*k̑(u)u̯on-
‘dog’
- Language
- Proto-Anatolian
- Class
- Base
- Grammar
- subst, common gender
- ID
- 255
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/zuwan(i)-/, zú-wa/i-ni-, CANIS
‘dog’1.1.1Transmission
The earliest attestation of the Luwian lexeme /zuwan(i)-/ is in KARKAMIŠ A6 dated around 800 BCE. The two other attestations occur in later inscriptions of the 8th century.
[D.S.]
1.1.2Forms
| nom. sg. | /zuwanis/ | zú-wa/i-ni-i-sá | KULULU 1, §11 (Tabal) |
| nom. pl. | /zuwaninzi/ | (“CANIS”)zú-wa/i-ni-zi | ASSUR letter b, §8 (Assur) |
| CANIS-ni-i-zi | KARKAMIŠ A6, §31 (Karkamiš) |
1.1.3Graphic Features
We accept the reading of the sign L.448 〈zú〉, which was previously read as 〈sù〉, cf. Melchert 1987a:202, Hawkins; Morpurgo Davies 1988a:179. For a summary of the arguments, cf. Simon 2008a, Yakubovich 2010d:66f. n. 58, Melchert 2012a:210.
[D.S.] [E.R.]
1.1.3.1Phonological Interpretation
Based on the Proto-Anatolian and Proto-Indo-European reconstruction of the paradigm of the lexeme for ‘dog’, this root never had a full accented e-grade, whence Čop’s law cannot have applied here. Thus, one does not expect the nasal to be geminate **/zuwannis/, and accordingly the lexeme is here phonologically interpreted as /zuwanis/. On account of this conclusion, the personal name mZu-wa-an-ni in KUB 10.96, 3 (Laroche 1966b:214 nb. 1581) attested in a Hittite colophon cannot be equated with HLuw. /zuwan(i)-/ ‘dog’.
[D.S.]
1.1.4Meaning
The semantic assignment of the lexeme was facilitated by the hieroglyph “CANIS” (L.98) picturing a dog, see Hawkins 1981a:148, 172f. with further literature. The lexeme for ‘dog’ is used twice in curse formulas, where it refers to the pet(s) of a deity, who will inflict pain upon the offender (KARKAMIŠ A6, §31; KULULU 1, §11), cf. Hawkins 2000a:444. The third instance is found in a letter, where the author orders two dogs of a specific type (see /war(ri)muttall(i)-/), cf. Giusfredi 2010b:213-216.
[D.S.]
1.1.5Derivatives
1.2 Carian
qan
‘dog ?’1.2.1Transmission
Hapax in a graffito from Silsilis (Egypt) of unknown date (although it may be connected with the Nubian campaign of Psammetichus II in 593/592, similarly to the other “southern Egyptian” Carian inscriptions [Adiego 2007a:2, 31], if the personal name psmaśk of the inscription refers to this pharaoh).
[Zs.S.]
1.2.2Forms
| nom. sg. ? | qan | E.Si 2, 2 |
The attestation follows Adiego 2007a:406, no photo is available for checking (for a critical discussion of the published drawings and the documentation see Schürr 2000c:171f.). It might also be attested in qanor, if it is a compound, see the discussion s.v.
[Zs.S.]
1.2.3Meaning ‘dog ?’
Inspired by the hunting scene showing a dog chasing two birds (ostriches according to Legrain 1905a:129, although they rather look like flamingos, one of them is obviously flying or just lifting off away from the dog, Legrain 1905a:Fig. A) drawn under the inscription presumably by the same hand (Legrain 1905a:129) Schürr 2000c:173f. suggests identifying the meaning of qan as ‘dog’, but since this would be phonologically irregular (PIE *k̑ > Carian s), he proposed a loanword from Lydian *kãn- (Adiego 2007a:406 called this a “somewhat risky interpretation”). Setting aside the problem of the existence of the Lydian word (see s.v.), the main problem with this interpretation is that the inscription remains fully opaque with the single exception of the personal name Psmaśk (that may refer to the ruling pharaoh, as per above). Nevertheless, a connection between the drawing and the inscription is plausible and the similarity of qan to the Anatolian words for ‘dog’ is remarkable. In fact, it may even be the regular Carian reflex: First, the change PIE *k̑ > Carian s must have been Proto-Carian, since it is not Proto-Luwic (the Luwian reflex is ‹z›) and it cannot be Proto-Caro-Lycian either, since the Lycian A change *s > h preceded this change (Melchert 1994a:328). Second, we know now that one of the sources of Carian ‹q› was *qu <*ku (Simon 2011b:544f., Simon 2017f:387). If the Proto-Carian change *ku > qu included *k̑u (which is phonologically plausible) and if this change preceded *k̑ > s (what we do not know, but it is entirely possible), then qan is the regular reflex of the Anatolian word for ‘dog’.
