*ku̯o-/*ku̯i-/*ku̯e-
‘who, what, which’
- Language
- Proto-Anatolian
- Grammar
- pron
- ID
- 2889
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/kwi-/kwa-/, kwa/i-i-
‘who, what, which’1.1.1Transmission
This lexeme is attested across the entire corpus.
[D.S.]
1.1.2Forms
For the grammatical forms, cf. Oshiro 1983a:51-55. Only a representative number of examples are given in the table.
[D.S.]
1.1.3Graphic Features
The use of the logographic reading *329 as REL has been replaced by its syllabic value kwa/i- (Hawkins & Morpurgo Davies 1993a), unless it is really used as a logogram.
[D.S.]
1.1.3.1Phonological Interpretation
The sign does not distinguish the lenis from the fortis variant, but the consistent absence of rhotacism in attestations of the late period indicates that the alveolar in the dat./loc. sg. ending was fortis (Goedegebuure 2010b:83f.).
The instances of the relative pronoun spelled with hwa/i- (*508) instead of kwa/i- (*329) are attested in late inscriptions and indicate a late phonemic merger between prevocalic /Hw/ and /kw/ (Hawkins & Morpurgo Davies 1993a:59f.).
For the HLuw. nom./acc. sg. n. kwa/i-za, Kloekhorst (EDHIL:490) interprets /kwat=sa/ instead, although the enclitic particle after a dental stem is /=sa/ with loss of /-d/, i.e. *kwa/i-sa, would be expected; cf. stems in /-ahid-/. Therefore, a reshaping in /-anza/ after the nominal stems is preferred here (thus also Melchert 2003b:191).
[D.S.]
1.1.4Meaning
The lexeme /kwi-/kwa-/ can be used either as a relative or as an interrogative pronoun (Melchert 2003b:191). There are some instances with the reduplicated pronoun functioning as a free choice relative, e.g. /kwis kwis/ ‘whoever’. One instance of a free choice relative function kwa/i-i(a)-sa… i(a)-ma kwa/i-i(a)-sa ‘whosoever’ is found with the reinforcing particle /imma/ in KARAHÖYÜK, § 22. Additional literature on Luwian relative clauses is given under the CLuw. lemma below.
[D.S.]
1.2 Cuneiform Luwian
kwī-, kwatt-
‘who, what, which (relative / interrogative pronoun)’1.2.1Transmission
The word is attested in Luwian text passages from MS including rituals, incantations, songs, and festivals as well as a MS mythological text (CTH 764) and a MS letter (CTH 769).
[A.B.]
1.2.2Forms
| nom. sg. c. | ku-iš | KUB 35.107 obv. ii 2’ | (CTH 764, MS) |
| KUB 35.107 obv. ii 5’ | (CTH 764, MS) | ||
| KUB 35.139 obv. i 14’ | (CTH 773, MS) | ||
| KUB 35.139 obv. i 17’ | (CTH 773, MS) | ||
| KBo 14.121 rev. iii 4’ | (CTH 773, MS?) | ||
| KBo 14.121 rev. iii 7’ | (CTH 773, MS?) | ||
| KBo 13.260 obv. ii 12 | (CTH 766, NS) | ||
| KBo 13.260 obv. ii 14 | (CTH 766, NS) | ||
| KBo 13.260 obv. ii 16 | (CTH 766, NS) | ||
| KBo 13.260 obv. ii 8 | (CTH 766, NS) | ||
| KBo 13.260 obv. ii 20 | (CTH 766, NS) | ||
| KBo 13.260 obv. ii 22 | (CTH 766, NS) | ||
