Beyond “Fūrinkazan”: What the Kōyō Gunkan Says About Shingen’s Flag

A fresh translation

Original text

Words and concepts

• 其疾 / 其静
  “Its swiftness” and “its stillness.” Each comparison describes a condition or quality; the wording is more than a list of emblematic nouns.
• 侵掠
  Incursion and taking or seizure. “Attack” is smoother English, but the compound has a harder material edge that the translation preserves.
• 四ツの語
  “These four phrases.” The source itself groups the clauses as four units, although their now-familiar one-character label does not appear here.
• 四方
  Literally “four-sided” or square. In this sentence it describes the form of the flag, not a four-direction deployment.

The source begins with an object

“Fūrinkazan” often reaches modern readers as a compact name: four characters, four images, one famous commander. Chapter 43 of the Kōyō Gunkan gives us something more concrete and more limited. It places the words under a heading for Shingen’s flags, quotes four clauses, and then specifies their physical presentation. The ground is black, the writing is gold, and the flag is square.

That order matters. The passage is not introduced as a timeless summary of strategy. It is part of an inventory embedded in a much larger chapter about military offices, counsel, flags, rules in camp, rewards, and the people valued or disliked in an army. The four clauses belong to the material and organizational world of that chapter. They are words made visible and carried.

The surrounding list also prevents the black-and-gold flag from swallowing the entire Takeda display. Immediately before the quoted lines, the text names two red flags of Hachiman, two red flags of Shōgun Jizō, and one ancestral Takeda flag. It then counts six flags in all. The famous object is therefore one member of a described group, not the only standard the source associates with Shingen.

Four clauses, not four nouns

The popular label joins 風, 林, 火, and 山: wind, forest, fire, mountain. That compression is memorable, but it removes the grammar that tells us why each image appears. The first clause concerns speed. The second concerns stillness. The third concerns incursion and seizure. The fourth concerns not moving. The natural images are comparisons attached to different kinds of conduct.

Reading only the four nouns can make the series look like four interchangeable virtues. The syntax resists that. Wind and forest qualify contrasting tempos. Fire marks an action that spreads through what it reaches. Mountain gives an image for remaining fixed. The lines are arranged around changes of condition, not a single command to move rapidly at all times.

The subject is also left implicit. Classical Chinese permits the clauses to work without repeatedly naming an army or commander. The fresh translation uses “its” to preserve that openness rather than inserting “Shingen” into every line. The chapter’s heading connects the words to his flag; it does not convert each comparison into a report of what he personally did in a named campaign.

The selected words from Sunzi

The four clauses come from the “Military Manoeuvring” chapter of Sunzi. In the publicly available Chinese text, they stand within a sequence of six comparisons: after wind, forest, fire, and mountain come being difficult to know like shadow and moving like thunder. The Kōyō Gunkan flag passage gives only the first four. That is selection, not a quotation of the entire sequence.

There is also a visible textual difference. The public Sunzi text reads 其徐如林, commonly understood as being measured or deliberate like a forest. The Kōyō Gunkan transcription used here reads 其静如林, “its stillness is like a forest.” The two versions point in a similar direction, but they are not the same written character.

It would be rash to turn that single difference into a story about Shingen deliberately rewriting Chinese strategy. A transmitted work can preserve variant wording through copying, printing, editorial choice, or an earlier form available to its compilers. What can be stated securely is narrower: the cited edition and transcription of Chapter 43 have 静, while the compared Sunzi text has 徐. A close reading should show the difference without inventing its cause.

What black and gold contribute

The prose after the clauses is short, but it changes our understanding of them. Color and material contrast make the writing legible as an emblem. Black is not treated here as a moral symbol, nor does the sentence explain why gold was chosen. The source simply records the combination. Any interpretation of the colors beyond that description would need other evidence.

The same restraint applies to the square form. The text tells us what shape it attributes to the flag; it does not provide dimensions, fabric, maker, date of manufacture, or a chain of custody to any surviving object. A later representation may resemble the words in the chapter, but resemblance alone cannot establish that the represented banner is the very object the text describes.

This is a useful example of how few words can carry several different kinds of evidence. The quoted clauses give the inscription. The following sentence gives color and shape. The chapter setting gives a place among other flags and military offices. Each level supports a different claim, and none should be asked to prove more than it says.

A house tradition, not a battlefield photograph

The Kōyō Gunkan is a layered military text associated with remembrance and instruction in the Takeda tradition. Its value does not depend on pretending that every line is a contemporary inventory made beside Shingen’s equipment. The public edition behind the transcription was printed in the Meiji period from an existing textual tradition; another public manuscript witness confirms that the work circulated in many scrolls and copies.

For this passage, the safest formulation is therefore “the Kōyō Gunkan describes” or “Chapter 43 attributes.” Those phrases identify the actual evidence. They do not deny that such a flag existed, and they do not claim that the description is a modern invention. They keep the text and the historical object related without making them identical.

This distinction also protects the passage from a familiar mistake: using a famous emblem to explain every event in a commander’s life. Chapter 43 does not link each phrase to a named battle plan. It does not say that four units were assigned four separate roles. It does not offer a code by which “wind” always meant one maneuver and “mountain” another. Such readings may be imaginative, but they are not contained in these lines.

What these three lines do not prove

The excerpt does not prove that the flag was carried in every campaign, that every Takeda soldier could read its literary source, or that the four images summarize all Takeda military practice. It does not establish the authenticity of any surviving banner. It also does not demonstrate that “Fūrinkazan” was the ordinary name used for the object by the people described in the work. The compressed four-character label is absent from the quoted wording.

Nor does the passage prove that the Kōyō Gunkan copied one uniquely stable version of Sunzi. The 静/徐 difference is exactly the kind of detail that warns against that assumption. The comparison is valuable because the sources can be placed side by side, not because one short variation settles the history of transmission.

What the lines do prove is precise and substantial. Chapter 43 presents four Sunzi-derived clauses as an inscription associated with Shingen’s flag and describes that flag as gold writing on black in a square form. That claim is less sweeping than the legend, but it gives the legend a verifiable textual anchor.

Reading the emblem back into sentences

The most useful move is to reverse the familiar compression. Begin with “wind, forest, fire, mountain,” then restore the comparisons and their verbs. Restore the sentence that says where the words appeared. Restore the neighboring flags and the chapter that contains them. The emblem becomes not smaller but historically denser.

A four-character name is excellent for recognition. A source-led reading asks a different question: what did this particular text actually place on the page? Here the answer is an inscription, an object, and a carefully bounded description. The flag’s force comes not from four free-floating symbols, but from four clauses carried together.

Source

Kōyō Gunkan (甲陽軍鑑), Scroll Fifteen, Chapter 43, section “信玄公御旗” (“Lord Shingen’s Flags”). The quoted wording is checked against the public Chapter 43 transcription and its linked National Diet Library scan, image 192. The NDL bibliographic record identifies the 1892–1893 Onkodō edition and marks the digital item as publicly available and public domain. A separate National Institute of Japanese Literature manuscript record identifies a twenty-scroll, twenty-three-volume witness. For the wording of the six-part comparison, see Sunzi, “Military Manoeuvring,” in the Chinese Text Project.