|(It makes sense if you read the post)|
The argument from evil traditionally works with two categories of evil: (i) natural evil, i.e. evil that is brought about by the lawful unfolding of natural processes; and (ii) moral evil, i.e. evil that is brought about by the morally responsible conduct of free will-possessing agents. But what if there was a third category of evil, one lying between the natural and the moral? In other words, what if evil resulted neither from the lawful unfolding of natural process, nor from the morally responsible conduct of free will-possessing agents?
Ted Poston’s article “Social Evil” — which appeared a couple of years back in Oxford Studies in the Philosophy of Religion — argues that there is. It claims that there is such a thing as social evil, i.e. evil that results from the problems of collective action. A collective action problem arises when private individuals, each acting rationally from their own perspective, can bring about socially sub-optimal outcomes. Because the outcome in such a case is the result of collective action, it cannot be categorised as natural evil; and because no one individual is morally responsible for the outcome, it cannot be categorised as moral evil. Consequently, it is a distinct category of evil, one which must be considered independently of the traditional categories.
Or so the argument goes. Over the next couple of posts, I want to take a look at what Poston has to say about all this. I do so as part of ongoing mini-series on the disambiguation of evil, which is a trend I’ve observed in the literature on the problem of evil. I want to say at the outset that I started off being pretty sceptical of Poston’s claim to have identified a distinct category of evil. But I ended up persuaded that there is something unique about social evil. I hope to explain my initial scepticism, and the reason for my conversion, in what follows.
In the remainder of this post, we'll cover Poston’s initial characterisation and defence of social evil. This will require us to consider an example of the phenomenon, along with his responses to critics who claim that it is just another form of natural or moral evil.
(Interpretive Note: “Evil” is assumed throughout this post and the next to be any form of pain or suffering.)
1. So what is Social Evil anyway?
The phenomenon of social evil is best illustrated by reference to game theoretical examples. We can start with the simplest, and most famous, game theory puzzle of them all: the Prisoners’ Dilemma. You all know the set-up. Two accomplices, arrested on the same charge, both offered the same deal. If they confess and give-up their accomplice, they can go free while their accomplice spends ten years in jail. If they remain silent, and their accomplice gives them up, it’s the same thing only in reverse. If they both remain silent, they both get two years. And if they both confess, they both get five years. What should they each, individually, do? The four possible outcomes are illustrated in the diagram below (I won’t explain how to read this diagram now - see my posts on game theory for more details).
The classic game theoretical analysis is that, if you were one of the prisoners, you should confess. The reason is that confession strictly dominates remaining silent. That is to say: confessing yields a better outcome for you no matter what your accomplice does. It’s true, of course, that staying silent can yield a better outcome, but that’s contingent on what the other prisoner does. You can’t rely on him/her remaining silent. Therefore, you’re better off confessing. The only problem with this analysis is that your opponent will reason in the same way, and so we can expect the outcome of the game to be that both prisoners confess. Collectively speaking, this is worse than would have been the case if both remained silent. Hence, the collective outcome in this game is sub-optimal, despite the rational behaviour of each player.
The prisoners’ dilemma yields a modicum of evil. There is, after all, the pain and suffering of five years in prison. But, to be clear, the two-player prisoners’ dilemma does not quite capture the phenomenon of social evil. For that, we need to switch to massively multiplayer versions, wherein hundreds and thousands of players (maybe even millions) interact. The classic illustration of this is some form of the tragedy of the commons.
Poston uses the following illustration. Suppose the city of LA is undergoing a water shortage. As long as a significant proportion of the population reduces their water consumption — e.g. by not washing their cars, not watering their flowers, ensuring all taps are turned off etc. — the problem can be averted. Otherwise, a great deal of suffering and hardship will result. Structurally, the game is similar to that of the prisoners’ dilemma. There are many more players in this version, of course, but once again the collective outcome in such a scenario is likely to be sub-optimal.
