Novel phylogenetic methods reveal that resource-use intensification drives the evolution of “complex” societies

Abstract

Explaining the rise of large, sedentary populations, with attendant expansions of socio-political hierarchy and labor specialization (collectively referred to as “societal complexity”), is a central problem for social scientists and historians. Adoption of agriculture has often been invoked to explain the rise of complex societies, but archaeological and ethnographic records contradict simple agri-centric models. Rather than a unitary phenomenon, “complexity” may be better understood as a network of interacting features, which in turn have causal relationships with subsistence. Here we use novel comparative methods and a global sample of 186 nonindustrial societies to infer the role of subsistence practices in shaping complexity. We also introduce a phylogenetic method for causal inference that generalizes beyond two binary traits, lifting a major constraint on comparative research. We found that, rather than agriculture alone, a suite of resource-use intensification variables leads to broad increases in technological and social differentiation. Our study provides evidence that resource intensification is a leader, not a follower, in the rise of complex societies worldwide.

Coevolution of relative brain size and life expectancy in parrots

“…These hypotheses could not be disentangled in this analysis, as both predict a direct relationship between life expectancy and cognition, albeit with reversed directionality. Future studies could additionally try to use process-based approaches (where evolution is modelled explicitly), such as generative inference (43) or Bayesian ancestral state reconstruction (44) to disentangle the direction of causality. However, we found no evidence that the relationship between relative brain size and life expectancy was explained by the need for longer development times (here measured by incubation to fledging time, and by age of first reproduction), or by increased parental investment (here represented by clutch size), as predicted by the Expensive Brain Hypothesis .…”

“…We would, however, also expect a strong effect of diet on life expectancy, since this hypothesis argues that long life spans allow species to invest more time in learning foraging skill which requires larger brains, and only pays off with an extended juvenile period. To fully explore this hypothesis we would need data on post-fledging parental care and future studies could additionally try to use process-based approaches (where evolution is modelled explicitly), such as generative inference (47) or Bayesian ancestral state reconstruction (48) to disentangle the direction of causality. We found no evidence that the relationship between relative brain size and life expectancy was explained by the need for longer development times (here measured by incubation to fledging time, and by age of first reproduction), or by increased parental investment (here represented by clutch size), as predicted by the Expensive Brain Hypothesis .…”

Coevolution of religious and political authority in Austronesian societies

“…https://doi.org/10.1038/s41562-022-01471-y ordinal variables represents a latent continuous trait and modelling their coevolution using a recently developed Bayesian phylogenetic method that allows inferences to be made about the influence of two or more traits of any distribution on each other, as well as the role of 'selection' and 'drift' in the evolution of each 33 . Since linguistic distances between societies were positively correlated with geographic distances (r = 0.31; 95% confidence interval, (0.28 0.33); d.f.…”

“…Many authors have implemented this approach using the Discrete component of the software package BayesTraits 56 , but that method is limited to binary traits. To avoid having to dichotomize our ordinal variables, we used a recently developed Bayesian method for dynamic coevolutionary analyses that can accommodate any number of traits of any distribution 33 . With this approach, ordinal traits are modelled as latent continuous variables evolving under selection (both autoregressive selection and cross-trait selection) and drift, similar to a multivariate Ornstein-Uhlenbeck model.…”

Text and the City: How the City Shaped Language

“…The surplus agricultural produce became large enough to sustain a specialised workforce for building permanent infrastructure such as irrigation canals, storage facilities, city walls, central temples and squares, and a political elite that maintained the infrastructure, distributed the agricultural produce, and converted the surpluses into merchandise for trading (cf. Ringen et al 2021). And just like that, the city was born.…”