/kwi(ya)/, kwa/i-i
‘when’
- Language
- Hieroglyphic Luwian
- Grammar
- adv
- ID
- 2991
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/kwi(ya)/, kwa/i-i
‘when’1.1.1Transmission
This lexeme is attested 31 times in inscriptions from Tabal, Malatya, Karkamiš, Tell Ahmar, Commagene and Assur, dating from the 10th or early 9th to the late 8th century BCE.
[A.P.]
1.1.2Forms
| indecl. | /kwi/ | kwa/i | BOHÇA, §10 (Tabal) |
| IZGIN 1, §2 (Malatya) | |||
| İVRİZ frag. 3, l.2 (Tabal) | |||
| kwa/i-i | KARKAMIŠ A1a, §24(Karkamiš) | ||
| ANCOZ 5, §3 (Commagene) | |||
| SULTANHAN, §10 (Tabal) | |||
| BOYBEYPINARI 2, §7 (Commagene) | |||
| ASSUR letter e, §10 (Assur) | |||
| BOHÇA, §10 (Tabal) | |||
| TELL AHMAR 1, §10 (Tell Ahmar) | |||
| TELL AHMAR 1, §18 (Tell Ahmar) | |||
| TELL AHMAR 6, §25 (Tell Ahmar) | |||
| hwa/i-i | SULTANHAN, §4 (Tabal) | ||
| /kwiya/ | kwa/i-i-a | KIRŞEHİR, §4 (Tabal) | |
| kwa/i-i-ia | KARKAMIŠ A6, §8 (Karkamiš) | ||
| kwa/i-ia | TELL AHMAR 6, §27 (Tell Ahmar) | ||
| kwa/i-i-a | KARKAMIŠ A6, §18 (Karkamiš) | ||
| KARKAMIŠ A6, §23 (Karkamiš) | |||
| kwa/i-a | KARKAMIŠ A1a, §13(Karkamiš) | ||
| kwa/ix-i[...?] | TOPADA, §37 (Tabal) |
The above table offers a representative selection, omitting broken attestations.
Spellings with final /a/ have previously been transliterated as 〈-’〉, e.g. in KARKAMIŠ A6, §18 (see Hawkins 2000a:125), cf. the discussion of the phonological interpretation below.
[A.P.] [AH.B.]
1.1.3Graphic Features
1.1.3.1Phonological Interpretation
The phonological interpretation of this word is provisional, but the allomorphy appears to be real, i.e. a change between /kwiya/ and /kwi/.That the final /a/ is real was first indicated in the transliterations offered in Hawkins 1972a:92 and Hawkins 1980d:132, but later he changed them to 〈-’〉, implying that the sign was only used as a space filler, see also Oshiro 1988b:15 who suggests a cognate link with Hittite kuwat ‘because’. From the available data, it cannot be said which of the forms represents the older variant. On the one hand, the sequence /iya/ occasionally undergoes contraction to /i/ (see Melchert 2003b:183), on the other hand, it is possible that /ya/ represents a phonological enlargement for this very short grammatical word, but cf. the discussion of the stem below. That the first vowel must be /i/ becomes clear from the attested plene spellings 〈kwa/i-i(...)〉 listed under Forms above. It is unlikely that the spelling should be interpreted to mean /kwai(ya)/, although the vocalically underspecified first sign in principle permits such a reading.
[AH.B.]
1.1.4Meaning
The semantic interpretation of this adverb became clear quite late, not least because the allomorphy of the forms was not recognised as such, cf. e.g. Oshiro 1988b:12-14, 15, who assumes them to represent two different conjunctions. Still, the temporal meaning was clear to Oshiro and others, cf. e.g. Melchert 2003b:207. Only relatively recently did it become clear that /kwi/ and /kwiya/ are probably variants of the same word, see Hawkins 2006a:22, who points out that they mean the same. In fact, the semantics of the two variants can be observed in one and the same inscription, viz. in KARKAMIŠ A1a, §17: ARHA-pa-wa/i kwa/i-i PES-wa/i-i-ha-a, approx. /ahha=ba=wa kwi awiha/, translating as ‘When I came forth, (these gods all came forth with me)’. KARKAMIŠ A1a, §21, then, reads: ARHA-pa-wa/i kwa/i-a PES-wa/i-i-ha, approx. /ahha=ba=wa kwiya awiha/, again translating as ‘When I came forth, …’ (cf. Hawkins 2000a:88 and Oshiro 1988b:12).
/kwi(ya)/ is not the only adverbial conjunction with a temporal meaning in the HLuw. corpus. Hawkins 2000a:127 regards the change between /kwi(ya)/ and the semantically similar HLuw. /kwanza/ ‘while’ in KARKAMIŠ A6, §18 and §12 resp. as not self-explanatory. Still, even near-synonyms are not at all unusual in the realm of adverbial conjunctions from a cross-linguistic perspective, cf. e.g. English while and as, because new words are constantly recruited for this function by way of grammaticalisation.
[AH.B.]
1.1.5Stem
While it is clear that the word belongs to the family around the relative and interrogative pronoun HLuw. /kwi-/kwa-/ ‘who, which’, the precise formation is subject to speculation. The most straightforward assumption is to take it as a lexicalisation of the old nom./acc. pl. n. form of the pronoun, which would have been */kwiya/.The attested allomorph for the nom./acc. pl. n. of the pronoun is /kwaya/, but this has very likely been reformed on an analogical basis and according to the i-mutation pattern that mandates a stem ending in /-a-/ for neuters in the nom./acc. sg. and pl., just like the nom./acc. sg. n. /kwan=za/, previously */kwi/, cf. CLuw. kui. Subsequently, the form /kwiya/ was sporadically shortened to /kwi/.
[AH.B.]