Ongoing debates among historians of early modern philosophy are concerned with how to best understand the context of historical works and authors. Current methods usually rely on qualitative assessments made by the historians themselves and do not define constraints that can be used to profile a given context in more quantitative terms. In this paper, we present a computational method that can be used to parse a large corpus of works based on their linguistic features, alongside some preliminary information that can be retrieved from the associated metadata. The goal of the method is to use the available information about the corpus to create broad groups that can work as sub-contexts for better understanding different sorts of works and authors. In turn, this makes it possible to better profile each group and identify its most distinguishing linguistic features. Once these features are clarified, it will eventually become possible to also identify what the most representative works and authors in each group are and which of them may be worth exploring in greater detail. This classification method thus allows historians to integrate their qualitative assessments with quantitative studies in order to better define the relevant context for any given work.