*k(h₂)et-

The comparison between HLuw. /katt(i)-/ ‘hostility (?)’, Hitt. kaddu- ‘enmity, strife’, and CLuw. kattawatnalla/i- ‘vengeful’ implies a priori the reconstruction of a Proto-Anatolian root *ket(H)- meaning something negative or hostile (cf. Laroche 1965a:51 “*kot-”, HEG I-K:545, HED K:140 “kat-”, EDHIL:455 “*k̑h₂et-”).The root vocalism can be either *o or *a from the Anatolian perspective. Based on the attractive Sanskrit cognate śátru- ‘enemy’, a palatalized *k̑ has been preferred for the reconstruction, which would however be expected to yield Luwian **/zatt(i)-/ following the view of the unconditioned reflexes of *k > Luw. k and *k̑ > Luw. z (ebenso Oettinger 2000c:360 n. 9; pace Mottausch 2011a:73 “k̑ depalatalized before *a”). Assuming a Hittite borrowing into Luwian is not convincing based on the available data, since Hittite only preserves the u-stem, whereas Luwian shows genuine old formations (pace EDHIL:466). On the other hand, if the reconstruction with initial *k̑h₂- is correct, the laryngeal might have prevented the development *k̑ > z in Luwian and caused *#k̑h₂- to fall together with *#k- (in other words depalatalized before *h₂). Along this line of thinking, the reconstructed root *kh₂et- would have some merit.

Whereas HLuw. /katt(i)-/ can go back to either a thematic o-stem (such as a verbal abstract noun *kót-o-), a consonantal stem or an i-stem, Hitt. kaddu- shows the reflex of a u-stem, which would also be preserved in the Luwian derivative kattaw-atn-alla/i- (Weitenberg 1974a:54). The HLuw. derivative /kattun(i)-/ attested to an old *-u̯er/n- abstract formation or to an infinitive that has been substantivized in *-o- or *-i-. This would point indirectly to the existence of an Anatolian root verb */katt-/.

[D.S.]

It is not quite clear whether the verbal root that underlies all the nominal formations in the table is indeed attested in Vedic ŚAT ‘chop up, cut up’ or if this is a separate root (see EWAia II:606 s.v. ŚAT: “Unklar.”); even if ‘chopping up’ is something that might happen in a battle, the implied semantic shift does not seem very intuitive. As all the nominal forms in the table share the notion ‘enmity, ill-will’ or its palpable result in the physical world, viz. ‘battle, fight’, it is perhaps preferable to assume the former as the basic meaning of the root (preserved in the verb Czech katiti se ‘to be annoyed’ < *kōt°?) and leave Vedic ŚAT ‘chop up, cut up’ aside. At least Vedic śátru- m. ‘enemy, foe, rival’ might then still be considered a cognate (but see below).

