Looking at the definitions for the many gismu, we find things like bakri:

chalk; x1 is a quantity of/contains/is made of chalk from source x2 in form x3

Why limit it to just x1-x3, why not x1-x30? Just joking, but how did they land on these seemingly arbitrary models of relations/actions?

I can construct a sentence about chalk which has many more than 3 parts:

[x1: foo] bakri [x2: made from limestone] [x3: from the beach] [x4: that evolved out of the geological processes from the ocean] [x5: in the form of a slab] [x6: located at the top of a cliff].

Just making stuff up, but couldn't it be said that these gismu/verbs could take way more arguments than they are showing/allowing? Why did they limit them the way they did? What was their thought process or motivation? How did they know the language would "work" with such limitations/decisions on what x arguments to include in a gismu?

Basically, I don't understand why/how they landed on the number and structure of the arguments to each of the gismu, and looking for some insight on how it doesn't break down.

By "break down", I mean like my example, I can show for every gismu more or less arguments than alloted. I can come up with a million other example hypothetical gismu with other argument structures. So why did they land on these as the final ones with their final structure?

  • 1
    Part of it is because Lojban has only five strong vowels (not counting the schwa), which limits the number of FA particles (the ones that label arguments so you can change them around) to five. But even then, most gismu have only two or three arguments, and follow a specific generalized pattern to make the vocabulary easier to learn. Bakri follows the "stone" pattern - x1 is STONE from source x2 in form x3.
    – No Name
    Aug 12, 2023 at 18:40
  • 1
    Also, chalk is gypsum - not limestone - and always "evolved out of the geological processes from the ocean", so those two arguments would always be the same and thus clutter the structure. And location isn't really fundamental to an object in the way source and form are and if you really need it, has it's own gismu ("stuzi", I think)
    – No Name
    Aug 12, 2023 at 18:50
  • How can I improve my answer?
    – Corbin
    May 15 at 21:49

1 Answer 1


First, it's worth pointing out that some selbri, notably {du}, do permit more than five arguments. Five is the maximum for gismu due to the FA series ({fa}, etc.) which only has five members; if gismu routinely accepted more arguments, then they would not be easily addressable.

Lojban is meant to combine six different source languages, each with their own cultural history, into a single synthetic vocabulary. As a result, we have a ready source of parameters from each source language. When a source language contributes a verb for a gismu, it also may contribute a subject, an object, and more; these are candidates for gismu arguments.

That all said, there are two kinds of "break down" or "problem" that we might want to consider, and Lojban has tools for addressing both of them.

First, as you've mentioned, it seems like there are dozens of arguments that we might want to add to a selbri in order to augment the described relationship. We can do that with modal place structures, as described in CLL, chapter 9. The classic example from CLL is {viska sepi'o lo kanla} "seeing with my eye;" the modal compound {sepi'o} "using a tool" is a conjugation of {pilno} "x1 uses tool x2 for purpose x3." The exact semantics of modal phrases are not specified, other than that they involve the particle {jai}; see CLL for nasty details.

The second kind is irregularity within the vocabulary. My favorite example involves animals. There are dozens of words like {danlu} "x1 is an animal of species x2," all with a regular two-place structure, and all are subrelations (in the mathematical sense of "binary relation") of {danlu}. So, for example, {mlatu} "x1 is a cat of species x2" is a subrelation of {danlu}, and {cinfo} "x1 is a lion of species x2" is a subrelation of {mlatu}. However, some gismu have extra places, like {tirxu} "x1 is a tiger of species x2 with coat markings x3," and this causes a lot of sophistry around the nature of big cats. If we were to remove the third place, with {tirxu fi zi'o}, then we obtain a subrelation of {danlu}.

  • 1
    I like how the subrelations for animal-type gismu map onto biological taxa. That is very good design.
    – Fomalhaut
    Sep 28, 2023 at 6:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.