History is usually told as the story of what we built.This is the story of what we refused to let break.
Entries marked MYTH are stories, not records. Entries marked FROM THE RECORD or DOCUMENTED rest on documented sources, cited at the end of the book. Where legend and record meet, the text says so.
Prologue
Mythic time · China — earliest written telling: the Huainanzi, 2nd century BCE
MYTH
Prologue — When the Sky Broke
Before there was history, there was repair. In the oldest Chinese telling, the sky itself broke — the cosmic pillar Buzhou was shattered, the heavens tilted, the earth flooded. And it was Nüwa, the goddess, who mended it: she smelted five-colored stones and patched the torn sky, and set the world right again.
The earliest written telling we have comes from the Huainanzi, in the second century BCE — but the story is older than writing. Most civilizations begin with a creation myth. China begins, in a sense, with a repair.
c. 2500 BCE · Mesopotamia — the Royal Cemetery of Ur
FROM THE RECORD
1 · The First Joint
Five thousand years ago, in the Royal Cemetery of Ur, goldsmiths were already doing what every repairer still does: joining what was separate, strengthening what would break. Archaeometallurgists studying the goldwork — the Queen’s lyre among it — found two distinct soldering techniques in the Bronze Age metal: metallic solder, and solder formed by chemical reaction at the joint itself.
There is no single inventor of soldering, no dated breakthrough. There is only this: the moment human hands learned that heat and metal could make two broken edges hold as one. The first joint in the longest craft in history.
The Queen’s gold lyre, Royal Cemetery of Ur, c. 2500 BCE (Iraq Museum; the head is a modern replica) — openly licensed via Wikimedia Commons.
1st century BCE – 1st century CE · Rome — mechanism confirmed by MIT research, 2023
FROM THE RECORD
2 · The Concrete That Heals Itself
The Pantheon’s dome has stood for two thousand years — the largest unreinforced concrete dome on earth, still holding. For centuries we credited Roman engineering. Then MIT researchers found something stranger: the Romans hot-mixed quicklime into their concrete, leaving reactive lime clasts scattered through the walls. When a crack lets water in, the clasts dissolve and recrystallize as calcite. The concrete heals its own wounds.
Rome did not merely build to last. It built materials that knew how to mend themselves. Two millennia later, we are still relearning the recipe.
The interior of the Pantheon’s concrete dome, Rome — CC0.
Every empire falls. Almost nothing built by human hands has worked continuously for twenty-two centuries — except this. In the third century BCE, Li Bing and the laborers of the Qin state split the Min River with a fish-mouth levee: no dam, only the current’s own logic — an inner channel for irrigation, an outer channel for flood, a spillway that flushes its own silt.
The Chengdu Plain has been watered by it ever since, through dynasties, wars, and revolutions — because every generation repaired it. Dujiangyan is not a monument. Monuments are finished. It is a maintenance program that never ended: the longest-running repair in human history, and still on duty.
Panorama of the Dujiangyan Irrigation System, Sichuan — CC BY-SA 4.0.
In late-sixteenth-century Japan, the tea master Sen no Rikyū taught that a bowl’s worth lay not in perfection but in presence — wabi-sabi, the beauty of the imperfect, the impermanent, the incomplete. The lacquer craft was older than the philosophy: broken things had been joined with tree resin for millennia before gold ever touched a seam. But Rikyū made the world in which the golden seam could be understood: broken pottery joined with urushi lacquer, the seams dusted with gold.
The damage is not hidden. It is illuminated — the object’s history made visible, and more precious for it. (The beloved tale of a shogun’s bowl sent to China for repair is legend, not record; the craft itself is real, and old.) Few doctrines of repair were ever stated this plainly: a repaired thing can be more beautiful than a whole one.
Kintsugi-repaired sake cup — openly licensed via Wikimedia Commons.