| KBo 22.254 rev. 9’ | (CTH 762, NS) | ||
| KUB 35.130 obv. i 1 | (CTH 770, NS) | ||
| KBo 22.254 rev. 9’ | (CTH 762, NS) | ||
| KUB 9.6 rev. iii 12’ | (CTH 759, LNS) | ||
| ⸢ku⸣-iš | KBo 13.260 obv. ii 10 | (CTH 766, NS) | |
| k[u-i]š | KBo 13.260 obv. ii 18 | (CTH 766, NS) | |
| ku-⸢iš⸣ | KBo 29.38 obv. 3’ | (CTH 769, MS?) | |
| KUB 9.6 obv. i 23 | (CTH 759, LNS) | ||
| KBo 29.13+ KUB 35.59 obv. ii 7’ | (CTH 760, NS) | ||
| KUB 35.107 obv. ii 1’ | (CTH 764, MS) | ||
| ⸢ku⸣-[iš? | KBo 14.121 rev. iii 12’ | (CTH 773, MS?) | |
| ku]-⸢iš⸣ | KBo 29.27 obv. i 4’ | (CTH 766, NS) | |
| k]u?-i[š? | KBo 29.38 obv. 1’ | (CTH 769, MS?) | |
| [ku]-i-š=a-an | KUB 35.14 obv. i 14’ | (CTH 760, NS) | |
| ⸢ku?-i?-š=a⸣-an | KUB 35.115 obv.? 1’ | (CTH 770, NS) | |
| ⸢ku⸣-i-š=a-an | KUB 35.115 obv.? 4’ | (CTH 770, NS) | |
| ku-i-iš | KBo 4.11 rev. 43 | (CTH 772, NS) | |
| KUB 35.139 obv. i 9’ | (CTH 773, MS) | ||
| KUB 35.139 obv. i 10’ | (CTH 773, MS) | ||
| [ku-]⸢i⸣-iš | KUB 35.14 obv. i 7’ | (CTH 760, NS) | |
| ku-⸢i⸣-iš | KUB 35.48 obv. ii 14’ | (CTH 760, NS) | |
| ku-i-⸢iš⸣ | KUB 35.45 obv. ii 21 | (CTH 760, MS) | |
| KUB 35.115 obv.? 2’ | (CTH 770, NS) | ||
| ku- ⸢i⸣-[iš? | KUB 35.115 obv.? 3’ | (CTH 770, NS) | |
| acc. sg. c. | ku-i-in | KBo 4.11 43 | (CTH 772, NS) |
| ku!-i-in | KBo 7.68 obv. ii 16’ | (CTH 764, NS) | |
| KBo 7.68 obv. ii 18’ | (CTH 764, NS) | ||
| ku-in | KUB 35.137 obv.? 7’ | (CTH 773, NS) | |
| nom.-acc. sg. n. | ku-i | KBo 29.61 obv.? 7’ | (CTH 754, NS) |
| KBo 29.61 obv.? 8’ | (CTH 754, NS) | ||
| ⸢ku⸣-i | KUB 35.133 obv. ii 3’ | (CTH 665, LNS) | |
| ku?-i | KBo 4.11 obv. r. e. 24 | (CTH 772, NS) | |
| dat.-loc. sg. | ku-wa-at-ti | HT 1 obv. ii 7’ | (CTH 394, NS) |
| KUB 9.31 obv. ii 31 | (CTH 757, NS) | ||
| KUB 35.54 obv. ii 3’ | (CTH 758, MS) | ||
| ABoT 2.241 r. col. 3 | (CTH 758, MS?) | ||
| ku-wa-at-⸢ti⸣ | KUB 35.54 obv. ii 2’ | (CTH 758, MS) | |
| nom. pl. c. | ku-i-in-zi | KBo 13.260 rev. iii 5’ | (CTH 766, NS) |
| KBo 29.25 rev.? iii 11’ | (CTH 764, MS) | ||
| ku-in-zi | KBo 22.254 rev. 11’ | (CTH 762, NS) | |
| KUB 9.31 obv. ii 23 | (CTH 757, NS) | ||
| KUB 9.31 obv. ii 24 | (CTH 757, NS) |
The attestations follow Laroche 1959a:55, 60 and CLL:118f. The transliterations have been checked against the photos of the Mainzer Fotoarchiv.
Two further attestations (ku-i-iš and ku-iš) are given in CLL:119 for KUB 32.5(+8) rev. iv 21’ that have not been included above for the following reasons: There is only one uncertain instance in line 3’ of the fragment KUB 32.5 and it is partially restored in Laroche 1959a:55 as [ku]-iš-(dur); whereas Starke 1985b:120 restored this passage as [tar-mi-i]š-du-ur. According to the photo of the Mainzer Fotoarchiv, the sign IŠ cannot be identified with certainty and the beginning of the word is completely broken away.
