*s(h1)r̥h1-meh2-
‘division, temenos’
- Language
- Proto-Anatolian
- Grammar
- subst
- ID
- 1212
1. Individual Anatolian Languages
1.1 Lycian A
hrm̃ma-
‘temenos’1.1.1Transmission
This lexeme is found in two grave inscriptions.
[D.S.]
1.1.2Forms
| acc. sg. | hrm̃mã | TL 84.4 (Sura) |
| TL 84.5 (Sura) | ||
| TL 149.13 (Rhodiapolis) |
Formally, a dative-locative plural hrm̃ṃẹ (N 348.4) referring to more than one temenos could be posited, although it could be taken instead as a personal name (thus Onur & Tekoğlu 2020a:19). Therefore, this attestation is not listed in the table above.
[D.S.]
1.1.3Meaning
The substantive hrm̃ma- refers to something that can be “built” for religious purposes. In TL 84, the author of the inscription made himself a hrm̃ma-: s(e)=ed(e) adẽ hrm͂mã ijase atlahi (TL 84.3-4) ‘He made his own hrm̃ma- for/in the ijase.’. This is followed by instructions regarding religious offerings pertaining to the hrm̃ma-. An etymological equation with Lyd. širma- c. ‘temenos’ is both phonologically and semantically adequate. The meaning ‘temenos’ is also appropriate for the context (thus DLL:25, contra Neumann 2007a:101). For further evidence see under Reconstruction.
[D.S.]
1.1.4Stem
The Lycian word hrm̃ma- is a common gender a-stem (contra Starke 1990a:280f., Martínez Rodríguez 2018a:276-278), which is confirmed by the genitival adjective atlahi (acc. sg. c.) in TL 84.4 and pñtbahi (acc. sg. c.) in TL 84.5. For the rare use of the enclitic pronoun =ede (instead of =ẽne) with the common gender singular; cf. TL 21.4-5. For further evidence regarding its gender assignment see under Reconstruction.
[D.S.]
1.1.5Derivatives
1.2 Lydian
širma-
‘temenos, precinct’1.2.1Transmission
This noun is attested in inscriptions of various genres, all of which are found in Sardis. Most of its occurrences belong to the mutually interconnected inscriptions LW 23-24, found in the precinct of Artemis. The inscription LW 71 is dated to the 4th century BCE, whereas other texts can be only dated within the broad range of 5th - 4th century BCE.
[I.Y.]
1.2.2Forms
| nom. sg. | šyrmas | LW 23.1 (Sardis) |
| dat. sg. | širmaλ | LW 12.7 (Sardis) |
| LW 23.2 (Sardis) | ||
| poss. adj. nom. sg. c. | šrmliš | LW 24.3 (Sardis) |
| LW 24.16 (Sardis) | ||
| šrmli[š] | LW 71.2 (Sardis) |
1.2.3Graphic Features
The restoration of [šer]ḷiš šrmlị[š] in LW 71 is reliable.
[I.Y.]
1.2.3.1Phonological Interpretation
According to Gérard 2005a:37, the use of the rare grapheme
[I.Y.]
1.2.4Meaning
The inscription LW 23 makes a reference to ‘this šyrma-’, while the related inscription LW 24 informs us that širma- possesses a supervising authority, šerl(i)-. Given that both inscriptions are found in the precinct of Artemis in Sardis, a reference to this location imposes itself, as was first hypothesized in Brandenstein 1929a:272. But it was not clear whether the word refers to a temple building, as accepted in Gusmani 1964a:1964 and much of the earlier literature, or to a sacred plot of land (temenos), as this term is consistently translated, for example, in Melchert 1994a. Two independent points support the second solution. First, the inscriptions LW 23-24 presumably belong to the Achaemenid period, when the precinct of Artemis at Sardis still lacked an imposing structure that could be compared to the later Hellenistic temple. Second and no less important is the phonetically impeccable comparison between Lyd. širma- and Lyc. A hrm͂ma- ‘temenos’.
[I.Y.]
1.2.5Stem
Lydian širma- is an a-stem, or perhaps a/e-stem. The disappearance of the stem-final final vowel in the possessive adjective is due to a productive syncope (cf. above).
[I.Y.]
