The PA form suggests that this root could have prefixed class markers in PEC; however, they are absent in all other languages (even in Av.) and may be in this case a PA innovation.

    The precise nature of the initial PWC affricate here is hard to determine, since both in PAK and Ub. the initial clusters of the type 'affricate+fricative' are substantially restricted (in Ub. all affricates change to fricatives in this position; in PAK palatalized/non-palatalized affricates are neutralised before fricatives). Position within a cluster also explains the change of uvular to velar in PAK (no combinations of front affricates + uvular fricatives are allowed).