*sh3ógu̯-o-
‘eye’
- Language
- Proto-Anatolian
- Class
- Derivative
- Grammar
- subst
- ID
- 233
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/taw(i)-/, ta-wa/i-, COR/LITUUS
‘eye’1.1.1Transmission
The Hieroglyphic Luwian lexeme for ‘eye’ is only attested in late 8th century inscriptions.
[D.S.]
1.1.2Forms
| acc. sg. | /tawin/ | (“LITUUS”)ta-wá/í-na-´ | KAYSERİ, §3 (Tabal) |
| acc. coll. | /tawa/ | ta-wa/i | KARABURUN, §8 (Tabal) x2 |
| abl./instr. coll. | /tawadi/ | (COR)ta-wa/i+ra/i | KÜRTÜL §4b (Maraş) |
The instance of (“COR”)ta-wa/i-sà-ta-ti does not belong to this lexeme; see the derivative /tawasta-/.
[D.S.]
1.1.3Graphic Features
The example in KÜRTÜL, §4b was previously read as (COR)ta-wa/i, but the presence of the additional sign <+ra/i>, i.e. (COR)ta-wa/i+ra/i, is confirmed by Hawkins 2000a:272. For the new reading of the logogram L.341 as (COR) see van den Hout 2002a:182-185 and for the use of the determinative L.341 with the lexeme for ‘eye’, see Hawkins 1980d:135f.
[D.S.]
1.1.4Meaning
The semantics of the lexeme /taw(i)-/ were assigned by Morpurgo Davies 1980a:127 by means of a comparative analysis between the adverb HLuw. /tawiyan/ with CLuw. dawiyan. For the sense in KAYSERİ, §3 see Morpurgo Davies 1987a:210f.; cf. Hawkins 1980d:135f., Hawkins 2000a:482.
[D.S.]
1.1.5Compounds
1.1.6Derivatives
1.2 Cuneiform Luwian
tāwa/ī-
‘eye’1.2.1Transmission
Widely attested in Luwian rituals both in MS and NS.
[Zs.S.]
1.2.2Forms
| nom. sg. | da-a-u-i-iš | KUB 35.49 iv 9’ | (CTH 761, MS) |
| abl./instr. | [ta]-a-ú-wa-ti | KUB 35.43 ii 7 | (CTH 761, NS) |
| coll. pl. | da-a-u-wa | KBo 7.68 ii 15’ | (CTH 764, NS) |
| IGIḪI.A-wa | KUB 35.107 iii 19’ | (CTH 764, MS) | |
| dist. pl. nom./acc. | IGIḪI.A-˹wa-an-ta˺ | KUB 35.88 ii 5 | (CTH 766, NS) |
| dist. dat./loc. pl. | da-a-ú-wa-an-˹ta˺-[an-za] | KUB 35.70 ii 22 | (CTH 762, NS) |
| ta-wa-an-ta-an-[za] | KUB 35.71 ii 2 | (CTH 759, MS) | |
| ta-˹wa˺-an-ta-an-[za] | KBo 29.63, 5’ | (CTH 762, NS) | |
| gen. adj. nom. sg. c. | IGIḪI.A-wa-aš-ši-iš | KUB 35.107 iii 17’ | (CTH 764, MS) |
| gen. adj. nom. sg. n. | IGIḪI.A-wa-aš-ša-an-za | KUB 35.107 iii 16’ | (CTH 764, MS) |
| IGIḪI.A-za | KUB 35.88 iii 15 | (CTH 766, NS) | |
| KUB 35.89, 13 | (CTH 766, MS) | ||
| nom./acc. pl. n. | ˹ta˺-a-u-wa-aš-ša-an-za | KUB 35.43 iii 8’ | (CTH 761, NS) |
| gen. adj. abl./instr. | ta-wa-aš-ša-ti | KUB 35.43 ii 8 | (CTH 761, NS) |
| da-a-u-wa-aš-ša-ti | KUB 35.43 iii 25’ | (CTH 761, NS) | |
| [da-a]-u-wa-aš-ša-ti | KUB 32.14 + KUB 34.247 iii 7’ | (CTH 761, NS) | |
| da-a-u-wa-aš-ša-[an-za-ti] | KUB 35.24, 9’ | (CTH 761, MS) | |
| ˹da-a-u-wa-aš-ša-an-za-ti˺ | KUB 35.25, 8’ | (CTH 761, MS) | |
| da-a-u-wa-aš-˹ša-an˺-za-ti | KUB 35.43 ii 14 | (CTH 761, NS) | |
| ta-a-ú-wa-aš-˹ša˺-[an-za-ti] | KUB 35.43 iii 12’ | (CTH 761, NS) | |
