Networks, case studies and the big picture: some reflections

Following on from our successful conference last month, News Networks is busy once again, this time in producing a two-volume edition that aims to re-evaluate the history of news in Europe. The aim of the project overall could be summarised as forging its own network in order to link and so affect scholars working in the field, discussing shared problems and different methods in order to come up with genuinely new approaches and cast light on the old.

One of the minor difficulties involved in writing about the News network project has been the proliferation of the word network. We’re a scholarly network looking at early modern news networks, with some using the ideas of network theory and some the technology of network analysis to make sense of them. This is not just a stylistic coincidence, and it’s provoked me to reflect a little on the importance of the term network in the constitution of our scholarly collaboration, our subject, our method of approaching it, and the tools we have for analysing it.

Instead of recapping forty papers from the conference (for a snappier, more organic review see our Storify) or detailing the volumes, I wanted to offer some personal reflections on methods and approaches that have featured in both. The volumes themselves speak to a division and its potential problems: the first is roughly methodological and the second focused on case studies. When dealing with essays of 8000 words or so this isn’t always a straightforward distinction, which makes me think of a discussion that has arisen a few times in the project’s life: the relationship between the ‘big picture’, the isolated case study, and computer analysis.

Some questions come to mind. What are our options when wanting to find out about any given subject? Is there anything between the grand narrative and the detailed case study? Is the point of the latter to help find the former, or should it stand alone, better reflecting the complex and irreducible nature of the ‘big picture’? Further, what do we mean by ‘big picture’; do we mean something fundamentally simply or fundamentally complex, or does this depend on whether we think we can capture it?

When computers and the so-called digital humanities come in, we’re confronted with a similar choice: do we want to focus in on the detail, the story, or have big data show us overarching themes, maps and networks? During the conference, it was pointed out that to think of these as oppositional is misleading; they can be complementary.
Taking the first pairing, perhaps they are not as remote from one another as first thought : the case study contains and contributes to wider narratives, and the wider narratives carry exemplars and the sometimes invisible support of myriad smaller details that can be excavated at will. I would say that the same can be said of computer analysis: it is not alien to the narrative case study, it is just one of several tools that contribute to writing history. We can use data analysis as well as narrative writing to uncover key trends, to build a case study, or to do both at once.

Likewise, digital possibilities should not be posited as the saviour of truth in history; there won’t be a computer programme that we can put all our collective data into and churn out the exact picture of the past, because it is not lying there waiting to be discovered. But it would churn out something new, new angles and a new approach, or promote for further study aspects that previously had seemed marginal, or had been hidden entirely.

It struck me that there are several things one could be talking about when using the term ‘network’ in relation to this or any other project. Firstly, there’s the group itself, the network of scholars that we’ve created. Then there are the networks we’re studying, the early modern connections that facilitated the transfer of news, thinking about postal routes and social groups (political, mercantile, confessional and so on). This, when considered in terms of network theory, divides again: we can think about these early modern networks in light of the ideas of network theory, so applying to them what we know about modern social networks (their non-random nature and tendency to cluster, the importance of hubs, that there are factors other than distance that influence the shortest routes across the network) and/or we can conduct network analysis on our historical data , like Ruth Ahnert’s impressive data analysis and visualisation of an underground network of Protestant correspondents.

So, you don’t need a computer science degree to get networks: you can start thinking in terms of them without doing computer analysis. A popular introductory text is Linked: The New Science of Networks, by Albert-laszlo Barabasi, and Franco Moretti’s Distant Reading and Matt Jockers’ Macroanalysis also have relevant chapters (with thanks to Ruth for the recommendations).

To start exploring data you need two things: the data itself (in a useful and valid format) and questions you want answers to. Exploring pivot tables in Excel is a good start for getting used to manipulating data, and seeing how it responds as and when you change your terms. Then there’s a variety of data analysis and data visualisation open source software, including Gephi and OpenProcessing (there are also massive corporate tools, for example Palantir, used a lot in law enforcement, and i2 Analyst Notebook, as used in the TV series The Wire).

One interesting DH project on the horizon will make use of the Centre for Editing Lives and Letters’ Correspondence of Thomas Bodley digital edition: Joined Up Early Modern Diplomacy. Dr Robyn Adams is starting the second phase of research on a database of letters to and from Bodley, the first phase of which involved myself as research assistant and Dr Adams as editor transcribing and putting online over a thousand of his letters. I’m going to end this blog post with a few thoughts on this data.

Exploring the data in the aforementioned excel pivot tables allows you to quickly and easily summarise specifics. For example, using this I can create a couple of charts that show how many letters Bodley received from each of his contacts, and how many he sent.