All formations in the table require or are in line with a root shape *kat- or *kh₂et-; only Vedic śátru- seems to demand a palatal *k̑- as the root-initial consonant, which, however, is at odds with Luwian k- (in /katt(i)-/ c. ‘hostility (?)’) and Slavic k- (in Old Church Slavonic kotora f. ‘fight’, etc.). A way out of this quandary would be to assume that the root was *k̑h₂et- (see above) and that *h₂ impeded the development *k̑ > z in Luwian and *k̑ > s in Slavic. However, one would have to accept at the same time that *h₂ did not impede the development *k̑ > *ć in Indo-Iranic, despite the fact that this branch otherwise preserves the most obvious effects of an influence of *h₂ on a preceding stop, namely laryngeal aspiration (as in *rot-h₂-o- ‘thing having wheels’ > Vedic rátha- ‘chariot’, Avestan raϑa- ‘id.’; *pn̥t-h₂-ós ‘path’ gen.sg. > Vedic pathás, Avestan paϑō; *meg̑-h₂-ó- ‘big’ > Vedic mahá-; see Mayrhofer 1986a:135f.). At least there does not seem to be a counterexample suggesting a development *k̑h₂- > Vedic kh-Indeed, one could claim that Vedic śaśá-, Proto-Germanic *hazan-, *hasan- ‘hare’, Latin cānus ‘gray’ point to a root *k̑h₂es- ‘gray’, and not *k̑as- or *k̑h₁es-, as it is usually assumed.. On the other hand, one could hypothesize that an inherited *katru- or *kh₂etru- > Vedic *kátru- or *khátru-, being totally isolated within the language, was remodeled by folk etymology to śátru- in analogy to ŚAT ‘chop up, cut up’ (śātáyati ‘to knock out, cut off, kill’). A similar instance of folk etymology probably changed an expected but synchronically unanalyzable *ákman- m. ‘stone, rock’ (< *h₂ékmon- from an unknown root) to Indo-Iranian *áćman- (Vedic áśman-, Avestan asman-) in analogy to the root *h₂ek̑- > *AĆ ‘sharp’ (Vedic áśri- f. ‘the sharp side of anything, corner, edge (of a sword)’) or, specifically, the stem *h₂ék̑mon- ‘sharpness, edge’. That *h₂ékmon- ‘stone, rock’ and *h₂ék̑mon- ‘sharpness, edge’ were two separate and etymologically unrelated formations originally is suggested by their being individually continued in Baltic: Lithuanian akmuõ, -eñs m. ‘stone’, Latvian akmens m. ‘id.’ (cf. Ancient Greek ἄκμων m. ‘meteoric stone, thunderbolt, anvil’) versus Lithuanian ašmuõ, -eñs (usually in the plural ãšmenys) m. ‘blade, edge’, Latvian asmens m. ‘id.’.Note that the communis opinio departs from a single etymon *h₂ék̑mon- that meant ‘stone, rock’ already in PIE (cf. NIL:290f. note 7), which, however, is questionable on the semantic side and leaves the Baltic -k- unexplained.

[S.H.]

Hittite kaddu- n. ‘enmity, strife’, Hieroglyphic Luwian /katt(i)-/ c. ‘hostility (?)’ Cuneiform Luwian kattawatnalla/i- ‘vengeful’.See above and EDHIL:466 s.v. kattu-.

(?) Vedic ŚAT ‘chop up, cut up’ (śātáyati ‘to knock out, cut off, kill’ AV+), śátru- m. ‘enemy, foe, rival’ (RV+).Cf., e.g., EWAia II:606 s.v. ŚAT, 607 s.v. śátru-, Werba 1997a:240.

Ancient Greek κότος m. ‘grudge, rancour, ill-will’ (Il.+), κοτέω, -έομαι ‘to grudge’ (Il.+).Cf., e.g., GEW I:931f. s.v. κότος., EDG:761 s.v. κότος.

Proto-Celtic *katu- ‘battle’ (Gaulish catu-° in names such as Catu-rix, Catu-marus, Catu-uellaunus, etc.; OIr. cath u, m. ‘battle, fight; troop’, Middle Welsh cad, pl. cadau, cadoedd ‘battle, war; army’, Cornish cas), *kat-ro- (Welsh cadr ‘handsome; powerful’, Old Breton cadr, Breton kaer ‘beautiful’), Welsh cadarn ‘powerful’, Breton kadarn ‘brave, strong’.Cf., e.g., Delamarre 2003a:111 s.v. catu-, EDPC:195 s.v. *katu-, GPC s.vv. cad, cadr, cadarn.

Proto-Germanic *haþuz m. ‘battle, fight’ (Old Norse hoð f. ‘battle, fight’, Old English heaþu-lāc ‘fighting’, heaþu-rōf ‘brave’, etc.), German Hader m. ‘dispute, quarrel’ (perhaps *haþara-?).Cf., e.g., de Vries 1977a:278f. s.v. hoð, EDPG:214f. s.vv. *haþarō- and *haþu-.

Old Church Slavonic kotora f. ‘fight’, Russian kotorá ‘fight’; Czech katiti se ‘to be annoyed’.Cf., e.g., Vasmer I:646 s.v. которá, GEW I:931f. s.v. κότος.