The pair that refuses to be one age
At the Metropolitan Museum of Art there is a pair of Japanese swords whose component dates run almost backwards through the history one might expect. The katana blade is sixteenth century. The shorter wakizashi blade is dated to the late fourteenth or early fifteenth century. Their mountings are nineteenth century. The two objects in front of us belong together, yet the shorter blade existed before the longer one, and both existed before the visible fittings united them as a pair. Few exhibits are better at explaining the word daishō. Historical unity need not mean common manufacture or even a shared century.
The word joins the ideas of large and small. In accounts of samurai dress it usually means the long and short swords carried together, most recognizably a katana and wakizashi. It can also direct attention to the matching appearance of their mountings. Three closely related questions follow. Which type of blade does the curator identify? What does the mounting say about how it was presented? What evidence connects the two items as a pair? A tape measure can help with the first question. It has almost nothing to say about the last. A period inventory might do precisely the reverse.
The distinction matters for common English names. Samurai sword can describe a whole collection of things without telling a visitor whether the object is a tachi, katana, wakizashi or tantō. A phrase such as short katana occasionally conveys the visual resemblance that struck its speaker, but it obscures the history of the wakizashi as a named form. Long sword and short sword in a label are often translations designed for readers without specialized vocabulary. They should send us back to the Japanese term, the named component and the stated measurement before we draw conclusions about historical practice.

The shaku and the measurement that counts
A commonly used modern equivalent for the shaku is about 30.3 centimetres. One shaku and two shaku therefore give practical reference points near 30.3 and 60.6 centimetres. A familiar classification puts tantō below one shaku, wakizashi between about one and two, and long swords such as katana or tachi generally at around two shaku or more. These figures are useful for reading a collection. They are less useful if treated as a single statute covering every community and every century of Japanese history. Museum titles also preserve established type names, earlier usage and details of an object's particular history.
| Type | Customary nagasa | Common form | Historical mounting | Dress or ceremonial role |
|---|---|---|---|---|
| Katana | Usually ≥2 shaku (≈60.6 cm) | Often shinogi-zukuri, with a longitudinal ridge | Through the sash, edge up | Long member of an Edo daishō. |
| Tachi | Often ≥2 shaku; no separate length cutoff | Typically ridged, curved and tapering | Suspended by fittings, edge down | Court, ceremonial and earlier warrior dress. |
| Wakizashi | Usually 1 to under 2 shaku (≈30.3 to under 60.6 cm) | Both ridged and flat-sided forms occur | Through the sash, usually edge up | Short member of a daishō. |
| Tantō | Usually under 1 shaku (≈30.3 cm) | Often hira-zukuri, with a flat-sided profile | Small sash mounting; guardless aikuchi examples survive | Personal and ceremonial dress; mounting varies. |
These are customary cataloguing boundaries. Historical dress regulations depended on the period and social setting; the one- and two-shaku conventions are not a universal legal code.
Sources · Japanese Sword Museum · Tokugawa collection guide
The relevant length is usually nagasa, the conventional blade length measured as a straight line from the point to the notch on the back at the base of the blade. This is a chord between landmarks. It does not trace the curve of the cutting edge. It excludes the tang that lies inside the hilt. It also excludes the grip, scabbard and other parts that make an assembled sword longer. Historical catalogues can render the Japanese term by more than one English expression. The sensible comparison is between the definitions used by the two records, not between the first numbers that happen to appear after length.
A wakizashi in the Met provides an unusually clear demonstration. The overall length in its record is 67.9 centimetres. The length described there as that of the blade is 59.4 centimetres. The length given for the cutting edge is 45.3 centimetres. A reader who extracts only the first number would think the object exceeds two shaku. A reader using the last number recovers a perfectly plausible wakizashi measurement. The intervening number is a warning to inspect the field names carefully. It would be easy to populate three rows of a spreadsheet with three apparently different swords when the museum is describing just one.

Measurement also has a history of its own. The conventional conversion lets readers working in centimetres compare records. It does not retroactively give every period ruler the same precision as a modern instrument. A rounded value in inches may lose information that the museum retained in centimetres, and an old catalogue may report a length without specifying its reference points. What looks like a meaningful difference between two descriptions can be a rounding convention. Close to a boundary, it is better to record both the original measurement and the curator's identification than to replace either with one's own more decisive-sounding category.
What curvature and construction contribute
Sori describes curvature. A conventional depth of curvature is measured in relation to a straight reference line rather than by tracing the entire edge. An outline can curve more strongly toward one part of the blade than another, so two blades with similar maximum depth can still have observably different silhouettes. The Kyoto National Museum's illustrated introduction is especially useful here. A carefully labelled diagram allows a visitor to find the back, point, tang and reference line without pretending a photograph is equivalent to a physical examination. It also makes clear why tilting a camera changes the outline available for measurement.