| [da-a-u-wa-aš]-˹ša˺-an-za-ti | KBo 29.20, 8’ | (CTH 762, NS) |
Attestations following CLL:224 (cf. also Laroche 1959a:96) checked against the photographs of the Konkordanz and slightly corrected. Melchert adds “coll. pl.” [da]-˹a˺-u-wa in KBo 7.68 i 1’ (similarly Starke 1985b:361), but its existence cannot be proven, since the entire context is broken away, thus it is unknown to which word ]˹a˺-u-wa belongs. Laroche 1959a:96 suggests adding ˹ta˺-x-[ (read by him as ta-u-˹wa˺-[) in KUB 35.102 i 2 (CLL:224: “perhaps”), which, however, cannot be verified due to the entirely fragmentary context. The restoration [da-a-u-wa-aš]-˹ša˺-an-za-ti in KBo 29. 20, 8’ follows Starke 1985b:199, Starke 1990a:194. For nom.-acc. pl. n. ˹ta˺-a-u-wa-aš-ša-an-za in KUB 35.43 iii 8’ instead of ˹ta˺-a-u-wa-aš-ša-an-za suggested by Starke 1985b:146, see Yakubovich 2010d:49f..The dat.-loc. distributive plural forms (restored by Starke 1990a:135 with discussion) are “quite uncertain” according to CLL:224, although they are based on partly duplicate and partly parallel texts and Starke’s restoration presents the only morphological possibility. Nevertheless, the context is obscure, thus the identification of these forms rests merely on formal grounds.
Forms in -ant- are plurals (with CLL:224, Melchert 2000b:61, 65), not another stem contra Laroche 1959a:96 (cautiously; Tischler 2008a:195f., Tischler 2016a:372 treats them both as plurals and as separate stems). The terminology of the plurals adopted here is that of Melchert, although it requires further research. Note that Melchert’s emphatic rejection (CLL:224: plurals in -a are “NOT dual, contra Starke, StBoT 31.136 et aliter”) is a misunderstanding, since Starke 1990a:135f. clearly calls these forms plurals and only historically explains them from duals having changed his earlier views, when he indeed called them duals (Starke 1986a:col. 161).
[Zs.S.]
1.2.3Meaning ‘eye’
There is a general agreement regarding the meaning (Laroche 1959a:96; Josephson 1979a:182; Carruba 1982a:46; Melchert 1988b:217, CLL:224; Asan 1988a [2014a]:110f.; Starke 1990a:42, 47; Haas 1994a:440, 901, Haas 2003a:731; Kloekhorst 2008c:704; Tischler 2008a:195f., Tischler 2016a:372; Goedegebuure 2010a:308), due partly to the logographic spelling and partly to the context.
[Zs.S.]
1.2.4Derivatives
1.3 Luwian in Hittite transmission
tāwa/i-
‘eye’1.3.1Transmission
Hapax in a Hittite medical ritual in NS.
[Zs.S.]
1.3.2Forms
| dat.-loc. sg. | ta-a-u-i | KUB 8.38 ii 12’ | (CTH 461, NS) |
Attestation following CLL:224 (cf. also Laroche 1959a:96) checked against the photograph of the Konkordanz. Note that the final vowel is written plene in fact: ta-a-u-i-iš-ši, i.e. tāwī=šši. It is impossible to decide whether the logographic writing IGIḪI.A in line 8’ of the same text conceals this word or its Hittite equivalent.
