An open API service indexing awesome lists of open source software.

https://github.com/lynn/toki-wile

a constructed language, that combines the parts of Ithkuil I find interesting with the parts of Toki Pona I find cute and learnable
https://github.com/lynn/toki-wile

Last synced: about 1 month ago
JSON representation

a constructed language, that combines the parts of Ithkuil I find interesting with the parts of Toki Pona I find cute and learnable

Awesome Lists containing this project

README

        

# toki wile 0.0.4

toki wile is a conlang that combines the parts of Ithkuil I find interesting with the parts of Toki Pona I find cute and learnable. This document, for now, assumes familiarity with Toki Pona and maybe some familiarity with Ithkuil.

I made this because I want to play with it, and find out what the interesting parts of Ithkuil feel like in a cute and learnable context. Maybe it's a lot of fun! Maybe it doesn't work at all! If you're interested, [**join the toki wile Discord**](https://discord.gg/bBcnEbxA9C).

Ithkuil is [©2004-2021 by John Quijada](http://ithkuil.net/). Toki Pona [was invented by Sonja Lang](http://tokipona.org/).

## Phonology
It's like toki pona (aeiou ptksjnlmw), but with consonant clusters and geminated consonants.

w/j do not appear in coda position, and instead there are diphthongs (`au ai eu ei oi`). Other vowel pairs like `ea` are disyllabic.

Stress is on the first syllable of each word.

/w/ has the allophone \[f\~v\] before /u/, and /j/ has the allophone \[ç\~ʝ\] before /i/.

## Words and sentences
There are two kinds of words: **pronouns** and **formatives**.

Formatives have an Ithkuil-like fusional slot structure, and are further divided into **nouns** and **verbs**.

There are no particles. Instead, there's a case system (replacing `li`, `e`...) and a "namako" formative slot (replacing `ala`, `anu`...)

The sentence structure is:

((Nla) Vla) (N0) (V1 N1) (V2 N2) (V3 N3)...
↑ ↑ ↑ ↑ ↑
la-clause head tail tail tail...

`N0` is called the "head", and each `Vi Ni` is called a "tail". The meaning is basically:

(When / Given that "Nla Vla":) <-- optional
"N0 V1 N1",
and "N0 V2 N2",
and "N0 V3 N3"...

That is, the nouns in the head `N0` "distribute" over all the verb-noun "tails". This is a generalization of the toki pona `S li V1 O1 li V2 O2`.

If present, the "la-clause" offers a clausal "context" or "condition" for the whole sentence. Its presence is marked by a verb ending in `-a` (see "Illocutions").

If there are no tails, then the nominative-case nouns in the head are linked copula-wise, "N1 is N2 (and is N3...)". So, it's a bit easier to distinguish "I eat" and "I am food" in toki wile.

## Pronouns
These are modeled after Ithkuil's PRAs. The shape is either C(C…)V, or VCCV. (When I use something like "VCCV" to describe the shape of a word, that means "vowel, consonant, consonant, vowel".)

The consonants indicate pronoun roles, and the vowels indicate cases.

m - mi
s - sina
n - ona
l - ni (demonstrative)
t - "ni:" (next sentence)
k - seme?

For example, `mi` is "mi li" (`-i` marks the nominative case), and `se` is "e sina" (`-e` marks the accusative case).

A consonant cluster combines referents: `sne` is "e sina e ona".

The VCCV shape rolls two pronouns into one funky word. `esmi` means the same thing as `se mi`. This is just an aesthetic thing from Ithkuil I like.

## Formatives
These are the building blocks of the language: nouns and verbs.

As in Ithkuil, the structure is divided into consonant/vowel "slots":

(Vn +) (preverbs +) Cr + Vs + Ca (+ suffixes) + Vk

### Required slots
The **Cr** slot is the root, found in the lexicon. For example **sk** means "lukin / kule".
Each root in groups two content words from toki pona, called its two "stems".

The **Vs** slot selects a stem from the root, and marks whether this formative is a noun or verb:

-a- Stem 1, noun.
-e- Stem 2, noun.
-au- Stem 1, carrier noun. The next word is a name. (laule Lin = "e jan Lin")
-eu- Stem 2, carrier noun. The next word is a name. (leule Kiti = "e soweli Kiti")
-o- Stem 1, verb.
-u- Stem 2, verb.
-i- Pronoun: interpret Cr as a pronoun. This allows applying Vn, Ca, suffixes... to pronouns. (omitwe = taso-RPV-KLP-ACC)

The **Ca** slot says some things about the "configuration" of the noun or of the event described by the verb. It is described in its own section below.

