Unraveling the London Moment: Navigating Financial and Economic Experts in Exile through Digital Network Visualization

Examining financial and economic experts during the London Moment, as I intend to do in my dissertation project , is a challenging endeavor. During the time in exile, numerous working groups, committees, and panels were formed to address questions about the post-war economic order. Within the complex exile framework, it´s easy to lose track, overlook key players, and not perceive connections between actors and/or ideas. To simplify this process, I have started creating a digital network visualization, aiming to complement my prior qualitative analysis. Furthermore, I aspire to formulate new research hypotheses, informed by the insights derived from this network visualization.

In this blog post, I will first outline some methodological and theoretical principles of network analysis, then show the initial results of my visualization and explain which further research hypotheses can be derived from them.

First of all: what exactly does network analysis / network visualization mean? Historical Network Analysis (HNA) has increasingly become a buzzword in contemporary historical research publications. The diversity of its applications is vast, and the community of scholars working under the umbrella of Historical Network Research is steadily expanding. However, the term itself often remains ambiguous.

Prominent proponents of HNA, Marten Düring and Ulrich Eumann, advocate for using the concept in alignment with Social Network Analysis (SNA) from the field of social sciences. SNA has its roots dating back to the 1970s, with its own set of tools and software. It employs formal, relational methods to reveal connections between actors.[i] When HNA is applied in this tradition, it distinguishes itself from merely using the network concept metaphorically for describing actor relationships descriptively. Similarly, Claire Lemercier, a sociologist who also engages in network analysis, emphasizes that the value of formal network analysis lies not in concluding that networks exist and are important, but rather in presuming their existence and then meticulously describing their patterns, understanding how they were created, and discerning their consequences.[ii]

Before delving into the practical aspects, it is necessary to provide a brief overview of the advantages as well as potential pitfalls of HNA. It’s worth noting that HNA is not limited to revealing relationships between actors; it can also illustrate relationships between individuals, places, or time points.[iii] Such relations – as who was where at what time, the overlaps in personnel at specific moments among selected groups, or the persistence of certain constellations over different time points – can quickly become highly complex. Software-assisted HNA can assist in visualizing this information graphically. As Lemercier puts it, “Network visualization, as a genuine method of Historical Network Research (HNR), offers the ability to analyze, interpret, and communicate highly complex social structures to the external world in ways that no other tool can.”[iv]

However, it should be critically noted that a graphical representation cannot claim to be exhaustive nor intuitively readable. Instead, a source-critical analysis of the results is imperative. In line with the nature of HNA, it cannot determine intentions behind relationships between actors, locations, or actors themselves. In this context, HNA generates more questions and hypotheses about action and interaction potentials, which can complement and expand qualitative source analysis. Questions like why Actor X didn´t leverage their strong connections across multiple groups to advance their economic ideas or how Actor Y, despite their peripheral position, exerted major influence in the post-war period arise.[v]

Also, manifest interaction and interaction potential, as Lemercier also observes, are not equivalent.[vi] Similarly, HNA cannot clarify the extent to which actors were aware of their respective positions, the connections they perceived, or how they positioned themselves within the web of relationships. Consequently, a data based HNA exclusively functions in close collaboration with my qualitative analysis of the source corpus and is subject to the veto power of the sources.[vii]

Following these theoretical explanations, let’s turn to the practical aspects: As an initial attempt at such data visualization, I used the open-source software Gephi. To do this, I created an Excel spreadsheet containing all the data about the active actors within my selected groups and their connections, which I then imported into Gephi.[viii] The implemented layout algorithm, “Fruchterman-Reingold,” aided me in representing clusters and groups within the overall network. In a first step, I visually modified individual nodes (representing the actors in my project) based on their frequency of connections within the network. They vary in size (from small for less connected to large for highly connected) and color (from dark for less connected to light for highly connected). Additionally, I colored the edges of the network (representing connections and their types) based on how frequently a connection between actors was utilized (within group meetings, etc.), with darker colors indicating higher usage.