Zukuri refers to form or construction in names describing the shape of a blade. The terms shinogi-zukuri and hira-zukuri direct attention, among other features, to the presence or absence of a distinct ridge along the blade. They describe an organization of surfaces. They are not a second set of names for the one-shaku and two-shaku classifications. A specialist label can give both a length category and a form precisely because the two convey different information. It is possible to recognize a particular geometry in a published image without being able to recover an exact length from that image.
Museum displays influence which of those features a visitor notices. Light grazing one plane can make a ridge appear sharply defined, while a reflection in an adjacent plane can nearly erase it. An illustrated guide can mark the ridge even where the photograph alone does not resolve it. Conversely, the actual object may display changes along its surface that a line drawing necessarily simplifies. The most rewarding approach is to move between the two. Use the diagram to identify what feature is under discussion, then return to the historical piece to see how the particular object varies from the idealized form on paper.
A long sword can be a tachi or a katana
A length above two shaku does not settle the choice between tachi and katana. Both can be long swords. The history of their mounting and wear supplies a distinction that a chart of lengths cannot capture. A tachi in its characteristic mounting was suspended with its edge down. A katana was conventionally carried through the sash with its edge up. These conventions also help explain the orientation often chosen for museum display and the relationship between an inscription and the side visible to the viewer. The classification of an actual blade can still require more evidence than its present orientation in a case.
A blade can outlive a mounting several times over. A changed mounting can put an old object into a new social setting, and historical shortening can alter the blade itself. Kyoto National Museum's guide discusses substantial shortening and the consequences for the tang. Such a history can affect whether a signature survives and how later viewers describe the object. The result need not fit neatly into an account that assumes a sword left one workshop in its present appearance and never changed. The museum record may preserve the older identity precisely because the present configuration cannot tell the whole story.
The Tokyo National Museum's Uesugi tachi offers a useful point of reference for the imagery of a long sword. The name is attached to a particular collection record. It is not a generic illustration that can be relabelled katana whenever a long curved form is needed. If an author wants to compare it with a katana, the informative caption preserves the name tachi and explains which feature the reader is invited to compare. A similar discipline applies to a photograph of a mounting without its blade. The long shape in that image might belong to the scabbard and its fittings. A blade measurement cannot be taken from it.

The dress of the Tokugawa period
Under the Tokugawa governments, wearing paired swords became closely associated with the social identity of the samurai. This is the historical setting in which discussions of daishō as a public sign of rank make particular sense. But people did not live in a diagram of ranks. Audiences at court, formal processions, journeys, household service and ordinary visits presented different occasions. Dress served as an intelligible language because those present recognized distinctions among these settings. A statement about the ceremony of an audience does not automatically describe the everyday behaviour of everyone who owned a sword.
Laws and household rules are particularly liable to lose their context in translation. A prohibition records what an authority wanted to regulate. It cannot, without further evidence, tell us how consistently the rule was observed, who possessed a dispensation or how people applied it in a private dwelling. A later etiquette manual may preserve a highly formalized account of an earlier custom. A period image may instead choose the most readily identifiable version for its intended audience. These sources can illuminate each other. Treating them as interchangeable quotations from a single timeless samurai code makes real differences of date and place disappear.
The familiar indoor and outdoor distinction illustrates the difficulty. Leaving a long sword at an entrance could be part of the etiquette of a visit, and the shorter companion had a different place in accounts of dress indoors. Such conventions depended on the host, the visitor, the room and the occasion. The sequence observed in a formal setting should not be promoted to a claim that every samurai in every domestic room followed an identical rule. A useful caption identifies the specific scene, notes what the participants visibly wear and describes the relevant custom as far as the evidence permits.
Daishō also needs to be distinguished from the history of schools associated with the use of two swords. A matching set can attest to rank, taste and the work of several makers. Its existence supplies no record of the owner's training. In the other direction, an account of a school cannot authenticate a particular pair of museum objects as the personal possessions of its founder. The two histories sometimes meet in a securely documented person or object. More often the association exists only in a modern reader's mind, encouraged by the ease with which English shifts between two swords, a pair of swords and a named tradition.
Three museum pairs, three different answers
Return to the Met pair whose katana bears an inscription naming Muramasa. The published lengths of the two blades are 63.9 and 35.4 centimetres. The catalogue gives their depths of curvature separately, 1.5 and 0.45 centimetres respectively. The nineteenth-century mountings bring the much older blades into one visual composition. A visitor interested in their origins should follow the separate blade entries. A visitor interested in how someone chose to appear in the nineteenth century should give the later fittings their full share of attention. Looking only at the oldest component erases the person or people who commissioned or assembled the pair.
