Who Were Japan's Maps Made For?

In 1838, the Tokugawa government's mapmaking program produced a map of Kii Province nearly six meters across. Roads, rivers, villages, and information about agricultural assessments occupied a sheet far too large for an ordinary traveler to unfold at a junction. Its dimensions were part of what it was made to do.^[1]

Seventeen years earlier, the survey project associated with Inō Tadataka had already delivered its remarkable maps of Japan's coasts and routes. Why did the government still need a different kind of map?^[2]

The question challenges a tempting history of cartography: first people drew picturesque approximations, then a surveyor measured the country correctly, and the old maps became unnecessary. Tokugawa Japan's surviving maps tell a less orderly and more useful story. Measurements improved, but different people continued asking different things of a sheet of paper. Accuracy mattered. So did the information selected, the format, and the people able to obtain it.

The Kii map belonged to a long tradition of kuniezu, maps organized by province and compiled under shogunal orders. In the Genroku program, ordered in 1696, the required scale was approximately one to 21,600. Villages appeared in oval forms, colored by district; their names were accompanied by kokudaka, assessments expressed in rice. Castle towns were identified differently, including the names of their lords.^[3]

These conventions made a particular kind of landscape legible. A village was not drawn principally as a cluster of houses in which one might find a bed. It was identified as a named unit within a district, associated with an assessment. The map joined physical location to information needed for administration.

An assessment was also a different thing from a measured harvest on the day the map was completed. The figure belonged to the administrative description of the village. Giving it a place on paper did not make every household's resources visible, or resolve every question about collection. It established a reference that officials could relate to other records.

The scale served that accumulation of information. A province represented on a very large sheet could carry many named settlements without reducing them to an unreadable cluster. Its size made comprehensive viewing difficult while allowing local detail to remain available. One format could favor inspection and comparison over convenient carriage.

That trade-off helps explain why a more recognizable national coastline could not replace it. A small national map might establish where Kii lay in relation to the rest of the archipelago. It would not necessarily preserve the village-level distinctions required by the provincial document.

Neither format made the other unnecessary. They organized different questions. One could be excellent at showing the shape of a country while being an inadequate substitute for the records through which parts of that country were governed.

A province was not necessarily the territory of a single lord. Producing its map could require cooperation across several jurisdictions. The National Archives identifies three domains involved in the Genroku map of Ōmi: Hikone, Zeze, and Minakuchi. The provincial sheet joined ground administered through separate houses.^[3]

That division made mapmaking an exercise in gathering knowledge as well as drawing it. The central government could specify a scale and symbols. People elsewhere still had to identify the places to which those conventions would be applied.

Shibata's municipal archive preserves a particularly concrete example. Northern Echigo's Genroku map was assigned to Murakami and Shibata domains. Information for an area belonging to Aizu had to come from Aizu before it could be incorporated. The local archive also describes requirements extending well beyond a decorative outline: river crossings, harbor depths, anchorage, currents, and whether mountain routes could carry cattle or horses in winter.^[4]

A mountain pass was consequently more than a gap between two hills. Its usefulness depended on a season and a means of transport. A river line could establish a barrier, but information about crossing it could make the same line part of a route. The map's value lay partly in converting local practical knowledge into something another official could consult.

The resulting picture should not be mistaken for direct observation by a single central survey team. It was assembled through distributed responsibilities. That structure could broaden the information available, while making the final map dependent on what contributing authorities supplied.

Later revision did not always require starting from blank paper. Shibata's account describes copies of the earlier mapping being distributed for correction during the Tenpō program. The older representation became a working document through which changes could be reported. In that local process some administrative detail was simplified, rather than every new edition simply containing more information.^[4]

The continuity lay in an organized comparison: what had changed since this account of the province was made? A map could be useful precisely because it was already old enough to require correction.

Other maps entered a commercial world. In 1779, Nagakubo Sekisui's national map was published in Osaka. A surviving copy at Nagoya University preserves publication information naming the publisher and engraver. This was a printed work, not a presentation manuscript made solely for an official office.^[5]

The distinction widened the possibilities of access. Someone could obtain the geographic representation as a product rather than qualify to inspect a particular administrative original. Printing also gave the maker a way to distribute many versions of substantially the same image.

