The digitization of vast corpora of texts is enabling quantitative, data-driven approaches to the study of literature and culture. Research in the Digital Humanities has tended to prioritize Modern English literature for several reasons, including the availability of high-quality natural language processing resources for English and the central position of Anglophone authors and English translations in literary criticism. In this talk, I will describe the development of computational text analysis techniques for premodern literary traditions such as Latin, ancient Greek, and Old English, focusing on two complementary approaches that enable deep diachronic profiling: quantification of aspects of writing style beyond word choice, and sequence alignment for mapping inexact verbal relationships across corpora. I will explore how these methods can be applied both to address longstanding questions in the humanities - including the development of Latin prose style across the Roman Republic and Empire, the compositional unity of Beowulf, and the “intertextual” influence of Vergil on later poets - and to characterize the cultural evolution of literature over long time scales.