Number: 99
Proto: *čančV ~ *čačV
English meaning: to walk, step, go
German meaning: schreiten, gehen
Mordovian: šašto- (E M), čašto- (E) 'sich bewegen; schieben, zuschieben', šanšne (E), šančk (ardo-) (M) 'galoppieren; in vollem Galopp fahren', šanža (M) 'langer Schritt, Sprung'
Khanty (Ostyak): soč- (Trj.), čuš- (DN), sos- (O) 'schreiten, zu Fuß gehen'
Mansi (Vogul): šōš- (TJ), šoš- (KU), šuš- (P), sūs- (So.) 'waten'
Nenets (Yurak): tańetā- (T) 'traben (langsam)'
Enets (Yen): toddeʔe- (Ch. B) 'laufen (vom Rentier)'
Nganasan (Tawgi): tantā- 'laufen'
Selkup: caaca- (Ke.), čaadša- (N), taača- (Čl.) taada- (OO) 'fahren, gehen', tanda- (Ta.), candsa- (Ke.), čăndša- (N) 'hinausgehen'
Kamass: tōn- 'gehen, wandern'
Addenda: Mot. тандалным 'скочу'
K. Reshetnikov's notes: Kms. and partly Selk. (to other etymological groups within Sam.); Nen., En., Ngan., Mtr. and partly Mord. > rec. 1800. In the corresponding entry of UEW, we evidently deal with four (!) different roots erroneously united. The Mord. root represented by Erz. šanš-ne, Mks. šanč-k 'im vollen Galopp', Mks. šanž-a 'langer Schritt, Sprung' etc. is likely to be a cognate of Sam. *tǝntV- (~ *c-, *-c-) 'laufen, traben (von Tieren)' (Janh.SW 147) > Kms. thon- 'galoppieren' (not mentioned in UEW), Mtr. танд-, NSam. *tǝnt-> Nen. тӑн-ета-, En. todd-eʔe-, Nga. tant-ā (accepting this comparison, we must reconstruct the Proto-Sam. form as *cǝncV-, supposing *č/a/nčV for PU), while Kms. t(h)ōn 'gehen, wandern' goes back to Sam. *tantǝ- (~ *c-, *-c-) 'treten' (Janh.SW 151) and the Selk. word attested as STa. tanda-, SKe. candsa-, SNm. čăndša- 'hinausgehen' - to Sam. *cǝnca- 'steigen' (Janh.SW 31); it should be stressed that all the three roots are to be distinguished from each other (so, in fact, we deal with material from three - not two! - etymological groups presented in Janh. SW). As for the Mord., OU and Selk. words listed in this record, they are (contrary to Rédei) apparently not related to the roots diskussed above, but it seems that they are really related to each other, going back to a protoform like *č/a/čV. Note that in this case (as well as in URAET 96) we have no variants of the type *CVCV/*CVNCV to deal with, since it appears more adequate to separate such forms as belonging to different etymological groups. See also URAET 1800, SAMDET 37 and MRDET 30.
References: FUV; DEWO 241; Paas.Beitr. 123, 129; Coll.CompGr. 53
