Comments:EAS 96, TMN 3, 308; Лексика 168. A Western isogloss. The root seems plausible, but has some phonetic irregularities. Instead of Evk. suǯakī one could perhaps compare Tung. *šolčikī 'ferret' (ТМС 2, 405). Shor čolasčaɣa, Khak. čolǯanaχ 'ласка, хорек' (VEWT 115) is probably < TM.
Comments:Doerfer MT 241. The Mong. vocalism is not quite regular: perhaps -i- instead of u~o under the influence of the following *-j- (which has to be reconstructed to account for Kor. -i-).
Comments:Basically a Turk.-Mong. isogloss; the Kor. form is late attested and somewhat insecure. Cf. perhaps also Manchu suŋǵen 'clever, smart' ( < *'perceiving'?).
Comments:The Korean reflex is somewhat dubious, being traditionally derived < *sār- 'to live' (sară-čap- = 'to take alive'). This, however, may be a secondary folk etymology, the original meaning of the compound being 'to take booty'.
Comments:Лексика 449. Mong. cannot be a Turkic loanword, despite Щербак 1997, 150. The form must be traced to *süt-sün > *sü-sün (with a regular dissimilative development > *üsün). Kor. has a frequent loss of vowel between a fricative and a stop. The original meaning must have been 'liquid' or 'milk-like liquid' (with drops or bubbles on the surface) - whence, on the one hand, "sweat" and "soak", on the other - "milk" (in the Turk.-Mong. area).
Comments:EAS 108, VEWT 424, Doerfer TMN 3, 325-326 (with quite unjustified doubts), MT 240 (Turk.-Tung.), Miller 1970, 129, Street 1980, 300, АПиПЯЯ 288, Дыбо 13, Мудрак Дисс. 41-42, Лексика 228. The comparison with Kor. and Jpn. seems very probable, although here we have a rather rare case of tonal irregularity: one would expect a low tone in Kor. and a high tone in Japanese. In Jpn. the tonal irregularity could be explained by a contamination with some different root - e.g., with the root reflected in PT *sanč- 'to pierce' (ЭСТЯ 7); but the Kor. high tone is still unexplained. It is interesting to note the common Turkic-Mongolian derivative meaning "2-year-old sheep or camel": Turkic *siĺek / *diĺek (see VEWT 424), MMong. šilegu, Khalkha šilbe, Bur. šülge (certainly not a loan in Mong. < Turk., despite Róna-Tas 1970, 215 with a detailed analysis of the words, see Poppe 1974, 123). The variant *diĺek (attested in MK tišek and in Yak. tisege, see EDT 563) is an obvious innovation in a part of Old Turkic dialects, following the replacement of *siĺ 'tooth' by *dīĺ q. v., just like WMong. sidüleŋ "2-year-old sheep" is an innovation following the replacement of original *sil 'tooth' by the suffixed formation *sil-dü > sidü.
Comments:ТМС 2, 76, АПиПЯЯ 79. Despite Poppe 1966, 198, Doerfer MT 52, TM cannot be borrowed from Mong. The irregular voicing in Jpn. ( < PA *-p-) must be due to a confusion with a very similar root, *sìmpá ~ *símpá 'firewood' < PA *sắp`í; Jpn. dialects show a great deal of confusion between these two roots.
Comments:Cf. also PT *sɨ̄pɨŕgu 'flute' (ЭСТЯ 7). In TM one may note Neg. sịwsaxawụn 'wire (for pipe cleaning'), Nan. sịosaqo 'wire (for burning a hole), sịosan- 'to burn (a hole with a wire)', Oroch siapti 'mouthpiece (of a tobacco pipe)' (ТМС 2, 74, 75). These words are hardly borrowed from Mong. sibsiɣür, but are rather reflexes of original *sibe- 'pull out, pick out' (see *zi̯upi) influenced by the Mong. word.
Comments:АПиПЯЯ 275. The Jpn.-Kor. comparison see in Martin 235 (with a somewhat fantastic reconstruction). Actually Korean drops -g- quite regularly, while -r is an obvious suffix; the vowel -u-, however, is a problem (*-a- would be expected), and should be probably explained as a result of contraction (thus *sù'ɨ̀r < *sagV-bVr or the like). Existing TM forms (see ТМС 2,106) are clearly borrowed from Mongolian, see Doerfer MT 39, Rozycki 186. Cf. also Karakh. suɣdɨč 'winter parties, arranged by turns'.