[Zs.S.]
1.2.4Derivatives
1.3 Lydian
*kan-
‘dog (?)’1.3.1Transmission
The Lydian lexeme for ‘dog’ is not attested in the Lydian corpus. It is reconstructed based on the name Κανδαύλης and the appellative κανδαῦλα (acc.) found in Greek transmission. The Greek Iambic Poet Hipponax of Ephesus (6th BCE) equates the Greek word κυνάγχα with the Lydian word κανδαῦλα and refers to them as epithets of the god Hermes (Hipponax 2, Degani 1991a:24). This equation offers a reliable source for the meaning of this epithet. On the other hand, the origin of the name of the Lydian king cited in Herodotus (1.7.2) is a speculative matter. In spite of the fact that the name Κανδαύλης and the epithet κανδαῦλα (acc.) sound similar, they must not have the same origin.
Their etymologies have been strongly debated. A thorough discussion of the problem and a summary of the literature are offered in Sh. Hawkins 2013a:167-182.
[D.S.]
1.3.2Graphic Features
1.3.2.1Phonological Interpretation
Based on the Lydian word for ‘dog-strangler’ in Greek transmission, Melchert 1994a:331, 359f. reconstructs the Lydian lexeme *kãn- ‘dog’. He posits the following development: *k̑ṷónV- > *kṷónV- > (+ delabialization) *kãn-. In spite of the plausibility of the scenario, no weight for the Lydian phonology or for the shape of the Proto-Anatolian reconstruction can be given to this lexeme, as long as it is not attested in an actual Lydian text; cf. Oettinger 1995b:39-43, Schürr 2000c:174.
[D.S.]
1.4 Palaic
kuwan-
‘dog (?)’1.4.1Transmission
This lexeme is a hapax legomenon attested in the festival for the Palaic pantheon (CTH 751).
[D.S.]
1.4.2Forms
| nom. pl. | ku-wa-ni-iš | KUB 35.165 rev. 6’ | (CTH 751, OS or MS?) |
Checked against the photograph of the Konkordanz.
The word ku-wa-ni-iš in KUB 35.165 obv. 16 does not belong here, see kuwani- ‘woman (?)’.
[D.S.]
1.4.3Meaning
The context refers to three plural entities, kuwāiš, kuwaniš, and warlaḫiš, being offered or sacrificed in different designated areas. The first two of them can be identified as ‘cows’ and ‘dogs’ on etymological grounds, the identifications mutually support one another. For the sacrifice of puppies in Hittite rituals see Collins 1990a:218-226. Since human sacrifices were quite rare in Hattusa, while the designated sacrifices of women appear to be altogether unattested (Collins 1990a:220), the context speaks against translating kuwaniš as ‘women’ in the passage under discussion.
[I.Y.]
1.4.4Stem
The classification as an n-stem is not assured synchronically, but supported by the comparative data.
[D.S.]
1.5 Hittite
kuwa-/ku(wa)n-
‘dog-man’1.5.1Literature / Comments
Melchert 1989a, Rieken 1999a:28-30, Kloekhorst 2008c:505f.
1.6 Miscellaneous
Armenian šown
‘dog’1.6.1Origin
Greppin 1978a:119–121, Greppin 1980b:205, Greppin 1982a:70 suggested that Armanian šown ‘dog’ is a loanword from Hieroglyphic Luwian zuwan(i)- ‘dto’. This was rejected by Simon 2013b:114 referring to its impeccable Proto-Indo-European origin (note also the unexplained difference of the initial consonants), on which see most recently EDAIL:520f. with discussion and ample refs. (cf. also below under PIE Etymology).
