In her recent book, L. A. Paul presses a serious problem for normative decision theory (Transformative experience, Oxford University Press, Oxford, 2014). Normative decision theory seems to be inapplicable when the values of potential outcomes are unknown, or when our preferences may change as a result of our choice. Paul then offers a framework for overcoming these problems, known as the revelation approach. I argue that, contrary to what Paul suggests, this approach is unhelpful in the large class of cases where the decision at hand centrally concerns persons other than the decision-maker. Unless Paul’s account is supplemented in order to handle such cases, many decisions will remain rationally intractable.