double modals (was: fronted /oU/)

[Bethany K. Dumas]

Some observations from a lightly-edited paper from the past (from a
native double modal speaker):

I have mentioned that studies of Southern Mt.. English based on attested
forms have all concluded that double modal constructions are rare. Early
participant observers (Carr 1905, Randolph & Wilson 1953) reached the same
conclusion. However, the intuitions and reported experiences of recent
participant observers (ELgin, Dumas, Bailey and Montgomery) are
otherwise. Montgomery and I, in particular, agree that multiple modal
occurrence is at least occasional except when a tape recorder is running.

I am convinced on the basis of my participant observer experience in
northwest Arkansas and East Tennessee that the rarity of multiple modal
construction in Linguistic Atlas and sociolinguistic interviews is an
artifact of the nature and structure of such interviews, not of the
frequency of occurrence of the constructions in naturally occurring
conversations. Atlas interviews are highly item-oriented; sociolinguistic
interviews are heavily narrative-oriented. Multiple modals do not
naturally occur with frequency in item-oriented or narrative-oriented
conversational interaction. As Rickford 1973 points out, some
stigmatized forms are rare in sociolinguistic interviews "because the
semantic conditions which they are normally introduced to express may
occur rarely, if at all, in the course of a sociolinguistic
interview" (p. 163).  (See also Butters 1973.)

Rather, they occur in such discourse environments as the following:

A. Possibility, with respect to past action:
   10. We might could have gone if he had gotten tickets. (#20 on the handout)
B. Possibility, with respect to present state:
   11. They may still could be all right U. T. fans.  (#30 on the handout)
C. Possibility, with respect to future action:
   12. I might could go.  (#23 on the handout)
D. Mitigation of a directive:
   13. Might could be you'd want to take a look out your east window
        toward my rose bushes. (#13 on the handout)

A number of interesting questions have been raised by the
literature. Briefly, they are these:
1. What base constructions are used?
2. How do the constructions vary with respect to provenance, frequency,
        and acceptability?
3. How are tags and questions formed?
4. What is their structure? I.e., is the first element a modal or an
        adverb?
5. What is their significance?  I.e., are multiple modal constructions
        anything more than lexical variants?

We have in the literature partial, sometimes conflicting answers to all
these questions. None of the answers are based upon a large corpus of
attested forms. The most interesting question of all seems to me to be the
following: How have these constructions spread in English and how is it
that their greatest frequency today appears to be in Southern and
AAVE varieties?

Historically, the double modals appear to have developed from Middle
English. Many constructions are reported from Scottish [?] and other
varieties of Northern British English. The settlement history of the
American south suggests, as Guy Bailey has pointed out to me recently,
that the construction spread from South Midland into Southern English,
possibly thence into AAVE and Atlantic creoles.

If this is true, than it is mysterious that many more attested examples
are reported for the lower south and the creoles, including Black English.

My tentative hypothesis for the present state of affairs is this: As
double modals developed in early Modern English, they grammaticalized and
developed a purely lexical function; i.e., they served as lexical variants
of phrases like "May be I can" or "I might be able to."  They retained
that lexical function, but as they spread into the lower south, they came
to function as politeness markers more than as simple lexical variants.

Bethany