The **Vk** slot marks case for nouns, and illocution for verbs. See the tables below.

### Vn
Vn is an initial vowel that I stuffed random bits of meaning in. The n stands for namako!!!

a- "ala" on verbs. also on nouns: "I ate not-the-bread (implied: but I ate something else)".
e- no meaning, use when an initial consonant cluster is too spicy
i- "kin" (also), on any formative
o- taso (only), on any formative
u- anu, on any formative, links disjunctively with previous formative.

### Preverbs
You can put as many of these as you want in front of the root slot. They behave exactly as in Toki Pona. They're not currently allowed in a noun.

-wa- kama...
-wi- wile...
-we- awen...
-wo- sona...
-wu- lukin...
-je- ken...
-jo- open...
-ju- pini...

### Suffixes
The **suffixes** are each of the form "VxCx", where "Cx" is a root from the lexicon and "Vx" marks how it applies:

-a- Stem 1, modifier.
-e- Stem 2, modifier.
-o- Stem 1, modifier, as if preceded with "pi".
-u- Stem 2, modifier, as if preceded with "pi".
-i- Ca stacking.

This is the equivalent to compounds in Toki Pona. As an extra feature taken from Ithkuil, a "pi" on the last item marks that the modification of the root concept is "incidental" (like "jan pona = good person"), whereas its absence marks that the modification is "lexicalizing" (like "jan pona = friend").

lale "e jan" (the person)
lalape "e jan pona" (the friend)
lalope "e jan pi pona" ! (the good person)

The vowel -i- is used to stack/nest Ca configurations. For example, `lalwe` (group of people) can become `lalwilje` (a pair of groups of people).

If the consonant cluster in a suffix is -st-, it is a case accessor suffix. This always makes a noun even if Vs marks verb status.

-ist "one who V-s"
-est "one who is V-ed"
-ust "haver of N"
-ast "context for N or for doing V"
-aust "goal of N, purpose of V-ing"
-eust "what is used by N, or used for V-ing"
-ost "place where N is, or where V-ing happens"
-east "manner of V-ing"
-aist "similar to N, or what V-ing is done like"

## Cases
These mark the role of the noun in the sentence.

-i nominative, subject (tp: li)
-e accusative, object (tp: e)
-u genitive (scopes on the previous word. covers one sense of pi)
-a X la (as in "tenpo pini la". see -a illocution)
-au tawa X
-eu kepeken X
-o lon X
-ea tan X
-ai sama X
-oi "o!" (vocative)

The "genitive" is really a broad appositive case, not limited to "genitive" in the Ithkuil sense.

## Illocutions
Illocution marks whether this verb is an assertion, question, command...

Also, for assertions, it marks evidentiality (the source of the information you're asserting).

-i "mi lukin / mi pali"
I'm asserting this based on direct, present, outside experience.
This includes statements about what I'm doing.

-o "mi sona"
I'm asserting this based on indirect knowledge.
Memory, hearsay, hunches, conventional wisdom...

-e "mi pilin"
I'm stating my opinion or internal feelings.

-oi "o!" (hortative)
Not an assertion: I'm expressing a wish that things be this way,
or a command for the listener to make things be this way,
or (with "moi") expressing that I should make it be this way.

-ai performative
This is true by virtue of me saying it's hereby true!
For example, "I hereby apologize" or "I hereby greet you" (=hello).

-u "anu seme?"
Marks a question. Yes/no, or wh-question if there's a -k- pronoun.

By the way, you can use "nwo-" (lon) to answer yes/no questions,
or you can repeat the verb. Either way, you should probably copy
the Ca from the question verb, and use appropriate evidentiality.

-a marks la-clause
Not really an illocution, just marks the end of Vla.
See "Sentence structure".

## Lexicon
I've only listed the toki pona content words associated with the two stems.
* When used as a root in a noun formative, these have the toki pona noun meaning.
* When used as a root in a verb formative, these have the toki pona verb meaning.
* When used as a suffixes, these have the meaning they have as toki pona modifiers in compounds.

