"The extra twist she [Mari, me] adds is to do away with traditional temporal
limits on the tense in the indicative, and as a fellow programmer of
sorts I can appreciate the appeal this has in seeming to provide a
more consistent way to deal with problem passages."
The only temporal limitation on the indicative verb forms that I do
away with is with respect to the aorist and the present. This does
not have the force of a computational 'quick and dirty fix'. On the
contrary, my goal was to provide a monotonically increasing model such
that information that was provided could NOT be eliminated (the other
forms DO carry temporal tense significance which may not be removed),
but information could be added (i.e. temporal significance of the
aorist can be 'added' by pragmatic implicature associated with a
perfective unmarked for tense, but nothing can make the future refer
to something past with respect to the deictic center of the discourse
(generally speech time)).
On Andrew K's observations about the present, NONE of the 60+ langauges in
O"sten Dahl's survey of tense/aspect systems marked present tense on a
verb alone. THat is, all the morphemes associated with present tense
are doing other work (in English, marking person: He work(s), in
Greek marking imperfective aspect), and the association of present
time with these forms comes as a matter of pragmatic implicature:
since the writer/speaker didn't use a form marked for past or future,
(s)he must have meant to refer to the present. THis implicature can
be cancelled in the appropriate context, cf. the historical presents
in Mark, as well as the English (relative) counterpart, as in the
following true statement, where yesterday marks it clearly as past.
Note that a similar implicature is available even without 'yesterday'.
I get home yesterday and find out my daughter has thrown up.
Daughter #1 also threw up, at 3 a.m.....