On Mon, 18 Nov 1996 13:37:30 -0800 Larry Beason said:
>Burkhard,
>                                                            But I think the
>concept of "quasi-modals" is a legitimate category even if--like all other
>categories of grammar--not leak-proof.  A quasi-modal is (to take a stab at
>an informal definition) a group of words, normally two, that function as a
>auxillary verb.
 
The key part of this this definition is the term "function."  My impression is
that "quasi-modals" are a category because they perform some of the semantic
functions of true modals: "have to" has a meaning very similar to "must" and
"be able to" has a meaning very similar to "can."
 
I think Burkhard makes a very good point that syntactically the quasi-
modals have none of the characteristics of true modals.  This is very easy
to demonstrate.  True modals, I believe, have inherent tense.  Thus, they
do not take the agreement -s (a tense marker) and they never occur in
non-finite constructions.
 
     1) I want to be able to explain grammar better to my students.
     2) *I want to can explain grammar better to my students.
 
     3) Having to leave early, the student sat by the door.
     4) *Musting to leave early, the student sat by the door.
 
Bob Yates, Central Missouri State University, [log in to unmask]