2. Reconstruction
2.1Reconstruction ‘division, temenos’
The equation of Lycian A hrm̃ma- c. and Lyd. širma- c. allows us to reconstruct a Proto-Anatolian formation in *-meh2- c., contra Melchert 1994a:288, who reconstructs a neuter men-stem, although that would be expected to yield a neuter plural stem in -ma, e.g. alãma ‘names’, and not a common gender one (but cf. Melchert 1999e:371). The reconstruction of an original stem in *-men- is excluded by the Lydian cognate, because the Proto-Indo-European suffix *-men- in Lydian is continued by stems in -mẽ- c. (Melchert 2004b:141f.). The root is related to Hitt. šārr-/šarr-(ḫi) ‘to divide up, to separate’ (and the stem-variant šarriye/a-(mi); HEG S:867 with further literature, Martínez Rodríguez 2018a:277 n. 14). The geminate -rr- in the Hittite root of šārr-/šarr-(ḫi) requires the reconstruction of a Proto-Anatolian root ending in a laryngeal, which may depend on the Proto-Indo-European etymology (cf. EDHIL:727-729 with further literature). Contra Starke 1990a:280f., the Hittite/Luwian word šaramman- n. ‘acropolis’, which contains a different root (thus also Martínez Rodríguez 2018a:277 n. 14), has nothing to do with the Lycian word hrm̃ma- ‘temenos’. From the Anatolian perspective alone, a root of the shape *serH- would be reconstructed, but from the Indo-European perspective advocated below, a root of the shape *sh1erh1- is adopted; see Part 2: PIE.
[D.S.]
3. Proto-Indo-European
3.1Semantic Reconstruction
Judging from the present Hitt. šārr-i/šarr- ‘to divide up, separate’, a PIE root *serh1- meaning ‘to divide, separate’ vel sim. can be reconstructed.
3.2Morphological Reconstruction
According to Kloekhorst, EDHIL:729, Hitt. šārr-i/šarr- ‘to divide up, separate’ “can only go back to a root *serh1- […]. I therefore mechanically reconstruct this verb as *sórh1-ei / *srh1-énti.” As for nominal formations, Lyc. A hrm͂ma- c. ‘temenos’ and Lyd. sirma- c. ‘id.’ point to a noun stem PAnat. *sr̥h1-meh2- (or *s(h1)r̥h1-meh2-) c. ‘division, temenos’; see above sub PAnat. Reconstruction and below. Outside Anatolian, however, there seem to be “no good comparanda”, according to EDHIL:729.
Interestingly, beside PIE *serh1- ‘to divide, separate’ there is evidence for a similar-sounding and synonymous root *h1erh1- ‘to separate’; cf. Watkins 2000a:24, who reconstructs a root PIE *h1erh1- ‘to separate’, which, in his view, was metathesized from older *h1reh1- ‘id.’. According to Watkins, this root underlies formations such as Lat. ratis f. ‘raft’ < *h1rə1-ti- ‘grating, latticework’,Differently EDLIL:514f., where Lat. ratis is related to the PIE root *h1reh1- ‘to row’ (LIV²:251f.): “ratis might be interpreted as *Hrh1-ti- ‘the rowing’ > ‘raft, boat’; note that rafts are normally propelled and steered by means of a pole” (EDLIL:515). Lat. rēte n. ‘net’ < *h1reh1-ti-,But cf. EDLIL:521 with different views and further lit. and Gk. ἐρῆμος adj. ‘empty, desolate, lonely, bereft’ (Il.+) < *h1reh1-mo-.Cf., e.g., EDG:456f., where also Lith. ìrti ‘to disintegrate’, Latv. ir͂t ‘to disintegrate, crumble’ are connected. Differently LIV²:271f. and ALEW I:400 (deriving the Baltic forms from PIE *h2erH- ‘to disintegrate, disappear’) as well as EDBIL:205 (reconstructing an aniṭ-root “*Hr-”). Cf. perhaps also Lat. rārus ‘of loose structure, sparse, rare’ (Plaut.+) if from *h1r̥h1-ro- (cf. EDG:457, Schrijver 1991a:310f.).
The formal and semantic similarity of *serh1- and *h1erh1- could suggest a common origin, thus pointing to one PIE root *(s)h1erh1- ‘to divide, separate’ (with initial s-mobile). Some of the above-mentioned forms might furthermore point to a root variant *(s)h1reh1- with full grade II (e.g., Gk. ἐρῆμος adj. ‘empty, desolate, lonely, bereft’, which is most straightforwardly explained from a preform *h1réh1-mo-).
[T.S.]