With recent advancements in systems engineering and artificial intelligence, autonomous agents are increasingly being called upon to execute tasks that have normative relevance. These are tasks that directly—and potentially adversely—affect human well-being and demand of the agent a degree of normative-sensitivity and -compliance. Such norms and normative principles are typically of a social, legal, ethical, empathetic, or cultural (‘SLEEC’) nature. Whereas norms of this type are often framed in the abstract, or as high-level principles, addressing normative concerns in concrete applications (...) of autonomous agents requires the refinement of normative principles into explicitly formulated practical rules. This paper develops a process for deriving specification rules from a set of high-level norms, thereby bridging the gap between normative principles and operational practice. This enables autonomous agents to select and execute the most normatively favourable action in the intended context premised on a range of underlying relevant normative principles. In the translation and reduction of normative principles to SLEEC rules, we present an iterative process that uncovers normative principles, addresses SLEEC concerns, identifies and resolves SLEEC conflicts, and generates both preliminary and complex normatively-relevant rules, thereby guiding the development of autonomous agents and better positioning them as normatively SLEEC-sensitive or SLEEC-compliant. (shrink)

In his recent book The Dimensions of Consequentialism, Martin Peterson defends, amongst other things, the claim that moral rightness and wrongness come in degrees and that, therefore, the standard view that an act’s being morally right or wrong is a one-off matter ought to be rejected. An ethical theory not built around a gradualist conception of moral rightness and wrongness is, according to Peterson, unable to account adequately for the phenomenon of moral conflicts. I argue in this paper that Peterson’s (...) defence of this claim is not convincing. Over and above this negative result, a careful assessment of Peterson’s case for degrees of rightness reveals that the theoretical corridor for accounting for moral conflicts without a gradualist conception of rightness and wrongness is relatively narrow. As I show, the only way of avoiding the conclusion of Peterson’s argument is to reject his conception of the ‘final analysis’ that an ethical theory provides, i.e. of what the theory ultimately has to say about individual acts and their normative properties. According to Peterson, such a final analysis should be seen as comprising the all-things-considered judgements yielded by the theory, and nothing else. As it turns out, the only alternative to this account that is compatible with the standard view about moral rightness and wrongness is to conceive of the final analysis as also containing judgements about morally relevant factors, or aspects, and the way in which they are normatively relevant. (shrink)

This paper argues against Martin Peterson in favour of the ‘standard view’ of rightness, according to which rightness does not come in degrees. It begins with a defence of the standard view against the charge that it is committed to ‘deontic leaps’. It goes on to claim that greater conceptual parsimony would allow Peterson to avoid certain problems involving equality and related matters that arise out of his conception of moral value, and that Peterson should take the same instrumentalist attitude (...) towards the norms of practical rationality as he does towards the norms of common-sense morality. The paper closes with some doubts about Peterson’s programme of consequentialization and its alleged advantages. (shrink)

In his recent book, The Dimensions of Consequentialism, Martin Peterson puts forward a new version of consequentialism that he dubs ‘multidimensional consequentialism’. The defining thesis of the new theory is that there are irreducible moral aspects that jointly determine the deontic status of an act. In defending his particular version of multidimensional consequentialism, Peterson advocates the thesis—he calls it DEGREE—that if two or more moral aspects clash, the act under consideration is right to some non-extreme degree. This goes against the (...) orthodoxy according to which—Peterson calls this RESOLUTION—each act is always either entirely right or entirely wrong. The argument against RESOLUTION appeals to the existence of so-called deontic leaps: the idea is that endorsing RESOLUTION would not give each relevant moral aspect its due in the final analysis. Our paper argues that, contrary to Peterson, all moral aspects remain visible in what can properly be called the final analysis of a moral theory that involves RESOLUTION, moral aspects do not have to remain visible in judgements of all-things-considered rightness or wrongness, respectively, introduction of what Peterson calls verdictive reasons does not change the overall picture in favour of DEGREE. We conclude that multi-dimensional consequentialists should accept RESOLUTION rather than DEGREE. (shrink)