This article presents a model of cultural evolution simulating the accumulation of tools in specialized and non-specialized populations under different demographic and environmental scenarios. The model predicts that the relationship between population size and repertoire size is nonlinear and can differ between non-specialized and specialized populations. For small population sizes, the non-specialized populations maintain knowledge better and therefore reach higher average repertoire sizes. In large populations, specialized populations can reach higher average repertoire sizes. This is because non-specialized population's total repertoire size is limited by the capacity of individuals to accumulate knowledge of different skills, while in specialized populations, each individual needs to know only a fraction of the population's repertoire. However, the model also predicts that specialized populations are more susceptible to information loss due to their subdivision of knowledge, and this can be amplified by demographic and environmental factors. The authors also use ethnographic data to analyze the relationship between population size and degree of craft specialization of societies, and how this may be influenced by ecological factors.

This article presents a model of cultural evolution simulating the accumulation of tools in specialized and non-specialized populations under different demographic and environmental scenarios. The model predicts that the relationship between population size and repertoire size is nonlinear and can differ between non-specialized and specialized populations. For small population sizes, the non-specialized populations maintain knowledge better and therefore reach higher average repertoire sizes. In large populations, specialized populations can reach higher average repertoire sizes. This is because non-specialized population's total repertoire size is limited by the capacity of individuals to accumulate knowledge of different skills, while in specialized populations, each individual needs to know only a fraction of the population's repertoire. However, the model also predicts that specialized populations are more susceptible to information loss due to their subdivision of knowledge, and this can be amplified by demographic and environmental factors. The authors also use ethnographic data to analyze the relationship between population size and degree of craft specialization of societies, and how this may be influenced by ecological factors.