The problem is as follows. Because a significant proportion of the population needs to reduce their water usage if the problem is to be averted, it is doubtful that it is rational for any individual to do so. For example, if I were one of the citizens, I might reason thusly: by not using any water I will not avoid the problem (because my individual action is not causally sufficient for avoiding the shortage); but by using it to water my flowers I can add to the aesthetic value in the world; therefore, it is better that I use the water than that I not use it. The problem, of course, is that every citizen can reason in roughly the same manner and thus the equilibrated in such a scenario is non-reduction in water usage. This leads to social evil, i.e. pain and suffering on a large scale.
To be clear, the water shortage example is just that: an example. It illustrates the logic underlying many collective action problems, from healthcare provision to overfishing to global warming. Governments frequently struggle to provide rules and regulations that alter the incentive structures in such “games”. Their hope is that this will make collective cooperation the rational choice, but the rules and regulations are not always effective. Poston’s point is that it is disconcerting — from a theistic perspective — that social interaction exhibits this feature.
Remember, Poston’s claim is that the outcome in the water shortage game is a distinct type of evil, one that is neither natural nor moral in nature. This is where my initial scepticism of his argument comes into play. When I first encountered the idea, I was convinced that social evil simply reduced to moral evil. If we are aware that the structure of the social dilemma is such that if everyone uses the water great suffering will come about, then we are obligated not to use it. To do otherwise would be to knowingly contribute to an evil end.
Poston seems to be able to respond to my scepticism, as we are about to see.
2. Is it really a distinct type of evil?
There are three initial concerns one might have about the novelty of Poston’s proposal. First, one might worry that social evil is really a sub-species of natural evil, and hence not distinct from that already-debated category. Second, and slightly more subtly, one might worry that the classic game-theoretic analysis of rational behaviour in such cases is flawed; that people are stupid not to cooperate. And third, one might worry that social evil is a sub-species of moral evil. This last worry, of course, is my own.
I won’t spend too much time on the first of these worries. The complaint, in essence, is that problems like the one exhibited in the water shortage game are brought about by the scarcity of natural resources, and this scarcity is a result of natural evil, nothing else. The obvious reply to this is that although scarcity is a feature of the natural world, the problems that result from it — particularly in the case of a renewable resource like water — are hugely exacerbated by failures in collective action.
The second concern gets into some technical aspects of rational choice and game theory. The complaint is that the decision to defect in these collective actions problems is a manifestation of human stupidity, which is a subtle form of natural evil. The response to this is complex. In essence, Poston argues that the complaint is driven by misleading intuitions about what the individual causes or brings about through their actions. Specifically, the complaint is driven by the belief that the individual’s choice somehow causes the sub-optimal outcome and is stupid for that reason. But this is a “gross misunderstanding” of the logic of the water shortage game. If, say, it takes somewhere between 300,000 and one million people to bring about the problem of water shortage, then one more or one less non-cooperator won’t make any difference. This is true for any case in which the threshold number needed to realise a particular outcome is vague. Thus, the individual’s decision not to cooperate makes no causal difference and is not irrational or stupid for that reason. Now, there are some technical rebuttals or alternatives to this that Poston considers in the article, but I’ll skip those because I agree with his take on it.
This brings us, at last, to the third complaint: that social evil is simply a sub-species of moral evil. This complaint relies on the notion that the individual’s choice to use the water in such a game is morally reprehensible. The easiest way to defend this claim would be to argue that the individual was knowingly and causally responsible for the negative outcome. But this defence is not available. As we just saw, the structure of the game is such that no individual choice is causally necessary or sufficient for the outcome. Hence, a different account of the moral link between the individual’s choice and the evil outcome is needed. Poston considers three possibilities.
The first relies on the Kantian principle of universalisation. According to Kant, it is morally wrong to act on a maxim of the will that is not universalisable. And since individual non-cooperation in the water shortage game is not universalisable, this allows us to make the following argument:
- (1) It is morally evil to act on a maxim of the will that is not universalisable.
- (2) Individual non-cooperation in the water shortage game is not universalisable.
- (3) Therefore, it is morally evil not to cooperate in the water shortage game.
