Douglas G. Wilson douglas at NB.NET
Sat Feb 27 06:57:16 UTC 2010

Wilson Gray wrote:
> It's been over fifty years since I took two quarters of Japanese at
> UCLA from a native of Tokyo, but, if memory serves - and it very well
> may not - ç”° spells [ta] and not [da]. It has the meaning, "field," and
> is the first syllable of the name, _Tanaka_, "field middle," reputedly
> the most common Japanese name.
> IAC, according to the rule by which /i u/ are devoiced between or, in
> word-final position, after voiceless consonants, /matuta/ > [mats ta].
> Of course, there may be other peculiarities of the phonology of
> Japanese here relevant, of which I have not the slightest knowledge.

The peculiarity here is rendaku ("sequential voicing"), I think. The
Wikipedia article on "rendaku" seems (to this naif) pretty good for its
size. In short, the rendaku (voicing) involves the second part of a
'word', so -- given the basic kun'yomi (pronunciation) "ta" -- Tanaka
would be expected, but rendaku can make the same kanji be voiced "da" in
the last part of a name, e.g., Matsuda or Toyoda or Honda ... but it's
not so simple, especially with names, so there exist alternative
surname-pronunciations without rendaku, Matsuta, Toyota, Honta [of these
few only Toyota is (I think) relatively prevalent]. Even aside from
names, rendaku seems to be a bag of worms.

Aha! The Kubozono reference in the Wiki article is interesting and
readable on-line. It includes (section 2.2) a summary of a study (Sugito
1965) specifically addressing rendaku affecting exactly this "ta" in
names! The rendaku can be predicted (sort of) after all, I guess.

-- Doug Wilson

The American Dialect Society -

More information about the Ads-l mailing list