In a guide produced for a Kyoto exhibition, two dark strips of metal occupy the space normally given to a single illustration. The photograph on the left is labelled back. The one on the right is labelled front. A date runs down one face; the other credits a man with making the blade and its carving. Someone reading from left to right encounters the date first. Someone reading the English object title meets the maker first. Neither sequence is inherent in the work. They are two perfectly reasonable ways for a museum to present writing that no single view can show in full.
The work is the long sword by Awataguchi Ikkanshi Tadatsuna identified as number 160 in the Kyoto guide. The date falls in 1709. A different museum, the Metropolitan in New York, describes the writing on both sides of a katana by Masazane dated 1526. Together they make an unusually good introduction to tachi-mei and katana-mei, terms that can sound far more mysterious than the objects require. Three things can change independently in these examples. The words on the metal can change, the surface on which we find them can change, and the order in which a publication invites us to read them can change.
Which way is front?
A signature on a Japanese blade is often on the tang, the portion hidden by the grip when the work is mounted. Historical descriptions orient its faces by reference to dress. Tachi and katana were conventionally presented in different orientations at the wearer’s side. A signature placed on the outward face would therefore occupy opposite faces in the two conventions. This is the historical point behind the vocabulary. A label that calls one side front is giving the reader a relationship between an object and its presentation, much as the obverse of a medal remains its obverse when a museum photographs it second.
It follows that the left edge of a computer screen cannot identify the front of a blade. A publisher may show the photographs vertically to give a line of writing sufficient height. Another may turn the overall view horizontally to fit across two pages. A third might put the two faces next to each other at the same scale. Left and right describe a layout. Front and reverse on a museum caption describe faces of the object. An account of dress explains the convention. The three descriptions can use different directions without anyone having made an error.
Kyoto’s guide lets readers see that relationship unusually clearly because the two directions occur on one sheet. Back appears to the left of front; the reverse carries a date and the front credits two activities to one named maker. Take away a single caption and it would be easy for a reader remembering the guide a week later to reverse the order. Retain the captions and the paired illustrations become a document in their own right. They preserve which physical face supports each part of the museum’s reading as well as what the two texts say when combined.
A date with an actual day
Masazane’s katana in New York has its maker’s statement on the front. The reverse records the twelfth day of the eighth month in the sixth year of Daiei. The Met gives the year as 1526 and renders the date in English. Here a day number is preserved. Compare the Tadatsuna example, which specifies an auspicious day in its eighth month. The second date can be precise about the month without giving a numerical day. A modern caption that supplied a calendar day in both cases for the sake of typographic uniformity would give one of them a precision it never had.

There is a second editorial choice even when all the old date elements are present. An English-language reader may take a familiar month name as a statement about a modern civil calendar. A careful account of an inscription can preserve the sequence of era, year, month and day, then give the year adopted in the museum catalogue. A precise conversion to a particular day on a different calendar is a separate calculation. A caption seldom needs to do all of this. It does need to tell the reader which part of the information is a reading of the metal and which part was introduced to orient a modern visitor.
Notice also that the maker’s words are a claim about a particular act. Masazane’s name occupies one place in a short statement about making. Tadatsuna’s statement in the worksheet credits blade and carving. The second object therefore preserves a connection between two kinds of work that a catalogue might otherwise record under separate headings. A database designed to hold only one short artist name could find both works. It would be much less effective at explaining why the second inscription matters to a historian of names, specialized labor or the presentation of an artist’s achievement.

The life of a photographed letter
The Metropolitan Museum has published the Masazane work in a catalogue of notable acquisitions as well as in its online collection. The printed account includes a detail specifically identified as the inscription on the front of the tang. An overall photograph gives the reader the arc of the whole piece, but a letter that barely occupies a dot in a printed overview can occupy a substantial portion of a separate figure. That change of scale is a decision about what a reader should be able to inspect. The publication’s figure number is useful evidence alongside the museum accession number.
In a narrow picture, adjacent strokes may seem continuous even when they are separated by dark metal. The edge of a reflection can gain more visual weight than a shallow line. Someone working from an English catalogue does not have to offer a new reading of the characters in order to discuss them responsibly. The printed transcription, the photograph and the museum’s English rendering can be used as three distinct sources of information. One records what the publisher read. Another lets readers reconsider the visual evidence available in that edition. The third explains the sense selected for an English-speaking audience.
