How We Got Here

In February 1796 the Astronomer Royal dismissed his assistant David Kinnebrook for marking star transits eight tenths of a second late. Bessel later established that observers differ from one another systematically and irreducibly — and astronomy's answer was the personal equation: publish a constant for each observer, build the variation into the model, keep the observer. This zine follows the measuring apparatus out of the observatory: Wundt's laboratory, Cattell's “mental tests,” Thorndike's puzzle boxes and his 1918 credo that whatever exists at all exists in some amount, Kelly's Kansas Silent Reading Test, Pressey's machine built from typewriter parts, Skinner's teaching machines, and Lovaas at UCLA in 1987 — while a second rail runs alongside from Galton through Kraepelin to the DSM. The two meet at a table with a child at it. The first zine built as a Burke-style chain, and every joint is marked documented, contested or leap — six solid, three dashed, one dotted — because two of them are weaker than the story wants. The turn: school found the same human variation astronomy found, and made it the thing to remove. Accommodation is the older answer. Fourteen spreads, twelve diagrams, a running chain spine, a “Not:” refusals set.

In March 2020 the WHO posted “FACT: COVID-19 is NOT airborne” and called the alternative misinformation. Not a lie — doctrine, faithfully applied, resting on a number that had quietly changed questions. William Wells put the boundary between droplets that fall and droplets that evaporate into lingering nuclei at 100 micrometres, and never wrote 5; his own table gives a 1 µm particle sixteen hours to fall two metres. The 5 µm figure came from industrial hygiene asking which particles reach the deepest lung — deposition, not suspension — and Alexander Langmuir fused the two in 1951, in a lecture preparing public-health students for biological warfare months after Korea began. Randall, Ewing, Marr, Jimenez and Bourouiba reconstructed it in Interface Focus under the title “How did we get here,” and refuse to solder the load-bearing joint, writing only that the deposition focus “likely” popularised the cut-off — so this zine marks it contested too. Meanwhile Flügge's team in Breslau had grown colonies nine metres from a sneeze in the 1890s, meant “droplets” to cover what we now call aerosols, waited five hours before reading their plates, and recommended ventilation. The 1–2 metre rule has its own separate lineage and turns out to match the width of a 1940s camera's field of view. The second chain zine, every joint marked documented, contested or leap. Fifteen spreads, thirteen diagrams, a running two-rail spine, a “Not:” refusals set.

How much evidence is enough before you do something to a child? Particle physics answered that question carefully and wrote it down. It will not permit the word “discovery” below five sigma — a p-value of 3 × 10⁻⁷, one chance in three and a half million — and the reason is empirical rather than principled: too many 3σ and 4σ results kept evaporating. Louis Lyons, whose paper is the spine of this zine, goes further and argues the bar should not be uniform at all, grading it by degree of surprise, impact, look-elsewhere effect and systematics: single top production 3σ, the Higgs 5σ, supersymmetry 7σ, faster-than-light neutrinos above 8σ. He also prints the statisticians' objection that his own gold standard is unsound. Against all that stands p = 0.05 — about 2σ — chosen by Fisher in 1925 with the word convenient, and unexamined since. The gap is roughly 170,000×, and the asymmetry runs backwards: if physics is wrong it retracts a paper and the particle is unharmed; if we are wrong a childhood already happened and there is no more data to collect. Run Lyons' four dials on an intervention and every one reads high — then the systematics column turns out to be empty, because 150 intervention studies for young Autistic children produced 11 that mentioned adverse events at all. Not an argument against science. An argument against scientism, which keeps the authority and drops the practice. The first chain here with no leap joints. Sixteen spreads, fourteen diagrams, two rails that never meet, a “Not:” refusals set.

For roughly two centuries the strongest objection to a moving Earth was that the stars showed no parallax. The reasoning was valid, the observations careful, and the conclusion wrong — partly because the distances required seemed absurd, and partly because telescopic star discs are an artefact of the aperture rather than a property of stars. Then in 1838 Friedrich Bessel measured 61 Cygni at 0.3136 arcseconds from Königsberg — a two-centimetre coin seen from thirteen kilometres. The method carries the whole argument: the baseline is the observer's own displacement across Earth's orbit, so a viewpoint with no location has no baseline, no angle and no depth. A view from nowhere cannot measure distance even in principle; it is the blindest available position. And it is the same Bessel, at the same observatory, who gave us the personal equation in No. 38 — twice showing that knowledge requires a located, particular observer. Every rung of the cosmic distance ladder above parallax is calibrated by the one beneath, and rung two — the period–luminosity relation — was built by Henrietta Swan Leavitt, employed as a Harvard “computer,” paid by the hour, not permitted to observe, and published under her director's name: “has been prepared by Miss Leavitt.” That is the third woman in three chains to do the work and lose the byline, and the zine stops to name all three. Donna Haraway gave the fantasy its name in 1988 — the god trick, the unmarked category claiming the power to see and not be seen — and her positive claim is the one usually dropped: only partial perspective promises objective vision. She refuses relativism just as flatly, calling it “a way of being nowhere while claiming to be everywhere equally.” Then the numbers: 96% of behavioural-science samples from Western industrialised nations, 68% from the US, reported as facts about humans. Then research done on Autistic people rather than with them, and double empathy as a finding that only exists because somebody stood at the other end of the baseline. The correction is neither removing the observer nor declaring all views equal: publish your coordinates. Fifteen spreads, thirteen diagrams, a single rail with a marked baseline, a “Not:” refusals set.

