Hallo conlangers!

On Sunday 20 November 2011 16:36:01 R A Brown wrote:

> On 19/11/2011 18:12, BPJ wrote:
> [snip]
> 
> I have excerpted bits from BPJ's email in a different order
> in which they were written because one or two problems
> discussed below are not problems in the off-line version of
> the Brx 'Phonology & Orthography' page.
> 
> [PALATALIZED PLOSIVES & SIBILANTS]
> 
> > [^5]: But note that you probably should abandon the
> > transcription of *x* as [çi] in favor of [ɕi] or even
> > [ʃi].
> 
> and:
> > I've heard of the sibilant [s] palatalizing and ending up
> > as [ɕ], [ʃ] or even [ʂ], but never actually as the
> > non-sibilant [ç].
> 
> I have _already_ abandoned that transcription; indeed, I had
> substituted /ʃi/, but of course phonemically it is /si/. At
> the moment I have it noted, slightly unphonemically, as
> /sʲi/, with a note saying it realized as either [ʃi] or
> [ɕi]; indeed, IMO [sʲi] itself would be quite satisfactory.

Good.

> /ç/ was a hangover from Bax (Piashi) where it did make sense
> since in that language both /si/ and /χi/ had fallen
> together as [çi].  But this is not so in Brx. There is only
> one series of fricatives, namely _sibilants_.  For that
> reason /çi/ and/or [çi] no longer make any sense and have
> been dropped [see above].

If there are no non-sibilant fricatives in Brx, there indeed is
no reason for transcribing it as /çi/.  It may have made sense
in Bax, but Brx is not Bax and Bax is abandoned, as far as I am
informed ... On the other hand, /ç/ of course nicely pairs with
the /c/ discussed below.
 
> > or you'll get a lot of [si] for {[ci]} and {[çi]};
> 
> I can't imagine any linguist mispronouncing either [ci] or
> [çi] as [si]!
> 
> The page is, I would have thought, fairly clearly meant for
> those with some linguistic savvy.  Obviously, if I were
> describing Brx for a general anglophone readership, I would
> use _shee_, and if for a francophone readership _chi_    :)

Sure.
 
> > [tʃi] and [ʃ i] have the virtue of not looking familiiar
> > 
> >  from English/French/Spanish orthography, while their
> > 
> > values are somewhat familiar to ordinary dictionary
> > users.
> 
> Brx /ci/ however _does_ represent _both_ /ti/ and /ki/. It
> seems to me reasonable that, if both palatalized /t/ and
> palatalized /k/ have fallen together as the same sound, I
> denote it phonemically as palatal plosive /c/.  This may, of
> course, be pronounced as [c], [tʃ] or [tɕ]

This makes sense.  I think, /c/ is a serviceable transcription
here.  It is midway between /t/ and /k/.
 
> [HIATUS]
> 
> >> Can the syllable break be a phoneme?
> > 
> > A hiatus can -- at least in French --, and /./ is the
> > closest thing to a hiatus symbol you got in an IPA-based
> > phonemic transcription.
> 
> and:
> > Sure, but h aspiré is commonly interpreted as a phoneme
> > with <hiatus> as one of its realizations. On that I
> > think there's agreement. h-dropping English and speakers
> > of h-dropping Swedish dialects also evidence [h] as a
> > realization of <hiatus>, so it goes full circle!
> 
> AFAIK hiatus is the normal between words ending a vowel and
> a word beginning with h aspiré.  But it is more complicated
> in that h aspiré also prevents liaison. One cannot IMO deal
> with h aspiré without considering the whole phenomenon of
> liaison.  I would rather leave the questions of hiatus &
> liaison in modern French to Christophe; I do not think it is
> helpful in this discussion.

I agree with you.  What does French liaison have to do with Brx?
 
> I am, of course, in no position to comment on h-dropping
> Swedes, but IME h-dropping anglophones treat such words just
> like any other words beginning plain vowels.  I am not aware
> of /h/ being a realization of hiatus except maybe as a
> result of hypercorrection or, of course, in some
> stereotypical depictions of Cockney.  In Brit English a
> common way of avoiding hiatus is "intrusive r' - but that's
> another matter.
> 
> OK: *all Brx syllables are CV; there is no hiatus.*

OK.
 
