Did Shingen Really Say “People Are the Castle”?

A fresh translation

Original text

Words and concepts

• 城郭 (jōkaku)
  A fortified castle complex. The sentence contrasts this with an yashiki-gamae, a compound arranged as a residence.
• 或人の云ふ (aru hito no iu)
  “A certain person says.” The text transmits an attribution; it does not present the verse as a document written in Shingen’s hand.
• 人は城
  “People are the castle.” Repetition extends the metaphor through the stone walls and moat, making human relationships perform the work of defenses.
• 情 (nasake)
  Compassion, humane feeling, or considerate treatment. “Compassion” is used here without turning the word into a modern political slogan.
• 讎 (ada)
  An enemy, hostility, or enmity. The line can point both to a hostile person and to the hostility produced by conduct.

The missing words before the famous words

The celebrated line is often detached from everything around it. In isolation it sounds like a direct declaration: Shingen announces his governing philosophy, and the quotation passes intact into the present. Chapter Thirty-Nine gives us a more complicated object. Before the verse, the narrator says that Shingen did not construct a fortified castle within Kai for his protection. Then comes the small but decisive phrase, “A certain person says.”

Those four words do not make the verse worthless. They tell us what kind of evidence it is. The Kōyō Gunkan records a saying attributed to Shingen by an unnamed speaker. It places that saying inside a much longer account of his final discourse. A careful translation should preserve the wrapper instead of presenting only the jewel.

This distinction separates three claims that are easily collapsed. First, the verse exists in the textual tradition of the Kōyō Gunkan. Second, the work attributes it to Shingen. Third, Shingen himself composed and spoke precisely these words. The selected passage securely supports the first claim and explicitly makes the second. By itself, it cannot settle the third.

A castle made of relationships

The verse works because a castle is divided into parts. The first clause names the whole; the next two name its defensive structures. People are not merely said to live inside a stronghold. They become the stronghold, its stone facing, and its ditch. Repetition gives the sentence its compact force while moving the reader from the abstract idea of defense to things that block, support, and surround.

The final pair explains how that human fortification is maintained. Nasake becomes an ally; ada becomes an enemy. The movement is from people as material to relationships as cause. A following may protect a ruler, but the loyalty of that following cannot be treated as inert masonry. Considerate treatment creates attachment. Hostility creates danger from within the same field of human relations.

That does not make the verse a sentimental rejection of every wall. It is an argument about what physical defenses cannot supply by themselves. Stone does not interpret an order, carry information, judge a changing situation, or decide whether to remain loyal. The metaphor assigns decisive defensive value to people because people can act—and because their action depends on how bonds are formed.

The verse belongs to a portrait of Shingen

Placement matters. The selected lines occur during an extended speech assigned to the ailing Shingen. The surrounding prose praises a career in which, according to the work, he did not depend on another great commander, abandon a besieged position, or allow one of his own castles to be taken. The claim about living without a fortified castle in Kai arrives as another sign of security and personal authority.

The verse therefore does more than offer general counsel. It helps construct a posthumous image. The ideal ruler needs no imposing fortress at the center because his household and retainers supply a stronger defense. Compassion is not introduced as private softness; it is presented as one reason those human defenses hold. Enmity is dangerous because a ruler who produces it manufactures his own enemy.

This is the Kōyō Gunkan arranging memory. The book’s value lies partly in that arrangement. It shows later readers what qualities the Takeda tradition wanted to gather around Shingen’s name: confidence, command of retainers, restraint in treatment, and protection founded on people rather than architecture alone.

Not an argument against fortifications

A literal reading can make the saying collide with the rest of the same chapter. Chapter Thirty-Nine includes a heading for “five essentials of taking castles” and gives a compact list of defensive features. It also praises Shingen for preventing enemies from taking his fortified places. The work plainly understands castles as practical objects and treats knowledge of them as part of military judgment.

The nearby claim is narrower: within Kai, Shingen supposedly did not build a castle complex for his own protection and remained in a residential compound. Even that is presented as praise within the narrative, not as a survey of every fortification in the province. The verse then turns the claim into a memorable explanation of rule. Its target is not masonry everywhere; it is the belief that masonry can replace dependable people.

Reading the aphorism as “Shingen opposed castles” therefore loses both textual scale and rhetorical purpose. The contrast elevates the ruler’s human foundation. It does not erase the book’s detailed interest in fortified positions, siegecraft, or defensive design.

What can “a certain person” tell us?

An unnamed source may initially feel like a defect. For historical method, however, the anonymity is useful because the text admits mediation. The writer could have introduced the verse simply as Shingen’s words. Instead, the sentence leaves another voice between the lord and the page. That voice may preserve a saying already in circulation, but the passage does not identify who transmitted it, when it was first heard, or how its wording changed.

The label “verse” also tells us how the line is presented, not how its authorship was authenticated. Its parallel phrases make it easy to remember and repeat. Memorable form can help a statement survive, yet the same quality allows it to travel independently of its first context. The modern fame of the first half demonstrates that process: “people are the castle” circulates more readily than “a certain person says.”

The responsible conclusion is neither “Shingen certainly said it” nor “Shingen never said it.” The passage gives evidence for an attribution inside a named work and a specific chapter. Where the chain of testimony stops, the article should stop too.

What the four lines do not prove

They do not prove that Kai lacked castles or defensive works. The first sentence is part of a eulogizing account focused on Shingen’s own protection. Establishing the province’s built environment would require other documentary and archaeological evidence.

They do not prove that every retainer experienced compassionate rule or that the Takeda house functioned without conflict. An idealizing statement tells us how authority is praised and explained. It does not turn the entire social history of a domain into one harmonious line.

They also do not authorize a timeless management slogan. The metaphor is embedded in a warrior-house text, inside an account of a ruler’s final words and reputation. Its compact insight can be understood by modern readers, but its historical setting should not be traded for an easy lesson about organizations.

Most importantly, the lines do not support deleting their own introduction. “A certain person says” is not clutter around the quotation. It is the part that shows us how the quotation reaches the page.

A more careful way to repeat it

The famous words remain powerful after their uncertainty is restored. In fact, the caution makes them more interesting. We are no longer reading an inscription suspended outside history. We are watching a tradition attach an ideal of human defense to Shingen and preserve it in a form designed to be remembered.

A precise summary would therefore be: Chapter Thirty-Nine of the Kōyō Gunkan reports that “a certain person” attributed the “people are the castle” verse to Takeda Shingen. That sentence is less tidy than a quotation card. It is also closer to the source.

Source

Kōyō Gunkan (甲陽軍鑑), Chapter Thirty-Nine (品第卅九), Scroll Twelve, the verse introduced during Shingen’s final discourse. The wording was checked against the public Wikisource transcription of Chapter Thirty-Nine and its linked National Diet Library image, frame 159, from the 1893 Onkodō edition. For an earlier printed witness and bibliographic context, the NIJL/CODH record describes a twenty-scroll, twenty-three-booklet Kyoto edition printed in 1656 and bearing an attribution to Kōsaka Danjō.