How Did Japan Turn Disagreement into a Decision?

On 5 February 1897, a committee drafting Japan's commercial law returned to a provision it had already debated. Ume Kenjirō explained that the principle had been settled by a majority; the remaining business concerned its wording. Hasegawa objected that the proposed language still failed to make the arrangement clear. Ume answered him. No one supported Hasegawa's proposal, the record states, and it lapsed. The draft provision was approved.^[1]

The record does not say that Hasegawa changed his mind. It does not need to. For this meeting, a proposal could cease to command the committee's time without its author declaring himself persuaded.

That small ending opens a much larger question. How could people act together when discussion produced competing judgments? Japanese governing institutions offered different answers: consultation to find the right course, frank argument intended to restore unity, and procedures that allowed a decision to stand while disagreement remained.

These were not simply successive stages in the discovery of reason. Each depended on who could participate, what the meeting was supposed to achieve, and whose authority made its conclusion effective. Their differences become clear when we compare early court prescriptions, late-Tokugawa arguments about deliberation, and Meiji procedures for closing a debate.

The work conventionally called the Seventeen-Article Constitution is preserved in the eighth-century Nihon shoki, which places it in 604 and attributes it to the crown prince remembered as Shōtoku. The document's attribution and transmission should not be confused with a surviving original issued in that year. What the chronicle preserves is a program of conduct for government, rather than minutes showing how its officials actually voted.^[2]

Its final article advises against deciding important matters alone. Weighty business should be discussed with others because there is a danger of getting it wrong. Small matters need not undergo the same process. Consultation is justified by the difficulty and consequences of the question, not by a requirement that every act of administration receive collective approval.

Article 10 supplies a reason to listen. Other people have their own inclinations; what seems right to one person can seem wrong to another. The text warns against treating oneself as certainly wise and one's opponent as certainly foolish. It also urges restraint and a willingness to follow the larger company even when one believes oneself right.^[2]

This recognizes both an intellectual problem and a practical one. An individual may be mistaken. An individual who remains convinced may nevertheless need to act with others. The text does not turn that second possibility into a voting rule. It specifies no electorate, threshold, or method for counting a divided assembly.

The consultation passages also sit beside Article 3's emphatic demand for obedience to imperial commands. The ruler and subordinate occupy different positions, compared to heaven and earth. Advice and command belong within the same prescribed order.^[2]

There is no need to remove one side to make the other intelligible. A ruler could require consultation because governing demanded more knowledge than one person possessed, while retaining the authority to issue the resulting command. Officials could be told to question their own judgment without acquiring an equal right to determine every policy.

Nor does the praise of harmony prove that disagreement was rare. The instruction is useful precisely because people could resist one another, become angry, and mistake personal inclination for public reason. The text seeks to discipline those difficulties. It does not make them disappear.

Its answer to disagreement is consequently a demanding form of conduct: listen more widely on important business, distrust self-certainty, and keep private resentment from obstructing collective action. The arrangement still leaves a question for later institutions to confront. Who can make it safe for a subordinate to say that the proposed course is wrong?

In 1787, the scholar Hosoi Heishū addressed a request for advice from Tokugawa Munechika, lord of Owari, a major domain in central Japan. He recommended open discussion of governmental business, large and small. Senior executives and junior officials involved in governing should gather before the lord and speak frankly about advantages and difficulties, without concealing what they thought.^[3]

The proposal did not invite the whole population into the room. Its participants were people already connected with the domain's administration. Within that company, however, the demand for candor was substantial. A junior official might possess information that a superior lacked. Merely collecting people in order of rank would not make that information available.

In another passage preserved among his writings, Heishū explained the obstacle through the etiquette of approaching a superior. A person of lower standing could hardly step forward before the person above had invited him closer. The first movement toward a more open relationship therefore had to come from above. Consultation should temporarily allow people to state their judgments without the usual difference of rank governing every exchange.^[4]

The image identifies a practical condition of speech. Permission cannot be assumed merely because a ruler says that good advice is welcome. Participants need to know that the present encounter permits a kind of frankness that might be risky elsewhere. An invitation, issued by the person able to punish discourtesy, could help establish that condition.

It also placed a burden on the lord. Heishū did not make the successful meeting depend only on subordinates becoming braver. The superior had to initiate the relationship through which their knowledge could reach him. A closed manner at the top could defeat an instruction to be open below.

