*páHu̯r̥ / *pHu̯én-, coll. *páHu̯ōr
‘fire’
- Language
- Proto-Anatolian
- Class
- Base
- Grammar
- subst, neuter
- ID
- 292
1. Individual Anatolian Languages
1.1 Cuneiform Luwian
pāḫur-
‘fire’1.1.1Transmission
Attested in the Luwian passages of a Hittite ritual with mythical scenes.
[Zs.S.]
1.1.2Forms
| nom.-acc. sg. | ˹pa-a-ḫu-u-ur˺ | KUB 35.107 rev. iii 4’ | (CTH 764, MS) |
| pa-a-˹ḫu˺-[u-ur | KUB 35.107 rev. iii 24’ | (CTH 764, MS) |
Attestations following CLL:163, checked against the photographs of the Konkordanz. The restoration in KUB 35.107 rev. iii 24’ is based on the parallelism with KUB 35.107 rev. iii 4’ (Starke 1985b:239).
The form is generally assumed to be singular (Laroche 1959a:77; Starke 1990a:571; CLL:163; EDHIL:613; Jasanoff 2010a:172), but then the lack of the expected particle =ša/za is remarkable. Jasanoff 2010a:172 suggests that this particle was assigned on lexical basis in case of r- and r/n-stems (pāḫur- is an r-stem as indicated by its derivative, pawari(ya)- ‘to light a fire’ (see s.v.), cf. Oettinger 1976a:103 [unless one wants to operate with the /n/ → [ɾ] rhotacism, cf. Melchert 2003b:180], contra Rosenkranz 1952a:27, who assumes an r/n-stem, cf. Starke 1990a:570f.).
There is no evidence for the methodologically dubious assumption of Eichner 1988a:146 that this form reflects a performance-based deformation in magical context instead of *păḫḫŭr or *păḫḫuwar.
[Zs.S.]
1.1.3Graphic Features
Starke 1990a:571 calls attention to the unusual spelling of Hitt. pa-aḫ-ḫu-u-ur in the parallel text KBo 9.127 + KUB 36.41 i 6’ and 29’, which he attributes to Luwian influence (similarly, but implicitly HED Pa:24).
[Zs.S.]
1.1.4Meaning ‘fire’
There is a general agreement regarding the meaning (Meriggi 1957a:218; Laroche 1959a:77, 160; CLL:163; Tischler 2008a:130, Tischler 2016a:246; Ünal 2016a:966). However, EDHIL:613 adds a question mark, pointing out that the word is attested in a broken context, thus its meaning cannot be independently determined (and also pointing to a historical phonological problem). Nevertheless, Starke 1990a:571 claims that the meaning is assured due to the word pa-aḫ-ḫu-u-ur ‘fire’ of the parallel Hittite version (KBo 9.127 i 6’ and 29’, implicitly already Laroche 1959a:160), although he also assigns a question mark. However, the parallelism of the two versions was not entirely assured: though the words indeed appear at seemingly corresponding parts, and the topic of the versions was indeed similar (a report about gods invited to a feast and about an uninvited one and his following wrath), they were clearly not translations of each other, since the details of the story were different. Moreover, unlike the Hittite version, the Luwian one referred to pāḫūr twice and it was not clear, if this is simply due to the loss of the relevant Hittite section or yet another dividing feature of the two versions.
However, the discovery of the join KBo 43.223 allows a more precise reconstruction of the Hittite text (see Groddek 2007a:317–326), which in turn allows the precise identification of the Luwian parts with the Hittite parts: the Hittite text shows the typical Hittite narrative technique in mythical texts that first describes the event in a detailed way, which will be reported in the subsequent part by one of the protagonists (on this narrative technique see Siegelová 1971a:85; del Monte 1979a; Groddek 1999a:45, Groddek 2000-2001a:23f., Groddek 2007a:318; Schuster 2002a:372; Giorgieri 2004a:414 with n. 23). The Hittite text now shows the sequence of the invitation of the gods, the wrath of the uninvited god and the closing scene of the burning fire, followed by a report repeating these events. In other words, the two Luwian scenes with pāḫūr now have the Hittite equivalent and it is therefore assured that the two texts represent two variations of the same mythologeme.