China answered the broken bowl differently: not with gold, but with iron. The staple repairer drilled paired holes on either side of a crack and set in small metal cramps — staples — drawing the shards tight, then sealed the seam. Bowls, vases, and teapots too loved to discard were given a second life, held together by visible stitches of metal.
No one knows exactly when the craft began; claims reach back centuries, and the record keeps its silence. What survives is the principle, and it is unmistakable: a mend need not be beautiful to be honorable. Sometimes repair is not philosophy. It is refusal — the refusal to let go.
Underside of a Meissen plate with 19th-century staple repair — the same mechanical principle as Chinese staple repair (no openly licensed photograph of a Chinese staple-repair piece exists) — CC BY-SA 3.0.
Classical antiquity · Athens — recorded by Plutarch (c. 46–119 CE)
FROM THE RECORD
6 · The Ship That Was Rebuilt
Plutarch tells us the Athenians preserved the ship of Theseus for centuries, replacing each rotted plank with new timber — until the philosophers asked the question that still haunts every restorer: when every plank has been replaced, is it still the same ship?
The paradox was never really about the ship. It is about us: about whether a mended thing keeps its soul, about what repair does to identity. Every conservator who lifts a brush, every owner who patches a hull, works inside this question. Theseus sails on — not because the wood survived, but because the mending never stopped.
The reconstructed Athenian trireme Olympias — an illustrative stand-in, not the mythical ship — public domain.
For most of history, repair was not a movement. It was Tuesday. In Edo-period Japan, farming families who could not afford new cloth layered rags into boro — patch upon patch — and held them together with sashiko, the small running stitch that turned reinforcement into art. In Europe, girls learned darning on samplers; an English sampler from 1785 survives, covered in neat mended squares — homework in the craft of keeping.
Nobody wrote manifestos about it. They simply could not afford to throw things away — and so they developed, stitch by stitch, one of humanity’s deepest technologies: the refusal to waste.
Child’s sleeping mat (boro), layered and stitched cotton, late 19th century, Japan — Cooper Hewitt, Smithsonian Design Museum, public domain.
In June 1941, Britain introduced clothes rationing. Every adult received a small annual allowance of coupons, and a single suit could swallow a large share of it. The arithmetic was simple and brutal: the nation’s cloth, thread, factory capacity, and shipping were needed for the war, and what reached civilians would not be replaced. Rarely in an industrial society has an entire population been told, as a matter of policy, that their clothes had to last.
The state did not merely ration. It taught. In the autumn of 1942, the Board of Trade gave official backing to a “Make-Do and Mend” scheme, launched with the Women’s Institute, the Women’s Group on Public Welfare, and the National Council of Social Services. Instructional pamphlets followed — the famous Make Do and Mend booklet of 1943 — full of practical counsel: how to darn, patch, and alter; how to unravel old knitwear and reuse the wool; how to turn worn garments into something else entirely. Classes in dressmaking, needlework, and knitting ran across the country. And presiding over it all was Mrs Sew and Sew, a wide-eyed animated doll who demonstrated sewing tips in pamphlets, on posters, and in Ministry of Information cartoon films — “a woman of resource and ideas,” the newsreels said. A surviving leaflet in the Imperial War Museum’s archives, “Darning Do’s and Don’ts,” tells its readers that “a neat darn is a real badge of honour.”
This is the hinge chapter of the modern half of our story. For most of history, repair was what people did because they were poor — the boro stitchers of Edo Japan, the darning-sampler schoolgirls of England. “Make Do and Mend” made it something else: a civic virtue, administered by the state, performed as patriotism. Thrift became a contribution to the war effort, measurable in coupons unspent and wool unraveled. The government that could conscript men could also conscript needles — and did, complete with a mascot. It worked, in the way such things work: not because the posters were charming (Mrs Sew and Sew was, by one modern account, “rather terrifying”), but because the alternative was going cold.
