Abstract
Two approaches to the statistical analysis of texts are suggested, both based on the study of numerals occurrence in coherent texts. The first approach is related to the study of the frequency distribution of various leading digits of numerals occurring in the text. These frequencies are unequal: the digit 1 is strongly dominating; usually, the incidence of subsequent digits is monotonically decreasing. The frequencies of occurrence of the digit 1, as well as, to a lesser extent, the digits 2 and 3, are usually a characteristic author’s style feature, manifested in all (sufficiently long) texts of any author. This approach is convenient for testing whether a group of texts has common authorship: the latter is dubious if the frequency distributions are sufficiently different. The second approach is the extension of the first one and requires the study of the frequency distribution of numerals themselves (not their leading digits). The approach yields non-trivial information about the author, stylistic and genre peculiarities of the texts and is suited for the advanced discourse analysis. This paper deals with the application of the second approach to the literary texts in Turkish. We have analysed almost the whole corpus of works by are illustrated by examples of computer analysis of the literary texts by O. Pamuk and Y. Kemal – two of Turkey’s most prominent novelists. The hierarchical cluster analysis based on the occurrence of numerals in the texts by Pamuk and Kemal shows the author, genre, and chronology differences of numerals usage in the literary texts of these authors.
Highlights
The scope of this research relates to stylometry
The first approach is related to the study of the frequency distribution of various leading digits of numerals occurring in the text
The frequencies of occurrence of the digit 1, as well as, to a lesser extent, the digits 2 and 3, are usually a characteristic author's style feature, manifested in all texts of any author. This approach is convenient for testing whether a group of texts has common authorship: the latter is dubious if the frequency distributions are sufficiently different
Summary
The scope of this research relates to stylometry (statistical study of texts in order to find individual features of the author's style – in particular, for attribution of texts). [1] – often lead to contradictory results, and the very abundance of methods indicates a lack of reliability of each of them individually. In this case, the emergence of new stylometric techniques is not redundant, and they are all complementary rather than mutually exclusive. We have proposed the idea of studying numerals found in the text as a means of characterizing the author's style [2–6]. The use of numerals in the text is directly related to its authorship, style, and genre (see below)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have