Christian replying to Marcus said:
> It certainly would be one. Another possible route within the TEI
> would be to employ the <choice> element together with <seg>. There
> has been some work on similar cases, e.g. linguistic annotation of
> Japanese or Sanskrit segmentation. (see http://www.tei-c.org/Activities/CE/cew12.pdf)
As far as I can tell, this report "cew12.pdf" does not fit the case
here, though I am not sure I understand the position taken. And the
"ruby" markup is only for the case where there are small phonetic
characters directly above the line to show how to read the characters
underneath. This works fine for modern Japanese.
> However, the issue David brought up is in fact a bit more complex (and
> that is why I said it had not been done to my knowledge), since it
> (specifically, the "kaeriten" mentioned above) also involves
> reordering chunks of the text to account for the differences between
> Chinese and Japanese sentence order.
To clarify, as Christian well knows, the "reordering" is not done in the
printed form, rather the kaeriten is a way of showing how to parse the
Chinese, by marking the subject and object, or verb and verbal prefix,
and so forth. So, it is essentially a way of diagramming the sentence.
But concretely it more-or-less reorders the characters into the same
order they would have if transformed into Japanese style sentence. The
kaeriten are all spacial and number glyphs. E.g. a reverse glyph (the
shape of a check mark) to the left of and between two characters means
to understand them as if the second came first, a number one and number
two pair of glyphs to marks a whole group of characters to be inverted,
and so forth.
Quite separately, on the right hand side of the column of Chinese
characters there are Japanese phonetic glyphs that insert prepositions
(post-positions actually), verb endings that show mood, subject markers,
and so forth, that are inferred from the meaning of the Chinese
sentence. These are between Chinese characters, and they in part
duplicate the kaeriten: a sentence subject-object division is shown with
a "one" and "two" kaeriten, but the Japanese sound that acts as a
sentence object marker (written as wo in small kana to the right between
two Chinese characters) is also sometimes written, just to be clear.
These phonetic glyphs are grammar indicators. Quite in addition, on the
same right side, but not between Chinese characters, there are also
readings of words that are difficult to pronounce or perhaps have a
special pronunciation here.
This seems rather absurdly complicated to describe, but it is an
extremely well worked out system that has been an important tool of
Buddhist learning in Japan for many centuries. It is definitely not
translation: the underlying Chinese is right there on the page. But when
read aloud, the Chinese sentence usually comes out in Japanese style
word order with verbs and particles inserted. And this is often felt to
be enough of a translation (which is of course often criticised nowadays
as stopping before a real translation.)
This interplay of different levels of language and precise placement of
different sizes is perhaps part of the reason that Japanese printing
turned away from moveable type back to woodblock printing in the early
17th Century. It certainly seems to be holding back Japanese use of
markup, and I fear is in danger of confining Japan to its own island of
proprietary computer software.
David Riggs, Kyoto