Mike Ayers asked:
> > From: Edward Cherlin [mailto:Edward.Cherlin.SY.67@aya.yale.edu]
>
> > The 'tsu' sign in reduced form is traditionally used in Japanese for
> > consonant doubling (chyotto is written chi yo tsu to), but
> > has been adapted
> > for glottal stops at the end of words.
>
> Odd. I've always considered Japanese "double consonants" to be
> glottal stops. Could anyone please explain the difference?
and Rick McGowan responded:
> They are glottal stops. But Japanese writing doesn't have a (standard)
> means of expressing a glottally stopped vowel pair. It only can express
> consonants. One supposes that a small "tsu" would suffice, e.g.
> ハヴァイッイ => hawai'i... And probably has already been used somewhere
> to that effect. As Ed Cherlin pointed out, "tsu" has been adapted for
> word-final consonants... in that sense, "tsu" is effectively used as a virama
> already.
Well, not exactly.
The Japanese double consonants are prototypical examples of what
linguists call *geminate* consonants. They are not glottal stops,
but rather normal consonants with prolonged occlusions at the
points of articulation of the consonants.
So you have cases with voiceless stops:
hakken 'discovery'
hatten 'development'
happa 'to blow up (with explosives)'
But you also get a geminate distinction on fricatives:
hassei 'utter a sound'
hassha 'departure'
What you don't get (in Japanese) are geminate voiced stops. Instead,
the same morphemes in front of syllables starting with voiced stops
result in the full form of the '-tsu' syllable:
hatsugen 'a statement'
hatsudoo 'to put in motion'
hatsubai 'to put on sale'
The reason for this is historical -- the forms in question are mostly
borrowed in from Chinese, and the relevant Middle Chinese syllables
ended in final -k, -t, or -p. (In some Middle Chinese, and in more
recent Chinese dialects, these final consonants sometimes collapse,
more or less completely into final -?, i.e. final glottal stop, in
Chinese.) The Japanese phonological reanalysis on borrowing interpreted
these Chinese final consonants as causing gemination of following
voiceless consonants, but before voiced consonants, as -ku, -t(s)u,
and -0 (zero, but vowel-lengthening) respectively.
Modern Japanese has no segmental glottal stop (meaning no distinct
phoneme for a glottal stop, contrasting with other consonants).
However, it *does* have prosodic glottal stops. (This is actually
a similar situation as for English, which also lacks a normal
segmental glottal stop, but which makes use of it for sounds
marginal to the sound system -- cf. "uh-oh!") Japanese is rich in
prosodics to convey all kinds of paralinguistic information. And
one common pattern is lengthening utterance-final vowels, combined
with various pitch changes, and then abruptly cutting them off
with a gottal stop. The nonstandard (but widespread) orthographic
way of representing that is to use a length mark (sometimes a very
looooooong [<== example of nonstandard orthographic representation
of prosodics] length mark *hehe*), followed by a small 'tsu' to indicate
the abrupt cessation with a glottal stop. This doesn't mean that
small 'tsu' is a glottal stop -- it certainly isn't in its standard
usage in the orthography -- but it is being used to represent
one in this particular context.
And Rick is correct that in the use of Katakana to attempt to
represent pronunciations of foreign vocabulary (at least that
which has not been nativized to the Japanese sound system), the
use of small 'tsu' to indicate consonant gemination has been
generalized to indicate simply final consonants. In that, it
is working like the Burmese killer, to drop off, what to Japanese
ears, at least, sounds like the inherent vowel (since *every*
syllable is supposed to end in a vowel or -n in Japanese).
--Ken
This archive was generated by hypermail 2.1.2 : Mon Jul 09 2001 - 18:35:59 EDT