[Zs.S.]
1.3.3Meaning ‘eye’
There is no doubt regarding the meaning (Laroche 1959a:96; Burde 1974a:34; CLL:224; CHD L-N:202; Polvani 1988a:65; Haas 2003a:117; Puhvel 2004a:195, HED M:90; Kloekhorst 2008c:704; Klinger 2010a:180; Singer & Gestoso Singer 2014a:318), since it can be corroborated independently from the homonymy with the Luwian word for ‘eye’ (see under Origin) by the context, which is a ritual against an eye disease (cf. esp. lines 8’-10’), where the goal is to induce lacrimation. The only exception is Ünal 2007a:709 (“a medical term”), for unknown reasons.
[Zs.S.]
1.3.4Origin
Although Laroche 1959a:96 already identified it as a Luwianism based on the homonymy with Luwian tāwa/ī-, see s.v. (followed by Burde 1974a:34; CLL:224), which is fully supported by its meaning (see above) and by the fact that the Hittite word for ‘eye’ is šākuwa-, it was not included in the list of Luwianisms of Melchert 2005a and van den Hout 2007a.
[Zs.S.]
1.4 Lycian A
tawa-
‘eye’1.4.1Transmission
This lexeme is found in grave inscriptions from Limyra.
[D.S.]
1.4.2Forms
| nom./acc. coll. | tawa | TL 106.2 (Limyra), TL 131.5 (Limyra) |
1.4.3Meaning
The meaning of tawa ‘eyes’ is assigned by Eichner 1985a:19 n. 26 comparing CLuw. dāwa ‘eyes’, cf. DLL:64. Furthermore, he compares the curse formula trbbala hati tawa with the Hittite formula parnaššea šuwaizzi, in both of which he sees the sense of ‘looking’, i.e. Lyc. tawa- ‘eyes’ and Hitt. šuwa(i)-(mi) ‘to spy’. Even though the signification of the two compared formulas now differs, his semantic assignment remains convincing. For various interpretations regarding the Lycian formula, cf. DLL:69, Schürr 2006a:119, Christiansen 2009a:49f., Schürr 2010a:156 n. 20. On the meaning of the Hittite formula see Giger-van den Heuvel 2009a with further literature.For an overview of the older literature and outdated hypotheses regarding the meaning of Lyc. tawa-, see Neumann 2007a:339.
1.4.4Compounds
1.5 Lydian
*šaw(i)-
‘eye’1.5.1Compounds
1.6 Hittite
šākuwa-
‘eye’1.6.1Literature / Comments
For the attestations, see CHD Š:65-77; cf. Rieken 1999a:59-61, Kloekhorst 2008c:704.
2. Reconstruction
2.1Reconstruction ‘eye’
The comparison between Hittite šākuwa- coll., Luw. /taw(i)-/ c., Lyc. tawa- coll. and Lyd. *šaw- leads to the reconstruction of a Proto-Anatolian root *sh3ogu̯-. The opposition between initial t- in Luwian and Lycian versus initial s- in Hittite and Lydian speaks for a reconstruction with initial *sh3- (Eichner 1973a:82, Oettinger 1979a:395, Rieken 1999a:59-61, Kimball 1999a:434, Melchert 2007-2008a:187 n. 14, Oettinger 2013a:170, pace Olsen 2006a:243f, Cohen & Hyllested 2012a:63).
Kloekhorst 2006b:132 (and Kloekhorst 2008c:705f.) explains the lenition of PIE *ku̯ > PA *gu̯ as a result of the preceding accented *-ó-. This sound law is a priori more attractive than the previous one assigned by Melchert 1994a:61, in which the PIE *ku̯ always becomes PA *gu̯ in medial position. The lenition of a consonant after *-ó- should be restricted to the second laryngeal following Melchert 2012c. However, Melchert 2012c:183 n. 16 concedes to the possibility of *ku̯ > PA *gu̯ /ó_, but makes the reader aware that an analysis of all the relevant data is missing. For a positive criticism on the leniting capacity of original short *-ó- see Kümmel 2012a:309f. On the other hand, one cannot exclude a root with a lengthened o-grade, which would have lenited regularily the following consonant.
