prafrš(i)-

The key attestations of this noun form a figura etymologica in a decree found in the precinct of Artemis in Sardis (LW 23). For its dating cf. the discussion under afar(i)-. The other two attestations belong to an inscriptions found near the cemetery of Sardis and the lower Hermos valley.

[I.Y.]

nom. sg.	prafršiš	LW 23.7 (Sardis)
dat. sg.	prafrlλ	LW 23.7 (Sardis)
gen.-dat. pl.	prafraν	LW 47.2 (Middle Kaystros)
unclear	prafr[-	LW 44b.8 (Manisa)

The attribution of the last two forms to the paradigm listed above is likely on formal grounds but cannot be contextually corroborated due to the fragmentary character of the respective inscriptions. In the past, these two forms were assigned to a separate stem prafra- (Gusmani 1964a:85).

[I.Y.]

For the suffix alternation between the nominative and the dative cases, see Gérard 2005a:87, for the diachronic interpretation of the same phenomenon; cf. below.

[I.Y.]

The noun prafršiš constitutes the subject of a clause featuring the verb ‘to watch’ as a predicate and the pronominal object, which probably refers to the trade or transfer of property (cf. under afar(i)-). It stands to reason that the noun prafr- refers, at least in this case, to the proprietor or overseer of property, which agrees well with the likely etymological relationship between afar(i)- and prafr-, proposed in Carruba 1969a:67f., n. 21. I interpret the second stem as a possessive compound pr-afr-, literally “(responsible for) house-transactions” (cf. Lyd. pira- ‘house’ and typologically Greek οἰκονόμος ‘steward, manager’). The alternative semantic interpretation of the noun under discussion as ‘(member of the) religious union’, offered in Haas 1962a:172f., does not avail itself of its synchronic connection with afar(i)- and appears to be primarily based on its speculative comparison with Gk. φράτρα ‘religious union’.

The hypothesis that the combination prafršiš prafrlλ expresses reciprocity (Melchert 1997a:37, n. 7) seems problematic in view of another dative complement =m occurring in the same clause. Alternatively, one can propose that prafrlλ is used here in the adjunct of manner “prafrš(i)- will safeguard it for me as befits a prafrš(i)-” (Aleksandr Potemkin, pers. comm.). Another example of a Lydian dativus modi is attested in LW 24.10 niwislλ qelλ-k ‘in some impious fashion’

[I.Y.]

The consonantal stem prafr- must have originally been characterized by the nominative and dative endings *prafr-s and *prafr-l respectively, which were secondarily renewed with productive inflectional markers, yielding *prafrs-iš > *prafršiš and *prafrl-λ (cf. Yakubovich 2019c:308-309). The gen.-dat.pl. prafr-aν, if it belongs here, had a productive ending and therefore did not undergo any restructuring.

[I.Y.]

For the lemma head see Lyd. afar(i)-.