- Old Japanese Pronouns [pdf] by Bjarke Frellesvig
19 Aug 2007
Old Japanese Pronouns
All I do is search, search, search. Here's another link that somebody might appreciate:
Labels:
language,
old japanese,
pronoun,
proto-japanese
Subscribe to:
Post Comments (Atom)
Very nice!
ReplyDeleteJapanese is so good at completely changing in a matter of 1200 years. Haha. Of those original pronouns there's next to none that made it into modern Japanese with their original semantic values.
It must be so much easier to write haiku's with those monosyllabic pronouns though. :D
Anyway, this reminded me of my paper I wrote about historical causes of Rendaku for my Historical Linguistics class. I think I'll make an entry on my blog giving the run down of my 30 page paper.
phoenix,
ReplyDelete> Japanese is so good at completely
> changing in a matter of 1200
> years.
I do not think that Japanese has really changed all that much over the centuries, especially in regard to pronouns.
> Of those original pronouns there's
> next to none that made it into
> modern Japanese with their
> original semantic values.
Most of them survive to this day undergoing none or very minimal phonological and semantic changes.
-ware (and wa ga..., wareware)
-tare > dare
-onore (and onodukara > onozukara)
-koko
-soko
-sore (Frellesvig neglected this long form which is conveniently attested in Nara-period man'yoogana: 正倉院文書: 万葉仮名文 (c. 762) and 歌経標式 (772))
-idure > izure
-iduku > iduko > idoko > doko
I think ware is hardly a minimal change of meaning. Although it is still used as a singular at time, ware mostly took on the meaning of wareware. I had completely forgotten about "wa" though.
ReplyDeleteI wonder what create the shift from tare > dare. It isn't exactly the kind of word that is likely to undergo rendaku. After all it will hardly ever follow "no" nor is it easily the second element of a compound noun.
And I'm curious how the iduku > doko shift took place. Would you happen to know more about that?
Consider all the new variants of pronouns (mostly from nouns or contractions of older pronouns) in Modern Japanese: わたくしwatakushi, わたしwatashi, 僕 boku, きみ kimi, わし washi, 俺 ore, あたし atashi, 拙者 sessha, 手前 temae, てめー temee, 吾輩 wagahai, うち uchi, 某 soregashi, あたくし atakushi, お前 omae, 貴方 anata, 其方 sonata, 己 onore, あんた anta, 貴様 kisama, 汝 nanji, やつ yatsu, 彼女 kanojo, etc., etc., etc.
ReplyDelete(Phew! What a list. Did I miss any?? And don't forget some crazy ones used in animé like ore-sama "my esteemed self")
Which of these pronouns above are present in the oldest stage of Old Japanese? If few or none are, then how is the claim that the pronominal system "changed little" over the centuries a valid one? The contractions alone (e.g. watakushi > watashi > washi) show that a lot did change.
> I think ware is hardly a minimal
ReplyDelete> change of meaning.
The pronoun ware is still used in numerous dialects (particularly in the Kansai regions of Kyo:to and Nara) to mean "I".
> how is the claim that the
> pronominal system "changed
> little" over the centuries
Because of the the original OJ pronouns, many (as I listed) still exist today with little or no changes.
Sure, there are new ones. Lots of new ones. But the fact remains that many of the original ones still exist as well.
> Phew! What a list. Did I miss any??
Yeah, a lot. A few years ago I enumerated a list of pronouns. Including all variations, there were well over a hundred.
> I wonder what create the shift
> from tare > dare.
The voiced dare form is rather new and begins to appear during the 18th century. Most likely it is due to analogy with "doko", ie reanalyzing the initial /d/ as expressing interrogative. It most definitely had nothing to do with rendaku.
> I'm curious how the iduku > doko
> shift took place. Would you
> happen to know more about that?
Extant forms found in the literature show the following progression, with early dates in parentheses: iduku (712) > iduko (late 9th century) > idoko (935) > doko (1099). The phonetic changes seem to be for ease of pronunciation.
Professor of Historical Linguistics, Japanese
To:kyo:, Japan
Extant forms found in the literature show the following progression, with early dates in parentheses: iduku (712) > iduko (late 9th century) > idoko (935) > doko (1099). The phonetic changes seem to be for ease of pronunciation.
ReplyDeleteI find it hard to accept that these changes 'just' happened, if they did not happen in any other words. Although the commonness of the word does allow it to undergo some irregular sound shifts, you'd still expect at least some similar changes to occur in other words. Is there any such proof of this?
Perhaps a link to La phonologie du Japonais (p.112) would be helpful in this discussion:
ReplyDelete«On trouve en langue moderne quelques cas de voisée initiale dans des lexèmes qui ne sont pas porteurs d'une connotation péjorative. Celles-ci sont dérivées historiquement d'une transformation phonétique impliquant une nasale, [...] ou bien elles résultent de la chute d'une voyelle haute, par example bara « rose » (dérivé de ibara / ubara), doko « où » (< iduku) ou daku « tenir dans les bras » (< idaku) (Yanagida, 1985, Kishida, 1984).»
English translation:
"We find in the modern language some cases of voiced initial in lexemes that do not carry a pejorative connotation. These are derived historically from a phonetic transformation implying a nasal, [...] or are the result of the deletion of a high vowel, for example bara 'rose' (derived from ibara / ubara), doko 'where' (< iduku) or daku 'hold in the arms' (< idaku) (Yanagida, 1985, Kishida, 1984)."
Might I also add that anonymous's statement, "The phonetic changes seem to be for ease of pronunciation", is a rhetorical one since all phonetic changes now occuring and in the past the world over can be qualified as such.
ReplyDeleteThere is a three-way locational demonstrative system:
ReplyDeleteproximal: koko
non-proximal: soko
interrogative: iduku
Notice that two of the three end in -o. Rather, I should say, two of the three end in -oko. Also notice that two of the three are two mora in length. iduku is the odd one out.
In an attempt to regularize (analogy), iduku first changes to iduko (now all three end in -ko). Then, either with further regularizing and / or due to the final -o, the medial -u- changes to -o-: idoko (now all three end in -oko). There are plenty of examples of the initial i- dropping off. However, again, it may more regularizing with the two mora koko and soko that justifies the drop of i- to doko (now all three end in -oko and are two mora in length).