Share this project


Share this project

A translation of a unique 17th-century compendium of samurai heraldry, annotated with the symbolism and stories behind the banners.
A translation of a unique 17th-century compendium of samurai heraldry, annotated with the symbolism and stories behind the banners.
Created by
389 backers pledged $14,101 to help bring this project to life.

The Process

Posted by Xavid (Creator)

First, the obligatory appeal.  We're well on our way to the golden $10,000 milestone!  Can you wonderful individuals spread the word and help us get there in the next 11 days?

And now, as foreshadowed, the think you've all been waiting for.  I'm going tell you what goes into translating O-umajirushi.  To keep this post a semi-reasonable length and to avoid boring all of you, I'm not going to give you a literal blow-by-blow of the translation process; consider this more of a high-level overview.

Let's look at some random page.  This is page 22 of the second volume.

To start with, it's in Japanese.  It's written vertically, from right to left, as is traditional in Japan.  It uses two of Japanese's three types of characters: kanji (logographic characters imported from China in ancient times) and hiragana (simplified kanji used to represent syllables phonetically).

First, lets talk kanji.  These are kanji that read "Andō Ukyō" (Andō, Capital Administration [Minister]):

Kanji are both the hardest and the easiest.  They're the hardest because there are thousands of them, and most have multiple possible pronunciations.  They're the easiest because they haven't changed much over the centuries.  I can look them up in a modern Japanese dictionary, like this one:

Or, especially for more obscure kanji, I can look them up online in WWWJDIC.  There are some kanji in O-umajirushi which have changed form in the following centuries and have to be understood based on context or a pronunciation given in hiragana, but these are relatively few.  In O-umajirushi, kanji are used mainly for names, titles, and digits in volume, page, and entry numbers; a few common kanji are also used repeatedly in captions.

Now, hiragana.  These are hiragana that read "tsukaiban" (messengers):

Hiragana seem like they should be much simpler to read than the kanji.  A modern Japanese student would tell you that there are only 46 of them, after all.  But unfortunately, things weren't quite that simple in the 17th century.  There are several differences that make reading O-umajirushi's hiragana more challenging, chief of which is the writing style.  Unlike the kanji in names and titles, which are printed in much the same style as kanji on a computer today, the hiragana in O-umajirushi are written in cursive (kuzushiji, "broken characters").  This involves both using a more stylized, less easily distinguishable writing style and sometimes using variant forms that are completely unrelated to the modern hiragana.  (Kanji used in captions and prose are also written in cursive.)  I've gotten far enough in my translation that I've developed a good key to the variants commonly used in O-umajirushi, but to build this, and when I encounter something new, I've had to rely on a few good resources.  One of these is:

Kuzushiji Kaidoku Jiten (Cursive Deciphering Dictionary) contains thousands of examples of cursive characters, organized by the first stroke, which makes it uniquely useful for identifying a character when you have no idea what it is.  A companion to this which is particularly useful for historical works is the Electronic Kuzushiji Dictionary Database, which lets you search for a particular character and see examples of that character from historical works, with dated citations.  This is invaluable when you can predict the hiragana in a name's pronunciation based on the modern reading of the name's kanji, and more generally for checking your work.

Most of the text in O-umajirushi is names, titles, and captions, not prose.  But for the small amount that's actually in complete sentences, I need to pull out the big guns:

Classical Japanese grammar is very different from modern Japanese: different verb forms, different particles, different usages for common words.  Getting the general gist isn't that hard, but getting a precise understanding of what verb forms, helper verbs, and particles are being used (since they often chain together and no spaces are used) is more daunting.

Still, I've completed my translation of Volume 1 and am well on my way though Volume 2.  Your unfailing support has been an inspiration as I wrestle with these cryptic characters, and I can't wait to emerge victorious and share the results with all of you.

Chris Olsen, Brandon Bradley, and 6 more people like this update.


Only backers can post comments. Log In
    1. Xavid Creator on

      If two people choose the same institution, I'll choose one of them at random and send them a message asking for a second choice.

    2. Rachel K Wright on

      Quick question about the "$50 donates a book somewhere" option. What if two people designate the same institution?