Reviewed by William Caraher
Oxford Studies on the Roman Economy. Pp. xx + 362, figs. 79, tables 54. Oxford University Press, Oxford 2011. $135. ISBN 978-0-19-960235-8 (cloth).
Historians and archaeologists of Greece and Rome have long recognized the relationship between population and the ancient economy. Karl Julius Beloch's 19th-century efforts to estimate population of the Roman world have continued to stimulate debate (Die Bevölkerung der griechisch-römischen Welt [Leipzig 1886]). Over the course of this debate, scholars have tended to accept either "low" or "high" population estimates for Augustan Italy—6–7 million and 12–14 million, respectively—which they have then extrapolated to the entire Roman empire. These positions have specific, complex consequences for how historians and archaeologists understand the Roman economy, urbanization, and ultimately social and political organization. Bowman and Wilson's Settlement, Urbanization, and Population contributes to these conversations by bringing to bear recent archaeological research across the Mediterranean basin. It is the second volume produced by a research program called "The Economy of the Roman Empire: Integration, Growth, and Decline," funded by the Arts and Humanities Research Council in the United Kingdom. The first volume (A.K. Bowman and A. Wilson, eds., Quantifying the Roman Economy: Methods and Problems [Oxford 2009]) considered the potential of quantitative methods for shedding light on the Roman trade, demography, and settlement. This second volume continues in this vein by applying quantitative methods to archaeological data to critique long-standing arguments on population and settlement organization.
The book is divided into two sections, with the first section being more explicitly methodological and the second being more oriented toward specific regions and projects. The first four contributions examine the role of various forms of regional survey in providing evidence for demographic change in the Roman world. Regional survey, with its emphasis on documenting sites of all kinds in both urban and rural areas and its use of the intensive pedestrian method to produce robust data sets on a large scale, has long held the potential for expanding settlement patterns beyond the limited perspectives offered by excavation. This new and growing data set, however, requires particularly critical reading. Several contributions emphasize the methodological challenges of using this data to estimate regional and Mediterranean-wide population figures from survey data. Price's methodological consideration of Greek population demonstrates the value of grounding estimates in evidence from well-preserved excavated classical and Hellenistic sites such as Halieis and Olynthus and then applies these to results from intensive survey on Crete. This work on the Greek world forms a point of comparison for considering the more lacunous evidence for Roman settlement and demography. Witcher's contribution considers the tricky issue of site recovery rates, which depend on variables such as artifact identification, long-term taphonomic processes, ancient habitation practices, and modern land use. Variability in site recovery rates across the Mediterranean basin has muddied archaeologists' ability to document ancient settlement patterns consistently and introduces significant uncertainty to the very data used to produce population estimates. Mattingly's contribution offers a striking testimony to this issue when he notes that exceedingly high surface visibility and site preservation present in arid zone surveys in North Africa regularly produce higher site densities than surveys in more temperate regions of the Mediterranean such as Greece and Italy. While this might suggest differences in settlement practice, it seems more likely to hint at the limitations of site recovery in more temperate areas of the Mediterranean basin. Attema and de Haas mitigate this issue by drawing on data collected through a variety of survey and excavation projects to reconstruct settlement on the Pontine Plain south of Rome. They identify a number of different settlement types in the region, ranging from isolated farms to villae, villae maritimae, and villages. They then consider the potential population of each of these sites and use this to model the population of the region. Like many authors in the volume, they present population estimates contingent on variables such as the average size of household and site recovery rates. At times, the authors' willingness to enumerate the number of significant variables that remain disputed among survey archaeologists, such as those related to site recovery rates and household size, obscures the potential of these methodological exercises to produce widely accepted conclusions.
The second group of six contributions takes these methodological discussions and applies them to understanding the complex relationship between population and urbanization across the Mediterranean. These chapters use quantitative methods to advance the long-standing debates concerning the size and significance of urban populations. The size of the population at major cities such as Rome and the numerous smaller cities of the Mediterranean and the role of these cities in the Roman economy remain topics of intense debate and significant uncertainty. As a result, the overarching tone of this section is that of caution.
