In this article, we argue that representation of science in informational networks is problematic. It causes debates and disagreements settled in the opposition between scholars and research management. On the one hand, it is impossible to imagine how research would proceed without presenting results in the form of publications and distributing information about scientific research through the Internet. On the other hand, this representation in the form of scientometrics over-empowers those who control how research is conducted. Scholars are forced to make use of the only means of scientific communication, which relates to journal impact factors and citation indexing. We look beyond the usual reasons used in these debates and offer original insight into this problem using an epistemological argument. From the epistemological view, we claim that representation of science in networks, just as representation in science, should incorporate variability. In this context, we devote our attention to altmetrics. Altmetrics give scholars the possibility to communicate and promote their work and allow various social and political actors use this representation of science. We examine this problem in the Russian context too.