Japan's Nesting Dolls of Power

In 1219, Minamoto no Sanetomo, the third shogun of the Kamakura government, was assassinated.

For the powerful Hōjō family, his death created a succession crisis. They did not resolve it by making one of their own the next shogun. Instead, they brought a small boy from the aristocratic Kujō family in Kyoto to Kamakura. Known as Mitora, he would later take the name Yoritsune. His formal appointment as shogun did not come until 1226.^[1]

If the Hōjō could help decide who would occupy the position, why did they not take it themselves?

Japanese history repeatedly invites questions like this. An emperor occupied the throne, while his grandfather exercised enormous influence. An emperor retired, but continued to govern. A regent held power beneath a shogun, then relinquished the regency without necessarily relinquishing power.

At first glance, the arrangement resembles a set of nesting dolls. Open the figure that appears to be in charge, and another seems to be waiting inside.

The image is memorable. It is also incomplete.

The question is not why the Japanese supposedly preferred to pull strings from the shadows. It is why a person who could already decide so much might still need someone else's title—and what allowed that person to remain powerful after giving up a title of his own.

In 1008, Fujiwara no Michinaga's daughter Shōshi gave birth to a prince. The child's father was Emperor Ichijō. Michinaga was his maternal grandfather. Michinaga's diary records the events surrounding the birth; the boy would eventually become Emperor Go-Ichijō.^[2]

A marriage and a birth were not merely private family events in this political world.

Becoming the emperor's maternal family offered a route into the life of a future ruler. It could bring access to his upbringing and connect relatives, attendants, and allies to the imperial household. For an ambitious court family, becoming indispensable to the next emperor could be more useful than trying to abolish the throne.

But kinship was not a certificate of unlimited authority. A prince still had to succeed. Other people at court had to support him. Rival families might have competing candidates. Being a grandfather mattered because of the relationships it made possible, not because blood ties could conduct the business of government on their own.

Michinaga is particularly revealing here. Although he is often treated as the supreme example of Fujiwara government through imperial regents, he never became kampaku, regent to an adult emperor, and held the regency for a child emperor only briefly. More enduring sources of influence included his position as Minister of the Left and his right, known as nairan, to inspect documents before they reached the emperor.^[3]

Historian Ōtsu Tōru emphasizes the work Michinaga did within the court's administrative and deliberative machinery: discussing provincial affairs, scrutinizing governors, managing revenue, and dealing with palace construction. This was not simply a man issuing secret instructions to a helpless monarch. It was a senior statesman operating through offices, documents, councils, and cooperation with other aristocrats.^[3]

Calling the emperor a puppet obscures more than it explains.

The political world Michinaga inhabited was organized around the imperial house and its hierarchy of offices. Destroying that center would also have diminished the value of being the emperor's grandfather or his leading minister. The throne's continued recognition helped make those positions worth holding.

Some power comes not from replacing the center, but from becoming indispensable to it.

From the late eleventh century, a different arrangement became prominent. Retired emperors such as Shirakawa, Toba, and Go-Shirakawa conducted political business through their own establishments. This form of government is known as insei, usually translated as rule by retired emperors or cloistered government. A retired sovereign's residence could be much more than a retreat: it had administrative offices, supportive courtiers, and provincial estates to manage.^[4]

The person on the throne and the person exercising sustained influence over appointments and decisions did not have to be the same.

It is tempting to explain this as a clever escape. The emperor, burdened by the restrictions of his office, retired and became free to do as he pleased.

Yet the decisive fact was not retirement itself.

A former emperor who lost his income, trusted officials, and political supporters would not necessarily govern merely because he had once occupied the throne. Effective retired rule depended on the means of acting surviving the departure from office.

This was already a different mechanism from Michinaga's. The maternal grandfather gained influence through proximity to a reigning emperor. A powerful retired sovereign maintained an establishment of his own outside the reigning emperor's immediate household.

Nor was the retired emperor a subordinate to whom power had simply trickled down. He was an imperial elder, a former sovereign, and potentially the head of a competing center of decision-making.

Power was not moving steadily down an organizational chart. It was being rearranged among office, family standing, and resources.

When Minamoto no Yoritomo established his warrior government at Kamakura in the late twelfth century, he did not begin on an empty stage.

The familiar shorthand is that warriors displaced courtiers, shoguns sidelined emperors, and Japan entered the age of the military governments. But the imperial court, aristocratic households, and religious institutions did not all cease to function when Kamakura rose. Historian Jeffrey P. Mass's work places the new warrior government within an older political world that remained active, rather than treating its establishment as the instant disappearance of Kyoto's authority.^[5]

To see what this meant, leave both capitals for a moment and consider a piece of land.

