In this article, we deal with similarity between epigenetic marks in the DNA and the so-called hapaxes in language. A grammar description based on hapax legomena is designed. We reflect hapax analysis of Czech language provided by Novotná (2013) and avoid random selection of the corpus. For this reason, we analyze the corpus of 12 authentic books from 12 authors who elaborated the theme “What’s new in…” concerning their field of science, assigned by Nová beseda publishing. By analyzing middle-sized corpus, we expected results similar to those in case of large-scale national corpus (see Novotná 2013). We chose to classify hapaxes into different categories in comparison to Novotná, yet the results show similar language productive categories. This kind of language potentiality seems to be analogical to epigenetic processes in biology, which is briefly introduced.
In this study, we aim to introduce the analytical method bag-of-words, which is mainly used as a tool for the analysis (document classification, authorship attribution and so on; e.g. [1, 2]) of natural languages. Quantitative linguistic methods similar to bag-of-words (e.g. Damerau–Levenshtein distance in the paper by Serva and Petroni ) have been used for the mapping of language evolution within the field of glottochronology. We attempt to apply this method in the field of biological taxonomy – on the Brassicaceae (Cruciferae) family. The subjects of our interest are well-known cultivated crops, which at first sight are morphologically very different and culturally perceived as objects of different interests (e.g. oil from oilseed rape, turnip as animal feed and cabbage as a side dish). Despite the phenotypic divergence of these crops, they are very closely related, which is not morphologically obvious at first sight. For this reason, we think that Brassicaceae crops are appropriate illustrative examples for introducing the method. For the analysis, we use genetic markers (internal transcribed spacer [ITS] and maturase K [matK]). Until now, the bag-of-words model has not been used for biological taxonomisation purposes; therefore, the results of the bagof-words analysis are compared with the existing very well-developed Brassica taxonomy. Our goal is to present a method that is suitable for language development reconstruction as well as possibly being usable for biological taxonomy purposes.