Conservatism in laboratory microsocieties: unpredictable payoffs accentuate group-specific traditions

Theoretical work predicts that individuals should strategically increase their reliance on social learning when individual learning would be costly or risky, or when the payoffs for individually learned behaviors are uncertain. Using a method known to elicit cumulative cultural evolution in the laboratory, we investigated the degree of within-group similarity, and between-group variation, in design choices made by participants under conditions of varying uncertainty about the likely effectiveness of those designs. Participants were required to build a tower from spaghetti and modeling clay, their goal being to build the tower as high as possible. In one condition, towers were measured immediately on completion and, therefore, participants were able to judge the success of their design during building. In the other condition, participants' towers were measured 5 min after completion, following a deliberate attempt to test the tower's stability, making it harder for participants to judge whether an innovative solution was liable to result in a good score on the Final measurement. Cultural peculiarity (i.e., the extent to which a design could be identified as belonging to a particular chain) was stronger in the delayed measure condition, indicating that participants were placing greater reliance on social learning. Furthermore, in this condition, there was only very weak evidence of successive improvement in performance over learner generations, whereas in the immediate measure condition there was a clear effect of steadily increasing scores on the goal measurement. Increasing the risk associated with learning for oneself may favor the development of arbitrary traditions. (C) 2010 Elsevier Inc. All rights reserved.