Paul Johnston paul.johnston at WMICH.EDU
Mon Apr 7 16:26:12 UTC 2008

Hold it, with man and male.  It is true there are dialects out there
with [eI] for both--Muddy Waters's AAVE, for instance  (Listen to
either "I'm a Man" or "Hoochie Coochie Man", both [meI~(n)]).  But
most dialects with "tensed ash" have an INgliding diphthong starting
with something in the [e~I~E] range, or maybe, in the South, a
triphthong made up of an upgliding diphthong plus schwa; male, on the
other hand, has an UPgliding diphthong.  I can conceive of some
Southern dialects having [aeI@] for both, with the /l/ of male
generating a schwa before it (though I think of this as a Northern
feature)/  There are also dialects with [E] in male (the gel=jail
merger of the Midlands produces this).  Michael, do you really merge
the two under one of these realizations?

This is where Labov falls down in his account of the Northern Cities
Vowel Shift, with his account of raising/tensing of /ae/, and
"crossover" of /E/.  In NCVS territory, /ae/ is DIPHTHONGIZED, and
the first element raises, and /E/ centralizes but remains
monophthongal.  The same sort of crossover happened in the Great
Vowel Shift in a whole bunch of dialects in England, particularly in
the far North and far South.  Long /a:/ diphthongized to an ingliding
diphthong with a raised first element, before /E:/ did anything (it
eventually raises as a unit to [e:] or [i:]).  No crossover was
involved, since the vowel nuclei were of different types.

The way I was trained, a diphthong acts like a unit phonemically, but
has a V1 and a V2 (first and second element) that are not identical.
If V1=V2, you have a long vowel.  And, yes, I realize that the
description in  this way is something of an idealization anyway,
since in real speech, segments shade into each other.   A diphthong
ending in schwa  (or anything non-high and not likely to be analyzed
as a glide) is just as much of a diphthong as any other one.

This is a side point.  If the vowel in man=male for you, this is
probably another case of phonemic neutralization, which different
schools of phonologists would treat differently.  Generative
phonologists might well analyze it as you do.  Others might posit
archiphonemes, which is something like Tom Zurinskas seems to do,
probably without realizing it.  Others might posit lexical transfer,
depending on what the surface realization is.  I doubt it is a case
of total merger, since to my knowledge, no English dialects merge the
WHOLE CAT class with the whole MATE/BAIT class.  But it's a debatable
point in phonology anyway.

On Apr 5, 2008, at 11:04 PM, LanDi Liu wrote:

> ---------------------- Information from the mail header
> -----------------------
> Sender:       American Dialect Society <ADS-L at LISTSERV.UGA.EDU>
> Poster:       LanDi Liu <strangeguitars at GMAIL.COM>
> Subject:      Re: yahoo
> ----------------------------------------------------------------------
> ---------
> On Sun, Apr 6, 2008 at 6:29 AM, Michael Covarrubias
> <mcovarru at>
> wrote:
>> Scot LaFaive wrote:
>>> Essentially, phonemes are
>>> in your head and allophones are the actual production of phonemes in
>>> specific environments. I don't know of any other definitions for
>>> phoneme and allophone that phonologists use, so I'm not sure what
>>> you
>>> mean by "search engines give several definitions." I think I'll go
>>> with what phonologists mean by phoneme and allophone instead of
>>> Google
>>> and Yahoo.
>>> Scot
>> Something "in your head" isn't much of a definition. Phonologists
>> probably have more definitions of phoneme than Google and Yahoo
>> combined.
>> michael
>> Why not?  It's just a less technical way to say that phonemes are
> perceptual.  Allophones are the way that phonemes are physically
> expressed.
> A tapped /t/, an aspirated /t/, and an unaspirated /t/ are all the
> same
> phoneme, /t/, but are different allophones.  /t/, however, is a
> very clean
> example.  When you get into vowels, especially diphthongs, it gets
> tricky,
> and murky.  For example, some of you may think Mary, marry, and
> merry have
> different phonemes, but they're all the same for people like me who
> have
> merged them.
> And if you're American, you probably say "man" and "male" with the
> same
> vowel sound -- "man" with a tensed ash, and "male" the same way,
> but you
> also probably consider "man" to have a short a sound, and "male" to
> have a
> long a sound.  If you say them that way (and I do), you are using
> the same
> sound in two different phonemes.
> When defining an accent one must make some arbitrary decisions
> about what
> constitutes a phoneme in that accent, and what constitute
> allophones in each
> phoneme.  Usually we try to apply an objective rule to this
> process, like
> the idea that changing phonemes affects meaning, but changing
> allophones
> doesn't, but this is not always cut and dry.
> --
> Randy Alexander
> Jilin City, China
> ------------------------------------------------------------
> The American Dialect Society -

The American Dialect Society -

More information about the Ads-l mailing list