Interestingly, these appear to follow something called a power law distribution, where there are a few massively key players, which decays to many more contacts who play much smaller roles. This is as opposed to a bell curve, for example, which indicates that what’s being measured is random, with a common middle range, such as people’s height or IQ – there is a norm. A power law distribution is an indicator that the group is not random. It is what we see when we look at social networks, and can also be understood in terms of the perhaps better known 80/20 rule. This is a phenomenon that crops up repeatedly in the social, natural and economic world: that 20% of activity tends to result in 80% of the outcome. The figures for the Bodley letters seem to behave similarly: 18% of writers to Bodley account for 80.2% of the letters he received.

In social networks, these important people are hubs – those who are connected to much more people than most, and help facilitate interaction across the whole network. It seems that there are always some people who are hugely more influential and well-connected than others, and they don’t just have to be those at the top of the social or political tree, either.

Whilst it’s gratifying that the graph reproduces something we would expect, and essentially that we already know, it’s only the tip of the iceberg. The next phase of the project is to represent some of the data visually, to ask it new questions and raise new points of interest by presenting the information in new ways. Perhaps the project will come up with a time-delay animation of the people and places Bodley writes to, showing the changing relationships and importance of different hubs on the political scene. Perhaps it can track the movements of the bearers of letters across the geographical and social landscape.

To come back to the earlier point about case studies and wider pictures, the Bodley correspondence is limited by what defines it: it is the correspondence surrounding one man, and so begins as a partial network. Can its specificity be a positive, though? Its criteria mean that it has boundaries, and so can be held in one place and looked at from all angles. Perhaps, just as in a power law distribution, there is no norm in the relationship between big picture and specific case study, and so we cannot take them as representative of the whole. But perhaps that is the wrong way to think of it: the whole is a complex set of constantly moving relationships and processes, and each case study, analysis and narrative recovers one version of what went before.

Elizabeth Williamson

The European newspaper?

What is a newspaper? What is Europe?

These are thorny, and potentially interlinked questions. We know what a newspaper is in its modern form — though it can be surprisingly hard to articulate the implicit knowledge that shapes our expectations of newspapers and instructs us in how to read them — but the nature and role of printed news in earlier societies is harder to establish.

Harder, because the definitional criteria we use do not always mesh perfectly with the historical facts, or the stories that we tell about early printed news. We might propose, for example, that a transhistorical definition of the newspaper would rely on at least the seven following criteria:
1. regular periodicity (i.e. exact frequency of publication, whether that is daily or weekly)
2. seriality (several issues appearing separately, each intended to follow on from the previous)
3. numbering — seriality has to be indicated typographically to assist in consecutive purchasing and collecting. This also implies:
4. continuity in physical appearance and title
5. a heterogeneity of news, from different sources (perhaps including both foreign and home news)
6. topicality of content
7. publication (i.e. making available to the public, not exclusively by printing).
These criteria for form and content represent one possible, minimal definition of the newspaper. An eighth, possible criterion, though it may be no more than a description of a tendency, is that the newspaper must contain some admixture of three kinds of content: news, advertising, and editorial. These criteria would provide a reasonable basis for a rigorous account of the history of early printed news.

But these bare formal criteria, though (fairly) uncontentious in themselves, begin to come under strain when tested against the sheer breadth and multiplicity of contexts in which early modern newsprint emerged and upon which it acted. The history of news, written by practitioners in various European countries, wrestles with such criteria, and bends them to tell a better story, a story that then begins to twist the evidence in favour of clear milestones. Who wants a history full of qualification and muddy waters?

Historians of news are confronted by problems of scale as well as definition. Thomas Carlyle wrote (in the 1840s) of the superabundance of pamphlets and news and printed evidence that the C17th bequeathed to research libraries: “Dreariest continent of shot-rubbish the eye ever saw. Confusion piled on confusion to your utmost horizon’s edge: obscure, in lurid twilight as the shadow of Death; trackless, without index, without finger-post, or mark of any human foregoer;—where your human footstep, if you are still human, echoes bodeful through the gaunt solitude, peopled only by somnambulent Pedants, Dilettants, and doleful creatures, by Phantasms, errors, inconceivabilities, by Nightmares, pasteboard Norroys, griffins, wiverns, and chimeras dire! There, all vanquished, overwhelmed under such waste lumber-mountains, the wreck and dead ashes of some six unbelieving generations, does the Age of Cromwell and his Puritans lie hidden from us.”

