kala Asi

Pondering XerographicPaper's video

In June 2025, XerographicPaper made a video titled “Why I DON’T Like Toki Pona”, which sparked discussion on the flaws of Toki Pona (some solid, some shallow). I watched it as part of my sin kulupu reporting.

A lot of the ideas in the video spark conversation topics in my head, but the comment section is already positively full of people like me, and I don’t feel like joining the ruckus. Nevertheless, unless I put my thoughts into words, I won’t be able to clear my mind, so this page is going to be my infodump. I don’t know who’s ever going to read this, but welcome.

Tokipono ne kreigxis por priparoli malsimplajn aferojn. Gxia kreinto diris tiom.

Okay, the fact the video starts with espeak-ng voiceover was really funny, props for that. Double props if this was done through espeak-ng itself and not through some wrapper around it, but who knows!

While people who are skilled in the language tend to give it glowing reviews

It’s a good time to ask two questions:

1. Am I skilled in Toki Pona? Eh. Short answer, I’d give myself a B1-B2 on the CEFR scale.

Long answer, I’ve been around since 2019. My biggest strength is comprehension and slow, deliberate phrasing. My biggest weakness is reading large blocks of sitelen pona. I only practice oral speaking with myself (same with English, actually), so I can’t truly evaluate myself on those.

I have listened to most of kalama sin (it was once difficult to listen to and keep up, not anymore), read some of lipu tenpo (my attention span is too short to truly binge-read it), but most of my exposure has been through lurking as people talk. I have binged most of the toki pona music there is, including actively pondering the lyrics. I have translated songs in the past, and now turning my attention to writing original prose.

I certainly know people who are more skilled, but I dare say their count is in the double digits.

2. Do I give it a glowing review?

N/A, I think. Languages are just what they are. I can speak Russian not because I like it, but because of where I was born. I can speak English not because I like it, but because of gestures broadly. I have ~A2-B1 German skills not because I like it, but because I was repeatedly given the opportunity to drill my skills.

I was introduced to Toki Pona because jan Misali liked it, sure. I learned it because learning is fun. I continue to use it because it’s something to do alongside people I like. (Also as a linguistics student, having deep knowledge of an academically understudied language is a treasure trove.)

Similarly, I don’t try to sell Toki Pona to people who come to the course I wrote, Wasona. If someone has decided they want to learn it, here you go, feel free to start. If not, have fun doing other things. It’s fine.

Toki Pona is a language which contains between 120 (Toki Pona: the Language of Good) and 181 (nimi.li) words depending on the source.

Through being the Linku pollster, I am kinda the source of that 181 figure, so I’ll add some caveats.

181 is the current cutoff between Linku’s obscure words and sandbox words, defined at 5% of users reporting they use the word. That cutoff is, of course, arbitrary; if we lower it to 0%, we might as well state that Toki Pona has between 120 and +inf words. But that’s even less helpful to understand the realities of Toki Pona you’ll see in the wild.

I’ve previously crunched the numbers, and the mean number of words that a person reports to use is ~135. Put another way, for an average Toki Pona speaker, 135 is the size of the active vocabulary. If we want to include like 95% of speakers in this, I recall a range of like 120-150. Unfortunately I am just remembering this off the top of my head.

In teaching an “averaged, unmarked nasin”, both I and seemingly many other proficient speakers have arrived at the 60% boundary, which today stands at 131 words.

Ironically given that I’ve just written several paragraphs about this, I think overall the community thinks too much about the word list, and we will all benefit from some chill and acceptance of acceptable ranges.

So, in order to let you judge for yourself, I translated the entirety of tu kuntu into English. [in the description] I LIED A LITTLE BIT though its not complete

tu kuntu is probably one of my favourite niche literary pieces I’ve ever read, given how many factors needed to collide for someone to enjoy or even comprehend the jokes.

The translation is fine! Though it’s hard to judge reading comprehension skills by tu kuntu. It’s relatively long, and full of 2020ish in-jokes, but all in all it’s not very complex. I’d love to see more back-translations in the community.

It taught me how to learn languages, and for that it has value.

nice!

… discard unimportant information. So what does that look like in practice? Well, one example that comes up a lot is fruit. If there are a bunch of fruit I can tell you which one I want by describing it by its unique traits. Let’s say I want the banana, and it is the only curved yellow fruit available. Then I can just say I want the curved yellow fruit.

Funnily enough, to me this looks like a good example of not discarding unimportant information. Why does it matter the fruit is curvy? Does that affect your enjoyment of the fruit? I would change ‘unique traits’ to ‘unique relevant traits’. This might seem like a pedantic point, but…

Toki Pona does not have a general word for animals

Yeah! When Toki Pona doesn’t have a semantic prime for something, it’s challenging you by suggesting that there is no common everyday relevant feature that unites this class, and excludes all other concepts. Put simply, that when we say animal in English, most of us don’t actually care about “the set of related living beings that is narrow enough to exclude fungi but broad enough to include corals”. We are probably conveying a related but distinct property, like the ability to move about, or having four limbs, or having feelings, or being fun to pet, or whatever. Toki Pona says, well then, talk about that property instead. Don’t be indirect.

