Computational Methods for Semantic Analysis of Historical Texts
This chapter presents an overview of the state of the art in the analysis of semantics phenomena in historical texts at scale, highlighting its critical aspects and proposing a new approach which joins together the expertise of computational specialists with that of humanities scholars. Semantic phenomena are grounded in linguistic, cognitive, social, cultural, and contextual factors, occur due to specific linguistic contexts or language-external events, and their dynamic mechanisms are often led by innovative language users and communities. Manual semantic annotation is very time-consuming and cannot usually be extended from one set of texts to another. The basic idea behind computational methods in historical semantics consists in building semantic spaces from text data to reflect the historical period of the corpus in question, with its conceptual and cultural frame of reference. Truly cutting-edge computational research in historical semantics should involve the development of innovative and impactful methods, which are built to answer questions relevant to humanists.