free variation in pronunciation

[Douglas Bigham]

Given this pesky business of free variation, has anyone used distinctions between using

A.) variant x
B.) (possible) free variation between x & Y
c.) variant Y

in dialect work?

For example, in my use (born & raised Southern IL) "drug" and "dragged" are both past of "drag".  I know one marks Appalachian English (supposedly) and the other is standard, but I can't recall which is which.  So, for me the two are in free variation (so free that I avoid them because I can't remember which is 'wrong').

This free variation would seem to coincide with the notion that this area is viewed as both North and South speech depending on whom you ask.  So. . . Has this kind of thing ever been used to determine geographic dialect boundaries?

-dsb