Lojban - a logical (human) language | Hacker News Reader

Lojban - a logical (human) language

neptune.spaceports.com

20 pointsgnosis29 comments

Descended from Loglan, whence came the old joke,

How many Loglan speakers does it take to fix a broken light bulb?

Two. One to fix it, and another to figure out what kind of bulb emits broken light.

In all seriousness, though, one of the things I treasure most about our squishy human languages is how much ambiguity they contain. Being able to construct sentences that mean two (or three, or more) things at once and playing around with the intersections of polysemy (http://languagelog.ldc.upenn.edu/nll/?p=2079) and listener context is great fun. I feel like you lose some of that when you take inspiration for your language as much from predicate calculus as from Chinese.

"Lojban's powerful tanru (combinations of selbri into novel concepts) and word-building features make it easy to make fine distinctions between concepts. This discourages the tendency for individual words to acquire families of meanings. Lojban's tanru metaphors are themselves ambiguous; they specify a relationship between concepts, but not what the relationship is. That relationship can be made explicit using unambiguous logical constructs if necessary, or can be left vague, as the speaker typically desires. Similarly, portions of the logical structure of a Lojban expression can be omitted, greatly simplifying the expression while causing some ambiguity. Unlike in the natural languages, though, this ambiguity is readily identified by a reader or listener. Thus all ambiguity in Lojban is constrained and recognizable, and can be clarified as necessary by further interaction.

This precision in no way confines the meaning of a Lojban sentence. It is possible to be fanciful or ridiculous, to tell lies, or to be misunderstood. You can be very specific, or you can be intentionally vague. Your hearer may not understand what you meant, but will always understand what you said."

http://www.lojban.org/publications/level0/brochure/lojbanmo....

Lojban has all of the wordplay and fun constructions of any natural spoken language. The difference is just that it parses cleanly. This means you can actually create richer constructions that are arbitrarily cleverly convoluted... you just have to hope the reader has enough stack space in his head to parse it!

Personally, I won't miss the self-parodies in English like the sentence "Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo." (http://en.wikipedia.org/wiki/Buffalo_buffalo_Buffalo_buffalo...)

> one of the things I treasure most about our squishy human languages is how much ambiguity they contain

Agreed. Designed languages lack...poetry.

I understand that you mean "poetry" in the metaphorical, not literal sense. Still the point about literal poetry being possible in Lojban is an important one:

"Original poetry has already been written in Lojban, and some has been translated into the language. Lojban's powerful tanru metaphor structure allows you to build concepts into words easily, as you need them, and has been used to create colorful images and to convey moving emotions. A Lojban speaker doesn't need a dictionary to use and understand millions of words that can potentially exist in the language. The absence of cultural constraints makes consideration of new ideas and relationships easier than in natural languages, spurring creativity. Lojban aids in communicating abstractions by identifying their nature explicitly. Lojban is thus a very powerful language, not only for poetry, but for discussing such abstract fields as philosophy, physics, metaphysics, and religion.

Lojban poets are already experimenting with new (and old) forms of poetry that seem especially well suited to the rhythm, sound, and flow of the language. Rarely do poets have such an opportunity to affect the development of a new language as they now can with Lojban. Lojban's rich and powerful. Lojban unleashes the full potential of poetic expression to communicate both concrete and abstract ideas."

http://www.lojban.org/publications/level0/brochure/lojbanmo....

> Lojban unleashes the full potential of poetic expression

Whoever wrote that is lying to himself and others. Please refer to my lower post (the one containing "wherefore art thou Romeo?") for a partial but sufficient explanation.

False. http://lmgtfy.com/?q=lojban+poetry

Thanks for proving my point about ambiguity.

And p.s.-- that's not poetry, it's "poetry"

It's more of a proof of concept than high literature -- keep in mind that Lojban is a very young language, and there aren't many people even truly fluent in it.

Any language's poetry draws its meaning, its form, its essence from the words of those millions of people who expressed their ideas and themselves in that language. No contrived language is ever going to be able to express a thought with the same weight, for example, that "wherefore art thou Romeo?" is expressed in English. It's just not possible, and translated meaning isn't enough. You will never have Giants' shoulders to stand on.

Translated meaning isn't enough? So I suppose all those translations of Shakespeare in to hundreds of human languages isn't enough either; and anyone who claims Shakespeare can be appreciated in anything but English is just "lying to himself and others" (as you said in an earlier post).

Sure, any young language isn't going to have a rich history of expressions weighed down with meaning. But I don't think that has anything to do with whether the language is deliberately invented or not. With time, assuming the language is used by a significant number of people, it will acquire such a history.

