Back Vowel Phonemes

[Matthew Gordon]

I agree with Dennis' suggestion if you're looking at potential merger. But, also
interesting in the Upper Midwest is the westward expansion of the Northern Cities
Shift. Category assignment is crucial in interpreting the progress of a given
speaker's participation. For example, if someone produces a low back unrounded
vowel in 'water', you can't tell whether it's an innovative (i.e. NCShifted)
pronunciation of the 'open o' phoneme or a conservative pronunciation of the /a/
class.

If you're dealing with speakers from a well-defined geographical area, I would
suggest doing some polling to try to determine how these words are classified
locally. You can ask people (other than your subjects) directly whether, e.g., the
vowel in 'water' is more like the vowel in 'cot' or the vowel in 'caught' (assuming
you've already established they have a distinction in this pair).

You may still run into problems with "idiolectal" variation. In Michigan, for
example, I found differences within the same small town in the assignment of the
stressed vowel in "Chicago".

In the North, I'd expect to find the following assignments:
dog = )
lost=)
horror=)
on=a
upon=a
water=)
wash=)

"Dennis R. Preston" wrote:

> Terry,
>
> If you are assuming a Labovian paradigm, then the "historical" class of the
> vowel gets first consideration. As the two merge, however, I wonder if you
> are not overdoing the necessity of assignment? Perhaps if you took all
> tokens of both (remembering environmental influences of course) and showed
> their "different" disptribution in younger speakers (opposed to old South
> Midlanders like me [Louisville] who have them quite distinct), wouldn't
> youn have done the task? (Remember, Labov also said (something) like "Once
> you know how to code your data, you've solved the prpblem.")
>
> To be more direct, hwoever, an old-time Kentuckian like me (without a hint
> of merger) has them as:
>
> dog=)
> lost=)
> horror=)
> on=)
> upon=)
> water=a
> wash=) (with "intrusive r")
> wash = a (without "intrusive r")
>
> Best,
>
> dInIs (who recently found that his vowels [with exception such as E-I
> conflation before nasals] match Peterson and Barney nearly perfectly)
>
> PS: Idolectal class-swapping is, however, common, even in those with clean
> separation. I seem to have a rule which says /a/-/)/ before 'g' is /)/ if I
> learned the word as a kid (dog, hog, log, frog, ...) but /a/ (or variably
> /)/) for "learned" words (togs, cog, bog, smog, ...). (I know, some of the
> words are only "learned" for guys from Louisville; don't write in!).
>
> >I am currently working on a project that involves acoustic analysis of
> >low back vowels in a regional variety of English.  As I move toward
> >plotting of F1/F2 values, using a Labovian approach, I need to make some
> >decisions regarding how to treat a subset of my data phonemically.
> >
> >All of the tokens vary, of course, but I have been able to place several
> >items clearly into the ) category phonemically and the a category
> >phonemically.
> >
> >The following set also all vary (which is indicated in major
> >dictionaries, such as AHDE Edition 1):
> >
> >dog, lost, horror, on, upon, water, wash
> >
> >My query is, what vowel would you consider to be the underlying phoneme
> >in each case, ) or a?  (I know, I know, choice of phoneme may depend upon
> >idiolect, but remember that I am assuming a Labovian paradigm to
> >facilitate one approach to studying change.)
> >
> >
> >
> >
> >Virtually, Terry
> >(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)
> >Terry Lynn Irons        t.irons at morehead-st.edu
> >Voice Mail:             (606) 783-5164
> >Snail Mail:             UPO 604 Morehead, KY 40351
> >(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)=(*)
>
> Dennis R. Preston
> Department of Linguistics and Languages
> Michigan State University
> East Lansing MI 48824-1027 USA
> preston at pilot.msu.edu
> Office: (517)353-0740
> Fax: (517)432-2736