Collective formations are found in Hittite, Luwian and Lycian, which have been interpreted as reflexes of the former dual of a thematic stem, i.e. PA *sh3ógu̯oh1, cf. Starke 1986a:col. 161, Starke 1990a:136, Melchert 2003b:185, Oettinger 2013a:170. The existence of the lexeme *šakui- c. and the reading behind the logogram IGIḪI.A-i- c. are a highly debated matter, which is not relevant for the present discussion; cf. Kloekhorst 2008c:705f. with further literature. Kloekhorst 2008c:705f. reconstructs a neuter plural in *-eh2, although there is no neuter singular form attested. Moreover, since Luwian /taw(i)-/ is common gender, the reconstruction of a common gender stem is more economical. In my opinion, it is clear that the lexeme for ‘eye(s)’ in Anatolian entered the system of alpaš/alpeš/alpa (Eichner 1985c:147-150) and that the function of the dual number was replaced by the comprehensive plural (collective) in -a. This is directly proven by Lyc. A tawa ‘eyes’, where the final -a goes back to *-eh2, instead of the phonologically expected outcome of the dual *tewe (< *sh3ógu̯oh1).
Regarding the stem formation, we choose to reconstruct a τόμος-Type, i.e. *sh3ógu̯-o-, for Proto-Anatolian (Melchert 1992a:40 n. 14). Rieken 1999a:59-61 reconstructs a neuter consonantal stem and finds evidence in the Hittite lexeme šakuni- c. ‘source’, where she posits a derivative in -ni- from a consonantal stem *šāku̯-. In spite of a parallel in the Akkadian language, the connection of šakuni- c. ‘source’ to the same root as šākuwa ‘eyes’ is considered tentative. A different approach would be to reconstruct a consonantal stem, which would have been secondarily thematized. The Anatolian data alone is very unclear and the formation chosen here must be considered provisional until further evidence.
[D.S.]
For the lemma head see PAnat. *sh3ógu̯-/sh3gu̯-´.
3. Proto-Indo-European
3.1Semantic Reconstruction
The Anatolian, Indo-Iranian, Germanic and Balto-SlavicLat. oculus ‘eye’ can be adduced, too, if it is to be reconstructed as *oku̯e‑lo‑, viz. as a derivative of an o‑stem *(s)h3oku̯‑o‑. EM:458 doubt that *‑lo- is a deminutive suffix here. For the semantics, cf. Gk. ὄμμα ‘eye’ → ὀμμάτιον ‘little eye’ > ModGk. μάτι ‘eye’.
Nussbaum (apud Weiss 2011a:195 n. 9) suggests, that the masculine o‑stem Lat. oculus ‘eye’ “may have been backformed from an originally neuter dual *okweloi” ‘both eyes’: (nom.‑acc.) du. n. → (nom.) pl. m. oculī ‘the eyes’.evidence straightforwardly points to a noun meaning ‘eye’. If it is an action noun (τομός-type), the meaning can be traced back to ‘the seeing one’ (< ‘the attaching one’, cf. Gk. *ὀπός ‘follower; following’).On the connection of ‘to see’ and ‘to follow’, cf. PAnat. *sHógu̯-o-/*sHgu̯-´.The deviant meanings in Vedic (‘?dice’), Greek (‘hole’) and Baltic (‘hole’) are, potentially independent, metaphorical interpretations of the ‘eye’ as a dark spot on a light surface (‘dice’, cf. ModE snake eyes, which is when the outcome of rolling two dice is two) or as a circular aperture (‘hole’, cf. ModE bull’s eye ‘side light’).
