Phonotactics

From sona pona, the Toki Pona wiki

The phonotactics of Toki Pona, that is the rules for putting sounds together, are well defined. Unlike its phonology, knowing the phonotactics is not needed to speak the language orally, since the main set of words already adheres to them. It is most useful for creating names and nimi sin. An asterisk marks a sequence that is not allowed phonotactically.

Rules[edit | edit source]

In the lesson "Proper Names", pu lists the phonotactic rules of Toki Pona in the following order:[1]

  1. (C)V(n) syllable structure: "Each syllable consists of a consonant plus a vowel, plus an optional n."
  2. Null onset is word-initial only: "The first syllable of a word does not need to begin with a consonant."
    As a corollary, every syllable after the first does need to begin with a consonant, thus is CV(n). There are no adjacent vowels or diphthongs. The sequence ana is syllabified as a‧na (2 morae), not an‧a (3 morae), following the maximal onset principle.
  3. No *ti: "The syllables *ti and *tin become si and sin."
    This rule concerns palatalization.
  4. No *wo or *wu: "The consonant w cannot appear before o or u."
  5. No *ji: "The consonant j cannot appear before i."
    These two rules cover the same thing. *wo, *wu, and *ji are disallowed because the semivowel in the onset, w or j, can be hard to distinguish from the vowel. The syllables would sound too close to o, u, and i, respectively.

At least one other rule is often noted:

  • No adjacent nasals.
    In sequences like *anna and *anma, the first syllable's coda -n would assimilate to the second syllable's onset nasal: respectively ana and ama.

Coda nasal[edit | edit source]

The coda nasal or moraic nasal is an archiphoneme of syllable-final -n, sometimes transcribed as /N/ and realized as [m ~ n ~ ɲ ~ ŋ] depending on context. It can assimilate to the place of articulation of the following consonant.

For example, anpa may be pronounced as [ˈampa]; linja may be pronounced as ['liɲja];[2] and ma Sonko (a place name for China) may be pronounced as [ma ˈsoŋko].

Other speakers try to carefully pronounce the coda nasal as its written value [n], just like n- at the start of a syllable. This is not always consistent between all words, or all instances of pronouncing them, and assimilation is more common in rapid speech.

Palatalization[edit | edit source]

The syllable *ti is disallowed because of palatalization, a common sound change that has occurred in many languages. An English example is the suffix "-tion", as in "motion". The previous phonemes /ti/ have been palatalized to /ʃ/, so it sounds like "-sion" as in "mission". For the same reason, Toki Pona *ti is changed to si.

As the phoneme /k/ can also be palatalized, many languages merge [ki ti si]. There are no minimal pairs between ki and si in the 120 nimi pu and among all commonly used words, except for the word kin which could be mistaken for sin.

wuwojiti[edit | edit source]

Pronunciation /ˈwu.wo.ji.ti//j/ sounds like English Y, as in "fjord" or "hallelujah".
Usage 2024: Obscure (22% → )Caution: Most speakers don't use this word.2023: Obscure4, Uncommon6 (22% ↗︎ )2022: Rare (13%)
Book and era No book (post-pu)
Part of speech Content word

wuwojiti is a mnemonic word for the disallowed syllables in Toki Pona phonotactics: *wu, *wo, *ji, *ti, and their equivalents with coda nasal. Some uncommon nimi sin flout this rule, often as a joke. The word is also uncommonly extended as wuwojitinmanna to showcase the "no adjacent nasals" rule.

The word wuwojiti possibly originated as a name by Discord user jan Wuwojiti in the ma pona pi toki pona community.[3] Later, the name was used as a word or mnemonic.

The following table shows some accepted alternatives to the disallowed syllables.

Disallowed syllable Alternatives
*wu(n) u(n)[i] ju(n) wa(n)
*wo(n) o(n)[i] jo(n)
*ji(n) i(n)[i] wi(n) je(n)
*ti(n) si(n) te(n)

Notes:

  1. 1.0 1.1 1.2 Chiefly word-initially.

wuwojiti also inspired an obscure, self-demonstrating content word, wawajete, which relates to acts that appear to break rules despite not actually breaking them.

Syllables[edit | edit source]

The following chart shows all 92 syllables under Toki Pona phonotactics. Despite that all of these syallbles are hypothetically possible, only 68 of these are used in the nimi pu. Additionally, 4 marginal syllables are used in nimi ku suli and 3 more in the nimi ku lili.[4]

Null coda Coda nasal
a e i o u an en in on un
j ja je jo ju jan jen jon jun
k ka ke ki ko ku kan ken kin kon kun
l la le li lo lu lan len lin lon lun
m ma me mi mo mu man men min mon mun
n na ne ni no nu nan nen nin non nun
p pa pe pi po pu pan pen pin pon pun
s sa se si so su san sen sin son sun
t ta te to tu tan ten ton tun
w wa we wi wan wen win

Possible words[edit | edit source]

The number of phonotactically allowed words, up to n syllables, is given by the following equation, where σ is the syllable count:

Syllables Range Count Running total
1 a–win 92
2 aja–winwin 6 624 6 716
3 ajaja–winwinwin 476 928 483 644
n

Morae[edit | edit source]

The moraic analysis is an alternative phonological analysis in which the basic unit of timing is the Mora as opposed to syllables. A mora in Toki Pona, similar to Latin or Japanese, may either be a:

  • single vowel (V)
  • consonant–vowel pair (CV)
  • coda n (N)

See also[edit | edit source]

References[edit | edit source]

English Wikipedia has an article on
Phonotactics.
  1. Lang, Sonja. (25 May 2014). Toki Pona: The Language of Good. Tawhid. ISBN 978-0978292300. OCLC 921253340. p. 37.
  2. jan Juwan [@juwan.]. (9 October 2023). Message in #sona-kulupu. ma pona pi toki pona. Discord. Retrieved 12 October 2023. "how do you pronounce 'linja'? (multiple answers are fine)"".
    Option Votes
    [ˈlin.ja] 42
    [ˈli.nja] 3
    [ˈliɲ.ja] 24
    [ˈli.ɲa] 20
    Other 4
  3. Netscape [@Netscape#5728]. (28 April 2018). Message in #toki-pona. ma pona pi toki pona. Discord. Retrieved 11 December 2023. "mi jan Wuwojiti".
  4. jan Talija [@nehemiagurl]. (17 November 2021). "Every toki pona syllable". Discord. Archived from the original on 13 November 2023.