The historians of the future will face problems which, if we don’t equip them with the right tools, they will struggle to tackle. Instead of struggling with patchy records, they will be sifting through masses of data looking for patterns and trends.
When did the Industrial Revolution really begin? Why did the German public elect Hitler? Why did the stock market crash in 1929? Such vexing questions are prime examples of questions historians struggle with today, and they are all difficult to answer due to the ambiguity created by a lack of evidence and data.
Historians’ current methodologies have been constructed to deal with a lack of complete information. Sources with known biases are used, and the methodologies, through source criticism, attempt to mitigate the impact of these biases and establish a plausible narrative, with varying degrees of success.
Economic historians, for instance, are limited in their ability to answer fundemental questions about the industrial revolution due to a profound lack of economic data. Estimates of GDP per capita, unemployment and wage levels are made on the basis of patchy price data and can vary according to the technique used.
In the future, however, historians will not struggle with a lack of data. The calcuation of basic economic indicators like GDP, unemployment and inflation has become standard practice across the world, meaning economic historians seeking, for instance, to explain the causes of the 2008 financial crisis, will have much more data to evidence their arguments. The emergence of mass data collection fundementally transforms the mission of historical methodology; instead of trying to fill in the blanks left by missing data, it must analyise large datasets and identify overarching patterns and trends. In order to do this, however, historians must be equiped with the appropriate skillset to manage large swathes of data.
This need is not only limited to economic historians; social historians looking to sift through the billions of social media posts and messages that may be available to historians in the future would greatly benefit from the emerging AI and machine learning technology that can facilitate the detection of patterns in datasets of this size. This is not a new idea either; this team from the University of Bristol used AI to analyse the news from 100 different British regional newspapers over the past 150 years, and managed to pinpoint the exact years when, for instance, mentions of electricity overtook those of steam (1898).
Data regarding demographic issues such as population growth rates, rates of birth, death, marriage and disease, occupational and education distributions, and migrations and population changes will all allow historians to constrcut much more elaborate and nuanced narratives. The rapidly advancing nature of AI technology means we cannot be certain what the technology will and will not be able to do, but there is no doubt that computers’ capacity to identify patterns and trends in data will develop. Failing to provide the historians of the future the ability to exploit this new technology will slow the discipline’s progress.
The history of historians’ favoured methodologies already suggests a transition to more quantitative and rigorous techniques, away from the vague pontification and speculation of our historian forebears. The emergence of social and economic history in the 70s necessitated the adoption of new technqiues, usually borrowed from the social sciences. Indeed, in the history departments of British and Irish universities in 2014, 26% identified themselves with social history while political history came next with 25%. This led to the introduction of such sub-disciplines as labour history, gender history, ethnic history, historical demography and a plethora of others. There is no reason why historians cannot again, in the face of technological developments, reinvent their methodologies to utilise these new analytical tools.
If we truly are to help the historians of the future develop a more complete and accurate understanding of the past, we must equip them with the appropriate quantitative tools to do so. To not modernise our methodologies is to rob future historians of the opportunity to gain unique insights into our period and learn from our mistakes.