[lojban-beginners] Re: Word seperation

["Jorge Llambías" <jjllambias@gmail.com>]

On 5/30/06, Alex Martini <alexjm@umich.edu> wrote:
> When streams of sylables are (formally) broken up into words, is
> stress accent involved?

Yes, stress is necessary to be able to tell where a BRIVLA ends.

Stress is totally irrelevant in CMENE, and you never _need_ to stress
CMAVO (in some relatively few cases cmavo can't be stressed, but in
general stress is free for cmavo too), but for BRIVLA it is essential to
get it right.

> I'm looking at cases like {dabroda} where
> both {dabro da} and {da broda} form two words with allowed
> structures. The only thing is that in the first (incorrect) example,
> the stress would be off.

That's probably the simplest example, but there are plenty of others.
{klamAda} is a fu'ivla, {klAmada} is {klama da}. {blablablAbla} is a
single lujvo, {blAblablAbla} is the tanru {blabla blabla}, etc.

> Would this cause problems parsing Lojban as
> spoken by speakers from languages that don't use stress accent or use
> only very weak accents, such as Japanese?

For humans probably not, because we use a lot of other clues to
break speech into words. If the "correct" parsing makes no sense,
we automatically explore other possibilities. For a mechanical parser,
I suppose it will depend on how intelligent it is made.

I'm not sure stress will be the worst thing to worry Japanese speakers
though, given the consonant clusters Lojban has. Lojban was not designed
to be particularly easy to pronounce.

mu'o mi'e xorxes