--- In
cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...>
wrote:
> Which would be a problem, again, if the semi-thematic paradigm
didn't
> belong to pre-PIE, which is does, by hypothesis.
That is fully logical of course. If the suggested semi-thematic type
is pre-PIE it may show pre-PIE phonetic changes, sure. But what
chance is there that it is pre-PIE?
The 1sg present ends in unlenited -imm in Old Irish in the verbal
types that do not continue the thematic conjugation (in Middle Irish
also in the thematic verbs). There is no other credible source of
the -mm than influence from the -mm of 'I am' where it arose
regularly from *-sm-. The OIr. copula is 1sg am or amm, 1pl ammi,
certainly reflecting *-sm-. Thus, there is *-sm- in Celtic, and
there is -sm- in Sanskrit ásmi and Hittite esmi. How could there not
have been *-sm- in Indo-European? So the root *H1es- met the
desinence *-mi directly in PIE, meaning that this form was not
thematic. There is also *-sm- in the OIr. 1pl ammi and in Sanskrit
smás, so how could there not have been *-sm- in PIE here? Then this
was not a thematic form either. I used to believe the semithematic
story myself, since sum is so aberrant, and "la grammaire comparée
doit se faire en utilisant les anomalies" (Meillet). But also esmi
is irregular in some of its synchronic systems, and sum is in fact a
bit less irregular than that in the Latin system. So all in all, one
will only ascribe pre-PIE age to the structure of sum if one has to.
And one doesn't have to.
Jens