*d(h)Eu-nt-mEn-
‘(a type of architectural enclosure of a spring)’
- Language
- Proto-Anatolian
- Class
- Base
- Grammar
- subst, neuter
- ID
- 1894
1. Individual Anatolian Languages
1.1 Cuneiform Luwian
duntum(a)n-
‘(a type of architectural enclosure of a spring)’1.1.1Transmission
A hapax, the word appears once in the incipit of an Ištanuwa festival song.
[E.R.]
1.1.2Forms
| dat./loc. sg. | du-un-tum-ni | KUB 25.38, 5’ | (CTH 773, MS?) |
1.1.3Graphic Features
Checked against the photographs in the Mainzer Fotoarchiv.
[E.R.]
1.1.4Meaning
CLuw. duntuman- is left untranslated in DLL:99, CLL:234, HEG T:440f., HHw:s.v., and Ünal 2007a:736. The semantic assignment made here is based on the meaning of Hitt. (É)dundumar, duntumnaš elucidated below. Ünal 2007a:736 disfavors the identity of the two words. A first indicator is the determinative É in the phrase LÚMEŠ Édu-un-tum-na-aš at KBo 42.14 rev. 6’, which points to a building, presumably one that is large enough to need attendants (LÚMEŠ). Secondly, at KBo 23.27 i 14’, ii 7, ii 8, ii 9 it is said that each dundumar belongs to a town, at i 11’ and ii 8 it is called by a name broken away, and at ii 9f. (probably also at i 11’ff.) it receives an offering together with specific rivers. At i 18 the action anda waḫnu- is performed on “that spring,” whereas the water at the dundumar is already anda waḫnu-ed (at ii 10’f.) and they do not anda waḫnu- other rivers (at ii 16’f.). Ünal 1998a:87f. n. 30 claims that anda waḫnu- denotes the action of encircling water wells in these cases. This would agree well with the assumption that the (É)dundumar is a building of considerable size, and its occurrence in connection with rivers and springs and the descriptions of the sites point to an enclosure of springs; cf. Steitler 2019a:11-13 (with references) on monumental architecture in spring sanctuaries such as Eflatunpınar or the SÜDBURG.
[E.R.]
1.1.5Stem
Melchert analyzes the stem as reduplicated from *dum-, showing an assimilation *dum-duman- > dun-duman- (CLL:234, applying Laroche’s explanation of dundumama/i- to the present lemma). In contrast, van den Hout 2006a:250 posits an i-mutated dunduma/i- in Hittite transmission. However, given the new semantic interpretation as an architectural structure and the existence of the Hittite -umar/-um(a)n- stem, it is more likely that the locatival suffix -umar/-um(a)n- (for which see Melchert 1983a, e.g., in ḫalhaltumar ‘corner’) forms a substantive with locational semantics. Here, this would be the site of a spring. An attractive side effect of this analysis is a base with a normal shape, i.e. *dunt-. It suggests a comparison with the synonym šarunt(i)- ‘well’, possibly a participle of the root *sru- (see under Lemma). An analogous interpretation yields a root *du- (see under Reconstruction) plus participial -nt- plus locatival -umar/-um(a)n- in Hittite vs. -um(a)n- without heteroclisy in Luwian.
[E.R.]
1.1.6Derivatives
1.2 Hittite
dundumar, duntum(a)naš
‘(a type of architectural enclosure of a spring)’1.2.1Literature / Comments
HEG T:441Hitt. (É)dundumar, duntum(a)naš is attested as nom./acc. sg. du-un-du-ma-ar KBo 23.27 i 14’, ii 7, [du-u]n-⸢du⸣-mar ii 8, du-un-du-mar ii 9, Hitt. gen. sg. du-un-du-ma-〈〈aš〉〉-na-aš KBo 23.27 i 11’, Édu-un-tum-na-aš KBo 42.14 rev. 6’.
[E.R.]
2. Reconstruction
2.1Reconstruction ‘(a type of architectural enclosure of a spring)’
The root du- can be connected to PIE *dheu̯- ‘to run’, from which a participial formation was built, i.e. *dheu̯-nt- (> Luw. *dunt-) followed by the locative suffix -umar/n- (originating in the suffix *-men-, for which see Melchert 1983a), i.e. PAnat. *d(h)Eu-nt-mEn-. One should also compare Luw. šarunt(i)-.
[E.R.]
