|Nasal||m /m~m̥/||n /n~n̥/||ñ /ŋ~ŋ̥/|
|Stop||p b /p b/||t d /t d/||c /c~ɟ/||k g /k g/||q /q~ɢ/||' /ʔ/|
|Ejective||p' /pʼ/||t' /tʼ/||c' /cʼ/||k' /kʼ/||q' /qʼ/|
|Affricate||ts dz /ts dz/||tł dł /tɬ dɮ/||cj /cç/|
|Fricative||f v /f v/||s z /s~ʃ z~ʒ/||ł /ɬ~ɮ/||x /x~ɣ/||h /h/|
|Approximant||w /w/||l /l/||j /j~ʝ/||r /ʁ~ʀ~r~r̥/|
- /s/ and /z/ become /ʃ/ and /ʒ/ respectively before a palatal consonant.
- Unvoiced consonants are realised as ejectives at the beginning of words if they're followed by a stressed vowel with a raising tone. This rule is derived from some kind of allophonic preglottalisation which is in some way present in nearly all dijålekts of łátoskì Uxykascù except Tàkvíkiscë. There are however some exceptions to this rule, yielding minimal pairs like tán /tán/ vs. t'án /t’án/. Similarly, there are also some rare instances of unexpected /s’/, /ts’/ and /tɬ’/, though these have no minimal pairs.
- /j/ becomes /ʝ/ next to /i/.
- Consonants in clusters tend to agree in voicing. Only consonants which do not have phonemic voicing distinction can assimilate.
- The exact realisation of /r/ is quite different among various speakers. As for the voiced allophone, [ʁ] and [ʀ] are the most common realisations, though there are still quite some people who have [r], expecially among the older generations. The voiceless allophone is however still always an alveolar [r̥].
- Palatalised (see section on vokálallofonía) alveolars are realised as postalveolars or (more commonly) alveolo-palatals. This palatalisation also takes place next to long /ï/.
- /ʔ/ is not phonemic word-initially, though it does appear there in words beginning with a vowel. This non-phonemic /ʔ/ is dropped if the preceding word ends in a consonant.
Nearly any combination of two consonants is possible. There are however some restrictions:
- Ejectives don't like to be clustered, although there are some exceptions, such as mák'tìme.
- Plosives at different dorsal POAs can't occur next to each other.
- /h/ can't follow consonants. However, it can precede them in intervocalic clusters.
- /tl/ and /dl/ are impossible. Wherever such a cluster would occur due to compounding, affixation, or borrowing, it becomes a lateral affricate, though still being written <tl> or <dl>.
Possible combinations of three consonants include:
- Fricative (except /h/) + stop (except /ʔ/) + /r/ or /l/
- (only intervocalically) /r l m n ñ/ + stop (except /ʔ/) + fricative (except /h/)
|Close||i y /i y/||ÿ ü /ɨ̹ː ʉː/||u /u/|
|Close-mid||e ö /e øː/||o /o/|
|Open-mid||ë /ɛː/||å /ɔː/|
Ï, ÿ and ü have more or less the same height and frontness and are also distinguished by the presence and type of rounding:
- Ï is unrounded.
- Ÿ has slight endolabial rounding.
- Ü has full exolabial rounding.
- a = /a/ (neutral)
- á = /á/ (high)
- à = /à/ (low)
- ä = /æː/ (long)
- â = /ǽː/ (long high)
Since the low tone is accompanied by vowel reduction it's probably derived from reduced/unstressed syllables in (pre-)PB. This is consistent with the fact that long vowels can't have low tone and that the following syllable, when not having low tone as well, receives a small increase in volume.
We don't really like diphthongs. Nevertheless /ʲ/ and /ʷ/ can occur next to vowels and form kind-of diphthongs, written as <i> and <o> respectively. /ʲi iʲ ʷu uʷ ʷo oʷ/ don't exist.
/a/ is raised to /æ/ next to [ʲ] or if there's an /i/ in the following syllable. Since /c/ derives from historical /kj/, this raising takes place next to /c/ too.
Unstressed final /e/ becomes /ə/.
Low-tone vowels are pronounced with a reduced quality:
- /à/ > [ɐ̀]
- /è/ > [ɛ̀]
- /ì/ > [ɪ̀]
- /ò/ > [ɔ̜̀]
- /ù/ > [ʊ̜̀]
It is because of this reduction that /ỳ/ does not exist.
Word prosody depends on two things: stress and tone. I've already mentioned that a syllable has one out of three tones so there's not much interesting to say about that anymore. That leaves the stress. Every syllable is either strongly stressed, weakly stressed, or not stressed.
The primary ("strong") stress usually falls on the first syllable or on a non-initial syllable with a high and/or long vowel. The position of the stress remains the same in inflected forms. Low vowels never receive strong stress; should a stressed syllable have low tone, the stress is always weak. Strong stress includes a slight increase in length.
All remaining syllables with high/long vowels are pronounced with weak stress. In addition to this, non-low syllables following low syllables also have weak stress.
- dijålekt [ˈdiˑˌʝɔːlekt]
- takavíhki [takæˈvíˑhki]
- càlli [ˌcɐ̀lːˌi]
- càllù [ˌcɐ̀lːʊ̀]
Sentence intonation works as follows:
- "Normal" sentences don't mess with the words' tones.
- Everything that isn't a core argument of the verb (including appositions and the like) is however pronounced with a lower tone.
- Questions start with a falling intonation on the first word and then continue with a rising intonation.
- Subclauses start with a rising intonation and end with a falling intonation on the last word.
In a sentence, verbs and head nouns (i.e. nouns that aren't modifying any other nouns) are more prominently stressed than other words.
Unstressed words are reduced. This basically means that strong stress becomes weak and that weak stress becomes even weaker or disappears entirely.
- Burenia ságo micu na nortox Ócane Pásfike, bý másxu Ałutsxí.
- [ˈbuˑʀeɲæ ˈsʼáˑgo ˈmiˑcu ↘ nɐ ˌnoˑr̥tox ˈócanə ˌpʼǽsfikə, ↘bý ˈmásxu ʔaɬut͡sˈxí]