A kanzashi is one of the few Japanese crafts a traveller is almost guaranteed to see and almost certain to misread. It appears in a shop window in Asakusa, in a museum case, in the hair of a woman at a New Year shrine visit — a spray of silk flowers on a pin, obviously careful, obviously expensive, and completely unexplained.
The word covers more than the flowers. It covers a family of objects, a technique with four named stages, a period when what you wore in your hair reflected things about you, and a trade that Tokyo's official travel guide now describes as "close-to-extinct."
This guide is built around four questions: what the word kanzashi actually names, how much work one piece takes and what that work consists of, what changed between the Edo period and now, and where in Tokyo you can still see the real thing.
Most English sources translate kanzashi as "hairpin" and stop. That translation is not wrong, but it is narrow enough to be misleading: it suggests a single object with a single function, when the word has always covered a range.
The Victoria and Albert Museum, describing a lacquered hairpin in its Japanese collection, sets out the range plainly: "During the Edo period (1615-1868), women used a wide variety of combs (kushi), bodkins or hairpins (kanzashi) and hairpins (kogai) in a wide range of materials, such as wood, ivory and tortoiseshell, which were most commonly decorated with lacquer."
Read that list slowly, because it is doing something English translations usually flatten. The museum reaches for the same English word twice: kanzashi is "bodkins or hairpins," kogai is "hairpins." Two separate Japanese terms, one shared English one. A distinction the Japanese holds, the English drops. That is the first sign that "hairpin" is the wrong size of word for what is in the case.
The pins themselves are worth looking at closely, because on some of them the movement is the design. Several nineteenth-century hair ornaments in the Metropolitan Museum of Art, worked in silver and silver gilt, carry a decorated head above a curtain of thin hanging elements — one of them is catalogued as silver, silver gilt and gun metal. Standing still, such an ornament reads as a brooch on a stick. Walking, it reads as something else entirely.
The range tracks a change in how hair was worn, and the museum records both. The Victoria and Albert Museum's account of that shift is specific about the timing: "From roughly the twelfth to the late sixteenth centuries, it was customary for women to wear their hair long and loose without ornamentation. Thereafter hair was put up with increasingly elaborate hair ornaments."
Hair worn loose needs nothing holding it. Hair worn up needs structure, and structure invites decoration. What began as a functional problem — how do you keep this arrangement in place — became, over several centuries, a category of craft.
The museum adds one more constraint that shapes everything in the next section: "At first hair ornaments were mostly confined to women of the elite but, from the mid eighteenth century onwards, they were increasingly available to all strata of society."
Of everything the word covers, one technique dominates what is still made in Tokyo today, and it has its own designation. Edo tsumami kanzashi was designated a traditional craft of Tokyo on 24 December 1982, and the Tokyo Metropolitan Government's craft register describes both the material and the method in terms precise enough to use as a checklist.
The register is explicit about where the name comes from: the craft is named for the process of taking habutae silk, cut into squares, pinching each square with tweezers, and folding it small to make the shape.
It then names four stages, in order.
Tachi, the cutting, uses a cutting knife and a wooden rule to cut the cloth into squares on a cutting board. Tsumami, the pinching, is done as maru-tsumami (round), kaku-tsumami (square), suji-tsumami or uragaeshi-tsumami (reversed). Fuki, the setting, uses tweezers to place each folded piece onto a paper base. Kumiage, the assembly, uses gokuten thread to bring the whole into shape.
Two of those fold types carry a described character. The register says maru-tsumami gives a soft impression while kaku-tsumami suggests strength, and that a maker switches between them to build up motifs — chrysanthemum and plum blossoms, butterflies, phoenixes.
The materials are specified as tightly as the method: the cloth is to be habutae, and the wooden body is boxwood, pear, magnolia, or timber of equivalent quality.
Nothing in that sequence can be hurried, and nothing in it scales. A flower is not stamped; it is assembled from individual folded squares, one at a time, each placed with tweezers onto a paper base. A single ornament of any complexity is a count of petals, and each petal passes through the cutting, the fold and the setting before the whole is assembled.
That is the arithmetic behind the scarcity. Tokyo's official travel guide, describing the Tsumami Kanzashi Museum, puts the same point in one sentence: "The process of making a kanzashi is very meticulous and painstaking. In fact, only very few craftsmen master this piece of art nowadays."
Hold that sentence. It returns later in this guide, when the question becomes where you can still see the craft at all.
In the Edo period, a kanzashi was doing more than decorating. What the Victoria and Albert Museum records is that the arrangement reflected things about the person wearing it.
The Victoria and Albert Museum states the range of what the arrangement communicated: "Not only did the hairstyle and its ornaments reflected the age, social class and marital status of the woman, but the hair ornaments also reflected the individual's taste."
