How Much Does a Katana Weigh? What Museum Measurements Actually Record

How much does a katana weigh? The Nagoya Sword Museum lists an Osafune Motoshige blade at 658 grams and a Horikawa Kunihiro at 724 grams. Their listed cutting lengths differ by only two millimetres. Both figures are quite plausible answers to a narrow question about individual blades. Neither tells us what a person would have received on lifting an entire sheathed sword from its box. The difference between those questions is large enough to change the history we tell. It begins before anyone opens the box, with the noun on the inventory card.

Two blades from Nagoya

Start with the two entries side by side. The first is assigned to the Kamakura period and has a length of 68.3 centimetres. The second belongs to the early Edo period and measures 68.1 centimetres. If the only information surviving on an old card were those two lengths, a researcher would have little reason to expect a sixty-six-gram difference. Yet the weights are there in the table. The blade with the slightly shorter recorded length is the heavier of the pair. A single length cannot describe the quantity or distribution of material in an irregular three-dimensional object.

The comparison has particular force because it comes from one institution and one table. The column headings stay the same between entries. The museum identifies its objects, gives the units once, and places each row in a period. Across two unrelated websites, the researcher would first have to discover whether both writers included the same components. Even within Nagoya's table there is a further distinction to keep. Some rows describe tachi, others katana. Grouping them together can be illuminating when the question concerns long Japanese blades. It changes the question if the heading of a chart promises an average for katana alone.

Consider the entry for Suishinshi Masahide. Its blade measures 69.4 centimetres and weighs 928 grams. This gives us three examples within a narrow band of lengths and a much broader spread of masses. It does not give us a time series. The three surviving objects had different histories. Their makers did not submit them to an experiment in which every other variable was fixed. A historical claim that swords became heavier over these centuries would need many more objects, clearly defined selection criteria and information about what happened to each blade after its manufacture.

What the Met measures on a signed blade

At the Metropolitan Museum of Art, accession 2001.574 is attributed to Masazane of Ise. The Met gives its date as 1526. The artist's name and a date survive on the tang, and the museum calls attention to the unusual pattern visible in the steel. The catalogue reports an overall length of 91.8 centimetres and a cutting-edge length of 75.1 centimetres. It does not supply a weight in that dimensions field. This is an unusually informative object for a historian of sixteenth-century workmanship and a blank cell for someone compiling published blade masses. Both statements can be true at once.

Blade for a Sword (Katana); 2001.574, Metropolitan Museum of Art, view 1
Masazane’s signed katana blade, dated 1526. The Met, 2001.574. Image courtesy of The Metropolitan Museum of Art, Open Access.

Look at the difference between the two lengths. The hidden part of a blade, the tang, is still part of the piece of steel a registrar might weigh. It is absent from a measurement explicitly restricted to the cutting edge. Dividing a published total mass by a length copied from a different field would produce a number, but the number would describe neither a complete blade nor the visible portion. The problem survives conversion to inches or pounds. Changing the units simply gives the same poorly defined quantity another name.

The object is particularly instructive as a photograph. The long unmounted form can be shown against a neutral ground without the furniture that frames a sword in a portrait. Steel reflects light. A band that appears broad and brilliant in one exposure can look almost hairline in another. Texture, illumination and photography help us see how an object was studied. They supply no independent reading of its mass. The photographic record and the dimensions record are complementary evidence, each with a different boundary.

Blade for a Sword (Katana); 2001.574, Metropolitan Museum of Art, view 2
A second catalogue view of the same Masazane accession. A change of view supplies another kind of evidence from a change of scale. Image courtesy of The Metropolitan Museum of Art, Open Access.

One sword, two cultural attributions

An eighteenth-century mounting at the Met provides a more complicated example. Accession 32.75.301a, b is classified as a peidao. Its blade is Japanese and its mounting Chinese. The museum assigns the approximate dates 1750 to 1800 to the mounting. The record gives 703.1 grams under weight and 238.1 grams for the scabbard. It also distinguishes the lengths of the blade, the sword without the scabbard, and the sheathed assembly. Those headings are part of the evidence. Removing them makes this object look like an exceptionally well documented Japanese katana of a known date. That would be a serious catalogue error.

Sword (Peidao) with Scabbard; 32.75.301a, b, Metropolitan Museum of Art, view 1
Peidao with a Japanese blade and Chinese mounting. The Met records the scabbard weight separately, accession 32.75.301a, b. Image courtesy of The Metropolitan Museum of Art, Open Access.

