*h₁E(h₁)s-

The Hittite adjective āššu- ‘good’ has been convincingly connected with the medio-passive āššiye/a- ‘to be dear to’ (HEG A-H:90, Lehrman 1998a:180, EDHIL:215f.). In the other Anatolian languages, we find the following cognates of Hitt. āššu-, i.e. Lyd. asfãν (Gusmani 1975d:169, Melchert 1994a:336, Hajnal 2004a:197) (i.e. PAnat. *(h₁)o(h₁)su-) and possibly ašνo-^(d) ‘to make good(?)’, HLuw. /assunuwa-⁽ⁱ⁾/ ‘to make good(?)’; see under lemma, and also of āššiye/a-, i.e. Luw. /azza-/ ‘to love’ and Lyc. A asa-^(ti) ‘id.’. The causative formations can be compared directly to Hitt. aš(ša)nu-^(mi) ‘to take care of’, whose etymological relation to āššū- is however debated; cf. Weitenberg 1984a:94, EDHIL:217f. with further literature. Since the meaning of the Luwian and Lydian cognates are not assured, no Proto-Anatolian formation in *-nu- will be reconstructed until further evidence has surfaced. In sum, a (nominal and verbal) root meaning ‘(to be) good, dear’ can be reconstructed for Proto-Anatolian (along with an adjectival u-stem), *h₁E(h₁)s-. The exact shape of the root however is not obvious. In order to explain the geminate -šš­- seen in Hittite, an assimilation between *h₁ and -s- has been posited, e.g. *h₁oh₁s- (EDHIL:223). Differently, Melchert 1994a:63 assumes a limited version of Čop’s Law for Proto-Anatolian, i.e. *#é.CV > *#áCCV; not accepted in EDHIL:224. Further evidence would be welcome.

[D.S.]

The PIE u-stem adjective *h₁ē̆su-, *h₁e(h₁)su- vel sim. means ‘good’. The exact shape of the root is unclear (*h₁e(h₁)s-?). Some consider a connection with the PIE root *h₁es- ‘to be’ (LIV²:241f. s.v.). In EIEC:235, a stem *h₁(e)su- ‘good’ is reconstructed that is “[u]ltimately related to *h₁es- ‘to be’”. Similarly Watkins 2000a:24 s.v. “(e)su- Good”, “[o]ldest form *ǝ₁(e)su-”, which, in his view, was originally a “suffixed form of *ǝ₁es-, to be [< ‘being true, being good’]”. 

To account for the u-stem forms attested in Anatolian (cf. Hitt. āššu-/āššau̯- adj. ‘good’, āššu-/āššuu̯- sbst. n. ‘good’ < pre-PAnat. *h₁E(h₁)su-), Indo-Iranian (Ved. sú adv. ‘well’, su- ‘good, well’ [in compounds] = Av. hu- ‘id.’, OPers. u(v)- ‘id.’ < pre-PIIr. *h₁su-), Greek (ἐύς adj. ‘good’ beside ἠύς adj. ‘id.’, ἐύ, εὖ adv. ‘well’ < pre-PGk. *h₁(ē̆)su-), Celtic (OIr. su-, so-, MW hy-, etc. ‘good, well’ [in compounds] < pre-PCelt. *h₁su-) and perhaps other IE languages, various derivational scenarios have been proposed. While the etymological connection of these forms is generally accepted, certain phonological and morphological details, especially regarding the Anatolian material, still remain unclear. Cf. the discussion and literature in NIL:239f. n. 1. As for Anatolian, “[t]he biggest problem of the Hittite word is the geminate -šš-,” according to Kloekhorst, EDHIL:224, who critically discusses the explanations proposed so far. In his view, the best reconstruction for Hitt. āššu-/āššau̯- adj. ‘good’Beside āššu-/āššuu̯- sbst. n. ‘good’, which “is a mere neuter substantiv[iz]ation of it”, as per EDHIL:224. is *h₁oh₁s-u-, with *-h₁s- > Hitt. -šš-. Laryngeal h₁ is suggested by the non-colored e- of Gk. ἐύ- ‘good, well’. This solution also allows to maintain the etymological connection with the above-mentioned u-stem forms of the other IE languages. As far as the stem formation is concerned, Kloekhorst (EDHIL:224) considers “an o-reduplicated u-stem adjective” *h₁o-h₁s-u-, to which he compares Gk. ὠκύς ‘quick, swift’ = Ved. āśú- ‘fast’, which “must reflect *h₁o-h₁k̑-u-” (EDHIL:224). Differently, e.g., Schaffner 2005a:276 n. 603 (following Schindler 1986b:389 with n. 16), who derives Gk. ὠκύς, Ved. āśú-, Av. āsu- ‘quick, fast’ from a possessive compound PIE *oh₁-h₁k̑ú- ‘having swiftness; swift, quick’.Somewhat differently LIPP II:323, 330 with n. 34: *h₂o-h₁k̑ú-, with a local adverb *h₂o ‘next to, at; toward’. For further discussion and lit., cf. NIL:200f. with n. 1, where an unspecific basis *HeHk̑- ‘fast, quick’ is reconstructed. It was formed to an unattested noun stem *h₁ók̑u-/*h₁ék̑u- ‘swiftness, quickness’, which could also have been the derivational base of PIE *h₁ék̑u̯-o- m. ‘horse’ (Ved. áśva-, Av. aspa-, Lat. equus, etc.), a substantivized possessive adjective *h₁ek̑u̯-ó- ‘provided with swiftness’ (Cf. Schaffner 2005a:276 with n. 604, LIPP II:330 n. 34, each with further lit.). According to Schaffner 2005a:276 n. 603, Gk. ὠκύς, Ved. āśú-, Av. āsu- are not likely to continue a primary formation (as assumed, e.g., in EDHIL:224), because no feminine forms are attested for Ved. āśú-, Av. āsu- and their comparative and superlative forms. This fact, however, fits the reconstruction of a possessive compound. – Whether Gk. ἠύ- ‘good’ reflects a lengthened-grade adjective *h₁ēsu-, an e-reduplicated adjective *h₁e-h₁s-u- (cf. EDHIL:224) or a metrically lengthened form of a stem *h₁esu-, remains unclear. Cf., e.g., NIL:242 n. 14 with further discussion and lit. 