3. Proto-Indo-European
3.1Semantic Reconstruction
The noun Luw. duntum(a)n- is interpreted above as ‘place of the running (water)’, with the locatival suffix Luw. ‑um(a)n- added to a participial formation that is supposed to mean ‘spring’ < *‘running, flowing water’.Cf. as a parallel for the latter the Lat. ti-stem fōns, ‑tis ‘spring, fountain’ < abstract *‘running (of water)’ : PIE *dhenh2- ‘to set in motion, slip away’ (LIV²:144f.), cf. EDLIL:230f. For more sources of words for ‘spring, well’, cf. Buck 1949a:44–46.The derivational base of Luw. duntum(a)n- may thus be reconstructed as PAnat. *dheu̯‑n̥t- ‘running’, derived from the root PIE *dheu̯- ‘to run, hurry’ (LIV²:147f.). This connection is very plausible considering the semantics of Ved. DHĀV ‘to run, flow’, Ved. dhautí- ‘river’, the possible Phrygian gloss δύαν ‘well, spring’ as well as the existence of river names related to this root.Cf. the Greek river name Θόας, better known as Achelous, cf. Strab.10.2.1).
3.1.1Root etymology
The meaning of the root underlying the Luwian formation is reconstructed in LIV²:147 as “laufen, eilen”. This semantic reconstruction may be specified by adding the information that the root is often used to denote the running (thus, flowing) of water in various languages, cf. Ved. DHĀV ‘to run, flow’, etc. However, since this connection is trivial (cf., e.g., PIE *teku̯- ‘to run, flow’),The CLICS³ graph FLOW shows 37 colexifications with RUN and 15 colexifications each with GO and WALK.the root is reconstructed here merely as PIE *dheu̯- ‘to run, hurry’ (as per LIV²:147).
3.2Morphological Reconstruction
As shown above, Luw. duntum(a)n- may be derived from a participle of the verbal root *dheu̯- ‘to run, hurry’. The root vowel is reconstructed in the full-grade *dhEu̯‑n̥t- ‘running, flowing’, which is made plausible by the underlying “Narten present”, PIE *dhḗu̯‑/*dhéu̯-, which is continued in Ved. act. dhā́vati vs. midd. dhávate ‘runs, flows’. Although the attested participles in the non-Anatolian languages (cf., e.g., Ved. dhā́vat‑, Gk. θέοντ‑) are, at least primarily, not connected with water, the existence of the participle PIE *dhéu̯‑n̥t‑ ‘running’ is beyond doubt.
3.2.1Root etymology
Based on the Indo-Iranian and Greek continuants, the root in questions can unambiguously be reconstructed as PIE *dheu̯- ‘to run, hurry’ (LIV²:147).
[A.O.]
4. Transmission
PIE ‘Narten’-present *dhḗu̯‑/*dhéu̯- ‘to run’, cf. Ved. dhā́vati ‘runs, flows’, midd. dhávate ‘id.’, MPers.(+) daw- ‘to run’ (EWAia I:789f., Grassmann 1996a:679f.), Gk. θέω ‘to run’ (GEW I:668f., DELG:433, EDG:544f., LSJ:796,), (?)Goth. *diwan ‘to die’ (EDPG:90, Orel 2003a:72, GED:93, Seebold 1970a:147–149)
LIV²:147f., IEW:259f., Jasanoff 2003a:75
Ved. dhautí- ‘river’ (hapax, RV.2.13.5)
EWAia I:789f., Grassmann 1996a:696
Gk. θοός ‘fast’, Myc. personal name Pe-ri-to-wo /Perithowos/ nom. sg. (hapax, KN Vc 195), Hom. personal name Πειριθόου gen. sg. (Il.12.129+),The first syllable of the first compound member *Περι- is metrically lengthened.Gk. river name Θόας (in antiquity and today Achelous, cf. Strab.10.2.1)
GEW I:668f., GEW III:105, DELG:433, EDG:544f., LSJ:803, LSJSuppl:192, DMic II:112
Phryg.? ἄδ-δεε /ad-deu̯e/? impv. 2sg. ‘(Gk.) ἐπείγου, press, urge, hurry’ (Hsch. α 1024 Latte), δύαν acc. sg. ‘(Gk.) κρήνην, well, spring’ (Hsch. δ 2474 Latte)
Obrador-Cursach 2020a:423
Lat. favēre ‘to be favorably inclined’ (Naev.+), (?)fove (CIL I².573, 250–200 BCE)Vine 2006a:225f. assigns the Latin verb not only to a different root (PIE *gu̯heu̯‑, cf. OCS gověti ‘to worship’, etc.), but also doubts the reliability of the form fove.
WH I:464–466, EM:220f., EDLIL:206, OLD²:747
ON deyja, OS dōjan, OHG touwen ‘to die’
EDPG:90, Orel 2003a:72, Seebold 1970a:147–149, de Vries 1977a:76
ON dǫgg, OE dēaw, OHG tou ‘dew’
EDPG:91, de Vries 1977a:92f.