/sann(a)i-^(di)/

The lexeme is attested 18 times in the HLuw. corpus, and the relevant inscriptions originate from the centre, east, south and far south of the HLuw. area. All attestations date to Iron Age Luwian between the 10th and the 8th century BCE.

[AH.B.]

1sg. pres. act.	/sanniwi/	(“SA4”)sa-ni-wa/i-i	MARAŞ 14, §12 (Maraş)
3sg. pres. act.	/sannidi/	(SA4)sá-ni-ti-i	BABYLON 1, §13 (Aleppo)
		(SA4)sa-n[i]-ti-i	TELL AHMAR 5, §16 (Tell Ahmar)
		[(SA4)]sa-ni-ti-i	ANCOZ 7, §12 (Commagene)
		(SA4)sá-ní-ti	KARKAMIŠ A11a, §23 (Karkamiš)
		(SA4)sá-ni-ti	KARKAMIŠ A11a, §24 (Karkamiš)
		(“SA4”)sa-ni-ti	KARKAMIŠ A18e, §5 (Karkamiš)
			KULULU 2, §5a (Tabal)
		(SA4)sa-ni-ti	BOYBEYPINARI 2, §12 (Commagene)
			ADIYAMAN 1, §5 (Commagene)
		(SA4)sa-˹ni-ti˺	ANKARA 1, §5
		(SA4)[sa]-ni-ti	KARKAMIŠ A 29d+i, l. 2 (Karkamiš)
		sa-ni-i-ti	ERKİLET 2, §2 (Tabal)
		sa-ni-ti	ERKİLET 1, §3 (Tabal)
		[sa?]-ni-ri+i	PALANGA, §9 (Malatya)
		SA4-ni-ti	MEHARDE, §7 (Hama)
3pl. pret. act.	/sannainta/	(SA4)sá-na-i-ta	KARKAMIŠ A1a, §4 (Karkamiš)
3pl. impv. act.	/sannantu/	(“*69”)sa-na-tu	ASSUR letter e, §19 (Assur)

One further attestation can probably be assumed for KARKAMIŠ A11b+c, §23 on a contextual basis, but the form is badly damaged with only 〈[...] ||-i-t[i]〉 remaining, evidently written across a line break.

[AH.B.]

In spite of its frequent occurrence, the meaning of this verb proved elusive for a number of years, until Laroche 1961a:34 connected the verb with Hittite šanna-/šann-^(ḫi) ‘to hide, conceal’; cf. Meriggi 1963a:2-3 for a brief overview of the suggestions made until then. Later, the interpretation ‘to overturn’ was put forward by Starke 1980b:80f. with n. 28 for the CLuw. verb, which is followed by Sasseville 2021a:226f. That the meaning assumed for CLuw. does not fully tally with the HLuw. evidence was already discussed in Hawkins 2000a:89. In HLuw., the verb mostly occurs in the protases of curse formulae, denoting a negative action performed on an installation or place for which the authors of the inscriptions invoke punishment by various gods. The meaning cannot be wholly negative, however, because it clearly has a positive effect for the author of KARKAMIŠ A1a, §4, who states *a-mu-pa-wa/i za-a-zi DEUS-ní-zi-i (SA₄)sá-na-i-ta, approx. /ammu=ba=wa zanzi massaninzi sannainta/ ‘These gods concealed (?) me’ in a context describing combat, resulting in protection for him. This contexts leads Rieken & Yakubovich 2010a:209 n. 10 to suggest tentatively a broader meaning such as ‘to remove’. However, none of the contexts force that interpretation. Rather, an examination of all the Luwian evidence strongly suggests that Laroche´s 1961a:34 interpretation ‘to cover; to hide, conceal’ is the most accurate one because it fits all of the HLuw. evidence as well as the one usable context in a CLuw. birth ritual, although it is still debated; cf. under lemma.

Outdated is Hawkins´ 2000a:548 attempt to separate 〈(“*69”)sa-na-tu〉 as attested in ASSUR letter e, §19 from the other forms.

[AH.B.]

The verb is a stem in /-(a)i-^(di)/, which is analyzed as a causative formation by Sasseville 2021a:226f.

[AH.B.]

Both attestations of the lexeme are found in the context of birth rituals, once in an etiological myth, once in the ritual instructions proper (in a MS copy).

[E.R.]

3pl. pret. ind. act.	ša-na-˹a˺-en-ta	KBo 29.26 l.col. 5’	(CTH 768, NS)
3pl. impv. act.	ša-an-na-i-in-du	KUB 35.103 iii 5	(CTH 766, MS)

Checked against the photographs in the Mainzer Fotoarchiv.

[E.R.]

The semantic interpretation of the verb must take into account its correspondence in Hieroglyphic-Luwian used to describe a destructive act on steles (see under Lemma) and possible cognates in Hittite. Within the Cuneiform Luwian corpus, only KUB 35.103 iii 1-7 provides contextual information about the meaning. The text describes the role of the Moon deities during child birth and afterwards. They are supposed to deliver the child, then to lift it, to perform šannai- (za-am-pa-ku-wa DUMU-ni-in wa-al-li-in-du ša-an-na-i-in-du) and finally to place it at the mother’s breast. Several interpretations of šannai-, the first three of which have already been advanced, make sense here:

            1) The child was lifted and overturned (in order to empty the lungs of amniotic fluid); see Starke 1980b:80 with n. 28, Sasseville 2021a:226f.

            2) The child was lifted and, thus, removed (from the body of the mother and the umbilical cord); considered in Rieken & Yakubovich 2010a:209 n. 10 for HLuw. /sann(a)i-^(di)/.

            3) The child was lifted and covered (with cloth); see Laroche 1961a:34f., considered also in Rieken & Yakubovich 2010a:209 n. 10 for HLuw. /sann(a)i-^(di)/; cf. Hitt. šanna-/šann-^(ḫi) ‘to hide, conceal’.

            4) The child was lifted and emptied (of any amniotic fluid remaining in the lungs); cf. Hitt. šannapilaḫḫ-^(ḫi) ‘to empty’ probably derived from the same root.

The asyndetic juxtaposition of wallindu šannaindu seems to imply that the two verbs describe a coherent action like those listed in 1), 2) and 4). The fact that the same action is performed on “steles in their places”, i.e. that they could be emptied by erasure for a damnatio memoriae, but left where they were, tips the scales in favor of solution 4) ‘to empty’.

[E.R.]

The 3pl. forms in -ainta and -aintu are ambiguous as to their stem class, but relevant information comes from the HLuw. cognate /sann(a)i-^(di)/. It makes it very likely that also the Cuneiform Luwian verb belongs to the class of the denominatives in -(a)i-^(di).

[E.R.]

For the possible lemma head, see Hitt. (and Pal.) šanna-/šann-^(ḫi).