[Zs.S.]
1.1.5Derivatives
2. Reconstruction
2.1Reconstruction ‘fire’
The paradigm of Hittite paḫḫur/paḫḫu(we)n- ‘fire’ preserves the inflection of a proterodynamic paradigm (Schindler 1975a:10, Rieken 1999a:331f.). The suffix of the oblique cases is reduced in Hittite (-wen- > -un) in analogy to -ur-, but the frequent preservation of the full grade in the suffix -wen- can be explained as an influence from the lexeme wātar/witen- ‘water’ (Rieken 1999a:333, differently Starke 1990a:531). The Hittite form paḫḫuwar, which is attested once (KUB 7.60 ii 11 (NH)), and the derivational base of the Luwian denominative pa-wa-ar-it-ta have been interpreted as collective forms by Oettinger 1986a:14, differently Rieken loc. cit., although here withdrawn.
The Luwian lexeme for ‘fire’ pāḫūr is spelled with a single -ḫ-;, which calls for a different phonological development. Melchert 1994a:257, 264 tentatively interprets this instance as a late lenition. He suggests that the accented vowel of the first syllable was secondarily lengthened under the accentuation and then caused the lenition of the following laryngeal by Eichner’s law. However, a word like a-aḫ-ḫu-wa-a-aḫ-ḫu-wa-a-mi-in (KUB 35.145 iii 10) would speak against this hypothesis. Besides, there is a more plausible scenario. By starting from the pre-form *pHu̯én- of the weak cases, one can assume an analogical spread of the zero-grade of the root to the strong cases. Afterwards, an anaptyptic vowel in the initial cluster occurred. The simplex notation of fortis initial cluster in Hittite and Cuneiform Luwian is in order; cf. Watkins 1982b:455, Melchert 1994a:32f., Rieken 1999a:206. The plene spelling in the initial syllable, like in pāḫūr, is not an argument against the original cluster *pH-, since it may be used to render the anaptyptic vowel. For examples in Hittite see Rieken 1999a:206.
For the attestations of the Hittite form pāḫḫur with plene writing in the first syllable and its explanation, cf. Rieken 1999a:332 n. 1623, Kimball 1999a:149, 395, Melchert 1994a:147. No solution is offered in EDHIL:613 for the Luwian lexeme. For an overview of outdated hypotheses see HEG P:366-370.
[D.S.] [E.R.]
3. Proto-Indo-European
3.1Semantic Reconstruction
PIE *péh2u̯r̥/*ph2u̯én- n. means ‘fire’ (EIEC:202, NIL:540, Watkins 2011a:63). Beside the neuter noun, there was a masculine word for ‘fire’ that is preserved in Ved. agní-, Lat. ignis, Lith. ugnìs, Latv. ugùns, and OCS ognĭ. Its exact reconstruction is still under debate: PIE *hxn̥gu̯ni- (EIEC:202), *h1óngni-/*h1n̥gnéi̯- (ALEW II:1146), *egni- : *ogni- (IEW:293), *h1o/egu̯-ni- (implicitly Balles 1999a:11), *og-ni-/*eg-ni- (Schindler 1975a:5), *h1ógni-/*h1égni- (EWAhd III:334). As for the semantics of masculine *h1o(n)g(u̯)ni-/*h1é(n)g(u̯)ni- vel sim. and neuter *péh2u̯r̥/*ph2u̯én-, “both mean ‘fire’ but have been seen to have originally constituted an opposition. The first word, regularly masculine in gender, is seen to represent the concept of ‘fire’ as something active, hence its deification in OInd Agní- ‘fire (god)’ who is one of the most invoked deities of the R̥gveda. In contrast, *péh2ur is neuter in gender and traditionally regarded as the inactive conception of ‘fire’ as purely a material substance” (EIEC:202; cf. also EWAhd III:334 with further lit.). A similar semantic opposition can be observed with the words for ‘water’, with PIE *u̯od-r̥/*u̯ed-(e)n- n. (NIL:706–715) denoting ‘water’ as a material substance and PIE *h2óp-/*h2ep- f. ‘(running) water, river’ (NIL:311–317) being “‘living water’, i.e. ‘water on the move’”, according to Mallory & Adams 2006:126. Cf. also the language-and-culture note in Watkins 2011a:5 s.v. ap-2.
