--- In
cybalist@yahoogroups.com, "Richard Wordingham"
<richard.wordingham@...> wrote:
> If one's willing to accept analogical explanations though, what
> stops secondary *-os being remodelled as *-es after primary *-
esi?
> This seems the simplest explanation if one won't accept a
> contrast /s/ ~ /z/.
It may be a matter of choice in the end:
EITHER *-es > *-ez > *-oz > *-os -> *-es after prim. *-esi, AND
3sg *-et > *-ed > *-od > *-ot > *-et after prim. *-eti
OR /s/ : /z/, no analogy needed.
This is aggravated however, by the parallel difference:
EITHER *d(e)y-ew-s > *d(e)y-ew-z > *d(e)y-e:w-z > *dy-e:w-s, AND
*k^l(e)-ne-w-s > *k^l(e)-ne-w-z > *k^l(e)-ne:-w-z > *k^lne:ws ->
*k^lnews after prim. *k^lnews-i
OR /z/ (lenghtening) : /s/ (not lengthening), no analogy needed.
IF influence from primary on secondary forms of the verb is assumed,
this raises another question: Why are the long-diphthong roots
consistently processed on the basis of the secondary forms? This is
seen in, e.g., Skt. pá:ti 'guards' from *peH2-t with regular loss
of /y/ in *peH2y-t, while no such reduction should occur in *peH2y-
ti. These reductions are younger than the Schwundablaut (loss of
unaccented short vowels), while the lengthening in *dyé:ws is older
than that. That presupposes two different attitudes to the status of
the "primary marker" *-i, which in long-diphthong verbs is treated
as not belonging to the word, while the inflectional endings pay a
lot of attention to it.
There is also the /s/ of the s-aorist which turns up with
lengthening of the root vowel, but is never word-final. Is this an
incomprehensible case of lengthened grade (which cannot be a marker
of durativity since it is an aorist), or is it the effect of a
lengthening sibilant, the same morphophoneme as in the nominative?
The easy solution is /s/ : /z/ at the relevant stage.
Strange things do happen, but why choose them in bundles at all
costs?
Jens