Comments:KW 163, Poppe 17 (Doerfer TMN 3, 479: "aus phonetischen Gründen unsicher"). The meaning 'head' in Kor.-Jpn. (whence 'brain' in Kor.) is secondary, with a universally usual semantic shift ( < 'bald spot'; cf. Slavic *golъ 'bare, naked' and *golva 'head' etc.; see Фасмер 1, 429). -čhí in MKor. may either reflect the variation of the reflexes of *-ĺǯ- (it is interesting to note also MKor. kắrčhjǝ̀ŋ 'young skin of a plant', which may actually reflect the same root), or a compound with -čhi 'stuff, thing'. Cf. also *kèĺčo 'to scrape, rub': the two roots tend to be confused with each other.
Comments:KW 225, АПиПЯЯ 112, 274, Лексика 262. Borrowing in Mong. from Turk. is impossible (despite Щербак 1997, 126); the Manchu form, however, may be < Mong. (see Rozycki 137).
Comments:One is also tempted to compare PJ *kákí 'persimmon' ( < *kem-k`a), but the word can be alternatively compared with TM *gaka-kta 'cranberry' (ТМС 1, 136). Cf. also other similar plant names: *ki̯omga, *gi̯aŋu, *k`éma.
Comments:ЭСТЯ 5, 107 (Mong.-Tung.). In Mong. the root has several expressive variants (cf. also PA *k`eŋa). In Kor. an early palatalization occurred (*-ŋi- > -ń-). See also notes to PA *k`ḗmŋV 'wide'.
Comments:Дыбо 5, Лексика 267. A Western isogloss. The root is sometimes difficult to distinguish from *kḗp`V 'belly' and *k`ébá 'corpse', due to natural contaminations.
Comments:АПиПЯЯ 281, Дыбо 5, Лексика 277. Ramstedt (SKE 89) compares the Kor. form with Mong. qarbiŋ etc. (see *k`i̯ắŕme), but this seems improper (the modern form kalbi goes back to MKor. kari-spjǝ). The Japanese form is somewhat insecure due to its late attestation and unclear suffix. In Mong cf. perhaps kerseŋ 'brisket'.
