Why Did the Shogun Keep the Lords on the Road?

A nineteenth-century handscroll takes more than fourteen meters to show one lord's procession.

About 650 figures appear in it. The family crests suggest a lord of the Honda house, though the identification is not certain. The procession belongs to sankin kōtai, usually translated as “alternate attendance”: the repeated movement of territorial lords and their entourages between their domains and the shogun's city of Edo.^[1]

Unrolled as a picture, the company becomes an orderly spectacle. As a journey, it required food, lodgings, transport, money, and people who knew what had to happen next. At either end were households that had to continue functioning when some of their members were elsewhere.

The familiar explanation is that the shogun forced the lords to travel so that the expense would leave them too poor to rebel. It is memorable, economical, and incomplete.

To understand alternate attendance, we need to ask a less tidy question. How did making powerful people repeatedly leave home help to govern them—and what happened to everyone drawn into those departures?

Tokugawa Japan was not administered entirely by officials sent from a single capital. The great territorial lords, or daimyō, retained domains, revenues, and bodies of retainers, while acknowledging the shogun's supremacy. Their relationship with the central government therefore had to be maintained across distance and through the continued existence of their own households.^[2]

A lord's attendance in Edo made that relationship concrete. He did not merely send assurances from his own castle. He came into the shogun's presence and remained available for service.

The practice developed before it received its best-known formal statement. In 1635, the third Tokugawa shogun, Iemitsu, incorporated alternating attendance into the regulations for the military houses. The familiar pattern was a year in Edo followed by a year in the domain, although arrangements varied and changed over time.^[3]

The repetition matters as much as the journey.

One visit could celebrate a settlement that both parties hoped to interpret differently afterward. A recurring obligation kept renewing the relationship. Leaving home, appearing at the center, and waiting there made submission something observable rather than a claim made at a safe distance.

The lord nevertheless remained a lord. His own retinue accompanied him; his domain had to govern in his absence. Alternate attendance did not eliminate the territorial house. It required that house to organize part of its life around service elsewhere.

This was a powerful compromise. The shogunate did not need to replace every local governing relationship in order to impose a recurring national obligation. The domains, in turn, could retain substantial institutions without being free to decide that the shogun no longer required their presence.

The arrangement was never a guarantee of loyalty. It made disregard more difficult to conceal and more costly to explain. That is already a significant kind of control, even before anyone counts the travel bill.

A procession had more than one audience.

To the shogunate, it demonstrated attendance and service. To people along the route, it displayed the standing of a particular lord. A daimyō could acknowledge someone above him while insisting, through the size and order of his company, that many people stood below him.

That helps explain why expense cannot be attributed to the shogun's intentions alone. The domains had reputations to maintain. A journey that displayed obedience could also become a competitive display of their own resources.

The authorities repeatedly tried to limit excessive expenditure. Regulations also addressed the size of retinues.^[3][1] Those efforts do not prove that the system was financially harmless. They do make it difficult to explain every costly feature as a deliberate attempt by the center to spend the domains into ruin.

The British Museum's account of procession culture describes a revealing economy: some domains traveled with reduced companies, then hired townspeople to enlarge their ranks as they approached Edo. The appearance at the destination could be grander than the traveling establishment that had borne the cost of the whole route.^[4]

A hired participant did not make the procession meaningless. He exposed one of its practical requirements. The house had to appear in a recognizable form before an audience capable of judging it. A cheaper arrangement might preserve that appearance without maintaining the same numbers on every stretch of road.

Here, governance and display were entangled. The shogun imposed attendance. Domain leaders made choices about how to fulfill it. Those choices could add costs, shift them, or disguise economies.

The result was neither a perfectly controlled machine nor an empty parade. It was a political obligation carried out by people who also cared about status, embarrassment, and the impression they made on others. Their ingenuity could help sustain the system while frustrating attempts to keep it economical.

The name “alternate attendance” directs attention to those who moved. Some of the most important constraints concerned those who stayed.

Lords' principal wives and heirs normally lived in Edo. Their residence helped bind the territorial houses to the shogunate: a lord returning to his domain left members of his immediate family within the capital's political order. These households provided guarantees of a kind that a promise of loyalty could not supply by itself.^[5]

The arrangement is often described as a hostage system. That identifies a real coercive function, but it should not make us imagine every woman living in a prison cell. A high-ranking household could have substantial resources and an elaborate domestic life while its members remained subject to restrictions on where they could live and travel.

Privilege and freedom of movement were not the same possession.

