Koiguchi and Habaki Craftsmanship: Water-Buffalo Horn Collars, Wedge Fit, and Honoki Scabbard Precision

Where a museum label changes materials

A long sword in the Walters Art Museum looks, at first glance, as if it has a bamboo scabbard. The collection record describes brown lacquer in imitation of bamboo. It goes on to identify a gold blade collar and a dragon ornament in silver. Three different questions arise within a few inches of the same object. What does the surface represent? What material did the maker use? Where exactly does the catalogue locate that material? The mouth of a Japanese scabbard, the koiguchi, is a particularly good place to ask all three. Colour, substance, and function can be difficult to distinguish at precisely the point where several crafts meet.

Hashimoto Ichizo, mounting with imitation bamboo scabbard, Walters Art Museum 51.1218
Hashimoto Ichizo, mounting with imitation bamboo scabbard, Walters Art Museum 51.1218. Hashimoto Ichizo (via Wikimedia Commons).

A saya is a scabbard. The koiguchi is its mouth, a name traditionally understood through the image of a carp's mouth. The habaki is a metal collar at the base of the blade. When the blade sits in its scabbard, the collar comes into contact with the mouth. The familiar description of a wedge fit refers to this seating relationship. It is a description of a feature of the ensemble. It is not a measurement, a guarantee of present condition, or a claim that every historical mouth contains the same insert. The difference between these statements runs through almost every museum record on the subject.

A boundary shared by two crafts

Look at a side view of a sheathed object and the koiguchi can seem to be little more than a line. In fact the line belongs to an opening with an exterior surface, an interior, and a rim. A lacquer artist, a woodworker, and a metalworker would have reasons to attend to different aspects of that opening. The person studying an object many generations later may have access to only one view. A photograph of the exterior may show a narrow dark band and a clean boundary in the lacquer. It cannot ordinarily tell us how much of the concealed surface currently makes contact with the collar.

The word collar itself can cause confusion in English. A catalogue can use collar for the metal band at the hilt, known as a fuchi, as well as for the habaki at the base of the blade. A reinforcement around the mouth of the sheath might also be called a collar in an extended description. If a line drawing labels three adjacent shapes without naming the component to which each belongs, the same English word can migrate from wood to hilt to blade in the reader's mind. Good captions give the Japanese term and the part of the ensemble it modifies. The caption then remains useful even when the image is cropped.

A further distinction concerns the koiguchi and the kojiri. The second term designates the opposite end of the scabbard. Both ends may have a finish that contrasts with the main stretch of lacquer, and both can appear in a catalogue as separate decorative features. In an isolated photograph with no sense of scale or orientation, a glossy dark end piece is ambiguous. An overall view restores the position that gives it a name. This is one reason an acquisition photograph of a whole assemblage can be more valuable for identification than a much more beautiful close view of the surface alone.

Where water-buffalo horn belongs in the account

Water-buffalo horn is associated with the reinforcement of Japanese scabbard mouths. Horn can take a smooth surface and presents a different material at the rim from the wood beneath the lacquer. On a finished exterior the change can be deliberately quiet. A polished dark band may harmonize so well with black lacquer that the boundary registers only as a slight change of reflection. The term koiguchi nevertheless applies to the opening itself. It is entirely possible to discuss the koiguchi of a plain wooden storage mounting without thereby identifying any separate horn component on that particular piece.

Horn is especially vulnerable to overconfident identification in photographs. The Canadian Conservation Institute distinguishes horn from bone, antler, and ivory. Horn belongs to the family of keratinous materials, whereas antler is a form of bone. These differences concern how the materials grow and behave, including how an object can respond to its environment. They cannot be settled by putting several polished black surfaces beside one another on a screen. Dark lacquer can cover wood. Horn can have translucent areas. Strong light on either can produce a narrow pale reflection that a compressed image makes look like a separate layer.

The distinction between horn and water-buffalo horn is a further step. A museum's identification of horn is an affirmative piece of knowledge, even if the species remains unspecified. Replacing its careful general term with a familiar specific species would imply evidence the entry may not contain. Sometimes a workshop record identifies a material by its traditional trade name. Sometimes a conservator has examined an accessible broken edge. Sometimes the catalogue gives no reason for its choice of term. An accurate caption can retain those different levels of information without asking any one object to represent the whole history of the craft.

What the collar actually supports

The Japanese Sword Museum's terminology for mountings distinguishes the sheath and its fittings, and its account of preservation identifies the seating of the habaki at the mouth. The physical principle is local contact between an individual collar and the corresponding part of the scabbard. It explains why the collar can be discussed meaningfully even when a view of the blade gives no information about the interior further down the sheath. It also explains why changes in the condition of a relatively small area can matter to the condition of the ensemble. The decorated surface alone tells only a part of that story.