An estate could contain several kinds of rights. The person entitled to a share of its income might not be the person managing it locally. The person charged with policing or enforcing a decision might be someone else again. When Kamakura appointed shugo, provincial military officials, and jitō, estate stewards, it inserted warriors into existing arrangements. It did not automatically erase the claims of aristocratic and religious proprietors. The government could also be called upon to restrain its own stewards' encroachments.^[6]

Who ruled this place? The answer depended partly on the question being asked.

Who should deal with a theft? Who was entitled to receive the dues? Which document established a claim when a dispute arose? A temple in distant Kyoto might rarely be represented in person, yet possess rights that people on the estate could not simply ignore.

Keeping older institutions, then, was not only a matter of courtesy or reverence for tradition. Warriors encountered a world in which people were already collecting revenue, preserving documents, maintaining connections, and asserting claims.

Force could change the balance. A victory could not, by itself, settle the business of every day that followed.

The warrior government acquired new powers while using, modifying, and sometimes eroding older relationships. Behind the continued presence of both emperor and shogun stood far more than two distinguished men. There were institutions and interests that had to be accommodated, contested, or overcome.

That brings us back to the child brought from Kyoto in 1219.

The Kamakura government had not been built by the Hōjō family alone, with a private army of its own followers. It emerged around Yoritomo and a coalition of warriors. His directly bound vassals, known as gokenin, had their own relationship with the shogunal lord. Service to Kamakura did not automatically mean that they had always regarded the Hōjō as their personal masters.^[5][7]

Hōjō Masako's position as Yoritomo's widow, together with her family's administrative and personal connections, helped the Hōjō navigate the succession crises. But preserving the government was not the same as announcing that it now belonged to the Hōjō in an entirely new sense.

A recognized shogun offered continuity. Around his household, the Hōjō could exercise influence over access, appointments, and the conduct of business without requiring every vassal to accept that a former fellow retainer had become his new lord.

This helps explain why an aristocratic child was not a pointless ornament. The child supplied a prospective successor around whom an established relationship of service could continue, while senior figures conducted government. The arrangement separated the recognized lord from those exercising much of the authority around him.

Nor can the Hōjō's decision be explained by a supposed rule that only someone named Minamoto could become shogun. Later Kamakura shoguns included members of the Kujō family and imperial princes. Birth and political connections constrained the choice, but not through that simple surname test.^[1]

Preserving the position was not risk-free.

A child grew up. A shogun could develop relationships of his own, and dissatisfied men could gather around him. The later replacement and removal of shoguns show that the office was not an inert object the Hōjō could put on a shelf and forget.^[1]

The title was useful precisely because people continued to recognize it. That was also what made it potentially dangerous to those governing in its name.

In 1221, the retired emperor Go-Toba ordered the punishment of Hōjō Yoshitoki.

The choice of target mattered. The order named the regent, rather than simply condemning every warrior associated with Kamakura. It offered a possible distinction: a warrior might oppose the Hōjō while claiming loyalty to a higher authority. That was a dangerous opening in the crisis that became the Jōkyū War.^[7]

The Hōjō could not resolve it by declaring that they held real power and Kyoto possessed only empty titles.

Later accounts, notably the chronicle Azuma Kagami, give a central place to Masako's appeal to the vassals. Its substance was to recall Yoritomo's protection and the benefits his followers had received. The celebrated words are not a surviving verbatim record of the occasion. But the political problem the account presents is clear: defending Yoshitoki had to become a reason for the warriors to defend their own position within the Kamakura settlement.^[8]

This was where acknowledged authority and practical interests met.

Go-Toba's command could give opponents a cause. It could not automatically provide him with an army. The Hōjō possessed means of organization, but still had to secure support. Neither side could make everyone else's choices disappear.

Kamakura won. Go-Toba was sent into exile on the Oki Islands. Court and warrior government continued to coexist, but the balance between them had shifted decisively in Kamakura's favor.^[7][8]

The war exposes the weakness of two opposite assumptions.

Imperial authority was not meaningless; otherwise the Hōjō would not have needed to answer the challenge with such care. It was not omnipotent either; otherwise the retired emperor's command would have settled the outcome.

Being willing to listen to an order, being willing to fight for it, and being able to win are three different things.

By now, it may seem that we have finally identified the person who really governed: not the shogun, but the shikken, the regent who stood beneath him and came to lead the Kamakura government.

