Page 1 of 1

Computer simulation of group selection and regality

Posted: Tue 2024-01-23 7:20:42
by agner
Group selection theory postulates that natural selection can act not only on individual organisms, but also on groups of animals or humans. Entire groups can go extinct if they are unfit, or a group can produce new daughter groups if the group as such has high fitness. This process can possibly lead to the evolution of traits that benefit the group while reducing the fitness of the individuals.

Group selection theory is highly controversial. Proponents believe that group selection theory is necessary for explaining altruistic behavior, while sceptics argue that a group of altruists can be invaded by egoists who will outcompete the altruists.

Mathematical models and simulation models of group selection have hitherto relied un unrealistic assumptions and approximations. I have worked many years to develop a more detailed simulation program, and the result is now available as an open source program named 'Altruist'. This program can simulate several different models and different mechanisms of group selection. Furthermore, the program can automatically rerun the simulation with different values of the relevant parameters in order to find the parameter ranges that lead to altruism or egoism.

The conclusion is that traditional group selection models can lead to altruism only under very restricted conditions. Altruism cannot evolve unless the rate of migration between groups is very low.

The conditions for group selection to work may be satisfied for social insects, but certainly not for humans. We need another theory for explaining, for example, collective fighting in humans. The answer is, of course, regality theory. If people support a strong leader in case of war, then the leader can overcome the collective action problem by rewarding brave warriors and punish cowards and defectors. The behavior of the leader is easy to explain because there is a strong fitness advantage in being the leader of a strong group.

Simulation of this model confirms the assumptions behind regality theory. The hypothetical genetic trait that makes people support a strong leader in case of war can easily evolve under realistic conditions. The simulations show that such a trait can evolve even under a low intensity of war, a relatively high rate of migration between groups, and even if survivors of a defeated group are adopted into the winning group.

The results of simulating the different models are summarized in the working paper: Computer simulation of group selection models https://www.researchgate.net/publicatio ... ion_models

The simulation program is available at https://github.com/AltruistSim