This is an important book. But it is not an easy book. Network Analysis in Archaeology brings together a number of the leading figures in archaeological network analysis to explore what ‘networks’ mean for archaeology, what a network perspective might offer the practice and interpretation of archaeology, and to offer the results of case studies ranging in space and time from across the globe. Newcomers to archaeological network analysis might find it daunting, but if they read this book after reading survey articles by individuals such as Tom Brughmans (2010, 2012, 2013) or Knappett’s 2011 volume, An Archaeology of Interaction: Network Perspectives on Material Culture and Society (Oxford: Oxford University Press), they will be well equipped to begin their own forays into archaeological network analyses.1
The book is timely. ‘Network’, as a metaphor, suffuses everything we do these days, to the point where the word is almost meaningless. This collection returns the analytic power to the concept. This collection shows what sustained and careful use of formal network methods can offer archaeology, beyond what are sometimes called the ‘spaghetti monsters’ of network visualization (Malkin 2011, p18).2 Indeed, social network analysis does not need to visualize networks in order to provide useful insight; sometimes the metrics alone are sufficient, as Coward’s contribution in this volume amply demonstrates (247-280).
The volume is arranged in four parts. Part I sets some of the background to the use of networks by archaeologists over the last half century (Terrell), and many of the theoretical potentials and pitfalls of the approach (Isaksen). Isaken’s contribution is especially important, as he draws attention to the ‘deceptive’ way the phrase ‘network analysis’ conceals ‘a wide arrange of techniques designed for different purposes’ (43). The network analysis tools that are widely available (and often freely available) come to us from studies of the structure of the internet, of social media, of mathematical graph theory. Their metrics and approaches have to be used with some caution and with understanding of the contexts for which they may be appropriate. Isaksen also points out that a major theme in archaeological network analysis is that of experimentation, of iteratively exploring both real and hypothetical networks and the interplay of variables that make one match the other. Given the necessary incompleteness of archaeological data, this experimentation or exploration seems to me to be a major attraction of network analysis for archaeologists.
Part II explores network analysis over geography, examining ways networks can be deduced amongst, within, and across sites and settlements. The contribution by Scholnick, Munson and Macri on Mayan political rhetoric (95-124) is particularly noteworthy, in that it brings to the fore the notion of ‘structural equivalence’ in networks deduced from epigraphic materials. They are looking, not just at connections, but at the holes, the patterns of non-connections. Given the rich epigraphic records of Greco-Roman antiquity, readers of BMCR might find much to inspire them in this piece. Rivers, Knappett, and Evans explore ideas around ‘centrality’ and deducing ‘what makes a site important’ in their contribution (125-150). They evaluate and discuss several models of assessing site importance, and offer their own, ‘Ariadne’, a simulation package that operationalizes their perspective (the Ariadne package is open source and is available for download). That these authors share their code is exciting; network analysis needs re-usable data to learn with and build upon, and such sharing is to be applauded.
Part III turns to material culture, and networks generated from artefact distributions themselves. The contribution by Mills et al. is extremely important in not only its sheer scope, some 1,600 settlements with approximately 4 million artefacts across 415 ceramic types in the American South West, but also in terms of the lucidity of their approach. Using Brainerd-Robinson co-efficients of similarity, they are able to explore and map regional migrations of people through the region, seeing both large-scale and micro-scale patterns play out. They demonstrate the ways in which their research confirms or is in accordance with existing findings but also the ways in which network analysis pushes understanding of this region and period forward. Mol and Mans reverse the telescope and use social network analysis to understand the social world within a single village in the Caribbean. They suggest that intra-site relationships as evidenced in material culture can shed light on larger inter-site relationships. Collar uses epigraphic evidence of the Jewish Diaspora, constructing networks via proximal-point analysis, to show how conceptions of Jewish ethnicity changed after the destruction of the Temple. It would be interesting to extend Collar’s approach to take temporal proximity rather than geographic proximity into account (which could be done with, for instance, Scheidel and Meeks’s ORBIS The Stanford Geospatial Network Model of the Roman World, which allows for the reconfiguring of Mediterranean geography by time rather than space).
The book concludes with a meditation on ‘Archaeology, networks, information processing, and beyond’, by Sander van der Leeuw. He remarks that virtually every paper in this collection uses network analysis in an analytic/descriptive mode, and so he frames his remarks to discuss where network analysis may go in the future. He explores ideas concerning ‘societies as dissipative flow structures’. He writes, ‘we can model the interaction between the society and its environment as a dynamic structure consisting of two flows going in opposite directions—a flow of organization (‘information-processing capacity’) that emanates from the society and flows towards the surrounding environment, and a flow of energy that is extracted from that environment and that flows inwards into the society. The resultant feedback loop dissipates entropy by organizing the environment and the society’ (338-9). He goes on to discuss multi-nets, or networks where the nodes have different kinds of relationships and the nodes themselves may not all be of one kind, pointing to the analysis of these complex networks as a future direction for archaeological network analysis. In this sense, it might be worthwhile for the person interested in network analysis to consider archaeological work on agent-based modeling, as it seems to me that the two approaches can speak to each other usefully.
To repeat; this is an important book. But it is not an easy book. It demands close attention, and familiarity with many of the terms and issues of network analysis. If the reader perseveres, she or he will find much stimulating material to engage with in all of the papers, across cultures and across time and space. It is appropriate to conclude with van der Leeuw’s parting remarks:
“…so far [archaeologists] have only explored and exploited the tip of the iceberg of the network perspective on society. But if the signs do not betray me, that exploration promises to be a fascinating and very rewarding trip that may bring our discipline much closer to a truly dynamic understanding of the past societies.”(346)
Notes
1. T. Brughmans (2010), Connecting the dots: towards archaeological network analysis, Oxford Journal of Archaeology 29(3), 277-303; (2012), Thinking through networks: a review of formal network methods in archaeology, Journal of Archaeological Method and Theory 19(2) doi:10.1007/s10816-012-9133-8; (2013), Networks of networks: a citation network analysis of the adoption, use, and adaptation of formal network techniques in archaeology. Literary and Linguistic Computing: The Journal of Digital Scholarship in the Humanities 28(4), 538-562; C. Knappett (2011), An Archaeology of Interaction: Network Perspectives on Material Culture and Society. Oxford: Oxford University Press.
2. I. Malkin (2011), A Small Greek World: Networks in the Ancient Mediterranean. Oxford: Oxford University Press.