Miwonša [mi’wɔ̃:ʃä] is an a priori conlang, which I have been developping since 2011. It is a personal language and besides that not yet connected to a conworld, although it cannot be excluded that the conlang may be linked to a fictional world in the future. Furthermore, Miwonša was designed to be as naturalistic and suitable for everyday communication as possible. Therefore the grammar is neither very simplistic nor complex, but comparable to the average complexity of European languages. One unique feature of Miwonša is that it can be written with two writing systems. The first one is a modified version of the Latin alphabet, which includes some diacritics.. The other alphabet resembles Hangeul in that strokes representing sounds are put together so that they form blocks. Each simple block represents one closed or open syllable (maximal complexity: CCCVC). Since all blocks for open syllables are symmetrical, you can form ligatures by combining the left half of the first block with the right half of the second block so that it is possible to display a CCCV.CCCV structure with only one ligature.



Miwonša has 27 consonant phonemes and 18 vowel phonemes (5 oral vowels, 5 nasal vowels, 8 diphthongs). The sounds [b, d, g, ŋ, v, x, dʐ, dʑ] are allophones of other phonemes or occur exclusively in loanwords from natlangs, respectively.


1: These sounds occur only in loanwords and can be replaced by their voiceless counterparts in casual speech.
2: Automatically inserted between a nasal vowel and a velar consonant, not phonemic
3. Can be pronounced as labialized post-alveolar sounds as in German or as retroflex sounds as in Polish or Mandarin Chinese (free variation)
4. These sounds are in fact alveolo-palatal and not palatal
5. Only in dialects: Realization of „k“ in coda positions (especially before nasal consonants)
6. Only in dialects: [ɬ] can be replaced by [ɕ]
7. Only in loanwords from natlangs in the letter combination <di> as in <dilema> (dilemma), pronunciation [dʑilɛma]
8. The consonant clusters <pl> and <kl> are pronounced as [pɬ] and [kɬ]. If <l> appears in the end of a syllable, it is pronounced as <ɬ> if the next vowel is a voiceless stop. For instance, „klapsi“ (crazy) is pronounced as [kɬapsi], not as [klapsi]. Furthermore, „walpi“ (snail) is pronounced as [waɬpi], not as [walpi]. However, the onset rule is not applied when /p/ or /k/ are preceded by /s/ or /ʂ/. For instance, „šklami“ (saliva) is pronounced as [ʂklami], not as [ʂkɬami].


1: These sounds as close-mid vowels [e ẽ eɪ̯ eʊ̯ o õ oɪ̯ ou̯] in some positions, especially near palatal consonants (free variation). However, the difference is not phonemic. 
2: Some speakers may pronounce the low vowels as back vowels: [ɑ ɑ̃ ɑɪ̯ ɑʊ̯] (free variation)
3: A frequent mistake by people who try to pronounce my conlang: <an> is a central vowel and more similar to Portuguese „ã“ (as in „manhã“) than to French „an“ (as in „dans“), unless you pronounce all low vowels as back vowels.To put it bluntly: The letter combination „an“  is usually not pronounced as in French.
4. In stressed positions, nasal vowels tend to be slightly longer than oral vowels. However, length is not a distinctive feature in Miwonša. The lengthening just occurs to highlight the nasal quality of the vowels. Between a nasal vowel and a stop or affricate, you need to insert a nasal consonant. The nasal consonant always has the same place of articulation as the following consonant (nasal assimilation): wanpja (answer) – [wãmpja]; antso (something) – [ãntsɔ]; wanca (girl) – [wãɲtɕa]; žanka (text) – [ʐãŋka]

Syllable Structure in Miwonša:[]

Syllables in Miwonša can have the following structures:

(S)C(G)V(V) --- (S)C(G)V(N) --- (S)C(G)V(C)

Onsets can consist of up to three consonants. Codas can only consist of one consonant. Coda consonants cannot appear after nasal vowels or diphthongs (exception: some syllables end in -ans or -anš)

C = Consonant; G = Glide /l, r, j, w/; S = sibilant /s, ʂ/; V = Vowel; VV = Diphtong; VN = Nasal vowel

Consonant clusters in the onset[]

The following chart shows how each consonant cluster in the onset is spelt. In general, onsets can consist of up to three consonants. All consonant clusters in grey can only appear in loanwords.

