Annals of Repair
Vol. VI — Iron Stitches
Two holes, one iron stitch.This is the history of the staple.
Entries marked MYTH are stories, not records. Entries marked FROM THE RECORD rest on documented sources, cited at the end of the book. Where legend and record meet, the text says so.
This volume is about the staple — the small iron stitch that held broken porcelain together for centuries before modern glue.
What follows is the record of an unglamorous repair: the Chinese texts that describe it, the traveling menders who practiced it, the European collectors who prized it, and the ancient holes that prove how old it is.
For years, craft histories have repeated an attractive claim: that the earliest picture of a Chinese staple-repairer — a jujiang at work — appears in Zhang Zeduan’s Along the River During the Qingming Festival, the great Northern Song panorama of street life. It would be a lovely origin: the craft caught in ink in the early twelfth century, mid-cry, mid-repair. But the claim does not survive a look at the evidence. Researchers note that no such figure appears in the Northern Song original at all; a similar figure turns up only in Qiu Ying’s Ming-dynasty copy, and the image most often circulated as “proof” actually comes from Zhou Kun’s Qing-dynasty Village Market Life album. A Song origin for the craft, on this evidence, is asserted often and documented never.
What is documented begins in ink of a different kind. In the tenth juan of Li Shizhen’s Bencao Gangmu, the Ming pharmacopoeia, the entry on diamond reads: its sand “can drill jade and repair porcelain” — the drill and the mend already paired in a single line. Half a century later, in 1558, the thirty-seventh year of the Jiajing reign, the painter Wang Wen set a tea-brewing scene on paper, and in it stands a jar unmistakably joined with metal staples: the craft not described but depicted, mid-life, among the teaware. Then in 1583, the eleventh year of Wanli, Matteo Ricci — the Italian Jesuit newly arrived in China — wrote in his China journals a comparatively detailed account of the staple-repair process itself. Three Ming witnesses, three media: the materia medica, the painting, the foreigner’s notebook.
Taken together, they describe a trade already mature by the sixteenth century: itinerant, skilled, equipped with diamond drills, serving households whose good ceramics were worth a mend rather than a replacement. A technical study from the Palace Museum pushes the craft’s existence back further, holding that the trade was established no later than the Yuan period — but the specific documents behind that dating have yet to be brought into the open, and the claim waits on evidence. The honest account, then, is this: firm written and painted evidence from the Ming; a Yuan horizon claimed by scholarship and still to be shown; everything earlier an open question.
The record only thickens from here. Later centuries add what the Ming could not: photographs of the itinerant repairman on Beijing streets, film of the bow drill in motion, and finally the craft’s entry onto intangible-cultural-heritage lists — Shanghai’s Fengxian district in 2019, Ürümqi in 2025. Each layer of evidence corrects the one before it. The Qingming-scroll claim may have been wishful, but it was also unnecessary: the true record, beginning in Ming ink, is sturdy enough without it.
The craft’s truest archive is not paper but the street. In photographs published as scenes of early-Republican Beijing, a man sits at the roadside beside a low wooden chest, his tools laid out around him: the itinerant ceramic repairman, the jujiang, at his station. The exact provenance of these photographs has not been established — they circulate through online reposts of old-Beijing imagery, and the archive behind them is still unnamed — but the figure they show is corroborated from every side: the shoulder pole, the paired cabinets, the cymbals. This was a man whose shop was carried on his back.
His kit was standardized by centuries of use. Across his shoulder went the flat carrying pole; from each end hung a small lidded wooden cabinet — one for tools, one for the little bellows needed when the job was a pot or a vat rather than a bowl. From the pole hung his trade sign: the triangular pennant flag, and the small copper cymbals or double-faced hand drum he struck to announce himself down the lane. In his hands, the bow drill of Henan — a bamboo bow driving an iron spindle tipped with diamond — with the hammer, the shears, and the tweezers that set each staple. The work was done on the doorstep, in minutes, for coins.
