Raising of unstressed vowels in Eng from schwa to /I/

[Cecil Ward]

----------------------------Original message----------------------------
John Hewson wrote:
> Remember AA Milne's
>       They're changing the guard at Buckingham Palice;
>       Christopher Robin went down with Alice.

This development seems to me to be patchy. I was asking myself the question
whether the spread of this phenomenon is due to analogy.

It seems to me that there are two possibilities, one is that this is a
"phonetic" development, by which I mean that the conditioning factors are
predictable from the phonetic environment. Maybe, but I don't see it.

The second possibility is that speakers have come to have certain words
stored in their lexicon as containing an //I//, whereas the "correct",
written form has a different vowel.

If this is the case, selective analogy could be responsible on a
case-by-case basis, analogy with other "legitimate" word-elements, and this
might involve "knowledge" of morphology.

Possible candidates that I might expect to find, off the top of my head,
would be things like

        pre- /pri/- vs pro- giving forms like *"pretect" (=protect),
*"previde" (=provide) by analogy with "prevent", "pretend" for example.


John's example of "palice" = "Alice" is another good candidate, because -ice
(pronounced with non-schwa) is a legitimate suffix, and so is -ace. Although
I can't think of a great number of examples right now
        -ace    Horace terrace menace solace
        -ice    lattice malice crevice avarice

For example, one might expect there to be pressure on "Eustace", because of
Doris, Phyllis, Eunice, Norris, Morris, Willis, Harris, Davies.

If forms like disagreeable or igreeable or even igree (=agree) are heard,
then I would consider that as a problem for this "analogy" theory.

John Hewson wrote:
> And there are all the plurals in [-iz], horses, boxes, matches, pages,
the 3rd sing verb forms in similar circumstances: smashes, catches,
dodges, passes, buzzes, and the past tense markers after /t,d/: waited,
boarded.

Good point. Are these forms then sources for analogical spread?

Cecil Ward.

-----Original Message-----
From: John Hewson [mailto:jhewson at morgan.ucs.mun.ca]
Sent: 18 July 2003 01:47
To: Cecil Ward
Cc: HISTLING at LISTSERV.SC.EDU

On Thu, 17 Jul 2003, Cecil Ward wrote:

(snippet)

> The phenomenon may be confined to the syllable that immediately
precedes the stress, is that correct?

Not true. Remember AA Milne's
        They're changing the guard at Buckingham Palice;
        Christopher Robin went down with Alice.

I've often heard Canada pronounced Canida by British speakers.
And there are all the plurals in [-iz], horses, boxes, matches, pages,
the 3rd sing verb forms in similar circumstances: smashes, catches,
dodges, passes, buzzes, and the past tense markers after /t,d/: waited,
boarded.

The first vowel in _disaster_ is reduced to schwa in American English, as
is the last vowel in _Latin_; is the [i] that is heard in British English
in these words a reduced vowel? The American pronunciation [latn] has a
syllabic [n] in the last syllable and no vowel. (The converse happens in
_pattern_, where BE has [patn], but in AE, where there is r-colouring, the
second vowel is clearly heard).

The question is further complicated by the fact that American schwa is
often barred [i].

****************************************************************************
***
John Hewson, FRSC                               tel: (709)737-8131
Henrietta Harvey Professor Emeritus             fax: (709)737-4000
Memorial University of Newfoundland
St. John's NF, CANADA A1B 3X9
****************************************************************************
***