Skip to content

Instantly share code, notes, and snippets.

Last active August 12, 2023 05:33
Show Gist options
  • Save IGJoshua/fcaa1a1d271d30239390d5a98ffdef16 to your computer and use it in GitHub Desktop.
Save IGJoshua/fcaa1a1d271d30239390d5a98ffdef16 to your computer and use it in GitHub Desktop.
The lojban I speak

The lojban I speak

coi ro do mi’e la saski’o tu’a dei cu ve ciksi tu’a lo me mi moi jbobau


I’m a fluent lojban speaker who has many opinions about the language and how it should be spoken. The lojban that I speak is a collection of ideas from many places brought under one roof, unified under a few core guiding principles:

  1. My espressions should be practical to understand, and interpretations should be more or less obvious to anyone with experience in the language
  2. Any grammar that I use should have its meaning be understandable without explanation for anyone who knows zantufa lojban, and relatively easy to explain to those who don’t
  3. The logic of utterances should be evident by their structure, although I’m not too rigid with 1:1 correspondences to formal logic
  4. The language should be practical to use in everyday speaking, common things which are intended to be said shouldn’t require circumlocution

With these goals in mind, here is a quick overview of the basics which will get you most of the way to knowing how I speak, without the longer discourse:

  • I use sotyge’a lojban, a variant of zantufa lojban
  • I use gadganzu gadri
  • I use singular logic, but as gadganzu lacks sets, masses replace them


zantufa lojban is a dialect of lojban that seeks to make the grammar more beautiful and to ensure a strong logical foundation for the language. It makes many changes, but the three which are most likely to affect my speech are its inclusion of something very similar to the simplified connectives system, n-ary GI (with GIhI as the new terminator), and the creation of a mulno jufra as a new concept in the language.

mulno jufra

A mulno jufra is a “complete thought” more or less, and is most closely matched to a sentence in English. It consists of several bridi which are connected together with logical, non-logical, and bo-grouping style connectives, all share a single prenex, and are terminated with iau.

.i da zo’u da broda de .ije da brode iau ge’e

This alters how some cmavo behave, in that many cmavo which previously introduced a sub-bridi now introduce a sub-mulno-jufra, minus the iau terminator.

.i da poi broda .ije brode ku’o nu ko’a brodi .ibabo ko’a brodo kei bu’u ko’e

This increases the logical power of NU, POI, and similar selma’o to match that of the full lojban language, without requiring constant use of forethought connectives.


sotyge’a is a dialect of lojban that seeks to reduce the number of selma’o used in lojban, allowing new grammatical constructions to become valid, most often with somewhat obvious interpretations.

The most significant changes it makes which will affect the readability of my statements is that it merges many selma’o into BAI, including LAhE and FA. lu'u becomes the new BAI terminator, and BAI gain the ability to be composed together, as in zantufa lojban. This changes the grammar of BAI to allow it anywhere a sumti may be used, not just a full term.


gadganzu is a modification to the gadri which aims to make them all useful again, foregoing xorlo’s “use lo for everything” mentality. It includes a scale for the gadri to exist on, from more specific to less.


All the gadri fall into one of the following categories:

These gadri describe sumti which are being designated by contextual clues being given outside of speech, often a gesture. These are more or less analogous to ti noi ....
These gadri are used to refer to sumti which are already present in the conversation, although they need not have been named before, only implied by conversation. e.g. .i mi ctidu’a fi lo mi mlatu .i le’e se mlatu cu barda
These gadri refer to specific instances of a thing. If I say le mlatu then you know that not only am I talking about a cat, but I have a particular cat in mind.
These gadri are used more commonly than the rest, and based on context will always resolve into either a Knowable or General sumti. They do not have semantics separate from those two categories, but simply note that it may be either, based on context.
These gadri function nearly identically to their CLL equivalent, speaking about “things which broda in general”, while not refering to any instance in particular. e.g. lo'e mlatu cu nelci lo ladru indicates that “cats like milk” in general, while not stating if a particular cat likes milk or that liking milk is a part of what makes something a cat.
These are likely to be the least used gadri, and make a claim that all things which it describes must definitionally fill the bridi they are contained in. e.g. lo'i mlatu cu nelci lo ladru states not only that “all cats like milk”, but that liking milk is a part of what makes something a cat, therefore implying that anything which does not like milk must definitionally not be a cat.

Additionally, each gadri is either a Massified or Individual gadri. These create sumti as described in the CLL, with massified gadri creating mass sumti, and individual gadri creating (potentially plural) individual/distributive sumti.

Note the lack of gadri for sets. Sets have no useful properties which masses do not already posess, and masses are more useful for conventional conversation, meaning they were kept as-is.

Outer Quantifiers

Outer quantifiers are something of a hotbed topic among lojbanists. Most want to see them be useful, but at the same time most logical interpretations of them give nonsense semantics. In the lojban I speak, outer quantifiers more or less match what the CLL indicates they should be, disregarding the update by the BPFK.

While I don’t particularly like logical expansions in many cases, this may be useful:

PA SUMTI = SUMTI poi lu’o ke’a PAmei

While this meaning isn’t exact, it may assist with understanding.

