Kip: A programming language based on grammatical cases of Turkish

by nhatcher • 6 months ago

254 62 6 months ago

lolc3• 6 months ago

Haha I can read some casual Turkish and this made my day!

Funny how the case system of Turkish is both strong and standardized enough for this to work well. I don't know any other language where flexible argument order would work so well.

thaumasiotes6• 6 months ago

I don't know any other language where flexible argument order would work so well.

What kind of sample size is that? A case system and flexible argument order are largely the same thing.

Note also that flexible argument order is a robust phenomenon in English:

1. Colonel Mustard killed him in the study at 5:00 with his own knife.

2. Colonel Mustard killed him at 5:00 in the study with his own knife.

3. Colonel Mustard killed him in the study with his own knife at 5:00.

4. Colonel Mustard killed him with his own knife at 5:00 in the study.

5. Colonel Mustard killed him at 5:00 with his own knife in the study.

6. Colonel Mustard killed him with his own knife in the study at 5:00.

But if you insist on looking in other languages, there's a famous Latin poem beginning Quis multa gracilis te puer in rosa perfusus liquidis urget odoribus grato, Pyrrha, sub antro?

Translating this as closely as possible to a one-word-for-one-word standard, it says What slender boy soaked [in] liquid odors presses you among many rose[s], Pyrrha, beneath [a] pleasant cave?

(Notes: rosa is singular for unclear reasons. There is nothing corresponding to the in of "in liquid odors"; the relationship between the odors and the soaking is expressed purely by case. There is also nothing corresponding to the article in "a pleasant cave"; Latin does not mark definiteness in this way. Location inside a cave is expressed with "beneath"; compare English underwater.)

Anyway, the actual word ordering, using this translation, is: What many slender you boy among rose[s] soaked liquid presses [in-]odors pleasant, Pyrrha, beneath [a-]cave?

I've heard that Russian poetry is given to similarly intricate word orderings.

lolc3• 6 months ago

Yeah sure there are other languages where the cases allow flexible argument ordering. I should have written "another language" not "any other language" because I meant the languages I know. I can see how it'd work in Latin. Though poetry takes a lot of license with language, so by quoting from poetry, we're at the extreme end of what a language can do. From what little I remember in Latin there would be examples where one couldn't distinguish the cases.

The only variation in your English examples is with the prepositions. English is a language where you just do not get that kind of flexibility. Unless, again, you do poetry. Can't even put the inflected pronoun first even though it would be unambiguous:

    Him Colonel Mustard killed.

Let's look at this contrived example:

    The policeman bites the dog.

Notice how in English I can't reverse them easily. Have to get creative:

    The dog the policeman bites.

Stilted and prone to get misunderstood isn't it? In German, this is possible:

    Den Hund beisst der Polizist.

Though if you do that, people think you made a mistake and will misunderstand you. The cases are rather weak in German. No comparison to Turkish:

    Köpeği ısırıyor polis.

They do it for emphasis as part of everyday speech. You can put the subject first or last or in the middle and there is no confusion. Though admittedly that example is stretching it :-)

inkyoto3• 6 months ago

I don't know any other language where flexible argument order would work so well.

Any highly inflected language has such a property. Slavic languages, Sanskrit (or, more broadly, Indo-Aryan languages) are prime examples.

Speakers of Finnish and Hungarian will likely chime in and state something similar.

ronjakoi2• 6 months ago

Finnish case markers vary a lot from word to word, because of not only vowel harmony but other features of the word stem, and consonant gradation which is a weird feature of Uralic languages.

For the subtraction example, some numbers would be 50:tä 5:llä and others 6:tta 3:lla. Of course you could encode for all those possibilities and successfully parse them, but it would feel weird for a compiler to reject an expression because it's ungrammatical Finnish.

Also it would feel weird if you first write (vähennä muuttujaa 256:lla) but then realise you made an off-by-1 and have to change it to (vähennä muuttujaa 255:lla) but that doesn't compile because it should be 255:llä, so you have to remember to change two things.

But on the other hand, that's just how it is to write in Finnish, so in prose we don't really think about it. In natural language, it's normal to have to change other stuff in a sentence for it to continue making sense when you change one thing.

readthenotes1• 6 months ago

Of course, missing

https://www.euronews.com/my-europe/2024/10/25/what-is-turkis...

nhatcher2• 6 months ago

Gossip tense has been discussed here:

(276 points and 255 comments) https://news.ycombinator.com/item?id=41793485

joomy10• 6 months ago

It would actually be a good fit for event handlers...

ugurs• 6 months ago

Clicking the link with a prejudice in my mind, I found the definitions cleverly clean and easy to understand. I would be pleased to see a German version of it, just to have a good laugh.

abdusco• 6 months ago

Yep. I was expecting to see an existing language (usually Rust) with keywords swapped with Turkish translations, but no, this is actually really nice.

hexfran• 6 months ago

Not relevant to the language itself but to the grammar: https://languagelog.ldc.upenn.edu/nll/?p=73

celaleddin• 6 months ago

Got especially excited, since I also experimented on a similar idea a few years ago:

https://github.com/celaleddin/sembolik-fikir

Will check this out further in the following days. Thanks for sharing!

joomy10• 6 months ago

That's pretty cool! From what I can tell, it does a morphological guess based on the suffix. If you didn't have the apostrophe, it'd have issues with ambiguity (say "aşı", does it mean vaccine or does it mean "aş" in accusative case?) but the apostrophe solves that problem too.