Every region had its name for the man and its version of his cry. In the Mandarin-speaking north he was the jujiang, the itinerant bowl-mender, and his call — “Basins to mend! Bowls to mend! Great vats to mend!” — is the most quoted street cry in the literature of old trades. Lanzhou heard “Bowls mended! Dishes mended!”; the Dongxiang nail-smiths called “Pots mended! Bowls mended!”; Shaoxing’s version ran to a full rhyme about vats and jars, mended whether ground to powder or knocked to pieces; Taiyuan said “Dishes and bowls mended,” Suzhou “Bowl-mender here,” southern Min “Pottery-mender,” Cantonese “Wok-mender.” The trade left proverbs behind it — “Without the diamond drill, don’t take the porcelain job” — and riddles, and a turn in Shaanxi opera. The Beijing folk painter He Daqi, who painted the old street trades from memory, gave the repairman his verse: the copper discs clink, basins bowls and great vats are mended, the bamboo bow draws the diamond drill taut, and broken porcelain is made whole again.
The shoulder pole is gone from the streets; the cry survives in recordings, in folk painting, in the memory of old Beijingers. What remains of the trade has moved indoors — to workshops and heritage demonstrations — and onto the lists: Shanghai recognized the craft in 2019, Xinjiang in 2025, and the Dongxiang bowl-smiths of Linxia hold their own provincial listing. The itinerant figure has become what he always was beneath the romance: evidence that a whole economy of care once walked the streets on foot.
Every staple repair begins with a decision that looks like violence: the drill. The crack is there; the vessel is already broken. The repairman does not reach for glue. He reaches for the diamond point and bores two small holes flanking the fracture — the first deliberate wounds, made so the vessel can hold together again.
The sequence, as the craft’s own literature describes it, is a small liturgy. First, the shards fitted and matched back to their places; then the pieces tacked with adhesive and bound tight with rope. Then the staple positions marked along the crack, each one planned against the fracture’s run. Then the drilling itself: a starter pit picked with the diamond awl, then the drill — the hand-pulled drill of Shandong, the bow-driven drill of Henan, the flywheel drill of Hebei — boring a dovetail hole on each side of the crack, wider within than at the mouth so the staple’s legs cannot pull free, and never through: about two-thirds of the wall’s thickness, no more. Then the staple cut from sheet metal as a small lozenge, forged, and set with tweezers and hook into the paired holes, hammered flush with the body. Then the seam sealed with oil putty, or egg white or rice juice worked with gypsum, lime, and porcelain powder. And finally the test: the repairman asks the owner for a bowl of water, and drinks, to prove the mend holds. A modern Sichuan practitioner, Wang Peibo, describes the same work in five plain stages: bind the fragments, plan the staples, drill, set the staples, seal.
The staple itself — also called the “leech nail” or “centipede rake” — is a flat saddle-shaped bridge with hooked legs, cut and forged on the spot to fit the holes just drilled. To modern eyes it looks exactly like an office staple, and it works on the same principle. Iron and copper were the common metals; the Palace Museum’s technical study notes iron and copper on surviving Ming pieces and reserves gold and silver for the finer work of the Ming and Qing. The tools around it — the awl, the three regional drills, the setting hammer, the hook knife for the putty — form a kit barely changed in centuries, because the physics has not changed either.
In 2020 a video circulated of a Xi’an craftsman giving a light bulb “surgery” — drilling glass, setting a staple — virtuosity as proof that the technique answers to the hand, not the material. The craft’s afterlife is this: once the economy of doorstep mending, now the discipline of the specialist, its motions identical. The drill still stops at two-thirds. The staple still spans the crack. The water still gets drunk.
Turn the dish over and the story is on the back. A nineteenth-century Meissen plate — the pride of Saxon porcelain — photographed from below, its underside tracked with rows of small metal staples: the break mapped in iron. The image is the one modern craft histories reach for when they need the European face of the staple. It is published, captioned as a nineteenth-century Meissen dish repaired with staple technique; the catalogue record behind the photograph has not been traced, and the dish’s own museum entry, if one exists, is still to be found. What is beyond doubt is the practice it shows.
For two centuries, roughly the eighteenth to the twentieth, stapling was Europe’s ordinary answer to broken china. The work crossed the continent under its own names: in England the “china mender” or “china riveter,” in France the recouseur or raccommodeur de faïence et de porcelaine, in the Netherlands the porseleinkrammer, in Germany the Porzellanflicker or Porzellannieter. The method — drill paired holes, span them with wire or staples, fill and seal — was the same one the Chinese menders practiced, arrived at independently. Two holes and a bridge is not an idea that needs to travel; it is what the break suggests to any hand.
