Date:  Thu, 7 Sep 1995 14:52:58 EDT

From:  Douglas Bayer x3701 3NW <dbayer[AT SYMBOL GOES HERE]YUKON.HQ.ILEAF.COM>

Subject: /a:I/ vs. /[AT SYMBOL GOES HERE]I/



I'd like any information about the geographical distribution

of the /[AT SYMBOL GOES HERE]I/ allophone of AY, and similar centering/raising or

shortening of other vowels and diphthongs before voiceless

consonants.



Could anyone point me to papers, maybe even URLs on the 'Net?



dbayer[AT SYMBOL GOES HERE]ileaf.com



---------- LONG --------------



I've heard that distinction in the following pairs described

as "Northern" or even "chiefly Canadian." And I haven't seen

or heard it implemented in any speech generation systems:



     /a:I/ = /[AT SYMBOL GOES HERE]I/  (AAH-ey vs. UH-ey)

      eyes = ice

     rider = writer

 high/hind/hide = height

 why/wine/wide = white



Yet it's obligatory for me. Isn't it "General American"?



It's the sole distinguishing feature in the pair

rider = writer, and often even in eyes = ice

(because the voicing of the final "s" can't always be heard).



The rule seems to be long, open /a:I/ when the vowel is

followed by no consonant, a voiced consonant, or a nasal;

and short, mid /[AT SYMBOL GOES HERE]I/ before a voiceless consonant.



But in the Great Lakes region, even "-ire" can go to /[AT SYMBOL GOES HERE]jR/.

This seems less wide-spread. (I'm acutely aware because I

pronounce my German name /ba:jR/ but many people say /b[AT SYMBOL GOES HERE]jR/)



Does similar centering/raising occur with other vowels or

diphthongs? For example, is there a region where "OH"

differentially goes to /o:w/ or /[AT SYMBOL GOES HERE]w/ in the same environments?



Any pointers appreciated.

(And I'll gladly accept uuencoded or .ps.Z files by email... :)



Doug Bayer, Techdoc Toolsmith

Interleaf Inc.