Comments:Лексика 335. The match is acceptable (as one of the Common Altaic monosyllabic verbal roots) if -rǝ in Jpn. is a historical suffix. Cf. perhaps also TM *sali- 'esteem; price' - which may reflect a merger of this root with *sáĺo 'love' q. v.
Comments:EAS 85, KW 317, Владимирцов 194. Poppe 29, 59 (Turk.-Mong.), АПиПЯЯ 15. Mong. is hardly borrowed from Turk., despite Щербак 1997, 144. *-jg- is probable because of the TM reflex and loss of *-g- in Jpn.; Korean high tone (rather untypical for a verb) is also possibly due to contraction (see also *sắjV 'shallow place' - a possible source of contamination).
Comments:Дыбо 1993. A Turk.-Tung. isogloss. Not quite reliable because of poor attestation in TM. Note, however, that PM *sigaji 'knucklebone' would be a good match - if it were not derived (or secondarily contaminated?) from *siɣa, see under PA *sìŋù.
Comments:EAS 153. The Jpn. word may be related if one assumes an original meaning 'interval (in general)'. Mong. *söɣem < *seɣöm (with a frequent labialization transfer).
Comments:KW 314, SKE 225, PKE 173, Дыбо 12, АПиПЯЯ 297. The unexpected -b- in Mong. may be due to a mixture with *serbe- 'bristle, stand on end', see under *sirp`a. Medial *-j- is to be reconstructed to account for the loss of *-r- in Kor.
Comments:KW 320, Владимирцов 284, Poppe 29, 67, TMN 3, 236 ("möglich"). In Korean the original root may have interacted with *sắi- 'leak, ooze' < *săjgo q. v.
Comments:An Eastern isogloss; cf. perhaps also Oyr. sagalak 'рыбка-урюп', Mong. (L 657) saɣamqa 'salt water fish, perch(?)'. Like in many fish names, details are not clear: note that the TM forms may be actually borrowed from Jpn. sakana `fish' (originally `wine-side dish'). Cf. also *suku.
Comments:A not quite secure Tung.-Jpn. isogloss. Since words for 'hunt' often go back to names of hunted animals, one can think of comparing PT *sajkak ( ~ *sajgak) 'antelope' (attested since Chag., see VEWT 395, ЭСТЯ 7), with the resulting correction of the PA reconstruction to *sajk`a. Cf. also *sĭgò and *sūku.
