Message-Id: <199602190749.CAA09064@locke.ccil.org>
Date:         Mon, 19 Feb 1996 07:02:26 +0000
Reply-To: ucleaar <ucleaar@UCL.AC.UK>
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: ucleaar <ucleaar@UCL.AC.UK>
Subject:      Re: loglan rapprochement orthography
To: John Cowan <cowan@LOCKE.CCIL.ORG>
Content-Length: 2712

> >I understood that every letter corresponds uniquely to one phoneme.
> We are getting into lingusitic convention here. I have always presumed
> that diphthongs in  linguistics are a single sound, and hence it is
> meaningful to say that a diphthong is a single phoneme.

It depends on the language and on the phonological theory. English
diphthongs behave like the long monophthongs, and the initial segment
is, on phonetic realizational grounds at least, not clearly a phoneme
itself. So English diphthongs are phonemes - or at least a fairly good
case can be made for that analysis.

> If so, then no, neither TLI Loglan or Lojban has all phonemes
> represented by a single letter, because linguists I  THINK would say
> that the word "oi" consistes of a single phoneme.

But what are the grounds for saying that what is represented
orthographically as <oi> is not a sequence of /o/ and /i/? That is the
simplest analysis, and it preserves audiovisual isomorphism. I don't
see why these linguists would disagree.

> If you extend it into two phonemes, you will phonologically insert a
> glide of some kind, in which case the result is no longer 2 phonemes but
> 3, because the glides are all phonemic in Lojban as well: "o'i o,ui o,ii.
> are distinct.

Is this official prescription? A sequence of two vowel phonemes must
be separated by a glide phoneme?

I had not realized this. I'd always supposed that <oi> is /oi/ =
/o/ + /i/, but syllabification is contrastive, so <oi> contrasts
with <o'i> thus:

   Nucleus
    /  \
   o    i

      Nucleus  Onset  Nucleus         N     O     N
         |               |      ===   |     |     |
         o               i            o     h     i

> >Do I misunderstand? Does the second phoneme in <bo> (both orthographies)
> >occur in the word <lau> (standard standard)?
> Official Lojban says no, but we choose not to argue with JCB whether the
> sound in that diphthong is o or u.  The Lojban phoneme is represented by
> au/ao and is a single phoneme, if I understand what phonemes are (always
> a questionable proposition, especilly when arguing with you %^)

If <oi> is not /oi/ then either there is no audiovisual isomorphism
({goi} should really be spelt <g%>, or whatever) or there is audiovisual
isomorphism only of a complex, doubly articulated variety, wherein
complete sequence of graphemes are in one-one correspondence with sequences
of phonemes, but individual graphemes are not in one-one correspondence
with individual phonemes.

I recommend that <oi> be deemed a tautosyllabic sequence of /o/ and /i/,
and same for other diphthongs. As for ao/au, that should be analysed as
a difference between alternative **phonology** standards.

coo, mie and