Elaborated Relations. On the usability of automatic identification of clause relationships in stylistics and literary history

Abstract

The paper aims to develop a method that allows an automatic identification of the different types of complex and compound sentences in the Hungarian written language through the analysis of conjunction words, relative pronouns and their positions in the sentence. This method opens up new perspectives in the stylometric research: on the one hand, conjunctions and pronouns as function words provide a large amount of data for statistical analyses, and on the other hand they also carry meaning – about the relationships between the clauses (eg. opposition, conditionality). By examining the relative frequency of each type, it becomes possible to identify the most characteristic relationship of the clauses in a given text or corpora. Thus, the style and poetics of the novels can be grasped at the scale of the sentence, while also revealing the topological-logical structure of the texts, which is not usually reflected during the reading process. The paper analyzes the frequency of each type of relation in a corpus of 100 Hungarian novels. This provides an opportunity to (1) register stylistic tendencies of the Hungarian novel in the period between 1832–2005); (2) examine which texts are the most typical of each type compared to other texts; (3) to visualize the ‘internal structure’ of a text, based on the relative proportions of types to each other; and (4) to distinguish different stylistic traditions.