The Hiroshima Prefectural Museum of History describes Sekisui's map as popular from the late eighteenth into the mid-nineteenth century. It included lines of latitude and longitude and improved the archipelago's outline compared with an earlier commercial tradition. The museum is equally explicit that it was not the product of a comprehensive ground survey.^[6]

That combination is important. A coordinate framework and an attractive national outline did not mean that every position had been established by the same field procedure. Compilation could bring together geographical information without reproducing the labor of measuring the whole country.

Nor does compilation make a map useless. A reader relating a distant region to a familiar one could gain something from the overall arrangement even without the precision needed for another task. The map supplied a shared space in which names, routes, and relative locations could be considered together.

Its continued circulation also changes the chronology. The publication preceded the completed Inō surveys; its popularity did not immediately end when those surveys were finished. Better measurements held within a government institution were not automatically available to everyone who wanted a map.

Availability was therefore part of usefulness. A more precise sheet that a reader could not consult might be less relevant to that reader's immediate purpose than a printed map already obtainable. The history of geographical knowledge includes the means of reaching an audience, not only the means of measuring a coast.

At the other end of the scale stood Edo's kiri-ezu, maps dividing the city into separate areas. The National Diet Library catalogues an Owariya set issued between 1849 and 1862 with sheets around fifty by fifty-four centimeters, folding in the cited format to seventeen by nine. Their physical design made possible a relationship with the user very different from that of the enormous Kii map.^[7]

The Owariya publisher issued revised maps from 1849 into 1870. The sheets distinguished roads and bridges in yellow, water in blue, temple and shrine grounds in red, and townspeople's areas in gray. The principal residences of territorial lords carried family crests. Important sacred places could receive recognizable buildings and trees rather than being reduced to identical symbols.^[8]

These choices gave the city several kinds of address. A person could identify a road through its color, a major residence through a crest, or a destination through its name and visual prominence. Reading did not require every feature to occupy the same level of abstraction.

The small folded format was more than a convenience added after the map was designed. It made a sheet something that could be carried, put away, and consulted again. The provincial map asked for space in which to be opened. The neighborhood map could accompany movement through the place it represented.

The publisher's business helps explain the presentation. Owariya also dealt in popular printed books and pictures. Its mapmaking participated in a trade that knew how to combine lettering, recognizable imagery, and color.^[8] Practical information did not have to look austere in order to be useful.

Yet the sheet did not give every inhabitant an individual address. A named lord's compound and a shaded area of ordinary urban housing were different units of attention. The distinctions made the city easier to read for certain purposes while preserving the prominence of some occupants over others. A portable map could be widely usable without being a socially complete portrait.

Those neighborhood sheets do not join like uniform modern tiles. Their orientations and scales vary. Researchers at the Center for Open Data in the Humanities, comparing the Owariya maps with present geographical positions, identify larger distortions toward Edo's outskirts, where notable destinations were fitted into the available frame.^[9]

That finding makes the physical page part of the explanation. A publisher working with a bounded sheet could enlarge, compress, or rearrange its geographical coverage to include places a reader might seek. The paper had edges even where the city did not.

This was a different problem from depicting an entire province at a prescribed uniform scale. The neighborhood sheet needed to be intelligible within a particular patch of urban life. Its usefulness could depend on including a conspicuous destination that would otherwise fall outside the frame.

A distortion could nevertheless mislead. A map suited to recognizing a sequence of places might give a poor basis for comparing precise distances. Explaining the publisher's problem does not turn every inconsistency into a hidden achievement. It tells us which assumptions a reader could safely carry from one kind of map to another.

The contrast is especially revealing when the sheets are assembled into a single modern display. A contemporary viewer may expect all their edges to agree because they belong to one city. Their original division into separate products did not require such seamlessness. Local usefulness and effortless national—or even citywide—integration were different design goals.

