Re: THEORY: phonemics (was: RE: [CONLANG] Optimum number of symbols

Julien:
> >1. I reject the classical notion of a phoneme inventory, because there
> >are different sets of contrasting segments for different phonotactic
> >positions (e.g. syllable-initial and syllable-final). Criteria for
> >identifying members of different sets with one another rest on a set
> >of rather woolly criteria involving some notion of phonetic similarity
> >or details of morphological alternations. If the phoneme inventory is
> >to be established by rigid application of the principle of contrast,
> >then the total phoneme inventory should consist of the separate
> >inventories for the different phonotactic positions, with phonotactic
> >position treated as a phonological feature of each phoneme.
>
> Oh, that is really interesting, but I think your analysis rely on a very
> old conception of phonology. Of course, At the earlier developments of
> phonology, the phoneme was the ultimate unit.
Just bear in mind that the phoneme theory is 'a very old conception of
phonology'. I didn't intend to advocate a specific model of phonology;
I simply was trying to explain why I reject the theory of the phoneme
as inadequate.

> But, if you consider
> Chomsky and Halle's model of phonology (that no one would seriously use
> nowadays), you will see that the last phonological unit is not the
> phoneme but the feature. Now there are theories (like dependency
> phonology) who reject the notion of feature, and prefer notions like
> elements or particles.
This is part of what I was alluding to under (2) below.

Incidentally, I would make a distinction between what I espouse
under (1) above, which correctly captures the set of combinatorial
units that form words' shapes, and more advanced phonological
theory that is attempting to deal with a much larger range of
explananda (even to the point of neglecting the criterion of
lexical contrastiveness). The complex and elaborated theories
of modern formal linguistics do not entirely supplant in our
philosophy the methods and analyses of classical structuralism.

> Your proposal is really interesting in treating position as a feature,
> but I think it would be redundant. Why encode as a feature an
> information that you can easily deduce from the syllable position?
You are quite right. What I meant is that if you wish to preserve
the notion from phonemics that the phonology can be stated in terms
of a phoneme inventory plus a set of phonotactic rules, then
position must be treated as a contrastive feature. But I agree
that a better way to look at it is to see not one phoneme
inventory for the whole language but rather one inventory per
position.

> And there would be several problems, in such cases :
> Southern French "ourse blanche" "white female bear" pronounced either as
> [uR.s@.bl~a.S@] or[u.Rs@.bl~a.S@] The [R] has no fixed position in the
> syllable tree, so you cannot give it a featural status. A good
> representation in my opinion is to treat this consonant as floating in
> the underlying representation, as almost phonologist would do now to
> take account for French liaison or Sanskrit sandhi.
Good point. I don't know the rationale for the French syllabifications
you cite, but I can give you analogous examples from English:

_William_  [wIwj@m] or [wIl(i)j@m]
_failure_  [f&oj@]  or [f&Il(i)j@]
... where the difference in realization depends on whether or not the
/l/ is part of the onset of the second syllable.

The proper way to analyse this is to see the syllabification of the
/l/ as unspecified in the representation that encodes lexically
contrastive phonology. But if we were to stick with the 'revised
phoneme theory' I sketched, then I think the way to go would be to
define some sort of archiphonemic /l/ unspecified for position.

> More over, I want to point out that if you treat each context as
> distinct from others and with no "interrelations", then phonology will
> not predict anything at all. I mean, if phonology has just to notice
> that these sounds appear in this context, those in that context and so
> on, then every thing is already done in phonology, but we then cannot
> hope to offer good generalizations, a hole view of how system organize
> themselves.
Maybe yes, maybe no. But for some purposes a principled but unexplanatory
description is better than an unprincipled but explanatory one. This is
the point I was trying to make above about modern and structuralist
approaches.

> >2. The more usual objection to phonemics (as I gather most of us know;
> >please don't take my explanation to be patronizing) is that it overly
> >privileges the segment. A full analysis of a phonological structure
> >requires recognizing elements both below the level of the segment (i.e.
> >structure internal to the segment) and above the level of the segment
> >(e.g. at rime, syllable, foot level). So, to answer Julien, the
> >validity of the segment is not being denied; what is being denied
> >is its privileged, 'emic', status. To my mind, Julien wrongly
> >takes 'phoneme' to mean 'phonological segment'; 'phoneme' carries
> >extra theoretical baggage with it.
>
> I must certainly have been not clear because I assume that (as far as I
> know about phonology) there is no doubt by now that phoneme is not the
> ultimate phonological unit.
Correct, but this is a rejection of the phoneme. Your statement could
be rephrased as "there is no doubt by now that the phoneme is not a
valid phonological construct", since part of the definition of the
phoneme is that it is the ultimate phonological unit.

> As I told in the preceding message, Theories
> like autosegmental or three-dimensional phonologies clearly demonstrated
> that we needed to separate different tiers in the underlying
> representations : this new way of representing sounds allows deeper
> analysises and they can take account for prosodic phenomena.
> So I hope I managed to explain clerly what I wanted to say :)
I understand you and don't disagree with you. I gather that, schooled
in modern phonology as you are, you too reject the phoneme. Recall
it was my throwaway line about rejection the phoneme that prompted
requests for clarification.

--And.

Reply