On the Particle ταρ

In his book on Greek particles, Denniston claims that τε after an interrogative is always followed by ἄρ (ἄρα), as τ᾽ ἄρ. Anyone interested in this blog will probably recall line 8 of the Iliad: τίς τ᾽ ἄρ σφωε θεῶν ἔριδι ξυνέηκε μάχεσθαι. During a lecture on clitic placement in 1892, Wackernagel suggested that it should read instead: τίς τάρ σφωε θεῶν. The enclitic particle ταρ is attested in the Venetus A manuscript, and discussed in the LSJ as well as Chantraine's etymological dictionary. On page 150 of How to Kill a Dragon, 1995, Watkins agrees with Munro's opinion that "the ancient grammarian's ταρ is probably right".

He points to the Luvian language where there exists a locatival enclitic particle tar. The use of this particle bears a striking resemblance to Homeric Greek. He compares τίς ταρ with the Luvian usage following the indefinite relative pronoun: kuiš=tar. The Luvian particle is also used with a clause-initial finite verb, mammanna=tar, regard with favor! To this he compares verse-initial phrases in the Iliad (which appear as τ᾽ ἄρ in most editions): 11.254, ῥίγησέν ταρ ἔπειτα: 15.397, ὤιμωξέν ταρ ἔπειτα: 18.37, κώκυσέν ταρ ἔπειτα: 3.398, θάμβησέν ταρ ἔπειτα. Watkins points out the semantic connection between the verbs, shuddered, wailed, shrieked, was awestruck, and suggests that they are all developments of a single older formula. In the Odyssey only is it found with a positive sense, rejoiced, 13.353, γήθησέν ταρ ἔπειτα. He also considers the more common phrases αὐτὰρ ἔπειτα and αὐτὰρ ἐπεί, and concludes that αὐτὰρ should be thought of as αὐ + ταρ rather than αὐτ + ἄρ.

These phrases are not paralleled in other Indo-European languages, and it is probably not a coincidence that the Greek and Luvian cultures seem to have made contact on the western coast of Anatolia. It was there that the Luvian language was spoken, and Watkins himself has suggested previously that Luvian was the language of Wilusa or Ϝίλιος (or Troy as we know it today).