Aristotle's physics explained motion by what a thing is: heavy bodies fall because it is in their nature to seek the centre, faster in proportion to weight, slower in a denser medium. It is not a foolish theory — go outside and test it and every prediction checks out, because everyday observation happens in air. The error is where the explanation was put. Galileo showed the rule contradicts itself (tie a heavy body to a light one and it predicts both faster and slower), rolled balls down inclined planes, and made the strange move of reasoning about a fall with no medium at all — while the famous Leaning Tower demonstration turns out to come from a biography written decades later and is marked contested here. Newton finished it: acceleration is set by the forces acting plus one scalar property, and disposition disappears. Then, on 2 August 1971, David Scott released a geology hammer and a falcon feather together in the lunar vacuum and they landed together, on film. The difference was never in the feather. It was in the air. This chain argues that mindset marketing runs that two-thousand-year move backwards on children. Behaviourism looks like it made the same move but stopped one ring short — it deleted the interior and still left the child as the only adjustable thing. California's task force commissioned its own research, was told the associations between self-esteem and its expected consequences were “mixed, insignificant, or absent,” and shipped in 1990 regardless. Growth mindset fails the same way across 63 studies and 97,672 participants, with higher-quality studies less likely to show benefit — and Dweck and Duckworth both objected to the deployment while it kept growing. The dashboard completes it: a schema with fields for effort, focus and behaviour points, and no field for the room. Along the way the zine corrects its own source — the famous Eric Hoffer line about a movement becoming a business and degenerating into a racket is a misquotation that invents a moral arc he never wrote. It closes on scarcity and slack, on equity literacy, and on the first leap joint since No. 39, marked because it is the sentence a reader most wants to be true. Fourteen spreads, twelve diagrams, two rails running in opposite directions, a “Not:” refusals set.

In 1824 Sadi Carnot proved something strange: that a heat engine has a hard, provable ceiling on its efficiency, set only by the two temperatures available and beatable by no engineer who ever lives. And with it came the result that matters here — waste heat is not sloppiness, it is the fee. An engine with nowhere to dump it does not run at all. So “eliminate all waste” isn't an ambitious goal not yet reached; in thermodynamics it names the one outcome that is ruled out. (He reached this from caloric theory, which was wrong — Clausius and Kelvin supplied the formula later, so “Carnot derived the efficiency equation” is itself a smoothed retelling, and the joint is marked contested.) Then the word left the engine room, and what it shed in transit was the ceiling. Frederick Taylor's Principles of Scientific Management appeared in 1911, and Raymond Callahan's Education and the Cult of Efficiency (1962) dates what followed to the month: in July 1911, one month after Taylor's American Magazine series concluded, a school board member told the NEA that “scientific management” held a message for every teacher — and that if teachers did not voluntarily increase their efficiency, the business world would force them to. By 1916 Cubberley's textbook was listing what schools must take from the factory: good tools, specialized machinery, continuous measurement of production, and “the elimination of waste in manufacture.” Ninety-two years after Carnot, the phrase arrives in a book about children as a specification. By 1912 an educator was already complaining that teachers were “curiously protected in inefficiency.” And the question nobody asked sits in Callahan's own parenthesis: would Schmidt, the pig-iron handler, have agreed that science had reduced his fatigue? It ends at spoon theory — and then spends an entire spread arguing that this argument should not need physics at all, because you are not a heat engine and deriving rest from thermodynamics is precisely the move the collection condemns. Fifteen spreads, thirteen diagrams, a spine that crosses its own ceiling, a “Not:” refusals set.

