Message-Id: <m0qQDxs-000023C@xiron.pc.helsinki.fi>
Date:         Tue, 19 Jul 1994 13:04:43 BST
Reply-To:     C.J.Fine@BRADFORD.AC.UK
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         Colin Fine <C.J.Fine@BRADFORD.AC.UK>
Subject:      Re: lojban writing and speaking rules
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Content-Length: 3033
Lines: 75

Philip asks about the Lojban 'audio-visual isomorphism' - and
how it apparently breaks down.

The answers given (which I'm not sure if Philip will have received
since he apparently had a problem with his mail-box) are of
course correct - the limits of a brivla (predicate word) are
unambiguously determined by the mandatory stress on the penultimate
vowel of the brivla (and the compounding rules have various
special cases to make sure this works).

But I think Philip is right to question the 'audio-visual isomorphism',
as I have done myself in the past.

It is clear that there is in a strict sense an isomorphism: since the
stress and pause rules on the one hand, and the rules for word-spaces
on the other, serve to delimit the words unambiguously, there is a
one-to-one correspondence between the words in spoken and written Lojban:
by definition an isomorphism.

However I am troubled by the very different mechanisms used for the limiting.

If I summarise what I believe are the relevant rules:

Words
=====
Syntactically there are three classes of word in Lojban: cmene (names) which
are consonant-final; brivla (predicate words) which contain at least one
consonant cluster (possibly buffered with "y"); and cmavo, which consist of
one or more vowels possibly preceded by a consonant.
Note that compound cmavo (eg "sepi'o", "panonopa") are not strictly words,
but decompose into the separate cmavo.

In speech
=========
P1. All consonant-final words must be followed by a pause
P2. All vowel-initial words must be preceded by a pause
P3. Pauses are not permitted within words
P4. Pauses are optional at all other word junctures

S1. The penultimate vowel of a brivla must be stressed (this applies to
words which are morphologically brivla, and not cmavo functioning as brivla
such as "go'i" or "remoi")
S2. All other syllables may be stressed or not at will
S3. If the syllable preceding a brivla is stressed, it must be followed by
a pause

In writing
==========
W1. Brivla must be preceded and followed by white space
W2. Whitespace is not permitted within a word.
W3. Whitespace is otherwise optional at word junctures.

WP1. It is customary but not necessary to mark obligatory pauses with a dot.

The problem in my view is that it is far from obvious that these sets of
rules do define an isomorphism. It actually requires two intricate proofs,
one for speech and one for writing.

There are at least two ways of writing which more closely reflect the speech
rules:
1) Mark obligatory stress, and obligatory pauses only

lenumisnIgaulesArcuterbAsnapo'o .u'ucurInkaba'alenudosenAndulenutcIdudei

2) Mark obligatory pauses, and leave whitespace after all brivla

lenumipunji lekunti leselidne berolobrivla curinka lenumleca falenudosenandu
lenutcidu dei

I am not seriously suggesting these as alternatives; (I find the first
extermely difficult to read, and the second rather easier) but I do think we
need to be more careful to explain why our speech rules and writing rules do
perform the same task.

        Colin