The desired result was unity between higher and lower officials, enabling government to proceed. The invitation to disagree was a means of arriving at that result. Heishū did not provide a rule under which a counted majority could overrule the lord or close discussion while leaving the lord's judgment unchanged.

The comparison with the early court prescription is useful without implying direct institutional continuity. Both arrangements valued knowledge obtained from several people within a hierarchy. Heishū's account made a particular social difficulty unusually explicit: how the hierarchy itself could prevent the discussion it needed. The person presiding had work to do before anyone began presenting arguments.

A late-eighteenth-century text offered an even sharper warning against confusing quiet with agreement. Shōhei yawa, compiled in 1796 by Takano Yokei, a scholar serving the lord of Nagaoka, collected reflections and examples concerning rulers and retainers. Its discussion of thorough deliberation condemned officials who appeared peaceable in public while criticizing one another in private.^[5]

Open disagreement over official business was proper, the passage argued, provided participants did not introduce private animosity. People holding similar offices would never all think alike. Stating their views and disputing right and wrong need not leave lasting resentment.

The central image was a meal. Different officials were like different flavors; the ruler was the cook responsible for combining them. Sweetness and pungency were not themselves faults. Whether the mixture worked depended on the person preparing it.^[5]

This was more than praise for everyone getting along. The passage made differences useful. If the ingredients were identical, there would be little to combine. A ruler needed access to judgments that were not already copies of his own or of the most senior retainer's opinion.

At the same time, the cook remained in charge. The image assigned the ruler responsibility for bringing the mixture to an appropriate result. It did not describe an assembly in which every flavor had one vote and the most numerous ingredient determined the dish.

The account was particularly alert to false endings. Deference, favoritism, and reluctance to expose one's thoughts could make business drag on without a conclusion. Alternatively, a forceful person could insist until his position prevailed, leaving others unconvinced. A decision would have been announced, but the process had failed to produce the understanding on which willing implementation depended.^[5]

The warning helps explain why prolonged discussion could have a purpose beyond politeness. An official who would later carry out a policy needed to understand it and accept his part in it. Ending the meeting quickly might simply move the conflict into administration, where the resulting obstruction would be harder to see.

Yet the ideal also carried a demanding expectation: well-conducted argument would converge upon an appropriate common judgment. What should happen when sincere, informed participants still disagreed? The cooking image supplied a responsible ruler, but no independent test establishing that everyone had actually become persuaded. That difficulty appears in a later participant's account of a meeting celebrated as a success.

In late 1860, Yokoi Shōnan took part in a confrontation within the Fukui domain, also known as Echizen. Two groups associated with the reigning lord, Matsudaira Mochiaki, and his retired predecessor, Matsudaira Shungaku, had become divided. A dispute over appointments helped bring them together at Shōnan's residence for what became known as the Great Debate.^[6]

Shōnan's letter from early 1861 described the result in extraordinary terms. He reported that senior officials had come to understand the dispute, that tears accompanied the change, and that the government had at last acquired a united direction. Representatives then traveled to Edo, where lord and retainers acknowledged failures in their relationship. These are his descriptions, preserved in the letter, not independently observed reactions supplied by a later narrator.^[6]

His account made personal reconciliation part of political success. The participants had not merely selected a proposal. They had, in his understanding, corrected the suspicions and estrangement that had prevented them from governing together. Their changed relationship would allow subsequent business to proceed.

The journey to Edo extended the work beyond the original room. A meeting among participants in Fukui still had to reconnect decisions with authority located elsewhere. Deliberation was embedded in a governing household whose members were not all present at the same table.

Shōnan's celebration also establishes the limit of the evidence. A participant pleased with an outcome had reasons to describe it as a recovery of shared purpose. His letter preserves what he counted as success; it does not allow us to inspect every participant's remaining reservations. Tears in his account are evidence of his explanation, not a measure of unanimity.

The case nevertheless adds something that a manual of conduct cannot. It shows a political actor interpreting a particular debate through the repair of relationships. For him, convergence was an achievement requiring candor, acknowledgment, and renewed confidence among people who still had to work together. Counting supporters would not, by itself, demonstrate that achievement.

The later introduction of more explicit meeting procedures addressed another possibility: a body might need to continue its work even when that level of reconciliation had not been reached.

During the Meiji period, parliamentary procedure became an explicit subject of instruction. Kaigiben, a manual associated with Fukuzawa Yukichi, Obata Tokujirō, and Koizumi Nobukichi, survives together with rules connected to the Mita public-speaking society. Keio's catalogue dates the issued form to around 1874, while distinguishing it from earlier preparatory work.^[7] Conducting a meeting could be presented as something to learn, with written arrangements that extended beyond the goodwill of the people present.