Network visualization consisting of the actors from the groups selected for my project and their connections. Created using Gephi.

At the same time, I have already created another version that distinguishes the edges by their type. Since my visualization primarily focuses on state and non-state committees, working groups, and private groupings that I have identified as important for my project, the assorted colors here represent each respective group.

Network visualization in which the individual groups are now color-coded according to the type of connection.

In the future, I hope to identify actors who served as central links between groups (see the example below of Yugoslav agricultural politician Rudolf Bicanic) and to highlight actors who, while not necessarily having a high number of connections (i.e., not frequently connected), simultaneously possessed particularly strong connections. At the same time, I intend to utilize an advanced qualitative analysis to illustrate which actors, in line with their position within the network, were able to disseminate ideas and why some were unsuccessful in doing so.

Connections of Yugoslav politician Rudolf Bicanic within the overall network of the selected groups.

I look forward to continuing to work with my dataset and am eager to uncover further insights. In the next steps, I will complete the visualization and conduct an analysis of it.

[i] See Düring, Marten / Eumann, Ulrich: Discussion Forum Historical Network Research. A New Approach in Historical Studies. In Geschichte und Gesellschaft, 39, 2013, pp. 369-390, p. 370. Since the 2000s, a relational approach has been established in German-speaking historical research, utilizing the methods of Social Network Analysis (SNA). Important contributions in this phase include works by Carola Lipp on political culture in the city of Esslingen and during the 1848 Revolution, as well as investigations by Wolfgang Seibel and Jörg Raab on persecution networks during the National Socialist era. Recent efforts are slowly attempting to establish a common standard and provide an overview of the current state and applications in research. Perhaps the most significant publication in this regard is the Handbook of Historical Network Research – Foundations and Applications (edited by Marten Düring, Ulrich Eumann, Martin Stark, Linda von Keyerlingk), which was published in 2016. Simultaneously, an essay on the Formal Method of Network Analysis in Historical Studies by Claire Lemercier, published in the Austrian Journal for Historical Studies in 2012, as well as the above-cited essay by Marten Düring and Ulrich Eumann, are central to this discourse. These works collectively advocate for the use of HNA in alignment with SNA principles.

[ii] Lemercier, Claire: Formal Methods of Network Analysis in Historical Studies: Why and How? In: Austrian Journal for Historical Studies, 23, 1, 2012, pp. 16-41, p. 22.

[iii] Düring / Eumann: Discussion Forum Historical Network Research, p. 370.

[iv] Lemercier, Claire: Formal Methods of Network Analysis in Historical Studies: Why and How?, p.30.

[v] See Düring, Martin / Keyserlingk, Linda von: Network Analysis in Historical Studies. Historical Network Analysis as a Method for the Study of Historical Processes. In: Schützeichel, Reiner/ Jordan, Stefan: Processes. Forms, Dynamics, Explanations. Wiesbaden 2015. pp. 337-350, p. 347.

[vi] Lemercier, Claire: Formal Methods of Network Analysis in Historical Studies: Why and How?, p.30.

[vii] For further information on the veto power of sources, see: Jordan, Stefan: Veto Power of Sources, Version: 1.0, in: Docupedia-Zeitgeschichte, 11.02.2010. Link: https://docupedia.de/zg/Vetorecht_der_Quellen. The visualizations presented below should be understood as an initial attempt at visualization and are by no means finalized.

[viii] The visualizations shown below are just a preliminary attempt and do not claim to be complete.

Cite this blog post
Lea Levenhagen (2023, November 16). Unraveling the London Moment: Navigating Financial and Economic Experts in Exile through Digital Network Visualization. The London Moment. Retrieved May 28, 2024, from https://doi.org/10.58079/onnl

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Search OpenEdition Search

You will be redirected to OpenEdition Search