A 2016 exhibition of mended ceramics surveyed the practice from the seventeenth century to the mid-twentieth: not only plates but punch bowls, teapots, and export wares — the breakage of whole households, mended and kept. The exhibition’s premise is the modern scholar’s correction to the old story of consumption: repair was not the end of an object’s life but its middle. In England the trade was thriving by the 1700s, with advertisements for mending services surviving from as early as 1725. Broken did not mean finished. It meant a visit from the mender.
The Meissen dish survives as an image more than as an object — cited, captioned, passed from one craft history to another. That is fitting. The staple’s European afterlife is largely photographic and museological now: shelves of mended plates in folk museums, X-rays showing the old holes, the craft itself extinct as a trade and preserved as evidence. The dish, turned over, still makes its argument: that for two hundred years Europe mended its best porcelain the way China mended its bowls — with metal, and patience, and no shame.
In 1773 a French trades encyclopedia devoted an entry to a worker it defined with unusual respect: the raccommodeur de faïence et de porcelaine — “a person who can make broken china and faïence usable again.” The entry then did something rarer still: it described exactly how he worked, as if the trade deserved the same dignity as any guild craft.
The method was the through-and-through rivet: holes drilled along the break, the fragments drawn together with wire or staples, and the holes then filled and disguised with a mixture of olive oil, egg white, ground glaze, and chalk. The encyclopedia’s author, Pierre Jaubert, credited the discovery of riveting to one Delisle of Normandy, early in the eighteenth century. A century later the Englishman Charles Godfrey Leland, in his 1896 Manual of Mending and Repairing, credited a “humble and modest workman named Delille, of the little village of Montjoye, in Normandy” — and then demolished the idea that the method had an inventor at all. One might as well apply for a patent for mixing brandy with water, he wrote, since there was probably never a person with wire who did not know enough to mend a broken jar by sewing it. The attribution wobbles between Delisle and Delille; the technique does not wobble at all.
The trade did not please everyone. The manufacturers of faïence — whose business depended on replacement — went to court in an attempt to have the menders banned. The menders were defended by their customers: ordinary citizens who recognized the usefulness of the work, and the raccommodeurs, thus supported, were granted the legal right to repair china and porcelain. It may be the first recorded right-to-repair case. By then the itinerant china mender was a familiar figure across several European countries, ranked in the street economy with the traveling knife grinder and the tinker — the man who arrived with his tools and left things better than he found them.
As early as 1877 a French history book was already calling Delisle’s “discovery” what it was: the rediscovery of an old technique. The craft was older than its inventor, older than the encyclopedia that described it, older than the court that protected it. What survives of the episode is the precedent: that mending was once important enough to sue over, and popular enough to win.
A seventeenth-century Dutch delft ewer lost its spout and its handle — the two most vulnerable parts of any pouring vessel. What replaced them was not delft at all, but decorated metalwork: a jeweled replacement, made for the break rather than in spite of it. The ewer was shown in a 2016 exhibition of inventive repair, where it stood for the far end of the mender’s art — the point at which repair stops apologizing and starts designing.
Fine mending has its own economy. The trade card of Edward Coombs, china burner and mender of Bristol, advertised that teapots which poured badly could be made to pour smoothly “by tipping them neatly with Silver” — the repair as upgrade, the break as an occasion. In China the craft formally distinguished the two registers: coarse work — the iron staples of household pots and bowls done at the doorstep for coins; and fine work, where the staples themselves could be cut as decorative patterned nails, and silver or gold replaced iron. The ewer’s jeweled handle belongs to this register: the mend not hidden, but made.
Between the two registers lies the great middle of the craft: six Chinese export plates of the early eighteenth century, mended with hand-forged metal staples — ordinary ironwork keeping good porcelain in service, shown in the same 2016 exhibition as the ewer. The plates and the ewer together describe a complete spectrum, from the invisible economy of the kitchen to the visible luxury of the collector. What unites them is the refusal shared by every mending tradition: that a broken thing of value should be thrown away. The distinction was never only aesthetic; it was economic — iron for the many, silver and pattern for the few — and it is the same distinction a museum drawer enforces without meaning to, keeping the jeweled handle and losing the iron staple.
