Historians ‘have long been aware that the texts they employ offer only partial glimpses of life as lived’ (Osborne 2011 : 25). Osborne's statement occurs in a plea to historians to be more aware of evidence from sources other than the written, illustrated by his consideration of what can be learnt from Classical Greek representations of the human body. He argues that categories of thought with which text-based historians might choose to operate, such as free or enslaved, Greek or barbarian, may be reframed or refocused by paying closer attention to visual media. In this book, I have shown another way in which the historian can supplement the direct evidence given by surviving ancient literature, documents and inscriptions, namely through an examination of the languages and idioms in which texts were written. Authors, scribes, stone-cutters and engravers all made choices when they committed a text to a written medium. Some decided to write in a particular language or dialect which was not the same as their native tongue, others chose to record bi- ormultilingual versions of the same message. Even monolingual speakers made conscious selections of lexical items, constructions or idioms. The linguistic choices made by ancient writers are rarely without a wider significance, and I have demonstrated how language use can be revealing both of the imperial ambitions of cities and states and of local attitudes towards empires. The vocabulary of a particular individual can shed light on his or her class, gender, age and ethnicity; it may reveal religious or political sympathies and antipathies.

Inscriptions, documents and literary works do not allow a 360-degree view of ancient life as lived, and neither do they give a complete picture of language as spoken. All the languages discussed in this book are ‘dead languages’, an expression that neatly encapsulates the truism that languages live in the minds and mouths of their speakers.