Editorial Principles
About this channel
Fragments of Japan tells the stories that the official histories of Japan rarely told: the voices of individual people, recovered from the words they left behind on paper.
We do not focus on heroes or the grand narratives of the state. We focus on letters, diaries, marginalia, and personal records — the small fragments of paper that survived from a particular life in a particular time. These fragments are often left out of the mainstream of history. But they often carry the temperature of their era more accurately than any official record.
We are a small team. We take time with each video. We try to release only what we are confident in.
What we cover
The subjects we choose generally meet these conditions:
- A primary written source by an individual (a letter, diary, memo, or record) is publicly available
- The person or event can be cross-referenced in multiple reliable published sources
- The story carries something universal — something that can move a viewer who knows nothing about Japan
A wartime soldier's letter to his mother. The diary of an Edo townsman who went to see Perry's black ships. A scrap of writing left by a woman on the day her castle fell. The doodles in a child's notebook from a Tokugawa-era school. A broadsheet that spread news through Edo. The era and the social class do not matter to us. What matters is that the person who wrote it was, indisputably, there.
How we use sources
Every fact and every quotation in our videos comes from published sources or publicly accessible digital archives.
Our main references include:
- Academic and trade paperbacks from publishers such as Iwanami Bunko, Kodansha Gakujutsu Bunko, and Chuko Bunko
- The National Diet Library Digital Collections
- The National Archives of Japan Digital Archive
- The Japan Center for Asian Historical Records (JACAR)
- Web-published historical documents from regional governments, museums, and universities
- Open-access academic papers (J-STAGE, CiNii)
- Wikimedia Commons (Public Domain / Creative Commons)
The video description and our website's per-video sources page list every source we consulted, distinguishing primary from secondary materials.
Our use of AI
We use AI as part of our production process. We never present AI-generated material as historical fact.
| Area | AI role | Human role |
|---|---|---|
| Script research and drafting | Claude (Anthropic) for background research and prose drafts | Verifying against primary sources, final editorial judgment |
| Translation | Claude for first-draft translation | Editing, tone, proper-noun verification |
| Visuals | Adobe Firefly for paper textures and abstract backgrounds | Never used to depict historical events; flagged on-screen as "imagery" wherever used |
| Music | Suno Pro for original score | Selection per scene |
| Voice | Human narration as default; AI voice flagged on-screen wherever used | Reading, breath, tone |
We do not use AI-generated imagery to recreate the faces of real historical figures or the events they lived through. Even our reconstructions of letters as paper are clearly marked as visual representation, not the original artifact.
Copyright and quotation
We follow Japanese copyright law. Our quotations are made under Article 32 of the Copyright Act and follow these rules:
- We quote only from published works
- We quote the minimum needed for our purpose
- Our commentary remains the principal element; the quotation is supplementary
- We display the source on screen, in the video description, and on our website
- We distinguish quoted material clearly through voice, on-screen text, and visual treatment
If you have a concern or correction regarding our use of any material, please contact us at the address below. We will review and, if needed, update or correct.
On translation
We translate every video into English ourselves. Claude assists with the first draft, and we revise it over time.
For historical terms and proper nouns, we balance accessibility for English-language readers with fidelity to the original Japanese meaning. For example, we do not always render the Japanese word tokko as "kamikaze." In some contexts we choose tokko pilot instead, because the word kamikaze in English has accumulated a particular weight over many decades. When we are reading the letter of a single person, we want to set that weight aside, at least for the duration of one video, so the reader can encounter the person directly.
Corrections and updates
Anything made by human hands can contain errors. When we discover one, we do this:
- Add a correction note to the video description
- Record the correction in the video's sources page on our website
- For significant factual errors, re-upload a corrected version of the video, or publish a correction video
- Make the history of the correction visible
We welcome viewer feedback.
Contact
For factual corrections, copyright inquiries, or any other matter, please reach us at:
- Email: contact@fragmentsjapan.com
- The description box of any video
- A direct message on our official social channels
We are a small operation and may not be able to reply immediately. Thank you for your patience.
About us
Fragments of Japan is produced by a small editorial team. Research, script, filming, editing, translation, and publishing are all done by us.
What we are trying to do is to bring one person from a particular past — as much as possible in that person's own words — to a viewer in the present. We hope this can be a quiet place where someone's fragment of paper finds its way to someone else.
Last updated: April 20, 2026