This paper will argue that archaeological network researchers are not well networked themselves, resulting in a limited and sometimes uncritical adoption of formal network methods within the archaeological discipline. This seems to have followed largely from a general unawareness of the historicity of network-based approaches which span at least eight decennia of multi-disciplinary research. Many network analytical techniques that would only find a broader use in the last 15 years were in fact introduced in the archaeological discipline as early as the 1970s. This paper does not aim to argue that every archaeological network study should include a historiography. It merely wishes to stress the need to explore the full range of existing network techniques and models. I will illustrate that knowledge of the diversity of archaeological and non-archaeological network methods is crucial to their critical application and modification within archaeological research contexts.
This paper will for the first time trace the academic traditions, network concepts, models and techniques that have been most influential to archaeologists. I will do this by combining a close reading of published archaeological network applications with citation network analysis techniques (Batagelj 2003; Hummon & Doreian 1989; White 2011), an approach that has not been applied to archaeological literature before. A citation network was created consisting of over 10,000 publications and their internal citations. This network consists of all archaeological network analysis applications, all publications cited by them and the citations between those publications. This data was extracted from Web of Knowledge (http://wok.mimas.ac.uk/) and manually when the publications were not included on Web of Knowledge.
The analysis revealed a number of issues surrounding the current use of network methods in archaeology, as well as possible sources and explanations for these issues. They include the following: (1) Although many network techniques are rooted in graph theory, archaeological studies in graph theory were not influential at all to more recent archaeological network studies. The introduction of graph theory and social network analysis into the archaeological discipline happened largely independently and, unlike social network analysts, archaeologists did not collaborate with graph theorists to develop mathematical techniques tailored for their needs. (2) The potential of social network analysis techniques was explored (largely theoretically) by Cynthia Irwin-Williams no later than 1977. Many of the techniques she described were not applied in archaeological research until the last ten years, possibly because the limited availability of cheap and potent computing power and large digital datasets in the late 1970s. (3) Some social network analysis techniques (e.g. centrality measures) have received more attention than others (e.g. ego-networks). The popularity of these techniques seems to be related with their use by social network analysts rather than how archaeologists have used them before. (4) The archaeological use of complex network models is largely limited to the extremely popular small-world and scale-free models. Archaeologists have neglected to explore the potential use of alternative complex network models. (5) The availability of user-friendly software seems to determine the popularity of social network measures used by archaeologists. (6) On the other hand, simulations of complex network models by archaeologists are rare due to the required technological and mathematical skills.
The results of the citation analysis expose the insufficiently explored potential of formal network-based models and techniques and points out publications in other disciplines that might have interesting archaeological applications. The paper concludes that in order to move towards richer archaeological applications of formal network methods archaeological network analysts should become better networked both within and outside their discipline. The existing archaeological applications of network analysis show clear indications of methods with great potential for our discipline and methods that will remain largely fruitless, and archaeologists should become aware of these advances within their discipline. The development of original archaeological network methods should be driven by archaeological research problems and a broad knowledge of formal network methods developed in different disciplines.
References
Batagelj, V. (2003). Efficient Algorithms for Citation Network Analysis. Arxiv preprint cs/03090 23: 1-27.
Irwin-Williams, C. (1977). A network model for the analysis of Prehistoric trade. In T. K. Earle, and J. E. Ericson (eds.), Exchange systems in Prehistory. New York: Academic Press, pp. 141-151.
Hummon, N. P., and P. Dereian (1989). Connectivity in a citation network: The development of DNA theory. Social Networks 11(1): 39-63.
White, H. D. (2011). Scientific and scholarly networks. In J. Scott, and P. J. Carrington (eds.), The SAGE handbook of social network analysis. London: Sage, pp. 272-285.