The scheme ended. Rationing ended in 1949. Affluence returned, and with it the century’s great forgetting — the planned obsolescence this volume has already described. But the phrase survived, and it has lately come back: invoked by climate campaigners, by mending circles, by everyone who has noticed that “make do and mend” describes, almost exactly, what the twenty-first century needs to learn by choice rather than by compulsion. The wartime state taught darning because it had to. The question this chapter leaves is whether a peacetime culture can learn it because it wants to.
Source note: rationing date (June 1941), Board of Trade backing (autumn 1942), partner organizations, the 1943 pamphlet, Mrs Sew and Sew’s pamphlet/film appearances, and the IWM “Darning Do’s and Don’ts” leaflet are documented in the sources above. No claim is made about the scheme’s measurable effect on consumption — the archives record the effort, not the outcome.
Part III — The Chemistry of Holding
1839 (patented 1844) · New England, USA
FROM THE RECORD
9 · The Accident on the Stove
In 1839, Charles Goodyear dropped a lump of rubber mixed with sulfur onto a hot stove — and instead of melting into stink, it charred into something tough, elastic, indifferent to heat and cold. Vulcanization: the accident that made the modern world possible, from tires to seals to every rubber patch that ever held.
(The lone-genius version is tidier than the truth — Nathaniel Hayward had been working sulfur into rubber too — but it was Goodyear who systematized it, patented it in 1844, and died, characteristically, in debt.) Repair entered the industrial age the way it enters most ages: by mistake, and then by persistence.
In 1881, Auguste de Méritens used an electric arc to join lead battery plates; by 1885, Benardos and Olszewski had patented a practical carbon-arc process in Britain. Then came the war that welded the world: America’s Liberty ships, 2,711 of them, assembled from prefabricated welded sections — the first launched in 1941, built faster than the U-boats could sink them.
When some broke apart in freezing seas, the Cambridge engineer Constance Tipper proved the culprit was brittle fracture in the steel — not the welds. She exonerated the joint with science. Welding had joined the hulls of an armada; Tipper’s test made sure the joining was honest.
Victory and Liberty ships fitting out, California Shipbuilding Corp., April 1944 — U.S. War Shipping Administration, public domain.
In 1936, in Zurich, the Swiss chemist Pierre Castan began coaxing a new kind of molecule into existence — a resin that, mixed with a hardener, would cure into something harder than the materials it joined. Patented in 1938, taken up by Ciba, and launched in 1946 as Araldite, epoxy became the quiet infrastructure of the modern repair: the adhesive that holds aircraft, boats, and a million mended things together.
(Some accounts date Castan’s work to 1934; the verifiable record starts in 1936 — this book keeps to the record.) A century of chemistry, and we are still, at bottom, doing what the goldsmiths of Ur did: making two broken edges hold as one.
Harry Coover met cyanoacrylate in 1942, while working on wartime gunsights — and dismissed it, because it stuck to everything. Nine years later he met it again, and this time he understood: a glue that polymerizes on contact with the faintest trace of moisture was not a failed plastic but a perfect adhesive.
Filed in 1954 and sold from 1958 as Eastman 910, it became “super glue” — the household repair in a tube. Twice discovered, twice almost discarded. Some materials have to wait until the world is ready to need them.
On February 10, 1943, Vesta Stoudt — a mother working the line at an Illinois munitions plant, with two sons in the Navy — wrote to President Roosevelt. The paper tape sealing ammunition boxes was failing in the field, she explained, with diagrams: what the soldiers needed was a strong, waterproof cloth tape.
The War Production Board listened. The tape was made. It sealed the boxes, and then it sealed everything else — the most democratic repair material ever issued, born not in a laboratory but in a mother’s worry. Every strip of duct tape still carries, invisibly, her return address.
A duct-taped car, still on the road — CC BY-SA 4.0.
Nothing broke. That is the point. In 1923, Alfred P. Sloan became president of General Motors, and the most consequential breakage in this volume’s modern half never involved a single broken part. Sloan looked at Henry Ford’s Model T — one car, black, unchanged for years — and saw not efficiency but a missed opportunity. His answer, introduced across GM’s lines beginning that year, was the annual model change: every year a new model, new styling, new reasons the old car looked old. The Chevrolet Superior tells the story in miniature: Series B in 1923, then F, K, and V through 1926 — a nickel-plated frame here, drum-shaped headlights there, never a revolution, always a reason.
