[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Rép. : [s] Re: Do phonemes = sounds?

On Mon, 6 Jun 2005, Richard H. wrote:

> I keep seeing phonemes being referred to as if they are real sounds.
> However I recollect seeing a discussion in a book which identified phonemes as being
> abstract definitions/notations and NOT representations of actual sounds.
>Ithink it was invented in order to detach phonology from phonetics.
>For me this separation always seemed artificial.
The present discussion shows that the most  remarkable theoretical breakthroughs in linguistics and phonology ( I am referring to F. de Saussure's conceptualisation of language as well as the Prague School - and Trubetzkoy) have unfortunately fallen onto oblivion. They have not been proven wrong!
In a nutshel: a phoneme is one the possible ways of knowing speech sounds (sounds produced by human speech organs). Distinguishing  phonemes from each other is not a NECESSARY consequence of acoustical features of speech sounds.  Only a naive empiricist would believe that the way in which a speaker perceives the speech sounds is determined entirely and only by the acoustic features of the sounds themselves. If this were true the Japanese would hear the difference between /l/ and /r/ and the English speakers would hear the difference between /u/ in French "roue" and /y/ in French "rue". The contribution of Trubetzkoy was essential to the clarification of the distinction between phonology and phonetics:
He argued that phonology  studies the form (contrast, systemic patterning) and phonetics studies the  substance (acoustics, articulation).  Perception, as a cogntive act, implies identification, categorisation and differentiation. The result is a cognitive construct.  A phoneme is a cognitive construct. Operating with phonemes involves abstraction and retension of only those features that are relevant for distinguishing meanings to the exclusion of many other that are present but not relevant in a given language.  Like any cognitive construct it cannot be said to EXIST in measurable terms. Similarly when you operate with a concept of "animal" it is irrelevant whether it has fur or feathers etc. Obviuously a phone to be recognised as a  phoneme must present certain acoustic features that the listener will choose and recognise as part of the phonological system he is operating with. This is why speech synthesis can work.
I agree with Laszlo Toth when he says:

>The problem is that the word "recognition" inherently
>contains a mapping of a data item to an abstract class. So it does not
>deal with any of the sets but with the mapping between them.
Best to all!

Branka Zei Pollermann
 Branka Zei Pollermann PhD
Psychiatrie de Liaison
Hôpitaux Universitaires de Genève
51 Bvd. De la Cluse, 1205 Genève
tél. : 0041 22 382 48 81
Portable : 0041 79 203 92 17

This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom they
are addressed. If you have received this email in error please notify
the system manager.