A familiar claim says that a Japanese sword can be turned upside down and will stay in its sheath. As a description of every surviving historic mounting, the claim has no evidentiary value. The condition of the wood, previous repairs, the identity of the collar now associated with the blade, and the means of support in a photograph may all be unknown. A museum display with the opening pointed down is evidence of a curatorial arrangement. An unseen support can carry weight independently of the antique components. A published photograph therefore supplies neither a retention measurement nor a reason to invert a historical object.

Mounted tanto with a bronze-coloured blade collar, Auckland War Memorial Museum
Mounted tanto with a bronze-coloured blade collar, Auckland War Memorial Museum. Artist unrecorded (via Wikimedia Commons). License.

The Auckland example illustrated here shows the value of restrained material language. Its description calls the habaki bronze-coloured. That tells the reader what someone observed, while leaving open the question of composition. The Walters entry for the imitation-bamboo mounting uses a different kind of language when it identifies gold on particular parts. These records need not be brought into artificial agreement. One names a colour and the other names a material. An editor combining them into a comparative gallery should resist the apparently harmless convenience of shortening both captions to the same word, metal, or promoting the first observation into an identification of bronze.

Magnolia wood beneath the surface

The Japanese Sword Museum explains that honoki, a Japanese magnolia, became a common material for scabbards. Its public account also mentions earlier uses of materials including leather and bamboo. That historical range is worth remembering when a catalogue employs the general word wood. Honoki is important to the craft tradition, but a statement about customary material is not a botanical identification of every surviving example. The Walters scabbard illustrates the point with unusual clarity. A maker wanted the visible surface to call bamboo to mind. That intention belongs to the history of design whether or not the underlying wood has been identified to species.

A plain storage mounting changes what the viewer can see. With no elaborate lacquer design to command attention, the contours of the wood and the relationship between adjacent parts become unusually evident. A brushed inscription on the outside can turn a modest-looking surface into a document. It might identify a blade, preserve an attribution, or record a later opinion. Such writing has its own date. A signed or annotated storage sheath can therefore be valuable evidence about the history of a blade even when the sheath was made much later than the blade itself. The relation at the mouth is one part of this longer association.

Storage mounting with a brushed inscription, Rama
Storage mounting with a brushed inscription, Rama. Photograph by Rama, via Wikimedia Commons. Reproduced without editorial changes. License.

The magnolia wood associated with these mountings is an organic material. It exchanges moisture with its environment and its dimensions can change. Lacquer, horn, metal, and textile respond in different ways. For a conservator the most interesting feature of a composite object is often the boundary at which the demands of two materials meet. A dry-looking surface on the outside says little about the history of the concealed joint. A visible change at that joint might result from a movement in one material, an earlier intervention, or a change in how the piece is supported. A single photograph taken today cannot distinguish these histories.

Precision without an invented number

Words such as fit, close, and precise have ordinary descriptive meanings and a stricter meaning in engineering. In a workshop drawing a tolerance is an allowable range relative to a specified dimension. A museum may publish the length of a mounted sword to the nearest tenth of a centimeter without publishing any such specification for its scabbard mouth. These numbers describe the object at different scales and for different purposes. Overall length helps a registrar identify a work, plan storage, or distinguish it from a related piece. It does not furnish a historical design tolerance for a concealed part.

The Metropolitan Museum's short sword with a Tsunahiro inscription provides an instructive example. The entry distinguishes the length of the whole work, the length of the blade, and the length of the cutting edge. Its materials list includes horn among several other substances. Its text dates the black lacquered scabbards associated with a pair to a later century than the blades. These are useful, quite specific observations. Reading them together permits a history of later mountings for older works. Reading them as though they specified the point of contact between a particular piece of horn and a particular metal surface would ask the entry to do something it never undertook.

Precision in historical craft can instead be recognized in the way the parts were understood as a particular assembly. The identity of a collar matters because its relation to one blade and one mounting is individual. A registrar's decision to assign related component letters to pieces under one accession number preserves evidence that could be lost if similar-looking parts were stored separately. A condition record can identify a change that was already present on arrival rather than implying that it occurred during exhibition. These aspects of precision concern observation and continuity. They remain meaningful even where no original working drawing or dimensional specification survives.

Lacquer, imitation, and the direction of light

The second Walters mounting shown below has a black lacquer scabbard with a cherry-blossom design. The museum describes contrasting surface finishes. That difference can become dramatic as a light source shifts, even on an object whose overall colour changes very little from one area to another. A photograph taken directly in front of a surface may conceal a design that appears at an angle. Conversely a strong reflection can make a continuous surface look as though it contains a separate piece. The catalogue's account of the lacquer is an independent way of checking what seems to be a material boundary on the screen.

Short sword with black lacquer scabbard and cherry-blossom decoration, Walters Art Museum 51.1261
Short sword with black lacquer scabbard and cherry-blossom decoration, Walters Art Museum 51.1261. Kinmichi II / Takamoto Hidemune (via Wikimedia Commons).