The instrument was not the engineer but the stylist. Sloan’s GM built a pricing ladder — Chevrolet, Pontiac, Oldsmobile, Buick, Cadillac, “a car for every purse and purpose” — and dressed each rung in yearly fashion, borrowing the logic of the clothing industry, where no one would be caught dead in last year’s cut. The changes were deliberately superficial: the Series K was not fundamentally better than the Series B, but its sun visor told the neighbors you had not settled. This was obsolescence without breakage. The machine did not fail. The desire was manufactured, on schedule, every autumn.
It worked. By the early 1930s, GM had taken industry sales leadership from Ford and would hold it for more than seventy years. The annual model change became the operating system of twentieth-century consumer capitalism: the new as a calendar event, dissatisfaction as a product cycle. Three decades later the machinery got its theorists — Brooks Stevens naming “planned obsolescence” in 1954, Vance Packard anatomizing it in The Waste Makers (1960) — and eventually its historian, Giles Slade’s Made to Break (2006). The full telling belongs to Volume VIII of these Annals; here it is enough to say the villain’s history exists, in full. It is this book’s job to write the hero’s.
Sloan’s invention outlived the automobile. The annual model change is now the template for phones, fashion, software — anything that can be made to feel old without breaking. Against it stand the chapters on either side of this one: the dorm-room manuals, the café tables, the right-to-repair laws. They are not nostalgic. They are the counter-schedule — the insistence that a thing’s life should be measured in use, not in model years. Nothing broke in 1923. Everything since has been an argument about whether that was progress.
Source note: Sloan’s 1923 presidency and the annual model change are documented across the cited sources; the Chevrolet Superior Series B (1923) through V (1926) progression follows the cited model-year history. The chapter avoids “invented” as a sole-origin claim — Sloan pioneered and systematized the practice; the 1927 date appearing in some popular sources refers to later formalization.
2003 · California, USA
DOCUMENTED
15 · The Dorm-Room Manual
In 2003, in Yosemite Hall Tower 1 at Cal Poly San Luis Obispo, Kyle Wiens had a broken iBook G3. The power port was busted. The machine had cost him $1,600 — serious money for a college student — and there was no repair manual anywhere, because Apple did not include repair documentation with the device, and what little information existed online kept disappearing under Apple’s take-down notices. So Wiens and his friend Luke Soules did what everyone does when there are no instructions: they started the hard way. They tinkered. They fiddled. They broke some tabs and lost a few screws. And they fixed it.
The fix created a second problem: parts. The pair tried repairing other laptops and could not find components, so they bought a broken computer on eBay and harvested parts from it — and began selling the parts themselves. But every part they sold still required the buyer to do things the hard way. So they did the obvious thing the manufacturers would not: they wrote repair manuals, photographed each step, and posted them online for free — the first to pair the parts with free, photographic, step-by-step manuals anyone could follow. A person with no technical background could take apart a Mac by following them. The parts came with knowledge attached. In marketing language, that is called a value add. In plain language, it was the manual that should have been in the box.
iFixit’s founding complaint — that paying for something did not mean you actually owned it, down to the last screw — turned out to be the defining consumer grievance of the digital age. The company grew out of the dorm closet into a real business, but the more important growth was elsewhere: the manuals became a commons, the teardowns became journalism, and the argument became legislation. Two decades on, iFixit works with Repair.org and others pressing lawmakers, and state after state has passed self-repair rights requiring manufacturers to provide the repair information buyers were always owed. The lineage the founders claim runs back to 1968, to Stewart Brand’s Whole Earth Catalog — the mail-order toolbox of the DIY age. The through-line is the same sentence, fifty years apart: you should be allowed to fix what you bought.
