something this project needs and does not have: **how the\nunderstand-without-agreeing register is actually received.** The two most common misreadings are now\nnamed and attested, and neither is hostile — both commenters are agreeing with her.\n\nThe consequence for outward copy is concrete: any surface that says *we map reasoning rather than\ndelivering verdicts* will be heard by a meaningful fraction as *the truth is in the middle*. The\nteeth-and-confidence ruling already pushes against that; this says the pressure is real and comes\nfrom sympathetic readers rather than critics.\n\n---\n\n## What would be worth doing with it\n\n**Not extraction.** The founder's instruction was to document it, and the clearance given for the\nAndrés and Caleiro sources (*\"it's fine with the ppl involved\"*) was about those two. The commenters\nare private individuals and are not cleared at all.\n\nThree candidates, none ruled: adopt the two-sided sayability sentence as the statement of the\nmatching rule's *sayable* clause, since it is better than the corpus's own phrasing · open the\nsame-longing-different-vocabularies case as a concept-layer question · and treat § 6 as the first\nattested data on how the register is misread, which is the kind of thing the constellation's stranger\ntest is for and cannot produce by itself.\n\n---\n\n## Cross-references\n\n[vision.md](../vision.md) § Opacity Is a Cost, Not a Mystery (the conviction § 1 states from inside) ·\n[civilizational-vision.md](../civilizational-vision.md) § Love Through Understanding (the founding\nimage; her *\"förståelse betyder inte acceptans\"* is its naivety objection pre-answered) ·\n[islands-of-coherence.md](islands-of-coherence.md) § 5b-ii (the matching rule, and § 5a's sayability\nclause) · [incentives-analysis.md](incentives-analysis.md) (the eight conditions, and the ninth this\nsuggests) · [does-descending-change-the-descender.md](does-descending-change-the-descender.md) § 8\n(the disengagement mode she describes leaving) ·\n[core-periphery-and-the-concept-layer.md](core-periphery-and-the-concept-layer.md) § 5b (the\ninverted case) · [live-election-test-design.md](live-election-test-design.md) § Inbjudningstexten\n(the corpus's own best Swedish public copy, same register, same election season)\n"}