Lojban - a logical language

1. Welcome to Lojbanistan!

Lojban, in Lojban, means "logical language" (the j is pronounced as in French bonjour) Lojbanistan is both an imaginary country where Lojban is spoken and, in practice, the international community of Lojban-speakers.

Lojban itself came out of an earlier language project called Loglan, and it shares Loglan's interest in the "Sapir-Whorf hypothesis" - the idea that the language you speak affects the way you think. Most people who have learnt a foreign language, or have grown up speaking two languages, will be familiar with this idea, having found themselves thinking and speaking in one language or the other because something is easier to say in that language. One of the main ideas behind the Loglan/Lojban project was to create a language which is both highly expressive and as culturally neutral as possible, then see what people from different cultures do with it. To give an example, in most European languages time and gender are very important - you can say "She goes," "It went," "He'll go" and so on, but just to say "She/he/it go," with no particular gender or time in mind, sounds strange. In Chinese, on the other hand, ta qu (he/she/it go) is perfectly normal. In Lojban there are plenty of words to show the time of an action, its length, how it happens and so on - but you don't have to use any of them. If I really wanted to, I could say:

But normally I'd just say: Notice that you can translate the first example into English (more-or-less) but the second one just won't go into English, or most European languages. If you're European and this strikes you as odd, you may have just witnessed a Sapir-Whorf effect!

However, if Lojban only existed as some kind of experiment to test the Sapir-Whorf hypothesis, probably very few people would bother to learn it. In fact, Lojban's other uses have taken over so much that quite a few Lojbanists who don't believe much in Sapir-Whorf still use the language because of its other benefits. Some that I've heard are:

... and so on. Not all Lojbanists would agree with all of these statements - we are not language missionaries - but there's something in the language for almost everyone.

In the long-term, Lojban has more ambitious goals. Some ways that people have suggested Lojban can be used are:

This last point needs some clarification. There are already dozens of "international languages", the most famous being Esperanto. Although Lojban can (and in my opinion, should) be used as an international language, it is not in direct competition with any of these, because it is not only an international language. Of course it would be nice if the whole world started speaking Lojban, but if the United Nations decided to adopt Esperanto or Interlingua or whatever, most inhabitants of Lojbanistan would rush out and learn that language ... and still go on using Lojban for its other benefits.

2. Lojban Words

Constructed languages create words in two main ways. There are some, mainly fictional languages like Elvish or Klingon, which make up all the words from scratch. Most international languages, however, take their vocabulary from existing languages (in practice, European languages) and sometimes modify them to suit the spelling and grammar of the new language - Esperanto is a good example. Lojban falls half way between the two.

The root words (gismu in Lojban) were created by a computer from words in the six most widely-spoken languages in the world: Chinese, Hindi/Urdu, English, Spanish, Arabic and Russian. This is one of the advantages of Lojban - it doesn't give a privileged position to European languages. These are also languages that have had a lot of words in common other languages; for example, French and Italian share a lot of words with Spanish, and although Turkish is not on the list, I found in Lojban a lot of elements of Turkish which had come in from Arabic and Persian (which is quite close to Urdu). And of course English words like "television" and "taxi" have spread all over the world.

All root words have five letters. For example:


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.

There are 1,300 root words. This might sound like a lot, but it is nothing compared to the roughly 20,000 words a native speaker of a natural language knows. Because you can't say everything you want with such a small number of words, Lojban allows compound words (lujvo in Lojban). For example, there is no root word for "nurse", so I created a compound, kurmikce (c from kurji (take care of) and mikce (medic). Lojban speakers have the freedom to create these compound words to express anything they want to say - if it catches on, it passes into the general Lojban vocabulary, and might even make it into the dictionary. The important thing is that somebody learning Lojban doesn't have to learn every lujvo someone makes up, because you can nearly always guess what they mean (if you can't guess, there's a neat little program on the Web which does it for you - remember what I said about computers being able to read Lojban?).

This leaves the cmavo, or structure words. Many languages put their grammar into their words, so that, for example, in Latin you have amo, amas, amat, amamus, amantis, amant. This kind of thing comes naturally to native speakers, and is horribly difficult for the rest of us. Lojban (like Chinese) makes every part of the grammar a separate word. In English you have to change "go" to "went" to put it in the past, but in Chinese you just say qu le, and in Lojban you just say pu klama. This brings us to ...

3. Lojban Grammar

"Grammar" is a word with painful memories for many of us. I learnt German at school, and was constantly amazed at how complicated and illogical the grammar was (not that English is much better!).

Lojban grammar seems strange at first sight, but is actually quite simple. It is based on a system called predicate logic, which states that in any sentence you have a relationship (selbri in Lojban) between one or more arguments (sumti). An argument can be a thing, event, quality or just about anything (quite how you can have a relationship of one argument is one of the mysteries of predicate logic!). To give an example, the English sentence

Robin adores Juliette Binoche
has a relationship (or "function"), "adore", between two arguments, "Robin" and "Juliette Binoche". In Lojban this would be
la robin. prami la julIET.binOC.
or, if you prefer,
la robin. la julIET.binOC prami
(the capital letters show non-Lojban stress for foreign words, the full stops mean that you have to pause slightly to separate the words - anythingelseinLojbancanberuntogetherwithoutbeingmisunderstood).

