Hacker News new | past | comments | ask | show | jobs | submit login
Toki Pona: A Language with a Hundred Words (2015) (theatlantic.com)
126 points by Tomte on Oct 29, 2021 | hide | past | favorite | 63 comments



Toki Pona is great as an introduction into language learning.

As many English speaking people seem to be functionally monolingual they could greatly benefit from learning it.

It would probably be the human language equivalent of a pure Java programmer spending a weekend hacking Lisp in broadening your horizon.

Even if you don't end up using Toki Pona, you will still benefit greatly from it. Due to its small size you will be able to acquire a certain "mastery" over it in just a few days which can be greatly motivating for tackling bigger languages.


I do not fully agree with you. Although it is a small language, it has language constructs that are foreign to English (and some of which I do not know of being used in any natural language, although I expect that there is an obscure language that does have them, as it is my experience that many 'wierd' language constructs found in constructed languages appear in natural languages). Just having fewer words, does not make a language simpler to use. It simply means that a lot of vocabular is created by combining words. This is also seen in a language like Chinese, where more complex concepts are made by combining two or more characters into a new concept. Over time, some of these sequences are used for other things than orginal intended to the extend that native speakers no longer are aware of the origin. You also see this in English and other natural languages.

English is not my native tongue and besides my native tongue, Dutch, I also understand some German, a little French, and a very little Chinese (Mandrin). I made some attempt to learn Toki Pona (about 10 ago). Although, I am not particular good at learning languages, I am more of a visual thinker, I have always had an interest in languages and have a collection of foreign language books and linguistics, see https://www.iwriteiam.nl/BookShelves.html#1.1


I am not sure if I get your point. Yes, having a small base vocabulary has trade offs and yes it is true that you will have to not just learn the meaning of every single word but also common idioms.

You can be up and running very fast in Toki Pona being able to produce and understand simple sentences about broad topics. If you need to be very specific, you will have a hard time but the philosophy of Toki Pona is to strive for simplicity and to focus on your core message.

You would have a hard time writing a scientific paper in Toki Pona but that is not what is was designed for anyway.

What language people have trouble learning can be a bit subjective and also heavily depend on the learning material and other outside factors. Sounds like Toki Pona didn't "click" with you which is fair enough. Maybe you want to give it another try or maybe it just is not for you, there are other interesting conlangs.

Toki Pona is relatively easy to learn not just for the small base vocabulary but also the relative simple phonology (Only 9 consonants and 5 vowel, always stress on initial syllable). Learning how to pronounce and write Toki Pona is extremely easy and regular.

Vocabulary is probably what takes the longest time in language learning, so you get a huge head start. The grammar might be very different to English or Dutch but has many similarities to Asian languages and is not that complex. As I argued for learning Toki Pona for mind expansion, I would argue that the difference is a good thing anyway.


I agree with what you say. But instead of learning Toki Pona, you could also learn the 100 most common Chinese words and probably say as much as you can say in Toki Pona. If you study that list, you will see that just like in Toki Pona, many of the words have multiple meanings and can be used in multiple modes.

Of course, Chinese being a tonal language and the use of characters make it a lot harder than Toki Pona, but the number of people you can talk with, is a lot larger.

One of the problems that I had when learning Toki Pona, is that I did not have access to native speakers. Learning a language from a book is a lot harder than learning it through daily use. Recent studies have shown that adapt there language when they talk to their children and that that might be one of the reasons why children learn a language so seemingly without any effort.

I understand that for adults that is also te best method for learning a new language, having a native speaker that is able to adjust his language use to your comprehension of the language.


This is kind of like the learn a programming language to get a job vs expand your mind discussion

If you want to learn a language for being able to talk to more people then yeah, most conlangs don't make sense. (Except maybe Esperanto for having a huge international community that is bigger than some smaller natural languages).

If anything, conlangs are good for filtering for people that share the interest of learning conlangs (which might be positively correlated with people being interested in linguistics, other cultures, open minded, nerdy or whatever) but not much for expanding the total mass of people you can talk to. It is a pretty nerdy hobby that many people do for the pleasure if learning the language itself. (Similar to hacking in a non mainstream programming language)

As for learning, adults and children do learn differently. While both profit from language immersion, adults do profit much more by explicitly learning grammar. It is a bit of personal preference though. For example I dislike learning vocabulary so I rely more on immersion and looking things up as I go. This might be slower but is more fun to me.

