σεμίδᾱλις

There is a general agreement that Ancient Greek σεμίδᾱλις ‘the finest wheaten flour’ is a loanword from the Semitic languages, see e.g. Akkadian samīdu ‘flour or groats’, although the exact way of transmission and the origin of the part °ᾱλις remained unexplained (for detailed references see Rosół 2013a:89f.; add now Hyllested 2020a:132). Two scholars saw, however, the fingerprint of Anatolian transmission in °ᾱλις (nevertheless, none of the overviews of Anatolian loanwords took into account this word [Yakubovich 2010d:113f., 146f., Yakubovich 2013a:118–120; Sh. Hawkins 2010a:224f.; Gasbarra & Pozza 2012a:189–196, Gasbarra & Pozza 2019a:57–70; Hajnal 2014a:110f., Hajnal 2018a:2041f.; Oreshko 2018a], not even the most comprehensive one [Simon 2018a]): Szemerényi 1974a:156 identified °ᾱλις as a Hittite suffix and Starke 1990a:445 n. 1601 derived σεμίδᾱλις from Luwian zammitāti- ‘flour or groats’ (see s.v.) with the regular Luwian rhotacism of the second dental. The assumption of a Hittite suffix is not very plausible, since fitting suffixes are missing (there is only -ala- and -al-, Hoffner & Melchert 2008a:54) and this word is not attested yet in Hittite (only in Akkadographic spelling as SÍ-IM-MI-TA-A-TI, HEG U-Z:730). Starke’s derivation from Luwian is weakened by the unexpected /e/ of the first syllable (although only κάμηλος can be compared phonologically [if it is indeed from Luwian kamara- as per Yakubovich 2016a:87 n. 30, cf. here s.v.] due to the lack of Luwian borrowings, cf. Simon 2018a:381) and contradicted by the initial consonant, since Greek sigma, contrary to earlier assumptions, does not transcribe Luwian ‹z› (Schürr 2017a:1–3, cf. also Simon 2019c:122).

However, due to the semantic and formal similarity it is hard to avoid the assumption that some type of connection existed between the Greek and the Luwian forms. As Simon 2009b and Sasseville 2021e argue, rhotacism was known in Lydian, too, thus, a transmission from Luwian or Hittite via Lydian into Greek is also possible (Lydian loanwords in Greek are well attested, see the most recent overview in Simon 2018a:381). This hypothesis can be supported by the unexplained shortening of the /i/ in σεμίδᾱλις (the Luwian form is hapax, thus its short ‹i› is not decisive), although the /ā/ remained long. Furthermore, also a Lydian toponym, Simidala (Zgusta 1984a:§1217-3), was connected by Neumann apud Zgusta 1984a:570 and apud Starke 1990a:445 n. 1601 with this word, although it remains open if ‘flour’ or ‘groats’ is semantically fitting for a toponym.

There are, however, difficulties with this derivation. The initial consonant and the first vowel still remain unexplained, unless the Lydian toponym indeed belongs here. If this is the case, then there is no problem with the Greek borrowing, “only” the origin of the Lydian /s/ and /i/ vis-à-vis Luwian ‹z› and /a/ remains unclear (the Lydian word cannot be an independent borrowing, since the rhotacism requires *d, which appeared in this position due to the Hurrian transmission, see s.v. zammitāti-).

All in all, an Anatolian, perhaps Lydian, transmission into Greek is currently the most plausible etymology, although the details remain unclear.

[Zs.S.]

Last updated: 19 February 2020