Comments:In Turk. one has to accept secondary fronting (*titir < *tɨtɨr - if the attested form is not actually tɨtɨr, just wrongly vocalized). It is also tempting to compare OJ t(w)od(w)o, mod. todo `eared seal, Eumetopias jubatus' (for a very similar semantic relation cf. PA *tā̀ŋo), but the vocalism is not quite clear.
Comments:In Turkic languages the root may have been partially contaminated with a synonymic *t`ḕbà 'run' (q. v.) (PT *tabɨĺgan 'hare'); but the traces of PT *togus are still clearly visible.
Comments:The TM word is usually considered to be borrowed < Samoyed.; this seems somewhat dubious because the Samoyed. (Selk. togɨl, Kam. tuɣul, see UEW 451-452) words point not to *-ŋ- ( < Ural. *suŋV-lV), but rather to *-k-, and thus may themselves be borrowed < TM.
Comments:Дыбо 15. The word may be in fact identical (or derived from) *t`ŏk`V 'curved' q. v. Doerfer (TMN 2, 526) tries to dismiss the Turk.-Tung. match for reasons historical (the Manchu forms means "Chinese button" - but in fact it means "metallic button in general"), semantic ("Knopf!" - so what?) and phonetic ("das -n" - so what again? -n is a very frequent TM suffix). The obvious semantic development in TM seems to have been "buckle" > "button".
Comments:KW 398. There may have existed an expressive variant with *t- (cf. some Turkic forms, Mong. doki-ɣur 'drumstick', doki- 'to hit, knock down' (KW 93), Evk. (May.) dokto- 'to hit, chop'); see Дыбо 13. The root is of course expressive (see TMN 2, 460), but seems reconstructable for PA, and distinguishable from *tŭ̀gì 'to pound'.
Comments:Accent correspondences between Kor. and Jpn. are irregular: cf., however, the irregularity in Jpn. dialects (especially Kagoshima tómò, pointing to original high tone in the first syllable).
