Drawing the Curve: How Historical Catalogues Illustrated Sori

On one opening of Shūko jisshu, a Japanese collector’s illustrated survey, the long curved shapes stand almost upright. A reader coming from a modern museum website might instinctively turn the image sideways. Try leaving it as it is. The narrow drawings and the vertical writing share a direction. The eye moves between a line and its annotation over a much shorter distance than it would cross on a horizontal gallery label. What looks unfamiliar on a screen makes sense on a page held open for reading.

The National Diet Library holds the edition illustrated here. Its bibliographic record names Matsudaira Sadanobu as compiler, gives Ikubunsha as publisher and dates the edition to Meiji 36–38, catalogued under 1905. The book reproduces a much older collecting enterprise. Japan Search, in a gallery curated by the library, describes the larger survey as including 127 blades. That number describes the historical scope of the undertaking. It cannot be derived by counting the 53 digitized frames in this volume, some of which include two pages and several quite different sorts of illustration.

The word sori gives us a way into these pages. In its English teaching guide Kyoto National Museum uses the term for the depth of the arc of a blade. The subject here is what happens when that three-dimensional object becomes a drawn curve and when the drawing in turn becomes a catalogue image. Four complete openings serve as our exhibits. They show a reader the decisions about orientation, grouping and space that an isolated outline would conceal. They also show why a clear diagram can be historically important without being a scale drawing.

One opening has room for several objects

Two facing book pages with long narrow drawings and vertical Japanese annotations
Shūko jisshu, sword volume, frame 7. The figures are reproduced in the orientation of the book. National Diet Library, Japan. Public Domain Mark.

In digitized frame seven, the vertical divisions of the opening are conspicuous. What appears at first glance to be a single, complicated figure resolves into several long elements, some much darker than others, and text at their sides. The dark forms hold the eye from farther away. The fine lines require a closer look. A reader can move from one kind of evidence to the other without turning a page. An index of names would never convey that experience, and a crop of one attractive detail would discard the comparison that the layout affords.

A long object is awkward to put into a book of ordinary proportions. Showing it at its full extent leaves relatively little paper for local detail. Enlarging a short portion makes it legible, but the reader then needs a way to place it in relation to the whole. Turning one component, moving it sideways or aligning several ends can all free up space for writing. A historian needs to know which of those things the page actually does before treating empty space as an extension of a represented object. An apparently missing section may belong to another view.

The crease at the middle of the opening is an instructive interruption. We can follow the printed border and see which marks belong to the left or the right page. Without the border, an enlarged digital image could make two neighbouring views look like a single composite diagram. Someone preparing an exhibition label from such a crop might describe a feature as adjacent to the blade even though a binding separates the two. The complete opening gives us an inexpensive check on that error before we try to identify any individual component.

The baseline in a modern teaching picture

Kyoto National Museum’s illustrated guide takes a different approach. It introduces parts of a blade by putting the Japanese terms alongside English explanations. One group of labels distinguishes blade length from curvature. Others name the tang, the tip and the point where an inscription may occur. Every label directs the reader toward a particular question about the image. A line indicating the depth of the arc has a different job from a line describing the outline. Both can appear equally dark on the printer’s paper, yet only one represents a feature physically present along the object.

That distinction brings an often unnoticed third line into view. To indicate how far a curve bows, an illustrator can supply a straight visual reference. It belongs to the apparatus of the diagram in the same way that the arrow beside a label does. A reader trained by modern museum graphics may unconsciously supply that straight reference when looking at an older drawing. It is perfectly reasonable to do so for comparison. It is a different matter to report that an eighteenth-century compiler drew or measured against a particular line if none appears on the surviving page.

The Kyoto Museum has issued more than one educational guide. The earlier and later sheets belong to a tradition in which a museum translates specialist vocabulary into things visitors can locate with their eyes. Printed labels, leader lines, white space and a controlled angle of view do the work a guide might otherwise do by pointing. A facsimile of an illustrated collection serves another audience. It may assume that its reader already recognizes what is shown and devote more space to differences among specimens. The same gently bent line becomes evidence in two different conversations.

A turn of the page changes the comparison

An outlined curved form fills much of the left leaf opposite several smaller figures
Frame 12. A sweeping dark line on one leaf contrasts with the more delicate figures across the gutter. National Diet Library, Japan. Public Domain Mark.

Frame twelve has a particularly legible sweeping form in one half of the opening. On the other half the eye moves among relatively slight vertical marks. The apparent curve stands out partly because there is uninterrupted light paper next to it. The asymmetry of the opening makes it tempting to select the more spectacular shape for reproduction. Yet keeping the quieter side explains why the curve looks so emphatic. Strong contrast is an attribute of this composition, as well as a possible attribute of the represented form.