Another daishō in the same museum is assigned to the nineteenth century. Its long blade measures 69.2 centimetres and its short one 46.5 centimetres. Even in this small sample the proportions differ considerably from the earlier example. The shorter member is not simply a reduced copy produced according to a fixed ratio. Nor can a photograph that shows one pair close together and another pair separated supply a reliable comparison of scale. The catalogue makes a stronger argument than an attractive layout. Separate dimensions and a shared record show exactly what is being compared and what connection the museum is asserting.
A third catalogue title gives a particularly instructive exception to the usual shorthand. The Met identifies a pair for a boy. Its longer blade measures 49.4 centimetres and its shorter one 38 centimetres. Both fall numerically between one and two modern shaku. Yet the record describes a katana and wakizashi in their relation to the child's paired mountings. Calling both pieces wakizashi on the evidence of a conversion chart would conceal the very fact that makes the ensemble interesting. The period of the blades and the period of the mountings also differ. Age, size, function in dress and present museum terminology have to be read together.

What is paired in a set of fittings?
A daishō soroi-mono is a coordinated set of fittings. The word set is easy to pass over on the way to an image of two completed swords. In fact a set of fittings can reach a collection without either of the blades with which it was once associated. The artist may have planned corresponding motifs at different scales or treated a subject differently on the large and small components. A collector can study these relationships without reconstructing a hypothetical missing blade. A catalogue that identifies fittings for a pair makes a different claim from one that identifies a pair of blades and their mountings together.
Historical ornament can do a great deal of the work of making a pair legible. A small detail repeated on two fittings may be enough to connect them across a room. Closer inspection can reveal subtle differences in surface and composition that keep the repetition from becoming mechanical. This is a question with an exact parallel in portraiture. Two sitters in related clothes do not have to wear identical garments to be seen as belonging together. The point of coordination is to preserve individual proportions while communicating a relationship. The survival of such fittings gives the historian access to decisions that are invisible in a simple list of blade lengths.
The tantō belongs in this account even though a typical Edo daishō is described as katana and wakizashi. It occupies an important place in the vocabulary of shorter blades and in the history of elaborate mountings. A tantō blade in a mounting identified as chisagatana, for instance, makes the distinction between blade classification and the appearance of the mounted piece especially clear. The Met uses precisely that distinction in its record for a dagger with an ornate mounting. Its title gives the reader two pieces of information. Reducing the whole to small samurai sword deprives the reader of both.
Reading the label back to front
One way to enjoy a collection of Japanese swords is to read each label twice. On the first reading, look for the familiar terms and find the broad type in the case. On the second, begin at the end. Whose collection was it, and what entered under one accession number? Does a date apply to the steel, the lacquer or both? Does the listed length describe the exterior, the full blade or the conventional distance associated with nagasa? Is a place assigned to a maker, to a later mounting or to a previous owner? The answers can transform two apparently straightforward silhouettes into a conversation across centuries.
A size chart still has its place beside that conversation. One and two shaku give an English-speaking reader a way into a vocabulary that might otherwise seem arbitrary. The boy's daishō keeps that vocabulary honest. The Muramasa-inscribed blade and its much older companion explain why a pair can have several beginnings. The Uesugi tachi supplies an example of a long sword with a different history of wear. Put them together and daishō comes to describe something richer than a long blade with a shorter version beside it. It describes a relationship made visible through metal, lacquer, clothing, social expectation and the decisions of people who kept things together.
Frequently asked questions
What is a daisho?
Daishō means large and small. In samurai dress it commonly identifies a long and short sword worn together, and the term can also qualify coordinated mountings or fittings.
What is a wakizashi?
A wakizashi is a Japanese short sword, conventionally about one to two shaku in blade length. It is the familiar shorter member of an Edo-period daishō.
What is a tanto blade?
A tantō is a Japanese short blade, traditionally classed at less than about one shaku. Its mounting may have a separate name, as in the Met example illustrated here.
What is a tachi?
A tachi is a Japanese long sword associated with an edge-down suspended mounting. Blade length alone does not distinguish it from a katana.
Why did samurai carry two swords?
In the Edo period the long and short pair communicated social identity in formal dress. The rules varied with the circumstances, and possessing a pair does not establish training in a particular school.
What are samurai swords called?
Terms include tachi, katana, wakizashi and tantō. They distinguish different forms and histories of wear. Daishō describes a relationship between two swords rather than a further blade shape.
Films and talks
JASA: Japanese Art Society of America — Samurai Splendor: Sword Fittings from Edo Japan (Lecture by Markus Sesko)
The Met — The Japanese Sword and the Japanese Aesthetic
Collection records and further reading
- Kyoto National Museum, illustrated guide to Japanese swords
- The Met, paired mountings with a Muramasa-inscribed blade
- The Met, nineteenth-century daishō
- The Met, a daishō for a boy
- Japan Tourism Agency, Swords of the Tokugawa Shoguns
- The Daishō in Edo Portraits: When Does a Painted Pair Identify Rank?