Until 1960 a second was a piece of a day — one 86,400th of it — and the definition pointed at something. Not a prototype in a vault, but the sky: you could check it against a star. Then quartz clocks arrived and, for the first time in the history of measurement, the instrument was steadier than the Earth. Scheibe and Adelsberger reported the seasonal wobble in 1936; Stoyko put it at about two milliseconds in 1937. The Earth was never irregular to itself — it became irregular the moment we had something steadier to compare it to. So the tie was cut, twice. In 1956 the second became a fraction of the tropical year 1900, a year that had already ended and could never be re-measured; in 1967 it became 9,192,631,770 caesium transitions, and after that the definition points at no day, no year, no sky at all. Every one of those moves was an improvement, and this zine says so — the collection exists to reclaim science, not to indict it. Since 1972 the gap between the atomic standard and the turning planet has been closed in public by leap seconds: 27 of them, first on 30 June 1972, last on 31 December 2016, and none since, because the Earth started spinning faster and the correction now due is a negative leap second that has never been tested. Duncan Agnew showed in Nature in 2024 that melting Greenland and Antarctic ice slowed the spin enough to postpone it from 2026 to 2029 — climate change bought the world's timekeepers three years, and he refuses to call that good news. Resolution 4 of the 27th CGPM decided in 2022 to widen the tolerance by 2035, and the implementing vote is at the 28th CGPM this October; the proposal is a leap hour. The argument: a leap second is not a bug but an honesty mechanism — a standards body saying in public, on a named date, with a number attached, that its standard and the world disagree by exactly this much. Abolishing it is defensible engineering. What it costs is the announcement. And that is the one thing carried across to us, marked leap because caesium licenses nothing about a person: an accommodation is a leap second, and the objections to it arrive in the same order — it's irregular, it breaks the system, everybody handles it differently — with the same proposed remedy, widen the tolerance and stop announcing. Ends at processing speed and crip time: the Earth is not slow, the second left. Eleven documented joints, one contested, one leap. Sixteen spreads, thirteen diagrams, a running spine that is a plot of its own data, a “Not:” refusals set.

A chain about how a whole branch of physics stopped waiting for the Moon. The Sun’s outer atmosphere is always there and almost never available: the photosphere drowns it, so for most of astronomy’s history the corona and the prominences could only be studied during totality — a few minutes, roughly every eighteen months, along a track a few hundred kilometres wide, usually somewhere difficult, and one cloud ended the expedition. It begins in 1604, where pinhole solar diameters came out systematically too large and nobody could say why until Kepler supplied the theory of how an image forms behind a small hole: the method came first and worked badly; the understanding came second and fixed it. Then Guntur, 18 August 1868, where several observers caught a yellow line near sodium’s — and where what Janssen actually took away was not the line but its brightness. The next morning he observed the prominences again with no eclipse at all, spreading the sky’s glare thin with dispersion while the emission line stayed put. Lockyer, who had wanted this since 1866 and was blocked by glass rather than by ideas, got there on 20 October; both reports reached the Académie within days and it struck a medal for the pair — for the method. Neither report flagged the yellow line, which is the first of four contested joints: the famous sentence “helium was discovered during an eclipse in 1868” is wrong in almost every part, and what the compression deletes is years of argument. Lockyer alone later called it D3 and proposed an element; for a quarter century chemists refused it, not because the spectroscopy was faulty but because nobody could put the gas on a bench — and a thing visible only in light from 150 million kilometres away did not, to them, count as a thing. They were rigorous and they were wrong, and the zine insists on both, because the same method also produced two elements that were never there: coronium and nebulium, dissolved by Bowen in 1927 and Grotrian and Edlén in 1939 and 1942 as forbidden lines of ordinary oxygen and iron — which is also how we learned the corona is millions of degrees hotter than the surface. Ramsay got helium out of cleveite on 26 March 1895, twenty-seven years late, and it had been underfoot the whole time. Then the chain runs downhill, because capability is specific: in 1919 the Moon was needed again to see whether starlight bends, the Sobral astrograph plates were set aside, Earman and Glymour charged Eddington with seeing what he wanted, and Kennefick later established that the call was Dyson’s and that the plates could only agree with Newton if the instrument had also changed magnification — a reason to distrust them other than what they said, which is the whole difference between rejecting bad data and rejecting inconvenient data. Finally Bernard Lyot, chasing stray light up a mountain, photographed the corona with no eclipse on 12 July 1931: an eclipse at a time and place of human choosing, turning minutes per expedition into hours per clear day. We now fly manufactured occulters at other stars. The transfer, marked leap because it is one, refuses the obvious framing first: eclipses were computed years ahead, so the difficulty was never luck but a structural condition — the observation existing only there, only then, with the ocean, the freight, the years and the weather all charged to whoever needed it. An expedition demonstrates what is required and is a catastrophe as a delivery mechanism. So there are three states, not two: reach it, ask for it, or have it — and the middle one is what we named accommodation, which is the word Stimpunks' own page exists to criticise, since the problem is structural and the response is individual. Lyot did not apply for access to the corona; he built an instrument in which it is simply there, and the occulting disc is not a concession but where the precision comes from. Which is a curb cut: Michael Pachovas and friends poured cement into a crude ramp one Berkeley evening and rolled off into the night, and Berkeley cut its first official kerb on Telegraph Avenue in 1972 — while the sledgehammer version of that story is disability lore, not history, which a chain that spends a spread on smoothed retellings could not close on. Angela Glover Blackwell named the curb-cut effect in 2017, and the zine states plainly that it is an observation about what built-in conditions do and never the justification. Twenty-one spreads, twelve documented joints, four contested, one leap, a generated spine, and a “Not:” set that refuses the reading that barriers are valuable for what they reveal. A rhyme, not a proof.

