California research

Your Name Dialectap at AOL.COM
Tue Nov 26 17:22:11 UTC 2002

Yes, the /ae/ is backed by speakers with the cot/caught merger and the /E/ is
lowered by speakers with /ae/ backing.  It seems to me that this could very
well be a chain-shifty response to the merger of the cot/caught vowels.

My data show that the cot/caught merger is about 2 (20-yr.) generations ahead
of the new /ae/ and /E/ shifts.  And, even then, it is female speakers who
are at the leading edge with the /ae/ and /E/ innovations.  This is what we
would expect, of course.

Naturally, the /ae/ and /E/ shifts are not wholesale and even the youngest
speakers aren't consistent with them.  For example, the ______/s/ environment
is a strong predictor of a lowered /E/ variant.  (DARE Volume 1 mentions
this.)   A typical informant in the youngest group might have a lowered /E/
in "guest" "reception" and "best" but later on in the recording a /E/ in the
traditional position for "guest" "reception" and "recipe".

"Don" and "Dawn" were side-by-side throughout the reading passage because
this was the name of the protagonists in the story.  Many of the informants
said that it was strange to have a bride and groom with the same name!  I did
not get any perception data other than those reactions, however.  In fact, in
an effort to get the most "natural" reading style possible, I didn't put
other obvious cot/caught minimal pairs next to each other in the passage.
The passage is full of /aw/ and /ah/ items, not necessarily minimal pairs.  I
hope that that isn't a fatal flaw.  My oldest informant, age 77, had a
traditional vowel for "talk"-type words.  Other than that, she had the

I didn't mention everything in my memo to Allan.  I also tested for
"morning"/"mourning"-type words--all merged.
The "shouldEn't" and "hiddEn" innovation hasn't lost any steam since I
reported onit in 1999.
"want" and "watch", etc. have /ah/.

Did this answer some of your questions?  Thank you very much for your
comments and for your interest.


More information about the Ads-l mailing list