Two things are worth keeping straight here, because popular accounts blur them. The first is that this is described as reflection, not regulation — the museum says the ornaments reflected these facts, not that a rule assigned a particular pin to a particular status. The second is that ownership widened, and the museum dates that: ornaments were "mostly confined to women of the elite" at first, and "from the mid eighteenth century onwards, they were increasingly available to all strata of society." What is documented there is availability. Whether more owners also meant more readers is a reasonable inference and not something the museum says.
That is the part worth carrying into a museum case. The object in front of you was not only decorative.
One claim follows kanzashi through English-language writing more persistently than any other: that women carried them as concealed weapons.
Here is what can be said from the sources used in this guide, and what cannot. The Tokyo craft register says the word is also said to derive from kami-sashi, hair-piercing, and records what was once believed: that a sharp, slender rod held a protective power, and that placing one in the hair was thought to drive off misfortune. Note the tense. The register presents this as a belief held in antiquity, not as a use it is vouching for — and what it describes is warding off misfortune, not self-defence.
Neither that register nor the museum records consulted here describe kanzashi as weapons. The honest position is that the shape is sharp, the sources record a protective belief attached to that sharpness, and the leap from "protective" to "concealed weapon" is one the sources do not make. Treat the weapon story as a claim in circulation rather than a fact you can check.
The register is careful about origins, and its carefulness is worth borrowing. It says the craft as practised in Tokyo is said to begin when a petal-ornament technique made in Kyoto reached Edo in the early Edo period and developed there — an attribution it marks as tradition rather than record.
It is more concrete about when the thing became visible. Quoting Morisada Manko, an Edo-period account of social customs, the register reports that around the Bunsei era (1818-30) women gathered small pieces of crepe in white, blue, red, purple and other colours at the back of the shimada coiffure, in the shape of chrysanthemums or cranes, and wore them as kanzashi. The register does not call that piece tsumami work, and the cloth it names is crepe rather than habutae. What it gives is a dated sighting of gathered silk worn as a kanzashi.
Which sets up the question this chapter is actually about. Not when kanzashi was invented, but when it stopped being ordinary — and what stayed behind after it did.
Everything in the previous section depended on one condition: that women put their hair up in structured arrangements that needed holding and invited ornament. Remove that condition and the entire category loses its everyday function at once.
The hairstyle did change, and the change is documented. The Metropolitan Museum of Art, describing a work of about 1910, notes that the sitter's hair "is coiffed in one of the various types of sokuhatsu 束髪 styles, inspired by Western women's hairstyles," and that these "became popular around the mid-1880s and 1890s." A new way of wearing the hair arrived, and it was not the arrangement the older ornaments had been made for.
What follows from that is inference rather than a documented claim, and it should be read as such: the demand that had sustained ornament makers was not a taste for decoration in the abstract but a daily requirement created by a hairstyle. Change the hairstyle and the requirement changes with it. No source consulted here states that causal chain outright — what the sources record is the change on one side and the outcome on the other.
It is worth being precise about what this did and did not do. It did not end the craft. What it ended was the assumption that a woman would need one.
What remained were the occasions where the older arrangement is still worn on purpose. Tokyo's official travel guide lists them in the course of describing what kanzashi are for: ornaments that "decorate young women's hairdos as they put on kimono on important occasions such as New Year, coming-to-age ceremony, and their birthdays."
The Tokyo craft register gives a slightly longer list of the moments the craft now serves: New Year, shichi-go-san, jusan-mairi, coming-of-age ceremonies, and weddings. It also records that Tokyo is today the principal production area for tsumami kanzashi.
Read those two lists together and the change becomes concrete. The craft did not move to a museum. It moved to a calendar.
A craft that is worn a handful of times in a life supports a correspondingly small number of makers, and the sources are blunt about the scale. Tokyo's official travel guide calls kanzashi "a close-to-extinct traditional handicraft of Tokyo" and states that "only very few craftsmen master this piece of art nowadays."
One detail in the Tokyo craft register makes the size of this world unusually easy to picture. The register lists the trade association for the craft — the Tokyo Hair Ornament Manufacturers' Cooperative Association — and gives its registered address. The address it gives is a company in Mukojima, Sumida City. The body that represents the makers of this craft is registered at a business address inside the district where it is made.
The practical question a traveller actually has is not historical. It is whether the piece in front of them is the thing this guide has been describing, or a machine-made object shaped to look like it.
The four-stage sequence in the Tokyo craft register is a description of process, but it is also a list of places where evidence survives on the finished object.
Because each petal is a square of cloth pinched and folded, a handmade flower is an assembly of separate pieces, each one set onto a paper base with tweezers. The petals have edges that were folded rather than cut to shape. They are placed individually, which means their spacing is the maker's judgement rather than a mould's.
A pressed or moulded flower has none of that structure to show. Look at where the petals meet the base, and at whether the petals are identical to each other in a way that hand-folding does not produce.