Wiens would go on to testify before Congress for the right to repair. The dorm-room repair is now twenty years in the past, but its premise has not aged: every device you own contains a manual the manufacturer chose not to write. iFixit is what happens when the users write it themselves — and then refuse to stop. If this volume has a thesis about the modern era, it is here: the opposite of planned obsolescence is not durability. It is documentation.
Source note: the 2003 date, the iBook G3 and its busted power port, the broken tabs and lost screws, the eBay parts harvesting, the Yosemite Hall Tower 1 dorm, the $1,600 price, and Apple’s take-down notices are all documented in iFixit’s own published histories (20-year retrospective and community history).
2009 – present · Amsterdam, worldwide
DOCUMENTED
16 · Coffee, Cake, and Screwdrivers
Martine Postma was a journalist in Amsterdam writing a guide to producing less household waste. Somewhere in the research, she noticed something: most of her tips were not really about waste. They were about repair. The guide was telling people, in so many words, to fix things — and the people she met had, as she later told AFP, “simply lost the culture of repairs.” The break this chapter opens with is not a broken object. It is a broken habit, in an entire society.
On October 18, 2009, in Amsterdam-West, Postma opened the doors of the first Repair Café. The format was disarmingly simple: neighbors brought broken toasters, torn jackets, and dead electronics; volunteers fixed them for free, over coffee. No manifestos — just tables, tools, and the radical suggestion that mending could be social. The event had been planned as a one-off. It was such a success that Postma kept going, organizing more sessions across Amsterdam through 2010, and then established a foundation — Stichting Repair Café — to spread the model. By March 2012 there were twenty groups across the Netherlands; by the foundation’s fifth anniversary, in March 2015, more than seven hundred cafés worldwide. The work is always free, always communal, always in public.
The Repair Café arrived at the exact moment the previous chapters of this volume describe: the age of planned obsolescence, of sealed devices and missing manuals. iFixit answered that age with documentation; the Repair Café answered it with ritual. Postma’s own diagnosis is the chapter’s key sentence: people had not become lazy or stupid — they had lost the culture, the ambient knowledge of how things come apart and go back together that every previous generation absorbed by watching. The café restores more than objects. It restores the social standing of people with practical skill, and it transfers that skill across the table, one repair at a time. The fix happens in public, with cake, because repair was never meant to be a solitary shame. It was meant to be what neighbors do.
The movement has now spread to every continent — thousands of cafés, from New York to Tokyo — and its premise has entered the policy conversation: things should be made for repair, not for the landfill. If iFixit digitized the manual, the Repair Café re-taught the gathering. Between them they describe the two halves of the modern repair revival: the knowledge, and the company. The journalist who set out to write about garbage ended up founding a community. That is usually how it goes. Repair, done honestly, was never really about the objects.
Source note: the October 18, 2009 date and Amsterdam-West location are confirmed by Postma’s own account as quoted on repaircafe.org; her former-journalist background, the “lost the culture of repairs” remark, and the twenty-group figure are from the March 2012 AFP interview; the foundation’s establishment and international spread are documented in the cited academic paper. The 2016 “250,000 kg diverted” figure from earlier research could not be re-verified and is not used here.
2019–2024 · European Union
FROM THE RECORD
17 · Repair Becomes Law
In 2019, the European Commission wrote repairability into ecodesign rules for household appliances — the first time a major regulator demanded that things be made fixable. In March 2023 came a dedicated Right to Repair directive; it entered into force on July 30, 2024, giving Europeans a one-year guarantee extension after every repair, and a public platform for finding repairers.
Five thousand years after the goldsmiths of Ur, mending is no longer just craft, philosophy, or resistance. It is law.
This book ends where history always ends: with the present.
The history above holds the repairs we remember. But most repairs were never written down — the darned sock, the soldered wire, the bracket glued back at midnight. History kept the cathedrals and forgot the kitchen tables.
The sky was mended by a goddess. The rest was mended by people like you.