Erik Carlson

Whether or not intrinsic value is additively measurable is often thought to depend on the truth or falsity of G. E. Moore's principle of organic unities. I argue that the truth of this principle is, contrary to received opinion, compatible with additive measurement. However, there are other very plausible evaluative claims that are more difficult to combine with the additivity of intrinsic value. A plausible theory of the good should allow that there are certain kinds of states of affairs whose intrinsic value cannot be outweighed by any number of states of certain other, less valuable, kinds. Such``non-trade-off'' cannot reasonably be explained in terms of organic unities, and it can be reconciled with the additivity thesis only if we are prepared to give up some traditional claims about the nature of intrinsic value.

In value theory, the notion of an organic unity is usually associated with G. E. Moore. In his Principia Ethica, Moore provided two definitions of an organic unity. In the first section of this chapter, it is argued that both definitions fail to capture Moore’s intentions, as well as being inadequate for measurement-theoretical reasons. Section 15.2 briefly investigates whether the existence of organic unities, as conceived by Moore, entails that intrinsic or final value cannot be additively measured. In section 15.3, a number of definitions of an organic unity proposed by later authors are surveyed, and some criteria for organicity are discerned. In section 15.4, it is argued that one suggested criterion presupposes that some instances of final value are not instances of intrinsic value, in Moore’s sense. The fifth and final section contains a brief discussion of whether the existence of organic unities can plausibly be denied.

Several philosophers have argued that our cosmos is either purposely created by some rational being, or else just one among a vast number of actually existing cosmoi. According to John Leslie and Peter van Inwagen, the existence of a cosmos containing rational beings is analogous to drawing the winning straw among millions of straws. The best explanation in the latter case, they maintain, is that the drawing was either rigged by someone, or else many such lotteries have taken place. Arnold Zuboff claims that each person is justified in concluding that her existence did not depend on a particular sperm cell first reaching the egg. If it did so depend, her existence would be extremely improbable, and an incredible coincidence for her. Similarly, intelligent life would be an incredible coincidence for us, if this were the only actual cosmos. We reject both these purported analogies. Referring to the nonheredity of 'surprise value', we conclude that an evolutionary explanation of the presence of rational beings is sufficient; there is no further need to explain the basic features of our cosmos which make intelligent life possible. This point concerning surprise value also reveals a fundamental disanalogy between straw-drawing and cosmos creation.

Many philosophers have claimed that extensive or additive measurement is incompatible with the existence of "higher values", any amount of which is better than any amount of some other value. In this paper, it is shown that higher values can be incorporated in a non-standard model of extensive measurement, with values represented by sets of ordered pairs of real numbers, rather than by single reals. The suggested model is mathematically fairly simple, and it applies to structures including negative as well as positive values

In Consequentialism Reconsidered, Carlson strives to find a plausible formulation of the structural part of consequentialism. Key notions are analyzed, such as outcomes, alternatives and performability. Carlson argues that consequentialism should be understood as a maximizing rather than a satisficing theory, and as temporally neutral rather than future oriented. He also shows that certain moral theories cannot be reformulated as consequentialist theories. The relevant alternatives for an agent in a situation are taken to comprise all actions that they can perform in the situation. The defense of this idea necessitates certain modifications to the standard consequentialist criteria of obligatoriness, rightness and wrongness. The problem of whether agents should adapt their actions to their own future actions is also addressed. Further, a conditional analysis of performability is suggested, and it is argued that particular actions should in this connection be regarded as `abstract' rather than `concrete'. The final chapter sketches a consequentialist theory for collective agents.

A popular view of harming is the causal account (CA), on which harming is causing harm. CA has several attractive features. In particular, it appears well equipped to deal with the most important problems for its main competitor, the counterfactual comparative account (CCA). However, we argue that, despite its advantages, CA is ultimately an unacceptable theory of harming. Indeed, while CA avoids several counterexamples to CCA, it is vulnerable to close variants of some of the problems that beset CCA.