Dialect Mixing

[Matthew Gordon]

In reading various studies of the dialects of the Midwest and West, I've
been struck by the tendency to frame the analysis almost exclusively in
terms of the traditional dialect divisions of the East Coast: Northern,
Midland, and Southern. Several studies develop complicated formulas to
calculate the relative influence of these dialects on the area under
investigation. It's easy to understand the motivation for this approach
given the emphasis on settlement history in traditional dialect
geography. Still, this focus on historical retentions has resulted in
dialectologists missing some important new features in these areas
(e.g., The Northern Cities Shift).

Does anyone know of an explicit defense of this approach as a method of
studying American dialects off the Atlantic Coast? Did, e.g., Kurath
argue that all American dialects can be adequately described with
reference to his "original" divisions?

Are there parallels in the study of dialects in other countries? For
example, do studies of Hiberno-English count the number of Midland or
Northern English forms?