A chain about a limit nobody beats, and the rooms we build as though it were free. In July 1928 one volume of Physical Review printed the measurement and its explanation on facing pages — J. B. Johnson at 97–109, H. Nyquist at 110–113, whose first footnote is three words long: “See preceding paper.” Every conductor above absolute zero hisses, and Nyquist’s conclusion is that the effect is “a universal function of frequency, resistance and temperature and of these variables only” — not the metal, not the workmanship, since “one may be made of silver and the other of lead.” The available power in a band is kTB, and that is in the primary rather than a modern gloss on it. The part that usually goes missing is why it cannot be engineered away. Nyquist did not measure the floor, he deduced it — from the second law. Two resistors at one temperature must pass equal power both ways in every band, or a resonant circuit between them would move heat between equal-temperature bodies for free. The hiss is not permitted, it is required: a silent resistor would be a perpetual-motion machine, which is the same law that says you cannot un-stir the milk from the tea — and nobody files that under insufficient effort. Twenty years on, Shannon’s Theorem 17 turns the floor into a limit on meaning, with the sentence that closes the door: “It is not possible to transmit at a higher rate by any encoding system without a definite positive frequency of errors.” You cannot filter your way to clarity, and that is a theorem rather than a complaint. So engineering stopped hunting for a quieter resistor and changed the question: cool it, shield it, narrow the band — three moves, none of which asks the receiver to beat physics, all of which change the conditions the listening happens in. Which is exactly the instrument at Holmdel, where a maser and a reference load both sat in liquid helium at 4.2 K — a floor lowered deliberately so a remainder would mean something. A remainder was left: 3.5 ± 1.0 K, and the paper is mostly a list of what it was not, down to taping the seams with aluminium and walking a transmitter to ten positions on the ground. The pigeon story is corrected against its primary, and the correction is better than the version that travels: Wilson wrote that evicting them “obtained only a small reduction in antenna temperature” — the floor survived them, and that clause is the one that gets cut. The famous phrase “white dielectric material” is not in his Nobel lecture, and the birds are not in the 1965 paper at all. What could not be removed was “isotropic, unpolarized, and free from seasonal variations (July, 1964-April, 1965)”, and every property that made it impossible to eliminate is a property that made it the universe. Then the 1928 shape repeats mirrored — theory at 414–419, measurement at 419–421, thirty-seven years on. The most consequential discovery in modern cosmology arrived as noise two engineers spent ten months trying to get rid of. Meanwhile the transfer to rooms turns out not to be ours to make: ANSI/ASA S12.60 caps classroom background at 35 dBA because the profession already calls the quantity signal to noise ratio and wants about 15 dB of it — and its instruction is to lower the floor rather than amplify the teacher. Wilkins halved headaches by removing 100 Hz fluorescent flicker, with his own caveat that the harm fell on “a minority that is particularly affected”; Kim and de Dear found sound privacy the largest source of dissatisfaction in any building, and found open plan bought no more ease of interaction than a private office. The last two joints refuse to overclaim, on the page rather than in a footnote: the step to Autistic auditory experience is contested — there is no k for a nervous system — carried by Remington and Fairnie’s finding of increased auditory capacity, one mechanism producing both the distress and the ability, and by Rosas-Pérez’s 2026 acoustics thesis, in which no Autistic respondent among 311 noise-sensitive people found filtering easy — printed with her caveat that the whole sample was already sensitive. Her other finding is where the zine says the analogy breaks, and the break is worth more than the resemblance: no habituation, but an accumulation effect. Thermal noise is stationary; kTB is the same in the tenth hour as the first. The human pattern is cumulative. A resistor never gets tired of being warm — so the floor in a room is crueller than the floor in a circuit, not merely like it. Which also means “you’ll get used to it” is testable, and the exposure hypothesis under it rests on five participants, none noise-sensitive, and two pure tones. The closing analogy is leap, because physics entails nothing about who deserves a quiet room. What it does is remove an excuse. Fifteen spreads, ten documented joints, one contested, one leap, a spine the page calls schematic rather than letting it pass for a plot, and a coda pointing at the practice already written down. A rhyme, not a proof.