(Sometimes, you will actually talk about “the set of related living beings that is narrow enough to exclude fungi but broad enough to include corals”. And you’re not prevented from doing so! It’ll just take a lot of explaining, quite similar to how annoying this sentence is in English.)

It’s not hard to find people polling about whether or not X creature is this or that

Yea! I think most people take this as a fun philosophical exercise, not as an actual problem in need of solving. Think less “is the trinity of one essence”, and more “is cereal a soup?”. Sometimes discussion can actually be about the imperfect overlap of those features that make up the idea of a soweli, a waso, a kala, etc — then you start seeing full sentences exploring each of these ideas.

lipu vs supa; ko vs kiwen; linja vs palisa

As I said before, I speak the language that I speak, and I don’t need it to be the bestest at anything. But the matter of “optimising the design” is still interesting as an insight into Toki Pona’s past. Using that lens, I think XerographicPaper makes a good argument for linja vs palisa — some semantic spaces could have been arranged differently and made for a more compelling design. How exactly, I don’t know, but it’s fun to think about, and someone could totally make an interesting personal project out of this.

lipu vs supa, though, imo is simply a byproduct of people really wanting to rationalise things that never really needed it. lipu is for holding sona, supa is for holding stuff. Shape is not super relevant. When things are sometimes lipu and sometimes supa, it’s more to do with how they’re used, than how they’re positioned in space. This tidbit will be fun to throw to the community and see if they feel different.

Geometry

So I don’t particularly care about non-euclidean geometry, so I neither watched jan Telakoman’s video, nor want to dwell on it here. I think making a video on just the basic idea of Euclidean space was indeed sufficient. If they wanted to continue the topic and get into actual maths problems, I think they could; they would have to start introducing notation, however, much like the page shown in the end credits of the video. I think we can all agree that end credits page is completely uninterpretable to a layman precisely because it lacks the context where that notation and terminology is built up, much like it would be in Toki Pona.

And I don’t think Toki Pona by definition abhors all notation! nasin nanpa pona is a good example of notation that people actually felt they needed and that achieved broad enough adoption.

Toki Pona is the brainfuck of conlangs

One of the defining features that define the vibes of brainfuck is that any reasonable length program in some language becomes a monstrously long string in Brainfuck. Having done translation to/from Toki Pona, it doesn’t feel that way. The translation is usually similarly-sized, while still conveying the same overall idea. Cause yeah, a translation will drop some and will add some information from subtext. That’s fine.

English use is much less common in the Esperanto Discord than in ma pona pi toki pona. Often proficient speakers choose English to speak with each other.

Is using Toki Pona more effort for me than using English? Yes. I am sure my fluency in Toki Pona is lower than in English. Do I care to make myself native-level fluent? No. How many people do, and will toki pona taso 24/7? A non-zero amount, but not that many. “Is Toki Pona too difficult in practice for serious general use?” I leave it to the jan Tepos and jan Kitas of the world to decide.

(Note: this doesn’t mean that everyone who’s been around <5 years isn’t fluent. I know that people who stuck around tpt-communities, particularly VRChat, get good much faster than I did, which is only natural given a much higher dose of immersion and practice.)

A good argument to make here would be: how do Esperanto and Toki Pona compare in terms of the number of speakers who actually speak the corresponding language 24/7 in a given community, for extended periods of time? I bet you can use this method to show Toki Pona far below Esperanto. Can a survey like this be conducted at all? I don’t know, but you might. (And does it even matter?)

Meanwhile this server where only Toki Pona is allowed is almost a ghost town by comparison.

It is a ghost town! That’s because ma toki pona isn’t actually where toki pona taso thrives right now. You’ll mostly find it on ma mun on Discord, as well as in the VRChat community.

The zines themselves are entirely Toki Pona, good, but on the site a couple of things, including this bit which addresses people who would already be able to speak the language very well, only in English.

This is primarily an issue of available volunteers! There is an open issue for i18n. You can still raise the question, why is English the default in most sites? I think the prevalence of beginners compared to proficient speakers. (Which, again, if you can think of a good way to measure, both communities would benefit!)

lipu tenpo itself was seemingly also only created to prove a point and help get Toki Pona an ISO code.

And yet it’s still going! Rare achievements can be a good way to mobilise the community towards a goal, but after that’s done, people are still going, simply because they enjoy it. (lipu tenpo has a Patreon, but I’m sure it’s far from a self-sustaining business. That mostly covers physical prints.)

See, there are 5 random content words that can be used as verbs, nouns, or modifiers, which also function as prepositions, …

I agree disambiguating parts of speech in Toki Pona is confusing for learners. But once we drop the question of “designing a language” and take Toki Pona as it already exists, honestly this is rarely, if ever, an issue. Their use as a modifier is easily disambiguated by reordering modifiers. And…

I suppose you could clarify this with a comma, or by using that hideous sitelen pona underlining thing, but how exactly these workarounds would transfer to the spoken language is unclear.

