*pḗr/pr̥n-´
‘house’
- Language
- Proto-Anatolian
- Grammar
- subst
- ID
- 3889
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/parn(a)-/, DOMUS-na-
‘house’1.1.1Transmission
This lexeme is attested 51 times in inscriptions from Tabal, Tell Ahmar, Cilicia, Malatya, Karkamiš, Assur and Maraş, dating from the 12th to the early 7th century BCE.
[A.P.]
1.1.2Forms
The above table offers a representative selection, largely omitting broken attestations.
[A.P.]
1.1.3Graphic Features
1.1.3.1Phonological Interpretation
Despite just a single fully syllabic rendition, the phonological interpretation of this word can be deemed certain after being cautiously suggested by Gelb 1942a:50f. In HLuw., the word is part of a family with at least two more members deriving from /parn-/, and it is further connected to a number of cognates in the other Anatolian languages, e.g. CLuw. parn(a)- n. ‘house’ (cf. CLL:170) and Lycian prñnawa- c. ‘grave-house’ besides the well-attested Hittite family around pēr/parn- n. ‘house’ (cf. HEG P:485-488, 563-567).
The spelling of the word consistently involves use of the logogram DOMUS (sign *247).
[AH.B.]
1.1.4Meaning
The meaning of this word has long been clear, since the meaning of the logogram DOMUS became clear early on, see Hrozný 1933a:58 passim approximating the sense and then the certain assignment by Meriggi 1934a:116f. /parn(a)-/ constitutes a basic level term and can hence be used in a variety of different contexts. One is KÖRKÜN, §4: wa/i-ti ku-ma-na á-sa-ti-ru-sá REX-ti-sá wa/i+ra/i-pa-si DOMUS-na “AEDIFICARE”, approx. /wa=di kuman Astirus hantawattis warpassi parna tamada/, meaning ‘When King Astiru built houses of weapons, (I set up this Storm-god Tarhunt of Aleppo there)’ (cf. Hawkins 2000a:172).
[AH.B.]
1.1.5Stem
The attestations for /parn(a)-/ were first sorted into a paradigm by Mittelberger 1963a:98f., with the correct identification as a neuter noun.Note that, at the time, the sign 〈za〉 was thought to represent /ī/. On comparative grounds, the word was originally a consonantal stem, but the nom./acc. sg. has been recharacterised as a thematic neuter stem, enlarged by the regular neuter particle /=sa, =za/. This may be the reason for the unexpected spelling 〈DOMUS-ni-za〉, i.e. /parnin=za/, attested at KARKAMIŠ A2+3, §15 (cf. the remarks by Hawkins 2000a:111).
[AH.B.]
1.1.6Derivatives
1.2 Cuneiform Luwian
parn(a)-
‘house, household’1.2.1Transmission
The attestations of the word are widely strewn across the Cuneiform Luwian genres (from MS onwards).
[E.R.]
1.2.2Forms
| nom./acc. sg. | pár-na-an | KBo 32.126 obv. 9’ | (CTH 772, MS) |
| pár-na-an-za | KUB 35.103 ii 8’ | (CTH 766, MS) | |
| KUB 35.103 ii 15’ | (CTH 766, MS) | ||
| [pá]r-na-an-za | KUB 35.54 iii 26 | (CTH 758, MS) | |
| dat./loc. sg. | pár-ni | KUB 35.54 iii 23 | (CTH 758, MS) |
| nom./acc. pl. | párar-na | KUB 35.68, 8’ | (CTH 763, NS) |
| dat./loc. pl. | pár-na-an-za | KUB 9.31 ii 31 | (CTH 757, NS) |
| p[ár]-na-an-za-aš(=) | HT 1 ii 7’ | (CTH 394, NS) | |
| erg. pl. c. | pár-na-an-⸢ti⸣-in-zi | KUB 35.54 ii 49 | (CTH 758, MS) |
| gen. adj. nom./acc. sg. n. | pár-na-aš-ša-an-za | KBo 29.25 iii 8’ | (CTH 764, MS) |
| gen. adj. nom. pl. c. | [p]ár-na-a-aš-ši-in-z[i] | KUB 35.3, 2’ | (CTH 759, NS) |
| [pár-na-aš]-ši-⸢in-zi⸣ | KUB 35.52 iii 8’ | (CTH 758, NS) | |
| gen. adj. dat. pl. c. | [pa]r-na-aš-ša-an-za-an-za | KBo 22.254 ii 6’ | (CTH 762, NS) |
| unknown | pár-na-a[n- | KUB 35.68, 5’ | (CTH 763, NS) |
| KBo 8.129 obv. 17’ | (CTH 759, NS) |
Contra HED PE, PI, PU:89, the ending -anzanza is not a mistake, but marks the plural possessor in the dat./loc. pl. (see Melchert 2000a:176).