The chain the How We Got Here motive was written for, and the one where it turns on us. Darwin stated the two-level problem in Descent of Man (1871) and finished the paragraph "and this would be natural selection." Wynne-Edwards (1962) got the group-level version wrong and Williams (1966) was right to demolish it — a joint kept because any sympathetic account that skips it is selling something. Hamilton (1964) supplied inclusive fitness, which is good science and stays good science. Price (1970) then partitioned selection into within-group and between-group components, so by 1970 both frameworks sat in one formula, before the argument started. Then the other rail: Tom Maschler tried to talk Dawkins out of the title in 1975, and The Selfish Gene’s own introduction says the word to stress is “gene,” that the book “devotes more attention to altruism,” and — in the next sentence — that “altruism might well be favoured at other levels”, which is multilevel selection in everything but name. Chapter one refuses the naturalistic fallacy outright: “I am not advocating a morality based on evolution.” And Dawkins later retracts his own most-quoted sentence, calling “born selfish” misleading and asking readers to delete it. The correction was printed in the same book as the error and it did not travel. Wilson & Wilson (2007) then revived group selection on a careful formulation, stated “the main exception to this rule” immediately after their famous line — and forty pages later compressed it into a one-foot summary, echoing Rabbi Hillel, that drops both the exception and the word single. That compressed version is the one our own Altruism glossary carries, second-hand, and we found it while writing the spread about the mechanism. Nowak, Tarnita & Wilson (2010) and a reply co-signed by more than a hundred biologists show a field still arguing; and for standard models kin selection and multilevel selection turn out to be the same arithmetic added up in a different order. The turn is tagged leap: account at one level only and every other level’s contribution becomes invisible — not disproved, never entered — which decides what repair is thinkable. Nine documented joints, three contested, one leap. Refuses being read as anti-evidence, as anti-Dawkins, and as “group selection won.”

The chain that starts on a grass and ends on a specification. Sugarcane is propagated by pushing a cut piece into the ground, so every plant is the previous plant again — it can be multiplied forever without anything ever meeting anything, which is what Anna Tsing means by “comparatively self-contained, oblivious to encounter.” Portuguese Brazil assembled that cane with land emptied by conquest and with enslaved Africans taken far enough that there were “no established routes for escape” — and this chain withholds its own format on that link, visibly and on purpose, because what was done there is not a mechanism for illustrating an argument. Interchangeability was not discovered in people; it was produced in them, by force, because the mill needed it as an input. Sidney Mintz, read at pp. 47–48, documents Barbados holdings cut into equal ten-acre parcels harvested seriatim “assuring a steady flow of cane to the mill,” mills that “operated unceasingly,” a work week broken only from Saturday night to Monday — and part of the pace driven by a belief about boiling syrup he calls “somewhat misleading,” so the demand had already outrun the chemistry. His verdict is printed here with the hedge he put on it: “I believe it was probably the closest thing to industry that was typical of the seventeenth century.” Industry before the homeland — missed, he says, partly because scholars looking for the factory “began with the artisans and craftsmen of Europe.” Tsing names the resulting idea of progress scalability, the ability to change scale “without any change in project frames” — and states its price: it “banishes meaningful diversity, that is, diversity that might change things.” Then the rails cross. Matsutake has defeated seventy years and millions of yen; the 2020 Mycoscience review still reports cultivation “mostly unsuccessful,” and the only field successes used pine already colonised — so the only way to farm it is to already have the relationship and wait. It exists solely as an exchange with a tree, which is why it will not travel. And in Japan it was lost by abandonment: fossil fuel replaced firewood, the woodlands went shady, the pines died, and by the mid-1970s it was rare — a support removed by improvements aimed at something else. The turn is tagged leap: an accommodation is, by definition, a difference that requires the frame to change, which is exactly what a scalable system is specified to exclude. Nine documented joints, one contested, one leap. Refuses enslaved people as metaphor, ruin as an aesthetic, and “nature says be kind.”