The name Masazane also appears in a fuller artist description linked to Ise, a historical region. It would be easy to read that artist heading and assume that Ise enters the inscription on the illustrated face. The Met’s transcription gives a more exact account of what the front actually says. A modern maker biography can provide a place, a period of activity and an artistic affiliation without any of those items being written on an individual work. Transferring a useful catalogue fact into a transcription produces a particularly convincing kind of mistake because every word used in the mistaken account may be factually defensible somewhere else.
A convention with a history
The practice of placing a name on the outward face was never sufficient to decide the history of every surviving signature. The Japanese sword preservation society’s July 2023 account discusses Aoe examples and explicitly distinguishes more frequent placements from unusual ones. In one discussion, the number of characters and the face chosen for the inscription are considered together; the report also mentions dated examples that depart from an expected arrangement. This kind of specialist comparison puts an elementary rule in perspective. What works very well for introducing two terms can be too blunt an instrument for writing the biography of a single piece.
A simple thought experiment shows why a general rule cannot do all the work. Imagine an old exhibition print with two photographs but no surviving caption. A scholar proposes that the text on the left is on the front because the exhibition mostly contained katana. Even if the list of exhibits is accurate, that fact does not identify which object was photographed, which way the operator turned it or whether the publisher reversed the order of the photographs. An object number and the original caption would shorten the investigation enormously. Statistical familiarity with a convention is a much less direct connection between a picture and a particular piece of metal.
The museum records also preserve historical changes to individual objects. The Kyoto National Museum entry for E甲215 names Rai Kunimitsu and Umetada in its title, and the curatorial account connects Umetada Myōjū with an Edo-period alteration of the earlier work. The title is therefore a small history of different contributions. A reader who translates the conjunction between the two names as joint authorship in a single workshop would have to bring together people from different periods. The catalogue gives us a better solution. It records the earlier maker and names a later intervention as part of the same object’s documented life.
Three signatures for three things
One object in the 1891 Ives and Walters gift demonstrates another reason to keep component and face separate. Its blade has the name Kuniyoshi and a date of 1525. The museum identifies inscriptions associated with Toshimasa on the fittings and Kajikawa on the scabbard. The mounts date to the nineteenth century. The three names therefore answer a quite different set of questions from the front and reverse of a single inscribed tang. One account distributes information over two surfaces of one part. The other distributes evidence across a number of parts assembled into one object.

A museum might publish the second as a series of consecutive photographs. To a reader scrolling through thumbnails, movement from one frame to the next can feel the same in both cases. Yet moving from one face of the blade to another preserves the authorial question under discussion. Moving from the blade to the scabbard changes it. A well-designed caption can identify the transition in a handful of words. Each photograph should carry the accession, the illustrated component and, when the evidence warrants it, the relevant maker or location of the inscription. The reader should not have to infer any of these from a thumbnail’s position in a row.
That distinction can recover later artists from the shadow of an earlier name. A catalogue entry beginning with a famous smith is easily cited as though everything that follows were made at the same moment. If a nineteenth-century inscription is on a fitting, it may be the most direct available evidence for the person responsible for a feature that catches the modern visitor’s eye. The caption naming the blade first is serving an institutional convention. The object itself gives the historian good reasons to begin somewhere else on a second visit.
Reading a seal that looks back
A study of Gassan Sadakazu published in the Metropolitan Museum Journal makes still another use of paired photographs. Its first group of illustrations includes an object in the Friedsam collection, accession 32.100.470, and close views of the signed and dated faces. Among the marks discussed is an owl-shaped seal that contains the character sada. The reverse carries a date in the third year of Meiji, 1870, also expressed in the animal cycle. A reader following the catalogue will encounter an animal as the outline of a seal and an animal designation within the calendar. These details belong to different systems of meaning.
What can an owl do in a catalogue that a typed name cannot? In a photograph its outline distinguishes the seal from the ordered rows of lettering around it. In a transcription the name of an animal alerts the reader to something difficult to convey in type. In an attribution argument its occurrence on more than one work makes comparison possible. None of these functions requires the owl to become a universal sign by which anyone can assign every work that carries a similar outline. The published study gives an accession and figure numbers precisely so that the individual evidence can stay attached to the individual case.
An animal year also recurs in a cycle. In the example under discussion an era year accompanies it, supporting a close chronological reading. Another work considered in the article raises a different problem, and the author weighs possible years rather than treating the presence of an animal sign as a complete date. That comparison is part of the value of a research publication over a strip of neatly transcribed characters. It tells readers where a calendar expression is nearly self-sufficient and where the historian’s argument depends on evidence about an artist’s working life.

