In The Emotional Construction of Morals (2007), Jesse Prinz offers a coherent naturalist approach on the foundations of human moral normativity. He classifies his view as a sensibility theory, a realist approach conspicuously opposed to the anti-realist bent of traditional emotivism. I sympathize with Prinz’s general approach; nevertheless, I have some specific disagreements. The first concerns the “is-ought” problem. I don’t think it is possible to build a realist, and a naturalist, approach in ethics without answering the famous problem with a positive stance. That is, without showing that it is actually possible to make the suspect transition—at least if we take “normativity” as a natural fact (as Hume, I think, actually thought). Here I will evaluate what I think are the main issues of Prinz’s approach. One issue that deserves to be commented on is the contrast between Prinz’s view and John Searle's well-known approach regarding the is-ought problem. Prinz thinks that Searle’s argument favouring the passage from an "is" to an "ought" is not cogent, but the correct understanding of Searle's argument requires assuming “obligations” (and duties) as factual. Therefore, what a naturalist theory must explain is how it is possible to pass from the fact that one is under a particular obligation to the supposed required action (or, in more fashionable words, how duties can be “agent-relative reasons for action”). In effect, I will present my approach on the concept of “reason for action,” influenced by Judith Jarvis Thomson’s approach, and her distinction between “duties” and “oughts.” In the final section I will make some comments on the problem of relativism. I agree with Prinz that descriptive relativism is true; but Prinz is committed to a form of agent-relativism. I will present arguments against agent-relativism that depend on the acceptance of a clear distinction between “duty” and “ought.”