Re: Accent Terminology Question

At 9:27 PM -0400 10/11/02, David Peterson wrote:
>Responding to Dirk and Jeff:
>
>This is what happens when you try to apply imperfect knowledge to a
>situation.  :(  Sorry.
No apologies necessary; all knowledge is imperfect.

><<FR  42-24   42-24-24   42-24-24-24>>
>
>Wow!  That's amazing!  I'd really like to hear that!
The best I can do for you is point you to the original article. It
is: Yip, Moira. 1989. "Contour Tones." _Phonology 6: 149-174.

><<John Goldsmith proposed a default left-to-right spreading rule in his
>1976 dissertation (he called it the Universal Association
>Convention).>>
>
>And by "right to left", I of course meant "left to right".  ;)  Not
>really.  I was confusing it with something else that I learned
>roughly at the same time in the same class--namely, reduplicatory
>prefixes and infixing prefixes.  I think they also attempted to
>describe the Arabic triconsonantal system with right to left
>spreading (as far as consonant roles, and the gemination that occurs
>in certain forms, e.g., the causative).
As I recall (and I hope I'm right since I have to teach it again in a
few weeks), the Arabic system is also left-to-right; biconsonantal
roots show up as 122 rather than 112.

Patterns of association in reduplication and infixation will depend
on the position of the infix/reduplicative affix. If you have a
reduplicating prefix, the pattern of association from the base will
be left-to-right; if you have a reduplicating suffix, the pattern of
association will be from right-to-left. There are some oddballs;
Nancowry has prefixing reduplication which is anchored to the right
edge of the base, and Koasati has suffixing reduplication which is
anchored to the left edge of the base.

Nancowry
base: -yak, reduplicated form: ?uk-yak

Koasati
base: tahas-, reduplicated form: tahas-to:(-pin)

In each case, the directionality of copying under reduplication is
not what is expected from the position of the affix.

><<though the
>patterns of tune-text association in Gregorian Chant (which otherwise
>behaves like a tone language) are right-to-left.>>
>
>Bah!?  How can I learn more about this?
Matthew Chen published an article on this in the first volume of
_Music Perception_ (1983). It's called "Toward a Grammar of Singing:
Tune-Text Association in Gregorian Chant." In grad school, I wrote an
OT reanalysis of some of the data; it wasn't published anywhere, and
isn't likely to be, though I'd like to find out if the analysis of
melodic patterns in plain chant are amenable to a preference rule
system like OT. I suspect they are.

><<Not all languages
>tolerate contour tones>>
>
>Another "oops".  When I said that languages do this "generally", I
>meant that languages with contour tones can do that.  ~:D
>
><<The idea of underlying tone is similar to the idea of any underlying
>feature.>>
>
>To illustrate, if, for example, you have a language with a high
>tone, a low tone, a rising tone, and a falling tone, and you have a
>string of affixes that usually have no tone associated with them...
>(I'll continue with my 1, 2 system, since there are only two levels.)
>
>/ke(1)te(1)/ = "to eat"
>
>/-ma()/ = "I" (subject) (the () means there's no tone)
>
>/-le(*2)/ = "it" (object) (that * is to indicate that the tone is underlying)
>
>So, if you want to build these into the word "I eat it", where the
>first suffix is the subject suffix, and the second subject is the
>object suffix, you'd get...
>
>/ke(1)te(1) + ma() + le(*2)/
>
>Which would end up as...
>
>/ke(1)te(1)ma(1)le(21)/
>
>With this output, you can posit that the affix /-le/ has an
>underlying high tone, which causes the blip.  Did I get that right?
I think so. Your messages come to me in 8 point blue type, so it's
hard to read. In your system, is "1" low and "2" high? If so, the "1"
on _le_ is unusual; since the suffix already has tone, there's no
need for a default "1" to be inserted creating a universally
dispreferred (even if parochially tolerated) configuration.

Dirk
--
Dirk Elzinga                                             Dirk_Elzinga@byu.edu

"No theory can exclude everything that is wrong, poor, or even detestable, or
include everything that is right, good, or beautiful."    - Arnold Schoenberg