Re: Romanian verbal paradigm

--- In cybalist@yahoogroups.com, "Richard Wordingham" wrote:

>>>> Almost everything in the Romanian conjugations points to Latin
>>>> and Latin alone.
>>>
>>>> What other IE language has an e:-subjunctive for a:-stems, and
>>>> an a:-subjunctive for all other stems?
>>>
>>> It is strikingly reminiscent of that alternation -e- for thematic
>>> (-a-) stems versus -ya:- for athematic (non -a- :) stems in the
>>> Sanskrit optative active.
>>
>> "reminiscent"?! That would imply some kind of kinship between
>> Sanskrit alternance in optative (1st conjugation) and Latin (1st
>> conjugation) subjunctive alternance.
>
> No! All 'reminiscent' implies is that they look similar.

Up to some point, as you & Miguel agreed.

>> Why are you linking it with a different verbal mode alternance for
>> a quite distant IE language rather than to corresponding alternance
>> in the originating Latin language?!
>
> I was answering Miguel's question. He is claiming that individual
> elements of Romanian conjugation scream out that it derives from
> Latin; I am saying that one has to look more widely to be sure of
> the Latin origin.

Yes, but coincidental partial similarities without a theory
supporting
why one should look with much care at a different origin and verbal
mode
alternance in a remote language cannot be counted as "pointing to"
that
remote language. It is simply meaningless.

>>>> What other language has a present ptc. (gerund) in -nd-?
>>>
>>> If the present participle had survived in Albanian, wouldn't it
>>> show -nd-? As it is, I can only think of the Germanic languages.
>
>> Survived... where from?!
>
> PIE *-ont-.

AFAIK, Albanian doesn't exhibit that (I can think e.g. at Pokorny
1282
which suggest rather PIE *mend- > Alb. ment, or #490 PIE *ent- > Alb.
ent,
int, there is no word suggesting eventual voicing of PIE /t/ in
Albanian
and I never saw such a law).

>> The forms in -nd(o)- are specific to Latin & Oscan-Umbrian.

... the point being this is really a Italic specific innovation.

>>>> What other language has verbal forms (pqpf.conj.) in -assem,
> etc.
>>>
>>> -ss- has degeminated in Romanian, so it looks a rather like a
>>> thematic sigmatic aorist. Sanskrit again.
>>
>> Keeping in mind that Romanian analogically inserted -rã- as plural
>> marker (and sometimes an extra -se-), there is nothing to suggest
>> rather Sanskrit over Latin for pqpf:
>> Rom (pqpf ind.):
>> -Vsem, -VseSi, -Vse, -Vse(rã)m, -Vse(rã)Ti, -Vse(rã)
>> Lat (pqpf. subj.):
>> -(i)ssem, -(i)sses, -(i)sset, -(i)sse:mus, -(i)ss:etis, -(i)ssent
>
> Typos: -sse:- in the 2nd person forms.

My Latin grammar had length marks exactly as I wrote above, I hadn't
check other sources when writing it; but you're right, also the 2nd
singular has -e:-.

>> Skr (thematic sigmatic aorist):
>> -(ai)s.am, -(ai)s.i:h, -(ai)s.i:t, -(ai)s.ma, -(ai)s.t.a, -(ai)
> s.uh
>
> That's the athematic sigmatic arorist!
>
> For Sanskrit _dis'_ 'point', we have the thematic sigmatic aorist
> adik- -s.am, -s.ah., -s.at, -s.a:ma, -s.ata, -s.an .

Still doesn't exhibit -(s)se-.

> I was answering the question. I'm not claiming a better match with
> Sanskrit than with Latin.

See above, about different distant verbal paradigms, showing up some
partial similarity.

>>>> What other language mixes s-aorists with true perfects?
>> [...]
>>> How can you tell that the Romanian simple perfect derives from
>>> the PIE perfect?
>>
>> He does not say that.
>
> I dispute that.

You shouldn't. Miguel stressed indirectly what I wrote above:

>> As a matter of fact, Romanian simple perfect simply continues Latin
>> perfect

... by pointing out a specific feature of Latin perfect (mixing)
within
the IE family, feature observed also in Romanian.

> And accept this, if you replace 'perfect' by 'perfect' system. I
> qualify the acceptance because the plural of the Romanian simple
> perfect formally corresponds to the Latin pluperfect.

(Daco-)Romanian 1st & 2nd plural were analogically rebuilt (as
already
pointed out) after dialectal split, with -rã- of the 3rd plural
emerging
as plural marker. Only the 3rd plural, with -rã instead of
etymological
-ru can be seen as analogical reconstruction before Common Romanian,
by
internal reorganization of the paradigm in PBR or by reproducing the
pqpf
regular ending. So Romanian perfect doesn't reproduce Latin pqpf for
the
1st & 2nd plural (as even ancient texts still preserve etymological
forms)
and only for the 3rd person there is something to explain -- but not
the
/r/ (which is there in Latin perfect), only the final vowel (which can
be
perfectly well explained through analogy with other PBR verbal times,
not
necessarily through 3rd plural of Latin pqpf indicative, a time which
did
not really resist in PBR).

> In the context of the question, I presume 'true perfect' means a
> derivative of the PIE perfect.

... in Latin, hence in Romanian (through Latin paradigm).

Regards,
Marius Iacomi