IPA Diphthongs--repost

[Herb Stahlke]

Here's the message on diphthongs I posted in Rich Format, now in Plain
Text.  Thanks to Mark and LanDi for showing me how to change that in
gmail.

I'll give it a try.  First, it might be useful to distinguish between
a diphthong and a vowel sequence.  American English "idea" has a vowel
sequence, and the last vowel, a schwa, is a separate syllable.
Southern British "dear" has virtually the same sequence of vowel
sounds, but it's one syllable, and BE has falling diphthongs ending in
/@/.  AmE r-ful varieties don't.

When I teach phonetics, I make the same statement you had heard, that
the second element of the diphthong indicates direction of tongue
movement rather than endpoint, although this is a matter of degree and
depends on things like speech tempo and just where the second element
is articulatorily.  If the second element is a schwa or an /r/, then
the tongue is more likely to get pretty close to that position, simply
because of where it is.

Whether you transcribe the second element of a falling diphthong as
/I, U/ or as /j, w/ is a matter of choice.  The difference between
non-tonic /I,U/ and final /j, w/ is functional rather than formal.
They are for all practical purposes the same sound, one used in a
vocalic position and the other in a consonant slot.  It just happens
that we can notate them differently depending on what we consider
their function to be.  This is more obvious with /r/, which has nearly
the same position in the vowel space as /@/.  We tend to think about
it as a consonant, and it's hard sometimes to convince a beginning
student that it can also be a vowel.  As a vowel it can be transcribed
as schwar or as syllabic retroflex /r/.  I think most people prefer
the schwar, but for diphthongs /r/ gets used.

I can illustrate a number of features of AmE diphthongs from my SE
Michigan variety of Inland Northern.  I have a phonemic contrast
between lowered and raised /a-/ diphthongs.  In this particular case,
the first element of the diphthong will be either the inverted lower
case print <a> or something close to /@/.  They occur in their usual
Canadian Raising environments, but they contrast in open syllables and
before /-nd/, so

/ha:I/ "elevated"
/h at I/ (greeting)
/ka:Ind/ "warm-hearted"
/k at Ind/ "variety"
/ba:U/ "front end of a boat"
/b at Uwa:U/ (sound a dog makes)
/ra:Und/ "circular, spherical"
/r at Und/ (preposition)

Notice that the raised diphthongs are shorter, as CR predicts, even
though they occur where CR would predict the lowered diphthong.

Similarly, I have

/ba:rd/ "banned" (preterit of  "bar")
/b at rd/ "poet"

I have quite a few minimal pairs like these, but what they all seem to
have in common is that the two forms always come from different
lexical categories.  Clearly it's a partial phonemic contrast since it
doesn't occur in all environments.

Back to whether to represent the second element with a vowel or glide
for a moment.   The choice depends on the point you want to make.  I
tend to use  /I/ and /U/ for those diphthongs and /r/ for the /r/
diphthongs. If I'm comparing Upper South and Lower North
pronunciations of "drive," I'll use the vowel variant so that I can
more clearly demonstrate the lowering found in the Upper South (and
Lower South) diphthong.  Central Indiana speakers have /draIv/ while
Southern Indiana speakers have /draav/, where the vowel quality is low
central, not back.  I can't show that change as clearly if I use the
consonantal option.

What makes the /r/ diphthongs distinct from vowel + /r/ codas, at
least in my speech, is that they start with vowels that I have only in
diphthongs.  /a/ (lower mid central) does not occur as a monophthong,
and open <o> also occurs only in the diphthongs /OI/ and /Or/ and
never as a monophthong.  Where many varieties of AmE have
monophthongal /O/, mine has the low back rounded vowel.

Herb

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org