12.10 Digital Histories

Scholars, be they historians or humanists, are often charged with having a complex about computers. They shouldn’t be. If you examine the histories of fields such as history, classics, and literary studies, you will find scholars from the earliest days of computing who used it to better understand the writings of Saint Thomas Aquinas,^[1] to determine the authorship of anonymous documents relating to the creation of the American constitution,^[2] and to better understand the shape and size of families in Early Modern Europe.^[3] Humanities scholars, put simply, were present at the creation of the computer. There was, however, a common refrain in the early days of digital history: scholars used the computer to manipulate two things, texts and numbers.

And while those efforts produced contributions, the story I want to tell here is one that focuses on historians’ use of forms of expression that are not text and are not number. Here, I’m talking about expressive forms that have two-, three- or even four-dimensions (like, for example, a model of a heritage building). Here, I’m also talking about forms of expression that are dynamic — they move — and do so in a way that is autonomous. They perform behaviours without the direct intervention of an author or programmer. This capability enables scholars to create simulations of historic battles, economies, and even cities.

Now, why would we want to use forms like these? The simple answer is that —  sometimes — different forms of expression can express an idea, an historical event or a pattern more clearly than words or numbers. Consider for example the stock market. Many of us see reports on its daily progress on the news, and we watch its movements because we’re interested in its behaviour. Has it gone up? Has it gone down? Has its behaviour been stable, or have prices veered all over the place? In principle, there are two ways we can communicate that information. We could present our audience with a list of prices. But that option is ultimately not a very good one, in large measure because it forces viewers to look at that list and then visualize the performance of the market in their head. It’s a lot of work. A better solution is to use the formalism — the form of expression — that you see in most news reports, namely the graph. The graph is the better solution because it enables the viewer to obtain a comprehensive view of the market’s behaviour with a simple glance and little cognitive work. In short, then, historians are exploring the use of different forms to help them better teach, represent, and explore the past. They are using computers because doing so makes it easier to create and disseminate those forms.

One means of creating forms is through the use of software known as Geographic Information Systems (G.I.S.). The basic idea behind G.I.S. is simple: it combines two things: lists (otherwise known as databases) and maps. More specifically, it assigns each item in our database-list a spatial location, and plots that location on a map using a pin, a dot, or a polygon. The idea is simple but in practice, it often produces powerful results for historians. Plotting information on a map can often reveal important spatial patterns that can deepen our understanding of social, economic, urban, and environmental history.^[4] Using this method, for example, historians have been able to challenge the received wisdom on topics such as the circumstances behind the emergence of the dust bowl in the 1930s. Traditionally, historians have faulted farmers for the dust storms, pointing to methods of cultivation that were not sustainable. In the 1990s, however, the historian Geoff Cunfer, using G.I.S., was able to demonstrate that the actual cause of the storms was drought, not farmers. In almost every case, the storms emerged in sections of the United States that were not farmed. Hot weather caused the storms, not farmers.^[5] Scholars have used G.I.S. to qualify or challenge our received wisdom on other topics as well. Michael McCormick, for example, used G.I.S. to demonstrate that Europe’s economy revived much more quickly after the collapse of the Roman Empire than historians had customarily thought.^[6] Historians John Lutz and Pat Dunae and their colleagues used G.I.S. to argue that 19th century residents of Victoria were not as racist in their behaviour as historians have usually assumed.^[7]

Historians and historical scientists have also used dynamic, autonomous forms to deepen their understanding of the past, using software that creates historical models known as agent-based simulations. The basic idea behind agent-based simulations is simple. You (as a historian) define the agents you want for your simulation. The agent can be a person, a microbe, or even a multi-national corporation. It can also be all of the above, in which case you have different classes of agents. Once you have defined your agents, the next step is to define their behaviour: to specify how they interact with each other and how they interact with their environment. Once you have stipulated their rules of conduct, your next step is to flip the switch, run the simulation, and see what kinds of patterns emerge. Researchers in the social sciences and the humanities have found these simulations to be extremely powerful because they have produced patterns that match, or nearly match, patterns that scholars have located in data from the past and present. They present a way for us to test theories and deepen our understanding of how historic economies worked, the circumstances leading to the emergence of individual cities and systems of cities, and the trajectory of significant battles in history. One reason that agent-based simulations are important is they provide a way for historians to assess the relative importance of one cause versus another in initiating a historic event.^[8]

Take the Battle of Trafalgar as a case in point; it occurred in 1805 during the Napoleonic Wars and pitted the British Royal Navy against a combined fleet from Britain and France. Britain emerged as the decisive victor. It lost no ships, while the Franco-Spanish fleet lost 22 of the 33 ships it had deployed. In explaining Britain’s victory, scholars have had two possible explanations. First, Britain had bigger and better guns and ships. Second, the fleet’s commander, Admiral Horatio Nelson (1758-1805), was a tactical genius. What scholars have not had is a way to differentiate the relative importance of each cause. In 2003, researchers Giuseppe Tratteur and Raniero Virgilio created an agent-based simulation to see if they could determine which of the two causes was more important. Their simulations suggested that, ultimately, it was Britain’s equipment that was critical. Britain would have won the conflict, even without Admiral Nelson. What Admiral Nelson contributed was the best tactics, methods of fighting that minimized Britain’s losses in equipment and personnel, but he or another commander could have contributed less and the Royal Navy still would have carried the day.^[9]