This volume's objective is to advance applications of network thinking in archaeological studies of seaborne communication and exchange. The broader aim, stated by the editors in the informative Introduction, is to promote flexible frameworks, methods and theoretical approaches for variable data scenarios and epistemological positions. This is achieved by presenting different analytical models to discuss patterns and scales of human mobility in the Mediterranean.
The volume originated from a 2013 University of Toronto workshop, ‘Networks of maritime connectivity in the ancient Mediterranean: structure, continuity and change over the longue durée’. Braudel's influential view of the Mediterranean as an interconnected whole, to be approached in diachronic perspective, is noted explicitly only in the bookend chapters of the resulting volume. More common threads running through the chapters are studies that build on Braudel's approach: Horden and Purcell's The corrupting sea (Reference Horden and Purcell2000) and Broodbank's The making of the Middle Sea (Reference Broodbank2013). Also highlighted are analytical approaches recently applied to Mediterranean archaeological datasets: the mobility turn, connectivity and small-world networks. Eight case studies progress from the Bronze Age to the early Byzantine period (seventeenth century BC to eighth century AD), with the North-eastern Mediterranean receiving the bulk of the attention.
Tim Evans's and Ray Rivers's chapters provide the immediate theoretical background for an application of modelling networks, from scale, sites and vertices to interaction types, distances and edges. Evans's technical presentation lays out the basics of gravity models, proximal-point analysis, maximum-distance networks and a stochastic model, ariadne. Rivers then examines the diachronic effects of the Thera eruption in the Bronze Age Aegean using many of the approaches introduced in Evans's chapter, broken down into ‘most likely’, ‘easiest’ and ‘best’ networks to present five pre- and post-eruption models. Here, agency, or the lack thereof, is critical to defining, within models, the underlying assumptions as to why networks develop.
Thomas Tartaron sets data at the forefront in his review of social network analysis models of maritime networks and raises methodological problems: how to use archaeology, distances, geomorphology, winds and currents to reconstruct parts or characteristics of a network. Conceptual problems in configuring networks leads Tartaron to argue for models of nested geographic scales and he demonstrates the utility of this approach for interpretation of the changing locations of Bronze Age harbours and anchorages in the South-western Aegean.
The role of religion within the structure and management of economic interactions is not often addressed in discussions of network analyses. To test this as a possible framework, Barbara Kowalzig applies small-world dynamics and the role of preferential attachment to examine how religious-economic networks might have fostered a ‘cabotage mentality’ within the Classical-period coastal sanctuaries, linked with Artemis, of the Western Aegean and North-western Mediterranean.
Whereas the preceding chapters (2–5) focus on terrestrial sites, the next four focus on data deriving from ships’ cargoes, from wrecks and on land. By looking at shipwrecks dating to the seventh to sixth centuries BC from the Northern Mediterranean, Elizabeth Greene, in a very accessible chapter, demonstrates how these data can be utilised as nodes or edges related to terrestrial and other shipwreck sites within broader economic and social network models. Greene establishes ‘spheres of activity’ or ‘zones of interaction’ through social network graphing, using the Gephi program to arrive at an ego-network as an effective approach for interpreting the ship as an individual within a realm of immediate contacts.
More abstract is Mark Lawall and Shawn Graham's contribution on consumers’ selection of transport amphora shapes in the Aegean from the Archaic to Hellenistic periods, and potters’ responses to that demand. The authors’ aim is to assess how generic network models of preferential attachment and small-world networks can be reinterpreted to arrive at qualitative conclusions about hypothetical amphora-related trade in antiquity. Two approaches within the Netlogo Model Library are tested: the mimicry model (usually used to model the dynamics of birds and butterflies) and the stylistic change model (usually applied to language evaluation of competing grammars).
Tom Brughmans’s useful contribution focuses on the negative: demonstrating an approach in network modelling that does not work and explaining why. Distribution patterns in the Roman Eastern Mediterranean of Eastern sigillata tablewares, grouped in 25-year periods from the second century BC to the second century AD, serve as the case study. Using the commonly applied Brainerd-Robinson coefficient for similarity networks (to identify similarity of forms), Brughmans argues that the modelling does not reveal any new information that could not have been achieved by other analytical methods.
Paul Arthur, Marco Leo Imperiale and Giuseppe Muci investigate the distribution of the dominant globular amphora within the structure of eighth-century AD Byzantine Mediterranean exchange networks. Using contemporaneous artefacts, the authors first establish links using a Gephi program tool of affiliation networks, and then map globular amphora distribution within these. Amphora shapes became more homogeneous (globular) over time, probably due to the increasing economic isolation of the Byzantine Empire in the eighth century. The authors note that in the following centuries, when economies became more diversified, so too did amphora shapes.
The concluding chapter, written by an archaeologist who specialises in the American Southwest, is an excellent overview of social network analysis, and of the volume. Barbara Mills raises some instructive points about maritime networks and datasets in general. In comparing terrestrial and maritime networks, speed of transportation on land vs at sea is highly relevant. Recovery issues within the archaeological record are significant, as ships, with their snapshots of entire cargoes, are very different to the archaeological remains normally encountered on land. There are a variety of approaches represented in the volume, and some chapters are not explicitly ‘maritime’, such as the analysis of cargoes after they had arrived at their destinations or situated within harbours. Mills underscores that this is one of the volume's vital aims, to ask whether maritime networks should be seen as fundamentally different to terrestrial networks. The range of topics addressed in this instructive book illustrate that social network analysis, and how archaeology can be applied to make these connections, is applicable to a variety of situations, both on land and at sea. In this sense, Mills harkens back to Braudel's longue durée; in the Mediterranean, it may not necessarily be relevant, or useful, to separate terrestrial and maritime networks.