Hakone makes the restriction visible. Its checkpoint, established at the site in 1619, closely inspected women leaving the direction of Edo. The familiar formula for checkpoints refers to incoming weapons and outgoing women, but Hakone's own historical explanation specifies that this checkpoint did not inspect incoming guns. Its particular concern was the outward movement of women, including members of lordly households.^[6]

The difference is more than a correction to a tourist slogan. A system governing a lord's conduct could reach his family through rules applied on a road far from his residence. Domestic relationships became part of the geography of political control.

Other separations worked differently. Many of the Kaga domain's retainers served in Edo without their families, while the lord's principal household remained there.^[7] “Going home” could therefore mean different things within the same governing house.

There is no single family experience concealed behind the procession. For one person, the journey meant reunion; for another, another period apart. Women who stayed, children who grew up in the capital, and families waiting in the provinces did not all have the same choices. None disappears from the history simply because the scroll follows the men on the road.

The road itself was only part of the journey's infrastructure.

At Hiratsuka, a post town on the Tōkaidō road, the tonyaba coordinated transport. Its staff included clerks recording the movement of porters and horses, and supervisors assigning them to tasks. Getting a party through the town required an office as well as a road.^[8]

This is the part of a grand procession that an image can easily hide. Order within the company depended on arrangements outside it. A neatly placed load in the picture had to be carried by someone; a fresh horse had to be available at a particular place and time.

When a post town needed additional porters and horses, neighboring villages could be required to supply them through a system known as sukegō. That raised an obvious question: who decided when the extra help was necessary?

In 1821, forty-three villages challenged Hiratsuka's transport officials, alleging that they summoned village labor before exhausting the town's own resources. A settlement required the town's resources to be used first and village representatives to check the requests at the office.^[8]

The villagers were not merely scenery beside the road. They contested how an obligation was distributed, and secured a part in overseeing it.

For these people, a lord's arrival could be less an exciting spectacle than a demand competing with other work. Labor assigned to a road could not simultaneously be used elsewhere. Even when a transport service generated income for some participants, that did not establish that everyone bearing the obligation benefited equally.

The connection between center and province was therefore built partly out of the work of people whom the political ceremony did not honor.

Calling this “the development of the road network” is not wrong, but it leaves something important unfinished. Roads, offices, and transport arrangements were maintained through decisions about who had to provide what. The convenience experienced by the traveler could be an obligation experienced by someone along the way.

A history of stability needs both perspectives. The procession that made the lord's obedience visible also made demands on communities whose consent to the larger system was not what the ceremony was designed to display.

Attendance was expensive. Repeated travel was only part of the cost; a domain also needed an establishment in Edo while continuing to sustain its affairs at home.^[2]

But expenditure is not the same thing as money vanishing.

From a domain's accounts, a payment for lodging or provisions was an outflow. From the supplier's position, it could be income. A system that strained a lord's finances could simultaneously sustain businesses around his residence or along his route.

This does not turn the arrangement into an economic success story. A merchant receiving payment and a village meeting an imposed labor demand might have very different experiences. Nor can we calculate the overall effect by placing all expenditure on one side of the ledger and calling it either waste or growth. We would need to know who paid, who received, what obligations accompanied the exchange, and what those resources might otherwise have supported.

The authorities themselves altered the balance. In 1722, Tokugawa Yoshimune reduced the standard period in Edo to six months, followed by eighteen months in the domain.^[3] The concession accompanied a requirement that domains contribute rice to the shogunate; that arrangement was abolished in 1730.^[10]

Time in the capital and resources sent to it could be weighed against one another. The change did not simply remove an obligation. It altered its terms. Attendance was adjustable policy, not an unchanging formula handed down intact through the entire period.

There were also criticisms of the life it encouraged. The Confucian thinker Ogyū Sorai compared lords' stays in Edo to living at an inn, and their families' longer residence to an indefinite version of the same condition. His complaint formed part of a broader criticism of removing samurai households from the land.^[9]

His criticism was not a neutral household survey. It shows that a contemporary observer could see dependence where another person might see urban opportunity.

The point is not to choose one view and dismiss the other. Alternate attendance created situations in which political duty, household necessity, commercial exchange, and the desire to live respectably became difficult to separate. The expense was real. The claim that every expense served one coherent intention is much harder to sustain.

Arriving in Edo did not complete a retainer's education in how to serve there.

A surviving set of daimyō information cards, dated 1828–33, turns that education into a game. The 268 upper cards recorded such details as a lord's residences, income, and seating position in Edo Castle; the matching lower cards supplied the names. Guards could use them to learn the people whose arrival they might have to recognize.^[4]

This is an unusually compact image of the system. Men from different domains became entries in a body of knowledge organized around the capital. Their differences mattered, but those differences had to be made legible to people serving within a shared order.

