Lojban is a logical language without a logic or: how I learned to stop worrying and love the bear goo

cross-posted from: https://hexbear.net/post/3947641

Hii, idk why I’m writing this cuz there’s probably 0 Lojban speakers here but… idk you may like my special interest writing lol and i just had this urge to start writing this. Also wanna preface this by saying this is mostly composed of my opinions (which I will try to justify ofc) and I’m just some unqualified nerd

Soo, for the last few weeks I’ve been learning this constructed language called “Lojban”. The name Lojban is technically an exonym created by English speakers and some Lojban speakers will use it even while speaking Lojban but a better “native” name is probably “jbobau” or “banjubu’o”, the latter for use if someone is pilled on algorithmically deriving Lojban words for countries, cultures, and languages from their ISO codes (I think I am lol, but I think it’s also okay to have one of our own words for ourselves and use the algorithmically derived stuff for everything else).

Lojban is a in class of constructed languages called “logical languages” and the usual definition is that they are languages for people to communicate anything a natural language could but they are based on logic. Lojban itself comes out of a previous language project called “Loglan” (yes that’s literally just the english words for logic and language mashed together, Lojban’s name is similar but has a better justification you will see later lol) after its community became disillusioned with the slow progress of its development and that the specification did not even come close to reflecting its real use (not to mention being intended to just encode English with logical structures which many people were also dissatisfied with). Loglan’s creator, James Cooke Brown, was actually threatening people with legal action for trying to improve the language by making dictionaries or specifying grammar that they were already using. So a lot of Loglan’s community left the project and formed the “Logical Language Group” (the organization that actually defines what Lojban is) and created Lojban to try to improve on Loglan. For a while there was some kind of complicated legal battle over copyright between the LLG and James Cooke Brown, which the LLG won eventually but not before they had already created a new vocabulary for Lojban from scratch to avoid copyright restrictions.

It’s weird but I remember hearing about Lojban as a kid cuz I think I was briefly interested in constructed languages (I even had a little one of my own I can barely remember) and cuz I was allowed unlimited access to the internet basically since I could work a computer on my own lol. I was talking with a friend from here and the topic of languages came up and we talked briefly about if a language that people speak could also be used for programming a computer or something and I vaguely remembered Lojban. So that night I looked into it again, for real this time, and discovered it’s actually rly cool and I wanted to make a serious attempt at learning it.

Lojban has several “claims to fame”, and not all of them are true sadly and I feel like this isn’t made clear enough to newcomers. It’s very cool and I love it ofc lol, but there are some things about it that I don’t like and which I only learned of after I got a decent grasp on the language and the project.

Quick intro to Lojban

In Lojban, the basic structure of a text or any speech (“utterances”) is a series of “bridi”. A bridi is like a Lojban sentence but the word “bridi” literally means “predication”. Like in classical first-order logics and like many English sentences (“I like you”, I have the relationship of “liking” you), a bridi is a assertion of a relationship between some things. Bridi are composed are 2 things: a selbri, (a predicate word or basically like a verb) which has a series of “places” numbered x1 through x[however many places the selbri has], and sumti which are like arguments or parameters that fit in the numbered places of the selbri.

So “I like you” can be rendered in Lojban like this: mi [cu] nelci do

“nelci” is a word being used as a selbri which has the English definition of “x1 is fond of/likes/has a taste for x2 (object/state)”. On its own it’s not very useful ofc, unless you have sumti to pass to it. I used the words “mi” and “do”, which are like pronouns for the speaker and the listener, respectively. Mi and do are considered to be in a category of word called “cmavo” [SCHMA vo], which literally means “structure word” and which are small, very commonly used words used for lots of things like filling in the sumti places of selbri. There are a few other pronoun-like words defined by Lojban but those aren’t very useful unless we only want to make statements about ourselves! Btw, there is another cmavo called “cu”, it means that the word after it is a selbri in a bridi, the reason it’s in brackets is cuz you often don’t need to specify it. In this case we don’t cuz it’s obvious where the selbri begins and ends.

Here’s one of the most wild features of Lojban that you get to learn about almost immediately when you start learning about it. There are no nouns :3

What you have to do if you want something with the function of a noun to use as a sumti to a selbri is to use a few different mechanisms that are provided for by the cmavo, most commonly the cmavo “le” and “lo” which are similar to the English word “the”. They convert a selbri to a sumti like this:

le gerku cu nelci le panka (The dog likes the park, or extremely literally: all of the at-least-one described as a dog liked/likes/will like (tenses are optional :3) all of the at-least-one described as a park)

Notice in this case, the cu was necessary cuz otherwise it would be ambiguous where the first sumti to nelci ends. Lojban LOVES terminators to grammatical structures. Le has a corresponding terminator like a close parenthesis “ku”. I could have also replaced the cu with a ku (the former is pronounced like shu btw, ku is pronounced like any English speaker would pronounce it hehe) and it would have also been valid. Most of these terminators can usually be dropped where dropping them wouldn’t cause any ambiguity. It seems complicated at first but you do actually get a feel pretty quickly when you can leave out the terminators.