1. In loanwords, <ti> and <sti> are always pronounced as /tɕ/ and /stɕ/, respectively:                         „tigro“ (tiger) [tɕigrɔ], „stigma“ (stigma) [stɕikma]
2. As mentioned before, [di] is always pronounced as [dʑ].
3. <pl> and <kl> are pronounced as [pɬ] and [kɬ].
4. Some phonemes are spelled differently before the vowels <i, iu, in> than before the other vowels. In particular, <ti, thi, di, hi> are always pronounced [tɕi, tɕʼi, dʑi, ɕi]:                                         tiwa (hate) - [tɕiwa]; anhi (here) – [ãɕi]
5. The phoneme /j/ is spelled [y], if it is the only consonant in the onset, and <j> if it is a part of a consonant cluster. The different spellings help to prevent mispronunciations after nasal vowels:             kanja (raft) – [kaɲa]; kanya (far away) – [kãja]
6. Aside from all the consonant clusters mentioned before, there can be even more consonant clusters in loanwoards from natlangs. One example is the Miwonšan word for Albania, which is „Ščiperia“. (The onset cluster „šč“ is only found in loanwords).

Coda consonants[]

A syllable in „native“ Miwonšan words can only end in six sounds: /p, t, k, tʂ, s, ʂ, l/. However, loanwords can have other coda consonants like /r/ or /m/ or /n/. For instance, the Miwonšan word for „Portugal“ is „Portugal“ [pɔrtugal].


Miwonša is a non-tonal language with lexical stress. There are different rules concerning where to put the stress depending on the part of speech and the grammatical form of the word.

Phonological restrictions[]

The consonant [w] can never appear before <u, ui, un>. The consonants [t, d, h, j, nj, lj, tsj] can never appear before <i, iw, in>. The letter combinations <ti, di, hi> are pronounced as [tɕi, dʑi, ɕi]. The grapheme combination <ji> is impossible in „native“ Miwonšan words.

Orthography (Summary)[]

The Latin alphabet of Miwonša contains 27 letters. The letters Q and X are almost never used except in words of foreign origin (for example names): A B C Č D - E F G H I - J K L M N - O P R S Š - T U V W Y - Z Ž

The following chart contains a summary of how to pronounce each letter and each digraph. Letters in yellow appear only in loanwords from natlangs:


The four declensions[]

In Miwonša, all native nouns either end in "-a", "-i", "-o" or a consonant. They can be grouped into four declensions ("a"-declension; "i"-declension; "o"-declension; "n"-declension). There are four cases (nominative, genitive, dative, accusative) and two numbers (singular, plural). All declensions have two or more sub-types. Nouns that belong to the a- and i- declension are feminine, nouns that belong to the o- and n-declension are masculine. Another feature of nouns in Miwonša is that nouns can contain possessive markers and determiners. All in all, there are 8 interfixes:

{uš} - this that

Interfix Function Interfix Function
{uš} this/ that


{iwh} "his/her" (3SG.POSS)
{uy} any/ some (determiner) {aim} "our" (1PL.POSS)
{iwm} "my" (1SG.POSS) {aiš} "your" (2PL.POSS)
{iwš} "your" (2SG.POSS) {aiw} "their" (3PL.POSS)

The general morpheme order is:

{stem}+{possessive or determiner}+{case/number}

Notes: <> = letters/ {} = morphemes

The "a" declension[]

In the "a" declension, all nouns are feminine. There are three subtypes. The A1 declension is the standard declension. The "A2" declension contains all nouns with a stem ending in <c. j, w, z>. The "A3" declension contains all nouns ending in <r>. Click on the table to enlarge it:

Additional Comments:
1: In the A2 declension, the suffixes {i, iš} get lowered to {e, eš} in the nominative, accusative and dative plural, respectively.
2: [r] is always palatalized to [ž] before {i, iš}
3: Nouns that belong to the a-declension are usually stressed on the penultimate syllable. However, if the noun contains an interfix, you always stress the vowel of the interfix, thus the stress shifts to the next syllable: káika => kaikúša.
4: After the interfix {uš}, you do not need to lower {i, iš} to {e, eš}, because the vowel of the prefix is no longer preceded by a palatal consonant or <w>. Compare: mace (mothers) vs. macuši (these mothers), not macuše. Please note as well, that the interfixes {uš} and {uy} are replaced by {oš} and <oy> after a stem ending in <w> because <w> cannot appear before <u> due to phonotactical restraints:                                                                                                   siwa - water (stem: siw-) => siwoša - this water/ siwoya - any water 
5: Before the interfixes {uš} and {uy}, [r] does not need to be palatalized anymore. Compare: tara (day) - taži (days) - taruši (these days), not tažuši
6: After the interfix {uy}, the suffixes {i, iš} are always lowered to {e, eš}.
7: In the A2 declension there is a morpho-pohonological alternation before the interfixes {iwm, iwš, iwh}: If a stem ends in <pj, c, kj, fj, sj, zj, hj, tsj, lj, mj, nj>, the final letter is changed to <p, t, k, f, f, s, z, h, ts, l, m, n>. Examples: wanca (girl) vs. wantiwma (my girl) - kholja (ring) vs. kholiwma (my ring), wonsja (victory) - wonsiwma (my victory)
8: In the A3 declension, a final [r] turns into [ž] before the interfixes {iwm, iwš, iwh}.
9: The interfix <iwh> is pronounced [iw] before the suffixes {i, iš}. Therefore, the "h" is silent: kaikiwhi - his cats (pronunciation: [kai'kiwi], not [kai'kiwɕi]