> [BRX APPROXIMANTS]
> 
> > And Pulleyblank and others add /ɦ/ as the semivowel
> > corresponding to /a/ at least for Middle Chinese, if not
> > for some dialects of Modern. Apparently PST *ɣ actually
> > figures in some of these correspondences.
> 
> I haven't ruled out /ɦ/.
> 
> > But note /ɦ/ in the /ɰ/ slot in Middle Chinese
> > (apparently it does show up as /h/ in some exochinese,
> > but I'm unsure of the details. Also NB the frequently
> > allophonic relation between [h] and [ɦ]!
> 
> Oh dear, I wish I hadn't mentioned that in the off-line
> version of the page I was currently using /ɰ/     :(
> 
> Basically it is a _place-holder_ until I have made a
> decision - nothing more, and nothing less.

Fine.  So the decision about what those syllables begin with
is still pending.
 
> > But let's assume for a moment that
> > 
> > 1) Brx is a natlang and not an engelang. 2) Pre-Brx had
> > the three approximant phonemes /w/, /j/ and /ɰ/,
> > 
>     :-)
> 
> Pre-Brx - maybe even a proto-BrSc which branched off into
> pre-Bax and pre-Brx      ;)
> 
> Actually, at one time I did think about producing a
> pseudo-history of how Brsc/Bax/Brx came about.  I thought it
> would help in defining the lexicon.  But I realized a
> pre-history for such a language would be very contrived and
> the amount of tome spent doing it would not be worth it.  As
> for lexicon, it would probably throw up far too many homophones!

As much as I am into diachronic conlanging, I am of the opinion
that fake diachronics make sense only with conlangs that purport
to be natlangs in some fictional world, with a fictional
linguistic history.  There is no reason making up such a thing
with an engelang which exists in the real world and has no
fictional historical background behind it, especially that the
probability that such a language falls out from a set of regular
sound changes is vanishingly small.
 
> I have already questioned how likely [ɰ] is to be a
> _phoneme_. It seems to me the sound more often occurs as the
> allophone of something else.

That is my impression, too.
 
> > which all could occur before and between any two vowels.
> > 3) That they all could have voiceless allophones, --
> > voiceless approximants normally, not fricatives --
> 
> "Because voicelessness has comparatively reduced resistance
> to air flow from the lungs, the increased pulmonic pressure
> creates more turbulence, making acoustic distinctions
> between voiceless approximants (which are extremely rare
> cross-linguistically and voiceless fricatives difficult."
> see:
> http://en.wikipedia.org/wiki/Approximant_consonant
> 
> I don't always go along with Wikipedia, but I do on this one.

I am not a phonetician, but my own impression of these things
concurs with this observation.  Voiceless approximants involve
a balancing act between frication and inaudibility.
 
> [snip]
> 
> > 7) The distinctions thus lost tended to be
> > morphologically important,
> 
> That's going to involve pre-Brx morphology - yikes!

I have said above what I think of diachronics in engelangs ;)
 
> > which militated against these changes -- cf. how the *s
> > 
> > >*h > zero > krasis change failed to happen in s-aorists
> > 
> > in Greek, with which you are surely familiar.
> 
> Michel Lejeune (Phonétique historique du Mycénien et du Grec
> ancien", Paris, 1972) considers intervocalic -s- in sigmatic
> aorists to be "restauration de -σ-" after the analogy of
> βλέπω ~ ἐβλεψα.
> 
> Analogy place too much a part in the evolution of morphology
> in so very many languages. Personally, I think the whole
> notion of pre-Brx with developing phonology and morphology
> is making things unnecessarily complicated.

Amen.
 
> [snip]
> 
> > Now the *engelang* Brx need not have a 100% natural
> > phonology.
> 
> No - an engelang could have a weirdly unnatural phonology.
> But that's not what I want for Brx.

Right.  Engelangs can have unnatural phonologies, but that is
by far not a must, and the goals of a particular engelang can be
such that a fairly natural phonology is desirable. 
 
> There you need only:
> > a) state that theoretically \_a\_ is a pure vowel /a/,
> 
> I have already decided - see above - that all syllables are
> of the CV type.

Nice and elegant.
 
> >> I notice on:
> >> http://en.wikipedia.org/wiki/Approximant_consonant
> >> 
> >> that the approximant equivalent of [ɑ] is given as [ʕ},
> 
> [snip]
> 
> > I have seen [ɰ] (or "[ɣ]") described as the approximant
> > equivalent of [ɑ] or [a].
> 
> Quite likely.

What there is is a scale of openness running from stops via
fricatives, approximants and close vowels to open vowels,
I think.

--
... brought to you by the Weeping Elf
http://www.joerg-rhiemeier.de/Conlang/index.html
"Bêsel asa Êm, a Êm atha cvanthal a cvanth atha Êmel." - SiM 1:1