The Difference Principle Jackson observes that we speak not only of individual actions, but also of group actions. Indeed, we can count any “mereological sum” of individual actions counts as a group action. He argues that we should recognize not only the wrongness of individual actions, but also the wrongness of group actions. (Note that Jackson is concerned with objective wrongness, which depends on what did or will in fact happen. This is to be distinguished from subjective wrongness, which depends on what probably happened or will happen. It is worth asking how his conclusions might change if he considered subjective wrongness.) Accepting that group actions can be wrong, Jackson thinks, the only way to save the difference principle: “the morality of an action depends on the difference it makes; it depends, that is, on the relationship between what would be the case were the act performed and what would be the case were the act not performed.” (This is, roughly, consequentialism. Not to be confused with Rawls’s difference principle.) Two counterexamples to the difference principle: 1. Over-determination case • X and Y actually harm Z. • If either had not acted, then the other, by acting alone, would have caused exactly same amount of harm to Z. According to the difference principle, however, neither did anything wrong. So, according to the difference principle, nothing wrong was done. 2. The beans example • A thousand villagers, each with a thousand beans. • Horizontal stealing: a thousand bandits each steal one bean from each villager. • Vertical stealing: each of a thousand bandits steals a thousand beans from a different villager. • Assume that it is no better, and perhaps worse, to end up with one bean. According to the difference principle, a bandit who steals horizontally does nothing wrong. If he did not steal horizontally, then the other 999 would have, so each villager would be left with one bean rather than none. By contrast, a bandit who steals vertically does something wrong. Parfit’s response: Parfit’s response: An action can also be wrong because, though it itself does not make a difference, it belongs to a collection of actions that does make a difference. Problem with Parfit’s response: Modified over-determination case: If X does not act, then the harm to Z is worse. Intuitively, and according to the difference principle, X does not act wrongly. But according to Parfit, X does act wrongly, because his act belongs to a collection of acts that together makes things worse. Reply on Parfit’s behalf: An act that collectively makes things worse is wrong unless it individually makes things better, and an act that collectively makes things better is good unless it individually makes things worse. Individual score overrides collective status except where individual score is zero. Jackson: “That looks like making zero unbelievably special.” Further evidence that the individual does no wrong • Suppose that the bandit has the choice of (i) horizontally stealing from the villagers or (ii) vertically stealing from an unrelated hermit. The bandit should clearly choose (i). • Suppose that the bandit has the choice of (i) vertically stealing from a villager or (ii) vertically stealing from the hermit. Doesn’t matter what the bandit does. This suggests that horizontally stealing is not as wrong as vertically stealing. Jackson’s solution: group actions can also be wrong Our mistake: • We judge that wrong is done in overdetermination case and the horizontal stealing case. • The difference principle implies that, in these cases, no individual does wrong. • We assume that only individuals can do wrong. • So we reject the difference principle. What we should think instead: • We should reject the assumption that only individuals can do wrong. Groups can also do wrong. • The difference principle implies that, in the overdetermination and horizontal stealing cases, the group does wrong. • So the difference principle implies that wrong is done in these cases. • So we have no reason to reject the difference principle. The difference principle also explains the intuition that horizontal and vertical stealing are morally equivalent. As group actions, they are morally equivalent. The basic explanation: Different options are available to groups than are available to individuals. This is why individual actions can be right, whereas the group action they compose is wrong. Objections Objection: This licenses X and Y to collude, in the knowledge that they will be individually doing nothing wrong. Reply: They will be doing something individually wrong in colluding: namely, causing a wrong group action to take place. Objection: The group action can be wrong only if its constituent individual actions are wrong. Reply: Everyone is driving 80, but everyone ought to be driving 60. Still, it would be wrong for any individual to drive 60, since the others will still be driving 80. No constituent individual action is wrong, even though the group action is wrong. Objection: What is the point of counting group actions as wrong? We can’t advise or blame groups (at least not the sort of groups that Jackson has in mind). Is saying that a group action is wrong any different from saying that a natural disaster is bad? Does this really address our worry about the individual/collective paradox? From last time: Are there “discrimination threshold” cases? Suppose I have two options. • If I “focus,” I turn the switch N times on a single victim. (Like vertical bean-stealing.) • If I “disperse,” I turn the switch once on each of N+1 different victims. (Like horizontal beanstealing.) Suppose that I am the only person facing this choice. It seems plausible that: 1. It would be wrong for me to focus. Now suppose that N of us face the same choice. It seems plausible that: 2. No matter how many others disperse, it would be wrong for me to focus. 3. If all of us disperse, the outcome is that N+1 victims get the switch turned to N. 4. If none of us disperses, the outcome is that N victims get the switch turned to N. 5. Therefore, if all of us disperse, we have an as-good-as-wrong effect. For the time being, assume aggregation: it is wrong to produce a larger sum of pain when one could have produced a smaller sum. First case: Additional pain from each turn is constant. • Then 1 is false. My dispersing produces a larger sum of pain, and so is wrong. • 1 seems true only because we find it psychologically difficult to aggregate many tiny effects and compare them to a single large effect. Second case: Additional pain from each turn increases with each turn. • 1 is true, because the first turn causes little pain. • But 2 is false because making (say) the Nth turn causes great pain. • However, now we have an absolute threshold case. Third case: Additional pain from each turn is zero! • How can the additional pain from each turn be zero, but the additional pain from N turns be more than zero? • Because one cannot perceive any difference between the pain from X turns and the pain from X+1 turns. And one pain cannot be worse than another if one cannot perceive any difference. • Paradoxical: No matter what we together do, we together could have produced less pain! This means that, no matter what we together do, we together act wrongly, or at least produce effects such that, if any individual produced these effects, he would act wrongly! Suppose 0 disperse, N focus. We would produce less pain if 1 dispersed, N-1 focused. Suppose 1 disperses, N-1 focus. We would produce less pain if 2 dispersed, N-2 focused…. Suppose N disperse, 0 focus. Then we would produce less pain if 0 dispersed, N focused!