Edward Lorenz published Deterministic Nonperiodic Flow in 1963 — three equations for atmospheric convection whose solutions never repeat, and a real result checked exhaustively for sixty years. What travelled fastest afterwards was not the mathematics but the register: nonlinear dynamics acquired a rigorous-sounding authority outside physics, difficult for a non-specialist to check. In 1999 Marcial Losada fitted Lorenz-type equations to speech coded as positive or negative in business meetings — equations derived for fluid convection, applied to things people say at work, on the justification that both are complicated. Being complicated is not a shared mechanism. In 2005, in American Psychologist, Barbara Fredrickson and Losada joined broaden-and-build to that model and produced a number: people at or above a positive-to-negative ratio of 2.9013 flourish; below it they languish; above 11.6346 flourishing fails again. Four decimal places, on the ratio at which a human life begins to go well. The precision is doing specific work — a number given that precisely claims an instrument was involved. Fredrickson’s 2009 trade book then rounded it away on its own cover: the 3 to 1 ratio that will change your life. The four decimals bought credibility at the journal and were friction at the bookshop. Then a part-time student. Around 2012 Nick Brown, in his fifties, taking an applied positive psychology master’s at the University of East London with no mathematics past school, met the paper on his course and could not find where the data entered the equations. So he went and read Lorenz 1963. With the original in front of him it was visible: the equations had no connection to the emotion data, and would generate the same numbers whatever the participants had said. He wrote to Alan Sokal, of the 1996 Social Text hoax; with Harris Friedman they published in American Psychologist in July 2013. The nonlinear modelling and both ratios were withdrawn as invalid — Losada did not co-sign — while what survives remains contested, Carol Nickerson arguing in 2018 that there is no empirical evidence for critical positivity ratios at all. The chain judges the paper and not the people. Lorenz was right; borrowing across fields is how much good work happens; Fredrickson corrected her own paper, which is the system working, slowly and embarrassingly, which is the only way it works. The failure worth studying is not that somebody was wrong. It is that it took eight years and a stranger. The final joint is marked leap and argues that the precision was the persuasion, and the precision was the empty part — which matters most for people already being scored, since a ratio for flourishing is an instrument for finding those not flourishing correctly. Six refusals, including that this is not an argument against measurement (the method worked), not “do your own research” (Brown read the source the experts cited, which is the opposite move), and not a claim that anyone was a fraud. Spread two draws a boundary with our own shelf: Five Times a Crab borrows attractor from the same 1963 paper, openly and with the source named. Same paper, two borrowings, and the difference is whether you say which one you are doing. Fourteen spreads, ten joints — eight documented, one contested, one leap — and a generated spine. A closing spread that is deliberately not a link asks whether the number reached the classroom, where 4:1 and 5:1 praise-to-correction ratios are repeated throughout positive behaviour support. No citation from that literature to Fredrickson & Losada could be established, and the spread says so rather than inventing one. What is there is a genre with its own ancestry — Gottman’s empirical 5:1, Flora (2000) — and the field’s own verdict, from Sabey, Charlton & Charlton (2019): good face validity, and “a lack of empirical support for any specific ratio.” A chain can be cut at the retraction; a genre reproduces without a source. Positive greetings at the door is the worked example, and the tell is what it measured — academic engaged time and disruptive behaviour, not whether a child felt welcome. With Ehrenreich on positive thinking as an apology for the crueler aspects of the market economy, whose flip side is a harsh insistence on personal responsibility. A rhyme, not a proof.

The chain is chronological and the sections are as long as the things they describe. This is the collection’s second scroll zine, and the second time the form has been the argument: a section’s height is the duration it covers, so Chicxulub is one screen you fall past and the wait between oxygenic photosynthesis and the sky changing colour is three, most of it empty, with a rail down the side so the emptiness reads as elapsed time. Hank Green asked what the most important thing that ever happened was, and defined important usefully — how much depended on it, how easily it could have gone otherwise, and the sharper move underneath: importance is not a property of the universe, so the question is what had to happen for mattering to exist at all. With Cali Moore (PBS Eons) and Joe Hansen (Be Smart), the candidates were the origin of life, the eukaryotic cell, the Earth–Moon system, Chicxulub, the ribosome and oxygenic photosynthesis. Put them in order and something the list format hides comes out. Only three were moments, and all three are impacts — and even those are not three of a kind: Chicxulub is in the ground with a crater and an iridium layer, Theia is an inference from a moon that has never been found, and the rock that may have carried life is imagined, standing in for a transition nobody can describe. One documented, one inferred, one imagined. Everything else was a duration. Photosystem II learned to pull electrons out of water once, and the sky did not change for something like a billion years, because oxygen was consumed on contact by a planet with an enormous appetite; the invention and the consequence are two different events. Eukaryogenesis was, in Green’s own words, a bunch of things in succession, and the date is a neighbourhood rather than a date. Pangea was over a hundred million years of being a continent — roughly a C curled round the Tethys, with a desert down its equator so effective that one landmass still held two worlds — and its assembly dried out the wetlands where lycopsids grew a hundred feet tall, leaving only the ones that could shrink. Two corrections the source did not carry. Its citation for earliest life is Nutman et al. (2016) on 3.7 Ga Isua stromatolites; Allwood et al. (2018) went back to the same rocks and found deformation structures, some of the bumps pointing downward, which stromatolites do not do — the widely accepted floor is the younger 3.48 Ga Dresser Formation. And the famous thirty seconds later and it would have hit ocean is not a result: it traces to a 2017 BBC documentary, not to Kaiho & Oshima’s ~13% of Earth’s surface or Collins et al.’s 45–60° trajectory, and the piece refuses it rather than using it — a stopwatch is a different kind of claim from a probability across a planet. A globe on every link shows the land at that moment, dashed and faint before 1 Ga and reconstructed after it, because that is where Merdith et al. (2021) can hold a full-plate model together and a crisp coastline older than that is an invention wearing a map’s clothes. The art carries its own epistemics, the way the joint tags do. Nine links: six documented, two contested, one leap, plus a leap named inside Link 9 and declined. The turn is the demand for a moment — when did you know, when were you diagnosed, what happened to you — every version of which assumes an impact. Mostly there is no crater; mostly a difference met an environment with no room in it, over years, and the word arrived long after the thing it named. The diagnosis is not the event; it is the sinks finally filling. Ends on four models of the next supercontinent, drawn all four and chosen none, because Earth has never held an arrangement and is not holding this one. Endosymbiosis itself is No. 7’s argument and is pointed at rather than repeated. A rhyme, not a proof.