Also notice that “gerku” and “panka” look very similar to nelci. That’s cuz they are the same type of word called a “gismu” :3 and you can tell easily where the gismu in a Lojban utterance are cuz they have a defined, regular form. They always have 5 letters, always start with a consonant and end with a single vowel, always contain exactly one consonant pair, and they are are always stressed on the first syllable (which is a consequence of the fact that ALL Lojban words are stressed on their second-to-last syllable).

Gerku and panku literally mean “is a dog” and “is a park” (they have more places for more info but yeh hehe)

Intro to xorlo and bear goo

Oh, you thought the above section was me explaining a language I love so that you could maybe understand it a little better? Actually, it was all a trick to give you the context to just barely grasp the biggest Lojban struggle session that ever happened enough to understand how fundamentally cooked this “”““logical””“” language is from a logical perspective :3 (/j)

Actually, I lied in the last section. “le” and “lo” don’t work like that anymore. Lojban underwent a period of rapid development within the LLG from its initial creation until the publication of the official specification of the entire language in a red book called “The Complete Lojban Language” in 1997. Then development was mandatorily frozen for 5 years to see how people used this “baseline” standard language. And since then, the development of the language was left to a different committee of LLG members and Lojban users called the BPFK which was dissolved by LLG in 2018 to form the very similar committee called the LFK (I will spare you their Lojban names :3). Technically, LLG has to approve whatever the BPFK and now the LFK do, but the vibe I get is that most of the LLG members who developed the language either left or mostly checked out of further development of the language. I think the LLG bylaws say they have to meet once a year but… not a lot has happened since 1997 tbh lol

One of the BPFK’s first language reforms (maybe the first actually) was a reform called “xorlo”.

You know those meanings of le and lo I mentioned earlier? Something that was very important in baseline Lojban was this concept of those le and lo words (called “gadri”) having implicit quantifiers.

Like when I say “le gerku” how do you know how many dogs I’m referring to? The answer, in baseline Lojban, is that le has 2 implicit qualifiers which, if not overridden, are this: “ro (all of) le (what is described as) su’o (at least one) gerku (dog)”. So, to be exactly clear, what I said was that one or more dog(s) that I’m thinking of like/liked/will like one or more park(s) that I’m thinking of lol. It can’t be zero cuz that affects the hypothetical truth value of the bridi (more on that later) cuz there has to be an actually existing liker and liked in order for that bridi to be true. Another important thing is, why did I use le instead of lo? The answer, also in baseline Lojban, is cuz lo refers to things that actually are something, as in: my use of the selbri after “lo” is not restricted by my intentions. That probably sounds like nonsense lol. Cuz it is unless you know the secret thing that no one told me initially and which a lot of other people never learned about le and lo which is that they also specify sets. Before I tell you more on that, I should tell you that lo also has implicit quantifiers like this: “su’o (at least one) lo (things that actually truly are) ro (all) gerku (dog)”. So, on set specifying: “le gerku” is ALL MEMBERS of the set of dog(s) THAT I AM THINKING OF and “lo gerku” is at least ONE member of the set of ALL dog(s) IN THE UNIVERSE (of discourse, maybe, lol)

Here is a very good example of the semantics of that from the red book I mentioned:
Example 6.41.

[ro] le ci gerku cu blabi
[All-of] those-described-as three dogs are-white
The three dogs are white.

Example 6.42.

ci lo [ro] gerku cu blabi
Three-of those-which-are [all] dogs are-white
Three dogs are white.
Okay, cool, cool, seems easy right? Just have to remember those implicit quantifiers

Wait… Lojban is a logical language. Speaking extremely literally, “lo gerku” is those things which dog, as in things that are dogging. Wtf is a thing that dogs and also… wtf is a set?

Okay okay, what about the gismu “cribe” which means “is a bear”. Imagine: you’re in a forest with a friend. You and your friend are trying to find lo cribe (a bear, one or more members of the set of things that are bearing). You’re walking along and you come across a pile of bear goo (a bear that is a pile of goo, maybe it’s a bear in a late stage of decomposition or something). Did you and your friend just find lo cribe???A???A?A???A/fd/ffd/fdflfmf.f.c? If the goo bears, doesn’t that mean it is part of lo cribe?

The problem is, that from a classical first-order formal logical perspective… human intentions are worthless. But more than that LOJBAN HAS NO COMPLETE, FORMAL DEFINITION OF ANYTHING EXCEPT ITS GRAMMAR AND IT ISN’T EVEN CLEAR ON WHAT A SET IS WHILE USING THEM

So the BPFK nerds argued about this endlessly in one of their sessions and could not come to any kind of conclusion on whether bear goo is lo cribe or not so they decided to drop almost all the semantics described above of le and lo despite the fact that no human speaker could come across bear goo and describe it as lo cribe rather than le cribe (as in goo that is bearing in speaker’s idea of the goo) unless they were trying to make a point about it being lo cribe to a committee and we got the “xorlo” reform. Now you use lo for everything unless you have some specific things in mind and then you use le. And these have no qualifiers unless you add them so “lo gerku” could mean a dog, all dogs, or some dogs and “le gerku” means some unspecified amount of things I’m thinking of, each of which I describe to you as a dog.

