Three Kinds of Cooperation
Ecologists speak about two types of cooperation — mutualism and commensalism — which distinguish whether both or just one of a pair is benefiting. I’d like to look at a different dimension of cooperation that has to do with communication. There are at least three different types of cooperation along this dimension, though perhaps you can distinguish more (if so, please post a comment!)
This is normally what we think of when we talk about cooperation: two (or more) agents using explicit communication to coordinate their behaviors. Examples are trivial to come by when thinking of human agents or computer technology. In non-human ecosystems, we see explicit cooperation too. Communication is achieved via basic tagging mechanisms such as pheromones, or more complex linguistic processes such as the famous bee dance.
Sometimes agents end up cooperating “by accident”. Meaning that their actions are determined individually without any explicit communication. But because of the particular state of the environment (including their configuration with one another), they each happen to benefit from the actions of the other(s). The tit-for-tat strategies in the iterated prisoner’s dilemma illustrate the concept. Past actions of one agent is noted and factored into future actions of another. There is no explicit communication, and each agent is (by design) looking out for their own best interests and nothing else. When a tit-for-tat agent interacts with a “narrowly selfish” agent, cooperation does not arise, but when two tit-for-tat agents interact it does. The TfT agent does not get to choose its strategy, rather the environment that it finds itself in that determines whether it will be in a cooperative situation or otherwise.
Physical and biological systems can be seen as exhibiting implicit cooperation as well. Any structure in physical space that is “built” from more basic parts can be viewed as having emerged via cooperation between constituent agents. Bricks cooperate (i.e. mutually benefit) from their physical proximity and particular configuration in a wall, compared to say, being randomly strewn around a construction site. In the random configuration, each brick is exposed on all sides to destructive forces, but in a wall they mutually protect one another. Additionally, bonds between two connected bricks mutually reinforce bonds with a third brick.
In social systems, implicit cooperation occurs all the time, though we tend to assume that all cooperation requires explicit communication. This fallacy helps explain the popularity of conspiracy theories.
A third form of cooperation can be illustrated by the analogy, “the enemy of my enemy is my friend.” In a population of agents where a subset (C) are cooperating with one another to their mutual benefit over the general (non-cooperating) populace, the independent agents can gain relative benefit by simply disrupting the cooperation of agents within C. This month’s New Yorker has a cover story in which a highly diverse (and normally antithetical) collection of dissident political parties are presenting a unified front in their opposition to Putin’s all powerful ruling party.
Oncologists note how cancerous cells exhibit extreme heterogeneity and competitive behaviors both with themselves and normal cells. But in a sense — the disruptive sense — they are cooperating with one another to lower the fitness of normal cells which are ultra-cooperative with one another. I have written several times here about the centrality of cooperation in the emergence of new agents. Tumors are new agents that emerge from the interactions of cancerous cells. Perhaps disruptive cooperation is a key factor. Tumors are different than “positive” agents in that they emerge from a substrate of high complexity, feed off of it (in a thermodynamic sense perhaps) and ultimately break down the complex system and hence the source of their own existence. Lest we dismiss this view out of hand, it should be noted that social/cultural agents exhibit this dynamic all the time.