The chain runs on two rails — what was told about the brain and what was observed in it — and the shape of it is the argument: the observed rail is live from 1969, and the told rail keeps running for another fifty-one years without touching it. 1878: Broca calls a rim of cortex le grand lobe limbique; limbus means border, and he claims no function at all. 1937: Papez proposes a circuit for emotion, as a proposal. 1952: MacLean fuses the two and coins limbic system — a name for a place becomes a name for a job, and nothing was discovered at that step; seventy-four years later the 2026 paper notes the category still “has resisted functional definition or even a specific list of its contents.” The 1960s: three brains stacked by evolutionary age, each described as independently conscious, and chronology pointed at living animals becomes a rank. 1969: Karten shows the bird forebrain carries the same neuron types and the same wiring as mammalian neocortex, packed as nuclei instead of layers — eight years before The Dragons of Eden won a Pulitzer carrying the model into common speech. That book is Carl Sagan’s, and this site is named out of his line about star stuff; the piece publishes that rather than keeping a clean story. 1995: Finlay and Darlington measure 131 species and find one conserved developmental schedule, not bolt-on modules — run it longer and the late parts grow hardest. 2005: an international consortium renames most of the bird forebrain, because paleostriatum, neostriatum and archistriatum were arguments disguised as labels. 2020: of 20 introductory psychology textbooks sampled 2009–2017, 14 discuss brain evolution and 86% of those carry at least one inaccuracy; two match the consensus. 2026: Imam, Kielo, Trude and Finlay measure 182 mammalian species and find not three floors but two systems that trade — olfactory system against neocortex at r = −0.63, with the limbic components covarying positively with olfaction, so they do belong together, just not for the reason MacLean gave. And the division is not age but the shape of the information: sight, touch and hearing arrive with space already in them (spatial order 0.90, 0.86, 0.89) and get orderly maps; smell arrives with none (0.11) and gets a fractured one that has to reach everywhere. The nine-banded armadillo spends its room on smell — hippocampus twice its expected volume, neocortex smaller by a factor of 2.8 — and is not a failed monkey. The single leap is the last spread, which moves from the model is false to what the phrase does when it is said to a person about their own nervous system. Companion to No. 36, which takes the same fault one level down, inside the neocortex.

Spring 1958: a two-and-a-half-minute instrumental is kept off American radio for inciting violence. It has no lyrics and no voice at any point — the Library of Congress, adding it to the National Recording Registry, says plainly that stations banned it “for its incendiary title.” And the title was not the musician’s. The record was made as Oddball; the word Rumble was supplied afterwards by somebody who had heard it and thought of a street fight, and the two best accounts of who that was — the Library of Congress and Rolling Stone — name different people. That joint is tagged contested and left that way, because both accounts agree on the thing that matters: the name came from the listening side and was then treated as the record’s content. Then the chain goes back. The sound was an accident first — a record hop in Fredericksburg, a borrowed improvisation, and somebody sticking a microphone against an amplifier — and in the studio it had to be manufactured, with a pen through the tweeters. Not a pencil through the speaker cone: that is the version in the New York Times obituary and in Rolling Stone, and Wray’s own account contradicts it. What he was undoing had thirty years and a great deal of genius behind it: Harold Black, on the Lackawanna Ferry on 2 August 1927, inventing negative feedback on a blank page of his morning paper to stop vacuum-tube amplifiers compounding their distortion down a long-distance line. And the word had held its verdict since the 1580s — distortion, from distorquere, to twist completely, the same root as torque and torture; the signal sense dates to 1887. So one oscilloscope reading has two names, and only one of them is a fact. Why a wordless record could be dangerous at all is the 1954 apparatus: Wertham’s Seduction of the Innocent, the Kefauver subcommittee hearings, the Comics Code — and Carol Tilley’s 2012 paper, from an archive opened in 2010, documenting that the case notes underneath it were falsified. Whether any of that reached this record is tagged contested and refused, because no documented line runs from the Senate to a programme director in Boston. The chain would rather stop than feel inevitable.