[T.S.]

Ved. sú adv. ‘well’ (RV+); cf. also su- ‘good, well’ (RV+) as first member of compounds (e.g., Ved. su-śrávas- adj. ‘having good glory’ [RV+]) = Av. hu- ‘id.’ (e.g., Av. hu-xratu- adj. ‘having good strength [of mind]’), OPers. u- ‘id.’ (e.g., u-martiya- adj. ‘having good followers’)Cf., e.g., EWAia II:734–736 s.v. sú¹ and sú², NIL:239, 240f. n. 4, 250 n. 5. 

Gk. ἐύς adj. ‘good’, acc. sg. m. ἐύν (Il.[+]); ἐύ, εὖ adv. ‘well’ (Il.+); ἠύς adj. ‘good’, acc. sg. ἠύν, nom./acc. sg. n. ἠύ (Il.[+]); cf. also nom./acc. pl. n. ἤεα· ἀγαθά (Hsch. η 163 Latte); and cf. also εὐ-, ἠυ- (= Myc. e-u- in personal names such as e-u-me-de ‘Εὐμήδης’) ‘good, well’ in compounds such as εὐ-κλεής adj. ‘of good report, famous’, ἠύ-κομος adj. ‘with lovely hair’Cf., e.g., GEW I:594f., NIL:239, 240 n. 3, 241 n. 6, 241 n. 7, 241f. n. 13, 242 n. 14, EDG:484f. 

OIr. su-, so-, MW hy-, etc. ‘good, well’ in compounds such as OIr. so-nairt ‘strong’, MW hy-bryd ‘beautiful, well-shaped’Cf., e.g., NIL:239, 241 n. 8–9, EDPC:358. 

?Goth. swikns ‘pure, innocent’, ON sykn ‘innocent’< PGerm. *su̯ikna- ‘innocent, pure’, which could go back to pre-PGerm. *h₁su-Hig̑-no- “gut verehrt, kultisch rein”, according to Heidermanns 1993a:582 with lit., who considers a formation to the root PIE *Hi̯ag̑- ‘to worship’ (LIV²:224f.) to be possible. 

?Lith. sū́drus ‘thick, dense, solid, tight, (dial.) lush, fertile’< *h₁su-dru- (cf. Skt. su-drú- ‘made of strong wood’), i.e. “*h₁su- ‘good, well’ […] + dru- ‘wood’”, according to EDBIL:434 with further lit. Cf. also NIL:239, 241 n. 11. 

?OCS sŭ-dravŭ ‘healthy’, Russ. zdoróvyj, zdoróv m., zdorová f., zdorovó n. ‘id.’, etc.Cf., e.g., EDSIL:478f., NIL:239, 241 n. 12.