My solution for this problem in Kip was to go all the way with the morphological analysis using TRmorph (https://github.com/coltekin/TRmorph) for it, and then resolve the ambiguities in type checking / elaboration. (Therefore Kip almost never needs apostrophes.) Whether it was worth it, I don't know, but it was a fun problem to solve. :)

joomy10• 6 months ago

Hi all, Kip's developer here! I was going to wait until we had finished the playground and landing page before posting about the project more, but here's the browser-based playground we have so far (thanks to Alperen Keles) for anyone who wants to play with the language: https://alpaylan.github.io/kip/

(The work on JavaScript transpilation just started today and currently doesn't work, but running the language should mostly work, though it probably has bugs, which I'd love to hear about in the repo's issues!)

sedatk• 6 months ago

Fantastic work, an area I’ve always wanted to explore.

nhatcher2• 6 months ago

Sorry! It was too beautiful not to share it. But I'm sure you can do a show HN once you think is ready.

joomy10• 6 months ago

That's okay, thanks for sharing!

araes• 6 months ago

Question on some of the syntax. It's neat, and think the idea's cool. Would definitely be something if nothing else for security through obscurity. Is it even code?

However, for some of the number stuff, if you write something like:

  (5'le 3'ün farkını) yaz.
  (3'ün 5'le farkını) yaz.

How does it tell whether it is:

  5 - 3 =  2, or
  3 - 5 = -2  ?

Does it always just return 2 because of the meaning of "farkını" and the placement of 'le and 'ün? Like:

  (5 first, 3 second, difference) write, vs
  (3 second, 5 first, difference) write  ?

Google just gave back:

  Write (the difference between 5 and 3).
  Write (the difference between 3 and 5).

Not especially familiar with Turkish, and mostly had to use translation, yet it looks like a language for defining math theorems? Number following "zero" shall be called "one", number following "one" shall be called "two". Or is that more just a feature of using natural language for the writing syntax?

joomy10• 6 months ago

"fark" here takes two arguments, the first (the minuend) is in instrumental case (-le), the second (the subtrahend) is in genitive case (-in). Now, because of the suffixes of the cases, regardless of the order in which you give the arguments, the type system can figure out which one is supposed to be the minuend and which the subtrahend.

If it helps, you can think of it like named arguments where the name is inferred from the case.

pinkmuffinere• 6 months ago

Disclaimer: I grew up speaking Turkish, but never studied it. I think I can give a common-sense explanation, but can’t give a rigorous “proof” appealing to grammatical rules.

I read “(5’le 3’ün farkını) yaz” as “having 5, 3’s difference write” (of course this is not natural in English). Ie, you’re given 5, you want to take 3, and write the result. Likewise, “(3’ün 5’le farkını) yaz” would be “3’s difference, having 5, write”. Again we are given 5, and want 3’s difference. Because we’re starting with 5, i think there is no ambiguity in the operation to be done — start with 5, subtract 3.

Idk if that actually helps clarify it at all, maybe it gives some intuition

andro_dev• 6 months ago

The Turkish sentence's expansion would be pretty much like the mathematical expansion of the expression order.

gerdesj• 6 months ago

"Kip (meaning "grammatical mood" in Turkish)"

I'm not sure what a grammatical mood is, so I tried a couple of well known translation services and got: kip == "mode". However big G did also manage "modal", "paradigm", "tense" and "module".

For my money: "tense". Just to confuse the issue, tense has several meanings in english! Here I think we are talking about a verbal tense:

https://www.bbc.co.uk/bitesize/articles/zh4thbk#zyh2s82

Tense can also be synonymous with emotion: dangerous/exciting and also as a measure: tension/tight.

Twey7• 6 months ago

In linguistics, tense is a verb conjugation to indicate temporal information, while mood is a conjugation to indicate various kinds of metainformation about the speaker's relationship to the information in the sentence. It's not as common a term as ‘tense’ when discussing English, because English doesn't conjugate for mood, but it is the standard word for describing some features of Turkish morphology such as evidentiality.

