>>CB: My thought is that there was an arbitrary connecting between "g" and the 
>>thing, <<

Certainly not iconic, but not entirely arbitrary, but rather
'motivated' by human psychology (hence the use of a velar for much of
the same purpose across cultures and languages).

>>CB:  I don't have a feeling that repetition represents endearment.<<

Well it most basically represents enhancement, extension, plurality.
There is no plural form in Japanese like the one in English (a cat,
two cats), but, for example, ki is tree, ki-gi is trees (repetition of
the ki morpheme, morphophonological voicing of the [k] to [g]).
As for what I called endearment, that would depend on us hearing you
interact with a baby or a beloved pet.

>>CB: The point is that alphabetical writing is arbitrary and picture writing 
>>is not .<<

Well alphabets are typically used for glottographic writing, pictures
are not--that is, they do not represent a written language derived
from a spoken one. However, the alphabets we know today descend from
writing systems that had pictographic elements in their development.
Moreover, you can still see motivated relics in alphabets (such as the
letter o, the shape of letter r vs. letter l pointing out an
articulatory difference). Korean hangul is not an entirely arbitrary
writing system because it represents a systematic working out of how
to pronounce Korean--it has phonological motivation.

>>CB: But the p and g are not simultaneously articulated. They are in a
sequence.>>

But the point is they are not discrete segments, which is one popular
view of the phoneme--that is what gives it its theoretical 'unicity'.
In speech, in acoustics and in auditory perception, however, that
unicity does not exist.

>>CB: How are words distinguished based on the articulatory gesture
concept ?<<

One difference is that if we use the articulatory gesture as an
abstract model of speech, it is hard to put into written discourse.
The beauty of the phoneme is that it helps you turn dynamic,
co-articulated speech into static linear segments for reading (though
research on reading processes show that a text is not 'bar code
scanned', that reading processes are not straight forward linear
either).  So once you grasp the dynamic nature of speech, you can find
it easier to embrace the articulatory gesture as a 'unit'.

I would say a second difference is that we can use the articulatory
gesture to move away from a negative differential view of meaning to a
positive differential view of meaning. Phonological encoding is
actually (I hope) an integral part of a language's grammar. And most
likely the basic unit is a psycholinguistic ability to control, store,
process, encode and decode the spoken syllable, not the phoneme.


>>CB: Categorical perception is still fundamental to how language works,
and how symbolling and culture works.  In what sense is categorical
perception a house of cards ?  Categorical thinking is fundamental to
humans, and it is fundamentally social/communicative/communal, so that
it must use conventional or agreed upon uses of percepts.<<

As it has been identified and researched, categorical perception is
not limited to human language or humans--even though it has been
argued before as something that separates human language and humans
from the rest of the animal kingdom. Most importantly though as it has
been identified and researched, it appears to be an epiphenomenon.

>>CB: This paragraph in the article you give us below, says that Chomsky and 
>>Halle do use phoneme.<<

Well the explanation there is largely schematic, and an uneven mix of
linguistics and philosophical ontology. But as I understand their
position, they reject the way phonemes were derived from
distributional methods (in structuralism), they embrace the use of
minimal pairs (which I reject--due to the circularity of using your
lexicon to delimit your phonological unit and then using your
phonological unit to delimit your lexicon ), and then they analyze
phonemes into features. In this version, the analysis is still static,
but the contrastive phoneme becomes unnecessary and the contrastive
feature is given prominence. However, the weaknesses of the approach
include it is still static, confuses articulatory aspects with
acoustic and perceptual ones, fails to produce an exhaustive agreed-to
list of features. It is also important to remember that in a static
analysis, the key thing to remember about the feature is that it is
not a discrete element in that it spreads across segments, syllables,
and whole words (like, for example, the feature of nasality in the
word 'can'--contrast 'can' with 'cap'--the vowel is different and the
nasality starts in the transition between the onset ([kj-] and the
nucleus vowel).

Since then the contrastive feature (which is really, I believe, what
the structuralist Jakobson called them) has morphed into extremely
abstract models like 'metrical frames'. One British linguist was so
certain that it was this sort of thing that underlay speech processing
that he got a patent on his version of the model and sold it as a
surefire way to program computers to process rapid human speech. He
got the patent, I'm sure he sold it to some programmer who didn't know
anything about linguistics and speech perception. Have no doubt this
was one of the 'holy grails' of programming in the 90s (successful
speech recognition). However, the subsequent software didn't work any
better than the ones flailing about with phonemic models programmed
into them. I think it didn't work at all actually.

If a computer can not be programmed to process complexly
co-articulated speech, it can not process spoken human language. And a
static model is not going to work. Most likely up to a point anyway
increasing memory and processing power combined with a syllabic and
word model will work better.
CJ

_______________________________________________
Marxism-Thaxis mailing list
[email protected]
To change your options or unsubscribe go to:
http://lists.econ.utah.edu/mailman/listinfo/marxism-thaxis

Reply via email to