The two letters in the accession carry a quieter lesson. A collection may assign one accession stem to pieces understood as an ensemble. The components can have separate identities in storage and separate measurements in the database. A photograph containing both pieces may appear under a single title. A catalogue search can therefore return one result for an object under which the registrar has recorded several physically distinct things. Conversely, a fitting that once belonged to a sword can receive its own number if it enters a collection on another occasion. Counting search results is not necessarily counting original swords.

The mixed attribution also rescues a piece of history that a weight chart would erase. The Japanese metalwork and the Chinese mounting participate in the same surviving object. Sorting the whole entry into a single column marked Japan because the blade originated there would conceal the work of the people who furnished its later setting. Sorting it into China alone would lose the other half. A catalogue can hold both observations together. A spreadsheet can do the same if its author allows a separate field for each component and records how the museum assigned the date.

A Japanese-made smallsword at the edge of the table

The Met's accession 26.145.327a, b looks like a promising search result for Japanese sword weight. The weight fields are filled in, the institution is identified, the photographs are clear and the dimensions give both metric and customary units. The catalogue describes a smallsword with a scabbard, probably made for a Western market and possibly associated with Dejima. That uncertainty matters. Even if an object was made in Japan, its presence in a Japanese arms collection cannot make it evidence for the mass of a historical katana.

Smallsword with Scabbard; 26.145.327a, b, Metropolitan Museum of Art, view 1
A Japanese smallsword for a Western market, The Met, 26.145.327a, b. This comparison is outside the katana sample. Image courtesy of The Metropolitan Museum of Art, Open Access.

There is still a reason to keep this entry in a research notebook. The record separately gives 340.2 grams and a scabbard weight of 85 grams. The length of the sheathed object is only slightly greater than its length without the scabbard. A sheath can make a modest difference to overall length while contributing a readily measurable quantity to the mass. This is a lesson about the design of an inventory, not a proposed addition to the Nagoya sample. In a public exhibition the two pieces might fruitfully stand next to each other. In a table calculating an average they require different classifications.

A second detail concerns the phrase possibly Dejima. It records a degree of confidence that the photograph cannot settle. Anyone using the record for an exhibition about exchange between Japan and Europe should preserve that qualifier. If a label says that a specific workshop at Dejima produced this object, it has silently converted an attribution into an established address. The same discipline belongs in writing about measurements. A reported number can be precise even where the object category, place of production or extent of the assembly remains uncertain.

What can be said about the centre of mass?

An inventory's weight field reports a scalar quantity. The centre of mass is a location. To record the latter, the researcher needs a reference point, a stated assembly and an observation of that particular object. A painted portrait supplies none of these. A photograph taken from the side cannot show the full distribution of material behind its visible outlines. The Nagoya comparison already demonstrates that two nearly equal lengths can correspond to appreciably different masses. It provides no basis for assigning a balance point to either object.

Older photographs sometimes include stands, textile supports and visible ties. Those belong to the history of display. Their position may reflect the arrangement that kept a fragile surface clear of a shelf, or the composition that made an illustration legible in a printed catalogue. Reproducing a photograph with a vertical line dropped from a stand and calling the line a measured centre of mass would transform the photographer's arrangement into a fictitious experiment. The responsible entry for these records is simply that the centre of mass was not reported.

There is a productive historical question left after that blank has been entered. Which features did the catalogue maker consider important enough to document? For the Masazane blade the inscription and the pattern in the steel occupy much more of the label than its physical dimensions. For the peidao the separation between blade and mounting allows a story about movement between places and artistic traditions. The pattern of things recorded, as well as the measurements themselves, reveals what kinds of comparisons a collection invited its readers to make.

How to cite an object whose measurements disagree

Museum records often bring older measurements into newer interfaces. An imperial fraction can be transcribed incorrectly. A decimal point can move during data entry. The Met entry for the blade and mounting numbered 1975.158 a, b, for example, identifies a blade dated 1492 and a mounting from the late sixteenth or early seventeenth century. In the online dimensions string some fractions and metric equivalents plainly require clarification. The sensible response is to retain those object dates where relevant, link the catalogue entry, and omit the disputed dimensions from a numerical comparison.

It is tempting to repair a suspicious figure by calculating what the cataloguer must have intended. That may be a useful question for correspondence with the institution. It does not establish which original field was correct. The inch value may be wrong, the metric value may be attached to the wrong label, or two errors may have arisen in different editions. A transparent note can identify the discrepancy without attributing a finding to the collection that it has never published. It also gives a later reader a way to revisit the case if a correction appears.

