Communitarianism and sacrifice

One definition of communitarianism is that the individual should in some sense sacrifice her individual good for the good of society. Thus, in opposition, individualism entails that the individual should not sacrifice her individual good for the good of society. If, for example, society needs more shoes, the communitarian would say that an individual should not pursue the occupation she prefers, but should make shoes; the individualist says that the individual should do what she prefers, and if society is left without enough shoes, too bad for society. It's more complicated than that, of course: if society wants more shoes, it can somehow raise the level of compensation so that the individual, because of that compensation, wants to make shoes. The difference is who must "give way" to the other: communitarianism says that the individual must "give way" to society; individualism says that society must "give way" to the individual.

There are two problems with this definition. First, there is no such actual, concrete thing as "society". Society is just a label we give to some abstract sense of the preferences, desires and opinions of a group of individuals. It's a useful abstraction, but both communitarianism and individualism according to the definitions above really can't avoid the fallacy of reification. More importantly, the word "sacrifice" is both ill-defined and its negative connotation subtly biases the discussion.

In one sense, sacrifice just means giving something up. In this sense, all mutually exclusive choices entail sacrifice. If I choose vanilla ice cream, I must sacrifice the chocolate; if I choose chocolate, I must sacrifice vanilla. Sacrifice in this sense is trivially unavoidable: until we can live forever and have everything we want, everyone must constantly make sacrifices.

Ayn Rand offers a better definition of sacrifice: A Randian sacrifice is giving up a greater value for a lesser value*. If I choose vanilla over chocolate because I prefer vanilla, I am giving up a lesser value for a greater, and thus not making a Randian sacrifice. Rand has an objective definition of value, so in her account someone can be fooled into making a Randian sacrifice; in a subjectivist account of value, a Randian sacrifice must be physically coerced**. But Rand's definition introduces a subtle paradox.

*I could cite this, but I'm lazy, and I don't think that Rand defines sacrifice this way is controversial.
**The proof of this statement (and the enthymeme necessary to make a Randian sacrifice logically possible) is left as an exercise for the reader.

Rather than look at the tension between the individual and "society", I think it's more illuminating to look at tensions between individuals; in this sense "society" becomes the collection of agreements (which become institutionalized in real societies) that individuals make. The Prisoner's Dilemma* illuminates the paradox.

*Chicken games are, I suspect, more common in practice, but the Prisoner's Dilemma illustrates the paradox more clearly, and I suspect that Chicken games can be reduced to win-win games.

Ex hypothesi, each individual values the choice to defect more than the choice to cooperate. Regardless of what the other agent chooses, each agent finds it better to defect than to cooperate. In a strict, "local" sense, the choice to cooperate entails a Randian sacrifice: mutual cooperation requires that both agents choose to abandon the greater value of defection for the lesser value of cooperation. And yet mutual cooperation is preferable to mutual defection: By mutually defecting, both agents are making a Randian sacrifice of the greater value of mutual cooperation for the lesser value of mutual defection. Where there is an apparent paradox, we are not thinking clearly.

We don't necessarily want to resolve all Prisoner's Dilemma games in favor of mutual cooperation*, but sometimes we do want to resolve them. The essence, then, of political philosophy is whether to resolve Prisoner's Dilemma games in favor of mutual cooperation, and if so, which ones to resolve and how to resolve them.

*as Yoram Bauman points out in The Cartoon Introduction to Economics, mutual defection in price competition, for example, can have positive social benefits.

Which brings us back around to pragmatism vs. moralism. The pragmatist looks at a Prisoner's Dilemma game, and after resolving all the games-within-games, decides that as best he can tell, mutual cooperation is better than mutual defection because the outcome is better. The problem then becomes to determine what social structures and institutions are necessary to achieve mutual cooperation.

There are two kinds of "moralists". The "individualist" moralist says that the local sacrifice is inherently immoral, regardless of the benefits of mutual cooperation. Better mutual defection than the sacrifice necessary for mutual cooperation. The "collectivist" moralist says that because mutual cooperation is better than mutual defection, cooperation is better than defection, regardless of what the other agent does. This philosophy just leads to victimization. (And it is, I think, this sort of asymmetric cooperation that dominates Rand's thinking and lends real weight to her critique.) Amazingly enough, those who most heavily promote the intrinsic moral value of cooperation seem to end up "forced" into defecting themselves.

Because of the fallacy of reification, naive definitions of individualism and collectivism lose coherence. Because there is no such real thing as "society", the question of whether the individual should or should not sacrifice his interests to those of society doesn't make sense. What makes sense is when individuals should sacrifice their interests to those of other individuals, and the fundamental distinction is between pragmatism and moralism. Pragmatism says that individuals should give up their local interests to others when their global, mutual interests outweigh the local losses, i.e. make a Randian sacrifice in the local sense to avoid a Randian sacrifice in a global sense. Moralism says that regardless of the mutual benefit, the structures necessary to ensure that outcome are never justified.