[Zs.S.]
2. Reconstruction
2.1Reconstruction ‘dog’
The reconstruction of the palatal velar *k̑ for Proto-Anatolian is supported by the equation between Hitt. LÚkuwa(n)-, Pal. kuwan- and HLuw. /zuwan(i)-/ (see Melchert 1994a:94, 192, 234, Melchert 2012a). The o-vocalism in the root can be reconstructed based on the Hittite and Palaic material, both of which preclude the reconstruction with an e-grade.
An ablauting paradigm can only be observed among the Hittite attestations: nom. sg. LÚkuwaš, acc. sg. LÚkuwanan, gen. sg. LÚkūnaš. By means of Indo-European comparative data, Rieken 1999a:28-30 assigns to this paradigm a root ablaut alternating between o-grade in the strong cases and zero-grade in the weak cases, i.e. *k̂(u)ṷón-/k̂un-´. The spelling [LÚk]u-ú-na-aš with the sign <ú> would indicate according to Melchert 1989a:99 followed by Rieken (ibid.) a retraction of the accent as in Vedic śúmas. According to Kloekhorst 2008c:506, the spelling with plene <ú> is only used to stress the fact that it contains the phoneme /u/ as opposed to /o/. Since he expects the phoneme /u/ to be lowered to /o/ in front of nasals, this unexpected phoneme /u/ would have been reintroduced by analogy to the full grade stem. In the end, the simpler scenario would be the one with the retraction of the accent.
[D.S.]
3. Proto-Indo-European
3.1Semantic Reconstruction
PIE *k̑(u)u̯ō̆́n-/*k̑un-´ m./f. means ‘dog, bitch’, cf. EIEC:168, NIL:436, Watkins 2011a:47f. For the gender (epikoinon and masculine), cf. Matasović 2004:88. In Indo-Iranian, there is also a feminine stem formed with the suffix *-ih2- (Ved. śunī́-, YAv. sūnī-). In a figurative sense, PIE *k̑(u)u̯ō̆́n- represents a technical term at the dice game meaning ‘worst throw at dice’, which is confirmed by Gk. κύων, Lat. canis and indirectly also by the derivative Ved. śva-ghnín- m. ‘winning player, winner’, literally ‘the one who kills the dog (= the worst throw)’ (cf. EWAia II:674 s.v. śvaghnín-, Jackson 2007a:227f.).
3.2Morphological Reconstruction
The word for ‘dog’ can be reconstructed as an o/Ø-ablauting stem: PIE *k̑(u)u̯ō̆́n-/*k̑un-´ (cf. NIL:436), with the weak stem appearing as prevocalic *k̑un-V° (cf. Ved. gen. sg. śún-aḥ) and preconsonantal *k̑u̯n̥-C° (cf. Ved. dat./abl. pl. śvá-bhyaḥ). The protoform can be interpreted as a root noun (so, e.g., Schindler 1972a:34, Schindler 1972b:47, 69, 89, NIL:437 n. 1) or as an n-stem noun (*k̑(u)u̯-ō̆́n-/*k̑u-n-´, cf. NIL:437 n. 1 with lit.). In the case of a root-noun analysis, the etymology remains unclear. The interpretation as an n-stem (*k̑u̯-ō̆́n-), on the other hand, seems semantically more convincing and also makes sense in terms of the derivational history. As recently proposed by Höfler 2015b, the word for ‘dog’ could have been derived from PIE *pék̑-u- n. ‘livestock’ (Ved. páśu-, Lat. pecu, etc.): PIE *pék̑-u- n. ‘livestock’ → adj. *pk̑-u̯-ó- ‘being with livestock’ → sbst. (substantivizing n-stem) *[p]k̑-u̯-ó-n- ‘the one who is with livestock’ > ‘dog’ (Höfler 2015b:237 with n. 41). – The weak stem of PIE *k̑(u)u̯ō̆́n-/*k̑un-´ served as derivational base of other stems. Indo-Iranian, for example, points to a stem pre-PIIr. *k̑u̯n̥-k(u̯)o- adj. ‘doglike’ vel sim. (Skt. śvaka- ‘wolf’, YAv. spaka- ‘doglike’, MPers. sag ‘dog’). An already PIE derivative might have been *k̑un-tó- (NIL:437f. with n. 19), which is attested in Germanic (PGerm. *χunđa- m. ‘dog’ > Goth. hunds, ON hundr, etc.) and, in extended form, in Baltic (Latv. sùntene, sùntana f. ‘big dog’, cf. NIL:437, 439 n. 20). For further derivatives of PIE *k̑(u)u̯ō̆́n-, cf. NIL:436–440.