Fine work outlives coarse work — that is the museum’s bias, and the ewer benefits from it. But the coarse work was the craft’s true body: millions of iron staples in millions of bowls, nearly all of them gone with the bowls they saved. The ewer survives to speak for them, its jeweled handle glinting with the argument the plain staples made silently: the break is part of the object now, and it can be made beautiful on purpose.
In the Walters Art Museum stands an Attic black-figure kylix — a drinking cup — its sides painted with sirens. Around its handles run small holes, empty now, where metal pins once bridged an ancient break. The cup was mended in antiquity, with metal, by hands two and a half millennia gone. The holes are the whole document: no text, no signature, only the paired perforations flanking a fracture, exactly as a Chinese repairman would drill them two thousand years later.
Archaeology has catalogued the habit across the Greek world: copper, lead, and bronze pins and staples joining broken pottery, the metal long since corroded or lost, the drill holes remaining as the repair’s shadow. On Samos, the island’s red ware was bound with metal strips or joined with X-shaped rivets — the same two-holes-and-a-bridge logic, worked in the Aegean centuries before any Chinese text mentions the craft. The materials differ — lead where China would use iron — but the mechanics are identical, because the problem is identical: a brittle body, a clean break, and metal’s willingness to hold.
Two traditions, no contact, one solution. The Mediterranean mended its cups while the Chinese tradition was still unwritten; the Chinese tradition reached its mature form while Europe was relearning the method as a “discovery.” The staple is what engineers call a convergent solution — like the arch, like the needle, it is what the materials suggest to any hand that studies them. Its history is not a line of transmission but a scatter of independent inventions, each one obvious in retrospect.
The kylix needs no philosopher. Its holes are a plainer and better record than any anecdote — the craft’s signature, drilled in clay, outlasting the metal itself. A note on sources, kept honest: no passage about stapled pottery has been located in Plutarch’s surviving works. His famous repair story is the Ship of Theseus — a paradox about replaced timbers, told in the Life of Theseus, not a word about pinned ceramics. The Greek staple tradition therefore rests on the objects themselves, which is the better foundation anyway: paired drill holes, two and a half millennia old, still legible.
Set the two repairs side by side. A bowl joined with iron staples: the crack bridged by small metal saddle-nails, the seam filled with putty, the mend frank and unadorned. A bowl joined with gold lacquer: the crack traced in precious metal, the damage illuminated. Both refuse the same thing — the pretense that the break never happened. They differ only in what they ask the eye to do: the staple, in the humble trades this chapter follows, asks to be overlooked; the gold asks to be seen.
The staple adds nothing but force. It contributes no color, no meaning, no value beyond the holding itself — which is why it was the repair of the poor and the practical, the doorstep trade, the kitchen economy. The gold seam, the kintsugi of Japan, adds meaning to force: it declares the break an event worth marking. The Japanese bakōhan bowl — the great stapled tea bowl whose story is told in Volume IV of these Annals — sits precisely at the crossing: a vessel mended with the plain staple and then honored with a name, as if the iron had earned the dignity the gold confers by default.
The staple’s world is wide and mostly unmapped. It is documented in China from the Ming, in the Greek world from antiquity, across Europe from the eighteenth to the twentieth century — each tradition arriving at the same mechanics without the others. Against this scatter, kintsugi looks like the exception: a single tradition, in a single country, that made the mend philosophical. The staple never needed philosophy. It had economics: a bowl costs money, iron is cheap, and a household that mends is a household that endures.
Whether the staple’s afterlife lies in heritage lists and museum vitrines or in the workshop of a living practitioner, its argument has not changed. The gold says the break is beautiful. The iron, in those same humble hands, says the break is unimportant — what matters is that the bowl holds water again. These Annals have given gold its volume. This one is for the iron: for the honest, visible, unadorned mend, and for the hands — on doorsteps from Beijing to Normandy to Athens — that refused to throw the bowl away.
The staple was never beautiful. It was honest.
Two holes, one iron stitch, a vessel that could be used again. Long before the cult of the invisible repair, there was the visible one — and it held.
This volume ends. The history doesn’t.
By AOJEL · Vol. VI, September 2026.
A companion volume: Repair Lab — free guides for the repairs in your hands.
Other volumes in the series: Vol. I · Vol. II · Vol. III · Vol. IV · Vol. V · Vol. VII · Vol. VIII · Vol. IX