Comments:EAS 76, KW 275, Martin 232 (Jpn.-Kor.), ОСНЯ 2, 86-87, АПиПЯЯ 69, 292, Дыбо 12, Мудрак Дисс. 91. The original meaning must have been 'soft, tender' (generally, or particularly - of weather, fruits etc.). An expressive root with phonetic variation, tending to contaminate with *ni̯ombu 'thin' q. v. (hence vocalic variation in Mongolian and Tungus; but despite Doerfer MT 69 it is hardly appropriate to regard the TM forms as mongolisms). In Mong. cf. also nalmi-gar ( < *namli-gar?) 'excessively soft, weak'.
Comments:An interesting common Altaic root, with quite plausible phonetic and semantic correspondences; however, some contaminations with *nuŋu 'chief, master' were possible.
Comments:KW 476. A Western isogloss. Medial -m- in Mong. is probably secondary (assimilatory here). Note that all languages reflect a derivative *ńi̯ūno-gV (in Turk. cf. *jōn-ga 'chips, wooden shavings' = Mong. *ǯomu-gu-l = TM *ńüŋ(g)-).
Comments:SKE 172, Дыбо 10, Лексика 149. In Jpn. one has to suppose either a secondary dissimilation (*ùmì < *mùmì), or a preliminary assimilation *ńi̯ū̀ŋńu > *ŋi̯ū̀ŋńu ( > *ùmì).
Comments:KW 281, Poppe 1974, 121, АПиПЯЯ 293. A Tung.-Mong. isogloss. The TM form can hardly be regarded as a mongolism, therefore (despite Poppe 1966, 31) this root should be carefully distinguished from ТМ *ńuŋńakī 'goose' (= Karakh. juɣaq id.). Mong. has parallel forms niɣusun and noɣusun - possibly resulting from contamination.
Comments:Лексика 202, Дыбо 1995b. Tone in Kor.-Jpn. is irregular. In Jpn. one has to assume metathesis (*míná- < *nímá-), quite probable in a root with two nasals.
Comments:PKE 132, Poppe 38, АПиПЯЯ 72. Turkic *-u- is not quite regular here (but on the basis of the attested forms a reconstruction *jola is also not excluded). The Korean form may belong here only if the meaning "burnt smell" (not "rank" or "foul smell") is original.
