I think your suggested way of handling this (with the transcriber-defined shape indicators for both occurences) is brilliant!

Cindi