Notes :The -w-/-v- in Kuwi is obviously epenthetic; it is, however, strange that we do not find such a -w- in the forms with deictic stem *a- (we have Kuwi_F āasi, etc.) This epenthese prevents us from firmly reconstructing a glottal stop between the two morphemes, for the development *-ʔ- > Kuwi -w- is not observed anywhere else. It may be that the original form *īv- has been replaced in Kui by analogy with the forms of the pronoun *a-.
Kuwi (Schulze) :imba'asi, imba'ari, imbai, imbaɨ/imbaari "which man, which men, which woman or thing, which women or things"
Additional forms :Also Kui embe, imbe where; bāi who (< imbai); Kuwi_F imin what kind; imbi, imbi'a where; Kuwi_S imbinaɨ which thing, which things; adj. for all genders imbini; imbia where
Notes :One of two examples on PK *y-, this interrogative pronoun in PK had the meaning 'who', opposed to *yān- 'what' and *ec- 'which, how many'. Lost everywhere, the initial *y-, however, left serious traces everywhere, since in Kui it left the variants i- and e-, while in Kuwi_F *ya- > a- before the following -a- and > i- before the following -i-.
Kuwi (Schulze) :ēna, ēnaɨ "which thing, which things"
Additional forms :Also Kui ananju, anaru, anari, anai what man, what men, what woman or thing, what women or things; adj. ani; anavani, inavani where; anaṛiki, anaki, inaṛiki, inaki why; Kui_K ini- = ani-; inari = anari; Kuwi_F adj. eni; eneki how; Kuwi_S ēnari which men
Notes :The second case of PK *y-, this time before a long vowel. In Kuwi the reflexation *yā- > *e- is stable (although Kuwi_F has short e-, while Kuwi_S has long ē-); in Kui the usual reflexation is a-; however, the original *y- is betrayed by forms like inavani, which probably have preserved traces of *y- because of old reduction *yānavani > *yanavani > *inavani. The pronoun was used as the inanimated 'what' in PK.
Kui :kappa (kapt-) "to swallow liquid hastily, gulp, drink"
Kuwi (Israel) :kap- (-it-) "to eat fast like a dog"
Additional forms :Also Kui kavali giva to chew the cud; Kuwi_Isr gab- (-it-) to eat fast
Notes :In Kui kappa < *kabpa. In Kuwi_Isr gab- is the original form (< *kab- with assimilation), and kap- is the p-conjugation, although here -p- has become so closely connected to the root that it has lost its morpheme status and became part of the root (this explains the past tense -it-). The form kap- can be also considered as a borrowing from SDR.
Kui :kasa (kasi-) "to bite, sting; n. a bite, sting"
Kuwi (Fitzgerald) :kacali
Kuwi (Schulze) :kaccinai
Sunkarametta Kuwi :kacc- (-it-) "to bite"
Parja Kuwi :kaca kupi "scorpion"
Additional forms :Also Kui pl. action kaska (kaski-); Kuwi_S kaca kupi scorpion; Kuwi_S kaskeri centipede
Notes :The forms for 'scorpion' mean literally 'stinging crab' (see 1734). The lack of gemination here is irregular; it may be due to specific word-contraction reasons.
Additional forms :Also Kuwi_F kāhi kīali to amuse; kahū a game
Notes :Vowel length in Kuwi_F is due to compensatory lengthening because of impossibility of geminating -h-. In the form kahū the lengthening probably did not occur to avoid two long vowels (although this is not a strict rule in Kuwi, and etymologically long vowels are never shortened before long vowels of the endings). NB: the same lengthening is characteristic of Kuwi_Isr.
