Cumulative Culture in Nonhumans
- Taxa: Homing pigeons (Clumba liviva)
- Topics: Collective intelligence
In the ever morphing effort to delineate the uniqueness of humans, the concept of “cumulative culture” stands out. Kevin Laland, who has done so much conceptually to extend the sphere of culture to nonhumans, still insists that “cumulative culture . . . demands transmission of information with a degree of accuracy of which only humans are capable.” So it is notable when an instance of nonhuman cumulative culture emerges, especially when this research suggests further lines of inquiry. Takao Sasaki and Dora Biro, based on their study of cross-generational knowledge transfer among pigeons, suggest “that collective intelligence in animal groups can accumulate progressive modifications over time”; they “predict similar accumulation of knowledge in other multi-generational social groups, ranging from the establishment of ‘traditional’ foraging areas in social insects to long-distance migration routes in whooping cranes.” This assertion generated a considerable amount of media coverage (see links below) because their core findings seem to disrupt yet another attempt to limit the expanse of culture beyond the human.
Their claims are based on a savvy reading of this literature, through which they devised an experimental case-study that “demonstrates the presence of cumulative cultural evolution in a non-human species.” Sasaki and Biro focused on collective navigation by homing pigeons, which they framed via study methods developed on humans—an experiment testing how design information accumulates within, resulting in progressively more successful design. Individual pigeons develop and then recapitulate homing routes; flocks do something similar but collectively. Such “solutions” stabilize as more birds are accommodated in the group. This scenario allowed Sasaki and Biro to systematically gauge both the appearance of novel innovations from existing behaviors and their cross-generational maintenance. In an experimental group, “naïve birds” were paired with experiences ones, which, in turn, were cycled out as other new birds were added; eventually this grouping outperformed control groups, demonstrating the progressive transmission of culturally acquired knowledge—just as we recognize in humans.
Central here is the “ratchet effect” of cultural learning—the experimental group continued to improve its route accuracy, an unexpected result on current theories of homing navigation in pigeons. This suggests the birds pool information in a manner that each new generation can contribute “innovations,” such that they are learning through a collective intelligence. Notably, this entails an interpretive capacity: “pigeons are capable of evaluating the payoffs of their performance, such that when errors do get added by naïve individuals, these innovations can be ‘pruned’ on the basis that they lead to worse performance (while those that lead to better routes are kept).” This kind of filtering process is considered a basic means by which cultural knowledge accumulates; Sasaki and Biro add to it a “time-depth” dimension current navigation is contingent upon a previous history, shaping a collective behavior. Granted, this needs further examination; they point to the need for “data on the ancestral and intermediate states of the behaviors in question,” perhaps through life histories. Is there an ethnographic potential here? Perhaps—they suggest “future work should examine the precise nature of social transmission operating during route learning in pigeons.” Regardless, their findings are significant for showing how something we equate entirely with complexity—accumulation of knowledge—can be achieved via simple means. Enhanced cognitive capacities, as with larger brains, may not be necessary for culture to emerge and evolve. As Sasaki and Biro conclude, “even agents with limited cognitive abilities can make progressively superior decisions over time,” because they pool collective intelligence.
Takao Sasaki and Dora Biro, “Cumulative Culture Can Emerge from Collective Intelligence in Animal Groups,” Nature Communications 8 (2017), https://doi.org/10.1038/ncomms15049.
Studies of collective intelligence in animal groups typically overlook potential improvement through learning. Although knowledge accumulation is recognized as a major advantage of group living within the framework of Cumulative Cultural Evolution (CCE), the interplay between CCE and collective intelligence has remained unexplored. Here, we use homing pigeons to investigate whether the repeated removal and replacement of individuals in experimental groups (a key method in testing for CCE) alters the groups’ solution efficiency over successive generations. Homing performance improves continuously over generations, and later-generation groups eventually outperform both solo individuals and fixed-membership groups. Homing routes are more similar in consecutive generations within the same chains than between chains, indicating cross-generational knowledge transfer. Our findings thus show that collective intelligence in animal groups can accumulate progressive modifications over time. Furthermore, our results satisfy the main criteria for CCE and suggest potential mechanisms for CCE that do not rely on complex cognition.
- Elizabeth Pennisi, “Think Only Humans Can Build on the Knowledge of Previous Generations? Meet These Pigeons,” Science, April 18, 2017, doi: 10.1126/science.aal1065.
- University of Oxford, “Homing Pigeons Share Our Human Ability to Build Knowledge across Generations,” ScienceDaily, April 18, 2017.
- Henry Bodkin, “Bird Brains Not So Stupid: Pigeons Show Human-Like Ability to Build Knowledge through Generations,” Telegraph, April 18, 2017.
- John Michaelle, “Homing Pigeons Share the Human Capacity to Build on the Knowledge of Others,” Evonews, April 19, 2017.
- Space, “Pigeons Have a Cultural Evolution,” Earth Chronicles News, April 18, 2017.
 C. A. Caldwell and A. E. Millen, “Experimental Models for Testing Hypotheses about Cumulative Cultural Evolution,” Evolution and Human Behavior 29 (2008): 165–71.