Comments:Martin 244 (Kor.-Jpn.). PJ *kama- presupposes a nasal suffix (*ki̯ap`a-nV). A very similar fish name is found in TM (Oroch kiampai 'чилимс', Nan. qǟfu 'гольян (озерный)', see ТМС 1, 397); cf. also a quite isolated Turkm. kepir 'carp' (if not from Pers. kopur, see ЭСТЯ 5, 47). Whether these forms reflect the same root is yet to be determined.
Comments:An interesting root, demonstrating the acquaintance of the speakers with the way of life of the salmon (weakening and dying after releasing roe). Note a common derivative *ki̯áta-rV in the Turko-Mongolian area.
Comments:Reasons for voicing in TM and Turkic are not clear (assimilative development in a suffixed form *ki̯át`ó-gV?; note that in PTM we can also reconstruct *kudege - the -ge would have been lost in all languages where the root is attested).
Comments:АПиПЯЯ 111. For Jpn. cf. also Koguryo /xuče/ 'mouth'. The root must have basically meant 'mouth or throat cavity' (whence 'mouth', 'cheek' (originally probably 'inner side of cheek') and 'trachea'), with a secondary development > 'cavity in general' > 'well' in TM.
Comments:KW 158, Poppe 52. Despite TMN 3, 579, Щербак 1997, 132 Mong. is hardly < Turk. Cf. also Evn. kēde 'friend' (ТМС 1, 443). The Kor. word is compared alternatively (JLTT 438) with OJ kara `clan, family', but the latter seems to have a better match in TM *kal-, see *kala.
Comments:Martin 245, АПиПЯЯ 15, 16, 71, 103, 279 (but the Turkic parallel has to be attributed to a different root, see *gā́t`ì). In TM one would rather expect *kid- - but the PTM reconstruction is dubious anyway (based only on one dialectal Evk. form).
Comments:Владимирцов 210. *-jb- is reconstructed to account for Mong. -b- rather than *-ɣ-. Note, however, that modern Turkic and Mongolian forms interact actively: some Turkic forms may be backloans from Mong. quwa (which itself is probably a Turkism, see above).