3.2Morphological Reconstruction
The simple thematic noun PIE *(s)h3oku̯‑ó- is continued with an anlauting *s- only in Anatolian (cf. Hitt. šākuwa‑, HLuw. taw(i)‑, etc.)On PIE *sh3- > Hitt. š‑, Luw. t‑, cf. PAnat. *sHógu̯-o-/*sHgu̯-´.but everywhere else without.On the s mobile, cf. PAnat. *sHógu̯-o-/*sHgu̯-´ and the discussion of Gk. *ὀπός ‘follower’.The only exception to this could be Gk. *ὀπός ‘follower; following’ if it is regarded as a psilotic variant of *ὁπός < PGk. *soku̯‑ó‑, cf., e.g., EDG:1089. This assumption, however, is circular.
Not only the Vedic accent (cf. akṣá‑) is indicative of the oxytonesis of PIE *(s)h3oku̯‑ó- but also the meaning which points to an action noun of the τομός-type. This kind of accentuation, however, contradicts the barytone stem PAnat. *sHógu̯‑o- (cf. above). The problem dissolves under the assumption of substantivizing accent retraction, viz. PIE (adj.) *(s)h3oku̯‑ó- ‘seeing’ → PAnat. (sbst.) *sHógu̯‑o- ‘the seeing one, eye’.
However, post‑PIE derivation of the assumed o‑stem from other nominal stems or the verbal root is possible in Indo-Iranian (*akša- : *akš- ‘eye’), Greek (*ὀπός : ὀπ- ‘to see; eye’) and Balto-Slavic (*oko- : *ok- ‘eye’). Nevertheless, on the basis of at least Anatolian, Latin (oculus ← PIt. *oku̯o- ‘eye’) and Germanic (OHG aga° ‘eye’) it is permitted to reconstruct a proto-form PIE *(s)h3oku̯‑ó‑.
[A.O.]
4. Transmission
PIIr. *akša- ‘eye’, cf. Ved. (‑)akṣá- ‘?dice’, ‘‑eyed (as a compound second member)’, PIran. *axša- ‘eye’ in *abi̯‑axša- ‘overseer’← *‘to have an eye upon’. PIran. *abi̯‑axša- is assumed to be the derivational base of YAv. aiβii-āxšaiia- ‘to supervise’, cf. EWAia I:42f. and Lyd. šawtaar-.
EWAia I:42f., Grassmann 1996a:5, Bartholomae 1904:97, 310f.
(?)Gk. *ὀπός ‘follower; following’ or (?)Gk. *ὀπή ‘entourage, fellowship’, cf. Gk. ὀπᾱ́ων ‘companion’ (Il.+), Myc. personal name (m.) dat. sg. O‑qa‑wo‑ni /ō̆ku̯āu̯ōni/ (DMic II:46) etc.On the word formation, cf. Oettinger 2017 [2018].
Gk. ὀπή ‘(peep)hole; cave’ (Ar., Arist.+), ‘face’ (Kerk.4.23)The compound member Gk. °οπός ‘(looking) like (X)’ is secondary and forms possessive compounds < ‘looking (like X)’.
GEW II:401f., GEW III:162, DELG:807, EDG:1089, DMic II:45f.
OHG adv. aga‑wis ‘officially’
EWAhd I:88f.
BSl. *oko- ‘eye’, cf. Lith. (dial.) ãkas m., akà f. ‘hole (in the ice)’, Latv. aka ‘well’, (?)OPr. *akas ‘hole; eye’;This is the derivational base of OPr. accodis (EV) ‘rochloch = hole in the wall for the elimination of smoke, smoking gap’, cf., e.g., Schmalstieg 2001a, EDBIL:46. Mažiulis (PKEŽ² s.v. accodis) connects it ultimately with PIE *skedh2- (LIV²:550f.) ‘to split, scatter’.OCS oko, gen. oka (n.) ‘eye’,Beside these, there are also s‑stem forms, e.g. gen. sg. očese. The s‑stem is probably analogical after OCS uxo, nom. pl. ušesa, nom.‑acc. du. uši ‘ear’, cf. ALEW I:27, Stüber 2002a:62.Russ. óko, Cz. oko, SCr. ȍko
EDBIL:46, ALEW I:25–27, LEW I:5, PKEŽ² s.v. accodis; EDSIL:365, Vasmer II:259f.