Using some of the ideas provided by Nicholas Rescher and Charles Taylor, this paper tries to offer an answer to the question about the possibility of a rationality “of ends”. The first part of the paper examines the reasons given by Rescher’s pointing out the possibility and the convenience of an evaluation of preferences. Given Rescher’s formal conception of such an evaluation and his concept of a rational agent, I will show that his discourse ends in a dead-end. The second part intents to show how Taylor’s notions of an “strong evaluator”, “horizon of evaluations” and “selfinterpretation” could be used to prove that, after all, we ware not trapped in a dead-end. However this result implies that we must change Rescher’s ideas about the agent, and about the agent’s selfclarifying process. Finally, given the dialogic nature of this process of articulation -of what individuals consider the best and most valuable-, the paper tries to derive some consequences in order to answer to the initial question.