News Release

Selective social learning helps preserve complex cultural knowledge

Peer-Reviewed Publication

American Association for the Advancement of Science (AAAS)

When people can choose a successful teacher to help them learn a complex strategy, efficient strategies accumulate and spread through a culture more easily, according to a new social learning experiment conducted by Bill Thompson and colleagues. Their findings suggest that some of the most complicated cultural “abilities”—from cooking to sailing—may be preserved through selective social learning. It has been difficult to explain the cultural accumulation of innovative strategies, which are hard to pass on to new generations but have allowed humans to expand across diverse ecosystems and build multigenerational accomplishments. Mathematical models suggest that choosing to learn from successful individuals can preserve rare but important problem-solving strategies. To explore this problem, Thompson et al. recruited 3450 people online and organized them into 20 populations that searched for algorithms to solve a sorting problem, transmitting their solutions over 12 generations of the experiment. After the first generation, where individuals came up with solutions on their own, half of the participants were able to choose algorithm demonstrators from earlier generations based on information about those demonstrators’ successes. The other half did not know anything about their demonstrators’ successes. The researchers found that two main algorithms to solve the problem evolved to be used by many people, but the most efficient algorithm only spread through the populations when people were allowed to select their demonstrators based on previous success. In a related Perspective, Joseph Heinrich discusses how selective social learning could have helped preserve vital cultural knowledge over human evolution.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.