The Theology of Everyday Life: Is Gossip a Sin?, Part 2, "Interlude: Evolutionary Stable Strategies"

Okay, in my next post I'm going to make this claim: Gossip makes the world a better place! Conversely, if you refuse to gossip, you're hurting everyone.

Wild claim, right?

But to make and defend that position, I'm going to need to use the idea of an evolutionary stable strategy (or ESS). And since you might not know what an ESS is, I'm going to get you up to speed on the concept.

Evolutionary stable strategies were first investigated when theoretical biologists began to use game theory to understand behavioral dynamics within a population of Darwinian individuals. (For an overview of game theory, you might want to read my posts on "Game Theory and the Kingdom of God." See the sidebar.)

I guess the best way to understand an ESS is to see the concept in its most famous application: The critique of group-selectionism as the vehicle of cooperation.

Theoretical biologists have long been intrigued by mutual cooperation. This is because theoretical biologists work with Darwinian agents (you and I call them "animals") who seek to maximize their own "fitness." Given this penchant for selfishness, how can cooperation emerge? If Darwinian competition is ruling the world, wouldn't we always be stabbing each other in the back? I mean, there is no honor among thieves, right?

So, biologists over time have attempted to create explanations and models for why cooperation could be adaptive. The current consensus is that kin selection and reciprocal altruism get cooperation off the ground in the animal kingdom. But before that consensus emerged, a different explanation for the adaptiveness of cooperation was in vogue: Group-selection. Nowadays, most biologists don't think group-selection works as a means to create cooperation. And the reason they don't buy group selection has to do with the research regarding evolutionary stable strategies.

See, for selection to work, it has to act on something. Generally, the focus of selective pressure falls on the individual (or the gene). Thus, natural selection sorts the fit individuals (or genes) from the unfit. In this individual/genetic selection, it does seem hard to create cooperation between competing individuals. Nature being red in tooth and claw and all...

But some early evolutionary thinkers had this idea: What if selection could act on the group level? Selecting fit from unfit groups? Imagine two tribes separated by a river. Tribe A is comprised of selfish, Darwinian individuals (call them "Hawks"). In contrast, Tribe B is comprised of cooperative individuals (call them "Doves"). Now imagine each tribe is placed under ecological strain (e.g., a harsh winter, famine). The Hawk group refuses to pull together and they collectively suffer. Thus, their population dwindles over time. But the Dove group does pull together. They help each other out. And, although they might suffer some population losses, they do better than the Hawks. Generations pass and, eventually, we see the Doves flourishing and the Hawks gone extinct.

It's a nice theory, isn't it? A strictly Darwinian process producing a peaceful, loving world. The nice guys finish first.

But there is a problem. The Dove strategy isn't evolutionarily stable. Let me explain.

When we zoom in past the group level to observe the individual Doves and Hawks we see them employing (highly simplified) behavioral strategies. For the Doves this might be captured by "always share." For the Hawks it might be "attack and take as much as you can." These individual strategies, if Doves always play with Doves and Hawks always play with Hawks, will produce the group-selection scenario outlined above: Doves, via their sharing, thrive where the Hawks, due to their mutual aggression, self-destruct.

Now here is where the idea of an evolutionary stable strategy comes in. A behavioral strategy is evolutionarily "stable" when it is immune to "invasion" from the other available strategies. We have two strategies in our scenario: Hawk and Dove. Hawk is evolutionarily stable, whereas Dove is not. Let me illustrate.

Imagine a Hawk manages to swim the river and take up residence in Dove Land. Every time the Hawk encounters a Dove, he dominates the exchange. Thus, as an individual the Hawk increases in fitness relative to his neighbor Doves. If this translates into survival and reproductive success, the Hawk fathers more Hawks. Eventually, Dove Land is run-over by Hawks. The Dove strategy cannot fend off this invasion.

In contrast, the Hawk strategy is evolutionarily stable. If a Dove swims the river and ends up in Hawk Land that Dove is in big trouble. It'll get killed. The Dove strategy cannot invade the Hawk Land.

Let us note that the Hawk strategy will not wholly displace the Doves. As Hawks increase in numbers their chances of encountering each other also increase. And when Hawk plays Hawk the outcome, due to their mutual aggression, causes both to fare poorly. Hawks really only do well when they can take advantage of a Dove. Thus, what will happen is that Hawks will increase in population up to a certain Hawk-to-Dove ratio (the exact ratio is determined by the relative payoffs of the encounters). But the Hawks will not take over completely. In the end, Dove Land will be comprised of both Hawks and Doves. That is not as bad as Hawk Land, but Dove Land exists no more.

The point of all this is that group-selection fails in that the strategy it is built upon (i.e, Dove) is not evolutionarily stable. In large groups of cooperating Doves, someone is, eventually, going to defect on the group and reap the positive consequences. This is known as the Free Rider Problem. And the Free Rider Problem just kills the whole notion of group-selection. Bubbles of cooperation can emerge among groups. But they are simply that, bubbles. They are fragile and are easily popped. From at Darwinian perspective, these bubbles (cooperative "golden eras") just don't remain stable enough over time for selection to operate upon them.

Oh well, it was a nice idea while it lasted.

So, there we are. Has anyone in blogland stayed with me? Found this remotely interesting? Well, if you did stay with me in my next post I promise to give you a payoff for your work in this post. Specifically, I'll use the concept of evolutionary stable strategies to suggest that gossip is a great way you can contribute to making the world a better place!