[dbrock]

There are many people who can converse fairly well in real-time written Lojban (although vocabulary is limited, and some parts of the language are generally not used).

Due to audio-visual isomorphism, being able to converse in written Lojban almost implies being able (theoretically) to converse in spoken Lojban. However, people are generally much less used to speaking and hearing Lojban, so spoken conversations are usually somewhat difficult (definitely far from fluent).

[DanielStraight]

Right. It's been around since 1955 (in spirit at least, in the form of Loglan) and the best anyone can muster is to converse "fairly well" in "real-time" but not in spoken conversation. In other words, it's a language no one can speak. Klingon speakers are even more fluent than Lojban speakers.

I'm not trying to say Lojban is stupid. It's not. It's cool. It fills an interesting niche in the language world. However, I'm not convinced it's speakable. Remember the quote by Kernighan: "Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it." I would argue that speaking a language is at least twice as hard as inventing it or writing it in a non-real-time fashion, so if you create the most complicated language you can still manage to write (however laboriously), you are by definition not smart enough to speak it. Lojban is that language.

[RiderOfGiraffes]

Lojban is not designed to be complicated. The important parts of the BNF grammar are about a page, and easily understood in an hour or so of properly directed study.

The hard parts are the vocabulary and the mind-set. It is specifically designed to be different, but even then, there are sufficient similarities that it doesn't have to feel completely alien. Some people present it as such because they think it will attract people, but I know that sometimes familiarity is a better draw.

It can be presented either way.

And people tend to say uncomplicated things in Klingon, whereas most lojban speakers are exploring saying very complicated things - you're not comparing like with like. Further, although gatherings of Klingon speakers are fairly common - piggy-backing on Star Trek conventions - gatherings of lojban speakers are rare. Even so, there are several people who speak it fairly fluently, and the number is growing (in some sense).

It is often criticised (I'm not saying you are doing so) for being something other than what people think it "ought" to be. This is a marketing issue. I think it's interesting, and, like learning Lisp, it has expanded my mind in interesting ways.

[DanielStraight]

Lojban may not be designed to be complicated, but it's designed to be complicated. By that I mean, while the goal might not be "make it complicated," the goals necessitate that it will in fact be complicated. There is no way to remove all ambiguity from language without making it exceedingly complicated.

[RiderOfGiraffes]

This is a persistent, incorrect interpretation of what people mean when they say that lojban is intended to be unambiguous.

Parsing lojban is unambiguous. Interpreting lojban quite specifically is not unambiguous.

Given a grammatically correct construction there is a unique parse. That is what is meant by "unambiguous". There's no problems akin to "Machines need to wreck a nice beach."

There are manifold ambiguities, however, in the semantics. When one talks of "lo sutra tavla" there is no indication as to the sense in which the speaker is fast. Perhaps the speaker produces many words per minute, or perhaps the speaker runs past while talking. These ambiguities can be reduced by using more precise expressions. Metaphorical use is frowned upon, so what it does not mean is one who persuades in a fraudulent manner.

For example, we can say "lo gerku" which refers to a dog, or some dogs, but gives no idea of how many. We could say "re le ci gerku", which means "two of the three dogs." More precise.

I can say "mi tavla", which means "I speak" or "I will speak" or "I have spoken" and even leaves the audience, topic and language unspecified. I can say "mi ba tavla" or "mi pu tavla" which are future and past respectively. I can be even more precise if necessary or desirable.

You said in http://news.ycombinator.com/item?id=918042 :

  ... my understanding based solely on reading ABOUT
  Lojban, I don't know a single Lojban word... aside
  from Lojban

I think you have fallen into the trap of not reading enough, and mis-interpreting some of what you have read.

I'm not surprised, I think much of the early material written about lojban was written without regard for how it might be mis-interpreted. Politicians today generally say nothing, because everything they do say runs the risk of being taken other than intended. The early lojban writers (writing about lojban, not necessarily in lojban) needed "spin doctors" to ensure that what they said could not be mis-interpreted.

All that aside, lojban is intended to be an expressive language, suitable for communication. Therefore it will be complex, although the complexities are not necessarily those of natlangs. I suspect that we are not that far apart. We agree that:

- lojban is complex

- lojban is not currently suited for general use

- lojban is cool

I further believe that:

- learning lojban (at least beyond "mi tavla") is mind-expanding

- learning the structures of lojban teach more than just lojban, they teach about structure, syntax, and monolinguistic assumptions.

- it's not for everyone.

[DanielStraight]

I think we actually completely agree. I agree with all your further believes too.