The form ku-wa-at-ti could either be a dat.-loc. (hence a stem kwatt- with the ending -i) or an abl.-instr. (hence a stem kwat- with the ending -ti), the only sufficiently attested context remains ambiguous, even with regard to a possible lexicalization of this form (cf. under Meaning). However, as shown by Goedegebuure 2010b:84 (in a comparison of function and written representation of Luwian pronouns), a dat.-loc.sg. is more likely for ku-wa-at-ti since the geminate -tt- written in this form rather represents a (probably voiceless) singleton stop. Based on this analysis, the stem of the relative pronoun is assumed to be kwī-, kwatt-.
For the grammaticalized adverb based on unattested abl.-instr. forms of kwī-, kwatt-, cf. under the Lemma kwāti(n), kwatī(n). For the indefinite pronoun formed with the particle -ḫa cf. the Lemma kwišḫa, kwiḫa.
[A.B.]
1.2.3Graphic Features
According to the photo of KBo 7.68 (also indicated in the hand copy) the forms of ku!-i-in in obv. ii show abnormal sign forms. In line 16’ there are traces of one and in line 18’ even of two additional wedges before the sign KU. These additional wedges, however, seem to have been partially erased, probably in order to maintain spacing between the words.
[A.B.]
1.2.4Meaning ‘who, what, which (relative / interrogative pronoun)’
The pronoun is used as relative and interrogative pronoun (cf. Laroche 1959a:55, CLL:117f., Melchert 2003b:191, HEG A–K:612, HED K:231, EDHIL:489, Yakubovich 2015b:15). For a recent overview on the syntax of relative clause in Anatolian languages cf. Melchert 2016f (with further lit. revising some of the views from the main contributions of Held 1957a and Garrett 1994a).
Regarding the relative pronoun, the function and status of the form ku-wa-at-ti is controversial. Due to the lack of further significant attestation, it depends only on the context attested in obv. ii 30–32 of KUB 9.31 (a ritual of Zarpiya from Kizzuwatna), the transliteration is based on the photo of the Mainzer Fotoarchiv, furthermore, the hand copy as well as the transliterations of Starke 1985b:53 and Görke 2015a have been taken account of (for suggested translations cf. below):
(30)u-ra-az dutu-az ta-ti-in-zi dingirMEŠ-in-zi (31) dé.a-aš-ḫa pár-na-an-za-ta ku-wa-at-ti an-da ḫu-u-i-na-i-ma-an (32)la-la-an-ti pa-a u-za-aš a-⸢da⸣-ri-ta-an
Oh Great Sun-god, fatherly gods and Ea, in the houses, inside which they take (something?) ḫūinai-ed, you shall eat.
In this context it remains unclear, to what constituent ku-wa-at-ti actually refers. Görke 2015a translated: “In den Häusern, von denen sie drinnen ḫūinaiman nehmen, sollt ihr euch speisen!” assuming an abl.-instr., cf. CLL:118 ‘from which’ (?) (presumably because of the ablatival meaning implied by the verb lalanti ‘they take’). But as noted under Forms, an abl.-instr. for this form is less probable (cf. Goedegebuure 2010b:84), than a dat.-loc.sg (now also accepted by Melchert, personal communication).
Appearing directly after pár-na-an-za, ku-wa-at-ti could introduce a relative clause providing further information to this noun. However, it does not agree in number with parnanza (the expected form would be kwatanza). This could either be a scribal error or a sign of an (ongoing?) lexicalization process involved with this form. Already Schwartz 1938b:343, 352 translated “(…) Eat ye what-has-been-brought-hither” and considered ku-wa-at-ti to be “perhaps an adverbial” (cf. similarly, Laroche 1959a:60 assuming an adverb of place ‘ou, là’, at least in KUB 9.31; cf. also Carruba 2002b:102 “wenn” or HED K:231 ‘where, whence’)
Differently, Goedegebuure 2010b:86 (assuming that parnanza belongs to the relative clause and appears as fronted constituent before the relative pronoun, cf. Melchert 2003b:207f.) translates: “Oh Great Sun-god, fatherly gods and Ea, for whom do they take (that which is) anda ḫūinai-ed to the houses (=temples?)? Now, feed yourselves!” However, the meaning of “(that which is) anda ḫūinai-ed” is obscure since the meaning of the assumed participle is also unknown.
A translation of this passage remains uncertain; in theory, the form kuwatti could either be interpreted here as a regular dat.-loc.sg. or as being in a transitional state of lexicalization where the paradigmatic form is still used but full grammatical agreement is not necessary anymore.
[A.B.]
