Site operator · translator · editor
Shiki Tsushima
An independent translator and editor building a bilingual archive in which every historical claim can be traced back to a short, identified passage.
Shiki Tsushima (津島史紀) operates, writes, translates, source-checks, and maintains this site.
The work begins with complete classical texts rather than lists of famous quotations. Published projects include full modern translations of the Shinchō Kōki, Bansenshūkai, and Book of Five Rings, together with other works on Sengoku records, ninjutsu manuals, classical sword strategy, and the Seven Military Classics. Books are published independently through Kindle Direct Publishing in English or Japanese according to the source and project.
This page makes no claim to an academic degree or institutional affiliation. The basis offered to readers is narrower and verifiable: a record of published translations, precise source notes, links to library witnesses, and a correction process that asks for locatable evidence.
The texts covered here
The archive currently works across four groups of sources.
- Sengoku records — the Shinchō Kōki, Kōyō Gunkan, Zōhyō Monogatari, and documents associated with Toyotomi Hideyoshi.
- Ninjutsu manuals — the Bansenshūkai, Shōninki, Ninpiden, and modern writings associated with Seiko Fujita.
- Sword texts — Miyamoto Musashi’s Gorin no Sho and Yagyū Munenori’s Heihō Kadensho.
- Military classics — the Liu Tao, Three Strategies, Sima Fa, Sunzi, Wuzi, Wei Liaozi, and the Questions and Replies.
The source and citation policy identifies the principal printed editions, manuscripts, library catalogues, and digital repositories used to check these works.
Why the site is different from the books
A complete translation preserves the order and full architecture of a work. A site article does something else: it stops over two to four lines and asks exact questions. What does a difficult term mean in this sentence? Where does the passage sit in the scroll? Which parts of a familiar story are actually present? Which conclusions would require evidence beyond the selected text?
For that reason, every web translation and explanation is written afresh. Site articles are not extracts, rearrangements, or lightly altered versions of the KDP editions. The books and the website share primary sources, but the sentences, organization, explanatory questions, and limits are newly made for the page.
How a reading is made
- Read the source in full and select a passage that can stand as a focused question.
- Identify the printed edition or manuscript witness and verify its bibliographic record.
- Transcribe only the short passage needed for the article and record its location.
- Write a new English translation for the page, without copying a published translation.
- Keep the original, translation, vocabulary notes, and interpretation in visibly separate sections.
- Test every historical claim against the selected wording or a separately locatable passage.
- State what the evidence does not establish, especially when a popular retelling goes further.
- Check internal links, language alternates, source URLs, mobile layout, and cumulative quotation length before publication.
Examples of the method
The article on Nobunaga’s supposed skull cups starts with the verbs in the Shinchō Kōki. The entry records three heads, their hakudami preparation, their presentation, and a drinking party. It does not use a word for cup, pouring, or drinking from the object. The article therefore distinguishes a grim documented display from a more precise later image.
The reading of the water-spider diagram records the dimensions actually written beside the drawing. It does not supply an undocumented wearing method merely because modern demonstrations need one.
The reading of striking and hitting treats Musashi’s two verbs as terms inside a technical sequence rather than turning one sentence into general-purpose motivation. These pages are useful not because they reproduce famous texts, but because they expose the reasoning between a source and a historical claim.
Quotation limits and public-domain sources
Short passages from public-domain works are quoted for explanation. The site keeps an internal cumulative ledger for each source. A Japanese and English pair counts as one excerpt, not two, and a source closes to new excerpts before the running estimate reaches ten percent of the work.
Library and archive images are linked only when the holding institution provides a stable record and the rights status permits it. A repository’s image, a transcription, and this site’s new translation are different objects; the citation notes are written to keep them distinct.
English and Japanese are companion editions
The Japanese articles are not machine translations of the English articles, and the English pages are not automatic renderings of the Japanese. They share the same short original passage. The explanation, heading structure, examples, and definitions are written independently because readers in the two languages need different background.
Corrections and responsibility
Historical texts and their witnesses can remain uncertain. A reading may need revision when a better image, a variant manuscript, or a stronger bibliographic record becomes available. Corrections that identify a work, witness, page, scroll, chapter, or line are welcomed and checked before the article is amended.
Editorial questions, source corrections, privacy requests, and site-operation inquiries can be sent through the contact page. The site’s privacy policy explains data and cookie handling; the disclaimer separates historical discussion from practical instruction.