🚀 go-pugleaf

I agree completely with David's caveats about elicitation, and share
his experiences with aspectual data. I had to teach 2nd year Russian
a few years back and discovered that I was pretty clueless myself,
and the books aren't much help.

I am glad this discussion has brought to light a wider range of examples.
I am collecting some of them and awaiting the return of the native speakers
here in the fall. I don't trust my own intuitions on any of them.

As to the null modal and treatment of the copula in general, I hope to
have something to say eventually, since I am busy dealing with English
modal and auxiliary system at present, and have only recently delved into
the GB literature. As a class (if indeed there is a coherent one), modals
seem to show unified semantics (as operators), a variety of syntactic
characteristics (in English: verbs, adverbs, adjectives, with the verbs
requiring a variety of subcategorization frames), and in some, but not all,
cases, morphological properties.

I suppose that under the Minimalist Program a modal should be base-generated
under T(ense) and undergo SPELLOUT early in the derivation, to insure that
it gets tensed and is the leftmost member of the string of verbs. Under this
proposal (suggestion, thought? not analysis -- yet) the Russian case is 
interesting because if a past tense were lacking at some part of the paradigm
one would wonder why. This musing is prompted by the following passage in
DP's post:

"...Hypothesis: there is a null all-purpose modal somewhere in these sentences
...is there a past tense of any ofa these constructions?"

Eric Schiller