=r

This particle is attested in various texts.

[D.S.]

indecl.	=r	KBo 4.11 obv. 27	(CTH 772, NS)
		KBo 4.11 rev. 48	(CTH 772, NS)
		KUB 25.39 iv 7	(CTH 773, NS)
		KUB 32.5, 3’	(CTH 759, NS)
		KUB 35.125 r.col. 6’	(CTH 766, NS)
		KUB 35.125 r.col. 7’	(CTH 766, NS)
		KUB 35.125 r.col. 8’	(CTH 766, NS)
		KUB 35.98 obv. 8’	(CTH 770, NS)
		KBo 39.179, 3’	(CTH 766, MS)
	⸢=r⸣	KBo 29.28 obv. 6	(CTH 765, MS?)
		KUB 35.133 iii 1	(CTH 665, LNS)

The word lammaur/n- (and lammaur) is no longer considered to have this particle attached at the end; see under lemmata.

[D.S.]

Checked against the photographs in the Mainzer Fotoarchiv.*

[D.S.]

Laroche 1959a:83 recognizes an enclitic particle =r of unclear function, attached to other enclitics or enclitic pronouns of the particle chain such as =mu=r, =du=r and =ti=r (thus also CLL:182).

Giusfredi 2014a:312f. offers a contextual analysis of the particle and suggests either a temporal meaning ‘then’ or a topic-marker. He notices for example that in a series of parallel clauses (KUB 35.125 r.col. 5’-8’), the particle is not used in the first clause, but appears in all subsequent clauses. Yakubovich 2022a:489 compares parallel ditransitive passages displaying a distribution such as kuiš=du=r vs. kuiš=tar, which would speak for a similar function of the particles =r and =tar. Indeed, it is very likely that the two locatival particles would have functionally merged at some point, just like the Hittite locatival particles did, leading to an elimination of some of them form the language. As Yakubovich (ibid.) himself concedes, however, evidence for their functional analysis is limited.

[D.S.]

Giusfredi 2014a:313f. suggests that the particle =tar is made of =tta and =r, which is synchronically ungrammatical, but in origin likely. Along the same line, Yakubovich 2022a:489, 493 argues that =tar and =r are the allomorphic realization of the same particle, the latter of which would occur solely after enclitic pronouns, whereas the other would not. This is indeed supported by several attestations, leaving out a few obscure and debatable ones. However, it still does not force the interpretation of synchronic allomorphy. If the particles had functionally merged, their distribution could simply reflect one in which =r would be preferred after =du, =mu and =ti for phonotactic reasons. One should acknowledge that =r would be difficult to pronounce following an enclitic pronoun ending in a consonant such as =aš, =an, and that is precisely where =tar is found instead.

Whereas Dunkel (LIPP II:661) rejects the reconstruction of a PIE particle *r̥, there is enough evidence for the existence of such a morpheme attached to other particles or pronominal stems (e.g. PIE *k^u̯ó-r ‘where’ and *tó-r ‘there’; see LIPP II: 466f. and 791f. respectively). Within Anatolian, one should also cf. Pal. =kuwar and Pal. =nuwar.

[D.S.]

*The forms in the table were checked together with Marian Wehrstein.