On Mon, 2 Jul 2007 17:01:25 +0100, R A Brown wrote:

> Jörg Rhiemeier wrote:
> > Hallo!
> Hi!
> [Plan B etc]
> > On Mon, 2 Jul 2007 07:27:29 +0100, R A Brown wrote:
> [snip]
> >>The language *simply is not spelled phonemically*
> >  
> > It can indeed be argued that Plan B has 16 consonant phonemes and
> > 16 vowel phonemes with a rule that forbids both consonant clusters
> > and vowel clusters; 
> It most certainly has 16 consonant phonemes. Whether it has 8 or 16 
> vowel phonemes is a matter of interpretation. If phonemic status is 
> given to /r/, then we have 8 vowel phonemes plus 8 combos of /r/ plus a 
> vowel. The rule is then that syllables must be of the form: C(r)V(C)

I did not remember that the "16 vowels" included 8 /r/+vowel combos.
I'd say that /r/ is a phoneme, so there are 17 consonants, one of them
being /r/ that patterns fundamentally different than the other 16, and
8 vowels.

>  >
> > this is probably a better analysis than saying
> > it had 16 phonemes each with a consonantal and a vocalic allophone.
> That IMHO is a simply a ridiculous analysis!

Concurred.  It is very far-fetched, and if someone was to analyse
a natlang that way, he should ask himself what he is doing.  It is
like saying that English had a phoneme that is realized as [h] in
syllable onsets and as [N] in codas - only worse (surely, the
English word _hang_ is not a phonemic palindrome!).

> [...]
> >>Indeed, Jörg Riemeier has outlined an engelang using just this proposed 
> >>syllabary, see:
> >>
> > 
> > 
> > I have to concede that that project quickly lost momentum soon after
> > I started it.  Well, I am more interested in naturalistic diachronic
> > artlangs.  Don't expect X-1 to go beyond that outline anytime soon.
> i know the feeling - the same has happened to my 'experimental Loglan'   :)

It is a well-known (to me) mistake of mine to start all sorts of
projects even though I have enough to do already, most of which
soon fall by the wayside.  And X-1 is one of those projects.

> > The question is, what are the phonemes of X-1?  One way to analyse it
> > would be that it had 16 consonant phonemes and no phonemic vowels,
> I don't think that would be a sensible way (or, indeed, a correct way) 
> of doing it.

Again, concurred.

> > and that the orthography was phonemic; but it is probably more
> > reasonable to analyse it as having 7 consonant phonemes /p t k s m n l/
> > and 4 vowel phonemes /E i O u/.
> Basically I agree, tho I think I would include the 'zero consonant' as a 
> phoneme, i.e. 8 consonant phonemes. This simplifies the rule that all 
> syllables must be of the type CV

That's a legitimate analysis, though I'd avoid it - I am not all that
comfortable with zero phonemes.

> > [...]
> >  
> > Right.  After all, letters are just means to *represent* language;
> > perceiving them as the "basic building blocks" of a language means
> > falling way behind even 19th-century linguistics.  And in the case of
> > Plan B and X-1, the "deep level" is actually a stream of *bits*.
> Yep - and IMO in view of Plan B's odd use of each quartet to represent 
> either a consonant or a vowel, it would have been better to use the 
> plain ol' hexadecimal digits, i.e. have the "alphabet": 0 1 2 3 4 5 6 7 
> 8 9 A B C D E F.

Yes.  That would not really look like a pronounceable *language*,
but Prothero's sequences of consonants of which every other is
to be pronounced as a (seemingly) unrelated vowel doesn't, either.
(Nor does the all-consonant orthography of X-1, I have to admit.)

> [Phonemic & phonetic vowels]
> [snip]
> >>
> >>Surely not! I have come across analyses of Caucasian languages which 
> >>postulate no _phonemic_ vowels; indeed, I have met analyses of PIE that 
> >>postulate no phonemic vowels.
> > 
> > 
> > I have heard of such analyses as well; I don't think they are appropriate,
> > though.
> Yes, I have reservations such analyses

Me too.
> > ObConlang: Old Albic could be analysed as having just one vowel phoneme
> > with vowel features being suprasegmental (they indeed behave quite much
> > like tones do in some African languages), and it wouldn't surprise me
> > if some razor-witted phonologist could analyse even that single vowel
> > phoneme away :)
> Indeed - I'm sure you're right     :)

I haven't yet tried to analyse Old Albic vowel phonemes away, indeed,
I'd normally say that the language has seven short and seven long

> > 
> >>But no phonetic vowels? Is such a beast possible?
> > 
> > 
> > You at least need some sort of syllable nuclei; however, nasals and
> > liquids could be pressed into service for this :)  But I doubt that
> > such languages have ever evolved naturally.
> If these sounds are serving as nuclei of syllables then they are 
> _phonologically_ vowels.

This is very true!
> One does need to be very careful how we define vowel & consonant. These 
> terms have somewhat different meanings depending upon whether we are 
> talking in terms of phonetics or of phonology (Pike suggested using the 
> terms 'vocoid' and 'contoid' when speaking phonetically, reserving 
> 'vowel' and 'consonant' for strictly phonological meanings).

Yes.  Phonetics is not phonology, and phonology isn't phonetics.

>       And both 
> must, of course, be distinguished from the popular usage which applies 
> these terms to letters of the alphabet, classifying |a e i o u| and 
> sometimes |y| as vowels and all the rest as consonants, however they are 
> actually used phonetically or phonologically.

And orthography is neither phonetics nor phonology.  But many people
fail to realize.

> Didn't we discuss this on the Conlang list in February this year?


... brought to you by the Weeping Elf