(And, obviously, what applies to the water shortage game applies to all similarly-structured social dilemmas)
Premise (2) is clearly true. Although I may decide to water my flowers on the ground that this makes the world a more aesthetically beautiful place, I clearly could not will that everybody follow suit. If they did, we would have massive water shortage and much suffering would result. So the problems with the argument must lie in premise (1). And, sure enough, Poston is keen to rebut it. It is simply not true that non-universalisable actions are morally evil. For example, it is surely morally good (or at least permissible) that I decide to become a doctor. But it is also, surely, bad if everyone decides to do the same thing. Society requires some specialisation of labour in order to thrive. Thus, the universalisation case for moral evil is not particularly persuasive.
The second argument for moral evil relies on a rule-utilitarian principle of permissible behaviour. According to the rule utilitarian, the right act in any given decision problem is the one based on an optimific rule, i.e. the rule that yields the best outcomes overall. The claim would then be that universal cooperation is the optimific rule in this scenario. This gives us the following argument:
- (4) It is morally evil to act in a manner that is not endorsed by the optimific rule in any given decision problem.
- (5) Individual non-cooperation is not endorsed by the optimific rule in the water shortage game; universal cooperation is.
- (6) Therefore, it is morally evil not to cooperate in the water shortage game.
Whatever the merits of rule utilitarianism, this argument fails because premise (5) is false. One of the key features of multiplayer prisoners’ dilemmas is that there is no unique optimific rule, i.e. a rule that, if followed by everyone, yields the best overall outcome. No one individual makes a difference to the social outcome and, indeed, individual non-cooperation can make us better off overall.
That brings us to the third argument for moral evil, this one relying on another Kantian principle: the “ends-not-means” principle. According to this principle, it is morally evil to treat another autonomous human being as a means to some desirable end and not as an end in themselves. Maybe — just maybe — my decision to water my flowers treats everyone else as means to an end, not as ends in themselves:
- (7) It is morally evil to treat another as a means to an end, not as an end in themselves.
- (8) Individual non-cooperation in the water shortage game treats others as means not as ends.
- (9) Therefore, it is morally evil not to cooperate in the water shortage game.
Again, whatever the merits of the motivating principle, the second premise in this arguments looks circumspect. One can kinda see an argument in its favour: to justify my decision to water my flowers, I have to rely on the goodwill of a sufficient majority of others not to follow suit. Thus I am, in a sense, using them as a means to a more desirable end. But, as Poston points out, because of the nature of the game, my decision to do so does not actually interfere, in a problematic way, with their preferred ends, and is, in fact, consistent with viewing them as fully autonomous human beings. It is not at all like the case of the slave-master who coerces another into doing something that only benefits the slave-master. My decision to water my plants may benefit others, and does not involve coercion or manipulation of any kind.
Poston concludes this portion of his article by pointing out that, despite all the arguments just made, many people will persist in thinking that individual non-cooperation in these cases is morally evil. This is because they think there is some connection between the individual decision and the morally opprobrious outcome. My initial scepticism stemmed from this view. I thought that there was some “knowledge”-connection between the individual act and the social outcome that made that act morally questionable, perhaps on the grounds that they knew their defection raised the probability of the negative outcome. But as Poston points out:
This is a failure to realize the logic of a multiplayer prisoner’s dilemma. An individual doesn’t affect the social outcome. Further, an individual’s change of strategy doesn’t affect the probability that the social outcome is achieved. Removing one grain of sand from the beach doesn’t affect the probability that the beach has enough sand to make a sand castle.
To sum up, the traditional debate about evil and God has revolved around two distinct categories of evil: natural and moral. The claim being made by Poston is that there is a third category — social evil — which is neither the product of natural laws, nor the result of morally responsible individual action. As we have just seen, this claim looks pretty robust. Social evil doesn’t seem to be a sub-species of natural or moral evil.
If we accept that basic claim, the next topic to consider is whether social evil can be accounted for by the traditional responses to the problem of evil. We’ll be looking at that the next day.