Bret Beheim

@babeheim.bsky.social

2943 Followers 2469 Following

cultural evolution, behavioral ecology, math models, data provenance, MOSAIC group leader @ Max Planck Institute for Evolutionary Anthropology + faculty at the Leipzig School of Human Origins https://babeheim.com/

Miyagawa Shuntei's 1898 painting, "Playing Go (Japanese Chess)"

decision trees in human play in the game of go before and after the arrival of superhuman AI, showing gradual change rather than revolutionary disruption

Though these learning models have been supported by laboratory and field experiments in which participants are given controlled access to potentially valuable social information (Baum et al., 2004, Caldwell and Millen, 2008, Efferson et al., 2007, Mesoudi, 2011), studying cultural evolution in outside controlled settings is a much harder task. Michel et al. (2011) use Google book archives to trace word use frequencies over the last several centuries but, lacking information about who is using those words and under what circumstances, they are unable to explain these trends beyond phenomenological descriptions and qualitative associations to major historical events. Ideally, if we had individual-level phenotypic measures for an entire population over many years, we could identify and measure the driving forces of cultural change, as has been accomplished in non-human evolutionary systems (e.g. Grant and Grant, 2002, Ozgul et al., 2009).