a minimalist phonology
From: | Roger Mills <romilly@...> |
Date: | Tuesday, April 24, 2001, 22:34 |
DAMN! (IN 50-POINT TYPE). I knew this went out in UTF7 the minute I hit
"send". My fault. Let's see if I can correct it:
>This is the shape of the vowel "triangle" as I learned it years ago (and I
>really MUST dig into storage and find that old xeroxed text!):
_____________
\ |
\ |
\ |
\ |
\_________ |
>Further, it is divided into a grid that follows the shape, so that cells in
>the top row, e.g., are wider than those in the bottom row-- capturing the
>intuition that the low vowels are closer together (more difficult to
>differentiate). The vertical/back axis shows that for back vowels, the
>tongue moves from high to low in a more or less straight line, while for
>front vowels, going lower also involves retraction of the tongue.
>I forget how many vertical columns (front to back) there are-- certainly
>more than 3; horizontal (high to low) probably 6 if not more. So we have
>roughly:
>
>i/y ? i- ? M/u
> I ? ? ? U
> e/+APY- ? 3 ? o (and unrounded o)
> E ?+AEA-? O
> æ ?V? A
> ? ?a? ?
>(script a and backward script and print a fit in somewhere in the bottom)
>
>(Hmm, badly spaced.) And not sure about some of the symbols/details-- I
know
>I've omitted some of the rounded/unrounded varieties-- , but you get the
>point I'm sure.
>
>Further, each major cell was subdivided 3x3 (tic-tac-toe grid), with the
>basic symbol in the center. Diacritics allowed for raised/lowered,
>fronted/backed varieties within each cell. Thus, raised/fronted [I] came
>very close to lowered/backed [i], but not quite. Our teacher (June Shoup,
I
>think a student of Ladefoged's; she died young, sad to say) seemed able to
>produce most of these subtleties, and some of the students and TA's could
too,
>at least within the confines of the classroom and practice sessions.
>Altogether one of the most stimulating courses I ever took. (And I hope
>memory is doing it justice!)
>
(snip Steg Belsky's msg. that I was replying to)
(Messages to me are displayed in UTF7; when I hit "reply to author" the
reply also goes in UTF7 UNLESS I remember (as I do much/most? of the time)
to change it to "western" charset. There seems to be no way to tell Outlook
Express to _please_ use "western" automatically. Anyone know a way around
this?
It is true, that if I choose "new message" and get Conlang's address from my
address book, then "western" is chosen. Then that involves cutting and
pasting the msg. I'm replyng to, a bit laborious, but if that's the
solution, so be it.. Why can't a computer be more like...us? :-( )
Replies