Social Transmission along Multiple Pathways Promotes Information Fidelity and Reduces Divisiveness

A story transmitted through a chain of several individuals can become distorted, while information checked against multiple sources at various stages might be more accurate. Using a novel large-scale networked behavioural experiment, we demonstrate that redundancy of information presented by social networks is a demonstrable cause of higher fidelity and consensus of information. We show that even minimal structural redundancy can promote information fidelity and consensus formation in re-tellings of a long-form scientific text article transmitted through varying network structures. We find that transmission in multiple-pathway networks (i) reduces information loss as texts propagate between participants, and (ii) promotes consensus on information content between non-interacting participants in independently tested networks. A simple computational model, grounded in past social learning research, explains these effects and offers new insights into the mechanisms of multiple-pathway information transmission, contributing to the resolution of an outstanding puzzle in early social psychology research.