Comments:The root denotes a wild plant, probably thorny; the vocalic reconstruction is not quite certain. In TM one would expect a *čial-, but the diphthong may have been distorted in a long form with initial affricate; Mong. *čulukir may similarly represent a later assimilation < *čalu-kir. In that case one could reconstruct *č`i̯ála.
Comments:A Mong.-Tung. isogloss. Not quite secure: the TM forms can be alternatively derived from *čiKü-ki, a diminutive of *čiKi (*čiKü) 'pivot' (q. v. sub *č`ĭ̀k`o ); cf. with the same anatomical meaning Ewk. čiki, Ewn. čị̄q, Neg. čịx, Ud. cigi (ТМС 2, 391, Корм. 310). For this etymological direction see Дыбо 1988, p. 120.
Comments:SKE 20, EAS 64. Turkic forms are attested late and can be < Mong.; however, Turkm. čāGa with a long vowel is hard to explain as a loan. On the other hand, cf. Karakh. čekün 'young of marmot' (EDT 415), Evk. čekše 'tarbagan': if these words are related, the PA reconstruction should be changed to *č`ek`a.
Comments:EAS 64, KW 419, Poppe 26, Martin 244, Menges 1984, 266, АПиПЯЯ 76. Jpn. tone is irregular; it may be, however, due to an influence of another root (cf. *čĕk`a, OJ toko 'always, eternally' /accent unknown/). Mong. čag may be < Turk. (see TMN 3, 27-28, Щербак 1997, 112), but may as well be genuine. The Kor. reflex is quite regular (the attested čǝk is just an orthographic variant of the expected *čjǝk), despite Doerfer's doubts in TMN ibid.
Comments:A Western isogloss. The medial cluster is not quite secure (perhaps one should rather reconstructed *č`i̯ŏli with different suffixes). It is interesting to mention MKor. čjǝ́mpúr 'horse's mudguards' (Nam 426) - perhaps a (somewhat distorted) loanword < Mong. čilbur.
Comments:A common derivative *č`i̯op`è-rV is reflected in Mong. *čöɣe-rü-m and Evk. čupe-r; another suffixed form, *č`i̯op`è-k`V (originally diminutive) may be reconstructed on the basis of Evk. čop-kī and Kor. čabä-gi. The root must have denoted a big water container, perhaps both artificial and natural.
Comments:The vocalism is uncertain; it is possible that the TM form should be removed from here and compared with Jpn. *tùrù, see notes to PA *čălù. In such case a reconstruction *č`aro for the present root would be possible.
Comments:Mong. čike-n in names of plants may be a merger of this root and *č`ĭk`a q. v. Kor. soksăi 'horse-tail' is hard to separate from Jpn. tukusi id.; most probably we are dealing with an assimilation here (soksăi < *čoksăi).
Comments:An expressive common Altaic root; both in Turkic and Japanese its reflexes may have partially coincided with the reflexes of *č`ī̀mu 'top, edge' q. v., but in Mongolian and TM they seem to be distinct.
Comments:A Western isogloss, somewhat questionable phonetically. Unclear is -k- in TM and Mong. (*-g- would be expected). A possible solution is to reconstruct for all three subgroups a suffixed form *č`i̯ugu-kV.