I should not be that hard to find other people learning Toki Pona on the internet but of course meeting some in real life might be difficult. So i agree with your point that total language immersion like you can do when traveling to a country were your target language is spoken is not possible with Toki Pona. You need to be fine with learning from books and videos.


Yes. For a language with simple rules, you quickly run into situations where you must decide how to express a simple concept, and you're inevitably going to "fall back" on constructions from your native language which seem perfectly logical to you which you aren't even aware of - but which may be strange, or outright incomprehensible to someone coming at it from a very different language.


I don't see why the small vocabulary would help learn it, if more complex concepts are built from that small vocabulary.

English only has 26 letters! It must be really easy!!

And they're fooling themselves if they think being able to express lots of concepts with only 123 words means it's easy to communicate by only learning 123 things.

No. There are still a gazillion concepts. If you want to communicate "car" to someone you still have to learn that "space for movement" means car... Or is it a bus? Ok "car" is "small-moveable-space-on-four-wheels". Hmm, bit verbose. We'll just add one more word...


You can learn 100 words from any natural language.


True, but there will be very few documents you can read or conversations you can participate in with only a few hundred words in a natural language.

The steep learning curve of most natural languages means it takes months before I can read or say anything nontrivial.


Interesting that the creators of Esperanto and Toki Pona are both Polish. I wonder if there is something in Polish culture / "mindsphere" that encourages language creation.

Ironically, Polish, by comparison to some other European languages of comparably-sized nations, is very uniform and devoid of dialects / regional accents. There are a small few regions that speak a proper dialect (not mutually understandable with regular Polish speakers), but otherwise, with minor exceptions, you cannot tell at all where in Poland someone is from based on how they talk.

In the UK, by contrast, this is still a big thing. I believe German dialects are very diverse, heck, Dutch ones can pinpoint people to a village sometimes. French had very diverse language base before Napoleon. Spain still has regional languages that are very much alive and in use, Italian accents (dialects?) are too quite distinct.

Maybe Poland is compensating for this here, by inventing whole new languages?

EDIT my bad, the article mentions a Pole, Marta Krzeminska, but she is not the language's creator. It was a Canadian who created it. Oh well.


> but otherwise, with minor exceptions, you cannot tell at all where in Poland someone is from based on how they talk.

Not sure if that's what you had in mind by "minor exceptions", but there are certain shibboleths that can give away the area of Poland you're from (or at least causally associated with). For example, take the phrase "to go outside". In most of the country, it's "iść na dwór" (lit. "to go to court/manor"), but in the Lesser Poland voivodeship, especially around its capital city Kraków, we use "iść na pole" (lit. "to go to field") instead.

Additionally, a rough determination of origin can be gleaned from various verbs and nouns. Poland was partitioned[0] for over a hundred years by Austria, Prussia and Russia, with each of the historical empires performing their own attempt at assimilating the populations in their partition. In the process, Polish spoken in those partitions borrowed many of the words from their respective occupants' languages. As a result, the way you refer to many everyday things (like whether you call a garden hose "szlauch" or "wąż ogrodowy") can give away which of the three ex-partitions you're from.

--

[0] - https://en.wikipedia.org/wiki/Partitions_of_Poland


Yeah, that's kind of what I meant. Also, eastern Poland (countryside more than cities) has a drawn-out, more melodic accent.

By contrast, you talk to someone from East London, or Birmingham, or Leeds in the UK, and you're in no doubt as to where they are from, no need for special words. Never mind Scotland, West Country, Wales...

This is a little less obvious nowadays than it used to be, as people move around more, but still a big thing.


> speak a proper dialect (not mutually understandable with regular Polish speakers)

a language's dialects are mutually understandable by definition IMO, if they are not, then those are not one language's dialects but (dialects of) different languages.


The broad Doric Scots I spoke as a child is, like a lot of regional linguistic variations in the UK, notoriously impenetrable to outsiders but I've never heard it described as anything other than a dialect.

e.g. I might have enquired of a visitor when I was a wee loonie "Far div ye bide?" - "Where do you stay?"


I have heard it called language but, notably, by people from outside the UK


My personal favourite silly example of Doric being "fit fit fits fit fit" - "Which foot fits into which foot" (talking about shoes, apparently).


Depends on the definition, and final decision of linguists. It's an ongoing controversy in Poland for two languages / dialects: Kashubski language and Silesian language. Both are strictly used only by regional minorities, both are not understood well by Poles from other areas, and both have German influences.