In order to fit with modern historiographical standards the history of news must be written with a keen eye to evidence, not only the evidence of surviving printed items, but printing house practices, distribution methods, the commercial rationale, manuscript news, oral traditions, reading and listening practices, etc. etc. Because printed news touches upon so many elements of society and everyday life, because its history is so interconnected with other histories, it needs to be reconstructed minutely and meticulously. And yet printed news, as it is understood today, plays an important role in the emergence of nations, of the public sphere and public opinion, of national identity, of democracy; and so it needs to be written from a geographically and socially broad perspective, and over the longue durée. Newspaper historians must be both foxes and hedgehogs, in the dichotomy of Archilochus, knowing both many things and one big thing.

The problem is exacerbated from a European perspective. Between 1500 and 1700, European countries developed new means of producing and distributing news. These news media depended on communication that crossed linguistic, religious and geopolitical boundaries. Emergent forms of news were crucial to the modernization of European states, the appearance of modern politics, the evolution of discrete identities, and the development of national consciousness. Histories of news have almost exclusively been written from national perspectives; yet the emergence of news media took place across Europe, and in every nation the news media was closely intertwined with pan-European channels of communication, international trade networks, and war. Newspaper historians must recognize the inter- and transnational nature of newspapers; yet to do so, and to do so with the necessary scholarly intensity and evidentiary integrity lies beyond the capacity of an individual.

The Early Modern News Networks project — based at the University of East Anglia, but involving scholars from across Europe, and funded by the generous support of the Leverhulme Trust — seeks to resolve some of these issues of history and method. We seek to explore ways of writing this history collectively and collaboratively. The network of scholars will examine the movement of news across Europe, the forms it took, the routes it followed and the speed and accuracy of transmission, and how it was transformed as it moved between cultures, languages and religions, and by doing so pursue methods for understanding news and news networks in a pan-European context. In this blog we will look at some of the issues, present some of our findings, discuss topical issues through the perspective of the past, ask questions about the nature of news and of Europe, and hope to stimulate discussion of the role of news in early-modern Europe. Let us know if there is anything you would like to see.

It is a pressing project. Just as Europe made news networks, news networks made Europe. Europe was an idea, shaped by war, trade, languages, religions, and by communication. Today Europe — in the distinctive shape and nature that it has assumed since the Second World War — is imperilled. The increasingly formalised economic and legal ties that began with the creation of the European Economic Community and resulted in monetary union are threatened by the actions of a number of under-regulated financial institutions and individuals. Yet beyond these economic and legal frameworks, Europe has an identity that lies in the communicative networks that began centuries before Immanuel Kant’s Perpetual Peace (1795) and the League of Nations (1920): one answer to the question ‘What is Europe?’ lies in the history of news.


Antwerp Meeting

The Plantin-Moretus Museum in Antwerp will play host to the first meeting of News Networks in Early Modern Europe.  (We’re very grateful to the Museum for making their facilities available to us, and giving us such a fitting backdrop to our research.)  As well as papers from network members, including Carmen Espejo and Paul Arblaster, the workshop will feature contributions from the first of many network associates, including Professor Andrew Pettegree, of the University of St. Andrews, and Dr Helmer Helmers, of the Universities of Leiden and Amsterdam.

This is the first of a series of five workshops to be held over the next eighteen months, thematically organised around a loose progression from the general issues surrounding the histories of news and newspapers in early modern Europe, to particular case studies of specific channels of news transmission, sites of news production, and news events.  The Antwerp meeting will be organised around discussions of media and methods, including questions of definition (what constitutes newsprint in the period?), existing approaches and pitfalls in the history of news, and locality.  More details, and news of additional contributors, to follow…

Early Modern News Networks

Welcome to Early Modern News Networks, a blog featuring events, updates and matters of interest arising from the Leverhulme-funded project “News Networks in Early Modern Europe”.  The project brings together its own network of five leading scholars of early modern news culture from across the continent to establish new approaches to the study of news networks, their formation and their functioning, that will lay the foundations for a methodologically coherent European history of news and newspapers.

Participating in the network are:

Professor Joad Raymond (University of East Anglia)

Dr Paul Arblaster (Zuyd University, Maastricht)

Dr André Belo (Université Rennes 2)

Professor Carmen Espejo (Universidad de Sevilla)

Professor Mario Infelise (Universita Ca’ Foscari Venezia)

The project will also feature contributions from guests and associates at its various meetings over the next two years; the network will hold a series of workshops in five European cities, each an important centre of early modern news production and distribution, culminating in an open symposium in London in the summer of 2013.  The first meeting of the group takes place in Antwerp in November, and the project website is currently under construction.  Watch this space for news and announcements!