The Constitution promulgated in 1889 gave the two chambers of the Imperial Diet precise rules for reaching decisions. Article 46 required the presence of at least one-third of a chamber's total membership before it could debate or vote. Article 47 required an absolute majority for decisions and assigned the president a casting vote in the event of a tie.^[8]

These clauses solved different problems. The quorum established when the assembled members were numerous enough to act as the chamber. The majority requirement established how a divided company could produce a decision. The tie provision supplied an authorized ending even where counting first produced equality.

None required a declaration that every opponent had been convinced. The losing position could remain a position held by actual members. Once the prescribed process produced a result, the chamber could act upon it.

This changed the relationship between persuasion and closure. Persuasion still mattered because it could change votes. But a successful argument no longer had to transform every participant's judgment before the institution could move forward. Agreement about how to decide could perform work that agreement about the proposal itself had not accomplished.

A vote also narrowed the immediate question. Participants might hold complicated or partly overlapping views, yet the procedure required a decision about a specified proposal. Formulating that proposal mattered: which alternative came before the chamber, which amendment altered it, and when the question was ready to be put. Numbers acquired authority through these prior arrangements rather than floating above them.

The constitution did not apply one threshold to every decision. Article 73 required imperial initiative for constitutional amendment, attendance by at least two-thirds of each chamber's membership, and support from at least two-thirds of those present. Changing the fundamental rules demanded a different combination of permission and agreement.^[9]

That exception exposes the work hidden in the word “majority.” Before counting, an institution had to identify the people entitled to participate, the minimum company that could act, and the proportion sufficient for the particular question. These were choices about authority, not conclusions supplied by arithmetic alone.

Return to the commercial-law committee in 1897. Its members were preparing statutory language, not exercising all the powers of the Imperial Diet. The distinction matters: approval within the committee was a step in drafting, not by itself the enactment of a national law.

The opening exchange concerned representation of a company in proceedings against directors. Ume described the principle as already decided by a majority and presented revised language. Hasegawa thought the relationship between representation and bringing the action remained unclear. His proposed change found no supporter, so it fell away and the text before the committee was approved.^[1]

The record preserves his objection alongside the result. This is a different kind of institutional memory from an account of everyone discovering a common truth. It allows a later reader to see that a question was raised and answered without needing to say that the answer satisfied its questioner.

Hasegawa did not vanish from the committee's work. Later in the same session, he joined in questioning another provision. Ume also explained that an earlier Hasegawa proposal had helped determine the treatment of goods received with a rejected contractual offer.^[10] One unsuccessful intervention did not define his position on every other issue.

That sequence is important to the practical value of procedure. A participant could lose on a point without having to be expelled from the working company or reconciled through a ceremony of personal repentance. He could continue to ask questions, propose alterations, and have an earlier contribution recognized in another clause.

The surviving record gives no headcount for the opening decision. Its narrower evidence is enough: a previous majority decision constrained the drafting discussion; an unsupported proposal lapsed; the committee proceeded; and the same participant remained active. No invented tally is needed to explain how the disagreement ended for institutional purposes.

The later constitutional arrangements also placed limits around what a successful vote could do. The Imperial Diet contained two chambers with different bases of membership. One was an elected House of Representatives constituted under electoral law; the other included imperial-family members, peers, and appointed members. Legislation required the Diet's consent, while imperial sanction and promulgation remained necessary. A majority within one room was not sovereignty over the whole system.^[11]

The comparison reveals two different demands upon a meeting. One asks participants to arrive at a shared judgment strong enough to sustain common action. The other establishes when their divided judgments are sufficient to authorize a result. A ruler trying to repair relations among retainers could place special weight on the first. A chamber with a continuing agenda needed an answer to the second. Neither demand made the people outside the authorized company participants in the decision.

These answers could overlap. A formally constituted committee still needed people to explain themselves, listen, and work together afterward. A lord's consultation could still end in a command. The decisive historical change was not from Japanese harmony to foreign argument, but in what counted as sufficient agreement for a particular body to act.

Hasegawa's objection survived because closure did not require its erasure. The committee could approve its text and continue with a member who had questioned it. The record gives a concrete answer to a question the ancient injunction to consult had left open: how could the institution proceed when listening had not made every judgment the same?