The Future in Tertiary Education Is Now: Diachronic Corpus Analysis, AI-Assisted Parsing, and Data-Driven Learning in Higher Education

Diachronic corpus construction has become, in recent years, a central activity in the Digital Humanities. The field now combines archival digitisation, computational annotation, and linguistic analysis within shared research infrastructures, and the boundaries between research output and pedagogical resource have grown correspondingly thin. The present paper reports on the pedagogical dimensions of the Athens Digital Glossa Chronos (AthDGC) project. AthDGC is a diachronic corpus of influential texts and their retranslations. It serves at once as a research resource and as a training environment for doctoral students in linguistics. Over four semesters, students worked with the PROIEL dependency grammar framework and the CoNLL-U format. Their manual annotations were set against the output of state-of-the-art neural parsers for historical Greek. The comparison turns out to be productive in itself. The analytical work of reconciling manual and automated annotations operates, we argue, as an effective pedagogical mechanism for developing diachronic literacy, that is, the capacity to interpret linguistic phenomena within their historical context using the methods of contemporary computational linguistics. We close by drawing out the implications of this model for curriculum design and, more broadly, for the role of corpus construction in training scholars at the intersection of linguistics, philology, and the Digital Humanities.