*u̯āi!
‘woe’
- Language
- Proto-Anatolian
- Grammar
- subst
- ID
- 876
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/wai-/, wa/i-i-
‘woe’1.1.1Transmission
This lexeme is attested only once in an inscription from Aleppo with a tentative dating to the 8th century BCE.
[AH.B.]
1.1.2Forms
| dat./loc. sg.(?) | /wai/ | wa/i-i | BABYLON 2, §2 (Aleppo) |
1.1.3Graphic Features
1.1.3.1Phonological Interpretation
With only one attestation, the phonological interpretation of this lexeme is tentative; the spelling would also permit a reading as /wi/. However, the cognate link to CLuw. wāi- ‘woe’ (CLL:250 and Hawkins 2000a:395f.) and Hittite wāi- ‘woe’ supports the proposed interpretation as /wai/.
[AH.B.]
1.1.4Meaning
The lexeme occurs in a brief clause in BABYLON 2, §2: wa/i-tu-u wa/i-i kwa/i-za sa-há, i.e. /wa=du wai kwanza asha/, tentatively translated by Hawkins 2000a:395f. as ‘Because I was woeful? to him (scil. Tarhunza, the storm-god)’; the verb has since been identified as a 1sg. pret. act. of /sa-/ ´to release´. He links it to Hittite wāi- ‘woe’, cf. also CLL:250 for CLuw. wāi- ‘woe’, and it appears that the attested form is a dat./loc. sg., i.e. ‘I let go in woe to him’. Alternatively, it is possible that the word lacks an ending and constitutes an interjection that is used adverbially in the context.
[AH.B.]
1.1.5Stem
Like its CLuw. cognate, the noun is likely derived by conversion from the interjection used to lament woe, cf. CLL:250.
[AH.B.]
1.2 Cuneiform Luwian
(u)wāi-
‘woe’1.2.1Transmission
The word occurs twice in incantations within Luwian rituals.
[E.R.]
1.2.2Forms
| nom. sg. | ú-wa-a-i-iš | KUB 35.87, 7’ | (CTH 765, MS) |
| acc. sg. | [w]a-a-i-in | KUB 35.109 iii 13 | (CTH 765, MS) |
1.2.3Graphic Features
Checked against the photographs in the Konkordanz.
[E.R.]
1.2.4Meaning
Judging by the context in KUB 35.109, where (u)wāi- is listed with ašḫarša ‘blood-deed’ and lappiya- ‘fever’ and is also to be treated magically, the word denotes a negative action or state. This is confirmed by KUB 35.87, 7’, where it is said that (u)wāi- should not negatively affect the ritual client. Obviously, (u)wāi- is formally and semantically equivalent to Hitt. (u)wāi- ‘woe’ (CLL:250, contra Starke 1990a:184-186 with an implausible semantic interpretation as ‘slander’).
[E.R.]
1.2.5Stem
Pace Starke 1990a:184-186 and Weitenberg 1984a:389, CLuw. (u)wāi- is clearly a root noun, based on the exclamation wai. The alveolar of the alleged -it- stem †(u)wait- posited by Starke is part of the verb tiye/a-(mi) ‘to step’ in the phrase (u)wāi tiške/a- ‘occur as woe’ (since Götze 1930b:6).
[E.R.]
1.2.6Derivatives
uwai-1.3 Hittite
(u)wāi-
‘woe’1.3.1Literature / Comments
HW 2nd ed.:239, HEG W-Z:204-207, Tischler 2008a:192, EDHIL:939, Ünal 2007a:772f.
2. Reconstruction
2.1Reconstruction ‘woe’
The presence of the word (u)wāi- c. ‘woe’ in both Hittite and Luwian allows us to reconstruct a Proto-Anatolian pre-form. The onomatopoetic character of the word, which can be associated with a cry of pain, leads us to a reconstruction with the vowel -a-, i.e. *u̯āi!. Kloekhorst (EDHIL:939) agrees with the onomatopoetic origin of the word, but suggests that (u)wāi- is nevertheless a derivative of ú-i ‘whee!’ attested in Hittite (KUB 55.38 ii 19). Although it is morphologically possible that (u)wāi- is a morphological creation after (Hittite) diphthong stems in -āi- c., it is not necessary either to posit a derivation. Both ú-i and ú-wa-a-i can be independent onomatopoetic creations, in which the latter became fully lexicalized as a declining noun with the meaning ‘woe’. For the Hittite phrase a-i-in ú-wa-a-i-in see HEG U-Z:170f. and EDHIL:939.
[D.S.]
3. Proto-Indo-European
3.1Semantic Reconstruction
Besides PAnat. *u̯āi̯- c. ‘woe’, a noun is attested in distant parts of the IE world (Av., Celt., Germ., Alb.). Thus, the noun may be reconstructed for the proto-language as PIE *u̯ai̯- ‘woe’.
