Quantifying Religion: Messy Data and Scholarly Input in the Database of Religious History

willismonroe's picture

There is a certain amount of privilege in posting near the end of the month, because I get to build on the fascinating and lengthy discussion which has already occurred. My response will be based on my current role as managing editor for the Database of Religious History (hereafter DRH), based at the University of British Columbia.

I would like to start by juxtaposing two visions of the digital study of religion offered earlier by Nathan Schradle and Saikou Diallo. Nathan remarks that scientists have been attempting to “wrangle and contain” the nature of religion (perhaps to limited success), and Saikou agrees that for scientists and engineers religion can be “quite unsettling and foreign.” It is precisely this conflict between the messiness of religious practice and belief and the rigid tools of digital analytics that inspired the DRH to try and collect and analyze the historical dimension to religion. Returning to the two perspectives, Nathan correctly notes that while the volume of data might be large and the speed of creation rapid, the valuable products of interpretation are the real measure of success. Saikou similarly brings up the rather fundamental issue of mapping basic digital data as a spectrum rather than a zero or one.

The DRH attempts to address these concerns by basing our knowledge creation and consumption on existing scholarly practice. We meet the scholars in the medium that they are most comfortable with and represent their knowledge and data in familiar and accessible ways. Our project records data about the practices, beliefs, traditions, and social contexts of religious groups throughout history. Experts answer extensive questionaries that are created through an inclusive process, using our editorial staff to stretch and expand the scope of our questions until we have a document that reflects the current specialization of the project. Finished entries can then be consumed by a diverse audience as web-sites, PDF articles, or discreet data-points depending on the level of specificity or interest. Furthermore, the entire dataset can be visualized and queried in a variety of ways. Scholarly agreement and disagreement is encoded throughout the site by not treating any answer as infallible but rather the answer of a single scholar for a single point in time and space, and duplication and overlap is encouraged. This results in a dataset that is both deep and fluid, since an entry for a single scholar can overlap in small or large ways with another and the overlap of data points can also function as data.

The end goal is to answer questions like the one posed by Christopher Cantwell about religion acting as a social media company which uses the data generated by its believers to be “monetized for the [religion]'s wellbeing.” With a wealth of historical data, at least hints to this type of question are within our reach. In fact some of the founding members of the DRH are interested in precisely this question about religion's role in cultural evolution and the way in which religions can borrow from each other and change over time in response to their ability to support the wellbeing of adherents, attract new members, and compete with other religions. This approach is only one vantage point into the data of the DRH. There are of course numerous other avenues of research in a large dataset created by scholars for both scholars and the wider public. One potential derives from our lack of a definition for “religion” in the project itself; we leave the designation up to our experts as they enter their data. Further down the line we may discover that within the messiness of our digital approach, the data begins to offer a pattern which contains within it the concept of religion.