CapekDraCor database and some aspects of quantitative linguistic analysis of the Čapek brothers’ plays

This text is a methodological contribution to the possibilities of the quantitative analysis of plays. Specifically, we focus on the plays of the Čapek brothers and introduce here a new source of data for the linguistic analysis of their plays — the CapekDraCor database. Texts are segmented into the relevant sublayers: (1) text (character dialogues); (2) metatext and indexing of its type (situational or authorial comments, stage directions). The significance of adequate text segmentation and its impact on quantitative analysis is examined by comparing selected phenomena (proper names, keyword robot) in two corpora with different text processing — the Čapek corpus in the Czech National Corpus and the CapekDraCor database. The possibilities of quantitative analysis of plays are discussed, the literary character is treated as a central element. We describe the characteristics and specifics of plays as a genre with a multi-layered structure that oscillates between written and spoken language — this duality is further emphasized by the frequency distribution of word classes within the plays. We perform an illustrative keyword analysis complemented by a sentiment analysis of the text and the use of a specificity score in the plays The White Disease (Bílá nemoc) and R.U.R.