How Did Japan Decide What Could Be Printed?

In 1842, during the Tenpō Reforms, the Tokugawa government ordered the Shōheizaka academy in Edo to examine books. A surviving compilation known as Kaihan shishin preserves material connected with that system: instructions, cases, and records concerning the inspection and circulation of publications. The reform had turned an institution associated with learning into one of the places through which publishing could be judged.^[1]

That arrangement makes censorship sound straightforward. The government sets a rule; a censor checks a book; the printer either receives permission or does not.

The history is less tidy. Publishers altered works before an official saw them. Magistrates could punish books whose relationship to a rule was contested. The Meiji state replaced Tokugawa institutions with national press and publication ordinances, but it did not replace uncertainty with a single transparent boundary. By the 1930s, control operated through laws, police bans, informal guidance, resource allocation, and state publishing at the same time.^[2]

The recurring problem was scale. Once print became cheap and commercially valuable, no government could personally compose every acceptable sentence. It needed chokepoints: people and institutions that could be held responsible for what entered circulation.

So who decided what could be printed? Sometimes an official did. Often a publisher decided what an official might tolerate. At other times the state shaped the field so strongly that choosing what to produce became inseparable from anticipating control.

Tokugawa censorship grew inside a commercial publishing world. Woodblock books could be produced in substantial numbers because the carved blocks allowed text and illustration to be printed repeatedly. Publishers invested in authors, artists, block cutters, paper, printing, distribution, and the expectation of sales. A book was therefore not only an expression. It was an enterprise with identifiable people who could lose money, inventory, blocks, or the right to trade.

That fact gave authorities leverage.

Regulating every reader was difficult. Regulating the businesses that multiplied texts was more practical. The publisher sat between creation and circulation. If the publisher could be punished for a work, then the publisher had a reason to ask questions before paying to make it.

The Kansei Reforms of the late eighteenth century expose how unstable that arrangement could be. The writer Santō Kyōden and the famous publisher Tsutaya Jūzaburō were punished in 1791 for sharebon, fashionable books associated with the licensed pleasure quarters. Later accounts often linked the case to a publishing edict supposedly issued in the previous year.^[3]

Historian Yamamoto Hideki has challenged that neat chronology. His study of surviving magistrate materials argues that the supposed May 1790 edict was probably not formally issued then. He points to a town magistrate's letter indicating that an edict was delivered to book wholesalers in October, after the books at issue had already appeared. If that reconstruction is correct, Kyōden and Tsutaya could not simply have read a clear published rule in advance and chosen to violate it.^[3]

The disagreement matters beyond one punishment. Censorship is often described as if the line between permitted and forbidden content already existed in a complete code. Publishers merely approached the line, and officials caught those who crossed it.

But control can work even when the line is uncertain. A punishment teaches the trade what an authority now considers dangerous. A magistrate's warning, a confiscation, or a penalty can turn one case into guidance for the next. The cost of uncertainty falls partly on the publisher.

This produces a distinctive kind of discipline. The government does not have to predict every future joke, erotic scene, political allusion, or fashionable genre. It can make the people financing reproduction predict the government's reaction.

The Tenpō Reforms of the 1840s made this anticipatory pressure particularly visible. The reforms attacked forms of urban consumption and entertainment that officials associated with extravagance and moral disorder. Publishing was one arena in which those concerns became concrete.

Literary scholar Tsuda Mayumi compares two works separated by roughly half a century. During the Kansei Reforms, Santō Kyōden had written a strongly didactic story. During the Tenpō Reforms, his brother Santō Kyōzan reworked it into a still more moralizing version. Tsuda notes a comic but revealing visual change: figures who had originally been nearly naked were clothed in the later remake as part of excessive self-regulation in response to censorship.^[4]

No official had to stand beside the block cutter and order each garment onto the page. The expected reaction of authority had already entered the production process.

That is an important difference between censorship and deletion. If we look only for works formally banned after publication, we miss the books that were never drawn, never cut, or were redesigned before an inspector saw them. The missing version leaves a weaker archival trace precisely because control succeeded early.

At the same time, self-censorship did not make official review unnecessary. Shiroto Makiko's study of Kaihan shishin identifies the compilation as a manual and record associated especially with the censorship arrangements established under the Tenpō Reforms. In June 1842 the government ordered the Shōheizaka academy to examine books under the banner of improving learning. The surviving material includes cases concerning publishing, selling books produced outside Edo, and even trade in manuscripts.^[1]

The range is revealing. Control concerned more than the instant at which ink met paper. Authorities had to think about geography and circulation. A work produced elsewhere might enter Edo. A manuscript could circulate without having passed through the same commercial process as a printed book. A system centered too narrowly on one printing shop would leave other routes open.