An entry like `-PL- olin, suwi` means: pla- is olin/noun, ple- is suwi/noun, plo- is olin/verb, plu- is suwi/verb.

```
-K- ike, jaki
-KJ- pakala, utala
-KK- moli, kipisi
-KL- kute, kalama
-KS- musi, nasa
-KW- kiwen, ko
-L- jan, soweli
-LJ- kala, pipi
-LK- kasi, kili
-LL- telo, kon
-LN- luka, noka
-LNSKW- kijetesantakalu, /
-LS- soko, /
-LT- oko, uta
-LW- akesi, waso
-M- ma, tomo
-MJ- moku, pan
-MK- namako, misikeke
-ML- mije, meli
-MM- mama, /
-MP- nena, lupa
-MS- suno, mun
-MT- sike, leko
-MW- tawa, kama
-N- nanpa, nimi
-NJ- wan, tu
-NK- lawa, poka
-NL- linja, len
-NM- jasima, /
-NN- mute, ale
-NP- anpa, sewi
-NS- insa, selo
-NSJ- sinpin, monsi
-NT- ante, monsuta
-NW- lon, weka
-P- pona, epiku
-PJ- sijelo, unpa
-PK- pu, ku
-PL- olin, suwi
-PP- suli, lili
-PS- seli, lete
-PW- wawa, lape
-S- pilin, sona
-SJ- ijo, kulupu
-SK- lukin, kule
-SM- sama, tonsi
-SP- walo, pimeja
-SPJ- loje, jelo
-SPW- laso, unu (purple)
-SS- jo, lanpan
-ST- (marks case accessor suffix)
-SW- wile, alasa
-T- toki, sitelen
-TJ- pali, pana
-TL- ilo, kepeken
-TM- mani, esun
-TP- lipu, poki
-TS- open, pini
-TT- tan, nasin
-TW- tenpo, awen
-XZ- yupekosi, /
```

## Ca
It's the fun part of Ithkuil! Here's a table.

```
Extension Perspective Pronoun Configuration
========= =========== ======= =============
ALE: Ø NRM: l² (see ¹) WAN: Ø
INSA: s RPV: t TU: j
OPEN: k N: n KLP: w
PINI: p

¹ A "pronoun infix" in Ca is allowed *only when Ext is Ø*.
* On a noun, they express a genitive: lenmwe = "lenwe mu", twallo = "twalo lu"
* On a verb, they express a subject: pwutni = "ni pwuti"
² NRM -l- is omitted if preceded by non-null Ext or before pronoun "m".
This is equivalent to applying the substitutions sl→s kl→k pl→p lm→m.
```

(The order these apply in to a root is: Configuration → Extension → Perspective. What that boils down to is that -stw- means `root.w().s().t()` "hypothetical part of a group", not `root.s().t().w()` "a group of hypothetical middle parts".)

### Extension
* ALE (Ø) describes a whole thing.
* INSA focuses a middle or part of something. For verbs, it acts like a continuous tense.
* OPEN focuses the start/head of something. For nouns, this might be the source of a river or the entryway of a house — interpretation is a bit creative.
* PINI focuses the end of something. For nouns, this might be like the last bit of food left, or the butt of a cigarette.

### Perspective
* NRM describes a real specific instance.
* RPV describes a hypothetical, unspecific, imagined instance. On verbs, it's like a subjunctive.
* N describes general claims, like "birds fly" (as a rule). It makes "kinds" out of nouns (a bird → "birds in general") and acts like a gnomic aspect for verbs.

### Configuration
* WAN describes "one thing" at a time: one bird, or many birds each in isolation.
* TU describes a pair acting/occuring together, or many pairs each in isolation.
* KLP (kulupu) describes a group acting/occuring together, or many groups each in isolation.

The point here is that configuration is not pluralization. In Ithkuil terms, the distinction between Monadic and non-Monadic is unmarked. So, `lweli` means "a bird" or "birds (a bird here, a bird there)", while `lwelwi` means "a flock of birds" but also "flocks (a flock here, a flock there)". You can use suffixes -anj (one) or -ann (many) to be more explicit about number.

## Interjections

When you say a word that consists of just one letter, it's an interjection.

* a: "a!"
* e: "what!? seriously?"
* i: "same, me too, relatable, ikr, ikwym"
* k: "psh, yeah right, ugh, ew, wtf!"
* l: "yay! yum! how lovely!"
* m: "umm..."
* n: "I see, gotcha"
* s: "don't worry, sure, no prob"
* t: "lol, hahaha, nice, 💯"
* u: "mu!"