Abstract: Research in cultural evolution has focused on the spread of intuitive or minimally counterintuitive beliefs. However, some very counterintuitive beliefs can also spread successfully, at least in some communities—scientific theories being the most prominent example. We suggest that argumentation could be an important factor in the spread of some very counterintuitive beliefs. A first experiment demonstrates that argumentation enables the spread of the counterintuitive answer to a reasoning problem in large discussion groups, whereas this spread is limited or absent when participants can show their answers to each other but cannot discuss. A series of experiments using the technique of repeated transmission show that, in the case of the counterintuitive belief studied: (a) arguments can help spread this belief without loss; (b) conformist bias does not help spread this belief; and (c) authority or prestige bias play a minimal role in helping spread this belief. Thus, argumentation seems to be necessary and sufficient for the spread of some counterintuitive beliefs.
Diffusion of a counter-intuitive answer (green) in groups that initially give mostly the intuitive wrong answer (red). The size of the node represents the confidence of the participants.