>================
>
>The examples you gave
>are all tainted by the fact
>a schwa in unstressed positions
>could be coloured into o by a
>neighbouring w m or p.
The rules work the same in cases without labials. For
example, the suffix *-tér- ~ '*-tor-, or the "month/moon"
word:
**mát-nu:t-z > *méh1no:ts
**mát-nu:t-m > *méh1notm.
**mat-nút-a:s > *m(e)h1nésos [*]
Miguel Carrasquer Vidal
===========
I don't understand your graphic conventions.
1. mat > meH1
Does it mean t > H1 ?
Cf. Semitic words with m_H- "time" meaning.
2. Is <:> long meaning ?
It's anti-phonological to suppose that
unstressed vowels can be long
while stressed are short.
What are the proofs we have
non-laryngeal long stressed vowels .
ARnaud
===========
The only thing that matters is the weight of the preceding
stressed morpheme: if it contains a long vowel (**a: > o,
**i: > e: or **u: > o), or if it ends in a consonant
cluster, the svarita-lengthening does not take place (e.g.
in all neuters, except the s-stems).
===============
I am very much sceptical.
Because I am very far away from the point
where I could even make comments on that premiss.
Arnaud
=================
[*] The posttonic lengthening perhaps reflects an ancient
stress retraction (in the case of *méhnot-, this is
suggested by Kartvelian *(m)ttute "moon", which might
reflect the older accentuation ***mat-nút-).
Miguel Carrasquer Vidal
=======================
Georgian has
Month : tve
Lune : mtvari (obviously the month-maker)
I cannot see the connection between
*tve and meH1-nvt-
I think it is unsupported.
From Starostin kartvelian :
Proto-Kartvelian: *du(ś)te-
Russian meaning: луна, месяц
English meaning: moon
Georgian: tve (Old Georg. twite-, ttwe-)
Megrel: tuta-
Svan: došd-ul, došṭ-ul
Laz: (m)tuta-
Arnaud
===============