The history of generalizing studies of early states and their origins goes back to the development of theories of social evolution in the eighteenth century. These assumed that it was possible to abstract away the historical specifics of individual states to create a category, ‘premodern states’, that has some features in common, and went on to propose general factors that might account for their origin, such as population pressure, warfare, or the control of irrigation systems. Such ideas were especially prominent during the heyday of neo-evolutionism in the second half of the twentieth century, which saw inferences from comparative anthropology linked to gradually increasing amounts of archaeological information.
This book presents a twenty-first-century version of such comparative approaches, based on the adoption of ideas and methods from the theory of complex adaptive systems and on the increasing efforts to create large-scale databases of attributes describing key features of specific societies, past and present, using information in ethnographic accounts and archaeological reports. This latter work represents a continuation of the tradition established by Murdock and White (Reference Murdock and White1969) with the creation of the ‘Standard Cross-Cultural Sample’, but now on a much larger scale enabled by modern information technology.
Complex systems are networks of interlinked entities whose interactions result in outcomes that cannot be predicted or understood simply from a knowledge of the properties of their individual elements, for example the emergence of new levels of organization. One of the main tools for the study of complex systems is agent-based modelling, in which autonomous ‘agents’, low-level entities with a given set of properties, interact with one another and the results of the interactions are analysed with the aim of testing hypotheses about the features of the entities and their interactions that produce the outcomes. A characteristic of many such models is that they are stochastic: that is, there is an element of chance in what happens at each step so the results will never be exactly the same. The models need to be run many times in order to identify characteristic outcomes, something only possible with the availability of massive computing power. For scholars interested in explaining how premodern states arose, the theory and methods of complex adaptive systems have understandably proved attractive, since they offer new ways to understand the emergence of societies with novel organizational properties.
Since its foundation in the 1980s, the Santa Fe Institute (SFI) has been a pioneer and world leader in the interdisciplinary field of complex systems across domains from physics to ecology, economics and the study of human societies, and from its early years has had an interest in applying its approaches to archaeological and anthropological topics. Although small in scale in terms of the number of individuals for whom it is their main position, the SFI is the hub of a much larger network of affiliated academics at other institutions, many of whom have spent periods of post-doctoral research there. The book under review is a collection of papers from a group of archaeologists and anthropologists connected to the Institute in various ways. Co-editor Jeremy Sabloff was actually President of the Institute from 2009 to 2015.
In his chapter in the introductory section of the book, the history of attempts to define early states and explain their origins is briefly but lucidly traced by Henry Wright, who made a major contribution of his own with his characterization of early states in terms of the presence of ‘specialized control activities’ that could potentially be recognized in the archaeological record (Wright Reference Wright1977).
So why have the kinds of comparative approaches espoused by past social evolutionists and by the contributors to this book been greeted with such scepticism, if not downright hostility, by many cultural anthropologists, especially in the last two or three decades? This is the question addressed by Laura Fortunato in her contribution. Part of the answer is the legacy of Eurocentric values and assumptions of superiority inherent in nineteenth-century evolutionism, which led to a move towards a detailed understanding of individual societies in their own terms through participant-observation fieldwork. This emphasis on detailed description also led to the view that systematic quantitatively based comparison involved ripping specific cultural features out of their context. The attitude has strengthened further since the rise of ‘critical cultural studies’ in the 1980s and its attacks on ideas of objectivity. This is one of the reasons that scholars from disciplines with comparative generalizing traditions such as economics, ecology and psychology have ‘invaded’ the territory largely vacated by anthropologists, to pick what they see as the ‘low-hanging fruit’ of interesting questions that they can address with their analytical methods.
Archaeology has gone through milder versions of these trends, since it has always employed ideas from ethnography and anthropological theory to put flesh on the bones of its artefacts, and that makes an element of comparison unavoidable. However, the construction of systematic databases for comparative purposes in archaeology is a much more recent phenomenon than in social/cultural anthropology, no doubt at least partly because most archaeologists have regional interests and because most archaeological data is site-based and complex in structure. Fortunato describes the progress that has been made in constructing such databases and emphasizes both the potential, arising from the huge quantities of data that are already stored for resource-management purposes, and the data science challenges of bringing them together to address comparative questions.
The core of the book is the second section, on new research. Paula Sabloff and Skyler Cragg's contribution falls into the category of developing systematic description for comparative purposes. Using text-based information they characterize the roles, domains and agency of different kinds of actors in a number of premodern states and use a network analysis to explore how these relate to one another. States have patterns that distinguish them clearly from the ‘outgroup’ of two non-state societies included in the study, though they are by no means uniform. The authors show convincingly that their approach has the potential to reveal aspects of the operation of these societies that are completely missed by traditional taxonomies.
