Evolutionary Consequences of Conditional vs. Unconditional Sharing: A Case Study in Computational Philosophy

Agent-based modeling techniques have been used by researchers in the field of computational philosophy to provide insights into a number of philosophical questions. In the study reported here, we ran a series of agent-based simulation experiments to see whether groupwide adherence to certain specific moral rules and practices might provide an evolutionary advantage. In each experiment, two groups of individuals competed with each other to harvest food. The members of one group (“sharers”) then shared food with each other as specified in one of two sharing rules; members of the other group (“hoarders”) never shared food. Individuals who had at least a specified amount of food survived and reproduced, while those who did not died of starvation. Changes in population size were used as a measure of evolutionary fitness. In some experiments, the sharers followed an “unconditional” sharing rule under which they were always obligated to share food. In other experiments, sharers followed a “conditional” sharing rule that made an exception when sharing could put the sharer itself at risk of starving. We found that unconditional sharing had an evolutionary advantage over hoarding when food was plentiful, but a disadvantage when food was scarce. Conditional sharing also offered an advantage over hoarding in a high-resource environment, but unlike unconditional sharing, it did not result in a disadvantage in a low-resource environment. Our observation of these evolutionarily favorable characteristics of conditional sharing provides an interesting example of the potential use of agent-based modeling to inform discussions about the evolution of morality.