After working on translation of Greek and Roman plays over recent years, and on the history of translation into English of Greek tragedy and comedy for the last five, I have become aware how few translators pay, or ever have paid, much attention to the fact that the surviving classical canon, tragedy and comedy from Aeschylus to Terence (Seneca is a special case), was composed exclusively for the masked player. Not that this will make much difference to most contemporary directors who have no intention of confronting their audience with actors in masks, so that the whole idea of the mask becomes just one more arcane convention to ignore or sidestep. And for a modern audience, this is probably no bad thing. With a few notable exceptions, attempts in recent years to return to the mask in Greek tragedy have not been distinguished for demonstrating why an ancient actor would have found it virtually impossible to comprehend that anyone performing without a mask could even claim to be an ‘actor’. So, why should we expect translators to tackle what is regarded, if noticed at all, as a directorial matter? Part of the answer resides in the unresolved issue of how the actor in ancient Greece balanced the demands of a concentrated physical performance with a text which, in the written form we now have, is dense, complex and subtle: especially as some of the most active surviving mask traditions in Asia, where we might look for a parallel, disassociate the word from its physical presentation. What is seldom recognised by the classical scholar is the number of stage situations in Greek tragedy where the mask is a part of the classical dramatists’ inbuilt mise-en-scène. It is in language within such moments, I would suggest, that we can gain some additional understanding of what ancient playwrights expected from their players, and how that was cultivated.