ܦܰܪܫܰܓܢܳܐ pršgnʾ paršaḡnā exemplar, copy
OP *patičagniya- (Hinz 1975, 186 has *patičagnya-); MP pa(č)čēn [pcgn'] copy (CPD 62); Arm. LW pačēn, pačean, pačik, patčēn copy, exemplar (Hübschmann AG 224, no. 517); cf. Benveniste 1934, 180–185, who suggests that the MP form is related to (Gathic) Av. cag- to give, grant, bestow (cf. AirWb 575 f.). — Bibl. Aram. pršgn; Hebr. ptšgn copy (Greenfield 1987, 257); Talm. Aram. pršgnʾ exemplar, example (Telegdi 253, 118; 221); Md. pʾršygnʾ paršegnā (Widengren 1960, 97; Nöldeke MG 41); JBA ptšgn, pršgn document, copy (Shaked 1987, 260). To explain the initial cluster in Talm. Aram. pršgnʾ Telegdi thinks that it is necessary to reconstruct a different and unattested Ir. form, with a different preverb (OP **fračayana-, or **paričayana-). This hypothesis is rejected by Widengren, who quotes other MIr. phenomena of allotropy of pt / pr / pč (e.g. Parth. prxwdn, Sogd. ptxwt, pcxwd). According to Widengren, in Middle Iranian there are different alternative forms of the word, namely patšaγn (the model of the Hebr. form ptšgn), patčaγn (from which Arm. patčēn derived) and paršaγn (confirmed by Syr., Bibl. Aram., Talm. Aram., and Md. forms). The relation among the three MIr. forms hypothesised by Widengren is not further explained, save by the generic mentioning of the fact that in some NW varieties of Middle Iranian the intervocalic and postvocalic change of δ to r took place. More simply (and more correctly, in my opinion), Nyberg (1974, 147) explains Hebr. ptšgn as coming from *patšagn < *pati-cagn- ← OIr. *patičagniya-; Syr. and Jud. Aram. paršagnā have an inorganic -r- < -t-. In fact, the change -t- > -r- (extremely frequent and normal in various languages) seems to have taken place in Middle Aramaic, and it does not seem possible to take it back to Iranian, where none of the attested forms has -r-. It is notable that some of the older Semitic forms maintain -t- (Hebr. and JBA), while Syr., Md. and Talm. Aram. have -r-. However, Telegdi underlines that the form pršgnʾ, being common to three Middle Aramaic varieties, cannot be an independent loanword, but must be a very old one in Aramaic. This is certainly true, but the change -t- > -r- attested in some of the Semitic forms, being phonologically natural and not attested in all the Semitic varieties, is not in my opinion meaningful enough for reconstructing a different and unattested form of the Iranian model ● Dt 17, 18; Ezr 4, 23; Joseph 8, 32; ES 1, 300A; Jul 70, 24; Bh gr 1, 202, 14 ◆ LS 609a–b; Lagarde GA 79, 198; Id. 1877 § 1838; Gershevitch 1969b, 182