The transformation of scientific genres in the context of the general digitalization of modern culture is considered. It is shown that the speech genre content of
this process is based on the mechanisms of generation and transformation of the
text of two types, the interpretation of which can be useful in order to better understand the nature, tasks and tools of scientometry at this stage, and in order to
better understand the speech genre structure of scientific speech. Firstly, the
structural requirements for articles and monographs indexed in scientometric
systems (Scopus, WoS, DOAJ, RSCI, etc.) are approved and streamlined, and
thereby our knowledge of what an article is from its structure (i.e. knowledge
about the genre of the article). Secondly, the requirements of indexing systems
lead to the fact that the texts of articles change, they are “written differently”,
and sometimes redone after appropriate recommendations from publishers. The
points highlighted in scientometric systems can be understood as signs that an article must comply with in order to be assigned to the “speech genre of a scientific
article”. The largest quantitative indicators for these items are indicators of how
close to the core of the genre this or that text will turn out.