[E.R.]
1.2.3Graphic Features
Checked against the photographs in the Mainzer Fotoarchiv. – The use of a mater lectionis (here párar-) is an unusual feature for a Cuneiform Luwian text. According to Otten 1953a:55, it serves to disambiguate the potential readings pár- and maš- of the first sign.
[E.R.]
1.2.4Meaning
Otten 1953a:55 elucidated the meaning CLuw. parn(a)- by way of equation with HLuw. DOMUS-na- ‘house’, which has stood the test of time. The interpretation has been generally accepted (Meriggi 1957a:198f., Meriggi 1957b:72, DLL:79f., HW:385, HEG P:567, HHw:124, EDHIL:666, HED PE, PI, PU:89).
[E.R.]
1.2.5Stem
In contrast with the preservation of the heteroclitic stem per/parn- n. ‘house’ in Hittite, Luwian transferred the noun to the thematic stems, yet keeping intact the neuter gender (for the process, see Melchert 2004d).
[E.R.]
1.2.6Derivatives
- parnalla/i-
- parnawaiške/a-
- ḪUR.SAG/URUParnassa (HED PE, PI, PU:89, Yakubovich 2010d:10)
1.3 Lycian A
prñna(?)
‘house(s) (?)’1.3.1Transmission
The base noun for ‘house’, which is well attested in other Anatolian languages, is perhaps found for the first time in a new grave inscription from Limyra. However, the uncertainty of the reading demands due caution.
[D.S.]
1.3.2Forms
| nom./acc. pl. n.(?) | p̣?ṛ?ñṇa | N 339 (Limyra) |
Differently, Christiansen 2019b:106 takes prñna to be the nom. sg. c. of an a-stem, which is formally possible. A neuter stem (cf. Luwian) with “collective” semantics is tentatively assumed here, because it was common with words for ‘house’.
[D.S.]
1.3.3Graphic Features
For the autography and the edition, see Christiansen 2019b:104-107.
[D.S.]
1.3.4Meaning
The word seems to refer to the rock-cut tomb, except that prñnawa-, the usual word, is not used. Christiansen 2019b:106 posits tentatively that prñna, if correctly read, might also be used to refer to the building or monument. She also offers an alternative reading and interpretation, which is less attractive than the first one. Nonetheless, a second attestation of prñna would be necessary to confirm her attractive reading and interpretation; cf. Neumann 2007a:283 with further literature for the reconstructed base.
[D.S.]
1.3.5Stem
It is not possible, based on this sole attestation, to tell whether the word for ‘house’ was thematized in Lycian to *prñne-, like in Luwian, or transferred to the a-stems (implicitly Christiansen 2019b:106), or whether it is still consonantal (see under Reconstruction). More evidence is needed.
[D.S.]
1.3.6Derivatives
1.4 Lydian
pira-
‘house’1.4.1Transmission
This noun is attested twice, once in the curse formula of the Aramaic and Lydian bilingual inscription found at the cemetery of Sardis and datable to the early 4th century BCE and once in an agreement found in the precinct of Artemis in Sardis and datable within the broad range of 5th - 4th century BCE.
[I.Y.]
1.4.2Forms
| acc. sg. | pira(ν) | LW 24.19 (Sardis) |
| dat. sg. | piraλ | LW 1.8 (Sardis) |
For the disappearance of the accusative marker ν before =k, see Gérard 2005a:77.
[I.Y.]
1.4.3Meaning
The noun pira- is paired with aara- ‘estate’ in both of its occurrences. In the bilingual LW 1, it corresponds to Official Aramaic byt ‘house’. Accordingly, the translation ‘house’ was tentatively applied to pira- already in Littmann 1916a:36. Later this interpretation received support from etymological considerations (cf. Reconstruction), and this helped to dispel doubts that the collocation aaraλ piraλ=k was translated into Aramaic in the reverse order.
[I.Y.]
1.4.4Stem
This is a common gender a-stem, the collocation with aara- would make its barytonic interpretation more likely.
[I.Y.]