Kyoto National Museum's published corrections for its Kyoto sword exhibition offer a good model for readers. An errata sheet preserves a connection between a particular entry in a particular edition and its correction. The printed catalogue remains a document of its time. If an author relied on a superseded dimension, a later reader can see how the mistake propagated. It is worth keeping the object number, title, edition or access date, original unit, component name and any later correction together. A naked number pasted into a spreadsheet deprives the next reader of that path back.

Who chose the survivors?

Numbers lend themselves to averages, but a museum collection is also a record of decisions. Owners preserved some objects and parted with others. Donors brought particular interests to the institutions they supported. Curators selected objects for their inscriptions, unusual surfaces, regional relationships or documented histories. An illustrated article then makes another selection from the items with accessible photographs. If the surviving sample is filtered again to include only records with a weight field, the result has passed through several historical and editorial choices before the first calculation takes place.

Imagine two exhibitions drawing on the same store room. One curator selects signed blades with inscriptions that anchor a chronology. Another chooses objects whose later mounts demonstrate cultural exchange. Their selections might overlap, but a weight distribution calculated from each catalogue could look quite different. Neither exhibition has made an error. The error would come later, if a reader discarded the selection criteria and described either set of figures as representative of everything produced in Japan. Object-centred comparisons travel well because their claims remain attached to named objects.

Six questions from the catalogue desk

What measured weights does the Nagoya Sword Museum publish?

Two named examples are the Motoshige blade at 658 grams and the Kunihiro at 724 grams. Their listed lengths are close, at 68.3 and 68.1 centimetres. The same table gives 928 grams for a Masahide blade measuring 69.4 centimetres. Those values belong to the specified objects. They do not establish a standard weight for an entire historical period.

Does the Met give a weight for its Masazane blade dated 1526?

The dimensions field for accession 2001.574 gives lengths and curvature but no mass. An account of the object can cite its dated inscription and cutting-edge length with confidence. Anyone quoting a weight would need a separate, identifiable measurement with an explanation of the condition in which the object was weighed.

Does the scabbard count toward a museum sword weight?

Read the component label in the individual record. The Met gives a separate scabbard weight for the Japanese-bladed peidao discussed here. Other catalogues might give only an ensemble weight or a blade weight. Keeping the original heading next to each number prevents confusion if the data are later copied into another table.

Can a museum photograph reveal the balance point?

A catalogue view alone provides neither a measurement of mass distribution nor a reference point for reporting it. A stand or prop visible in the picture belongs to the display arrangement. Where the institution gives no centre-of-mass measurement, an honest comparison leaves the field open and explains why.

Why include objects that are not katana in this discussion?

The peidao and the smallsword illustrate how the same search can find works with Japanese components or Japanese manufacture but different historical identities. They make excellent comparisons for understanding how museums name and weigh components. Their measurements should be kept outside a sample specifically intended to describe Japanese katana blades.

What should a reader save when citing a museum measurement?

Record the institution, accession number, object title, component measured, units, and the catalogue edition or access date. Preserve qualifiers about dates and places of manufacture. If the object record gives conflicting equivalents, identify the problematic field and consult a correction or the institution before republishing that number.

Reading the record beside the image

The next time a label gives an unusually precise figure, follow it back to the institution and read the line on either side. Sometimes the adjacent words name the excluded scabbard. Sometimes they identify a later mounting. Sometimes there is enough information to set two signed and dated objects beside one another and explain exactly what differs between them. That is a stronger result than an average compiled from anonymous entries. It lets another reader identify the same works, repeat the comparison, and disagree with the interpretation without first having to reconstruct the evidence.

The four photographs accompanying this essay show objects in different relationships to their catalogue entries. Start with the Nagoya figures for actual measured examples; turn to the Masazane blade for the importance of a missing field; then look at the peidao and smallsword to understand the limits of a search by place of manufacture. Each case answers a different part of the original question. Together they recover something more interesting than the promise of one universal weight: the changing identity of an object as it moves from workshop to mounting, collection, photograph and printed or digital page.

Museum films

The Met — The Japanese Sword and the Japanese Aesthetic

The British Museum — How the samurai sword evolved over 1000 years | Curator's Corner S10 Ep11

Collection records and further reading

Related Articles