The chapters by Hooper et al. and Kohler et al. exemplify different approaches to agent-based modelling as a tool for explaining the development of hierarchies. The first creates an abstract model of strategies for obtaining resource sites, designed to explore the role of the use of economically defensible patchy resources in generating hierarchical differences and changing the scale of political integration. They show how the extent to which the ecology favours the existence of defensive coalitions has a major impact on the development of hierarchy and inequality. Perhaps less important than the specific conclusion, though, is the way that the modelling methodology enables the outcomes to be systematically related to the interactions between different elements of the model. This cannot be done with words. On the other hand, their linkage of their scenarios to real examples such as the Mississipian is via brief interpretive sketches.
Kohler begins by describing how new theoretical developments based on the use of formal methods, such as game theory, have opened out studies of the growth of hierarchy and inequality, for example by showing the complex dynamics involved in the generation and maintenance of group cohesion in different conditions. The core of the paper, though, is a detailed agent-based simulation based on the specifics of the archaeology of the North American Southwest. It starts by scattering households across a landscape with a resource distribution that changes through time on the basis of a climate record and explores the consequences of a variety of hierarchical and non-hierarchical social strategies on their success, including in warfare. However, Kohler recognizes that he is not in a position properly to evaluate the goodness of fit between his simulations and the historical context he is modelling. As with Hooper et al., the implications are qualitative: the models take us beyond vague verbalizations of processes to the generation of concrete outcomes given the model's assumptions, showing how large-scale patterns can emerge from micro-level processes.
Ortman and Peregrine return us to a comparative approach to the growth of social complexity, using a large-scale database of archaeological traditions, as opposed to ethnographically known societies, coded in terms of concrete archaeologically detectable attributes. By doing so, they escape from the traditional approaches of simply putting different ethnographic societies in order, on the one hand, and, on the other, focusing only on early civilizations and ignoring all the negative cases where this development never occurred. This is illuminating. It shows, for example, that it is the range of complexity that has grown. Just as with the evolution of life, the simple has continued to exist even as new levels of complexity have come into existence. But it can also tell us about process. For example, there is only a weak relationship between time since agriculture appeared in a region and the population size (logged) of the largest regional settlement at a given point in time, but a much stronger one between the latter and the duration of the associated tradition, indicating that larger-scale societies have tended to change faster than smaller ones. They make a strong case for their generalizing approach, not as a substitute for more specific studies of individual societies, but as complementary to them, by enabling the recognition of broad patterns in the scale and integration of the networks of energy and information that characterize different societies.
In the following chapter, Ortman addresses the question of why recent Darwinian (in the sense of ‘descent with modification’) approaches to cultural evolution have played little role in research on the origins of complex societies and been uncongenial to many archaeologists. Building on developments in the theory of evolutionary biology that have gone beyond the long-established neo-Darwinian synthesis, he proposes that what is needed is ‘a good model of cultural genotypes—the physical means by which humans represent information in the brain and utilize it in shaping behaviour.’ He develops and justifies the idea that such a model is provided by the universal human use of ‘conceptual metaphor’, the use of concrete phenomena to model more abstract ones. He goes on to propose that cultural complexity could not have emerged without the development of cultural genotypes that conceptualize social distance in a manner that makes people comfortable with being part of the ‘densely interacting social networks’ that give improved economic pay-offs. Others have discussed the kinds of social institutions and associated social norms that provide the glue that keeps large-scale societies together, but Ortman is more interested in identifying the specific content of the conceptual metaphors of actual social groups and how they have changed over time. He sketches a Southwestern US example where they can be traced using a combination of historical linguistics, oral traditions and material culture.
In the final substantive chapter, Peregrine strongly defends the position that the stages in the traditional social evolutionary taxonomies of Service and Fried represent real social patterns. He argues that they represent convergent ‘recurrent social formations’ that arise from a systematic relationship between the technology and scale of societies, demonstrated through the analysis of data from the Atlas of Cultural Evolution (Peregrine Reference Peregrine2003); only a limited range of stable social forms exists. These are conceptualized as peaks in an adaptive landscape that itself is changing through time so that as the technology and scale of societies changes, new social forms become more successful: early states represent one of those peaks.
In summary, this is a stimulating book that brings together developing approaches that open out the stale discussions on the origin of complex societies in interesting new directions. However, I am not sure that it will win over comparison sceptics.
