*u̯lṓ-/*u̯l̥-´
‘to die’
- Language
- Proto-Luwic
- Grammar
- verb
- ID
- 1046
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/wala-/, wa/i-la-
‘to die’1.1.1Transmission
There are ten attestations of this lexeme in the HLuw. corpus, of which six are found in inscriptions in Tabal, Karkamiš, and Tell Ahmar. The inscriptions date between the late 10th and the mid-8th century BCE. The remaining four attestations belong to the letters found in Assur, composed in the late 8th century BCE.
[AH.B.]
1.1.2Forms
| 1sg. pret. ind. act. |
/walahha/ |
(CRUX)wa/i-la-ha |
|
|
(“MORI”)wa/i+ra/i-ha-a |
|||
| /walahhan/ | MORI-ha-na | ASSUR letter a, §8 (Assur) | |
| 3sg. pret. ind. act. | /walatta/ | (“MORI”)wa/i-la-tá | |
|
(MORI)wa/i-la-tá-´? |
|||
| MORI-ta | TELL AHMAR 5, §8 (Tell Ahmar) | ||
| 3pl. pret. ind. act. |
/walaunta/ |
wa/i-la-u-ta | |
| /walanta/ | MORI-ta | ||
| part. acc. sg. c. | /walammin/ | wa/i-la-mi-na-´ | ASSUR letter f+g, §6 (Assur) |
1.1.3Graphic Features
1.1.3.1Phonological Interpretation
With most attestations written syllabically, the phonological interpretation of this lexeme has been reasonably clear from the start. The interpretation is further aided by the cognate link of the verb with CLuw. wa-la-an-/ú-la-an- ‘dead’ (cf. Hawkins 1980b:113). While HLuw. spelling conventions do not permit any conclusions as to the quality of the /l/, the cited evidence from CLuw. suggests it to be non-geminate. Based on historical considerations, it is possible that the vowel of the first sign is mute, i.e. /wla-/ (see under Reconstruction).
Most attestations of the verb are preceded or partly covered by the logogram MORI, whose reading was provided by Hawkins 1980b: 114 when he furnished the combination of the logograms VIR2 (*386) and MINUS (*381) with a new interpretation based on the sense of the verb. The other logogram used with the verb, CRUX (*309), is not restricted to this lemma and can also be found with the adverb /wala/ ‘fatally’ (but see also the section on meaning below).
The stem of the verb displays rhotacism, i.e. variance between /l/ and /r/, cf. (“MORI”)wa/i+ra/i-ha-a as attested in KULULU 2 in contrast to all other syllabically spelt forms showing -la- (cf. Rieken & Yakubovich 2010a:217). Another type of variation is present in the form attested in ASSUR letter a, §8, which ends in /-hhan/ rather than the more usual /-hha/; Carruba 1984a:65f. has explained these forms as nasalisation instead of a different verbal ending.Morpurgo Davies 1980b:97-101 had suggested these to represent a 1pl. pret. ind. act.
[AH.B.]
1.1.4Meaning
The lexeme has been examined in full by Hawkins 1980b, who also draws the cognate link to CLuw. ulant(i)-‘dead’ and further words from the same family (Hawkins 1980b:113). Even without adducing the CLuw. evidence, the meaning of HLuw. /wala-/ becomes clear from nearly all contexts it is used in, such that when the verb is combined with the pre-verb ARHA /ahha/ ‘back, away’, the meaning is very clearly ‘to die’, with /ahha/ usually adding a sense of ‘utterly, fully, to completion’.
However, the relevant clause in KARKAMIŠ A23 does not include that preverb, thereby suggesting a difference in meaning, which would lack the sense of completion that is provided by /ahha/. Also, the spelling of this form involves the logogram CRUX (*309) instead of the more usual MORI (*386.*381). Hawkins 1981a (and Hawkins 2000a:120) suggests to interpret it as ‘to suffer, be ill’, but in the absence of further evidence supporting such a meaning, this proposal may be reaching too far beyond the basic meaning of the word. Instead, a more straightforward interpretation such as ‘to be about to die’ vel sim. seems appropriate.
[AH.B.] [I.Y.]
1.1.5Stem
While the stem of the verb is clear, it cannot be said which of the two conjugation patterns it belongs to, because the relevant forms to decide that, i.e. 2sg. and 3sg. pres. act., have so far gone unattested.
[AH.B.]
