†maddun-

According to the traditional view, the Anatolian name Maddunāni consists of two Luwian words, maddu- ‘wine’ and nāni- ‘brother’ (see already Laroche 1959a:70; cf. s.vv. maddu-). However, this was rejected by van den Hout 2003a:302–306, since compound names with -nāni- are not spelled plene and proposed instead a segmentation Maddun-āni- and a meaning ‘Maeonian’. He believed to have found this word in the place name Maddunašša (as an adjectival derivation) and the personal name Madduwatta as well. The basis for this proposal was his derivation of the word *maddun- from *maiwon- (whence Μαιονία) / *mayun- via the Lydian sound law *y > d / V_V assuming that the sound law operated already before the 15^th c. This view has gained both supporters (Melchert 2008a:154 and Yakubovich 2010d:91f., 113f., though see his n. 18 against the case of Madduwatta on geographical grounds) and opponents (Weeden 2013a:81 n. 33 [without arguments] and Hawkins 2013a:8, 9, who rejects this derivation due the proposed early date of the sound law, which is, however, a priori not impossible).

Nevertheless, whatever the date of this sound law is (it has been argued that its early date is proven by Lydian mariwda ‘a theonym’ [see s.v.], but this is refuted by Schürr apud Gander 2015a:484 n. 191 on semantic grounds) and whatever the preferred segmentation of the above onomastic material is, as Gander 2015a:485f. rightly demonstrated, *maddun- cannot be derived from *mayun- since the quoted sound law leads to †madun-, i.e. to a word with a non-geminated lenis or voiced consonant, not to the attested geminated fortis or voiceless consonant. Note furthermore, first, that the spelling with geminates is entirely consistent, the only exception is KUB 57.114 r. col. 3’, which, however, cannot be used as a real argument, since the parallel manuscript (KUB 7.54 i 1) has Maddunāni. Second, the appearance of plene spelling in -nāni- cannot be treated as a problem, contra van den Hout, since this is exactly the expected orthography (for further philological arguments against the hypothesis of van den Hout see Gander 2015a:484-489). In other words, there is no reason to assume a separate word *maddun- and classify it as Proto-Lydian.

[Zs.S.]