Consider how a straight rectangular border alters the experience of a line close to it. When the gap narrows and opens again, the eye picks up the change immediately. Put the same printed line on an unbounded background and its direction becomes harder to describe without a point of reference. A book designer who reproduces an isolated silhouette on a blank page can thus change the apparent prominence of its curvature while preserving every pixel of the silhouette itself. No change in the historical object is required to explain the difference.

The effect matters in searches for historical types. A thumbnail grid may display a short, strongly marked object at exactly the same height as a long, slender one. The viewer can learn a good deal about the shapes of the two outlines from such a grid. The viewer cannot recover their relative physical sizes from the screen height. A museum may give precise dimensions a few lines farther down in each record. The number and the picture support each other only after the reader has checked that they refer to the same component and view.

What the compiler collected

Matsudaira Sadanobu is remembered in this connection as a collector of visual knowledge. The title Shūko jisshu signals a wider survey of antiquities. Japan Search describes it as a collection of copied illustrations made under his direction and singles out a drawing of the Kogarasumaru tachi associated with the Ise Sadatake family in its presentation of the work. This is already a story about more than one person and one act of seeing. A drawing could circulate from a household to a compiler, enter a published selection and be chosen again for a digital gallery.

Each stage gives the picture a new neighbour. In a household group it might have been stored beside letters or other drawings inherited together. In an illustrated survey it sits among examples selected for an intellectual purpose. In the National Diet Library’s online gallery it appears near material as varied as books of names, theatrical stories and photographs of museum objects. None of these arrangements is meaningless. A sequence established by a modern curator can illuminate a historical relationship while telling us very little about the order in which the original owner kept the material.

A second entry in the gallery makes the point from an unexpected angle. It connects the name Kogarasumaru with a pictured story of Taira no Koremochi, and explicitly signals that the identification involved is one of several traditions. Put that entry beside a line drawing with the same name. One image documents a named design in a work of antiquarian scholarship. The other places the name within narrative history. Their shared label makes comparison possible. It cannot establish that a narrative artist consulted the same physical blade that stood behind the antiquarian drawing.

Curves in company

Drawings of curved mounted forms with projecting fittings extend across an opening
Frame 14. Mounted outlines and the spaces around projecting elements occupy much of this opening. National Diet Library, Japan. Public Domain Mark.

Frame fourteen is crowded by comparison with frame twelve. Several long forms run through the opening, and projecting details draw the eye out toward the writing. The silhouette of a mounted object can carry the curve of its casing, the proportions of its fittings and the direction in which an attachment projects. Tracing all of that as a single undifferentiated line would answer a different question from comparing the backs of two bare blades. Here the written descriptions and the distinction between exterior and interior outlines matter more than the sheer elegance of a line.

A modern collection catalogue offers a useful check. The National Diet Library’s curated selection includes the Uesugi tachi. The accompanying entry distinguishes an attributed blade from a decorated mounting and specifies the designation that applies to the mounting. The reader who sees only a long exterior in a photograph might be inclined to transfer the description of one component to the whole. The catalogue’s division allows a more exact reading. A drawing of a mounted profile can be valuable testimony about the exterior even when it provides no direct view of the blade within.

Another object in the selection, Mikazuki Munechika, shows why a curved word in a historical name needs particular care. The name evokes the crescent moon. The museum relates that description to small crescent-shaped features along an area of the steel. It does not describe the great sweep of the entire outline. A picture editor seeking an attractive caption might accidentally join the two scales of observation. Looking up the institution’s explanation restores the local feature and also tells us what the striking nickname meant to the people who transmitted it.

Scale, orientation and the screen

A digital reproduction can change a page in two ways that are easy to confuse. It can enlarge the whole image equally, or it can stretch one direction more than the other. The first change leaves the proportions of the drawing intact while removing any direct correspondence between the screen and the book. The second changes the curve itself as it appears to the viewer. A facsimile fitted into a fixed website banner is especially vulnerable. Preserving the relation between the height and width of the full page protects the evidence more successfully than a caption promising a high-resolution detail.

There is a similar problem with rotation. A whole opening turned through a right angle retains its shapes, but its relationship between drawing and writing becomes difficult to read. Mirroring the image introduces a different distortion. It reverses the disposition of the inscriptions and exchanges the faces as the viewer encounters them. Small written characters can often reveal the reversal quickly. The historic orientation should remain available even if an exhibition also offers a rotated enlargement. Otherwise the visitor can end up comparing the modern designer’s convenient view with the original compiler’s quite different arrangement.

