The question of the existence of a deontic logic with which human beings operate in some domains must address the problem of the limited results of deontic abstract selection tasks. Recent research attempts to solve the problem by arguing that deontic abstract selection tasks have ambiguous instructions that are not easy to interpret, and by having participants work on new versions of these tasks with simple and less equivocal instructions. In this paper, I analyze the texts used in this research and the results obtained with them, and conclude that the revised approach does not really solve the problems of the initial deontic abstract tasks. I argue that the participants' answers should have been better in the new versions, and that there are pragmatic phenomena and phenomena related to textual comprehension that may have influenced the experimental results.
Conditional reasoning; Deontic logic; Pragmatics; Selection task; Textual comprehension