Comments:All forms reflect a suffixed *ènŋù-kV. A certain phonetic problem is raised by u- in Mong.; one has to assume PM *unu-gan (cf. the early TM loans: Evk. unukān 'foal', Sol. unuxũ 'kid'; see on it Poppe 1974, 128) with vocalic assimilation from earlier *anu-gan. Otherwise the etymology seems quite plausible, and appears to explain the Jpn. form better than the usual theory of Chinese borrowing (neither MC mạ́, nor OC mrāʔ can explain the initial um- in Japanese).
Comments:ЭСТЯ 1, 292. In TM cf. perhaps Evk. dial. erbeɣu 'name of a shrub with yellow flowers, used instead of tea' (ТМС 2, 462). Mong. *üril 'plum' should be kept distinct from *ölir 'small apple' (see under *ójle); however, since the sequences -l-r and -r-l in Mong. are liable to metathesis, it is difficult to decide which protoforms are actually reflected in *üril and *ölir. In any case, labialization in Mong. is probably secondary here (either *üril < *irü-l or *ölir < *elör < *erö-l).
Comments:Cf. Nan. erĩ 'rope for binding up the yurt with ritual purposes' (ТМС 2,463), suggesting that *-k- may be originally a suffix. Despite Щербак 1997, 162, Mongolian forms are hard to explain as borrowed from Turkic.
Comments:The root denotes wormwood, artemisia in Turkic and Mongolian; in Korean the meaning is not quite certain (probably some sort of shepherd's purse, Capsella bursa-pastoris). The phonetic match is precise, anyway, and the reconstruction seems reliable.
Comments:Цинциус 1972a, 37-40. Poppe 51, 56, АПиПЯЯ 296, Дыбо 6; a weak attempt of disproving the etymology see in TMN 1, 188. A "nursery" word, represented in all Altaic subgroups. Ramstedt (KW 129) compares the Mong. form with Turk. *eči / *eče 'elder brother', but it is obviously closer related to *ata (Mong. *eči-ge = Turk. *ete-ke / *ata-ka). The form *ete-k`e, reflected in Turkic, Mongolian and Tungus, is probably an archaic diminutive, and Doerfer's (TMN 2, 9) attempt to treat the Turk. form as a compound ata 'Vater' + äkä 'älteres Geschwister' is certainly faulty.
Comments:A Western isogloss. The medial cluster with *-b- should be reconstructed in order to account for labialization in Turk. and Mong. However, the Turkic form may be alternatively compared with Mong. ojun 'thought' - if the latter is not related to PT *ōd 'thought' (as suggested by Владимирцов 285 and supported in Мудрак Дисс. 103).