kišta-⁽ⁱ⁾

The etymological relationship between Hitt. kišta- and Pal. kišta-⁽ⁱ⁾, both with the semantics of extinguishing, is clear (HEG I-K:592f., EDHIL:482f.). The shape of the root is revealed by the attestation of kišnuške/a- ‘to extinguish’ found in the Old Hittite ritual for the disappearing deity with Palaic and Luwian incantations, which leads to the interpretation that the Palaic root verb kišta-⁽ⁱ⁾ and the Hittite root verb attested only in the passive diathesis as well as root noun kāšt- c. ‘hunger’ contain an “enlargement” with -d- (HEG I-K:592-94 with further literature). For parallels in Luwian, see Luw. /skalda-⁽ⁱ⁾/ and CLuw. kwayata-⁽ⁱ⁾ (Sasseville 2021a:348, 358). Adducing Hitt. kāšt- ‘hunger’ and perhaps HLuw. /asta(/i)-/The Hieroglyphic Luwian word /asta(/i)-/ sometimes translated with ‘hunger’ has received a different meaning here and thus different root assignment., we can reconstruct a Proto-Anatolian root of the shape *g/g̑^(h)es- + d^(h)óh₁-/d^(h)h₁-´ (> Hitt. and Pal. kīšta- ‘to extinguish’ and its reanalyzed neo-root *g/g̑^(h)esT-. The Palaic adjective kišta- ‘extinguished’ can be interpreted as *g/g̑^(h)es-to- or *g/g̑^(h)esT-o-, though it was surely no longer properly analyzable for speakers.

It is also tempting to connect HLuw. /kistam(i)-/ ‘east’ here, although this road is paved with semantic and phonological difficulties, if the voiced velar of the root is indeed correct. Disconnecting HLuw. /asta(/i)-/ from the Hittite and Palaic root kiš- would open the possibility of reconstructing an initial voiceless velar; see below.

[D.S.]

The first radical of the root underlying Pal. kašti‑⁽ⁱ⁾ ‘to extinguish’, etc. can thus be any (palato‑)velar. Two roots of the shape *Kes- listed in LIV² may be semantically appropriate on first sight: PIE ?1.*g^(u̯)hes- ‘to eat, devour, consume’ (LIV²:198f.) and PIE *k̑es- ‘to cut (off)’ (LIV²:329). The former is attested verbally only in Indo-Iranian.Perhaps, the PIE word for ‘guest’, PIE *g^hóst(h₂)i- ‘stranger, guest, enemy’ (NIL:173), is related, cf. NIL:173 n. 2 with lit. Furthermore, Neri 2013a:189 n. 41, with reference to Lipp 2009a I:86, adduces the word for ‘hand’, derived from PIE *g̑^hes‑ ‘to take’ which, according to him, can have developed to ‘(Germ.) zu sich nehmen, to ingest’ > ‘to eat’.While the connection with Hitt. kāšt- ‘hunger’ may, with reservations, be accepted (cf. Lipp 2009a II:284 n. 1),However, Lipp’s comparison with the Ved. desiderative jíghatsa- ‘to starve, want to eat’ (AV+) (cf. also Lipp 2009a I:109 n. 27), is not valid.the development to ‘to become extinguished’ via ‘(Germ.) sich verzehren, to eat away’ (cf. Lipp 2009a I:109 n. 27) is semantically too far-fetched. The root is thus excluded for the current lemma.
On the other hand, PIE *k̑es- ‘to cut (off)’ (LIV²:329) is certainly attested in various IE branches, cf. LIV²:329, IEW:586.The Italic continuants, Lat. carēre ‘to lack, be without’, Falisc. 1sg. fut. karefo ‘(Lat.) carebō, I will lack’, etc., are semantically suitable for a word for ‘hunger’, cf. OIr. noíne ‘hunger, famine’ besides Goth. nauþs ‘need’ (EDPC:285, EDPG:385, EWAhd VI:1036–1041). However, the attribution of Lat. carēre ‘to lack’, etc. to the root PIE *k̑es- ‘to cut (off)’ is semantically not trivial (“Semantisch unsicher”, LIV²:329 n. 4 s.v. *k̑es‑), though possible, cf. contexts like “this tree lacks the branch because it is cut off”.Although a semantic relationship between ‘to cut’ and ‘to extinguish’ may be established via ‘to kill’ (cf. CLICS³ Subgraph EXTINGUISH, 2022-01-18), this etymological proposal is too weak to be taken seriously.
Therefore, Pal. kišta‑⁽ⁱ⁾ ‘to extinguish’, etc. is left without PIE etymology for the moment.

[A.O.]