3.2Morphological Reconstruction
Since Schindler 1975a:7, 10, the word has usually been reconstructed as a proterokinetic r/n-stem, i.e. PIE *péh2u̯r̥/*ph2u̯én-.According to EWAhd III:333, the word might alternatively be reconstructed as an ó/é-acrostatic stem *póh2u̯r̥/*péh2u̯n̥- n.; *póh2u̯r̥ is also propsoed by Rasmussen 1989:77 n. 10, but see the discussion in NIL:542 n. 2 with lit.The exact morpheme boundaries (*péh2-u̯r̥ or *péh2u̯-r̥?) and the root etymology (*peh2- or *peh2u̯-?) are not clear (cf. NIL:541 n. 1, EWAhd III:333f.). Furthermore, an amphikinetic collective, PIE *péh2u̯ōr/*ph2un-´, can be reconstructed on the basis of, e.g., Goth. fon n. ‘fire’ and TochB pūwar ‘fire’ (cf. Schindler 1975a:10, NIL:541 and for Tocharian Malzahn 2014b:198).
[T.S.]
4. Transmission
Myc. pu-ra-u-to-ro /pū̆r-au̯strō/ nom. du. ‘pair of fire-tongs’, pu-ka-wo /pū̆r-kau̯oi̯/ nom. pl. m. ‘fire-kindlers’; Gk. πῦρ, gen. πῠρός n. ‘fire’ (Il.+), πυρ-καϊή f. ‘funeral pyre, scene of the fire’ (Il.+), ἄ-πῠρος adj. ‘without fire’ (Il.+), πῠρᾱ́, πῠρή f. ‘funeral-pyre’ (Il.+), πῠρετός m. ‘burning heat, fiery heat’ (Il.+), etc.Cf., e.g., DMic II:168 s.v. pu-ka-wo, 172 s.v. pu-ra-u-to-ro, GEW II:627–629, NIL:541–545 with notes, EDG:1260f.
Arm. howr ‘fire’, hnocʻ ‘oven, fireplace’Cf., e.g., Olsen 1999a:49, 533f., NIL:541, 544 nn. 14, 20, EDAIL:426f.
Goth. fon n. ‘fire’ (gen. sg. funins, dat. sg. funin), ON fýrr, fúrr m. ‘id.’, funi m. ‘id.’, OE fȳr n. ‘id.’, OFris. fiōr, fiur n. ‘id.’, OS fiur n. ‘id.’, OHG fiur, fuir n. ‘id.’, ModHG Feuer n. ‘id.’, etc.Cf., e.g., Casaretto 2004a:579f., Ringe 2006a:122, 137, 277, EWAhd III:330–334, NIL:541–545 with notes, EDPG:151.
OPr. panno n. ‘fire’, panu-staclan n. ‘fire steel’Cf., e.g., NIL:541, 545 n. 23, Larsson 2010a:49.
Cz. pýř m., pýří n. ‘glowing ashes’, Cz. pýřeti ‘to blaze, glow’, Serb. upíriti ‘to kindle, stoke’, etc.Cf., e.g., Trautmann 1923a:232, IEW:828, NIL:541, 544 n. 17.
TochA por alt. ‘fire’, TochB pūwar alt. ‘fire, beacon-fire; digestion’Cf., e.g., NIL:541, 543 n. 5, Adams 2013a:421f., Hackstein 2017a:1314, 1319.