This example and the imitation-bamboo scabbard make especially effective companions. One work asks lacquer to evoke the structure of another plant. The other exploits contrast within a dark surface. Both show why a materials list and a visual description ought to be kept together. On a cropped detail, a groove in the decorative representation of bamboo might read as the end of a structural component. An isolated bright line in the black finish might seem to mark the edge of a band. Return to the entire object and to the written description before deciding which surface the detail actually represents.

A reproduction can also change apparent scale. A close view of a narrow band printed at the width of a page makes the band feel like a large independent object. In an overall photograph the same feature may disappear into the visual rhythm of the hilt, guard, and sheath. Neither view is more true in every respect. One preserves local information, the other preserves relationships. A useful online museum record allows the reader to move between the two while keeping the object number and the part name in view. The person researching a scabbard mouth needs precisely that movement between detail and ensemble.

How much can be learned without disturbing an object?

Consider the questions a curator can answer from documents before any physical examination begins. When was the mounting accessioned? Did a blade, collar, sheath, storage box, and older photograph arrive together? Does a previous exhibition show a different exterior with the same blade? Does an earlier record locate a repair at the mouth, or merely note that the sheath was once restored? The answers distinguish a documented change from a difference between two photographers' choices. They can also reveal that what initially seemed an original combination has a more complicated and historically interesting association.

When direct examination by a qualified conservator does take place, it is guided by the condition and value of the individual object. A question about historic appearance might be answered by comparing older photographs. A question about present stability might be answered by studying the existing mount and its support. A claim about species might remain open if the necessary evidence is inaccessible. There is no scholarly advantage in turning an aesthetic curiosity into a physical trial of a centuries-old assembly. A careful report may end with more precise vocabulary and the same honest uncertainty about an interior surface that has never been documented.

Writing the record for the next viewer

Imagine a caption for a future exhibition bringing these works together. It could begin by naming the opening, identify the habaki as a separate metal component, describe the visible finish of the sheath, and give a material identification only at the level supported by the catalogue. A second sentence might explain why an earlier blade is now associated with a later exterior. The result would be readable to someone encountering Japanese terminology for the first time and useful to someone comparing several collections. Most importantly, the reader could return to a particular object number and test every part of the description against the available evidence.

A final detail to preserve in a caption is the origin of the photograph itself. An accession number identifies the object, while a photographic credit identifies a particular view of it. One object can have several separately dated photographs made under different conditions of light, with different parts visible. If the question concerns a change in the dark rim of a scabbard, the dates of those views matter more than the date on which an online page was last revised. Without that information, a newly uploaded image of an old photographic print could easily be mistaken for a new examination. Recording the distinction costs little and makes a future comparison much more reliable.

The koiguchi is an unusually good introduction to the history of Japanese scabbards because it is both visually modest and physically consequential. It brings the vocabulary of wood, horn, lacquer, and metal into one small space. It also brings together the work of the people who made the mounting, those who kept or replaced it, and those who now describe and preserve it. The most persuasive evidence of craftsmanship is found in those particular objects and relationships. A gold collar on one mounting, a bronze-coloured surface on another, and an inscribed wooden sheath each tell us more than a universal promise about how every historic sword ought to fit.

Frequently asked questions

What is the koiguchi on a Japanese scabbard?

Koiguchi names the opening of the saya. It describes a location and can apply whether the mouth is plain wood or carries a separate reinforcement.

What does the habaki do at the scabbard mouth?

The metal collar seats at the mouth and helps retain the blade. The intended relationship is local to the two components.

Does a dark scabbard mouth prove it is water-buffalo horn?

Colour alone cannot establish the material or species. A museum may record horn without identifying the animal or attributing it to a particular component.

Why does honoki appear in scabbard records?

Honoki is a Japanese magnolia associated with scabbard wood. An object entry that says only wood does not establish a species for that specimen.

Can a historic photograph establish the exact tolerance of a habaki?

An overall photograph shows exposed contours. A numerical tolerance requires original specifications or appropriately documented measurements.

Is an inversion test a museum standard?

An inverted display or a photograph supplies no retention measurement. Historic assemblies are evaluated from their individual condition and conservation records.

Films and talks

中之島香雪美術館 — 中之島香雪美術館 企画展「館蔵 刀装具コレクション 武家の嗜好品」<2022年12月17日(土)~2023年2月26日(日)開催>

ニコニコニュース文化部 — 刀身・拵に込められた日本人の思い。春日大社国宝殿の「刀剣の美を楽しむー刀身の美・鞘の美ー」を橋本麻里さんと巡ろう【刀研機構×ニコニコ美術館】#ニコ刀 #ニコ美

Collection records and further reading

Related Articles