The two faces in one sentence
The nineteenth-century Naotane record illustrated here has photographs taken from more than one side. The existence of those views is enough to demonstrate an elementary point about publishing. “Side two” in an image sequence identifies a photographic relationship. It does not, on its own, say where a particular inscription is or what it records. For that the reader returns to the object record or a specialist description. The most satisfying museum pages allow the two forms of inquiry to meet. Readers can choose their own visual route through the photographs and still recover the particular evidence behind a written claim.
A useful caption for the Kyoto teaching example could be written in a single sentence if necessary. It would identify the long sword and Tadatsuna, mention the credit for carving, locate the auspicious date on the reverse and preserve the year 1709. The Masazane example needs a slightly different sentence because it supplies a numerical day. The Kuniyoshi accession needs another kind of sentence altogether, one that assigns different names to the right components. The aim of learning the terms is the ability to notice when the sentence has to change. The surprise is how many kinds of history can survive on a pair of surfaces small enough to fit in the palm of a reader’s photograph.
An inscription in three editions
An inscription can acquire a history on paper as well as a history in metal. At the end of its entry on Masazane, the Metropolitan Museum’s acquisitions catalogue lists earlier publications, including references to rubbings. It also illustrates the front of the tang. A reader following those references would be able to ask how the same marks circulated before a museum put a color photograph on the web. A rubbing, a printed detail and an online photograph all translate a three-dimensional surface into an image. Each preserves a different relation between the letters and the space around them.
Consider the problem of scale. A photograph printed at the end of a long object entry might be given just enough space to identify the line under discussion. On a digital screen the same view can expand until the metal texture occupies most of the picture. In a rubbing reproduced in a book, the margins and the overall shape can be especially important for relating the writing to its support. The best account of a difficult reading would identify which of these reproductions the author actually saw. Giving only the date of the newest publication can make an older interpretation look like a judgment reached independently from a new image.
There is a small reward for taking these bibliographical details into the gallery. A visitor can stand at a distance from an object and understand how much of the exhibition’s scholarship is necessarily elsewhere. The display may show the whole blade to establish its shape. A guide can set its two inscriptions on the same page. A specialist publication can bring in other dated objects or an earlier rubbing. What seems like missing information in one medium can be entirely legible in another. Learning to connect them by the accession and the figure number is often more productive than trying to extract a complete historical argument from the one surface that happens to face the viewer.
Questions about inscription records
What do tachi-mei and katana-mei describe?
They distinguish conventions for locating a signature on opposite faces of the tang, understood in relation to historic dress. The terms are useful when reading catalogues or matching photographs. A named convention gives evidence about orientation; the attribution, object type and identity of the inscriber require their own evidence.
Which side of Kyoto’s illustrated Tadatsuna has the date?
The visitor guide labels the date on the back, reproduced on the left of its paired photographs. The front appears on the right and carries the statement crediting Tadatsuna with the blade and its carving. Those directions apply to this layout; in a different publication the images could change places.
What date does the Met give for the Masazane blade?
The museum renders its reverse inscription as the twelfth day of the eighth month in Daiei six and gives 1526. The front credits Masazane with making the work. In discussing calendars preserve the era expression or quote the museum’s rendering rather than supplying an independent modern day conversion without a calendar source.
Did a single person sign every component of the Kuniyoshi ensemble?
The Metropolitan Museum distinguishes the blade inscription from those associated with Toshimasa on the fittings and Kajikawa on the scabbard. The blade is dated 1525; the mounting belongs to the nineteenth century. A photograph captioned simply with the name Kuniyoshi can obscure what the other inscriptions describe.
What does an owl contribute to the reading of Gassan Sadakazu?
A Metropolitan Museum Journal study records an owl-shaped seal containing the character sada on one of the works it illustrates. The opposite face bears a Meiji-period date. The seal, text and date are separately described forms of evidence and can be compared through the figures in that study.
Does writing on the unexpected face establish a false signature?
Historical practice varied. The Japanese sword preservation society discusses differences within the Aoe tradition, including documented instances that depart from a frequent placement. A judgement about an individual object belongs with its complete museum record and specialist research. Orientation alone cannot carry that judgement.
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
- Kyoto National Museum, 2018 visitor guide, no. 160
- Masazane, 2001.574
- Kuniyoshi, 91.2.54
- Naotane, 91.2.56
- Rai Kunimitsu and Umetada, Kyoto National Museum
- Nihon Bijutsu Tōken Hozon Kyōkai, July 2023
- Metropolitan Museum Journal, A Study of the Works of Gassan Sadakazu
- Metropolitan Museum, Notable Acquisitions, no. 53
- Reading a Japanese Sword Exhibition Catalogue: Dates, Eras and Corrections