The response was institutional multiplication. Town magistrates, scholarly authorities, publishers, wholesalers, and other participants could all become points at which a work's movement was slowed, inspected, reported, or discouraged.

This did not create perfect control. Satire survived. Publishers tested boundaries. Rules could be criticized, evaded, or applied unevenly. What changed was the cost of pretending that publication was only a private relation between writer and reader.

A commercial book passed through hands. Censorship turned some of those hands into accountable intermediaries.

The Meiji Restoration dismantled Tokugawa political institutions while creating a much more centralized national state. Printing technology, newspapers, political debate, and the scale of circulation were all changing at the same time. The old publishing controls could not simply be left in place unchanged.

The new government responded with national legislation. Hara Hideshige's legal history traces Press Ordinances in 1869 and 1875, and a further Press Ordinance in 1887. He emphasizes that these measures developed alongside copyright rules and drew on foreign legal models: British law influenced the earlier ordinances, while the 1887 press legislation drew on the German Press Act of 1874.^[5]

Foreign models did not make the resulting regime politically neutral. Law supplied the government with nationally standardized ways to identify responsible parties and restrict publication. A printed guide to the amended 1875 Publishing Ordinance survives in the National Diet Library: fifteen folded leaves containing the ordinance, penalties, and forms. Its very existence shows that publication control had become something a printer or publisher could encounter as a formal body of national rules and paperwork.^[6]

This was a different legal environment from the negotiations of Tokugawa book wholesalers and Edo magistrates, but one administrative logic persisted. A government confronting thousands of pieces of print still needed to know whom to address.

Registration requirements, named editors or publishers, deposits, penalties, and prohibitions make responsibility traceable. They transform print from an anonymous object into an object attached to someone who can be summoned, fined, prosecuted, or ordered to stop distributing it.

That traceability can support more than censorship. Copyright law also needs identifiable works and claims. Commercial disputes need responsible parties. Hara's point is precisely that the legal histories of press freedom, copyright, and publishing institutions were entangled rather than developing as isolated stories.^[5]

The same machinery can therefore serve competing purposes. Making a publisher legible to law can protect an exclusive publishing interest and expose the publisher to state sanction. Requiring formal publication procedures can stabilize a market and create a route for restriction.

The question “Was printing free?” is too blunt to capture this. A publisher might operate in a vigorous commercial market, invest in new technology, issue politically significant material, and still work within laws that made certain content dangerous. Expansion of print increases the importance of regulation because more material can reach more people more quickly.

Modernization did not eliminate censorship by making information circulate. It altered the institutions through which circulation could be governed.

By the early twentieth century, prohibiting a publication could appear in one of the state's most routine printed instruments: the Kanpō, the Official Gazette.

The issue of 20 February 1902 includes Home Ministry Notice No. 11, described in its table of contents simply as a prohibition on the sale and distribution of one publication.^[7] The catalogue entry does not by itself tell us everything about the prohibited work or the enforcement that followed. Its value is more basic. A ban had a numbered administrative form, an issuing ministry, a date, and a place in the same official publication that carried other orders and notices.

That routinization matters. Censorship need not always arrive as a spectacular trial of a celebrated author. It can operate through repeated administrative decisions that are individually small enough to look bureaucratic.

The surviving records of the Home Ministry's Police Bureau show the scale of that work more clearly. The National Diet Library describes a 54-reel microfilm collection whose first major group consists of newspapers, pamphlets, and handbills banned, warned, or otherwise examined by the bureau between 1928 and 1940. The catalogue is arranged chronologically across years of material.^[8]

An archive like this changes the historical picture. The object of control was not merely “literature.” Newspapers, pamphlets, political handbills, and other short-lived forms could matter precisely because they moved quickly and addressed current events. Their low cost and immediacy made them useful to political movements—and therefore relevant to police.

The institution examining them did not need to produce a single timeless theory of unacceptable speech. It could build a file of cases. The file turned decisions into precedent, memory, and administrative knowledge.

This is another way uncertainty gets managed. Publishers and editors try to infer what will attract punishment. Officials compare a new item with earlier cases and current instructions. A changing political environment can tighten or loosen the practical boundary without rewriting every underlying rule.

The system is therefore better imagined as a process than as a blacklist. A blacklist is an output. The process includes surveillance, interpretation, warnings, negotiations, seizures, prosecution, and the expectations created by previous decisions.

