Louis Bissières
Time and Place: Thursday, 01.07., 10:40–11:00, Room 2
Session: Economic Transactions
Keywords: Economic history; Merchant networks; Temporal approach of network;, Multi-layered networks
Network analysis tends to produce snapshots of structures that can seem permanent, not affected by time or change. Still, the very structures approached by scholars mobilizing these kinds of methods are susceptible to change. Time matters in network history, and the observation of its effects is an issue[1]. As advocated by C. Lemercier, time needs to be integrated in studies about networks since nodes, ties and the overall network structure can be affected by change. In this talk, I will explore the evolution of a merchant network at the end of the eighteenth century. Measuring the evolution of networks through time is part of my broader research on networks of partners as a major resource for Early Modern merchants. My goal is to understand if this resource was stable (in terms of organization and personnel) and how it was managed, hence the study of change in networks.
In this communication, I offer an analysis of a network of transactions, and its evolution through time. My main source is a database that I have built from Levi Hollingsworth’ account books. It covers five years of activity (between the years 1785 and 1795) and includes roughly 28,800 transactions and 3,000 different accounts[2]. The years studied are 1785, 1786, 1787, 1791 and 1795. These sources allow for continuous analysis of the first three years, and observation of two later stages of activity (in which not all new relations can be traced back, but still allow for comparison). Levi Hollingsworth (1739-1824) was a major merchant in Philadelphia in the end of the 18th century. His activities had a wide range both in terms of space and type, and, from the previous stages of my research, it appears they were organized in a specific structure. In this structure, the network of partners was highly hierarchical, and structured around a few merchants with specific, attached clienteles. Most of these central actors were involved in the trade of several commodities (generalists). Unlike specialists, who could rely on their knowledge of one kind of merchandize and its trade, I argue that generalists needed their clienteles to deal with uncertainty on Early Modern markets. I intend to elaborate a longitudinal analysis of this network, to understand what were the changes that affected it.
In this paper, I develop a comparative analysis of five successive networks, one for each year of Hollingsworth’ activity. I consider that one year is an adequate unit since seasonality prevents comparison of shorter periods. My goal is to show the relevance of a series of indicators that can help historians shed light on change in networks. This methodological question is indicative of my broader research agenda to understand how individuals managed the (un)stability of a merchant network. First, I study the evolution of a merchant network’s composition through time, with an emphasis on the turnover of actors and the evolution of their centrality. Despite this turnover, I insist on the stability of the whole structure in the second section of my talk. I especially show how similar the five networks were in terms of their overall hierarchy and segmentation in clienteles. In the last section of the presentation, I push this study further by switching from single-layered to multi-layered analysis. Indeed, the understanding of networks (in terms of structure and, as I intend to show, change) can be deepened if the relations between nodes are specified[3]. In a transaction network, multilayered analysis allows us to distinguish different kinds of transactions (e.g. cash payment, sale of merchandize, book credit). I explore the changes affecting selected layers of the merchant network. I conclude with a discussion about the possibility of comparing temporal layers in the case of an Early modern merchant network.
[1] Lemercier, Claire. 2015. « Taking time seriously. How do we deal with change in historical networks? » In Knoten und Kanten III. Soziale Netzwerkanalyse in Geschichts- und Politikforschung, edited by Markus Gamper, Linda Reschke and Martin Düring (eds.), p. 183-211. Bielefeld: Transcript Verlag.
[2] Historical Society of Pennsylvania, Hollingsworth Family collection (#289). Journal K (1785), Journal L (1786-1787), Journal N (1791), Journal P (1795).
[3] Van Vugt, Ingeborg. 2017. “Using multi-layered networks to Disclose books in the Republic of Letters.” Journal of Historical Network Research 1 (1): 25-51. Dickison, Mark E., Magnani, Matteo and Luca Rossi. 2016. Multilayer Social Networks. Cambridge: Cambridge University Press.