Networking the Republic of Letters

Abstract

In recent years it has become common to speak about the republic of letters as a network. But this was not always the case. Rather, it is the product of a specific set of conditions: the confluence of readily available digitized documents, computational power to analyse that data, and a ready acceptance of the ‘network perspective’ in the popular consciousness. In our increasingly interconnected world we encounter networks at every turn. The Internet, public transport networks, and power grids make our everyday lives possible; our careers are dependent on networking; and social networking sites provide an online account of our professional and personal capital. Networks have become a metaphor for connectedness, but also a concrete framework for visualizing and measuring complex systems of knowledge in the era of big data.<p>

<p>Although scholars working in the humanities might not realize it, the network turn is due to the emergence of ‘network science’ as a field of interdisciplinary study. In a series of key publications in the late 1990s and early 2000s, scholars such as Albert-László Barabási, Reka Albert, Duncan J. Watts, and Steven Strogatz showed that a huge variety of real-world networks – such as, for example, neural networks, transport networks, biological regulatory networks, and social networks – share an underlying order, follow simple laws, and therefore can be analysed using the same mathematical tools and models.1 These publications build on work from various different disciplines, such as sociology, mathematics, and physics, which stretches back some decades; but the emergence of network science as a field in its own right was the product of certain conditions that did not exist before. Barabási and Albert explicitly cite the computerization of data acquisition as essential to their research. In other words, what they needed was numerous examples of big network data, which they could compare, and the computational power to analyse that data. In this field, thousands of publications every year describe the development of new quantitative network analysis methods, and the analysis of new types of network data.<p>

<p>The advent of large-scale digitization efforts in the humanities has given scholars unprecedented access to their research materials. Perhaps more importantly, however, it has also put quantitative analysis methods within the reach of this community. This is particularly true of large collections of metadata, as these represent structured information that is easier to abstract and quantify. Correspondence metadata, such as the data collected by the constituent members of the COST Action <i>Reassembling the Republic of Letters</i>, lends itself particularly well to quantitative analysis, as it is exactly the kind of data that network analysis was designed to study – a set of well-defined relationships, namely letters sent and received, between well-defined entities, namely individuals. As discussed in chapter II.4, some work may be necessary to establish the identities of the individuals, but correspondence is a social relationship that is particularly clearly defined, due to its physical manifestation in the form of the manuscript letter.<p>

<p>The value of the COST Action <i>Reassembling the Republic of Letters</i> additionally relies on a ‘network effect’ – a term employed in the context of modern technology companies, which means that the value of a software product rises with the number of people using it, as such products typically facilitate interactions between users in some way. By combining the metadata of a wide range of historical correspondence projects, and by making them compatible with each other, their combined value to the scholarly community is greatly increased. Consistent metadata allows for much more wide-ranging searches across correspondence collections, and the power of quantitative network analysis grows rapidly with the size and scope of the network under study.