Kashubski is mostly called a language, even thought it's not officially recognized as a regional language (you can't expect to use it in for official matters).

Silesian is mostly called a dialect/subdialect, but there's a regional movement to elevate Silesian to a separate language through political means.


Mutual intelligibility is a gradual scale, not a binary. So where you draw the line on that scale is a rather arbitrary (and usually political) decision.


A language is a dialect with an army and navy.


I feel the same way about Russian. Big country, large diaspora, but everybody's accents sound the same. If there are differences then they seem subtle. Can a native Russian speaker explain this?


If you mean Russian accent when speaking English, then it comes mainly from difficulties pronouncing sounds that have no direct equivalent in Russian (e.g. the sounds represented by R, W, TH), or those that do but where it's pronounced noticeably different (e.g. I/И). These are mostly the same for all dialects of Russian.

However, Russian itself has plenty of dialects that can be quite different from the language standard, and there are certain characteristic variations that can readily pin one's geographic origin. For example, standard Russian tends to pronouce unstressed О as А (or rather as shwa biased towards A), but this is most pronounced in southern dialects, and is missing from most northern ones. Similarly, Г in standard Russian is pronounced same as hard G in English, but in southern dialects it tends to become something more like GKH.

As with other countries, dialectal variety has been significantly reduced through universal education (that tends to enforce the standard and frown upon dialects as "uneducated speech") and mass media. The historical tendency for authoritarian and highly centralized governance, both during the imperial period and later in the USSR, exacerbated this.


Polish grammar is very complex.


To be fair, this goes for all Slavic languages.


This bit in the introduction has an interesting blind spot:

In Icelandic, a compass is a direction-shower, and a microscope a small-watcher.

The author didn’t have to reach for Icelandic since “microscope” is already from Greek with the same construction small/watching.


> The author didn’t have to reach for Icelandic since “microscope” is already from Greek with the same construction small/watching.

Is it really from Greek, through? My understanding is that it was coined in some non-Greek language from Greek roots, which isn't quite the same thing. Google translate gives "smásjá" as Icelandic for microscope, which looks like it might be a compound using the languages "native" vocabulary ("smá sjá" translates as "little see").


According to Wikipedia, the name Microscope was derived from the earlier "Telescope", which in turn was coined shortly after the object was invented:

> Galileo's instrument was the first to be given the name "telescope". The name was invented by the Greek poet/theologian Giovanni Demisiani at a banquet held on April 14, 1611 by Prince Federico Cesi to make Galileo Galilei a member of the Accademia dei Lincei. The word was created from the Greek tele = 'far' and skopein = 'to look or see'; teleskopos = 'far-seeing'.

The primary source for this is "Rosen, Edward, The Naming of the Telescope (1947)".


I made a game based on the idea of Toki Pona using a sub-set of English.

Instead of learning a new language (I tried to learn Toki Pona) I used Toki Pona as a template for basic English, Toki English, a subset of English.

The benefits of Toki English are this:

* English speakers already know all the meanings of all the words, and can use it right now.

* Non-English speakers learn BOTH some English _and_ a basic structure to speak to anyone else that learns Toki English and English.

* It's actually hilarious to try and speak with it.

The game I made (very rough) is like a form of charades, where you have 2 teams. One person gets a phrase/idea/object that they must convey to their team using only Toki English. The props are simple, a printed sheet of paper with all the Toki English words.

[edit for clarity]


I haven't even looked at your game. And I'm not particularly interested in conlangs.

But I want to say thanks for making a game that people might learn something from, rather than the mindless crap that the industry mostly produces.


There's nothing to look at for the game. It's just a spreadsheet where I picked the English words from the Toki-Pona lexicon, and printed them out on paper. :)


If anyone is interested in learning about conlangs, jan Misali has an interesting playlist[0] with reviews on various contructed languages including Toki Pona.

[0]: https://www.youtube.com/watch?v=eLn6LC1RpAo&list=PLuYLhuXt4H...


Note: he is a fraud, albeit a cute one.

https://www.youtube.com/watch?v=3aUx7ajQrys


He also has a whole playlist with lectures specifically about Toki Pona: https://youtube.com/playlist?list=PLjOmpMyMxd8T9lZjF36c4mn4Y... Great channel; I second the recommendation.


The only authoritative source is paid book. Closed, commercial languages like Borland's Delphi or Toki Pona have no future. Don't waste your time.


