Re: Russian vowel <bI>

Tommi Nieminen <tommiDOTnieminen@xxxxxxxxxxxxxxxxxxxxx> wrote:
> Peter T. Daniels kirjoitti:

> >> but a linguist might make the mistake of
> >> assuming that the [a] of language L1 is "the same" (in some mysterious
> >> way) as the [a] of language L2.
> >
> > If L1 and L2 both have the vocalic system /i a u/ (as many languages
> > do), then the /a/ of L1 is indeed "the same" as the /a/ of L2 (in some
> > very straightforward way), regardless of the phonetic realizations of
> > the low vowel.

> That's exactly what I meant by "mistake". After abstracting out a
> phonological system with maximised oppositions like /i a u/, the
> linguist can easily claim to have found a universal like eg. "every
> language has a low and a high vowel". Of course: if you first modify the
> data to suit your intentions, you shouldn't be surprised that is does
> just that. Phoneticians can only sigh.

> This what I find so baffling in linguists: even when they claim to work
> with "real" language data, they actually add layer after layer of
> abstractions to it so that in the end what they're studying is just
> their own theoretical constructions.

Please don't identify PTD with "linguists". In generative phonology
and its descendents, a phonological system does not consist in
"maximised oppositions", and underlying phonological forms are the
same sort of objects as the phonetic forms that can be directly
observed. They're not abstract (though they are not directly
observable). Specifically, in the example, it is possible to
distinguish many different language systems with a single low
vowel, other things being equal. (That might not be true with
the archisegments of early generative phonology, but archisegments
were given up on a long time ago.)

--
Greg Lee <greg@xxxxxxxxxxxxxxxxxxx>
.

Relevant Pages

  • Re: Phonemes
    ... phonology text for pre-Optimality Theoretic phonology. ... phonemes of a language. ... redundancy that linguists have previously been so eager to eliminate ... that have a third high vowel prefer a front round vowel rather than a ...
    (sci.lang)
  • Re: The word "evolution"
    ... > choose to, by lengthening the vowel. ... Most, or possibly all, linguists believe that we store representations ... underlying form into a "surface form", ...
    (talk.origins)
  • Re: Hebrew and tradition
    ... > Yes, linguists agree, and this matter has been discussed extensively. ... I assume Pinker's book (_The Language Instinct_) mentions it, ... If you had ever _read_ a book on phonetics or phonology, ... "Syllable theory" has nothing to say about what a syllable is, ...
    (sci.lang)
  • Re: Glottal Tension
    ... Cambridge Red books came out almost simultaneously (e.g. Gussmann, ... Phonology), and it appears that not only has the notion of phoneme been ... abandoned, but so has the notion of phonological system, and "discovery ... teach new linguists how to discover the sound system of a language, ...
    (sci.lang)
  • Re: Grammar question - is "hir" grammatically correct?
    ... vowel at the beginning of the next word. ... linguists think about this, but it seems to me that French terminal ... when they don't make it into the audible utterance. ... L'academie des beaux arts would agree. ...
    (rec.arts.sf.written)