Latin verbal system: how perfect and aorist joined in the new perfect?

[Steven A. Gustafson]

Wilmer "Xelloss" Ricciotti wrote:

> What I can't understand is the criterion which was used to choose for
> the new pefect an aorist form instead of a perfect one and vice versa.

Going mostly from memory, I doubt that any criteria can really be
pinpointed.  Latin verbs in a number of tenses show some strong signs of
being taken apart and reassembled after someone mixed up the pieces, and
lost a few along the way.  Classical Latin is a highly artificial
literary creation.

The most obvious such process occurred in the Latin 'future,' which in
some verbs was made with a verbal suffix from -bhu; and in other verbs
the "future" is really the old subjunctive, and a new subjunctive was
shoehorned in, made mostly of the old optative.

In Classical Latin, we associate the first future with the "first and
second conjugations," verbs with a stem-vowel of -a or -e:  but dialect,
vulgar, and archaic evidence shows that it was not quite so cut and
dried in the unpolished world of ordinary speech.

A Faliscan cup inscription from Praeneste goes "foied vino pipafo cra
carefo," CL "hodie vinum bibam cras carebo."  Obviously in this closely
related dialect the -bhu future attached itself to the cognate word of
-bibere-; in Latin that would yield *bibebo, which even the Pope
couldn't say with a straight face.

Similar confusion reigns during the vulgar period, where graffiti show
endless confusion between subjunctive and future endings, and forms like
-audibo- are attested.  These verb forms were apparently subject to much
confusion in the non-literary language, and had to be reorganized yet
again; even in Romance languages with relatively conservative verb
systems, like Castilian, the future has been made over.  (It is possible
that given the phonology of VL, -monebis- sounded like it was really a
contraction for -monere (h)abes-, since short 'i' and 'e' fell together
in many areas, and h- was dropped.)

Which may not directly answer your question, but the point is: the Latin
verb endings are reshuffled in a major way, in an ongoing process, in
which the attested forms of CL are but one stop along the highway.
There appears to have been a large random element when the final
decision came down as to what was considered classical.  Analogy and
shoehorning have largely refitted the grammatical options that were "in
the air" in the Latin dialects into the standardized literary language.

My vague general understanding is that the perfect personal endings
represent the old "secondary" endings, given a fairly extensive
analogical remodelling, which may have as much to do with the
substitution of -erunt for the archaic and poetical -ere in the perfect
as anything else.  (All the rest of the 3pl forms have -nt- somewhere in
there.)

I will see if I can find anything in any of my Latin books about where
the -i came from.  It may have been added by analogy in the first
place.  Again, vulgar and Romance evidence suggests that the -i- in the
-vi perfets was apparently not present on the lips of many speakers;
some Romance forms seem to require *amaut, &c., instead of CL -amavit-.

Given this further evidence of tinkering by arbiters of elegance, you
have to wonder how "reliable" the attested forms of the literary
language are on details like this.

"Truth is the successful effort to think impersonally and
inhumanly."
             --- Robt. Musil, -The Man Without Qualities-