Toki Pona is simple enough that you don't need the book to learn the language. The books ("pu" and "ku") are not a prescriptive manual but more like a descriptive dictionary, and indeed, a recent census [1] shows that those books weren't the most used learning materials.

[1] https://tokiponacensus.github.io/results/#how-people-learned...


That was an issue with the language Loglan where the author claimed copyright on the language. As a reaction the community created Lojban. Basically GNU/Lojban.

Whether you really can copyright a language is pretty dubious though, probably not. In the case of Loglan it was not enforceable but Lojban is now far more popular anyway, so it is moot.

As for Toki Pona in practice it is not much of an issue because the free learning material is good enough. The language is very small and not that strictly defined anyway but I get your concern.


The community of toki pona speakers has grown to the point where there is no such thing as an authoritative source, much like a natural language. If you want to learn in depth your only option is to join a chatroom and talk to people. For intro material, there are many free options beyond the official book. https://devurandom.xyz/tokipona/ is quite good.


buying the book is just a way to thank the author, all the contents are freely available in the official site:

sina toki ala e lon, toki pona li pona.


This is about a language like English or Frennch not a computer programming language. Did you even look at the link at all?


I know what toki pona is. Just like Delphi, you find little roadblocks. License problem here, can't legally give it to friend, can't put it on your website. These are all completely artificial problems that simply don't exist in free languages. That's why Delphi failed despite being superior at the time and that's why toki pona will fail.

You are knee deep in a cow piss on a remote farm in Nebraska, offline and some library has some version issue with your proprietary licensed language. A trivial solution if it was free language, but now it requires 300 miles trip. You are in Nagorno Karabakh and just spoke to some random stranger about this new funny speak-language but you cannot give him PDF because it's not free. These issues stings very deep because how artificial and easily solvable would they be. Some things must be free otherwise they cannot be used to their full potential. So you stand there, in Nebraska or Nagorno Karabakh and swear - Never again will I use proprietary language. I have nothing agains proprietary software or linguistic books but the language itself must be free.


> offline and some library has some version issue with your proprietary licensed language

How is this different from any proprietary book about a freely-licensed language?

If you're in a remote farm in Nebraska and you only have a copyrighted Esperanto textbook, you can't share it, even though Zamenhof's materials might be public domain.

I fail to see any problem with toki pona here.

Something similar can be argued about Delphi / Object Pascal. Borland's compiler is non-free, but the language can be used with FreePascal. I'm happily using FreePascal on my personal projects and haven't touched Delphi in more than a decade; it's a nice language and there's nothing wrong with it.

Sharing FreePascal is totally OK, just like sharing freely-licensed materials about toki pona.


IMO speaking TP is not restricted by licenses. i can speak it completely offline. and IMO it's designed primarily to be spoken, not written. so a paid or non-free book is not an obsticle, since TP's design philosophy wants to include everyone possible around the world (eliminating voiced/unvoiced plosive pairs, hard to pronounce vovwels, consonant clusters), so a latin based script does not belong to the language's core; there is even a Hangul script for TP.


Bit of a side-track, but if you ever get to go to Korea, spend a few hours learning how Hangul works. I am crap at learning how to speak languages (17 years in Germany, and I still have most people shifting to English as soon as they hear a few sentences from me), but I managed to learn how to use Hangul pretty much on the flight to Korea.

Helped immensely in getting around! Of course, I had absolutely no comprehension of what things meant, but I could match sounds to signs and avoid eating things I’m allergic to.


What would a shopping list in Toki Pona look like?


lipu esun mi: (shopping list, literally: my buying document)

1 kili palisa jelo mute (bananas, lit: many yellow stick-fruits)

2 kala pi loje walo (salmon, lit: red-white fish)

3 moku kasi laso supa (salad greens, lit: flat green plants for eating)

4 telo walo (milk, lit: white drink)

5 kili kiwen lili mute (nuts, lit: many small hard fleshy-plant-food)

6 pan supa sike (pita, lit: flat, round grain-product)

It's worth noting that the translations are ambiguous and personal; if I gave this list to a stranger I might end up with pizza crust instead of pita, or nut milk instead of cow milk. If I wanted to explicitly specify nut milk I'd have to say something like "telo walo pi kili kiwen lili" (lit: white drink of small hard fleshy-plant-food). Having said that, even in English people generally wouldn't go into that level of detail on a personal shopping list. When my partner writes "milk" on the shopping list I know they actually mean a "half-gallon of whole milk" based on prior experience.