You might be thinking "Well in that case a function is a verb and an argument is a noun, so why bother with special terminology like selbri and whatnot?" However, in Lojban I might describe my feelings about Juliette like this:

A philosophical / psychological point: some people (such as the philosopher Alfred Korzybski and the psychologist Albert Ellis) have claimed that the English verb "be" has a bad effect on our thinking. For example, if I say "Jim is bad" (or good, or a Communist or whatever) it implies, if only subconsciously, that there is something bad about his very essence, that he always has and always will be bad, and that everything he does is bad. More Sapir-Whorf effects, perhaps. In Lojban, there is a different word for "is" (du) meaning "the same as" (as in "Three and four is seven" or "That is the Eiffel Tower"). To say la djim. du xlali would meaning something totally different in Lojban ("Jim is equally bad"), and even la djim. xlali would be pretty bad Lojban - you would have to say:

... or be more specific and say, for example, le nu la djim. darxi le ninmu cu xlali - "the event of his hitting the woman is bad." Note that cu does not translate as "is", or indeed as anything in English. It is simply there to separate the function xlali from what goes before: le nu ko'a gasnu xlali would be "her/his/its doing-kind-of bad-event," just as le lunra gusni would be "the moon-kind-of light", or "moonlight".

Getting back to my obsession with a certain French actress, if there are no nouns, verbs, subjects or objects in Lojban, how do we know that la robin. la julIET.binOC prami means that I adore Juliette and not the other way round (a nice thought, but not realistic). Different languages handle this problem differently. In English it is done with word order, and when that isn't enough, with prepositions (words like "at", "from", "to", "with" and so on). In other languages, like Latin or Turkish, it's done by changing the form of the words e.g. Juliette'i Robin sever means "Robin loves Juliette", not "Juliette loves Robin."

In Lojban this is built into the meaning of the word. For example, the word dunda means "give", but its full meaning is:

x1 gives x2 to x3

So mi pu dunda le cuska le ninmu means "I gave the book to the woman" not "I gave the woman to the book" (of course).

But enough of grammar for now. The important point is that Lojban has a lot of what we would call "grammar", but nearly all of this is contained in the cmavo (structure words), and you can use as many or as few of them as you want.

4. Lojban is not Vulcan

As I've said, Lojban means "logical language". Unfortunately, in many cultures we have the idea that being logical means having no emotions, rather like Mr. Spock in Star Trek. In fact "logical" is not the opposite of "emotional", but of "illogical" - in other words, saying things that don't make sense.

In Lojban you can of course use "emotional" words like prami ("love/adore") but there is also a special class of words set aside for expressing feelings. Some examples are:


Each of these has a negative - .uinai means "unhappy" - and degrees of feeling - .uicai means "extremely happy" or "delighted". You can also combine the words in any way you like, which means you can invent words for emotions that don't have a word in your native language. For example, to express the feeling in a lot of Turkish pop songs (arabesk), I invented the word .iucai.uinaicai (pronounced "you-shy-we-nigh-shy") meaning something like "I am deeply in love and deeply unhappy."

It doesn't stop there, though. .iu means "love", but "love" can mean a lot of things. If we need to, we can modify these basic emotions. .iuro'i is emotional love, what we most commonly understand by the word. .iuro'a is social love - what you might feel for a good friend. .iuro'u, however, is definitely sexual, while .iure'e is spiritual love, the kind of thing mystics feel, maybe. You can even have .iuro'e - mental or intellectual love - if, for example, you had a passion for physics.

Lojban also has a lot of words for handling the general business of conversation. Some examples are:

Because there are no native speakers of Lojban (well, not yet) there are very few rules for Lojban conversation and writing. Anyone who has tried speaking in a foreign language will know how easy it is to be rude without intending to. In English, for example, you have to choose between "Can you come over here?" "Could you possibly come here, please?" "Come here, will you?" or even "Hey you!" Lojban simplifies this considerably by separating the emotional and social content of a sentence from its literal meaning. I can just say:


or I can make it polite by saying pe'u ko klama ti - I don't have to disguise it as a question about your ability to come.

5. The State of the Art

Lojban is a fairly new language, and it is possible that as it is used more, and used by more people, some parts of it will change. However, to avoid the endless discussions about how to "improve" the language that have plagued some other constructed languages (including Lojban's predecessor, Loglan), it was agreed, after an a long initial period of developing and testing Lojban, that absolutely no changes to the language would be made for at least five years after the publication of The Complete Lojban Language (which eventually came out in 1998). Because the language was tested and debated so much before this "baseline", it is also unlikely that any major changes will occur after that period - what will probably happen instead is that people develop the language by creating new lujvo (compound words) and exploring some of the less-used aspects of Lojban grammar. There are also a few areas of the language that have been deliberately left open for future developments.

In addition to the published The Complete Lojban Language, there is an HTML version of an earlier draft of the book, which is essentially the same. Other materials available from the Lojban website are wordlists, a draft textbook, and a large number of Lojban texts, both original and translated. There is also a small but rapidly developing body of Lojban software. The main plans afoot in Lojbanistan at the time of writing this are:

If you feel motivated by any of this to learn more about Lojban, please check out the Lojban homepage at http://www.lojban.org . There is also a Lojban beginners' course.

co'o mi'e robin.

4/3/99