Mvithizu is the language of a nomadic pastoralist people of the same name, who inhabit a wide area of tropical savanna. (Properly, in this language the people are Mvithizukye and the language is Mvithizutsi, but at least for the time being I've decided to use Mvithizu for both in English.) They are periodically in contact with the sedentary agricultural Chondru, as well as with other nomadic groups. To the small extent that I've envisioned their culture thus far, they are somewhat based on the Maasai.
Design Principles
Mvithizu, like Chavakani, isn't inspired by one real-world language group but by several, and unlike with Chavakani I'm having a harder time identifying a primary inspiration as ideas come to me. Typologically, the following things can be said about Mvithizu:
- It is heavily synthetic, favoring suffixes
- It is head-marking
- It has relatively free word order but defaults to SOV along with other head-final traits
- Its morphosyntactic alignment is primarily ergative and dative
- It has quite a few noun classes
Nasals: /m n/ m n
Lenis stops/affricates: /b d dz ɡ/ b d dz g
Fortis stops/affricates: /p t ts k/ p t ts k
Lenis fricatives: /v ð z ɣ/ v dh z gh
Fortis fricatives: /f θ s x/ f th s kh
Other: /w l r j/ w l r y
Vowels: /i u ɛ ɑ/ i u e a
Onsets are compulsory while codas are forbidden. Onsets consist of a base consonant optionally preceded by a homorganic nasal (written m before labials and n otherwise) or followed by a glide /w/ or /j/. All consonants except /m n w l r j/ may be preceded by a homorganic nasal, and all consonants except /w j/ may be followed by either glide. While this means that phonetically Mvithizu has prenasalized consonants, since nearly every consonant can be prenasalized in this language I think it's more parsimonious to analyze them as clusters of a nasal and a following consonant rather than as unitary phonemes as in Chavakani.
The exact difference between the fortis and lenis consonants varies between dialects. Although I've transcribed them with the characters for voiceless and voiced consonants for simplicity, for most speakers voicing is not the primary distinguishing feature. Fortis consonants are typically held for somewhat longer than lenis consonants, and for some speakers they are aspirated. Some but not all speakers partly or fully voice the lenis consonants. The overall situation resembles that of Ojibwe.
Phonetically, /d t ð θ/ are lamino-dental, and /n l r/ are apico-alveolar. /dz ts z s/ may be lamino-alveolar or apico-alveolar. The velars front significantly before /i/ and /j/, more so than is common in English, becoming more or less cardinal palatals. Although I've transcribed the vowels /i u ɛ ɑ/, each has a pretty wide range of realizations, dependent more on free variation than allophony: /i/ may be [i ɪ ɨ ɪ̈], /ɛ/ may be [e ɛ æ], /ɑ/ may be [ɑ ä ɐ ʌ], and /u/ may be any of [u ʉ ʊ o ɵ]. Vowels are generally nasalized before a following prenasalized consonant. For many speakers, utterance- or even word-final vowels devoice.
Mvithizu is more or less syllable-timed, with nasals and glides not adding to the length of syllables. Because of this, vowels are generally slightly shorter before clusters or fortis consonants. It does not have contrastive stress or tone, although most commonly speakers will pronounce the first syllable of a word with a pitch-based stress and pronounce weaker secondary stresses on the third, fifth, and any subsequent odd-numbered syllables after that.
Up next: Noun classes, and possibly also pronouns!