Root aorists vs. marked presents

[ECOLING at aol.com]

Concerning the long-continuing discussion of "root aorist" in Latin, etc.,
I would like to refer people to a simply wonderful article by Gene
Gragg in a volume of the Chicago Linguistics society many years ago,
about Ethiopic verb stems, where he shows a very clear case that the
choice of stems depends on the semantics of the verbs concerned,
which stem will be primary unmarked, which marked as derived.

In the discussion of root aorists and marked presents, etc.
did anyone compile this kind of semantic information?
I wonder whether I missed it in the discussion?

Did anyone compile the examples being offered for

Primary root-aorists with marked present stems
vs.
Primary present stems with marked sigmatic (etc?) aorists

to see whether the semantics of the verbs concerned
reinforces the validity of the two distinct sets of hypothesized pairings.

Use of this semantics would even permit the inclusion
in the data set of verbs which are only attested in later
languages in one stem, not in both stems in the same language.
This could greatly strengthen our ability to see whether the
hypothesized form categories really constituted paired oppositions
in early IE or pre-IE.

I think that was the point of the discussions?

There can of course also be semantic domains of verbs in which only
one of the stems would be common, we need not have only
two sets of verbs.  The question is rather whether we can
identify coherent semantic domains which predict which stem
forms were used for verbs in those domains.

We need to go beyond a purely formal approach.

Perhaps much of this has already been done?
Is it done in Rix: Lexikon der indogermanischen Verben?

(By the way, I would like to get a copy of that book in any event,
can anyone tell me an easy way?  I think I found it once
somewhere on the web, don't remember where now,
cannot seem to find it again,
could not find it recently at Amazon.com.)

Best wishes,
Lloyd Anderson