*pek^u̯-

We posit for Hittite pakku-šš(a)- ‘to crush’ an imperfective verbal formation with the suffix -šš(a)-, comparing pakkuššuwar to the verbal substantive ēššuwar of the archaic form ēšš(a)-/īšš(a)- ‘to make’. The suffix -šš(a)-, which originally had consonantal inflection in the plural (e.g. īštēni : iššatteni), conjugates according to the half-consonantal ḫi-conjugation. The fluctuation between a consonantal and a half-consonantal stem is reflected in the derivatives pakkuššuwant-: pakkuššawant-. For back-formations in -want- see Oettinger 1988a.

A sigmatic stem was suggested based on the verbal noun pakkuššuwar; cf. Oettinger 1979a:212.Kronasser 1966a:404f. suggests a foreign origin for the morpheme -uš-. Puhvel (HED PA:60f.) compares the Hittite verb tekku(šš)-, which he traces back to PIE *dek^u̯s-, and assumes that the final /s/ is part of the root. However, the fact that he omits the Lydian cognate paq- in his reconstruction casts doubts on his analysis.Oettinger 1979a:185 does not make any suggestion regarding the s-extension of this specific verb.Jasanoff 2003a:119 n. 70, 174 f. does not include pakkušš(a)- in his discussion of various s-extensions found in Hittite.

Melchert 1992a:41 was the first to connect Lydian wc-paq-ẽni-^(t) with Hitt. pakku-šš(a)- and thus reconstructed a labiovelar *k^u̯ for the root final consonant. A reconstruction *peg^u̯- would have been possible based solely on Hittite pakku-šš(a)-, since Hittite /g^u̯­/ becomes a fortis consonant before /s/, cf. akkuške/a- besides aku-. However, Lydian paq- precludes a reconstruction with *g^w- (Melchert 1994a:336). Whereas Lydian offers two possibilities, Hittite pakku-šš(a)- shows an original o-grade of the root. Since the accent in Lydian wc-paq-ẽni- was on the imperfective suffix -ẽni-^(t), the original vowel of the root may be either *e or *o. This comparison brings evidence against the assumption of an anaptyptic u-vowel in the Hittite verb, which was suggested byOettinger 1979a:212. Furthermore, Oettinger (loc. cit.) reconstructs *pah₂k- or *pah₂ĝ-, which is not supported by the Lydian evidence.Kimball 1999a:299 argues that the laryngeal drops in this environment causing compensatory lengthening, which in turn lenites the following voiceless stop. Thus, the geminates -kk- in Hitt. pakku-šš(a)- would speak against this reconstruction. However, this argument is not valid, because Hittite /g^u̯­/ becomes a fortis consonant before /s/, see above and cf. Kloekhorst 2008c:618, HED PA:59-61, HEG L-P:386-388.

[D.S.] [E.R.]

PIE 2.*pek^u̯- ‘to make ripe/cooked’ is well attested by verbal and nominal formations outside Anatolian (LIV²:468, LIV²⁺ s.v., IEW:798, NIL:548–552, EIEC:125). According to Janda 2000a:48–51, HED:60f., and Puhvel 2010b:213f., this root might be related to 1.*pek^u̯- ‘to pound, crack, crush, grind (grain)’. The latter meaning, which is attested in Anatolian, could then have been the original one, while ‘to make ripe/cooked’, which is preserved in the non-Anatolian languages, developed secondarily. As argued by Janda 2000a:49, the starting point for the semantic shift could have been “die Sphäre der Getreidezubereitung, in der der Bedeutungsübergang ‘genießbar machen’ → ‘kochen’ vonstatten ging” (cf. also LIV²:468 n. 0, Maier 2013a:115–117 s.v. paksuu̯ant-). This fits well the attested usage of Hitt. pakku-šš(a)- ‘to crush’ (see above sub PAnat. Reconstruction), which “involves the softening up or smoothing of hard or harsh substances, specifically cereals, to ready them for consumption” (Puhvel 2010b:213). Thus, under the assumption of a semantic development ‘to crush (grain)’ > ‘to make edible’ > ‘to make ripe/cooked’ (rejected by Kloekhorst 2008c:619 as “not […] semantically attractive”), the two roots 1.*pek^u̯- and 2.*pek^u̯- can be combined.Another root whose allegedly primary meaning is preserved in Anatolian, while a secondary meaning is attested in the non-Anatolian languages might be PIE *men- (extended form: *mneh₂-): the primary meaning ‘to see, look’ is restricted to Anatolian (e.g., CLuw. mimma- ‘to regard, recognize, favor’ < *mi-mn-, CLuw. manā- /mnā-/ ‘to see’ < *mneh₂-). Outside Anatolian, however, we find a meaning ‘to think, remember’ (e.g., Ved. mánya-^te ‘to think’ < *mn̥-i̯e/o-, Gk. μιμνήσκω ‘to remind’). As per LIV²:447 n. 1, “[…] dürfte die Grundbedeutung der Wurzel ‘sehen’ sein, woraus dann ‘innerlich sehen’ → ‘sich vorstellen’ → ‘gedenken’”; cf. also Steer 2016a:221f.PIE 3.*pek^u̯- ‘to trust’, however, has to be separated from both 1.*pek^u̯- and 2.*pek^u̯-.

[T.S.]

Only in Anatolian; for attestations, cf. above.

Verbal stems: PIE pres. *pék^u̯-e/o- (Ved. pácati ‘to cook, bake, fry’, YAv. -pacaiti ‘id.’, Lat. coquere, -ō ‘id.’, W pob- ‘to bake’, OCS pekǫ, inf. pešti ‘to bake, fry’, Lith. kepù, inf. kèpti ‘to bake, fry’Metathesized from *pekù, cf. LIV²:468 n. 2., Alb. pjek ‘to bake, fry’); PIE pres. *pék^u̯-i̯e/o- (Ved. pácyate ‘to ripen’, Gk. πέσσω, Att. πέττω ‘to ripen, digest, cook’)Cf., e.g., LIV²:468.

Nominal stems: PIE *pek^u̯-ti- (Ved. paktí- f. ‘cooked dish, cooking’ [RV+], Gk. πέψις f. ‘cooking, ripening, digestion’ [Hp., Arist.+], OCS peštĭ f. ‘oven’), PIE *pek^u̯-te/or- m. ‘cook’ (Ved. paktár- m. ‘the cooking one’ [AV+], Lat. coctor m. ‘cook’ [Petr.+], Gk. πέπτρια f. ‘female baker, cook’ [Hsch.]); for further nominal formations, cf. NIL:548–552.

Verbal stem: PIE *pék^u̯-u-/*p_ek^u̯-u-´ u-pres. (TochB päkw- ‘to rely on, trust’, pres. I 3sg. midd. paktär, 3pl. midd. pkwantärCf., e.g., LIV²:469, Malzahn 2010a:703.

Nominal stems: Ved. pāká- adj. ‘unsuspecting, foolish’ (< pre-PIIr. *pok^u̯-ó-), Gk. πέπον voc. sg. ‘my dear!’Cf., e.g., LIV²:469 n. 1, NIL:550 n. 4, each with further lit.