Two articles offer some general perspectives on population and urbanism. The first chapter in this section, by Morley, introduces vocabulary from sociologist Philip Abrams, who saw urbanization as an ongoing process of concentration, crystallization, integration, and differentiation. While this model does not necessarily illuminate specific causes for change, it provides a technical vocabulary to understand urbanism in a comparative context. Unfortunately, the other authors in this section did not carry this vocabulary through their work. In a similarly general vein, Wilson's article offers an overview of the population of the Roman empire. His figures tend to conform to the "low" count, but he cautions that lower population figures imply higher levels of urbanization and a more efficient economy.
The other contributions focus on individual regions. Articles by Marzano and Hanson use rank-size analysis, central place theory, and Zipf's Law to consider the relationship between the physical size of the city—which served as an analog for population—and economic integration of a particular region. When applied to population distribution, Zipf's Law postulates a knowable relationship between population size and settlement patterns by arguing that in a standard series, the population of a city is inversely proportional to its rank. Since the density of ancient populations in urban sites is difficult to determine, Marzano proposes that site size in Britain and the Iberian peninsula provide a useful indicator of population, with the assumption that population density was more or less consistent across settlements in a region. The imperfect fit between the number of settlements of a particular size and the distribution predicted by Zipf's Law either reveals problems with the data set or, more helpfully for historical analysis, understandable patterns in the organization of local settlement. In the case of the Iberian peninsula, for example, the convexity of the distribution of cities in relation to the linear relationship proposed by Zipf's Law suggests regional settlement systems with far less economic integration. Hanson's contribution on Asia Minor offers a more nuanced and complex perspective on urbanization. He considers the topographic context of cities in this densely urbanized Roman province as well as the location of roads and other transportation routes while offering a useful corrective on the sometimes exaggerated size of cities in Roman Asia Minor. A similar level of regional scrutiny appears in Bowman's systematic consideration of the papyrological and archaeological evidence available for Ptolemaic and Roman Egypt, which he argues could provide useful insights for demography across the Mediterranean as a whole despite the unique economy and geography of the Nile. Keay and Earl integrate epigraphic and landscape data to produce a detailed study of a subregion in Roman Baetica. Keay and Earl's project, as well as Hanson's work in Asia Minor, reveals the growing significance of GIS technologies placing more traditional forms of archaeological data (e.g., inscriptions of known provenance) into more complex topographic, geographical, and spatial contexts.
While few of the essays in this volume present definitive conclusions to complex issues surrounding population, the economy, and urbanism in the Roman period, they do provide a distinct perspective on the study of ancient demography and economy. In general, the contributors share a general reluctance to engage too fully with comparative data from premodern contexts—such as medieval or Ottoman census data—and this marks a departure from recent scholarship in the eastern Mediterranean, which has often seen this kind of data as offering important clues to long-term or systemic premodern demographic trends. Moreover, it is striking that the contributors to this volume generally avoided engaging recent debates surrounding siteless or artifact-level survey among practitioners of intentional pedestrian survey in Greece. These debates have focused on the ontological instability of sites as historical and archaeological realities in the landscape and would add an additional level of complication to many of these articles in the first half of the volume. The complexity of siteless survey evidence, particularly from "second wave" surveys in Greece and Cyprus, was generally overlooked in these studies, despite the unprecedented quality of this data.
The appeal to quantitative data and methods makes this volume a natural complement to the recent resurgence of quantitative research on the ancient economy typified by works such as Sheidel, Morris, and Sallers' Cambridge Economic History of the Greco-Roman World (Cambridge 2007). As the editors themselves admit in the introduction, there will always be significant limitations to the data and the conclusions, but only through careful scrutiny of the available data will scholars understand the parameters within which knowledge can advance.
Department of History
University of North Dakota
Grand Forks, North Dakota 58202