1.3 Lycian A
ti-
‘who, which’1.3.1Transmission
This lexeme is well attested across the Lycian corpus.
[D.S.]
1.3.2Forms
As correctly pointed out by Houwink ten Cate 1961a:70, the relative pronoun is regularly placed after the verb in Lycian A. He also takes the enclitic =ti found in the sentence introductory particle chains (e.g. m(e)=e=ti) as a form of the relative pronoun. However, Laroche 1958b:169 compares =ti with Hitt. =za and assigns instead a reflexive function to =ti in this latter position (followed by Melchert in DLL:65, Gehrisch 2018a, Christiansen 2020a, Melchert 2021c). As correctly explained in Neumann 2007a:356, there are two homonyms =ti. The first one is the reflexive particle, which occurs regularly before the verb in the clitic chain, whereas the relative pronoun =ti (nom. sg. c., nom./acc. sg. n.) stands after the verb (e.g. statti=ti) or the predicative noun (e.g. ijãna=tija). Further, he notices that there are some cases of the relative pronouns, which are fronted. What he does not specify is that these cases mark the protasis of conditional sentences, e.g. kbi : tike : ti=ñtepi=tadi : atlahi : tibe : kbijehi tibe=te : ala[h]adi=ti (TL 101.3-4) ‘Whoever places anyone else in addition, either one of his own or someone else’s, or whoever assigns…’ (see also Melchert 2021c). The conditional use of ti=… (nom. sg.) ‘whoever’ is used twice with a verb in the plural instead of expected singular (TL 111.2, TL 139.3). This is simply due to a constructio ad sensum, which can even occur with “no one”For an example in Hieroglyphic Luwian, see Sasseville 2021a:469f.. An overview of the older literature is given in Neumann 2007a:356-358 and a recent analysis of the various types of relative clauses is offered by Melchert 2021c.
Borchhardt & Eichner 1997-1999a:62f. (and Eichner 2017a:285f.) and Kloekhorst 2011a:14f. do not accept the interpretation of =ti as reflexive pronoun (preferring the relative pronoun view), because it is not used regularly (sometimes omitted) with the verb prñnawa-(ti) ‘to build a grave-house’, which is no basis for rejection. To take the =ti in e.g. m(e)=e=ti as a relative pronoun would go against the syntactic rules and, therefore, it should be excluded; see Melchert 2021c for a more detailed rejection. Moreover, the existence of a particle =ti, other than the relative pronoun, is assured by the sentence tibe=ti=p(e)=ewẽ pijeti=tdi : e[ … ] (N 324.27) ‘or to whom he gives for his own benefit …’, in which there is already the relative pronoun =tdi.Pace Melchert 2021c:73 n. 48, a reflexive particle used with a verb of giving is not problematic at all.
[D.S.]
1.4 Lycian B
ki-
‘who, which’1.4.1Transmission
The relative pronoun is found in both Lycian B inscriptions.
[D.S.]
1.4.2Forms
| nom. sg. c. | ki | TL 44c.56 (Xanthos) |
| TL 44d.9 (Xanthos) | ||
| TL 44d.22 (Xanthos) | ||
| TL 44d.51 (Xanthos) | ||
| TL 44d.69 (Xanthos) | ||
| acc. sg. c. | ki | TL 55.5 (Antiphellos) |
| dat. sg. | kdi | TL 55.8 (Antiphellos) |
| nom. pl. c. | ki | TL 55.7 (Antiphellos) |
As Melchert (DLL:117), Schürr 2005a:140 and Neumann 2007a:166 assumes, Lycian B ki- is a relative pronoun (contra Shevoroshkin 2000a:205, Shevoroshkin 2004a:515). The association between Lycian A tibe and Lycian B kibe shows the outcome of Lycian B *ku̯ before front vowels. Pace Schürr 2005a:140, the instance of kibe=uwe=neu (TL 44c.58) is more likely to be the “disjunctive” kibe rather than the relative pronoun followed by the particle =be. The relative pronoun in Lycian B regularly precedes the verb or the predicate noun. The instance of kdi (TL 55.8) is however syntactically unclear. But since we are dealing with poetical texts (adapted to fit the meter), the word order should be taken lightly. Furthermore, there is in TL 44c.56 one example of case attraction involving the relative pronoun; see Lyc. B wis(e)i-(di).
[D.S.]