Date: Fri, 29 Mar 1996 21:37:41 -0600

From: "Timothy C. Frazer" mftcf[AT SYMBOL GOES HERE]UXA.ECN.BGU.EDU

Subject: Re: dialect areas and the latest American Speech

On Fri, 29 Mar 1996, Baden Hughes wrote:

Noted your AS message. Is there any way of getting hold of a copy of What

determines a dialect area? by Lawrence M. Davis and Charles L. Houck as

appearing in this issue ?

I just read this article and it's very provocative. Larry and Charles

raise some serious questions about whether or not we can truly draw

dialect boundaries in any way that is statistically valid. I'd like to

throw several corollary questions out to this list.

First, granting for the sake of discussion that Larry and Charles are right,

which I can't judge, not being a math person, their article

talks about boundaries in the USA based upon LInguistic Atlas type data.

Are there dialect boundaries ANYWHERE that would be valid by these same

tests? In other words, if dialect boundaries are indeed more the

product of subjective perception than a relfelection of objective

realities, is this simply a something that is true for USA English, or is

it true everywhere?

Second, much of the discussion in this article is based on lexicon.

Instances of "dialect mixture" involve people who manifest two

regionalisms for the same object, like "frying pan" and "skillet."

Are these problems confined entirely to lexical geographies, or can we

make the same generalizations about pronunciation and morphology or syntax?

Sociolinguists in the Labovian tradition identify northern and southern

vowels shifts, along with a Midland area which participates in neither.

What is the relationship between "dialect mixture" and these phonological


Third, I would think if Larry and Charles' conclusions were valid, they

would apply to other forms of human geography besides dialectology. Is

it just as tricky--or downright impossible--when we try to map other

kinds of behavior? I'd be interested to see some geography department

types get into this debate.

Fourth, a comment rather than a question. Certainly Davis and Houck have

a point about regional differences -- at least, those of the sort we

measure in things like Kurath 1949 -- being harder to demonstrate west of

the Missississippi. Wolfram's 1991 map (based on Labov's work) show a

narrow Midland band in the east spreading everywhere as we approaches the

Rockies (from the east, that is). I suspect one reason for this as that

the the two non-Midland cultural hearths, the North and South, didn't

keep their identity farther west. Cotton farming--well, there was some

in Texas, but not in Oklahoma or New Mexico, so plantation culture and

"plantation southern" kinda peter out as you go west. (And you don't get

the concentrations of AFrican Americans whose dialect might contribute a

lot to the distincition between Plantation Southern and "South Midalnd").

In the north, Yankee culture gets more and more diluted as you go west.

You still find those little villages that look like they were transplanted

from New England in southereastern Wisconsin, but I suspect not so much

in North Dakota (although they turn up again in n. California and

farther NW). And Larry and Charles are right when they point out that

Inland Northern is not exactly uniform, although I think it probably has

a common set of features.

Fifth--if we give up on "dialect areas" or "dialect boundaries"--not that

I am going to, but if--does that still mean that there are not certain

features whose variation is at least in part regional?

Well, enough rambling for now.

Tim Frazer