Confusing the Modern Breakthrough: Naïve Bayes Classification of Authors and Works

Abstract

The Modern Breakthrough is widely considered to be one of the most important turning points in late nineteenth century Nordic literature, ushering in a period of literary experimentation predicated on a pivot toward naturalism. Georg Brandes’s iconic work, Det moderne gennembruds mænd (1883), provides a literary historical framework for the consideration of the movement, outlining in broad strokes the contours of this shift in literature and, in the portraits of a series of featured male authors, offering a touchstone for broader understanding of this movement. In 1983, Pil Dahlerup offered a corrective to Brandes’s work with Det moderne gennembruds kvinder. Here, Dahlerup surfaced the numerous female authors who were writing groundbreaking work in the shadows of the male-dominated literary world. A great deal of scholarship on the Modern Breakthrough considers the rich network of literary cross influence that characterized the period. Influence, however, is a complex phenomenon and one that is hard to formalize. In the following work, we propose to explore the related phenomenon of similarity, predicated on the notion that the most sincere form of flattery is imitation. To what extent do writers from this period share aspects of language? Can we capture this sharing in a useful manner computationally?