This paper works within a particular framework for reasoning about actions—sometimes known as the framework of “stit semantics”—originally due to Belnap and Perloff, based ultimately on the theory of indeterminism set out in Prior’s indeterministic tense logic, and developed in full detail by Belnap, Perloff, and Xu [3]. The issues I want to consider arise when certain normative, or decision theoretic, notions are introduced into this framework: here I will focus on the notion of a right action, and so on the formulation of act utilitarianism within this indeterministic setting. The problem is simply that there are two different, and conflicting, ways of defining this notion, both well-motivated, and both carrying intuitive weight. This problem was first pointed out in my [14], but here I address what I now think of as a mistake in that treatment. In that earlier book, in order to explain our conflicting judgments about right actions, I set out two substantially different accounts of the notion, which I labeled as the “dominance” and “orthodox” accounts. But here, there is only one account, only one theory of right actions, and our conflicting intuitions are instead explained by showing how this theory yields different results when actions are evaluated from different perspectives. In effect, a semantic explanation, which postulates an ambiguity in the notion of a right action, is replaced by a pragmatic explanation. The paper is structured as follows. In the next section, I review Prior’s indeterministic framework as well as the structures underlying stit semantics. Although these structures were originally introduced for the purpose of interpreting formal languages containing special modal operators—tense operators, agency operators—there is none of that here. The 1 concepts I am concerned with in this paper are defined entirely in terms of the underlying structures themselves; there is no need to introduce or interpret any formal language. In the third and fourth sections, I motivate the two ways of understanding the notion of a right action, and define the corresponding dominance and orthodox act utilitarian theories. Finally, in the fifth section, I show how these two theories can be unified, and how our conflicting intuitions about right actions can then be explained as resulting from the different perspectives from which actions might be evaluated...