--- In
cybalist@yahoogroups.com, "P&G" <petegray@...> wrote:
> But can you find any example where the imperfect subjunctive
stands in
> opposition to the prsent, and carries a past time reference? In
all the
> examples I can think of, the so-called "imperfect" subjunctive
carries a
> present reference:
> Utinam adesset!
> Si adessem, ....
> I think we are misled by the traditional label into thinking there
is some
> connection between the "imperfect" subjunctive and the imperfect
tense. But
> even in formation there is not.
Some may have been misled, I was not. I was very much aware of the
danger, so I checked up on the matter. And I found that the
imperfect subjunctive is indeed so often the past-tense replacement
of the present subjunctive (and the pluperfect subjunctive, of the
perfect subjunctive) that it fully justifies the traditional
denomination. Still, as always, a *term* cannot capture all
characteristics of what the term is used to refer to, so much in the
functions of the ipf.sbj. is not covered by its traditional name.
If the past-tense counterpart of the present indicative is called
the imperfect indicative, then I find it quite wise to call the past-
tense counterpart of the present subjunctive the imperfect
subjunctive. And it *is* the past-tense replacment of the present
subjunctive: Take an indirect question: scio quid faciat "I know
what he is doing (prs.sbj.)" : sciebam quid faceret "I knew what he
was doing (ipf.sbj.)". In absolutely parallel fashion, the perfect
offers scio quid fecerit "I know what he has done (pf.sbj.)" :
sciebam quid fecisset "I know what he had done (ppf.sbj.)".
Now that had nothing do to with its form, but the function is found
to be fine for whoever wants to call it the past of the subjunctive.
There certainly was a good reason to do so.
Now you say the form is against this. Well, that is preceisely the
reason I came up with my analysis which takes care of that problem.
If the stem-formation of a subjunctive is given a past-tense
marking, as it was in my focal point, 2sg sbj. *-s-esi => 2sg. past
sbj. *-s-e:s (copying what must have been at one time 2sg prs.
*esi : 2sg ipf. *e:s), then you've got it.
>
> >The Latin imperfect subjunctive is what the present subjunctive is
> >transformed into if the overall setting is shifted to the
preterite.
>
> Yes. But the subjunctive built from the infinitive marks action
> contemporaneous with, or anterior to, the main verb, depending
only on the
> choice of infinitive (or stem). Any time reference is carried in
the main
> verb.
If time is immaterial, then why does the subjunctive follow suit
when the tense of the main verb changes? Surely you wouldn't say
that, in the sentences quoted, the relation between faciat and
faceret is different from the one between scio and sciebam? Well, in
a way one might say that, for you get faceret instead of faciat, not
only if you say sciebam instead of scio, but also if you take a
different past form, such as the perfect scivi or even the
pluperfect sciveram. This means that the "ipf.sbj." is just some
*past* representative of the subjunctive, not necessarily exactly an
imperfect. Still, the imperfect is the past with the least amount of
marking, so it should be more obvious than any of the others, but we
could conceivably have surprises. It is my sugggestion that there is
no surprise in this, but that the form was in fact created as the
imperfect-like variant in past-tense use of a present-like form in
non-past use.
> >The IE basis of this is twofold: preterite and optative.
>
> I thought we agreed the formation was internal to Latin.
Yes, internal to a Latin language stemming from Indo-European. The
task is to specify how the ipf.sbj., form and function, can grow out
of the Indo-European background of Latin. New grammatical categories
are not made from scratch. I *looked* that way, and that was the
problem I wanted to solve.
>
> >I still do not insist upon it, for it is based on the second
person:
> >*weg^h-s-esi rhyming with *esi, therefore ipf.sbj. in *-se:s
rhyming with
> >*e:s. On top of this it works with the augment, which however
scares me
> >less.
>
> The augment in Latin scares me more, but then I'm rather
conservative.
Now the thing is, as I have seen now, you get an explanation of the
other half of the functional range of the ipf.sbj., the expression
of unreality ("irrealis"), precisely if you put in the augment at an
earlier stage. It must have been an augmented imperfect that was
replaced by the new creation of a past subjunctive, for that is what
we can find material to reconstruct. That being so, I see even less
cause for alarm if the augment is appealed to in the form also.
Jens