Automatic translators, while an impressive and convenient piece of technology, usually focus on providing a plausible gloss in the target language, so typically lose a lot of nuance. For looking up words a dictionary is usually a better bet; for example, Wiktionary has https://en.wiktionary.org/wiki/kip#Turkish with a link to the explanation of the English word as well.

inkyoto3• 6 months ago

A grammatical mood indicates the modality of the verb, and some languages possess rich inventories of the grammatical moods. They are called differently in different languages, but mood is an established term in English.

English has indicative («go», «is going» etc), subjunctive / conjunctive / conditional («went» in «as if they went»), imperative («go!»).

German has two conditional moods – Konjunktiv I and II, for example.

Finno-Ugric languages have many more.

lowdownbutter• 6 months ago

languagejones has a good overview in this video

https://youtu.be/xdpvR3kaXaQ?si=dcNS-3VcH8AMU7dt

yowayb• 6 months ago

I studied Turkish for a few years and remember thinking it could make an interesting programming language (due to the grammatical/agglutinative features). I was gonna call it Ç, but I was never seriously going to make it. Happy to see someone went for it!

mertd• 6 months ago

Ç is a brilliant name

joomy10• 6 months ago

Okay there is an updated web page and playground now: https://kip-dili.github.io/

hahahahhaah• 6 months ago

A language an LLM can choke on!

chuckadams• 6 months ago

Reminds me a bit of Lingua::Romana::Perligata.

joomy10• 6 months ago

Yes, that's one of my inspirations! I'm writing a short paper about Kip and I'm citing Perligata there for sure. The closest modern non-English programming language I know that also uses grammar features is Tampio, for Finnish. https://github.com/fergusq/tampio

SuperNinKenDo• 6 months ago

I love this kind of stuff - non-English programming languages, particularly when they utilise language features in unexpected ways like this.

My Turkish is pretty rusty - and was never any good anyway, but really cool stuff.

octoberfranklin2• 6 months ago

It would be really helpful if this page showed side-by-side comparisons of the same program written in Kip and some other language, like say Haskell.

I'm having a hard time seeing how this is much different from record types, except that you're limited to only eight fixed record field names (one for each grammatical case).

joomy10• 6 months ago

Cases essentially act like named arguments, except the names are inferred from the case of an argument, which is inferred through morphological analysis. And that analysis can be ambiguous, so the ambiguities are solved by the type checker / elaborator. It's different from record types in the sense that you can provide the arguments in any order to a function, and the system will figure it out because of the cases.

octoberfranklin2• 6 months ago

Cases essentially act like named arguments

... and there are only eight argument names. Which seems... limiting.

It's different from record types in the sense that you can provide the arguments in any order to a function

You can provide the fields of a record in any order, and call a function with it!

   f { foo=3, bar=5 }
   g { bar=5, foo=3 }

I still think the page needs a Rosetta Stone.

tristenharr• 6 months ago

Woah this is AWESOME! I am going to have to take a look! Love the thoughts, doing something similar at https://logicaffeine.com/studio

Check out Logos lang, would love to chat sometime. love that you chose Haskell!

Twey7• 6 months ago

Those of a certain vintage might remember Lingua::Romana::Perligata: https://metacpan.org/dist/Lingua-Romana-Perligata/view/lib/L...

Twey7• 6 months ago

I love when the ‘language’ part of programming languages doesn't get left behind :) Very neat!

Ruidy• 6 months ago

I have been learning the language because my girlfriend is Turkish.

Now I can use my programming brain to fast forward the learning.

Amazing

biddit• 6 months ago

Yep, this was a skillset combination I never envisioned. (am married to a Turk). Maybe a good reason to push past the intermediate plateau.

hasanabi• 6 months ago

Better call it "Kanak".

timonoko• 6 months ago

----

arikana• 6 months ago

It seems the order of the parent structure is still following the Anglo-saxon grammar even though lines and pharisees are Turkish , I would like to see something:

Fibonacci-Dizisi (n):

  n 0'a eşitse: 
    dur.

  değilse:

    bu-sayıyı yazdır.
    şu-sayıyı, (bu-sayı + şu-sayı), (n - 1) ile Fibonacci-Dizisi'ne devam et.

Ana-Akış:

  "Bir sayı girin: " yazdır.
  Girdiyi oku -> girdi.

  girdi'nin tam-sayı-hali:

    Yokluksa: "Geçersiz sayı" yazdır.
    Varlıksa (n): 0, 1, n ile Fibonacci-Dizisi'ni başlat.

Başlat: Ana-Akış.

nevertheless, that is a good start, congratulations.

nacozarina• 6 months ago

with AI every country can migrate development to a programming language that reflects the diction of their unique ethnolinguistic traditions, it's very empowering

Aerbil313• 6 months ago

Neat idea. I’m a native Turkish speaker, and when I was first learning English as a kid I remember always being baffled at the extreme chaos and irregularity in English. In the end I learned it mostly by exposure and not from school.

Kip: A programming language based on grammatical cases of Turkish

Related Stories