On Mon, 19 Jul 2004
enlil@... wrote:
>
> 'Fraid not. I don't know of a clear cut case where there wasn't a
> cluster-breaking vowel.
No, I'm sure you don't, but I do, and I have told you many times how that
is decided in favourable cases. The inflectional accent moves to the next
vowel there is in the underlying structure whenever a syllabic flexive
is added. So, when such a movement skips possible stations between
consonants there just was no vowel to catch it there. If you put in
phantom vowels after all consonants the accent shifting rule become
completely erratic in that the accent now moves to the following vowel,
now skips one, now skips two vowels for no apparent reason (except that
it apparently wants to go to a vowel *we* know about which should be
enough to arouse anybody's suspicion). If we write in only the minimum
number of extra vowels we get a regular movement to the next vowel there
is, which certainly looks like a better solution.
In *H2nér-m, *H2nr-ós, the accent moves to the following vowel by any
standard;
in *H1dónt-s, *H1dnt-ós, as in *dwéys-ti, *dwis-énti, it does not move to
**-nét-, **-yés-, but skips one theoretically possible station which was
then apparently vowelless;
in *yéwg-t, *yug-médh&2 (Ved. áyok, áyujmahi), it does not land in
the slot w_g, nor in g_m, where there apparently were no vowels to take
it.
In all of these examples we find the accent moved to the next vowel if
we do not put any more in to confuse the picture. If we do, we get
unpredictable degrees of skipping.
>I mean, there's *stex- "to stand" but how
> do we know it's not *sed- "to sit" plus *-ex-? It looks to me like
> that it's probably just that, given that the semantics work.
Work to show what? What do you get if you add the same spurious suffix
to *legh- 'lie down'? A word meaning 'stand out of bed'??
Jens