fak

The inscriptions where this particle is attested were found in Sardis, or in its vicinity. The most archaic of the datable inscriptions seems to be LW 54 (6^th), the most recent among them is LW 3 (reign of Alexander the Great). The majority of the inscriptions can be only approximately dated to the 6^th, 5^th or 4^th century BCE. The particle fak tends to be associated with curse formulae, usually located in the final parts of the inscriptions, but there are also instances of its use elsewhere.

[I.Y.]

As acknowledged already in Littmann 1916a:35, the particle fak always occurs in the absolute clause-initial position. Like other forms in such a position, it is capable of hosting chains of Wackernagel clitics. A non-trivial formal property of this particle is the impossibility of its co-occurrence in the same clause with the verbal prefix fa- Gusmani 1980a:53. This property perhaps arose by analogy with the clauses featuring fa.

The semantic discussion of fak in Gusmani Gusmani 1964a:117f. is inconclusive. Its crucial property, which seems to have gone unnoticed up to now, is its association with the imperative mood. The clauses where it occurs in better-understood inscriptions invariably contain wishes or commands. This explains the association of this particle with the apodoses of curse formulae, but there are also other clear examples, notably the placement of fak before the infinitive clause expressing a legal injunction in LW 22.4. While there is a residue of obscure or fragmentary contexts where the meaning of this particle cannot be established with certainty, none of them precludes the assumption that we are dealing with modal clauses. An additional argument in favor of the proposed semantic analysis is the likely genetic connection between Lyd. fak and the Luwian clause-initial particle pā, which likewise appears to have a modal value. Note that the Lydian language, unlike the majority of the other Anatolian languages, does not appear to distinguish between the present and imperative endings in the first person, so it is a priori likely that modality in such clauses could be expressed in Lydian by syntactic means. 

[I.Y]