The past was opposed to the injunctive (the unmarked "default form") in IE
by the presence of the augment in the past tense. When the injunctive grew
out of fashion, the augment became superfluous, so the short form with
only secondary endings became unambiguously past. That is the case in
Homeric Greek (where the augment is optional) and in Anatolian (where the
augment is lost). Gathic Avestan is in a state of transition, in that the
augment is rarely used - but still recorded a few times, so that the
unmarked form is, at the same time, the normal form of the past and also
the default form that may be used when temporal or modal specifications
are clear from the situation or not relevant. This may be an uncommon
turn of events, but the history is clear and fully understood. Perhaps
the history of Indo-European is just so much better known than that of
other language families that we are in a position to prove also odd
stories which would not be accepted in other fields for lack of plain
evidence.
Jens
On Thu, 17 Apr 2003, P&G wrote:
> Here is a cross-posting from the Indic list. Someone may have a helpful
> response to make. Peter
>
> A curious thing about Hittite and the RV injunctive is that in Hittite,
> secondary endings are said to function as marking a preterite.
> I can't see how the RV injunctive functions develop out of a
> preterite.
>
> Also, according to Bybee et al "The evolution of grammar", zero past
> vs marked present is unknown in contemporary languages. This makes
> Hittite seem quite strange. There have been suggestions that
> sentence particles in Hittite had an aspectual role. I wonder
> what the current thinking is.
>
> Nath
>