What Did Loyalty Mean to a Samurai?

In an illustrated account of the Mongol invasions, a warrior passes a defensive wall and calls to another man to witness his courage in the battle ahead.

The warrior is Takezaki Suenaga, from Higo, in what is now Kumamoto. The man on the wall is Kikuchi Takefusa. The picture gives their names, and Suenaga's appears in red. He wants more than an opportunity to fight. He wants his part in the fighting to be recognized.^[1]

This may seem an awkward beginning for a story about samurai loyalty. Aren't loyal warriors supposed to serve without thinking of themselves?

That expectation is precisely what needs examining. Loyalty could require a person to risk his life. It did not follow that his name, his household, or what he received in return had ceased to matter. To understand the commitment, we have to look at the relationship that made it meaningful—and at the other relationships it could put in danger.

The invasions of 1274 and 1281 brought forces of the Mongol empire and its allies against northern Kyushu. The scrolls associated with Suenaga are thought to have been commissioned to record his service. They present a participant's claim to recognition, not an impartial survey of everyone who fought.^[1]

That makes the appeal to a witness especially revealing.

A deed on a battlefield has an immediate purpose: defending a position, reaching an enemy, keeping someone alive. It can also acquire a second life as something reported afterward. Other people have to learn what happened and decide what it deserves. The danger a warrior faced does not automatically become a place in the record.

Suenaga's appeal addresses that distance. He wants someone else to be able to say that he was there and acted bravely. The picture then carries the claim beyond the people who could have seen him in person.

There is no need to choose between genuine courage and an interest in reputation. Seeking recognition does not make an arrow less dangerous. Nor does fighting bravely prove that the fighter was indifferent to what followed.

The relationship between service and reward depended on this problem of visibility. A superior could promise to acknowledge good service, but somebody still had to establish whose service counted. A witness, a report, and eventually an image could help turn an action into a claim.

The scroll therefore shows a social world around the warrior. His achievement needs other people who understand it and institutions through which it can have consequences. The solitary hero is already less solitary than he first appears.

Suenaga's service did bring a material return. In 1276, he entered Kaitō in Higo, where he had been granted rights as a reward for his service in the first invasion. Documents from 1293 and 1314 later record his donations and regulations connected with a local shrine and the temple Tōfuku-ji. Revenues were directed toward their upkeep.^[2]

A battlefield achievement had entered the continuing life of a place. What came back was not simply a purse of money to spend and forget. It involved rights, income, and responsibilities that could shape a household's future.

Documents from the Kamakura government show how carefully such positions could be defined. In 1187, an order confirmed a warrior's appointment as a local estate steward, or jitō, while requiring him to remit the prescribed dues according to the proprietor's instructions. His appointment did not wipe out everyone else's claims.^[3]

In 1215, another order confirmed a widow in the stewardship specified by her husband's will. Existing taxes and service obligations were to continue.^[3]

That widow belongs in a history of warrior loyalty. Service was not always a bond between two men that vanished when either died. It could be attached to an enduring household, supported through property and carried across generations. A woman maintaining its rights could participate in the arrangement without resembling the mounted fighter in a battle painting.

These records also make the other side of the relationship visible. A government that expected service had to recognize claims, settle disputes, and make its promises worth something. Its authority depended partly on being a place to which people could bring the question: what are we entitled to keep?

The arrangement could support trust without making the parties equal. Nor was it costless to people outside the warrior household. Revenue came from places where others lived and worked; a reward to one person could become an obligation for another.

Loyalty in this setting could include gratitude, affection, reputation, and calculation. Reducing it to wages misses the depth of the connection. Removing its material conditions makes it difficult to explain how the connection endured.

In 1600, the Sanada family stood on opposite sides of the conflict that culminated in the Battle of Sekigahara.

Sanada Masayuki and his younger son Nobushige opposed Tokugawa Ieyasu. Masayuki's elder son, Nobuyuki, supported him. The family relationship survived in their names; it did not produce a single political choice.^[4]

The consequences separated them further. Masayuki and Nobushige were sent into exile. Nobuyuki retained a place in the Tokugawa order and was awarded his father's territory around Ueda, in present-day Nagano. In 1622, he moved to Matsushiro, where his descendants continued as lords. Nobushige had meanwhile returned to war on the Toyotomi side at Osaka and died in 1615.^[4][5]

It is tempting to turn this outcome into a clever family insurance policy: choose both sides, and someone is sure to survive. But the survival of a branch does not, by itself, establish an agreed plan behind every decision. Knowing the result gives us an advantage the people making their choices did not possess.

