[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[lojban] Re: ambiguity in lojban



21.06.2006, 10:05 Chris Capel wrote:


> Many of the older texts mention that Lojban might be uniquely suitable

> for human-computer communication. The idea is that since computers are

> so bad at resolving the meaning of polysemous words, and at resolving

> ambiguity in grammar, (two things that humans are extremely good at,)

> that the absence of these things will make it much easier to allow

> computers to understand human speech. But the fact is that computers

> that don't have a flexible enough language model to resolve

> ambiguities from polysemy and grammar ambiguity are unlikely to be

> nearly subtle enough to correctly resolve ambiguities in extent. So

> creating a language like Lojban for that purpose (which was never the,

> or even a, primary purpose for the language, as I understand) is

> analogous to taking a course in mechanical engineering and building a

> shovel cleaner before starting the work on digging a tunnel to China.

> Any application in which some restricted subset of English could be

> used, except for problems with polysemy and grammar ambiguity, is an

> application that probably doesn't need to use a natural language at

> all.


The passage with the shovel to China is great! ;)


I think no one has ever pretended that Lojban can 10 times ease the understanding of the language by a machine.


Lojban just allows to throw away lots of debris present in any natural language (and constituting just 1% of the difficulties, IMHO) and concentrate on the main and biggest problem, namely the chain:


symbolic system -- [a lot of hidden, mysterious AI stuff, including context resolving] -- internal knowledge representation in a mind.



mi'e .ianis.

To unsubscribe from this list, send mail to lojban-list-request@lojban.org with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if you're really stuck, send mail to secretary@lojban.org for help.