Re: Proto-Altaic Phonology (inc. Vowel Harmony)

On Wed, 12 Jul 2000 01:18:14 -0700, Danny Wier <dawier@...> wrote:

>Not only are the databases online, but they can be downloaded and
>viewed and manipulated with Sergei's own program.
BTW, I think the program can have a lot of other linguistic applications.
Personally, I keep most of my conlanging stuff in Starling dbf files.
But unfortunately, it isn't enough user-friendly. I think only people
suited to DOS will like it. This winter I heard that a version for
Win95/NT was nearly ready, though.

>The DOS-based downloadable data contained a phonetic analysis (and
synthesis)
>of the member groups/languages of Altaic, with a reconstructed Proto-Altaic
>consonant and vowel system.  The results:
>
>Consonants:
>
>Labials: p' p b m
>Dentals: t' t d n r l s z
>Palatals: ^c' ^c ^3 'n 'r 'l ^s j
>Velars: k' k g N (= eng)
>Vowels: a e i o u "a "o "u
>(the last three could also be ja, jo, ju)
>
>Distinction between initial and non-initial is made for all consonants
except
>r, z, 'r and j.
Don't forget tones...

>Each of the vowels is shown with the five simple vowels in the
>following syllable, thus a-a, a-e, a-i, a-o, a-u, e-a, e-e... i_u-o, i_u-u.
>This demonstrates patterns of vowel harmony in Turkish (and Mongolian and
>Tungusic to a lesser degree).
>Turkic has patterns of front-back vowel pairs: a <> e, i <> I, o <> "o, u
<>
>"u.  The following vowel affects front-back status.  Rarely, round-unround
>status is also affected.
IIRC Starostin believes that the two vowels were partly directly
preserved in certain types of verbal stems in Proto-Tungus-Manchu.
Without that it would be difficult to choose between several alternative
reconstructions (e. g., similar reflexes in all languages could be
equally interpreted as _a <> i_ or _i <> a_).

In the other branches the disyllabic stems often contracted, with the
first vowel being modified by the second (i. e. reverse to 'vowel
harmony'). The vowel harmony in suffixal syllables (in Turkic etc.) is
considered secondary.

The initial variant of his reconstruction is explained in detail in
"Altayskaya problema i proiskhozhdeniye yaponskogo yazyka" -
downloadable from the same site, but AFAIK never translated into English.

This book contains a lot of other untrivial stuff, e.g. the demonstration
of the opposition 'voiced vs. voiceless' for word-initial consonants in
Proto-Turkic.

Note also that Starostin mostly accepts O. Mudrak's reconstruction of
Proto-Turkic vowel system (with oppositions e ~ ä, a ~ a2 and o ~ O).
These oppositions are not consequentially marked in the databases since
here the reconstruction depends on the Chuvash and Yakut data which are
not always available.

Funnily, it seems that Proto-Altaic reconstruction is now based on
a larger root list than Proto-IE... despite the fact that Starostin
includes in his main database only roots attested in three or more
branches (with some exceptions for the pronominal morphemes and the like).


Basilius