"control" in functionalist perspective and related questions

[E.G.]

Hi all,

I've been reading, with much interest and to the detriment of
dissertation-writing, the archives of this list. I want to pose a few
questions which may have been asked and answered before, but I'd be grateful
for any comments and/or references.

Has anyone written a full-fledged functional analysis of "control" phenomena
for any language or in typological perspective? It seems like all the pieces
are there, from Givón's work on event integration and clause union,
Haspelmath's frequentist explanations, and so on. There are two things that
I wonder about especially:

* To what does "control shift" or the fact that for a given verb, one can
find both "subject control" and "object control", correlate in functional
terms, cross-linguistically? Generative treatments seem to be silent on
this.

* With verbs like "beg (someone to do something)," would the "subject
control" construction ("He begged me to join my team") be interpreted as
causative, i.e., "to (let him) join my team"? And if so, would it be
considered syntactically/actantially more complex than the "object control"
construction ("He begged me to go with him")? Would the syntactic/actantial
status of "me" be the same in both constructions?

I'm interested in functional approaches that prefer "implicit argument" to
"gap" or "zero" analyses, and how they would handle this sort of thing. I
tend to think that if one applies Givón's work on manipulation verbs to
this, one might arrive at concepts like "internal" vs. "external" manipulee
(from the point of view of the target event/complement clause), the former
for the "object control" construction, the latter for the "subject control"
construction. This would reflect whether or not the manipulee is
coreferential with the agent of the target event or not.

* Has anyone tried to continue  Givón's cross-linguistic studies of the *
functional* aspects of formal oppositions between, e.g., more- and
less-finite complements? What about between equally-finite constructions
(e.g., two "infinitives" or two "subjunctives" that are found with the same
verb lexeme?).

* Has anyone offered a functional explanation for why languages might have a
diachronically stable system in which one has both overt- and
non-overt-subject complement constructions for verbs like "want" (see, for
example, Haspelmath's article on "want" in WALS)? For example, a language
with both "I want to go" and "I want that-I-go/for-me-to-go." If the answer
is that there's a communicative interest in maintaining a difference in
function, then what does one do with the fact that in other languages the
diachrony played out differently, and either the overt-subject or the
non-overt-subject construction was lost?

There are some interesting treatments of this in the framework of Functional
Grammar (e.g., Dik and Bolkestein), but they tend to replace formalism with
formalism.

I'm working on this problem in a language (Coptic), where the notion of
"control" doesn't seem useful, since for the same verb one often finds both
subject and object "control." Moreover, both more- and less-finite
constructions are almost always found in the same environments, and here I
don't refer just to typical "control" environments. Roughly, "I am ready to
go/I am ready that-I-go," "He cannot go/He cannot that-he-go," "I want to
go/I want that-I-go", etc., so there is no complementary distribution of the
sort often discussed. It seems highly unlikely that this is "just"
variation; rather, it looks like different functions are encoded by the
different constructions.  Interestingly and perhaps unsurprisingly, the
diachrony of the various complement clause constructions is a pretty good
diagnostic for the functional differences found.

I apologize if the questions are half-baked or naive. I would be very
grateful for any comments, and will gladly post a summary.

Eitan