/san(n)awi-/, sa-na-wa/i-

HLuw. /san(n)awi‑/ and the other Anatolian lexemes under this lemma point to an original meaning ‘good’, cf. also Hitt. sanezzi- ‘excellent’. Such meanings may have various sources (cf. Buck 1949a:1175–1177), one of which is ‘excellent’ < ‘outstanding’ < ‘standing apart’, cf. Fr. adv. à part ‘to one side, separate, aside’ → ModE adv. apart ‘id.’ vs. ModHG adv./adj. apart ‘unusual, elegant, attractive’. The local meaning ‘apart, distant’ can also be the source of prepositions meaning ‘without’, cf., e.g., Gk. χωρίς ‘apart (adv.); without (prep.)’ (LSJ:2016). Therefore, it is semantically plausible to connect the word family of HLuw. /san(n)awi‑/ ‘good’ with the prepositions Gk. ἄτερ, Lat. sine, TochB snai, etc. ‘without’ as well as with Ved. sanutár ‘apart’ and the other formations listed sub Transmission above.
Based on morphological considerations (cf. below), it is most promising to reconstruct the underlying root as a local adverb PIE *sen ‘apart’.Furthermore, it may be possible to connect PIE *sen‑o- ‘old (vs. young)’ with the current lemma (cf. LIPP II:714), e.g., by a semantic shift ‘distant (of age)’ > ‘old’, cf. similarly Lat. antīquus ‘lying in front’ > ‘old’. It should also be noted that Lat. comp. antiquior and sup. antiquissimus mean not only ‘older’ and ‘oldest’, but also ‘better, preferable, more desirable, more important’ (Lucil.+) and ‘most important, chief, first’ (Cic.+), respectively (OLD²:157 s.v. antiquus 10), thus showing the same semantic shift that is proposed for the Anatolian material above.

LIPP II:710–716 reconstructs a particle PIE *sń̥ ‘apart; without’ as underlying the above-mentioned formations and others.To justify the reconstruction of an accented zero grade, Dunkel (LIPP II:710f.) points out that there are further similar particles, some of which are communis opinio as, e.g., PIE *nú ‘now’. However, unlike PIE *nú ‘now’, a particle *sń̥ is not attested independently, as Dunkel himself admits (LIPP II:710). Furthermore, the only other particle with an accented syllabic resonant, PIE *kŕ̥ts ‘cross, through’ (LIPP II:433–435), is a deradical formation (from PIE *(s)kert- ‘to cut [through]’, LIV²:559f.) and thus either no valid parallel for “PIE *sń̥” or, if it is, the zero grade *sn̥ must also be interpreted as secondary.The full-grade source of this secondary zero-grade formation is either a local adverb *sen ‘apart’ or a root *sen- (adj.) ‘separate’.The verb Hitt. šann(a)‑^(ḫi) ‘to hide, conceal’, if it is a nasal-infix present, is best derived from an adjective or an adverb (factitive ‘to make separate/apart’ > ‘to separate’). Deverbal origin is not likely in view of the semantics of the verb.
Phonologically, especially Gk. ἄτερ ‘without’ speaks in favor of an aniṭ base *sen(‑), while PAnat. *sn̥‑nÉ‑H‑/*sn̥‑n‑H‑´ ‘to hide, conceal’ requires a laryngeal. There are various ways to explain the shape of this verb: it is either a factitive nasal present built on an adverb *sen‑h₁ ‘apart’,On the modal ending *‑h₁, cf. LIPP I:127–133.viz. *sn̥-nó-h₁‑/*sn̥‑n‑h₁- ‘to make apart’But cf. above sub PAnat. Reconstruction on the caveat regarding the required o/Ø-ablaut. Since no adverbial ending or suffix *‑h₂ or *‑h₃ can be reconstructed, the alternative solution is to reconstruct the basis as *senh_2/3 ‘apart’, which, however, is contradicted by Gk. ἄτερ ‘without’.or, according to LIPP II:714, it has to be interpreted as a “praesens departiculativum”, viz. *sń̥h₂o- derived from the univerbation *sn̥=h₂o “(auch) ohne” ← *sen ‘apart’ + 2.*=h₂o ‘also; and’ (LIPP II:334–345).Cf. similarly PIE *pr̥h₂o ‘in front, ahead’ (LIPP II:650–655) and *tr̥h₂ó ‘through’ LIPP II:799–803.Although the problems regarding the phonological reconstruction of PAnat. *sn̥‑nÉ‑H‑/*sn̥‑n‑H‑´ ‘to hide, conceal’ cannot be solved yet, the underlying basis is reconstructable as a local adverb PIE *sen ‘apart’.If the root PIE *sen- ‘old (vs. young)’ is connected – cf. for the derivational process PIE *néu̯‑o- ‘new’ ← PIE *nú ‘now’ –, the tentatively proposed identification of PIE *sen- and PIE *h₂en- ‘old’ as *(s)h₂en- ‘old’ (cf. s.v.) has to be reevaluated.

[A.O.]