Date: Sat, 16 Sep 1995 14:24:50 -0400 (EDT)
From: jorge@PHYAST.PITT.EDU
Subject: Re: lojb conv
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
To: Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
Reply-to: jorge@PHYAST.PITT.EDU
Message-id: <01HVCWXOUJSY000GNL@FIPORT.FUNET.FI>
Content-transfer-encoding: 7BIT
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Content-Length: 3305
Lines: 69

And:
> > > I didn't get the impression that the speakers were that bothered about
> > > pauses or glottal stops or whatever.
> > I think I ignore them completely. A computer may get confused by that,
> > but humans can cope well.
> But it also tends to suggest - interestingly - that they're not natural
> word-boundary signallers. If you look at indicators of word-boundaries
> in natural languages, you find that stress, vowel harmony, obstruent
> devoicing, etc. are used in the grammar, but not pauses. But at the
> same time, those examples from NLs do show that there are functional
> pressures on languages to come up with ways to signal word-boundaries,
> so Lojban is not being unnatural in trying to do that, and is being
> merely Lojbanic in trying to take it to extremes.
 
Maybe it is too extreme. For one, it is only a very limited type of
word that is signaled in that way. Mostly it is stress and consonant
clusters that do the word-boundaries. Of the few cases where stops
are prescribed, only a small fraction are truly needed for disambiguation.
 
The stops in front of .V and .VV are not really necessary at all.
Supposedly, they are necessary to distinguish between things like
{broda i} and {brodai}, but if the first is pronounced as three
syllables, which is how it must be pronounced, there is no confusion,
even without the stop.
 
That leaves names. The pause after a name is necessary, there is no
way around that, but that is the pause that least bothers me.
 
The pause in front of names starting with a vowel is the same as that
before .V and .VV, which is not really needed. I can say {la and}
without pause and it can't be anithing other than {la and}. It
is different from {land} and from {la'and}.
 
The pause in front of a name after a COI is supposedly necessary,
in order to avoid the COI being taken as part of the name. If it
was up to me, I would forbid {'} for names, and that would make
the pause there unnecessary for all names except those containing
the syllable {coi}, which could harmlessly be added to {la}, {lai}
and {doi} as forbidden syllables. It certainly is less frequent
than {la} as a possible syllable. I definitely don't make a pause
in the middle of {mi'e xorxes}. I tend to lojbanize English "h" as
{x}, so not having {'} for names doesn't seem to be a big loss.
 
In short, I think that pauses at the start of words are never
really needed. The only ones really necessary are the ones
after a name, but since in most contexts people know what are
the names they are using, hardly can there be confusion if
the pause is not made.
 
> > [ki'e=thanks; ki'a=???]
> kekkatso -
 
I believe that would be {malpinji!} in Lojban.
 
> I'm always getting those two confused. Like {kiu} and {kui}.
 
You could remember {ckire} for {ki'e} and {krinu} for {ki'u}.
 
> Ah! It suddenly makes sense. A lot of people these days are developing
> theories of comprehension that rely very little on syntax. My Lojban
> comprehension would seem to counterexemplify them.
 
Well, in my case I don't think syntax is the main tool I use. I already
assumed from context what Goran was going to say, and then I just saw
that what he said matched my assumption. That's why I didn't notice,
before you mentioned your interpretation, that there was a missing {cu}.
 
Jorge