Refining Phrygian diachronic phonology: The case of Phrygian (?) γάλλος ‘priest’

Sowa (2007: 157) casts doubt on the Diakonoff/Neroznak (1985: 109) derivation of alleged Phrygian γάλλος ‘castrated priest of Attis and Cybele’ from a PIE base *g2 hl3-los1 ‘cut short, shave’, chiefly on the ground that the analysis of the word as an *l-deverbative on this basis is “dubious”.2 Instead, Sowa tentatively suggests a connection with PIE *gelH- ‘gain power over’, as reconstructed in LIV2 (p. 185f.).3 Since an appeal to the *l-deverbative is the most obvious way of accounting for the medial geminate in both suggestions, which I shall refer to as [1] and [2], respectively, as well as being the most interesting from the point of view of my (2006) theory of Phrygian conditioned devoicing of mediae, my intention here is to assess the phonology of both suggestions in the light of that theory. Obviously for this purpose I am assuming the word is Phrygian.