More important, a strategy that preserves a family name may still divide the people who bear it. Exile and a brother's death do not become painless because another branch keeps its lands.

Which choice should count as loyal? The question changes with its object. Loyalty to a father, to the men one serves beside, to a political alliance, or to the future of one's own children may require different acts. A family can supply reasons for commitment without supplying one answer to all of them.

Calling Nobuyuki merely a survivor and Nobushige merely a hero would settle the matter through labels. Each had to act within relationships that a later portrait cannot fully display. An action that looked like fidelity from one position could look like abandonment from another.

This is not proof that loyalty meant nothing. It is a reason loyalty could become so demanding. People did not always have the freedom to honor one obligation without breaking another.

Warrior writings did not direct every demand downward.

In admonitions dated 1412, Imagawa Ryōshun rebuked a younger relative for arbitrary rewards and punishments, favoring personal interests, extracting wealth from the people, and refusing unwelcome advice. The lord, he insisted, should examine his own conduct rather than judge only his retainers.^[6]

The document is an accusation and a prescription. It shows what its author thought a ruler should correct, not how every warrior behaved. Yet the direction of its criticism matters. The person entitled to command could himself be failing the relationship.

Arbitrary reward undermined more than a disappointed individual's feelings. When favor rather than service determined recognition, a retainer could no longer know what conduct would be valued. Punishment without reliable judgment made obedience a less secure foundation for a life. A lord who refused advice could deprive himself of the knowledge he needed to rule.

Those are failures within an unequal order, not objections to hierarchy itself. Ryōshun was not asking his relative to abolish lordship. He was insisting that its privileges came with standards.

That gives advice an ambiguous position. A retainer who tells a lord something unpleasant may appear less compliant than the person who agrees with him. Yet the disagreement may be the more useful service. The lord's immediate satisfaction and the continuing good of his house need not coincide.

This does not make criticism safe. A principle that rulers ought to listen cannot guarantee that a particular ruler will. It does, however, prevent us from treating silence as the only imaginable form of fidelity.

The vocabulary of loyal service could be used to demand obedience. It could also be used to ask whether the person receiving that obedience was worthy of it. The relationship carried judgments in both directions, even when power did not.

In the seventeenth century, the Confucian thinker Yamaga Sokō posed an uncomfortable question about the warrior's place in society.

The samurai ate food he did not grow and used goods he did not make. If other people supplied these necessities, what justified his own position? Sokō's answer emphasized moral cultivation, service, and an exemplary life, alongside readiness to use arms. Duties to parents, children, friends, and a lord all belonged to the conduct he prescribed.^[7]

This was an argument for the usefulness of a privileged group. It should not be accepted as proof that farmers, craftsmen, and merchants lacked moral knowledge. Sokō assigned the warrior a special role partly by describing other people's opportunities more narrowly.

Even so, his question marks a change in emphasis. A person whose status rested on readiness for war needed a purpose that could occupy years in which no great campaign required him. The demand to serve had to become a way of living, not merely a promise about what one would do when summoned to battle.

Useful service might therefore be judged through steady conduct as well as a conspicuous exploit. Learning, judgment, restraint, and reliability could become part of a warrior's account of his obligations.

This also complicates the familiar idea of a single code passed unchanged from the age of mounted combat into the age of castle-town administration. Ethical writing responded to the circumstances in which its readers had to find a place for themselves.

The historian Olivier Ansart's study of seventeenth- and eighteenth-century samurai discourse likewise emphasizes competing accounts of loyalty, rather than one universally accepted definition.^[8] That disagreement was not incidental. Once people asked what service should achieve, they could disagree about whether an impressive act had actually achieved it.

One of Japan's most celebrated acts of revenge would make the disagreement impossible to miss.

In 1701, Asano Naganori attacked the official Kira Yoshinaka inside Edo Castle. Asano was ordered to die by seppuku, ritual suicide, and his domain at Akō was confiscated. The retainers who lost their master also lost the house that had organized their service.^[9]

Their response was not an uninterrupted rush toward revenge. Ōishi Yoshio, one of Asano's senior retainers, sought restoration of the family. His associates later agreed to await a decision concerning Asano's brother before acting. Preserving a future for the house and avenging its dead lord remained different possibilities.^[9]

In early 1703 by the Western calendar, the group remembered as the Forty-Seven Rōnin—masterless samurai—attacked Kira's residence and killed him. Forty-six men subsequently held in custody were ordered to commit seppuku.^[9]

The story is often offered as the answer to the question of loyalty: they had nothing left to gain, yet gave everything for their lord.