3.3Syntactic Reconstruction and Phraseology
The word for ‘dog’ is attested in a number of PIIr./PIE poetic phrases (EWAia II:675): cf., e.g., Ved. r̥jí-śvan- (name, RV; literally ‘having swift dogs’) ~ Gk. Hom. κύνες ἀργοί ‘swift dogs’, Kύν-αργος (name, Eretria; cf. Schmitt 1967a:16, 245, 246 n. 1427, EWAia I:251 s.v. r̥jíśvan-, each with further lit.); Ved. śvā́nau catarakṣau ‘the two four-eyed dogs’ (RV 10.14.10,11) ~ YAv. spānəm … caϑru.cašməm ‘four-eyed dog’ (cf. KEWA III:403 with lit.); Ved. śvá-pad- m. ‘wild animal’ (literally ‘having the feet of a dog’; AV+) ~ YAv. (if correct) spō.pad- m. (name; cf. Mayrhofer 1979a:78).
[T.S.]
4. Transmission
Ved. ś(u)ván- m. ‘dog’ (nom. sg. śvā́, śuvā́, acc. sg. śvā́nam, gen. sg. śúnaḥ, dat./abl. pl. śvábhyaḥ, etc.; RV+), also f. ‘bitch, female dog’ (RV 9.101.1) beside śunī́- f. ‘id.’ (AV+); cf. also śva-ghnín- m. ‘winning player, winner’ (RV+). – YAv. span- m. ‘dog’ (nom. sg. spā, acc. sg. spānǝm, gen. sg. sūnō, etc.), YAv. sūnī- f. ‘bitch, female dog’; cf. also YAv. spa-ciϑra- adj. ‘belonging to the race of dogs’, YAv. spō-jata- adj. ‘killed by dogs’, YAv. spā-bǝrǝz- adj. ‘having the height of a dog’Cf., e.g., AiG II,2:73, AiGr III:278f., EWAia II:674, de Vaan 2003a:289, 310, ALF:141f., NIL:436f., 438 nn. 5–6.
Gk. κύων m./f. ‘dog, bitch’ (acc. sg. κύνα, gen. sg. κυνóς, dat. sg. κυνί, etc.; Il.+), κυνά-μυια f. ‘dog-fly’ (Il.+); Myc. ku-ne /kunes/ nom. pl. ‘dogs’, first compound member in ku-na-ke-ta-i /kunāgetā(h)i/ dat. pl. ‘huntsmen’ (cf. Gk. κυνηγέτης ‘huntsman, hunter’ [Od.+], Dor. κυνᾱγέτᾱς [Pi. N. 6.14])Cf., e.g., GEW II:49 s.v. κυνάμυια, 58f. s.v. κύων, DMic I:402, Bartoněk 2003a:184, 366, 572, NIL:436f., 438 nn. 8–9, EDG:803 s.v. κυνάμυια, 811 s.v. κύων.
Arm. šown ‘dog’, first compound member in šanačanč ‘dog-fly’Cf., e.g., Olsen 1999a:133f., 794, 672f., NIL:436, 438 n. 7, EDAIL:520–522.
TochB ku m. ‘dog’, TochB kuñiye adj. ‘pertaining to a dog, stemming from a dog, canine’Cf., e.g., NIL:436f., 439 n. 14, 440 n. 25, Adams 2013a:190, 193.