The "o" declension[]

While the declension in the singular is equal for all nouns of the "o" declension, there are two sets of plural suffixes for animated and inanimated nouns. Animated nouns ending in <j> or <y> take the suffixes <e>, <ek> and <eš> in the plural. Final <r>s of noun stems are palatalized:

Example nouns of the "o" class (an., type 1): wonco (child), škunlo (thief) wonti (children), škunli (thieves)

Example nouns of the "o" class (an., type 2): anro (father) anži (fathers)

Example nouns of the "o" class (inanimated): žanjo (eye)  žanja (eyes)

The "i" declension[]

All nouns of the "i" declension require the same suffixes in the singular and plural. The only irregularity is the "i"-alternation (see section about phonology):

Example nouns of the "i" class (no alternation): raiči (opinion)  raič(opinions)

Example nouns of the "i" class (alternation): kaipi (mountain)  kaipja (mountains)

The "n" declension[]

With very few exceptions, native nouns of the "n"-class are very regular:




Overview: Verb morphology in Miwonša[]

Although the verb morphology is out and away the most complex part of Miwonšan grammar, it is not as complicated as one might assume at first glance. There are, however, some features which make it hard and easy at the same time: On the one hand Miwonšan verbs have an immense amount of of grammatical categories and affixes, which is the cause of extremly high morpheme/word ratios and very long words. On the other hand there is only one single and very regular conjugation, very little allomorphy and very strict rules concerning where to place which affix. This leads to the conclusion that it is much more reasonable and economical to obtain an idea of the structure of Miwonšan verbs by taking a look at the whole morphological scheme.

First of all it is important to know that there are three main verb types: 

(1) Standard Verbs (containing Modal Verbs); (2) Adjectival Verbs; (3) Copula Verbs

All three types can appear as finite and as infinite verb forms. There are several types of infinite verb forms:

(1) Infintives, (2) Adverbial Participles, (3) Adjectival Participles, (4) Gerunds, (5) Supines, (6) Gerundives

Furthermore, there are five voices (genus verbi), two aspects, at least five tenses, three grammatical moods and both subject and object agreement:

Verb morphology
Genus Verbi Active (1), Passive (2), Reflexive (3), Reciprocal (4), Impersonal (5)
Aspect Imperfective (1), Perfective (2)
Tense Present (1), Past (2), Future (3), Present Perfect (4), Past Perfect (3)
Mood Indicative (1), Subjunctive (2), Imperative (3)
Agreement Nominative (1), Accusative (2)

Of course, not all combinations of grammatical categories are possible. Nevertheless there is a fixed sequence of affixes which serves as a basis for a model with which it is possible to describe the verb morphology quite accurately:

Seven-Slot Model
Slot 7 Slot 6 Slot  5 Slot 4 Slot 3 Slot 2 Slot 1
(Stem) Interfix Tenses STEM

Subject Agreement

Object Agreement; Voice

Mood,  Verb Type

The Seven-Slot Model[]

The following table contains an overview of all used morphemes and their position in the verb:
























Used abbreviations:

E - epenthetic morpheme

GER - gerund

IMP - imperative

IND - indicative

INF - infinitive

INTR - intransitive

IPFV - imperfective

IPRS - impersonal

IRR - irrealis

PAST - past tense

PFV - perfective

PTCP - participle

REFL - reflexive

RECP - reciprocal

SBJV - subjunctive

SUP - supine















































Slot 1: Aspect and Mood[]

In Miwonša, there are several types of infinite verb forms:


(1) Yežiman lufiwa. - I want to sleep.

Gerunds (can be declined):

(2) Lufjanza nanhwo saipiran. - Sleeping often is healthy.


(3) Munsišanje čwonai čwonjan zužanti aš tonšuloš žwena Miwonšak. - The students read the book to learn more of the Miwonšan language.


(4) Munsišanje zužanti kwan aš tonšuloš žwena Miwonšak. - The students must know more about the Miwonšan language.


(5) Žai rumiwku hanzi lufjanši! - Do not wake up sleeping dogs!

Slot 2: Voice, Transitivity and Object Agreement[]

Slot 3: Subject Agreement[]

Slot 4: Verb Stem[]

Slot 5: Tense[]

Slot 6+7: The Perfect and the Future Tense[]

Tense, Aspect, Mode and Verb Type: Possible Combinations and restrictions[]


Finite verbs[]





Adjectival Verbs[]

The copula "kwa"[]

Copula kwa conjugation.png