The Metropolitan Museum of Art holds a seventeenth-century Shigaraki tea jar, and its catalogue describes what the object is made of like this: "Stoneware with natural ash glaze; gold lacquer repairs."
Read that semicolon again. The repairs are in the medium field, in the same sentence as the clay and the glaze — listed as part of what the jar is made of.
That is the whole of this article's argument, and it is a different one from the article you have probably read. The article you have probably read treats kintsugi as a philosophy: breakage is an opportunity, the mend is more beautiful than the unbroken thing. This one is about what a museum writes down, what it does not write down, and what you can therefore look for when you are standing in front of a mended bowl. Japan's crafts are a subject of their own, and our guide to the traditional crafts takes them one at a time.
Japan, Immersive Art Experiences — which crafts you can stand in front of, and which you can handle.
Museum catalogues are written for other curators, which is why they are useful to a traveller. They are terse, they are structured, and each field answers a fixed question. What something is made of goes in one field. How big it is goes in another. Nothing is there for effect.
So the placement is the finding. Three entries, three objects, one pattern:
The Shigaraki tea jar: "Stoneware with natural ash glaze; gold lacquer repairs (Kyoto ware, Shigaraki type)."
A small jar of the late sixteenth or early seventeenth century: "Stoneware with ash glaze (Karatsu ware); gold lacquer repairs."
An eighteenth-century red raku tea bowl: "Earthenware with red slip under milky glaze (Raku ware, unknown workshop); gold lacquer repairs."
In each case the repair sits after a semicolon, in the same line as the body and the glaze (and in one entry, after a comma). It is being reported the way a material is reported.
Above those fields, some entries carry a curatorial description in prose, and that prose is not guaranteed. Two of the six carry none at all — the Karatsu small jar and the red raku bowl have a record and no description — and one describes the object without mentioning the repair. The medium field is the part that is always there. Everything above it varies by object, which is a useful thing to know before you decide a museum has told you nothing.
What none of the entries says is when any of it happened. The dates in those lines — seventeenth century, late sixteenth, eighteenth — are the dates of the objects. The catalogue does not date the repairs, and neither will this article. A bowl made in 1600 could have been mended in 1620 or in 1920, and the medium field does not tell you which.
The word says gold. Ella Tennant of Keele University, whose account of the practice is quoted at length below, glosses kintsugi as "joining with gold."
Then there is this entry, for an early seventeenth-century sake cup: "Stoneware with ash and iron glaze (Karatsu ware); silver lacquer repairs."
Silver. Same field, same construction, same museum, different metal. The catalogue names the metal, and here it names silver. What that gives you is one recorded cup in which the metal in the mend is not the metal in the name.
This matters for looking rather than for arguing. The name of the practice is not a description of what will be in front of you: a mend recorded as silver is a mend the word kintsugi does not cover.
On this cup the museum goes further than the medium field, and it is worth seeing how far. The object page describes a "subtly flared rim that has been repaired with silver lacquer," and adds a reading: "The silver repair harmonizes beautifully with the warm brown tones in a 'perfectly imperfect' (wabi-sabi) aesthetic that appealed to tea masters of the time." Position, effect, and an aesthetic judgement — all of it above the medium line, in prose.
This article works from six entries in the Metropolitan Museum of Art's collection, each of which records a lacquer repair in the medium field. It is a selection rather than a census — the collection holds more, and these six were chosen because between them they cover both metals, both cultures and four different wares. A list of mended objects is also a list of things somebody refused to throw away. Read that way, the catalogue answers a question the philosophy does not: what was worth the trouble?
Two of the six entries are not Japanese ceramics at all.
A bottle vase of the thirteenth century: "Stoneware with celadon glaze (Longquan ware); gold lacquer repairs." And a tea bowl of the same century and the same kilns: "Stoneware with celadon glaze (Longquan ware), gold lacquer repairs."
The catalogue calls both Longquan ware and gives the culture as China.
On the tea bowl, the museum draws the inference itself, and marks it as an inference: "Because kintsugi is a technique unique to Japan, its presence on this tea bowl suggests that the object was once used and cherished there." Suggests — the repair is the evidence, and the museum will go as far as a likelihood and no further. The same note adds the contrast: "In China, by contrast, broken ceramics were typically mended with metal staples."
Which puts a floor under the practice that the philosophy tends to float above. Whatever was decided about these two bowls was decided about thirteenth-century Chinese celadon that had already travelled into a different repair tradition.
There is a standard origin story, and it is worth handling carefully, because the scholar telling it here hedges and the hedge is the first thing a retelling loses.
Ella Tennant of Keele University notes the dates first: "Kintsugi was fairly widespread in Japan around the late 16th and early 17th centuries," with origins going back to the Muromachi period, "approximately 1336 to 1573."
