CRGreathouse said:
What about near-altruists, who collectively punish defectors?
I think that's the key: you have to have punishers. But even then, those individuals are going to lose some fitness by punishing, and others (peaceniks, let's call them) have an incentive to free ride and get all the benefits of the punishers without any of the expenses of being one of them. How these types don't take over the population is kind of mysterious, and you get into some really esoteric stuff on population equilibria, which is really complicated and which I don't know enough about to explain, I'm afraid!
For more on this:
http://plato.stanford.edu/entries/game-evolutionary/
Yeah, self-sacrifice. It's easier if others do it ... It is hard to give up some privileges for the collective good ... We like to protect our immediate self-interest ... A Homo economicus would not sacrifice any thing that he/she values. Luckily, we don't live in a world populated by Homo economicus.
Maybe not a world populated by exact models of homo economicus, but we're closer than many think. Homo economicus has a great explanatory power. What's really cool are many of the studies in experimental economics, which is a sort of cross-discipline between psychology and economics (and maybe philosophy?). I kind of think that we would be much closer to homo economicus if we weren't so god awful at analyzing probabilities and so prone to cognitive biases!
This reminds me of the Prisoner's dilemma
because while it is comforting to think that we could live in a world of altruists ... but Game theory tells me that people will defect or cooperate in the beginning to defect at some point.
Well, the problem is that in a single iteration of the Prisoner's Dilemma, there is absolutely no incentive to cooperate. The Nash equilibrium is to defect, and that's the only rational move. Some people have tried to "solve" it, but all one really does is change the parameters of the game, and make it something other than a PD.
Look up iterated (or repeated) Prisoner's Dilemma games, though. Those get interesting, and there are a number of strategies that actually do allow for cooperation, such a "Tit-for-Tat," in which Player A will cooperate on the first move, and then (on her next move) do whatever B does. If B is also a "Tit-for-Tat" strategist, the game will work out nicely.
Also, look up some of the other games, like the Stag Hunt!
Sorry about all the Game Theory rambling, I just love the stuff so much!