1.1.6Derivatives
1.2 Cuneiform Luwian
*wala-
‘to die’1.2.1Derivatives
1.3 Lycian A
la-(ti)
‘to be dead’1.3.1Transmission
This lexeme is found in grave inscriptions of Eastern Lycia.
[D.S.]
1.3.2Forms
| 3sg. pres. ind. act. | lati | TL 88.2 (Myra) |
| ḷatị | TL 112.2 (Limyra) | |
| ḷa[ti] | TL 75.4 (Tyberissos) |
The attestations of this verb follows the ones of Melchert (DLL:34).
[D.S.]
1.3.3Meaning
On contextual ground, Thomsen 1899a:21 translates the verb la-(ti) as ‘to be dead’, i.e. se ẽke lati ddaqasa m(e)=ene ñtepi tãti ñtipa tezi ‘As soon as Ddaqasa is dead, they will lay him down into the sarcophagus.’ (TL 88.2). This semantic assignment is accepted by Carruba 1968a:14, who adduces as etymological evidence Luw. ulant(i)- ‘dead’ (thus also Hajnal 1995a:110 n. 100, DLL:34, Christiansen 2020a:242f.).
[D.S.]
1.3.4Derivatives
2. Reconstruction
2.1Reconstruction ‘to die’
The conjugation of the Hieroglyphic Luwian verb /wala-/ (or /wla-/) ‘to die’ is not secured, although a verb of the ḫi-conjugation is very likely (see Sasseville 2021a:366f.). For Lycian la-(ti) ‘to be dead’, one is either dealing with a factitive or a root verb, whose root ends with -VH, i.e. *u̯lÉH-; cf. stta-(ti) ‘to stand’, ha-(ti) and ha-(di) ‘to release’. Judging by the overall data, the latter seems more likely: In Lycian, the labial *u̯ was lost before the -l- (for the sound law see Carruba 1968a:14 n. 2, Melchert 1994a290, Schürr 2016d:163), which supports a reconstruction of the shape *u̯lÉH-. Taking the Luwian and Lycian verbs as reflecting originally the same formation, we reconstruct a root verb of the ḫi-conjugation, i.e. PLuw. *u̯lṓ-/*u̯l̥-´ < PAnat. *u̯lóH-/u̯l̥H-´. The laryngeal is regularly dropped in root verbs of this shape due to the following alveolar in the ending. Considering that we are dealing with an o-grade in the strong stem, one must assume that the Lycian root vowel -a- comes from the weak grade, i.e. la- < *u̯lá- ← *u̯lé-/*u̯la-´ < *u̯lṓ-/*u̯l̥-´; for the generalization of the weak grade in Lycian root verbs, see under Lyc.A mar-(ti).
A Proto-Luwic *u̯ means that it can come either from a Proto-Anatolian *u̯- or *gu̯-. Hawkins 1980b:119 suggests a connection to the Proto-Indo-European root *gu̯el- (followed by Melchert 1994a:238). Instead, the root *u̯elh3- is suggested here from the Proto-Indo-European perspective (see below). If this etymology is correct, Schwebe-Ablaut is necessary to account for the outcome of the Luwian and Lycian verb, i.e. *u̯lóH-/u̯l̥H-´← *u̯ólH-/u̯l̥H-´.
[D.S.]
3. Proto-Indo-European
3.1Semantic Reconstruction
3.2Morphological Reconstruction
As far as the root etymology is concerned, Proto-Luwic word-initial *u̯- (cf. CLuw. walant-/ulant- ‘dead’, HLuw. wa-la/i- ‘to die’, Lyc. la-(ti) ‘id.’ [< *u̯lÉH-, cf. above the Proto-Luwic reconstruction]) is ambiguous and can go back to PIE *u̯- or *gu̯-. The underlying PIE root is then either *u̯elh3- ‘to die’ (as per LIV2+ s.v., for example) or *gu̯elH- ‘to torment, sting, bite, ache’ vel sim. (LIV²:207; cf. also LIV2+ s.v. *u̯elh3- n. 1). Semantically, a connection with PIE *u̯elh3- ‘to die’ seems more straightforward. Hitt. u̯alḫ-zi ‘to hit, strike’, on the other hand, belongs to a different root PIE *u̯elh2- ‘to tear, pull’, which is also the root of Lat. vellere ‘to tear, pull (out)’ and has to be separated from *u̯elh3- ‘to die’, according to LIV2+ s.v. *u̯elh2- with n. 1 and s.v. *u̯elh3- with n. 1.
[T.S.]