Japanese vowels

["Dylan W.H. Sung" <mabr12@dial.pipex.com>]

It is as Dr. Lau has commented, that the Japanese vowels of present day
speech is similar to the vowels in Hakka of the Hong Kong region. But this
is reflected in the short vowels only, as in the english words can, get,
him, mop, sock.

My good friend Thomas Chan who supplied me with the Cantonese tones
earlier, also sent me a copy of his course work handout on the use of
Chinese characters for the transcription of the Kojiki (see earlier message
Re: Hakka Language and Japanese Go-On Borrowings).

In it, there were a list of characters that had been used to transcribe the
individual syllables of Japanese then. Remeber the Kojiki was finished in
712AD, and so had 8 distinct vowels in this old variety of Japanese. They
are the letters 

a i1 i2 u e1 e2 o1 o2

where the subscripts 1 and 2 for i, e and o show there to be
differentiation between these sounds. 

Mr. Chan said that there are very few words which had the vowels for e and
so there is not much certainty as to the exact use the ancient Japanese put
to these. 

We must remember too that Japanese kana (literally "false name") today
having some 40 plus symbols are derived from the Chinese characters. There
are also Japanese kokuji characters that are used for the transcription
also, and hence have only a (native Japanese) Kun reading and no On
reading. 

>From my comparison of the words, I am tenuous about the following.

I have suggested to Thomas Chan that the letters are as a result of
applying chinese tones to the character sounds and so giving the characters
an artificial distinction, whereas Japanese today only has different levels
of pitch rather than true tones on the fashion that Hakka or any Chinese
dialect has. We are agreed that the o2 vowel is predominantly represented
by characters which have the -ng ending in Chinese, suggesting that the use
of the sound for transcribing Japanese is subtle in its distinction of o1
and o2. 

I have looked at it from the the Hakka phonology viewpoint and 
here is an excerpt of my email to Thomas Chan 26 Dec 1997. He is a
cantonese speaker, interested in Hakka linguistically.

==========================================================================
i1 is like the modern i in Japanese
i2 would have the Hakka ~ui ending in my speech, but not true of MeiXian
(as far as I know) as in MX ~i is the ending. Some typical relations

MX 	H	E
fi	fui4	meeting, association
fi	fui4	a fee
fi	fui1	to fly
fi	fui2	fat
mi	mui4	flavour
mi	mui1	beautiful

the split in the e vowel is probably due to tonal effects. Japanese not
having a true tone system would try to imitate the Chinese through this
crude method, as its spoken sound has high and low levels anyway. 

e1 tonally lower in pitch
e2 tonally higher in pitch

o1 has the ~u sound in Hakka.
o2 (from the characters) seem mostly to have ~ng endings. I suspect this o2
has legato properties whilst the former o1 has staccato properties.

The e's I need more characters to see if this is true. I note that some of
the characters did not appear to have the sound as a borrowing in the
dictionary. I think this may possibly be due to the 'most common'
considerations.
============================================================================

===

After the Meiji Restoration until the present day, there has been continual
change in the language and education system in Japan. These modernisations
have gotten rid of many of these extra vowels and standard Japanese today
is the Kanto dialect of Tokyo rather than the western Japnese KanSai
dialects.

Dylan.