Still, while young languages don't have a rich history compared to older languages, they have other advantages (of which Lojban has plenty).

Also remember that the history that has accrued to older languages is a double-edged sword. Because of this "baggage" it's sometimes difficult to express precisely what you mean without being misunderstood (or perhaps to even know what you mean -- see the Sapir-Whorf hypothesis). Lojban is designed to help overcome such difficulties.

The other thing to note is that the weighty phrases like "wherefore art thou Romeo?" aren't used very much in everyday speech. It's not like we go around constantly quoting Shakespeare at one another (except occasionally for effect, or humor). We generally communicate using rather simple language and concepts that should have no problem being expressed in even a young (but expressive) language like Lojban.

>So I suppose ... anyone who claims Shakespeare can be appreciated in anything but English is just "lying to himself and others" (as you said in an earlier post).

It cannot be appreciated to the same degree, no. Just as a 6-year-old cannot appreciate Shakespeare to the same degree that a 50-year-old English literature professor can appreciate it. How this is at all debatable, I have no idea.

> With time, assuming the language is used by a significant number of people, it will acquire such a history.

This will of course never happen.

> Still, while young languages don't have a rich history compared to older languages, they have other advantages (of which Lojban has plenty).

There are alternative solutions to requiring everyone to learn a second language. For example, expressing oneself more clearly. This can be accomplished to any desired degree in modern languages.

> Also remember that the history that has accrued to older languages is a double-edged sword. Because of this "baggage" it's sometimes difficult to express precisely what you mean without being misunderstood (or perhaps to even know what you mean -- see the Sapir-Whorf hypothesis). Lojban is designed to help overcome such difficulties.

On one edge of the sword: the lasting effects of the most skilled manipulators of the given language. The other edge: occasional difficulties with ambiguity and linguistic relativity. To dull the second edge, Lojban requires that we flatten the first. No thanks.

  > > With time, assuming the language is used by a significant
  > > number of people, it will acquire such a history.
  >
  > This will of course never happen.

"Man flying? Impossible!" "A man walking on the moon? Preposterous!" "Slavery/segregation is the natural state of man and will never change." "The Third Reich will last a thousand years." "The Soviet Union will never collapse in our lifetime."

History is full of surprises.

Sure, today Lojban is a fringe language with a few thousand speakers, tops. And there's no indication of any kind of widespread movement to adopt it. But that could change with time. The language is really in its infancy.

Of course, it's not likely it'll ever be as popular as any of the major natural languages are today, but I wouldn't rule out the possibility of hundreds of thousands or even millions of people speaking it (or a future offspring of it) one day.

Every one of your listed false predictions held within it an inherently great value to prove wrong, a value so enormous and ungodly that anywhere from thousands to billions of spirited men and women pushed everything else in their lives aside for that one cause. They made that choice to give their lives singular purpose because the importance of the goals they pursued necessitated nothing less, and what they knew to be potentially massive benefit to all mankind outweighed everything else in their lives.

This goal of millions of speakers of Lojban would require that same monumental level of collective human effort and perseverance, but it does not necessitate anything of that kind to anyone, nor will it ever, because the potential benefit to mankind is just that marginal.

This argument also means poetry is better in languages spoken by lots of people, in countries with rich history, and later poetry is better than earlier one all other things being equal...

All other things are never equal.

Although these don't look much like any particular word in any language, you can see bits of different languages in each of them. For example, prenu has the "per" of English "person" and the ren of Chinese. cukta has the "ook" of English "book", all of Chinese shu(c is pronounced "sh"), and part of Arabic (and Turkish) kitap. vanju is like French vin and Chinese jiu. This makes learning words easier for the largest number of people.

I want to physically reach through my monitor and strangle him for writing that. Words in human languages are composed of morphemes, not letters. It's easier to learn words in foreign languages if you know how to dissect a new word into its morphemes so that you can read the meaning of the word.

By getting a five letter statistical average over completely different langauges, the morphemes of the original words will not be preserved, and therefore you cannot look at a lojban root word, dissect it into morphemes, and then read what it means.

"per" and "pre" are two completely different morphemes in english, the root of "book" is not "ook", it's "bo". "vin" and "van" are two completely different morphemes in French, etc, etc. The lojban words are not similar to their real-language roots in the same way that real words are to each other, instead they're similar based on the mathematical distance between letters, which is completely useless for learning it.

You can sort of figure out the above examples after the fact, after you have learnt the words, with the knowledge of what the word is in those eight languages. But given a lojban word you have never seen before, you have no way of knowing what it means, even if you know the word in all those languages.

