Page - (000095) - in Knowledge and Networks

87 Network as a Fashionable Metaphor Not surprisingly, the term network is polysemous. Its three most widely accepted definitions can be found in current geographical literature: network as a metaphor; network as a technical infrastructure; and—similar to, but distinct from, network studies in the strict sense—network as relational flows between places. The pre- dominant definition of the word is its metaphorical one: A network is a social pro- cess (migratory network) or a system of relations (network of cities) that crosses borders. It is fluid, moving, and dynamic. In this sense the network is always the opposite of an enclosed and stable territory. The study of networks from this per- spective does not involve any methodological choice, and the term network could simply be replaced by another (Glückler, 2013; Grabher, 2006). In political geogra- phy this metaphorical usage is especially strong given the presumed weakening of the level of the state, which is being overwhelmed by transboundary movements and the supranational nature of financial and economic actors. P. Taylor’s publica- tions of the Globalization and World Cities Research Network1 are quite representa- tive of this trend, which moves away from the state level to consider mainly relations among the so-called world cities. Although many of these papers are quite interest- ing from a thematic perspective, network remains an emblematic word that implies neither conceptual nor methodological change. More interesting for my topic is the coexistence of two intersecting research traditions, one dedicated to technical net- works and the other dedicated to flow studies. Two Geographic Traditions: Infrastructure Networks and Flow Studies As for the nonmetaphorical network, two main approaches exist in current geo- graphical research: technical networks and flow studies. Studies of infrastructure networks (transport, energy, communication) were predominant starting with the seminal doctoral dissertation by Kansky (1963) until the middle of the 1990s and remain at the core of network studies in geography today. Although a minority of geographers tried to mix methods and propose new ways to study transportation networks (Ducruet, Ietri, & Rozenblat, 2011; Gleyze, 2007), the canonical tradition using the series of indices by Kansky (different ratios between the number of edges, the number of vertices, and the number of cycles derived from graph theory) pre- vails. It should be noted that in the vast majority of cases, networks are planar— often valued and nondirected—a characteristic that may help explain the lack of methodological dialogue with social network analysis, in which a network is more often Boolean, directed, and nonplanar. Even though some methodological 1 http://www.lboro.ac.uk/gawc/ 5 Studying Networks Geographically: World Political Regionalization in the United…