It would work fine as long as everyone consuming the shopping list has a reasonable idea what normal things your shopping lists. Things like "kili sike loje" (red round fruit) are ambiguous but are no doubt close enough to jog your memory enough to work.

The toughest part that comes to mind would be particular spices. You would probably need to start borrowing words to distinguish between things like thyme and rosemary ("namako Tin", "namako Loseli")


One key characteristic of natural languages is that when they don't work, people invent and share new words.

kili sike loje isn't enough to distinguish cherries from cherry tomatoes, red grapes, cranberries, red currants, red plums or red apples -- every one of which is a perfectly reasonable thing for me to put on a shopping list.

And so you'll note that we have the words tomato, grape, cranberry, currant, plum and apple.

Now, French has a language academy to decree what is proper, and English --

"We don't just borrow words; on occasion, English has pursued other languages down alleyways to beat them unconscious and rifle their pockets for new vocabulary." -- James Davis Nicoll

A language which isn't useful will either change or be abandoned. Natural languages almost always change.


Languages often invent terms by combining existing words, though, and then some combinations become standard. Indeed, your own list has "red grapes" and "red currants"! For some other prominent examples, consider "pineapple" or "pomodoro" ("pomi d'oro").


Similar to Toki Pona but a bit more expressive is Mini:

https://minilanguage.com/


This is awesome, thanks for the link!


The article casually mentions Koko, and it's a bit sad that the story of Koko's "language" is still alive...

I can not remember book exceprt, but recently there was a great video about Koko https://www.youtube.com/watch?v=e7wFotDKEF4


For those interested enough to read but not to watch: what is wrong with Koko's language? I vaguely remember (and the article mentions) it was a primate who learned sign language, but your comment suggests the story is more complex?


TLDR: There's no evidence she was doing anything but brute forcing word combinations in order to get rewards.


For those interested, 'The Allusionist' had an episode where they learnt Toki Pona with a friend. I'm no longer subscribed, but I remember this episode fondly.

https://www.theallusionist.org/allusionist/tokipona



Thanks! Expanded list:

Toki Pona - https://news.ycombinator.com/item?id=23402374 - June 2020 (38 comments)

Ilo sitelen, a handmade computer for Toki Pona - https://news.ycombinator.com/item?id=22788011 - April 2020 (89 comments)

Toki Pona: A Language with a Hundred Words (2015) - https://news.ycombinator.com/item?id=22689959 - March 2020 (127 comments)

Toki Pona (a minimal language) - https://news.ycombinator.com/item?id=16018991 - Dec 2017 (1 comment)

How to Say (Almost) Everything in a Hundred-Word Language (2015) - https://news.ycombinator.com/item?id=14191186 - April 2017 (102 comments)

Toki Pona: A created language that has only 100 words - https://news.ycombinator.com/item?id=11153406 - Feb 2016 (88 comments)

Toki Pona: a human language with 120 words - https://news.ycombinator.com/item?id=9914534 - July 2015 (110 comments)

Creating your own language. For example, Toki Pona has just 120 words. - https://news.ycombinator.com/item?id=54392 - Sept 2007 (1 comment)


What you just did sounds like an interesting "Show HN" project. Automate the thing, so that a HN bot posts a comment with previous talk about the topic.


Related to [1] - since Toki Pona has such a limited set of sounds, it seems like it would be a great fit for stenography...

[1] https://news.ycombinator.com/item?id=29035644


There are twenty three keys on a steno machine. With only a hundred-word language the current machine would certainly work. But with so few words even the machine could be simplified.


The few Toki Pona speakers I know have this game where they make an appointment to meet for some purpose all in Toki Pona, and then when they next meet they compare where they were (and for what purpose) at the time they thought the appointment was at...

The limited vocabulary makes precision hard, even if you are not on Tanagra.


I found the ambiguity in toki pona too much a hurdle when I was trying to use it some years ago. The best minimalist conlang I ever found was Angos

https://qdoc.tips/angos-grammar-pdf-free.html

https://www.reddit.com/r/angos/

I couldn't really get into it though. Nowadays, it's all about ido!


Note: it's a constructed language. You can halfway through the article without realizing this.


I learnt this with a friend once, it was a nice experience. Forgotten pretty much all of it now though :D


Love Toki Pona.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: