The article presents the research into hedges and hedging strategy realization in evaluative academic genres viewed from a transdisciplinary perspective. To this end, 64 dissertation reviews written by Russian male and female scholars in humanities (linguistics, literature studies, history) and sciences (physics, chemistry, medicine) were analyzed. Hedging is understood as a system of linguistic means, which perform a mitigating function associated with expressing vagueness, uncertainty, skepticism and doubt. Perlocutionary effects of hedging are connected with reducing the illocutionary strength of the writer's claims or risks of opposition, face saving, positive and negative politeness. The following classification of hedges is proposed: self-mentions, mitigation hedges, hedged performatives, evaluations of propositions and vagueness hedges. The results of the comparative quantitative and qualitative analyses reveal that hedging strategy is determined by a number of discursive and personal factors such as sphere of communication, speech genre, gender and professional experience of the writer. In dissertation reviews hedges show high frequency of occurrence, which can be explained by the public nature of reviews, their potential traumatic effect on the reader, the existing requirements for scientific data presentation in a tentative and indeterminate way, as well as high academic competence of reviewers. The research proves that in humanities hedges demonstrate a higher frequency rate than in sciences. At the same time, male reviewers reveal higher hedging frequency than female reviewers.