I don't think anyone claims that Lojban phonology will help people guess the meaning of its root words. Even if you guess that {klama} means "travel", you probably can't guess that it means precisely (x1 travels to destination x2 from origin x3 via route x4 using conveyance x5). As a native English speaker who's been learning Lojban casually for a few weeks, I find the phonological resemblances helpful mostly as mnemonics - {prenu} is easier to remember for sounding a little like "person", {nitcu} like "need", et cetera. (And remembering the place structures does get easier, there is a fairly consistent rationale to how they are chosen.)

No, but the author of the article claimed that you can see "bits of different languages in them", which you can't. Neither morphemes nor phonemes are preserved, only letters. The method for creating these words make it probable that there will be at least one letter in the lojban word in common with the word in the six original languages, but there seems to be no guarantee that the letter or letters in common are from the actual root of the word, which are the important letters of the word.

Therefore, the way the root words in Lojban are constructed does not make the learning of words easier for the largest number of people, it just makes it really difficult for everyone.

Take your example of "klama". I can see that it shares "l" and "a" with "travel", but how do you construct a mnemonic for remembering it? It shares no morphemes and no phonemes with "travel", and there's no phonological resemblance either, so there's nothing to help you learn it if you are an English speaker, you just have to memorize it as-is.

I'm sure there are many other features of Lojban that are interesting, and I'm sure that compound words in Lojban are pretty easy to learn and construct, but the text I quoted from the article about root words shows a complete lack of understanding of how human languages work and grow and evolve.

It's not instantly clear from e.g. http://www.lojban.org/publications/etymology/etysample.txt what the algorithm was for the computerized part of the process of gismu construction, although I doubt it was as naive as you seem to be implying. In any event, the humans involved also played with the gloss words they were feeding it to try and find good compromises. {klama} does not in fact take the "l" and "a" from "travel" - it comes from (in Lojban's phonetic spelling) the English "kam" (come), the Hindi "ana" and the Chinese "lai".

At worst, learning Lojban vocabulary is no worse than learning any foreign language, and at best it's somewhat easier (which has been my experience) - I certainly don't see how its phonology "just makes it really difficult for everyone".

Compare it with Esperanto: At best it's really easy because the word you want to learn shares a lot of morphemes with some European language that you know, and at worst there's no similarity to any language you know. If you are Chinese it sucks all the way, but if you are Portuguese or Polish or English you can sort of guess what unknown words in Esperanto means, you can guess, you can wing it. It's biased towards European languages that way, and that is something the article about Lojban pointed out that they wanted to explicitly avoid when creating it.

But the result of their statistical average method is that everyone has to look up and memorize every single root word when learning Lojban. It might be easy to memorize them because of the simple spelling and that you know they're all five letters long, and that there might be letters in common with the word in your native language, but you have to do it for each of them.

Noone can look at a Lojban word and guess what it means, regardless of your language background. You have no chance at guessing it correctly, you can't wing it, you can't speak pidgin Lojban, you either know the root word for something, or you don't, and if you don't you have to look it up. If the root words had shared morphemes with other languages, then it would have been easier for some people and harder for others, but since it doesn't share with anyone, it's hard for everyone.

"Oh, cool, a new Clojure library," I thought to myself.

Time to stop coding for a little while.

http://xkcd.com/191/

While Lojban is "logical", human language is not. Or rather, it is, but it follows its own set of (well defined) rules. If Lojban does not follow these rule's, it's not truly learnable, and thus irrelevant. If Lojban _does_ follow these rules, and is genuinely learnable, then it offers little advantage except perhaps a pared down vocabulary that avoids ambiguity and redundancy, and a less flexible morphology and syntax. Assuming the latter, human languages change naturally, in order to facilitate learning and communication, and Lojban would change too, if actually used. Metaphor, ambiguity, imprecision, etc. all are the _products_ of language adapting itself to suit our needs better. So for all the "flaws" of natural languages, they work, and work well, and if Lojban were adopted by people, it would end up looking like a natural language, with all their "flaws". And that's a good thing.

On one of my consulting engagements, I met this guy who decided to try to raise his kid speaking klingon. The kid just showed no interest in it whatsoever, even though that is all that dad spoke to him for 2 years. I think a natural language has to evolve, perhaps something related to how our brains process language.

I don't know how much we can really infer from one kid's dislike of Klingon. It's not exactly a statistically significant sampling.

Fair point, but in the absence of empirical data to the contrary anecdotal evidence supporting the intuitive point of view is something.

Wow, so this is the lifehacker's Klingon?

Reply on news.ycombinator.com