Then the story. The third Ashikaga shogun, Yoshimitsu, "is said to have" broken a favourite tea bowl and sent it to China, and it "returned repaired with its pieces held in place by metal staples." Staple repair, she notes, "was a common technique in China as well as in parts of Europe at the time for particularly valuable pieces." The shogun disliked the result, and the search for a better answer is where Tennant says the practice begins.
Note the four words doing the work: is said to have. That is the author, a scholar, marking a piece of tradition as tradition. Drop those four words and the shogun becomes the documented inventor of a technique.
Two things in that passage are useful even if the anecdote is legend. The first is that staples were the alternative, and staples were normal — in China and in parts of Europe — for objects too valuable to discard. Mending prized ceramics was not a Japanese peculiarity; the question was what the mend should look like. The second is the period. Tennant places the spread of the practice in the late sixteenth and early seventeenth centuries. The Met's Karatsu pieces are objects of roughly that period — but that is a date for the pot, not for the mend, and the catalogue does not date its repairs. Two facts sitting near each other on a timeline are not the same fact.
Here is the practical part. You are in front of a case, the label is short, and the object has a line running through it. Three things are visible without any technical knowledge.
Follow the seam from one end to the other. A single line that runs from the rim and stops; a network of lines meeting at angles; a seam that encloses an area rather than crossing one — those are the shapes to sort what you are seeing into.
Those shapes change what you are looking at. A single line leaves you the original object, held together. A network leaves you the original object in several pieces, brought back to one. An enclosed area leaves you the object plus a region that is not ceramic at all. The medium field records the material and not the shape of the damage, so this part is yours to read.
What the seam will not tell you is why any of it was decided, and the catalogue does not supply that either. Whether a mend was meant to be admired or merely to work is not a question a medium field answers.
Look at the label's other fields, and notice what does not change.
All six of the Met entries above are classified as Ceramics. Not "Ceramics and lacquer," not "mixed media" — Ceramics, with the lacquer inside the medium line. A museum has to file each object somewhere, and the filing is a judgement about what the thing principally is. Here the judgement is that a mended pot is a pot.
That is worth carrying into any case in any museum, because labels bundle several kinds of statement into one short block. Some of it is description, some is classification, and they are not the same claim. The same habit pays off on a quiet route through Tokyo, where the information you need is public and structured and reading it properly is most of the skill.
Mended ceramics are harder to plan around than a building, for a structural reason: they are collection objects, and collections rotate. What follows is two levels, arranged by how much has to be arranged in advance.
The first level needs a ticket and nothing else. What it needs from you is a check before you travel.
The Tokyo National Museum describes its own working method on its change-of-exhibits page: "TNM Collection Exhibitions at Tokyo National Museum are rotated almost every week," with the qualification that "Some works are exhibited for a longer period of time."
So the durable thing to plan around is the room rather than the object. The museum's gallery listing gives each display with the dates it runs, and ceramics is one of the standing genre displays; the listing currently shows those running for months at a time. Read that listing against your own dates. Do not assume a particular bowl will be out, and do not assume that a mended one will be — a rotation is a selection, and the museum does not promise a theme.
The second level is not a different building. It is reading the record before you travel, which is free and which changes what a case looks like.
Every one of these Met entries states, at the top of the page, whether the object is on view and in which gallery. That single line answers the question a photograph never does — the piece you have in mind is either out or it is not — and it answers it before you have bought a ticket or crossed a city.
The rest of the page is worth two minutes. The medium field for the material. The dimensions, which are the most reliable corrective to a photograph: the Longquan bottle vase pictured earlier is under 15 centimetres tall, which a photograph of it will not tell you. And the curatorial text if there is one, which — as this article keeps finding — is where the museum says the things it will not put in a field.
Do that for two or three objects before a trip and a case stops being a wall of unfamiliar pots. You will be looking for particular things, which is the difference between visiting a museum and reading one. If you would rather have the context in which objects like these were actually handled, tea in Tokyo is arrangeable at several levels of formality.
Three things, and then the part that is still moving.
First, the catalogue puts the mend in the medium field: in the line that says what the object is made of, after a semicolon, next to the clay. That placement is a judgement, and it is the strongest evidence any of these sources give that a repair was treated as part of the object rather than as an injury to it.
Second, the metal is not always gold and the objects were not always Japanese. One Karatsu cup is listed with silver lacquer repairs, and two of the six are Longquan celadons the catalogue gives as Chinese. None of these six medium lines says when a repair was done or by whom, and the one place the catalogue ties a mend to a country it does so as a likelihood, in prose rather than in a field.
Third, when you are in front of one, follow the seam. Where it starts, where it stops, whether it joins two edges or fills a space where a fragment used to be. That is visible without the word, and it is more than a label will give you.
And the part that is still moving is the schedule. The six entries quoted here are in New York; what you can plan in Tokyo is a room rather than a particular bowl, because the Tokyo National Museum rotates its collection displays and publishes what changes and when. A mended bowl is a thing you catch rather than a thing you visit — which is a reason to read the room listing before you fly, and to build the day around a district and a theme rather than around one object you have decided to see.