Comments:EAS 55, KW 457, SKE 209, Lee 1958, 110, Poppe 12, 112, Цинциус 1984, 57, Martin 226. Not borrowed in TM < Mong., despite Rozycki 81. In Turk. the root has probably merged with *ŭĺi 'to freeze' q. v. (cf. the meaning 'to blow (of a cold wind)' in Chuv.).
Comments:SKE 1949, 209, Lee 1958, 109, Цинциус 1984, 61, АПиПЯЯ 291. The Mong. form can belong here only if the attested HY form o[li]mi is aberrant, reflecting the stage of language (or dialect) with h- already lost. Otherwise Mong. reflexes are to be separated from this root. On possible Turkic reflexes see under *ŭla.
Comments:EAS 55, Цинциус 1984, 53, Lee 1958, 110, АПиПЯЯ 75, Martin 228. Mong. *hurul regularly < *hulu-r. As for the Jpn. form, it may be a reduplication (as thought traditionally; in this case *pǝ- < *p`ùl(o)-gV), or reflect an already suffixed form, cf. e. g. Nan. polpị ( < *p`ùlo-pV).
Comments:EAS 55, Цинциус 1984, 61, АПиПЯЯ 78, 277, Rozycki 82. Despite Doerfer MT 80, TM cannot be borrowed from Mong. Initial *p`- (not *p-) is reconstructed because of *h- in Mong. and high pitch in Jpn. Cf. *póńe.
Comments:Mergers with *pā̀t`e were possible (esp. in Jpn., where otherwise -a- would be expected). Reasons of the loss of -n- in Mong. are not clear. Cf. PAN *-petpet 'firefly'.
Comments:In Turk. *üŋ would be expected; the vocalism was probably influenced by the homonymous *oŋ / *öŋ 'colour, exterior'. Kor. has a usual verbal low tone.
Comments:Mong. also has a back variant *(h)uri-, reflected in *(h)uri-ldu- 'to contest, fight' (WMong. urildu-, uraldu-, Khalkha uralda-) = PT *uru-ĺč- id.
Comments:EAS 53, 125, Poppe 101, Цинциус 1984, 55. A Western isogloss. Despite TMN 2, 87, Щербак 1997, 162, Mong. cannot be borrowed < Turkic; again, despite Doerfer MT 18, TM cannot be < Mong. (all because of archaic *p- preservation both in Mong. and TM). Quite enigmatic is the relationship of this root to synonymous *ŭk`urkV q. v.
