Comments:The root presents several problems: a) Mong. has ǯ- instead of an expected d-; b) the TM forms are few and rather controversial (it is not really clear whether the Manchu form is related to Evk.); c) the Jpn. form is attested late. There may also have been some confusion with the reflexes of *negre '(female) deer' q. v.
Comments:The Kor. match is glossed as "wild boar" by Ramstedt (in SKE), but as "jackal" or "Korean wolf" in modern dictionaries - thus it may be actually the same word as nɨktä < *l[ù]k`V̀ 'lynx'; but the Mong.-TM parallel is still valid. {Cf. PAA *li:k 'pig'}.
Comments:ОСНЯ 2, 57. A monosyllabic root, but, unlike the 1st p. pron. or the accusative particle, it did not undergo denasalization in PA. This may be explained by the fact that it was in most cases already incorporated into the verbal form as a suffix. It is interesting to note Mong. *büi, *bu 'neg. particle' - which may be originally the same morpheme, but functioning as a separate word and thus subject to the rule *mV > *bV.
Comments:In Turkic cf. *bȫnček 'insect' (Az. miček, Turkm. mȫǯek etc.) - if not, as usually thought, a contraction < *bȫgen-ček (see under *bi̯ūgi).
Comments:Mong. mačag cannot be < Turk., despite EDT 293, Clark 1980, 43; it is also difficult to suppose Man. mačixi (especially the verb maču-) < Mong. mačag. The root thus seems to be reconstructable for PA, with the meaning approx. "to fast, hunger with religious purposes".
Comments:Martin 229; it is not quite clear whether PJ *mat- 'wait' belongs here as well (or rather to *mude 'finally' q. v.). A Kor.-Jpn. isogloss; cf. perhaps Mong. *möči 'edge' (if not = *möči 'limb').
Comments:ОСНЯ 3, 51. An interesting common Altaic religious term (although within Turkic it is rather hard to distinguish from *bāj 'rich' < *bēǯu q. v.).
