The duration of /ae/ and /ai/

Tom Zurinskas truespel at HOTMAIL.COM
Sat Feb 2 01:08:39 UTC 2008

If an NC speaker said "ah" instead of "eye" for a stimulus word that you'd expect to be "eye" /aI/ as in "bide", they are saying two different phonemes.  If you said both were "eye" /ai/ it does not seem right.  I would think you need to specify that the phonemes were spoken correctly.  It may not be possible to get "eye" out of Southerners who substitude "ah" for "eye" consistently.  So you may be comparing "words" not "phonemes".

Also a Southerner may say ~bayud for the word "bad".  This adds two different phonemes and makes /ae/ longer.

Tom Zurinskas, USA - CT20, TN3, NJ33, FL5+
See - and the 4 truespel books plus "Occasional Poems" at

> Date: Mon, 21 Jan 2008 11:00:16 -0600
> From: jsalmons at WISC.EDU
> Subject: The duration of /ae/ and /ai/
> ---------------------- Information from the mail header -----------------------
> Sender: American Dialect Society
> Poster: Joseph Salmons
> Subject: The duration of /ae/ and /ai/
> -------------------------------------------------------------------------------
> The article in question is this:
> Jacewicz, Ewa, Joseph Salmons & Robert Allen Fox. 2007. Vowel duration
> across three American dialects. American Speech 82.367-385.
> The main finding is that, in our sample, vowels produced by speakers
> from westernmost North Carolina are significantly longer than those of
> speakers from southern/eastern Wisconsin (Madison and points east and
> somewhat northeast -- basically, the western edge of the Northern
> Cities area), while central Ohio (the Columbus area) takes an
> intermediate position.
> We did indeed find that /ae/, the vowel in 'bat' and 'bad', was longer
> than /aI/, the vowel in 'bite' and 'bide', in looking across the whole
> set of dialects. I don't have the numbers in front of me right now,
> but I think /ae/ and /aI/ were pretty comparable in duration for NC
> speakers, but /ae/ was longer for Ohio and Wisconsin speakers.
> Our NC speakers had pretty monophthongal /aI/, as you'd expect, but it
> is surely not a monophthong in Wisconsin. It's important to note
> that /ae/ is really diphthongal, especially in Wisconsin. Tom Purnell
> is doing really nice work finding the same kind of pattern across
> Wisconsin and into Minnesota. If you plot Upper Midwestern /ae/ in the
> F1/F2 space over multiple points of its duration, in fact, you'll see
> a striking U shape for many speakers. I guess the traditional reliance
> on single measurements at the mid-point of the vowel or even two
> measurements has tended to obscure just how much movement there is in
> this vowel.
> Joe
> On Jan 20, 2008, at 10:13 PM, James Harbeck wrote:
>> ---------------------- Information from the mail header
>> -----------------------
>> Sender: American Dialect Society
>> Poster: James Harbeck
>> Subject: Re: which would take longer to say /ae/ or /ai/
>> -------------------------------------------------------------------------------
>> I would think that in those dialects where the I
>> in /aI/ is elided, the tendency would be to
>> lengthen it in order to preserve distinctiveness
>> (those who are more used to Southern US speech
>> can comment on whether this seems actually to be
>> the case). I don't have the current issue of AD
>> (this reminds me: I need to go do some business
>> at the ADS website) -- is there an indication of
>> the sample base? It would be useful to break it
>> out by rendering of the diphthong, or at least
>> geographically.
>> Is any suggestion made in the article to account
>> for the relative times -- tongue movement time,
>> for instance, or typical context, or need for
>> distinction (in some Northern dialects now the
>> area around [æ] is getting crowded, I think)?
>> James Harbeck.
>> ------------------------------------------------------------
>> The American Dialect Society -
> ------------------------------------------------------------
> The American Dialect Society -

Shed those extra pounds with MSN and The Biggest Loser!

The American Dialect Society -

More information about the Ads-l mailing list