Teiwo

tentatively the phonology looks as follows

the abrupt release clicks are ǂ series before front vowels and ǃ series before back/central vowels. the ǁ clicks do not appear before front vowels

the prenasalization on the clicks is not really audible word-initially, also the nasal aspirated clicks are voiceless all the way through (i.e. [ᵑ̊ǀʰ] etc)

the tenuis plosive series is actually somewhat aspirated, and the difference between that and the aspirated series is a matter of degree as well as the fact that the ‘aspiration’ on the aspirated series is closer to an [x] or [χ] before /a o/. they do not contrast before high vowels.

/d/ is offset a bit from /t/, being more alveolar while /t/ is basically dental or dental-alveolar

/w l ɰ/ are /m n ŋ/ following nasal vowels. /w l/ are also [m n] word-initially (/ɰ/ is not [ŋ]).

/ɰ/ is written 〈g〉, except in the coda it is written 〈r〉. the ""aspirated"" series is written with h ie 〈ph th čh kh〉 (not sure im a fan) (maybe 〈px tx čx kx〉 looks cooler)

short vowels are kinda reduced, particularly /i o/ are generally [ɪ ʊ]

diphthongs /ai au ąį ąų oi ǫį/ come from coda /w y/ (see below)

long vowels are written kinda inconsistently, /iː aː/ are 〈ii aa〉 but /eː/ is 〈ei〉 and /oː/ is not currently distinguished from /o/ because i think most ways of writing long o look like trash (the name is /teːwoː/ btw) ↳ it's tentatively written 〈o:〉 or 〈å〉 or 〈ô〉.

as hinted above theres some nasal harmony involving the nasal vowels, /w~m l~n ɰ~ŋ/, and the nasalized clicks, essentially like so:

the nasal vowels are marked for nasalization (the clicks are not). nasalization spreads rightwards through vowels and rhinophilic consonants (/w l ɰ wˀ lˀ yˀ ɰˀ h/ + the clicks) and is blocked by rhinophobic consonants (everything else).

the nasalized clicks have a bit of an unusual distribution wrt nasality: they are allowed morpheme-initially and do not trigger nasalization on the following vowel, but otherwise they are only found following a nasal vowel. morphological processes can result in a click following a non-nasal vowel and they don't trigger nasalization of it (however they maintain their own prenasalization).

syllables are maximally (C)C(y)V(C)

the coda consonant is restricted to /p t k 7 š h w l y ɰ/

/Ṽl Ṽɰ/ ⇒ [Vn Vŋ] aw/ay/oy = au ai oi

possibly /b d/ are allowed after long vowels or ptk voice after long vowels or something

there are three types of initial 2-clusters:


 * C₁ is a tenuis click, C₂ is one of /k kʰ kʼ x xʼ/ (other clicks are not allowed in clusters)
 * C₁ is a voiceless sibilant and C₂ is a voiceless unaspirated plosive
 * sč is banned but šč exists
 * C₁ is one of /p t k b d h/ and C₂ is /y/

there is one type of initial 3-cluster:


 * C₁ is /s š/, C₂ is /p t k/, and C₃ is /y/

in general word-internal consonant clusters are kind of rare, ie in most cases a cluster onset occurs word-initially and coda occurs word-finally. word-internal 3-clusters from a coda + onset 2-cluster are possible but rare. 4-clusters do not seem to exist. internal 2-clusters consisting of a coda plosive + onset plosive have a tendency to become geminates.

single-syllable words are necessarily bimoraic ie their nucleus is either a long vowel or a diphthong.

tone.

example words (these have no meaning as of yet)

nąįzą hąųmį gisʼąį xʼo:ga šehai tłeho:

it is not a coincidence that they have 1 heavy syllable and 1 light syllable

the people are horse nomads, and avid practitioners of horse archery. the typical bow design is basically a penobscot bow, quite compact for its draw weight. it is basically a laminated wood/sinew recurve bow with a much smaller recurve bow glued in front of it with their ends attached by cables. they tend to prefer hickory as a wood. they have a rather unusual draw, more similar to the indians of northern california than the horse nomads of central asia. they hold the bow almost horizontally, hooking the string with the thumb and gripping the arrow between the thumb joint and the upper phalange of the index finger, with the ball of the middle finger coming over the thumbnail. the wrist holding the bow is then rotated downward as the arrow is released, dropping the bow out of the way of the arrow. when not on horseback they shoot from a sort of kneeling/crouching/genuflecting position.

Nouns
nouns have basically a free form and a combining form, where the former is derived from the latter by an unpredictable morphological process. (generally, prefixation, suffixation, reduplication, -7- infixation, …) i guess nouns actually have two free forms, singular and plural (there is probably also a dual, derived from either the singular or plural)

noun classes are basically phonological and more or less correspond with how a noun's free form is derived from its combining form

probably some semantically based noun classes for animates?

number of nonce noun classes for a few nouns + ""light nouns""

order within an NP is very free and they do not even have to be continuous. basically everything that agrees is stitched together.

the GIMMICK here is basically to get converbs out of agreement markers (?)

basically, we get relative and complement clauses with verb-GER + attributive agreement with the head noun. we also have certain nouns with fairly abstract meanings (hence: "light nouns") that are often used as the head of a complement clause but are rarely used without one. complemented light noun + oblique case act as a sort of adverbial clause. at least some of these light nouns also have bespoke noun classes, in which case at least some of the time we can omit the light noun altogether and verb-GER-agreement-OBL is essentially a converb form.

4 agreement types: A, P, X, attributive/possessive animates agree nom-acc non-animates agree erg-abs and dont have A

theres probably a certain level of direct-inversion going on

theres really only one marked case, OBLique, but there is also an adverbial derivation -ke which is sort of used as a case by the verb ‘become’, which takes a subject (the becomer) and a noun in the adverbial case (the became thing) (unclear if the latter is agreed with on the verb). the adverbial case also forms a general converb when used with the gerund of a verb.

tentatively, there is a topic marker, which is also the word for ‘if’.

Verbs
medium sized (~100? 200?) closed class

verbs are generally (exclusively?) single heavy syllables. as a side note they also have phonemic contrasts that are not found in morphology, eg the tenuis vs aspirated plosives, clicks are not found in morphology, etc

triple agreement


 * intransitive (S-V)
 * extended intransitive (S-V-X)
 * transitive (S-O-V)
 * extended transitive (S-O-V-X)

-GERund to which noun class agreement markers are added to form relative clauses (attributively agree with the head)

the verb template essentially consists of two parts: the verb root (possibly compounded, as Teiwo: allows verb compounding) + inflectional morphology close to the verb root and a proclitic satellite domain which is mainly adverbial in function and only loosely attached to the root+inflectional domain.

currently not clear where subject agreement goes—it might be in the inflectional domain (i think this is where O and X agreement is) or it might be at the beginning/end of the satellite domain

things that go in the inflectional domain:


 * O agreement
 * X agreement
 * preterite
 * "indirect" preterite (evidential)
 * ""inverse""
 * causative
 * visual-anticausative
 * directionals
 * gerund

things that go in the satellite domain:


 * various aspectual things that are less basic than imperfective/perfective (eg. progressive, habitual, iterative)
 * "body-part satellites" which are like quasi-incorporated body part nouns unrelated in form to the nouns for those body parts
 * these are used for possessor raising, applicatives, and unpredictable lexicalized change in the semantics of the verb
 * some adverbs (quickly, slowly, carefully, badly/aimlessly, adroitly, firstly, now, already, again, still, …)