Allomorphic Variability in the Middle Persian Continuants of the Old Iranian Suffix -ka-

Abstract

This paper investigates the Middle Persian continuants of a very widespread derivative suffix, OIr. -ka-, which is generally claimed to have had only one outcome in Middle Persian, namely -Vg, written ‹-k'›. As is known, this outcome gave raise to a number of different Middle Persian suffixes, for instance -ag, -āg, -īg etc., through the reanalysis of the preceding vowel as a part of the suffix. I wish to demonstrate that already in Middle Persian, and not just in New Persian, OIr. -ka- had many other minority outcomes that have never been recognised in the previous studies. I also wish to underline that these allotropes, not all of which are perceived, and function, as true suffixes in Middle Persian synchronically, are sometimes explicitly marked in the Pahlavi script, notwithstanding its well-known ambiguous and archaizing nature. Finally, I suggest that the presence already in Middle Persian of different outcomes of OIr. -ka- can partly depend on early borrowing processes among Middle Iranian dialects, and partly reflect different diachronic stages of the lenition process undergone by the OIr. voiceless velar plosive in internal and final postvocalic position.