Re: USAGE: <o> in spoke; gotten [was: Re: Hear Me! Hear Me!]

On Tue, 2002-06-25 at 02:47, Thomas R. Wier wrote:
> N.B. Three responses here:
>
> ==============================================================
> Quoting Adrian Morgan <morg0072@...>:
>
> > This reminds me of a certain distinction between some [dia|idio]lects:
> >
> > For me, there is a rule that the "o" in "spoke":
> >  - cannot precede an /l/ in the same syllable,
> >  - but *can* precede an /l/ that is the start of the *next* syllable.
> >
> > This can be contrasted with dialects in which either the first or
> > second half of this rule is void.
>
> Interesting. I suspect that it's not [l] that's the conditioning
> environment, so much as how heavy the syllable should optimally
> be.  An [@u] is bimoraic, while [@ul] is trimoraic.  Extraheavy
> syllables like the latter are marked structures, and there is a
> tendency to fix them somehow.  In your dialect, this involves
> velarization and assimilation of the [l] to the preceding [u].
> In many dialects of the American South, /o/ before liquids like
> /l/ and /r/ is monophthongized to [o] (or perhaps with only
> very slight diphthongization).  Words, like "old", though, still
> suffer from the same problem, so there the tendency is to delete
> the final consonant to become [ol].
The thing is, at least in my speech which Adrian claims follows the
first rule from my explanations of it, (a) it only happens before /l/ to
my knowledge and (b) it remains a diphthong-length (I guess bi-moraic,
but it feels more like it's one-and-a-half mora long).

> > Because of the second half, the diphthongs in the first syllables of
> > "solo" and "polio" are for me the same "o" in "spoke". However, in a
> > common alternative idiolect (i.e. common in Australia) this second half
> > of the rule is void and the vowel [O] is used instead.
>
> What is the quality of the /l/ in "polar" and "polling"? For
> me, they're different:  the first is alveolar, the second is
> velar.  (This may constitute a minimal pair if we assume underlyingly
> there is only one /l/ in each.  Of course, it may be that the
> verb "poll" actually has underlyingly two /l/s.)
For me, the /l/ in 'polar' is the velar (dark) one, in 'polling', it's
alveolar. I imagine the presence of a /u/ in 'polar' is what darkens it.
(In /pQl/ 'poll', it's stil alveolar).

> =====================================================================
> Quoting Tristan McLeay <kesuari@...>:
>
> > > I'm never sure about [E]. It's defined in my head as "halfway between
> > > [e] and [{]", and I think you hear a lot of teenage girls using it in
> > > place of [e] sometimes these days. But I'm pretty sure it doesn't
> > > exist in my idiolect.
> >
> > I have been led to understand that the vowel in the word 'get' is /E/.
> > It may well be that it is higher than [E], especially in Adelaide
>
> It certainly is in my (Texan) dialect.  <get> is flagrantly [gIt],
> no ifs or buts about it.  (Which is odd, since original /E/ doesn't
> usually shift to /I/ unless followed by a nasal. <let> is definitely
> [lEt], for example.)
I was actually talking specifically about Australian English, there;
(some) New Zullenders and (some) South Effrucans have it consistently
higher, perceived by Aussies as /I/ (thus: 'sex' is heard as 'six' which
is heard as 'sex' (it's /s@ks/), and makes for puns in poor taste).

Tristan