monovanlent wordsOmzinesý wrote: ↑24 May 2021 02:02 Ideas for a lang where I wanna avoid SAEisms
- nouns and monovalent verbs are one class (generic/verby vs. non-generic/nouny marked with article)
- bivalent verbs mark voice (direct, inverse, reflexive)
- no tense but realis but irrealis moods
- bunch of aspectual clitics that are semantically 'optional'
- head marking
- many stop-fricative clusters like /kf/
- pitch accent (Every word (by definition) has a domain of high-tone syllables. aááa is allowed but áaaá is not. A glottal stop may appear as a suprasegmental feature on the right edge of the high-tone doman.
i ɯ u
ɛ ʌ ɔ
Back-vowels have a harmony where only rounded or unrounded ones appear in one word (or maybe some sequence in a word).
Vowels can be short or long (phonemically two vowels). Rising diphthongs ɛi, ɔu, ʌɯ also appear.
(1) p t k q
(2) f θ s ɬ x χ
(3) ʋ ð̞ l ɣ̞ ʁ
(4) m n ŋ ɴ
(5) ʋ̃ ð̞̃ l̃ ɣ̞̃ ʁ̃
Nasal approximants nasalize the following vowel, which is the main phonetic feature to recognize them.
Allowed clusters are (1)(2), (1)(3), (4)(5).
(1) 'The boy is running.'
pógó tsitéa
pógo-´ tsitéa-`
run(er)-PREDICATE (is_)boy-PARTICIPANT
(2) 'The one running is a boy.'
tsitéá pógo
tsítéa-´ pógo-`
run(er)-PARTICIPANT (is_)boy-PREDICATE
They must be able to work as modifiers of participants ('N who is a boy,' and 'N related to the boy'/'N of the boy'), too. That has some other (tonal?) change. If all those distinctions are marked suprasegmentally, there must be some other means than just pitch accent. Maybe I could go a bit towards Semitic vowel alternations, but not too much.
There are aspect markers and modality/polarity markers. Aspect markers can well be clitics appearing in the end of the first word, but modality/polarity markers could be more "verby". Maybe they always appear in the end of the predicate.
-------------------------
Suprasegmental features
There are four suprasegmental features. They are mostly inflectional/derivational, but the first low-pitch domain is often lexically determined.
1) High-pitch domain
2) Glottalization
3) Nasalization
4) Vowel reduction
1) High-pitch domain
This lang can be described as a pitch-accent system.
There are two levels of pitch: high <á> and low <a> or <à>. Every vowel has one.
The high-pitch vowels appear in a domain that can be one or more segments long. There can be low-pitch domains in the beginning and in the end of the word. So the word stucture od domains is: (low-pitch domain) high-pitch domain (low-pitch domain).Every pitch domain can be one or more vowels long. Especially, predicates are long in this language, so the domains are often long too.
tsitéá has a low pitch-domain on tsi and a high-pitch domain on téá.
tsítéa has a low pitch-domain on tsi, a high-pitch domain on té, and a low-pitch domain on a.
2) Glottalization
Glottalization is surely the easiest of the suprasegmental features. It is a binary feature of every word. There can be a glottal stop in the end of the high-pitch domain. Glottalization is marked with the circumflex on the last vowel of the high-pitch domain, <â>.
mógûpla
3) Nasalization
Every domain is either nasal or oral. Nasalization is not as limited as pitches. A three-domain word can have any of the eight combinations of nasalization: OOO; OON, ONO, NOO; ONN, NON, NNO; NNN. Non-nasal domains are far more frequent, though. I'm not sure how to mark nasalization. The easiest way could be just <n> in the end of a nasal domain.
4) Vowel reduction
This is the suprasegmental feature I have though least about. The basic vowel phoneme inventory is:
i ɯ u
ɛ ʌ ɔ (probably near-low rather than mid-low)
All vowels can be reduced to the corresponding mid vowel: e̞, ɤ̞, o̞. That is, vowel height is reduced (or according to an alternative analyses, a third height added). It could be that vowel reduction can only appear in low-pitch domains. Reduction could thus be marked with gravis and no accent. It remains to be seen if reduced or non-reduced vowel is the more frequent one and thus left without gravis.
5) Vowel rounding harmony
Vowel rounding harmony can be seen as the fifth suprasegmental feature, but it is not based on the domains. A word cannot have both rounded and unrounded back-vowels.
How to call the domains?
pre-stress domain, stress-domain, post-stress domain (It is not a stress accent but stress is a handy word)?
pre-peak domain, peak-domain, post-peak domain (domains are more than just high pitch though)?
first low-pitch domain, high-pitch domain, second low-pitch domain (even more emphasizes pitch)?
first domain, second domain, third domain (all words don't have three domains, and thus second domain can be first, which is messy)?
Ideas?