Experimental Computational Philosophy: shedding new lights on (old) philosophical debates

Philosophy can benefit from experiments performed in a laboratory for philosophical experimentation (SophoLab). To illustrate the power of Experimental Computational Philosophy, we set up and ran several experiments on a part of Harsanyi’s theory on utilitarianism. During decomposition and translation of this theory in the experimental setting of Sopholab, we discovered that it is underspecified. We filled in some blank spots and found out that information and its costs are key in the effectiveness of act and rule utilitarianism. We also identified three further elements that have particular influence on the effectiveness of both strands of utilitarianism: group size of agents, decision-making around uncertainty, and social culture towards particular types of actions. We conclude having shown that setting up computational philosophical experiments is a useful way to gain new and deeper insights in existing argumentations used in old (and new) philosophical debates.