Lojban and human-computer interaction

[Minty]

What are the limits and possibilities of using Lojban for human-computer interaction?
For those who don't know, Lojban is a constructed language which was created and computer-tested to be 100% logical with no room for misinterpretation, but still (supposedly!) humanly enough to learn.
I like the idea if/when computers become intelligent in the sentient sense because one will be able to speak in English, software can translate English-Lojban and the computer will be able to translate Lojban to computer code. I was wondering if translation errors would still occur at human-Lojban stage and if they can be programmed out in time.

Also, will lojban-human print-outs read in lojban grammatical form or would it be possible to program another translation, if you will, into human grammar?
I think that it would be more accurate and easier if the human interpreted the lojban-form translation because humans have that edge when it comes to coming to conclusions with a lack of data, whilst computers can only deal with absolutes, 1s and 0s.

By the way, if you want a mind-fuck and to test your intelligence while simultaneously getting yourself the edge in tomorrow's world of computing, give lojban a try. I tried it and failed but I was only 15 then so I'll try again some time soon (yes, I'm a geek. problem?). It's actually not as difficult as I once found it.

P.S. I hope u liek my pwetty pictures.

 

[monitoradiation]

Lojban? That sounds cool...

I have a hard time believing that you can express subjective issues in any language without room for misinterpretation since innately within subjectivity there lies room for interpretation.

We already have a language that is rigorous - mathematics. I'm not sure if everyone can read it, though. I sure as hell cant read all of it.

Edit: That kind of reminds me of the Elcors in the Mass Effect games, whose signature trait (aside from looking like small meandering elephants) is to preface every sentence with the expression/inflection it intends for the sentence to take on. Like "Pleasantly surprised: Oh, I didn't know that chocolate was edible."

 

[borrofburi]

lojban to english will not be difficult, english to lojban however will be. The whole point of lojban is that english is ambiguous and hard (if not impossible) for computers to understand.

 

[Minty]

I have a hard time believing that you can express subjective issues in any language without room for misinterpretation since innately within subjectivity there lies room for interpretation.

You may be right. The fact that language is a way of describing the world and the only way one can describe the world is by experiencing it means that all language is subjective, or at least inter-subjective (generally agreed upon). However, ignoring the philosophical aspect of it, it is about as rigorously logical as you can get. For example, the sentence in English "The three men carry a piano" may sound unambiguous to you, but are the three men each carrying one piano? are all three carrying one in unison? In lojban there are markers for all of this so it becomes like "Those-over-there which-I-call-men in-unison carry the piano of-number one".

We already have a language that is rigorous - mathematics. I'm not sure if everyone can read it, though. I sure as hell cant read all of it.

But that's a language of numbers and relationships- it's not great for expressing the average human experience. How do you say "I am happy" in mathematics?

lojban to english will not be difficult, english to lojban however will be. The whole point of lojban is that english is ambiguous and hard (if not impossible) for computers to understand.

That's what I thought. I was thinking maybe one could make a computer program that parsed the English into its possible meanings and checked it against the previous sumti in order to find the most likely meaning. It doesn't seem like it has to be that advanced technology, especially when you consider what we can already do. I reckon it just needs some interest and funding (unless I'm being too optimistic).

 

[monitoradiation]

You may be right. The fact that language is a way of describing the world and the only way one can describe the world is by experiencing it means that all language is subjective, or at least inter-subjective (generally agreed upon). However, ignoring the philosophical aspect of it, it is about as rigorously logical as you can get. For example, the sentence in English "The three men carry a piano" may sound unambiguous to you, but are the three men each carrying one piano? are all three carrying one in unison? In lojban there are markers for all of this so it becomes like "Those-over-there which-I-call-men in-unison carry the piano of-number one".

I'm not disagreeing that sentences which are factual statements can be better represented than they are currently by the english language. I'm stating that I have doubts as to whether or not subjective sentences can be rigorously represented without room for misinterpretation.

But that's a language of numbers and relationships- it's not great for expressing the average human experience. How do you say "I am happy" in mathematics?

That's kind of my point. Even a language of rigor like mathematics would not be good for expressing subjective issues, which makes me have further doubt that a new language can. But I'm not rejecting the possibility outright.

 

[nasher168]

The problem is that as soon as you were to have large numbers of humans speaking it, they would make unconscious modifications over time. The language would evolve like all others have, until it becomes illogical and the original point is defeated.

 

[Minty]

The problem is that as soon as you were to have large numbers of humans speaking it

Well, thank goodness that probably won't happen, at least not soon. It was never designed to be some kind of international auxillary language like Esperanto, or to be spoken on a large scale. The only real purpose of it was to test the Sapir-Whorf hypothesis, which is my main reason for starting to learn it.
I suppose, if it does become wide-spoken and more 'natural' and less 'logical', at least it will have had a good start- meaning it will remain at least somewhat close to its roots and probably retain many of the benefits of that.