But it did not produce a single judgment even among writers committed to warrior ideals.

In Hagakure, Yamamoto Tsunetomo criticized the delay. Kira might have died of illness while the retainers planned. In that case, all their preparation would have deprived them of the chance to act. Tsunetomo favored an immediate willingness to risk death over a calculation designed to secure success.^[10]

The disagreement cuts through the familiar tribute. Patience and coordination can look like the virtues that made the revenge possible. Under Tsunetomo's standard, that same calculation compromised the commitment. The eventual outcome did not settle whether the decision had been right.

His position was an extreme ethical argument, not a neutral description of samurai practice. Recognizing it does not require accepting its glorification of death. It reveals how differently people could define the act they wished to honor.

The government faced another problem. Devotion to a former lord did not erase the public consequences of an armed attack. If a retainer's conviction could authorize revenge, the victim's household could have convictions of its own. Private fidelity did not automatically produce a rule under which other people could live.

Nor did the participants bear every consequence themselves. A willingness to surrender one's own life cannot, by itself, justify taking another person's. Admiration for resolve and judgment of the action remain separate questions.

In 1748, Chūshingura transformed the incident into puppet theater, adding fictional characters and situations under different names.^[10] Its emotional power belongs to its history too. But the drama is not a transcript of what every participant thought.

An event capable of sustaining such different readings tells us less about an unquestioned code than about an unresolved problem: when several obligations conflict, what makes one of them supreme?

Around 1900, Nitobe Inazō published Bushido: The Soul of Japan in English.

He was not translating a single medieval rulebook. In his preface, he described questions from foreign interlocutors that had encouraged him to explain Japan's moral education. His account drew parallels with European literature, philosophy, and chivalry. It was an interpretation addressed to readers beyond Japan.^[11]

The title itself performs a remarkable expansion. Traditions associated with particular warriors, houses, and periods become a way to describe a nation's character.

Nitobe gave loyalty a central place. Yet even his idealized account did not reduce it entirely to compliance: he defended conscience and described persuasion and remonstrance as responses to a lord's error.^[11] Treating him as the author of a simple formula for blind obedience would flatten his argument too.

The historian Oleg Benesch places modern bushidō discourse within the late nineteenth-century search for national identity and Japan's changing relationship with the wider world. Comparisons with foreign ideals helped shape the supposedly self-contained tradition. Modern advocates selected and reorganized the past for purposes their predecessors had not shared.^[12]

This does not mean that earlier warriors possessed no ethics, or that every act of fidelity was invented afterward. We have already encountered documents, admonitions, and arguments that made obligations important. What changed was the claim that these diverse materials expressed one stable national essence.

A household's instruction tells particular people how to behave within its affairs. A national portrait invites readers to recognize an entire people through the selected ideal. Moving from one to the other changes the scale of the claim—and makes differences harder to see.

That distinction matters whenever a modern list of samurai virtues is used to explain a medieval decision. The list may contain values that earlier people recognized. It cannot tell us, without further evidence, how they ranked those values, whom they owed them to, or what they believed a difficult situation required.

A memorable account of a tradition can become part of the tradition's later history. It should not be allowed to replace all the earlier histories from which it was assembled.

These cases do not yield a percentage of loyal samurai or a verdict that they were all secretly opportunists. They offer a more useful pattern.

Loyalty acquired force through relationships people expected to continue: service recognized, households maintained, promises remembered, and reputations judged by others. Those relationships could make cooperation possible when acting together was dangerous or costly. They could also create obligations that survived the person who first received the service.

But a commitment did not come with an automatic solution to every conflict. A father and a superior might demand different choices. The immediate wish of a lord might endanger his house. A private act of revenge could challenge the order under which everyone else was supposed to live.

The hardest question was often not whether to be loyal, but what fidelity required when its objects diverged.

This is why neither heroic sacrifice nor calculated survival provides a sufficient answer by itself. Courage does not guarantee good judgment. Concern for a family's future does not prove cowardice. A person may act sincerely and still cause harm; another may recognize an interest of his own within a commitment that is real.

Return to Suenaga calling for a witness. He is not asking to disappear into an anonymous ideal. He wants his service to be seen and remembered. Around the risk he takes are people whose recognition matters and a future in which the deed may support more than his own name.

Loyalty belonged to that world of continuing claims, not outside it. To understand a samurai's choice, ask what he believed he owed, what others owed in return, and who would have to live with the result.