3.1.1“Root” etymology
Various non-Anatolian languages (Av., Gk., Lat., Celt., Germ., Balt., Alb.) attest an interjection PIE *u̯ai̯!, which expresses pain, sorrow, grief, etc., thus ‘alas!’ vel sim.
3.2Morphological Reconstruction
The Anatolian languages continue a noun that on the basis of the additional evidence from Avestan, Celtic, Germanic, and Albanian is reconstructable as PIE *u̯ai̯- ‘woe’.PIE *u̯ai̯- ‘woe’ is not labeled as a “root noun” here because the underlying morpheme is not a verbal or (pro)nominal root but an interjection.
3.2.1“Root” etymology
The phonologically (and semantically) obvious source of the noun PIE *u̯ai- ‘woe’ is the interjection PIE *u̯ai̯ ‘alas!, woe!’. Although both directions of conversion are attested, viz. interjection > nounCf., e.g., Germ. Bavarian interjection/sbst. dig ‘chick! (call to summon or attract the attention of chickens); chicken’, diminutive pl. digai ‘(young) chicken’, cf. BWB III:1711.and noun > interjection,Cf. ModE sbst./interjection man ‘man (sbst.); (interjection expressing surprise, delight, disbelief, amazement, etc.)’, cf. OED³ online s.v. 16c.PIE *u̯ai̯ was originally used as an interjection.
The sound changes attested in the Celtic continuants of the interjection and of the collocation (PIE *u̯ai̯- + *dheh1- ‘to cause woe’) as well as in the Armenian and Celtic cognates of the lo-stem (PIE *u̯ai‑lo- ‘wolf’) prove that *u̯ai̯ has to be reconstructed already for PIE. Without this evidence, an independent creation could not be excluded, cf. EIEC:313. Besides the phonologically ambiguous cases (with question marks sub Transmission above), Gk. οὐαί ‘ah!, woe!’ and Arm. váy ‘ah!, alas!, woe!’ could be taken as examples of an independent (re)creation, or they may just have escaped sound change, cf. Olsen 1999a:787.Cf. also ALEW II:1172, where any genetic relationship of the interjections collected above is dismissed: “Die nur teilweise Befolgung der regulären Lautkorrespondenzen spricht dabei eher für Elementarverwandtschaft als für etymologische Verwandtschaft”; cf. similarly WH II:724, EDLIL:650, EDPC:405.
To sum up, it is perfectly possible to reconstruct the interjection PIE *u̯ai̯ ‘alas!, woe!’ as the basis of the noun PIE *u̯ai̯- ‘woe’, the substantivized adjective Core IE *u̯ai‑lo- ‘the howling one’, and the collocation PIE *u̯ai̯- + *dheh1- ‘to cause woe’.
3.3Syntactic Reconstruction and Phraseology
On the basis of Anatolian (Hitt. (u)u̯āi + dāi‑/tiya- ‘to harm’), Germanic (OHG (+) wê + tûon ‘id.’), Celtic (OIr.(+) faíd ‘shout, cry’), and Baltic (Latv. vaĩdi ‘sorrow, misery’), a collocation consisting of the nominal object PIE *u̯ai̯- ‘woe’ and the transitive verbal root PIE *dheh1- ‘to put, place’ can be reconstructed: PIE *u̯ai̯- + *dheh1- ‘to inflict pain, harm’, cf. Hackstein 2012e:15f. The non-compositional meaning (not †‘to put woe’) as well as the fact that in some languages (Celt., Latv.) the collocation was univerbated prove the PIE status of this collocation.
Furthermore, it is possible to reconstruct three syntactic constructions which are not attested in Anatolian but in Avestan, Latin, Baltic, and Germanic (cf. Barðdal et al. 2013a:328–337): bare exclamative *u̯ai̯ ‘alas!, woe!’, dative exclamative *u̯ai̯‑dat! ‘woe/misery to dat’ (cf. Lat. vae tibi! ‘Get lost!, Go to hell!’ [frequently in Com.]), and predicative dat‑is‑*u̯ai̯ ‘woe is to dat’ (cf. OE wa bið þæm “they have woe”, cf. Barðdal et al. 2013a:336).
