Lexical encapsulation consists of a series of abstract unspecific nouns (fact, plan...) referring to predicative antecedents. This study is based on a corpus of parliamentary debate in English, Catalan and Spanish (PD) and deals with encapsulation as a complex lexical cohesion device which allows the addressor to evaluate the information of the debates in various ways. Our work tries to highlight the role of lexical encapsulation to reflect the addressor's positioning with respect to the topic discussed, and to establish whether there are outstanding cross-linguistic differences. In order to delimit the diverse evaluative strategies, an approach from the perspective of prototype theory is adopted, since there are fuzzy boundaries among the different evaluative operations observed. Firstly, the distinction between implicit and explicit evaluation is established. As for explicit evaluation, two different operations are described: objectivizing and subjectivizing encapsulation. The objectivizing one works as a subtle strategy of persuasion addressed mostly to the opponents, whereas the subjectivizing encapsulating structures point to rally the unconditional allies. The results show similar trends in the three languages due to the homogenizing potential of PD as a distinct genre of political discourse.