A crest might identify the house. A name alone might not tell a guard where its lord belonged. Practical competence required knowing relationships, precedence, and the consequences of getting them wrong.

Much of this was useful precisely because the lords did not merge into one household. Edo brought territorial houses into recurring proximity while preserving their distinct identities. Learning the capital meant learning how to work among those distinctions.

The city offered more than official instruction. In his study Tour of Duty, historian Constantine Nomikos Vaporis treats alternate attendance as lived experience as well as political regulation. He emphasizes the movement of people, practices, and cultural material between Edo and the domains—not simply the capital transmitting a finished culture to passive provinces.^[2]

That distinction changes the kind of influence we should look for. A compulsory stay might create an opportunity to find a teacher, exchange an object, or encounter a practice brought by someone from a different domain. Such possibilities still depended on money, rank, permission, and personal interest. Proximity did not make everyone equally able to participate.

Nor was every encounter enriching. Some people could be tired of the city, short of funds, or anxious to return. A system's capacity to connect people is not evidence that they all enjoyed being connected.

The cards preserve a modest part of that experience: the work of learning a world larger than one's own house, because the duties of the house now required it.

One encounter in 1855 shows how far the consequences could reach beyond ceremonial attendance.

Maeda Nariyasu, lord of Kaga, ordered twelve retainers to inspect a Satsuma domain warship off Takanawa in Edo Bay. Among them was Ogawa Sennosuke, whose diary is discussed in the University of Tokyo museum's account of life at the Kaga residence. Kaga was considering a warship of its own. The visitors examined the Shōheimaru, noting its equipment and arrangements and comparing what they saw with their domain's plans.^[7]

This was not the shogun instructing a provincial lord in a technique already mastered by the central government. Men from one domain were learning from a vessel built by another, in waters near the capital.

The setting mattered. Kaga maintained people and an administrative base in Edo. Those people could be sent out to investigate something that had become available nearby. A residence associated with submission could also operate as a place from which to gather useful knowledge.

The episode does not prove that alternate attendance caused Japanese shipbuilding, or that contact automatically produced successful imitation. The diary records an inspection. What the visitors understood, what they could reproduce, and how their information affected subsequent decisions are separate questions.

It nevertheless reveals a practical possibility created by the arrangement. A common center made some encounters easier between houses that remained politically and geographically distinct.

The knowledge did not have to travel along the same line as authority. A command might move from shogun to lord. Information could move from Satsuma's ship to Kaga's observers, then into discussions within Kaga itself. The people who occupied the system could use it for purposes beyond the act of demonstrating obedience.

This is not an argument that restrictions secretly benefit everyone. The opportunities came with costs that the participants did not necessarily choose. It is an argument that an institution can create connections whose uses exceed the reasons for which those connections were required.

The same retainer could fulfill an obligation and learn something his lord valued. Neither part of the experience cancels the other.

Alternate attendance helped the shogunate turn a difficult relationship into a recurring practice.

Powerful houses retained their domains, but could not simply remain within them. Their lords appeared at the center, their households maintained an Edo presence, and their retainers learned to work within a political world larger than the domain. Distance did not disappear. It was repeatedly crossed under an obligation the shogunate could enforce.

That is the strongest part of the system's governing logic. It should not be confused with a claim that the journeys alone explain the long survival of Tokugawa rule. Landholding arrangements, military power, administrative institutions, and many other relationships also mattered. Neither can the pattern be reduced to a clever trick that impoverished every rival into permanent obedience.

The examples show different consequences operating at the same time. A procession could affirm subordination and proclaim its lord's dignity. A capital residence could keep a family within reach and give its members a place in urban society. A route could support a business and burden a village. An official stay could separate a retainer from his household while bringing him close to unfamiliar knowledge.

These were not accidental details outside the “real” political story. They were among the ways the institution worked and the costs through which it endured.

They also place a limit on what a designer of such an arrangement can control. The government could require an appearance in Edo. It could regulate parts of the journey and the stay. It could not fully determine the purchases, rivalries, friendships, frustrations, or exchanges of knowledge that would arise from keeping so many households in motion.

The handscroll offers an orderly image because the figures all belong to one procession. Life around it did not share that single direction. At the next station, someone was arranging labor. Elsewhere, a family was waiting. In the capital, a guard was learning names, and a retainer might be preparing to inspect another domain's ship.

The shogun's purpose helps explain why the lords had to travel. To understand what their journeys made, we have to follow the lives pulled into their wake.