Several vertical profiles aligned inside two rectangular page borders
Frame 21. The full spread preserves the grouping of figures within rectangular borders. National Diet Library, Japan. Public Domain Mark.

Frame twenty-one makes particularly good use of repeated narrow forms. On one side of the opening outlines are close enough to invite a side-by-side reading. On the other they share space with a relatively open field and separate areas of script. To ask which one is the most deeply curved is less fruitful than first asking which features define comparable views. Does each outline include the same extent? Are all the figures bounded at both ends? Does an adjacent drawing supply another face or another object? The arrangement makes these questions visible before an answer is available.

Beyond the illustrated book

A visitor standing before a museum case has other ways of seeing. The visitor can change position, follow a surface as light shifts, and discover which details disappear when a polished area reflects the surroundings. Those experiences are difficult to put into a line engraving. On the other hand, an open book can bring several objects into a proximity that no lender, exhibition timetable or case arrangement could guarantee. Historical catalogues retain special value precisely because their assemblies can survive the dispersal or temporary inaccessibility of everything assembled on the page.

Bashford Dean’s work in New York offers a later setting in which to think about the problem. The Metropolitan Museum has published an account of his role in presenting Japanese arms and armor. The historical gallery views reproduced there place individual objects within a much larger display. The visual experience of a long form in a case changes when stands, labels and other objects occupy the same field of view. A gallery photograph gives those elements back to the historian while losing some of the selective clarity an illustrator can achieve on paper.

The films accompanying this article let the comparison go one step further. The account of the Kusanagi-kan at Atsuta Jingū introduces a space arranged for visitors today. The Metropolitan Museum’s film discusses conservation of Japanese works in overseas collections. Neither film reconstructs the working room of Matsudaira Sadanobu. Each makes visible the modern institutions between a surviving object and its public image. Even a seemingly neutral catalogue photograph depends on decisions about lighting, supports, orientation, stage of conservation and the amount of surrounding space a reader will see.

Making a historical comparison accountable

A persuasive comparison names the edition and the view before it describes the line. In our case that means the Ikubunsha edition of the sword volume, now at the National Diet Library, followed by the digital frame. It means specifying whether the remark concerns one side of an opening, a whole exterior, a detail or the modern guide’s reference line. The source remains legible even if the reader disagrees about what the mark depicts. This small courtesy is more reliable than assigning a precise numerical curve to an image whose reduction history is unknown.

It also helps to preserve differences of vocabulary in translation. Depending on the sentence, an English caption might need “curvature,” “the arc,” “profile” or “the outline of the mounting.” Substituting sori for all four makes prose sound specialized while erasing distinctions the images have preserved. Kyoto’s diagram supplies a specific technical use. The illustrated openings allow a much wider investigation into why a line looks as it does. One of the pleasures of reading the two together is discovering exactly where an everyday word like “curve” stops being precise enough.

The enduring achievement of the old survey is that it lets us look across the book at once familiar and unfamiliar things. A piece of writing gives scale to a neighbouring drawing. A border reveals a change of direction. A projecting attachment complicates what appeared at first to be a simple silhouette. Those relations were selected and preserved by people with a reason to put the shapes on paper. Seen in that setting, the curve is more than a measurement to retrieve. It becomes a record of the ways successive readers learned to see an object together.

Questions about illustrated curves

What does sori mean in Kyoto National Museum’s diagram?

The guide uses it for the depth of a blade’s arc. It labels blade length separately.

Which edition supplies the four illustrated openings?

They come from the Ikubunsha edition of the sword volume of Shūko jisshu. The National Diet Library records Meiji 36–38 and indexes the volume under 1905.

Does the number 127 refer to the digitized frames?

Japan Search gives 127 as the number of blades in the historical survey. Frames count images in the digital copy and may contain several drawings.

Why reproduce the gutter and border?

They identify which figures share a page and show the straight references and empty space against which a curved outline appears.

Does the crescent in Mikazuki Munechika refer to the entire arc?

The museum account connects the name with smaller crescent-shaped features observed along the blade. It describes a local feature rather than the entire profile.

Can two equally tall thumbnails establish the original relative sizes?

The objects may have been reduced at different scales. Matching display heights do not establish matching object heights or curvature measurements.

Museum films

The Met — Conserving Japanese Works of Art in Foreign Collections

Nomura Group — 熱田神宮 剣の宝庫 草薙館

Collection records and further reading

Related Articles