Despite Abayev (3,15) and Shagirov (2,149) the comparison of PAK *śǝħV́ with PN *sag 'deer' is impossible for phonetic reasons. But there is an interesting parallel in Kartvelian (PK *šw-el- 'roe, chamois'; the comparison with NC see Trombetti 1923, 370).

    See Хайдаков 1973, 14.

    The meaning 'hunting' in Ub. is secondary, being influenced by a similar (phonetically and semantically) root represented in Ub. sʷā-ḳ́á 'hunter', PAK *š:a-ḳʷa- 'to hunt', PAT *šʷara-ća- 'to hunt' (the latter form demonstrates a similar merger of two originally distinct roots) < PWC *š́:ʷa- 'to hunt'. The components *-ħV in PAK and *-xǝ in PAT (*šʷara-xǝ) are unclear. They may represent originally independent roots (for PAK *ħV cf. PEC *wHɨ̆rχ_V 'mountain goat', for PAT *-xǝ cf. PEC *ɦānxV 'deer' q.v.), but since they are attested only in compounds and in different subbranches of WC, such etymologies at present seem too risky. The roots for 'deer' (*š́ʷV) and 'hunt' (*š́:ʷa) are correctly separated by Shagirov (1977,2,149) in an argument with Dumézil (1967,160; 1969, 112). Also correct is his separating of Abkh. a-čá 'female deer' and Ub. žʷa 'deer' (see PWC *č́ʷa / *ǯ́ʷa). However, his separating PAT *šʷVrV from PAK *śǝħV́ is hardly justified (the Ub. form ʎa-šʷá is not even considered).