1.5 Hittite
É-er/parn-
‘house’1.5.1Literature / Comments
CHD P:273-291
2. Reconstruction
2.1Reconstruction ‘house’
The Anatolian word for ‘house’ is widely attested as Hitt. É-er/parn- n., Luw. /parn(a)-/ n., Lyc. prñn- (mainly in derivatives, but cf. above) and Lyd. pira-. The formation is heteroclitic with -r-/-n- in origin (i.e. *pḗr/pr̥n-és as in Rieken 1999a:307f.), although the non-Hittite Anatolian languages have thematized the word based either on the weak stem as in Luwian or on the strong stem as in Lydian.Whether the -i- in Lydian reflects *ē or *r̥ remains open; cf. Sasseville 2017a:139 n. 20. Laroche 1965a:53f. etymologically connects the noun for ‘house’ with Hitt. peru-/perunaš n. ‘rock, cliff’ (HEG P:576-579, HED PE, PI, PU:33-36, 91, Rieken 1999a:308), which is culturally supported by the fact that many indigenous people of Anatolia lived in caves of rocks and cliffs. No etymology is suggested by Kloekhorst (EDHIL:666), and Tischler (HEG P:570f.) leaves it open. As pointed out in the literature, a loanword is contradicted by the archaic heteroclitic formation (Rieken 1999a:308,EDHIL:666); for a new Indo-European etymological proposal, see below under PIE lemma.
An adjectival derivative in *-u̯o-, i.e. *pr̥n-o-u̯o- (and a substantivization thereof *pr̥n-o-u̯eh2-) can be reconstructed at least for Proto-Luwic, see under Lyc. prn͂nawa-.
[D.S.]
3. Proto-Indo-European
3.1Semantic Reconstruction
The root etymology of the word family in question is not evident. The root ablaut and morphology of Hittite *pēr, gen. sg. parnaš seem archaic, which speaks against a loan from Egyptian pr ‘house’, as had been proposed in the early days of Hittite scholarship (see HEG P:569). A more plausible, though still uncertain, alternative is that the underlying Proto-Indo-European root is the same as in Greek περί, πέρι, Vedic pári, Old Avestan pairī ‘(all) around’, and Latin pariēs, -ētis m. ‘wall’ (as per Pedersen; see HEG P:570). The latter is better taken from a seṭ root (qua *pr̥hxi-h1it- ‘going around (vel sim.)’) than from *pr̥i-h1it- with unexpected vocalization in a non-Lindeman context, which suggests that Greek περί etc. are from *pérhxi (not *péri). This, however, is difficult to reconcile with Proto-Anatolian *pḗr, pr̥n-´ which shows no traces of any laryngeal. These reservations notwithstanding, the underlying formation could be a root noun *pér(hx)- ‘circular wall’ (“Ringmauer” as per Pedersen; see above) > Anatolian ‘house’, the locative of which, *pér(hx)-i ‘at the circular wall’ (“an der Ringmauer”; cf. Hittite dat.-loc. sg. pēri ‘in the house’ OH/NS, É-ri OH/OS) would then have been the basis for Greek περί, πέρι, Vedic pári, Old Avestan pairī ‘(all) around’differently LIPP II:617–621.
3.2Morphological Reconstruction
The heteroclitic inflection of Proto-Anatolian *pḗr, pr̥n-´ may go back to an actual heteroclitic stem *pér(hx)-r̥, gen. sg. *pr̥(hx)-n-és (cf. Rieken 1999a:306f., referring to Schindler 1973a:153), or it may represent an inner-Anatolian remodeling of a former root noun *pér(hx)-, gen. sg. *pr̥(hx)-és (see above). Under the latter view, the oblique in -n- (which was generalized as the base allomorph in all languages except Hittite and Lydian; see above) would be an innovation of Anatolian date, comparable to the heteroclitic s-stems of the type tunnakiš, tunnakišn- ‘inner chamber’ (on which see Melchert 2012e) and, more broadly, with heteroclitic stems such as Vedic dóṣ, gen. sg. doṣṇás n. ‘fore-arm’ (cf. Oettinger 1978a:83). A widely accepted explanation for these is that the *-n-forms originate in a reinterpretation of a locatival form in *-en (cf. Nussbaum 1986a:200ff.; Melchert 2012e:181), which is especially plausible for a word such as ‘house’.
[S.H.]
4. Transmission
Hittite É-er/parn- n. ‘house, household’, Cuneiform Luwian parn(a)- n. ‘house, household’, Hieroglyphic Luwian /parn(a)-/ n. ‘house’, Lycian A prñna(?) ‘house(s) (?)’, Lycian A prñnawa- c. ‘grave-house’, Lydian pira- c. ‘house’, etc.See EDHIL:666 s.v. per/parn-; HEG P:563–572; Rieken 1999a:306f.
(?) Vedic pári ‘(all) around’.Cf., e.g., EWAia II:91f.
(?) Old Avestan pairī, Young Avestan pairi, Old Persian pariy, etc.Cf., e.g., EWAia II:91f.
(?) Ancient Greek περί, πέρι ‘(all) around’ (cf. Mycenaean pe-ri-°).Cf., e.g., GEW II:512.
(?) Latin pariēs, -ĕtis m. ‘wall’ (Enn.+).Cf., e.g., WH II:254 s.v. paries.