The Hōjō's own history unsettles that answer too.

Hōjō Tokiyori resigned the regency and took Buddhist vows, yet remained a dominant political figure. Historian Takahashi Shin'ichirō's biography treats his government after religious retirement as a distinct part of his career, rather than assuming that his departure from office ended it.^[9]

Two positions need to be distinguished. The shikken held an office within the warrior government. The tokusō was the head of the main Hōjō lineage. One person could occupy both positions, but they were not identical. Giving up an official post did not necessarily mean handing over family property, retainers, or leadership of the house. Hōjō Sadatoki likewise continued to exercise authority as family head after resigning the regency.^[10]

A list of regents, then, is not enough to tell us who made the most important decisions in every period.

And there was another complication. If the family head's power depended on a household and its resources, the people who managed those resources, dealt with retainers, and conducted business could become powerful themselves. In the later thirteenth century, Taira no Yoritsuna, a leading Hōjō household official, acquired exceptional influence. In 1293, Sadatoki had him eliminated and reasserted his own authority.^[10]

Now the nesting dolls seem almost complete: emperor, shogun, regent, family head, household official.

But this was not a permanent chain of command, with power inevitably passing to the next smaller figure. Some of these positions were offices; others rested on family standing. Sometimes they were combined in one person. Sometimes they were separated. An official could gain power and then be destroyed by the master whose household he served.

Delegating the conduct of business can enable an agent to accumulate influence. That is a mechanism worth examining. An endless, irreversible descent of power is not what these events establish.

Several centuries later, Tokugawa Ieyasu offered another variation.

In 1605, he passed the title of shogun to his son Hidetada and became known as the ōgosho, the retired ruler. He later based himself at Sunpu, in present-day Shizuoka, and continued to take part in government and foreign affairs. Officials, retainers, monks, scholars, and merchants gathered around him. Sunpu was not merely a quiet place where an old man had gone to rest.^[11]

This was not a subordinate gradually squeezing his master out.

Ieyasu transferred the formal position to the next generation while retaining the people, standing, and institutions through which he could continue to act. Hidetada held the shogunate and its government in Edo. Ieyasu remained influential through the authority of a founder and the working establishment around him.

From the perspective of succession, the arrangement had an evident advantage. A successor could begin exercising responsibility while his predecessor was still alive to support him. The transfer did not have to begin with the founder's death.

But this is not a formula for retaining power while enjoying the comforts of retirement.

The important question is what survives retirement. Without support from influential people or the ability to mobilize resources, an honored title may secure respectful treatment and little more. Authority outside office does not spring into existence because somebody has made a clever gesture.

Ieyasu's example shows that formal succession and a predecessor's continuing influence can overlap. It does not show that everyone who gives up a position can do the same.

Together, these cases suggest an explanation that is useful because it has limits.

When a title remains a widely recognized basis for rank, succession, and rights, while another network of people and resources can conduct much of the practical business, title and decision-making power may separate.

Those handling affairs can work through an existing source of recognition rather than build an entirely new order. Other people do not have to abandon familiar allegiances, offices, and rules all at once. A new arrangement of power can develop within an older form.

This does not mean that the actors were all performing the same calculation. Marriage, inheritance, administrative ability, military success, and personal relationships mattered differently in each case. Michinaga, Shirakawa, the Hōjō, and Ieyasu were not carrying out a common plan across the centuries.

They did, however, repeatedly encounter a related difficulty: controlling what happens now is not the same as being able to remake the order that other people recognize.

Keeping the old form also had costs.

A preserved title could become a rallying point for opponents. People drawing authority from different sources could make incompatible demands. Arrangements that seemed workable in ordinary times might become subjects of armed conflict when succession or war forced a choice. The Jōkyū War revealed that danger. The struggles at the end of Tokugawa rule over who could speak for the court and who should lead a new government arose under very different conditions, but again showed that recognized authority could be contested rather than simply possessed.^[7][12]

The lesson of Japan's nesting dolls is not that titles were fake. It is not that the cleverest person was always hidden behind the throne.

A title could make cooperation possible and give a challenge its justification. Handling affairs could build influence, but did not guarantee that it would last. Leaving office could mean retaining power—or genuinely losing it.

To understand who governs, look beyond the person in the highest seat. Ask whose hands a decision passes through before it becomes an order, who must act to carry it out, and who can prevent it from happening.

Open the nesting dolls, and there may be no final, secret master inside. More often, there are people who need one another—and struggle with one another over what that need allows them to do.