The suppression of Ishikawa Tatsuzō's Ikiteiru heitai (Living Soldiers) in 1938 shows how high the stakes could become once war, mass media, and public-order law converged.

Ishikawa traveled to China as a correspondent connected with the publisher Chūō Kōronsha and wrote about Japanese soldiers after the fall of Nanjing. According to Onishi Yasumitsu's study of the case, the resulting work was suppressed and Ishikawa was arrested under the Newspaper Law on public-order grounds.^[9]

The episode differs from the Tenpō reformers clothing figures in a rewritten comic book. Here a modern author and publisher operated in a world of newspapers, magazines, correspondents, national law, and wartime military operations. The potential audience was far larger and faster to reach.

Yet one mechanism is recognizable. The publisher remained a point where risk had to be assessed before readers received the work. What could be printed depended partly on what editors believed could survive the legal and political environment.

The historical significance of the case does not require us to imagine that every censored work in the 1930s faced the same process or that censorship was suddenly invented by wartime authorities. The continuity lies in accountable intermediaries. The change lies in the concentration and reach of the institutions surrounding them.

A national police apparatus could coordinate controls beyond one city's magistrates. Mass-circulation media could amplify a text far beyond the scale of an Edo woodblock edition. War made the state more concerned not only with preventing particular statements but with maintaining a broader information environment compatible with mobilization.

At that point, censorship by prohibition was only half of the story.

Soone Park's study of the Cabinet Information Bureau distinguishes “negative” and “positive” mass-communication policy. Negative measures restricted newspapers, magazines, radio, and other media through law, censorship, guidance, and control of resources. Positive policy meant producing and disseminating information that the government wanted people to receive.^[10]

The distinction changes the question from “What was forbidden?” to “What occupied the space that remained?”

From 1936, the government published Shūhō (Weekly Report), a national-policy pamphlet intended to explain and promote state policy. Park shows that officials studied the format and readership of successful private current-affairs pamphlets and designed Shūhō to compete in the same informational market. Through advertising and established distribution routes, it eventually reached a very large circulation.^[10]

This is censorship's most expansive form: not merely blocking an unwanted message, but altering the conditions under which messages are produced and encountered.

Resource controls matter here. Paper, printing capacity, distribution, and access to news are material inputs. A publication can disappear because an official bans its sentences. It can also disappear because a publisher cannot secure what is needed to produce it, or because the institutional field rewards another kind of publication instead.

This does not erase the difference between propaganda and censorship. The difference is analytically useful. A government pamphlet is not the same act as seizing a newspaper. But the two policies can reinforce one another. Restrict private alternatives while expanding official material, and the reader's menu changes from both directions.

The wartime system therefore reveals something that had been present in weaker form much earlier. Control of print is never only control of words. It is control of relationships among authors, publishers, printers, distributors, officials, resources, and readers.

The more complex the media system became, the more points of intervention became available.

Across these periods, there was no single Japanese institution that continuously decided what could be printed. Tokugawa magistrates and scholarly examiners, Meiji legislators, Home Ministry police, courts, editors, and wartime information agencies belonged to different political orders. Their standards and powers were not interchangeable.

What recurred was a governing problem: print multiplied speech faster than rulers could inspect it person by person.

The practical answer was to make circulation pass through responsible intermediaries. Punish a publisher and other publishers learn. Require legal identification and an editor becomes reachable. Record prohibited pamphlets and a police bureau builds an institutional memory. Control paper and official information, and some publications become difficult to produce before a censor needs to open them.

This is why self-censorship belongs inside the history rather than outside it. When a publisher changes a picture, abandons a manuscript, softens a passage, or refuses a risky work because of expected official consequences, the state's decision has been partly delegated to the person bearing the commercial and legal risk.

Delegation does not make the result voluntary. Nor does it mean publishers were passive. They experimented, negotiated, disguised meanings, recycled older material, and sometimes accepted punishment. The boundary existed because people kept testing it.

The question “Who decided?” therefore has a deliberately plural answer. Officials made decisions, but they also built systems that induced other people to decide in advance.

That is what connects the 1791 punishment of a writer and publisher, the heavily moralized remake produced under the Tenpō Reforms, the national ordinances of the Meiji state, the Home Ministry's files of prohibited publications, and the wartime Shūhō. The mechanism changed from case to case, but control became more effective when it entered the production chain before a reader ever held the finished page.

A censor did not have to read every book. The state had to make enough people worry about the reader who might.