Comments:Basically a TM root, but cf. Yak. kej- 'to butt'. Jpn. *k(ù)i 'fang' may belong here, too - but see under *k`i̯ū̀ge). On the whole, rather dubious as a common Altaic etymon.
Comments:Лексика 78, Doerfer MT 143. Cf. also Mong. geji- 'to dawn' (KW 137), proving that *geɣe- < *geje-. The Korean form points to a cluster with *-j-, therefore a reconstruction *gi̯òjnu (with subsequent assimilative palatalization -jn- > -jń-) is perhaps more plausible; cf. also the variation *n/*ń in PT. The Jpn. reflex is somewhat problematic: loss of final resonant here may be explained by a standard development before a velar suffix (*ka < *gi̯òjn(u)-gV, cf. Mong. *gege-ɣe, Man. geŋ-ǵe); but one would rather expect a PJ form like *ku(i). The irregular vowel may be due to contraction, cf. a similar case in PJ *ká 'mosquito' < PA *kúńe.
Comments:The medial consonant behaves irregularly: the Turk. and Mong. forms point to an intermediate *gi̯ogo, probably a result of assimilation < *gi̯oko.
Comments:АПиПЯЯ 288, Miller-Street 1975, 108ff (with a somewhat different distribution of roots). MKor. kằr-hằi- < *kǝ̀r-hằi- through assimilation (?). The root tends to contaminate with a number of similar roots: see *k`ĭ̀rka, *k`ĭ̀régV, *kìro.
Comments:Martin 232, JOAL 119 (Kor.-Jpn.). Reflexes outside the Eastern area are scarce, but still the comparison seems probable, despite some phonetic problems (in TM *-ü- would be expected).
Comments:One of several similar roots (cf. *ki̯ájo, *ki̯àjú) tending to contaminate, which may explain some irregularities (notably, tone and irregular -a- in Jpn.).
