nak

The inscriptions where this particle is attested all come from Sardis. They are both poetic and in prose, and are found both at the cemetery and in the precinct of Artemis. All of them are approximately dated to the 5^th or 4^th century BCE, except for LW 80, which has been more narrowly dated to the 4^th century BCE.

[I.Y.]

indecl.	nak	LW 10.21 (Sardis)
		LW 13.5 (Sardis)
		LW 24.2 (Sardis)
		LW 24.17 (Sardis)
		LW 80.17 (Sardis)

The enclitic form =nak, despite its outward similarity with the tonic form, must be considered separately, pace Gusmani 1964a:169. See further under =ãk. The disjunctive (or perhaps conjunctive?) particle nak ‘whether / or’ (q.v.) must likewise be treated separately, despite the opinion of Gusmani 1964a:169.

[I.Y.]

This clause-initial particle is capable of appearing at the absolute beginning of the discourse (LW 24.2, 17), where it is best left without translation (Hawkins 2013a:184). The etymological considerations suggest that introducing a new discourse or discourse shift may indeed be its primary meaning, as opposed to the ubiquitous clause-initial ak. The occurrences of sentence-initial nak outside LW 24 are more difficult to classify, since they all occur in obscure contexts. The translation ‘so, thus’, adopted for some occurrences of nak in Sayce 1925a:34, 42 and Vetter 1959a:45, stems from reading too much into a clause-initial particle.

[I.Y.]

The particle nak can be most plausibly derived from the allegro reduction of the earlier *nu=ak, following (Carruba 1959a:38). This reconstructed combination appears to be structurally similar to the clause-initial complex fak < *fa=ak (see further under Lyd. fak). The Hittite morphophonemic change in clause initial particle chains, nu=aš → n=aš, nu=an → n=an etc., offers a formal parallel to the development postulated for Lydian, as also already seen in Carruba 1959a. The closest parallel to the reconstructed particle complex within Anatolian is Pal. nū=ku, which frequently appears in discourse-initial position. But the fusion *nu=ak > nak must have occurred after the separation of Lydian from the other Anatolian languages, unless one is willing to embrace a more complicated hypothesis that the inherited *nuk became nak under the analogical influence of the other clause-initial particles ak and fak.     

[I.Y.]

For the lemma head see Lyd. *nu.