Singular Logic

Non-Exclusive Numbers

How poi'i Gives tu'a and jai Superpowers

poi'i is a word in the selma’o NU. It allows you to take a complex bridi and extract a sumti from it.

lo poi’i ke’a mamta cu klama le zarci

The mother goes to the store.

As you can see however, there is a ke'a under the poi'i clause. This is the sumti which is extracted for use. It can be nested arbitrarily deep.

lo poi’i mi ta’e klama lo stuzi be lonu mi’a kelci ke’a noi volbolo’yselkei cu te jivna mi lo mi bruna

I often go to where we play volleyball; volleyball is a competition between myself and my brother.

This is much harder to translate into english while maintaining the structure, but hopefully it is somewhat clear that the x1 of te jivna is being filled with the game of volleyball. I’ve given a lot of additional information about it however, stating that I habitually go to the place where I and others play it.

Since poi'i is in NU, that means that grammatically speaking, it is an abstraction. The words tu'a and jai are both used to allow you to refer to an abstraction without specifying what it is, the classical example being this:

mi troci tu’a lo vrogai

I try the door.

In this case, we are creating an unknown abstraction related to the door. However, there aren’t many things you would normally try with a door, usually just opening or closing it, so the meaning is pretty clear. With that in mind, it may expand to something like so:

mi troci loka ce’u gasnu lonu lo vrogai cu co’a kalri

I try to cause an event of the door beginning to be open.

Based on this, we can see that tu'a expands into some usage of a gadri paired with a cmavo of the selma’o NU. This allows us to come up with other examples, like so:

.i mi’o ta’e simxu co skami tavla .i mi benji lo lisri tu’a do

We habitually mutually speak with computers. I sent a story to something related to you.

This implies that whatever fills the x3 of benji is going to be an abstraction, and many lojbanists separate poi'i from the other abstractions because while it may grammatically be in NU, it can still produce something which is an abstraction or something which is concerete.

In the lojban I speak, any value that poi'i might refer to is permissible as an intended value of tu'a. That means the above example might expand to something like this:

.i mi benji lo lisri lo poi’i ke’a skami jecu se ponse do

I send a story to that which is a computer and is possessed by you.

This means that tu'a no longer refers only to abstractions, but becomes a general operator meaning “something related to this”.

jai as a word also functions very similarly to tu'a in some cases; specifically if jai is placed immediately before a selbri, then the x1 of that selbri is considered to be as if it were wrapped in a tu'a, which means jai gains the same benefit of becoming a general operator as tu'a does.

Relative Numbers


this is gonna be about how za’u etc are related to xo’e and not pa

Common Idioms

Experimental cmavo

Copy link

tsani commented Jun 11, 2020

tu'a gaining super powers has more to do with how one interprets the word "abstraction" in its definition. Does it refer to a genuine abstraction (the things produced by nu, ka, du'u) or does it mean something that is grammatically an abstraction? In the time I've been a lojbanist, it's been understood that the purpose of tu'a is to raise a concrete object to a genuine abstraction. lo poi'i ke'a skami, although grammatically containing an abstraction, does not refer to an abstraction; it's the same as writing lo skami.

To me, you're reinterpreting tu'a as zo'e ne and justifying it by having a different interpretation of the word "abstraction" in its definition. I think it's reasonable to propose that tu'a ought to be interpreted as zo'e ne, but I think that the argument based on the existence of poi'i isn't compelling. I think that what are more compelling are your examples together with the idea that shorter ways of saying things ought to be more vague.

Personally, I'd be okay with:

  • tau is interpreted zo'e ne
  • tu'a is interpreted as always. That is, it produces a genuine abstraction.

But I'm more okay with the status quo of:

  • tu'a behaves as always
  • zo'e ne is used for the purpose you want

P.S. I find writing lo poi'i ... to be an unnecessary use of experimental cmavo, given that lo is defined as zo'e noi which gives you a ke'a. I think that the places where poi'i is really solving a problem is when making a tanru out of a poi'i-clause or using poi'i with something other than lo.

Copy link

There's a couple comments I have about this response.
First, tu'a is entirely unnecessary in the first place, seeing as it only serves to act as zo'e ne/pe in places that take abstractions, and it only disambiguates things like lo barda from tu'a lo barda where both are used to fill in an event (am I talking about a large event, or an event related to the large thing?). We could very easily strip it from the language with no change to semantics, and the only disambiguation it has is only ever useful in situations where a predicate can apply to both dacti and su'umdu. If we permitted arbitrary sumti raising that wouldn't change, it's just an extra required word in many cases.
Second, this article shows my thought process as much as it shows a particular view of the language. In my mind, the reason tu'a and jai have been expanded to function similarly to zo'e ne/pe is because of the inclusion of poi'i. Before I was aware of poi'i I did not want nor accept this expansion of usage of tu'a.
Third, there are many things which the BPFK asserted about the language that I disagree about. This article will in time explain most or all of them, but one part of it is that I do not believe that either poi or noi (or voi) is sufficient as a primitive for expanding lo (or other gadri), I consider the gadri to be something of primitives and do not like nor use the expansion to zo'e noi.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment