free variation in pronunciation

[Buchmann]

First of all there's the OLD joke:
A linguist asked a Scot if he used "eether" or "eyether."
The Scot replied "Ehther will doo."

THEN, there's my family usage
[over the last half century anyway] :
"Eether A or B is OK."
"You may use eyether A or B (but nothing else)."

"Neether product is very good."
"Neyether man can be trusted."

In MY usage the VARIATION IS NOT FREE.

    [ There is also a tendency to contrast with
similar vowels elsewhere in a construction.]
THIS might lead to an impression of free variation.

I have found the pron. to be CLASS, LOCALITY, and USAGE  specific.
I have NEVER heard free variation in the U.S.    I am too far removed in
time and space from today's British to have even an inkling. I do not
remember ever hearing "eyether" used by an American ( or American
character )  on TV -- I think such usage would have startled me into
attention, but some of us do use it naturally and without affectation.
[ Though "eyether" has, to a great extent, been presented as an
affectation in American movies.]

Lynne Murphy wrote:

> _Either_ and _neither_ have two pronunciations each, and as far as I can
> tell they are in absolutely free variation within certain speech
> communities and even speakers.  Sometimes I've heard the claim that one or
> the other pronunciation is more emphatic, but I've seen little empirical
> evidence to back this up.
>
> A few months ago here we discussed some other variant pronunciations and
> spellings (e.g, vase, theatre/theater, gray/grey), and came up with
> semantic distinctions that are (believed to be) made amongst the two
> variants.  (Thus they aren't free variants.)  Other variants are not 'free'
> because they mark regional, social, formal (etc.) distinctions.  Are there
> any other examples of absolutely free variants in lexical pronunciations?
> By my definition of "absolutely free", they must be equally acceptable in
> all semantic, social, (in)formal, stylistic (etc.) contexts for a single
> speaker/speech community.
>
> Thanks,
> Lynne
>
> M Lynne Murphy
> Lecturer in Linguistics
> School of Cognitive and Computing Sciences
> University of Sussex
> Brighton BN1 9QH
> UK
>
> phone +44-(0)1273-678844
> fax   +44-(0)1273-671320