Concluding remarks
Overall, the simulations confirmed the plausibility of the ‘cultural drive’ hypothesis. Under suitable conditions, the runaway coevolution of culture, social learning and brain capacity may start in a social species. The necessary conditions include a minimum starting level of social learning and a socio-ecological situation that ensures the possibility of sporadic invention of different, beneficial (for the individual or for the group) and cognitively demanding behaviours (memes). The chances for the runaway brain-culture coevolution increase when such memes belong to different categories: some are individually beneficial (e.g., ‘Machiavellian’) while the others are group-beneficial (e.g., ‘cooperative’). In this case, the runaway brain-culture coevolution can start and proceed under varying levels of between-group competition and migration.
Hominins, and especially the Early Pleistocene species of the genusHomo , probably found themselves in the right situation due to changes in their social and ecological niche. The changes of the social niche were related to the lower within-group aggression and competition, higher paternal investment in offspring and the proposed trend towards monogamy and social conformity. This probably resulted in new optimal strategies for reaching higher social status and reproductive success (individuals had to rely more on ‘Machiavellian intelligence’ than on physical strength and violence) (Lovejoy, 2009; Stanyon, Bigoni, 2014; Raghanti et al., 2018). The changes of the ecological niche were related to new feeding strategies such as confrontational scavenging and collective hunting (Bunn, Ezzo, 1993), which required high levels of within-group cooperation (Bickerton, Szathmáry, 2011; Gavrilets, 2015) and cognitively demanding behaviours such as the Oldowan stone tool making (Morgan et al., 2015).
The simulation results make it possible to suggest a tentative explanation for a somewhat counterintuitive pattern observed in anthropogenesis: in the Early Paleolithic, while cultural development was very slow (e.g., Beyene et al., 2013), the brain volume was increasing rapidly; in the Middle to Late Paleolithic cultural development accelerated greatly, while the brain ceased to increase and even somewhat decreased from the Late Paleolithic to recent (Holloway, 2015). In our simulations, brain expansion was efficiently stimulated by a ‘rough’, primitive culture consisting of a few large (difficult to learn) memes. Later, as the culture becomes more ‘sophisticated’ (that is, saturated with numerous small but efficient memes), the brain expansion slows down and even reverses. The mechanism underlying this pattern, which we dubbed ‘the vicious circle of meme simplification’, requires the combination of two types of limitations: for brain expansion and for the amount of knowledge that can be learned in a lifetime. Another requirement is that the available mechanisms of social learning must give a strong selective advantage to smaller memes. Under such conditions, increasing costs of brain expansion result in stronger selection for smaller memes, which, in turn, makes larger brains less beneficial.
However, brain expansion can receive additional boost if such means of social learning evolve that are at the same time (1) costly in terms of brain expansion (rely on many complicated neuronal curcuits) and (2) are tolerant to meme complexity, that is, make it possible to learn and teach difficult skills and concepts relatively easily. Such mechanisms of social learning can break or weaken ‘the vicious circle’, so that the brain expansion will be able to proceed further. Human language is probably just such a kind of a social learning mechanism. We tentatively suggest that the extraordinarily fast brain expansion in the course of human evolution, which later opened unique possibilitities for the development of civilization, was probably a kind of evolutionary accident: the cultural drive was just too strong for the brain to evolve in a more balanced way, e.g., by structural optimization rather than by disproportionate increase in volume and the number of cortical neurons.
The results also suggest that the runaway brain-culture coevolution can acquire additional acceleration from two positive feedback loops: via population growth and via lifespan extension (e.g., if cultural development results in lower mortality rate) (fig. 10).
The results also imply that between-group competition (1) tends to make culture less diverse, (2) facilitates the propagation of group-beneficial memes, (3) effectively removes parasitic memes from the meme pool. Cultural group selection (that is, selective survival and propagation of groups with more efficient cultures) shapes the content of the meme pool, while individual selection is almost unable to do so. Due to the horizontal spread of memes, culture is a group characteristic rather than an individual one.