5/ Why did people stick to their theory? Because social learning canalises exploration. Participants were less likely to test configurations that could challenge their received theory, reducing their chances of discovering the effects of variables not emphasized by that theory. 🔄

4/ Our main finding: Even when people had incentives to improve their solutions, they stuck to the theory they received—even if it was wrong.

3/ Some participants received a "theory" about what makes the wheel roll faster—either correct, partially right, or misleading. We then compared their exploration to participants who received no theory at all.

2/ We designed an experiment where participants had to optimize a physical system: a wheel rolling down a track. Their goal? Adjust the weights to make it roll as fast as possible. 🚀

New paper in #ProcB with @alexmesoudi.com, @jfbonnefon.bsky.social, @rmcelreath.bsky.social and Rob Boyd.

We investigate how social learning shapes the way we explore and show that it can even preserve useless theories! 🧵👇

6/ Ours findings reveal the unique challenges of individual vs. social learning:
🌀 Information loss in chains
🔄 Redundant effort in groups
😵 Cognitive & motivational fatigue in extended individual learners

We discuss the implications for experimental tests of cumulative cultural evolution.

5/ This highlights the effect of increasing acquisition costs:
📚 As knowledge grows, social learners in later generations must invest more time to acquire accumulated skills—slowing innovation.

But individual learners? 🚀 They get more proficient with each session, keeping innovation time stable.

4/ What did we find?
✅ In chains, participants improved over generations—but repeated individual learners performed even better! 🤯
Why? Social learners need more time to reproduce previous discoveries, leaving less time for further innovation ⏳.

3/ Using the Totem Task, where participants create complex “totem poles” by discovering and combining innovations, we compared 4 experimental conditions:
1️⃣ Transmission chains
2️⃣ Constant groups
3️⃣ Repeated bouts of individual learning
4️⃣ Extended periods of individual learning

2/ 🧪 Experiments have yielded groundbreaking insights, but critiques remain. A key debate:
what criteria should be used by experimenters to determine whether the outcome of an experiment constitutes cumulative culture? 🤔

1/ 🌟 @PNASNews Special Feature: Half a Century of Cultural Evolution 🎉

With @amesoudi.bsky.social, @glupyan.bsky.social & Pierce Edmiston, we review the field's key experimental methods, findings, and critiques—and present a new lab experiment. 👇

www.pnas.org/doi/epdf/10....