*šaptama-

Reconstructed from Anatolian personal names attested in Old Assyrian texts.

[Zs.S.]

Ša-áp-ta-ma-ni-kà	kt o/k 134, 5 (AKT 4.6)
	kt o/k 195a, 3
	kt o/k 195b, 2
	kt o/k 172a, 4.8
	kt o/k 172b, 5
	KTP 43, 10.15
Ša-áp-ta-ma-ni-˹kà˺	kt o/k 195a, 6
Ša-áp-ta-ma-‹ni-›ki	kt o/k 117, 4

The attestations follow Albayrak 2004a and Kloekhorst 2019b:176 n. 834 (cf. also Laroche 1966b:157 No. 1111; Zehnder 2010a:265, checking against the photos was possible only in case of AKT 4.6). Although the -a- preceding -nika could be interpreted as a so-called linking a, too (on this concept see Kloekhorst 2019b:80f.), due to the preserved *m (cf. under Origin) this consonant could not have been in final position, which implies that it must have been followed by (at least) this -a- (and thus Kloekhorst’s question mark in Kloekhorst 2019b:108 is not necessary).

[Zs.S.]

Šaptamanika is a compound name with -nika as its second member, identical to Hitt. nega- ‘sister’, and thus Neumann 1974b:279 suggested a meaning ‘seventh sister’ (followed cautiously by Eichner 1992a:85 and EDHIL:726, 756, Kloekhorst 2019b:163, 176f., 214 and hesitatingly by CHD Š:208 [“perhaps”]).

Zehnder 2010a:265f. emphasizes, however, that there is no assured independent evidence for Hitt. / Luw. *šaptama- ‘seventh’ due to the semantic problems of the alleged cognates (Hittite šaptaminzu [see s.v.], Cuneiform Luwian šappatammimma/i- [see s.v.], and Hittite šiptamiya- [CHD Š:400]) and thus Neumann’s interpretation cannot be verified, even if it is plausible.

Alternatively, Neu 1999a:251f. suggested analysing the name as *Šaptam-an-iga with *šaptam ‘seven’ and a suffix -iga as in the name ^fMuliga. Nevertheless, Zehnder 2010a:266 rightly argued that Neu did not explain the suffix “-an-” and that there is no evidence that “-iga” is a suffix in ^fMuliga. However, Neu referred also to Palaic divine names with the suffix “-iga” (ignored by Zehnder), and although the precise stem class of this suffix is debated (see the discussion in Starke 1990a:72–74, Sasseville s.v. Palaic wulašiniki-, and Simon forthcoming), the emergence of re-modelled, entirely a-stem forms remains possible. All in all, although Neu’s theory can be partly saved, the lack of explanation of the segment “an” still excludes his derivation.

Also HEG S:853, 1063 entertained a similar derivation claiming that this name “offensichtlich” belongs to the name ^mŠapta (Laroche 1966b:157 no. 1110), which would mean ‘seven’. However, HEG did not explain why and in which language the name ^mŠapta would mean ‘seven’ (note that the PIE form *septm does not lead to such a form in the Anatolian languages) and HEG did not explain the necessary derivation, either, and thus this theory can also be excluded.

In conclusion, while Zehnder is correct in emphasising the lack of direct evidence, the amount of cumulative evidence, i.e., the formal identity of *šaptama- with Indo-European words for ‘seventh’ (see under Origin), the plausible Hittite and Luwian cognates with meanings derived from ‘seven’, and the cross-linguistically well-known semantics, supports the analysis of Neumann.

[Zs.S.]

In view of the PIE form *septm and the Hittite forms šiptamiya- ‘seven-drink’ and šiptamae- ‘to septuple’, the vowel of the first syllable is unexpected and irregular in Hittite. Setting aside the ad hoc proposals of Eichner 1992a:85 (dialectal form or assimilation, with question mark), Neu 1999a:252 suggested Luwian origin, which would be phonologically regular (cf. also CHD Š:208 and HEG S:853, 1063). Nevertheless, it was not included in the overview of possible Luwian names in Old Assyrian transmission by Yakubovich 2010d:208–223.

However, Zehnder 2010a:266 rightly points out that if this is a Luwian form, then we have the problem of a hybrid compound name, since the second element -nika is certainly Hittite. For the same reason EDHIL:756 even claims that a Luwian origin is “unlikely” and suggests in Kloekhorst 2019b:177 that the word need not be Luwian, because it can be explained from Hittite: He assumes (EDHIL:726, 756, Kloekhorst 2019b:177 with n. 835) that the PIE word for ‘seventh’ was *sptmo-, i.e. a zero grade form, and that the regular result, *išptama- lost its initial *i- in analogy to šiptamiya- ‘seven-drink’ and šiptamae- ‘to septuple’.

Kloekhorst’s account is, however, demonstrably wrong in both assumptions. First, he expects zero grade in the ordinal on the basis of *pnk^w-tHo- ‘fifth’ (vs. *penk^we ‘five’) and *(s)uk̑s-to- ‘sixth’ (vs. swek̑s ‘six’), which is a non sequitur, since these forms are derived by other suffixes than *šaptama-. Besides, he ignores the identical derivation attested in Latin (septimus), Greek (ἕβδομος), and Vedic (saptamá-), which clearly demonstrate that the pre-form had e-grade. Second, the analogical loss of the initial *i- is impossible since analogy cannot result in phonologically illicit forms, i.e. here †šptama- (the parallels he quoted for this loss are also wrong, see the critical discussion in Simon 2019h:357 with ref. to Kloekhorst’s renewed and again ad hoc attempt to explain these forms after he has abandoned the former analysis).

Assuming that the meaning ‘seventh’ is correct, there are two possibilities to explain this form: First, modernizing Eichner’s attempt, it may be treated as a Kanišite Hittite form in contrast with the Ḫattuša Hittite forms. While recent research has identified several differences between both Hittite dialects, including phonological ones (see in general Starke 1993a:24 n. 12, Kloekhorst 2019b, esp. 233–245, and Simon 2019h), there is no independent evidence that one of these differences would have been the treatment of *e in this position. In other words, the assumption of a dialectal form remains ad hoc. Second, following the phonological rules, it can be identified as a word of Luwian origin. Not Luwian in strict sense, as a Luwian-Hittite compound would be indeed problematic as per Zehnder and Kloekhorst, but as a Luwian loanword in Hittite. Such an early Luwian loanword cannot be a problem, since Luwian loanwords are attested slightly later already in Old Hittite and Yakubovich 2010d:161–206 argued that Luwian and Hittite were in contact already in pre-historic times.

All in all, a Luwian loanword in Hittite provides a semantically, phonologically, and historically fitting solution.

[Zs.S.]