Re: [lojban] Re: lojban as an auxiliary language

[Brandon Wirick <brandon@yrick.com>]

> Hmm. It seems to me that the hardest part of making the semantic web
> really work is getting everyone to agree on the same intentional
> meaning for any given word. Lojban vocabulary tends to be a bit less
> overloaded than English, but the extensional boundry of any given word
> isn't much less fuzzy, in general. And when you start to think about
> lujvo, well, things go downhill pretty quickly. If Lojban were to ever
> be used in a widespread manner, many lujvo would acquire two distinct
> senses based on different interpretations of the tanru, or an
> assortment of slightly different place structures for the same
> meaning. So does Lojban really help this part of the problem?

I was under the impression that lujvo were vague, not ambiguous. A
{gerzda} is some sort of nest/house/dwelling with some sort of
dog-like aspect to it, right? It is not necessarily a doghouse as
English-speakers understand the term. If you look at one of the large
upper-ontologies like SUMO or OpenCyc, the terms there are vague as
well, but the names are contrived English labels. The makers of those
ontologies even disclaim the names they give classes, saying that the
actual meanings of the entries may have nothing to do with the way a
reader may interpret the name. In Lojban, that would not be the case.
In Lojban, the names motivate the meanings.

> I'm really not sure I understand you here. Are you arguing that people
> should voluntarily avoid using Lojban for non-technical communication
> so that it remains usable for technical purposes? And how would that
> make it more usable?

I'm saying, really, that for a system that already restricts
conversations to rather technical communication (e.g. something
Semantic Web-oriented) Lojban would be a wonderful internal medium.