The maps therefore invite two questions rather than one. How faithfully does this sheet preserve spatial relationships? And which practical relationship with the reader shaped its selection and arrangement? The first cannot be abandoned. The second explains why historical mapmakers did not all produce imperfect versions of the same ideal page.

The Inō project answered a demanding set of geographical questions through systematic measurement. Its parties recorded distances and directions along coasts and routes, used bearings toward visible features, and checked latitude through astronomical observation. The resulting work related places through accumulated measurements rather than solely through inherited descriptions.^[10]

This changed the basis on which positions could be trusted. A name copied from an earlier map might preserve valuable knowledge of a location. A series of measured connections could test its relation to another location. Repetition and comparison helped make a coherent larger outline possible.

The survey was not simply one man converting footsteps into a finished picture. Inō died in 1818. Takahashi Kageyasu and the project's collaborators completed the submission to the shogunate in 1821. The final set comprised 214 large sheets, eight medium sheets, and three small sheets.^[2][11]

Their scales disclose choices about use. The large maps were about one to 36,000; the medium and small maps compressed the work to about one to 216,000 and one to 432,000. A detailed part of the survey and an overview of the archipelago could not carry the same density of information on the same amount of paper.

An overview gains relationships by losing detail. A local sheet can preserve distinctions that disappear when the reader tries to see a much larger territory at once. The production of several formats was therefore part of the accomplishment, not an admission that the surveyors had failed to decide on the correct map.

The coast-and-route network also remained a particular kind of knowledge. It did not automatically replace every village assessment, identify every urban resident, or describe the seasonal condition of every mountain path. Its precision should be appreciated without making it answer questions it had not been organized to settle.

At the moment the national survey achieved its most coherent form, Japan still had practical reasons to make other maps.

The 1821 submission was not initially a commercial atlas placed on sale beside Sekisui's map. Its official custody affected who could use it. But the work's later history also warns against imagining one original permanently sealed away from every other institution.^[11]

A set of small Inō maps now at the Tokyo National Museum had been deposited at the shogunate's academy, Shōheizaka Gakumonjo, in 1822. Hoshino Yoshihisa's account identifies the academy's stamp and the needle holes used to transfer positions between superimposed sheets. Copying was skilled production with material traces of its own.^[12]

A copy could extend access without requiring a fresh survey. It could also place geographical knowledge within an institution concerned with learning rather than leave it attached only to the field party that produced the observations. Custody was one of the ways a map acquired an audience.

Later, a printed official map drew on Inō's smaller format. Tokushima University's catalogue identifies an 1867 edition issued through the Kaiseijo, the shogunate's institution for Western learning, and a reissue in 1870 using the same printing blocks under its successor institution.^[13]

The administrative transition did not make the measured coastline useless. Nor did using the same blocks make the two editions identical in political setting. The geographic work could continue while the institution authorizing its publication changed.

Copies eventually became essential to survival as well as circulation. The original presentation maps were lost in the palace fire of 1873. Other sets and copied sheets preserved evidence of the project.^[2] A history that values only the lost original misses the objects through which the work remained available to study and use.

The large provincial map, the commercial national sheet, the folding neighborhood guide, and the measured survey did not merely represent successive levels of intelligence. They organized different relationships between places, information, and users.

The provincial document could help officials relate a settlement to a district and an assessment. The commercial national map made a broader geographical picture obtainable. The neighborhood sheet combined portable form with names and visual distinctions useful within a city. The survey established connections whose reliability rested on a particular program of measurement.

Their differences do not mean that accuracy was optional. A wrong name could defeat one use; a wrong distance another. They mean that an accurate answer must still be an answer to a question somebody needs to ask.

That is why the immense Kii map could remain meaningful after the Inō project. A better outline of the archipelago did not abolish local government, and it did not fit every administrative fact into three small-scale sheets. What was left out of one map could be the principal reason for making another.

The most revealing question to put beside a historical map's legend is consequently a practical one: what did this arrangement allow its user to do? Follow that question, and the old sheet stops being merely a less exact view from above. It becomes evidence of a particular encounter with the country—made possible by some carefully chosen information, and limited by what the page did not contain.