The dog was first — before any plant, before agriculture, the only animal we domesticated while still hunting and gathering. It has lived in our houses ever since, and after 158 years of argument nobody can say where that happened, when, or how many times. Darwin did not wonder; he concluded. The Variation of Animals and Plants under Domestication (1868, vol. 1, p. 33) ends the chapter with a verdict — “the balance of evidence is strongly in favour of the multiple origin of our dogs” — and the retelling that turns his stated position into a musing is doing him a favour he did not ask for. Being wrong out loud is not the same as not having decided. In 1997 mitochondrial DNA from 162 wolves and 140 dogs settled the ancestor — wolves, only wolves — and put the origin past 100,000 years ago. Two of that paper’s nine authors were Peter Savolainen and Robert Wayne, who would spend the next fifteen years publicly contradicting each other: the dog wars began inside one set of results. Savolainen’s camp read 1,543 living dogs and found all ten haplogroups in one region only, south of the Yangtze, under 16,300 years ago. Wayne’s answer was to stop reading living dogs: 18 prehistoric mitochondrial genomes, a European origin, 18,800 to 32,100 years ago. Each accused the other of the same error in mirror image, and by 2013 Wayne said in print that they had reached an impasse and did not interact much. Meanwhile a skull sat in a Brussels drawer. Mietje Germonpré re-measured a Goyet Cave cranium catalogued as a wolf, called it a dog, and dated it to 31,680 ± 250 years — twice the age of anything else then called a dog. In 2015 Drake, Coquerelle and Colombeau put it through 3D geometric morphometrics and concluded it is a wolf. Neither side has conceded, and the joint stays contested, because early domestication is a process and a method that must return one of two labels will return one whether or not the animal was one. Then two outsiders bought a truce. Greger Larson and Keith Dobney, pig people, argued to funders that without dogs you have no other domestication and no civilisation, raised three million dollars, and got fifteen collaborators — Wayne, Savolainen and Germonpré among them — into a two-day meeting Dobney remembers as tense. The dog wars did not end with a winner; they ended with shared samples. What followed: a dual origin, suggested and hedged in 2016; then in 2020 at least five ancestry lineages by 11,000 years ago, a single common ancestry from a now-extinct wolf population, and the strangest finding of the lot — gene flow running out of dogs into wolves and almost none coming back, undetectable at current resolution. Why is marked leap and left there: the explanation on offer came from an interview, not a paper. And in March 2026 the oldest dog genome ever read, 14,200 years old from Kesslerloch in Switzerland, killed one hypothesis and then said the quiet part: “Beyond this, our results cannot resolve where in the world, and how many times, dogs were domesticated.” Seventeen spreads, thirteen marked joints, a running two-rail spine — bone against genome — and a “Not:” refusals set that declines to cash the leap. A rhyme, not a proof.

The correction was published in the same paragraph as the problem, by the person who caused it. Schrödinger introduced code-script in What Is Life? and immediately added that chromosomes “are law-code and executive power — or, to use another simile, they are architect’s plan and builder’s craft — in one.” Watson, Crick and Wilkins all credited the book; Crick wrote to him in August 1953 to say so. So the qualifier was not lost in transit — it was on the page the people who carried the idea forward were reading. The chain’s account of why it fell off anyway is the opposite of a conspiracy. One half became testable within a decade and the other did not: the copying mechanism produced tables, prizes and a three-billion-dollar project, while how a sequence becomes a body produced almost nothing measurable for fifty years. A qualifier survives only if something can be done with it. The Modern Synthesis then filed development as an “incidental blackbox”, blueprint became the ordinary word — with no first use anybody can source, which is why that joint is graded contested rather than documented — and in June 2000 the genome was announced as “our own instruction book, previously known only to God”. Then the prediction failed twice, from inside. A gene count estimated at up to 100,000 came in near 20,000, and an onion turned out to carry five times more DNA than a person while onions vary among themselves more than they differ from us. Pigliucci declared the metaphor finished in 2010 and proposed developmental encoding — which is Schrödinger’s sentence arrived at again, from the other end, sixty-six years later. The last three links go where this house lives. In 2018 the retired metaphor was a book title; the objection quoted is a working behaviour geneticist’s, because this is not an argument against heritability. In 2021 Spectrum 10K began collecting Autistic DNA, was paused within weeks after we objected, consulted on for years and closed in 2025. And the twelfth link is marked leap, because the chain cannot show the metaphor caused the searching — only that a plan is the kind of object that can contain an error, and that the picture is wrong on the evidence of the three links before it.