In this paper, I provide a general argument for the so-called X-First programme in the debate about normativity and defend X-First accounts against recent criticism. According to X-First accounts, there is a single normative primitive that is normatively fundamental and involved in explaining all other normative phenomena. Recently, Daniel Wodak (2020) has argued against X-First accounts by questioning the assumption that there is a single normative primitive. Wodak defends an alternative view, according to which normativity is the disjunction of multiple normative features. I argue that disjunctive accounts either force us to abandon large parts of our ambitions to understand normativity, or are themselves committed to the X-First methodology. It is argued that in the absence of local X-First accounts, disjunctive accounts fail to explain those parts of the structure of normativity which call for explanation, while their attempt to account for the unity of normativity is a postulate of what they set out to explain. These considerations, I argue, provide a positive argument for the X-First programme: We need to investigate X-First accounts in order to learn more about normativity.

In this talk, we offer a semantic diagnosis of “relativized ‘oughts’”, i.e. statements concerning what we, legally, morally, prudentially, etc., ought to do. According to this diagnosis, these are effectively equivalent to attitude ascriptions (“According to Creon, Antigone ought to…”). We propose to treat these ascriptions in the semantic framework of Gibbardian expressivism. In doing so, we offer a defense of the view that there is an “ought simpliciter” in addition to relativized “oughts” of legality, prudence, etc., against challenges from Baker (2018) and Coop (2007). Moreover, Wodak (2016) has argued that expressivists are particularly ill-placed to account for the relativization of “ought” to different standards. Our argument shows that this is not so.

Ranking theory is a well-established account of rational belief (= taking to be true) and its dynamics. It is, however, theoretically mandatory that a formal representation of epistemic states be extendable to a decision theory. Ultimately, epistemic states manifest themselves in rational action; each such representation should be able to say how. Alas, the decision-theoretic extension of ranking theory is still wanting. Giang & Shenoy (2000) is the only, but not well received proposal so far, for reasons to be explained. This talk will present a modification of this proposal. In analogy to standard decision theory, the upshot will be a theorem proving a unique representation of a preference relation over ranking-theoretic prospects (instead of lotteries) by an ‘expected utility’ function for those prospects (where expectation works in the ranking-theoretic, not in the familiar probabilistic way, whence the underlying utility function works like a ranking function too). What is this good for? One perhaps surprising, but instructive application is that this ranking-theoretic decision theory allows advancing the field of (conditional) deontic logic by providing the first substantial account of how (subjective) norms and beliefs interact. This is to show that this theory is a fruitful one deserving further development.

The view that one ought to do what one has most reason to do has served as a blueprint for a variety of more specific accounts of how reasons determine the deontic status of actions. I suggest a new such account with remarkable explanatory power. Over and above returning the intuitively correct results in a wide range of cases, the suggested account entails important deontic principles, including ones that axiomatise Standard Deontic Logic. These results can be turned into the core of an explanation of deontic principles in terms of facts about how reasons interact in determining oughts.