In the first part of the paper (sec. 1-4), I argue that Elqayam and Evan's (2011) distinction between normative and instrumental conceptions of cognitive rationality corresponds to deontological vs. teleological accounts in meta-ethics. I suggest that Elqayam and Evans' distinction be replaced by the distinction between a-priori intuition-based vs. a-posteriori success-based accounts of cognitive rationality. The value of cognitive success lies in its instrumental rationality for almost-all practical purposes. In the second part (sec. 5-7), I point out that the Elqayam and Evans's distinction between normative and instrumental rationality is coupled with a second distinction: between logically general vs. locally adaptive accounts of rationality. I argue that these are two independent distinctions that should be treated as independent dimensions. I also demonstrate that logically general systems of reasoning can be instrumentally justified. However, such systems can only be cognitively successful if they are paired with successful inductive reasoning, which is the area where the program of adaptive (ecological) rationality emerged, because there are no generally optimal inductive reasoning methods. I argue that the practical necessity of reasoning under changing environments constitutes a dilemma for ecological rationality, which I attempt to solve within a dual account of rationality.
Keywords: cognitive success; general vs. locally adaptive rationality; is and ought; means-end inference; normative accounts of rationality.