It is crosslinguistically common for function words to be less prominent intonationally. I don’t see why this wouldn’t simply apply to Toki Pona. It is interesting to think that people don’t think to teach that, though.

But I guess having dedicated words for all the official books is more important.

Skipping over most of the “what if Toki Pona grammar but different” conversation. This one’s interesting — I think the books are already either used in English as jargon, or in Toki Pona as proper names. At least, that makes up the ~majority of their appearances. Maybe some day we will see people stop counting them. That’s up to the community.

And you know what else sucks? Forming chains of verbs.

This one is also fun because I don’t think preverbs are chains of verbs. I mean, they are chains, and they are verbs, but they’re not serial verb constructions, they’re modal verbs. It makes sense for the auxiliary verbs to be both semantically limited (hence preverbs as a closed class), and to have different syntactic behaviour from the main verb (hence they can only be negated).

And to top it all of, there’s this cute little quirk of the grammar where any sentence of form ona li (Verb) can simultaneously mean “it does the verb” or “it has the verb done to it,” with no single reliable way to specify the latter, because verbs as nouns can refer to the generic object of the noun and li doesn’t specify the part of speech.

In practice words have inherent noun/verb/adjective meanings, and out of context they will receive those meanings. If a need arises to express one of the other parts of speech, people will circumvent it syntactically (like … li ijo X, … li X e ijo). The desire to explain things in “reliable ways” and without parts of speech comes more from Toki Pona syntax modelling, I feel, which is indeed entirely transparent to parts of speech, making it seem like interpretation is some Wild West, when usually it’s clear enough.

Etymology

Putting on the hindsight lenses again, jan Sonja never emphasised the etymologies in Toki Pona’s design. Rather, it’s the community that stepped in and put every single word under a microscope. Someone who wants etymology to follow design guidelines can, therefore, easily do a “better job” than Toki Pona, a language that wasn’t trying.

[mocking] “It has words from all over the world!”

Still factually correct, if not that interesting to me.

Anyway, on a related note, I once did a relex of Toki Pona where each relexed word is voted on by the community. You can find it here. The overall impression I can give was that it was fun, but the end product looks more oddly-designed than Toki Pona. The words are too short and too many start with a vowel.

Despite the fact that the system of particles in Toki Pona could theoretically allow a Subject-Object-Verb word order without much issue, at least for sufficiently simple sentences, so far such a method of speaking has only been proposed as a joke that should not actually be used.

It’s one of those things that I feel like someone, at some point, will start trying in earnest. It is indeed not strictly necessary for syntax to function, and it might open up possibilities in creative expression. The main problem is that SVO~SOV variation doesn’t just come about overnight (it will be rightfully seen by speakers as strange and ungrammatical), and presumably needs to be “seeded” by some situation where it’s crosslinguistically most common.

The prepositional phrases they form are supposed to go at the end of predicates, even though they often modify the subject all the way at the start.

mi moku kepeken uta

I am not gonna lie, this is perhaps the one statement in the video that just baffles me. Does the author also think that “I with a mouth” forms one syntactic node? Does the author not know the difference between predicates and adjuncts? I don’t know from what angle to approach this, so I’ll just move on.

tonsi

As time goes on, I see more people adopt tonsi sensu stricto. If I wrote the course two years ago I would probably highlight the umbrella trans meaning as a semantic broadening, but now (and through conversations with others) I leave that for learners to discover and consider on their own. I think the “just like Esperanto” problem hasn’t materialised because gender words in Toki Pona occur primarily as self-expression, not being as part of being gendered by others.

Look, it even affects the way the community presents its data!

It does! Some people chose to self-describe themselves with more than one gender word, and that’s up to them.

ona li sin ala

Yeah, this sentence wouldn’t be the go-to way to say “they don’t grow back”. This is either because jan Minasa has a comparatively old nasin, or to fit in the rhythm. I don’t remember being misled by it, though.

ni li kepeken mani luka

Okay, but seeing mani + a potential number word and assuming it’s not gonna mean a number is just a skill proficiency issue

In the montage of context fails at the end I realized I forgot to add another example! When I was reading the first entry of lipu kule (lipu kule li kama lon) I realized I had no idea what nasin jan was supposed to mean when it was casually dropped in the second paragraph. It was only after reading the third paragraph that I realized what it was supposed to mean (politics) but I only figured it out because of how many times the “we don’t want politics/everything is political” debates come about. I guess it worked for me, but would someone who is not already familiar with this know that “sitelen ale li nasin jan” means “all writing is political”? I highly doubt it.

I would understand this, because nasin jan for politics had been at that point a commonly seen lexicalisation on Discord, via a channel name. By now, the channel has been renamed to just nasin, so newer people are less likely to understand it. A cleaner way to talk about politics today might also involve lawa or kulupu or ma.

Conclusion

Wait, there’s supposed to be a conclusion? Oh heck right uhhhh. Toki Pona is a language that exists. Some people speak it, cool. People will learn it if they want and won’t if they don’t. For technical language, I think when there’s a Will, there’s a wahey. I think some people will continue seeing Toki Pona as a design project or a toy, and they’re free to go design more conlangs and build new communities. Idk stuffs fine.