Comments:ССЕ 201. Werner 2, 321 <*ud'ǝ>. The reconstruction of -d'- ( = our *-ǯ-) is based on the Yug form and is certainly incorrect: Kott. -t- can only point to PY *-t- (otherwise -j- would be expected). Ket -d- and Yug -d́- here are evidently results of assimilation < *ʔut-ǯV- (where -ǯ- is a widespread verb-forming suffix).
Comments:ССЕ 201. Initial n- in Ket and Yug forms is secondary, being derived from formations like da-nam-s-ut < da-d-mam-s-ut 'sucks the breast (mam)' etc. Werner 2, 27 *ut < *qut (with *q being reconstructed for no obvious reason at all).
Arin:u-ta (Лоск.) "long (adv.)" [-ta is probably a suffix, as in pata "far"]
Comments:ССЕ 201. Werner 2, 323 <*ug(ǝ) / *ug-dǝ>.
Proto-Yenisseian:*ʔuǯ- (~x-)
Meaning:up along the stream
Ket:ūta1 // uta5, ūtĺ1 'river source', ud-bej5 'wind from the river source' (i. e. blowing up along the stream)
Yug:út́ej; ud́-bej 'wind from the river source'
Kottish:ujā 'up along the stream', ujal 'above'
Comments:ССЕ 202. Werner 2, 368. The Ket and Yug forms reflect rather *ʔuč-, most probably as a result of secondary assimilation: *ʔuǯ-Ka > *ʔuč-Ka > ūta.
Proto-Yenisseian:*ʔuǯem (~x-)
Meaning:gudgeon (пескарь)
Ket:uŕem5 / ūŕem1, pl. uŕemn5 (South.); Bak., Sur. udǝm5, pl. udǝmn5
Yug:ud́im5 ,pl. -ɨŋ5
Comments:ССЕ 202. Werner 2, 355 *ud'ǝm. ?Cf. also Ket úran, MKet. údǝn 'snail', South. úrɔn 'leech' (Werner 2, 352, 354).
Comments:ССЕ 202. Werner 1, 267. The PY form is hard to reconstruct because of its expressive nature and poor recordings. Probably a compound of some sort - but Werner's analysis as "long leg" most of all resembles a folk etymology within Kottish.
Comments:ССЕ 202. Werner 2, 366, saying that the reconstruction *ʔūs- is questionable because a morphemic alternation us / uj is observed. It is, indeed, observed, being part of the productive Ket -s- / -j- alternation in verb forms and compounds (with -j- appearing before voiced consonants) - but does in no way contradict the reconstruction of PY *ʔūs-.
Comments:ССЕ 202. The attribution of the Arin and Pumpokol forms is somewhat dubious, not just because of unclear morphological structure, but also because of irregular -r-. On the whole the root is expressive and the reconstruction is somewhat difficult; one should also consider a possibility of a Turkic loanword (suggested in Stachowski 1997/2, 233); see Werner 1, 389.
Comments:ССЕ 203. Werner 1, 220 compares the Kott. form with Ket. ĺɔj 'to lean on' and reconstructs <*ʎok->. This is dubious because d́- in Kottish is definitely a prefix (see the paradigm above), while in the Ket form it is a part of the root, for which Werner himself (2, 11) reconstructs a protoform <*loʔǝjǝ / *loʔǝj> (see *ŕōj), thus contradicting himself. Besides, tone 4 (original length) in *ŕōj does not favour reconstruction of any intervocalic velar in the latter root. All these consideration do not allow us so far to make the reconstruction for Kott. d́-aki any more precise than *ʔVKV.
Comments:ССЕ 203. The comparison is possible if Arin t́- (t-) is a prefixed morpheme. Werner (1, 322) doubts the comparison without giving any reasons.
Comments:ССЕ 203. Werner 1, 16 (doubting the connection of the Kott. and Arin forms); the reconstruction is not quite clear, indeed - due to the absence of Ket and Yug forms.
Comments:In ССЕ 308 the Kott. form is attributed to *ʒVl- 'speak': this reconstruction is probably erroneous. We should probably reconstruct *ʔVĺ- and reduplicated *ĺVĺ- - cf. Werner 2, 6-7 <*ʎaʔǝlǝ / *d'aʔǝlǝ>. Velar q-, k- in Arin and Pumpokol are in this case most probably preverbial elements.