Btw, I didn’t even get into the plural quantification thing, as in there are predicates that can only be satisfied by groups of things and not by the things themselves and vice versa. Like one person cannot “gather” and a group of people cannot eat, the people in the group each eat

If you wanna melt your brain some more with that, here you go: https://mw.lojban.org/papri/gadri:_an_unofficial_commentary_from_a_logical_point_of_view :3

Also, Lojban has set and “mass” gadri like le and lo that do explicitly specify masses (a group of “individual” (that has its own bad definition) things that has the same properties of the individual things which compose it and may have other properties as well) and sets (a group of things that has entirely different properties from the individual things that compose it). Like one way to interpret that is that sets of bears have cardinality but they don’t have “bearness” and vice versa for sets) which, as a result, can be used to specify the stuff about plural quantification of variables in predicates (see the above link for info about “collectivity” and “distributivity”)

If this all sounds really confusing, that’s because it is and I’m still trying to figure out what Lojban means to me. Initially I was using xorloified Lojban cuz it’s an official reform (but technically not a “finished” official reform or something that is permanently binding lol, idek, it’s been 20 years lmao) and I hadn’t yet realized Lojban is in the process of splitting into multiple dialects cuz, to me it seems like, people who cared about the logical aspects of Lojban are now dissatisfied with the lack of progress on better formalism and the whole language actually and xorlo and similar BPFK reforms that make the language easier to speak while reducing its semantic richness and the people who don’t really care about logic in Lojban hate that stuff anyway and are cool with dropping most of it. I read some of the arguments for and against xorlo and I am using non-xorlo Lojban for now cuz I wasn’t convinced by the bear goo arguments. It doesn’t matter that much for communication most of the time, I feel like, usually, xorlo people will just miss out on some of the meaning of what a xorlo non-user will be saying but otherwise it’s okay. Although, sometimes I legit cannot understand wtf IRC users are saying at all lol cuz their dialect is so far from the baseline Lojban I’ve been learning and also I cannot make their utterances parse sometimes in the computer parsers cuz its so far away from baseline lol

Lojban myths and truths

Lojban is a logical language

This is true, as long as you can agree on what lo logical language ku means :3 Tbh, when I went in I rly was expecting SOME kind of real formal system behind it before I found out that Lojban is a language to speak some logic but not rly a language to do logic, I guess. Most of the parts are there but not much has materialized (yet?)

There are some attempts to formalize Lojban more, this new one is particularly impressive tbh, even though I think the approach is kinda wrong: https://brismu.systems/ (brismu: a relational interpretation of Lojban)

Lojban is an unambiguous language

This is true syntactically. There are real formal definitions of Lojban grammar out there based on BNF grammars and parsing expression grammers which is VERY VERY VERY NICE lol. It really helps to learn the language cuz if you’re doing something grammatically wrong a parser will reject it. Semantically, I think you can be even more vague and ambiguous in Lojban (like with tanru, which are metaphors based on combining multiple gismu) than probably most natural languages but you can also be painfully, beautifully non-ambiguous with your meaning :3

So… it’s mixed

Lojban is somewhat culturally neutral

It’s hard/impossible for me to say exactly cuz I was a monolingual English speaker before learning Lojban and I live in which I think is where most of Lojban’s creators are from. It does have some very nice features though. Like there is no set word order, I didn’t mention it before but you can move around the sumti places and selbri in pretty much any configuration you want so likee if you don’t speak a SVO language like I’m speaking now you can totally just do whatever you want and still be understood. Also, the parts which make up the gismu which form the root words of the language are algorithmically sourced from the world’s biggest languages weighted by how many speakers they have. So Mandarin is the most important source for the gismu

But at the same time Lojban’s phonology (how it sounds) and especially orthography (how it’s written, like with the Latin script) is quite European, although there are much less used writing systems for Lojban floating around. I should mention though that the language, while it has a very European phonology, was carefully designed so that speakers who might have trouble making some of the more difficult or culturally specific sounds can still be understood. Tbh that would be worth a post of its own but you can find details in the first few chapters of the Red Book if you’re interested

On perfection

Am getting tired now so I’m gonna finish this up before I get too tired to post it and then forget it but…

I don’t want anyone to put off of trying Lojban out cuz of the bear goo stuff or anything else I mentioned. It’s genuinely a very cool language project, it’s just unfinished imo even if most of its original creators are done with it

There are a lot of things in Lojban I wish we had in English, like spoken tone indicators, evidential indicators (cmavo that attach to things to specify how you came to know or believe something), actually working grammar checkers, etc etc etc

Would encourage you to look into Lojban if any of this stuff interests you. You’ll just have to figure out some/a lot of it on your own but if you figure out a good interpretation maybe we should all start speaking it hehe :3

I hope you liked reading, sorry if it’s messy I kinda just threw this together, little energy for more elaborating rn