While the bare exclamative construction, viz. the interjectional use of *u̯ai̯ ‘alas!, woe!’, is clearly primary, the predicative construction (dat‑is‑*u̯ai̯) came into being in PIE already. By fronting of *u̯ai̯ and ellipsis of ‘to be’, the latter gave rise to the dative exclamative construction (e.g., Goth. wai izwis “woe {be} to you!”), cf. Barðdal et al. 2013a:337–345. While the dative exclamative construction is only attested in Core IE, the noun PIE *u̯ai̯- ‘woe’, which originated in the predicative construction as wellThe transfer in this very context is not unexpected because such predicative constructions with non-canonical subjects in the dative, on the one hand, and possessive dative constructions, on the other, are more or less the same.and is also part of the above-mentioned collocation PIE *u̯ai̯- + *dheh1- ‘to inflict pain, harm’, is reconstructable for PIE (cf. PAnat. *u̯āi̯- sbst. ‘woe’).Differently, Barðdal et al. 2013a:349f. argue in favor of einzelsprachlich and late (e.g., OHG) origin of the nominal use, at least in the context of the predicative construction. However, they do not take into account the Anatolian material attesting exclusively the noun and its derivatives.
[A.O.]
4. Transmission
OAv. auuōi (hapax, Y.45.3) ‘woe’, Av. vaiiōi ‘alas!; (sbst.) wailing’, YAv. āuuōiia ‘id.’, bāuuōiia ‘alas!’ (hapax, Yt.3.14)The anlauting a- of OAv. auuōi may be due to stylistic reasons, because the verse consists only of words beginning with a‑: aēibiiō aŋhǝ̄uš auuōi aŋhat apǝ̄mǝm “Woe shall be to them / they shall have woe at the end of (their) life” (translation by Barðdal et al. 2013a:328).
On Ved. uvé (hapax, RV.10.86.7), cf. EWAia I:233.
Bartholomae 1904:180, 334, 955, 1359
Gk. αἴ, αἶ, αἰαῖ ‘alas!; wow!’ (Ar.+), αἰάζω ‘to wail, groan’ (E.+)
GEW I:30, DELG:28f., EDG:30, LSJ:34Gk. οὐαί ‘ah!, woe!’ (LXX+) may well be a loanword, cf. GEW II:343 (hebrewism or latinism), DELG:835 (hebrewism). Gk. ὀά, ὀᾶ ‘alas!, woe!’ (A.+) is excluded here because of the reduced phonological shape (/wa/, not /wai/).
On the graphematic realization of /w/ as Gk. 〈ο〉, 〈ου〉 (and also 〈υ〉), cf. Schwyzer 1939a:313.
MIr. fáe ‘woe!, alas!’, MW gwae ‘woe, misery’
EDPC:405, eDIL s.v. 1 fae
Lith. (Žem.) vaĩ, vaí, Latv. vaĩ, vaì ‘alas!, woe!’
LEW I:3, LEW II:1179, Karulis 1992a:467, Trautmann 1923a:338
Alb. vaj ‘alas!’, sbst. vaj m. ‘mourning wail’
Demiraj 1997a:405f., AED:492
PIE sbst. *u̯ai̯‑lo- ‘wolf < *the howling one’, cf. Arm. gail ‘wolf’ (Bible+), personal/river name Gayl, etc. (EDAIL:196–198, Olsen 1999a:34),As in Arm. gayl ‘wolf’, one would expect Arm. †gáy with the change *u̯- > Arm. g- instead of Arm. váy ‘ah!, alas!, woe!’ (Bible+) (EDAIL:591), which is also the source of a noun and a verb in Armenian.Gaul. personal names Vailo, etc., OIr. personal name Fáilán, MIr. fáel m. ‘wolf’,OIr. faílenn ‘sea-gull’ may belong here if it is interpreted as a diminutive (< PCelt. *u̯ai̯‑l‑ino-), cf. EDPCAdd. s.v.?W gwael ‘poor, wretched’ (EDPC:406, eDIL s.v. 2 fáel)Cf. also Av. vaiiōi sbst. ‘wailing’ of a wolf in V.13.8 as well as CLuw. wayam(ma)n- ‘howl, wail’, which occurs with the attribute ulipnaššanza ‘of a wolf’.
EIEC:647
PIE collocation *u̯ai̯- + *dheh1- ‘to inflict pain, harm’, cf. Hitt. (u)u̯āi + dāi‑/tiya- ‘id.’ (HEG U–Z:204f.), OHG (+) wê + tûon ‘id.’ (Otfrid+), the compound Goth. wai-dedja ‘robber’ (DWB s.v. weh E, GED:387f.) and the univerbations (object incorporations) OIr. faíd, W gwaedd f. ‘shout, cry’ (EDPC:406,Matasović in his addenda and corrigenda to EDPC notes s.v. *waydā that it might go back to a “compound *way‑dhh1‑eh2, with the root *dheh1- ‘do, make’ as the second element” but “no trace of such a compound is attested in any other IE language.” However, there is enough evidence for the collocation, which in the case of Celtic became the derivational base of the ā‑stem.de Bernardo Stempel 1999a:295, eDIL s.v. faíd), Latv. pl. vaĩdi ‘sorrow, misery’ (Karulis 1992a:468, Ulmann 1872a:326)
Hackstein 2012e:15f.