The pre-history of urban scaling

PLoS One. 2014 Feb 12;9(2):e87902. doi: 10.1371/journal.pone.0087902. eCollection 2014.

Abstract

Cities are increasingly the fundamental socio-economic units of human societies worldwide, but we still lack a unified characterization of urbanization that captures the social processes realized by cities across time and space. This is especially important for understanding the role of cities in the history of human civilization and for determining whether studies of ancient cities are relevant for contemporary science and policy. As a step in this direction, we develop a theory of settlement scaling in archaeology, deriving the relationship between population and settled area from a consideration of the interplay between social and infrastructural networks. We then test these models on settlement data from the Pre-Hispanic Basin of Mexico to show that this ancient settlement system displays spatial scaling properties analogous to those observed in modern cities. Our data derive from over 1,500 settlements occupied over two millennia and spanning four major cultural periods characterized by different levels of agricultural productivity, political centralization and market development. We show that, in agreement with theory, total settlement area increases with population size, on average, according to a scale invariant relation with an exponent in the range [Formula: see text]. As a consequence, we are able to infer aggregate socio-economic properties of ancient societies from archaeological measures of settlement organization. Our findings, from an urban settlement system that evolved independently from its old-world counterparts, suggest that principles of settlement organization are very general and may apply to the entire range of human history.

Publication types

  • Historical Article
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Archaeology / methods*
  • Cities*
  • Efficiency
  • Geography
  • History, Ancient
  • Humans
  • Mexico
  • Urbanization*

Grants and funding

Funded by National Science Foundation (1005075 and 103522); James S. McDonnell Foundation (220020195); John Templeton Foundation (15705); Rockefeller Foundation; Bryan J. and June B. Zwan Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.