Комментарии:Martin 236. Basically a Kor.-Jpn. isogloss, and the precise reconstruction is not quite clear (if *-mp- is to be reconstructed, cf. Man. sampa 'crayfish, crab'?). From other languages one could link a quite isolated Turkish form semek 'fish' (used as a synonym for balɨq).
Комментарии:A Mong.-Tung. isogloss. Note also Tar. saŋgila- 'herabhängen (von den Ohren)' (KW 312, VEWT 400) - hardly a loanword from Mong, but phonetically strange (why -g-?).
Комментарии:An Eastern isogloss; Mong. *sinu- 'to covet' would fit semantically, but phonetically rather continues PA *si̯ā́ńi q. v.; cf. also *sĭ̀ni.
Комментарии:PKE 177, Martin 227 (Kor.-Jpn.). The accent variants in Jpn. are due to a confusion of PA *sắp`í 'stick' and *sī̀pe 'swamp vegetation' q. v. As opposed to *sèp`u 'long twig, rod' (q. v.) this root must have denoted twigs and branches used as firewood and also suitable for making fences (cf. Mong. sibeɣe 'fence' and OJ siba-gakji 'twig fence').
Комментарии:An expressive root, but nevertheless with satisfactory correspondences. Korean has a usual vowel reduction in two of the root's derivatives. Mong. *sabir- may be a result of partial contamination with *sĕpo 'throw, scatter' q. v.
Комментарии:A Western isogloss. Note some of the Turkic derivatives (Kum. sarqɨm 'thin flow', Tat. dial. sarqɨm 'thick whey left after pressing curds' etc., see ЭСТЯ 7) which may speak in favour of the original meaning 'to ooze, be filtered'. In this case the root may actually be derived from *sā̀rV 'to be rare, thinned out' (q. v.) (assuming that length in TM was secondarily shortened); cf. especially the common Western derivative *sār-k`V 'maw used for filtering / fermenting'.