This is a way of looking, not a verdict. A trained eye reads these things quickly; a first-time buyer should treat them as questions to ask rather than a test to apply.
The most checkable signal is the cloth. The Tokyo craft register does not leave the material open: for Edo tsumami kanzashi, the cloth is to be habutae, and the wooden body is to be boxwood, pear, magnolia, or timber of equivalent quality.
That is a specification, and it gives you a question with a right answer — though it answers a narrow question. Ask what the cloth is. A piece sold as Edo tsumami kanzashi has to be habutae, because the designation requires it. A hand-made ornament in another cloth is not a fake; it is simply outside that designation.
A kanzashi is never worn alone. It sits at the top of an arrangement that includes the kimono, the occasion, and the time of year — and it is the smallest element in that arrangement, which is exactly why it carries the finish.
The Tokyo craft register names motifs such as chrysanthemum and plum blossoms, butterflies and phoenixes.
The register offers these as examples rather than as a closed set, and they do not all work the same way. Plum sits at the end of winter and the start of spring in ordinary Japanese usage; chrysanthemum belongs to autumn. The phoenix is an auspicious emblem rather than a seasonal one. The butterfly the register also names is the one this guide will not place for you: unlike plum and chrysanthemum, it does not sit at a single point in the year, and the register gives no ruling either way. Where a given motif sits is a question with an answer, and it is not always a month.
That is the useful thing to carry. The question in front of a case is not "what month is this for" but "what is this motif doing" — and the answer, whether it is a season or an emblem, tells you something about the occasion the piece was made for.
There is a reason a hand-sized ornament belongs in a conversation about Tokyo's architecture and art. The habits are the same at both scales: a preference for restraint over display, an attention to how a thing is finished where nobody will look, a willingness to spend disproportionate time on a small area. If you are building a trip around that sensibility, our guide to Tokyo's architecture and design covers the same instinct at building scale.
Access to this craft runs in three layers, and the difference between them is not price but permission.
The first layer is open to anyone, and Tokyo has a dedicated venue for it. The Tsumami Kanzashi Museum in Takadanobaba, Shinjuku City, is listed on Tokyo's official travel guide, which describes it as a place to see "a close-to-extinct traditional handicraft of Tokyo."
Three practical details are worth checking before you build a day around it. Admission is free; the museum is open on Wednesdays and Saturdays only, 10:00 to 16:00, with last entry thirty minutes before closing; and it is about thirteen minutes on foot from the Toyama exit of Takadanobaba Station. The official listing itself advises confirming hours and closures directly before visiting, which is sound advice for any small craft venue. If you are planning the rest of that day, our guide to places to visit in Tokyo sets out routes that pair well with a short museum stop.
The second layer is not hidden, but it is not signposted either. The Tokyo craft register lists the moments the craft is now made for: New Year, shichi-go-san, jusan-mairi, coming-of-age ceremonies, and weddings. Those are the days to be in a shrine precinct or outside a hall if you want to see kanzashi worn rather than displayed.
You will not be shown the craft in these settings. You will see it in use, which is a different and in some ways better thing — and it is the only layer where the ornament is doing the job it was made for.
What makes this layer available is knowing when and where to look, and being willing to observe rather than shop. Our overview of Japan's traditional crafts maps the wider landscape these occasions belong to.
The third layer is a working floor. Tsumami kanzashi is made in a small number of Tokyo workshops — the Tokyo craft register names Taito, Arakawa and Sumida among the principal production districts — and a workshop is not a venue. There is no ticket counter, because it is not an event.
Access here runs on introduction rather than availability. That is the same constraint that governs most of the workshops we visit, and it is the reason this layer exists in the guide at all: not to tease it, but so that you can tell which of the three you are standing in.
One thing about this craft is still moving, and it is the reason a visit has a date on it. The register places the craft here: Tokyo is now the principal production area for a technique the same register says is thought to have reached Edo from Kyoto. Tokyo's official travel guide, writing for visitors rather than for the trade, is the body saying it may not stay — "close-to-extinct," with "only very few craftsmen" left who have mastered it. A craft register and an official travel guide describe the same object, and only one of them is describing something settled.
Three things are worth taking away from this guide.
The first is that kanzashi names a family, not an object. The Victoria and Albert Museum keeps kanzashi and kogai as separate Japanese words while rendering both as hairpins, and the pins in museum cases differ from each other more than the translation suggests.
The second is that the silk flowers are a specified craft, not a general style. Four stages, named folds, habutae cloth: when you ask what something is made of and hear the right answer, you are not being sold a story.
The third is the arithmetic of scarcity. A craft worn at New Year, at coming-of-age, at weddings, is a craft with a small number of occasions and a correspondingly small number of makers — one whose trade association is registered at a single business address in the district where the craft is made.
That is the frame worth carrying into the museum case, or into a shop window in Asakusa. The object is small. What it is attached to is not.