May 23 2017

Making Religious History Digitally Native

Interview with Edward Slingerland, Project Director, Database of Religious History

screen shot from an entry in the Database of Religious History

The University of British Columbia, since 2016, has been working on the creation of the Database of Religious History, a crowd-sourced, interactive, dynamic, and searchable encyclopedia. The digital humanities project is premised on the Big Data approach—a comparative methodology popular among historians, linguists, and anthropologists, but as of yet, less prevalent with scholars of religion.

In an e-mail exchange, DRH director Edward Slingerland talked to RSN about the project’s theoretical groundings, obstacles to success, and the promise of data collection and the comparative approach to discovering some of the biggest questions in the development of religious practice across history and geography.

There’s digital humanities, and then there’s digital humanities for religious studies and religious history. What’s the discipline-specific theoretical grounding of the Database of Religious History (DRH)?

I see most digital humanities tools as subject-matter neutral. The formulation of the specific questions of the various DRH polls, and the variety of polls we will eventually create, is obviously bound up with discipline-specific assumptions about relevant and useful concepts and categories. 

We also think that the DRH, as a technological platform for gathering and sharing scholarly knowledge, is a major contribution the digital humanities, and we plan to make the platform itself available for other research groups to pursue their own projects.

While users can use the DRH similarly to an encyclopedia—that is, looking at individual entries—the database also reflects a Big Data approach to comparing and understanding large scale patterns or trends in the historical record.

Yes, although you can browse individual entries as in an encyclopedia, the DRH is actually a relational database with all of the powerful functionality that comes with the ability to manipulate data on a large scale. Answers to the various poll questions are ultimately grounded only in space and time, which allows users to analyze answers to specific questions within certain date ranges and geographical areas, to correlate answers with other types of geo-spacial data, or to visualize DRH data in a variety of ways.

Currently the DRH consists mainly of entries organized around “religious groups,” which are given labels and defined in space and time. Within any given entry, the expert may also be linking to external web resources such as references, primary texts, images, et cetera. Soon we’ll be introducing new types of polls organized around other ways to slice up the historical record, such as individual supernatural beings, particular places (temples, archaeological sites, landscapes), single texts or groups of texts, and rituals. No matter what the unit of analysis, the questions and answers (quantitative content) and the comments, links to images, texts, references (qualitative content) can all be analyzed in an extremely flexible way.

We’re planning projects where we’ll be using DRH entries as gateways into other qualitative data about religion on the web, creating a situation where the DRH is at the center of a rich ecosystem of qualitative, digital data.

What are the features that distinguish the DRH from online encyclopedia?

It is crowdsourced in the sense that multiple experts contribute overlapping entries, and we do not force consensus on any topic, but allow the full range of scholarly opinion to be displayed. In, say, the Stanford Encyclopedia of Philosophy, you have single entries on a given topic, by a single author, and are really only getting his or her view. The fact that experts [who write entries for the DRH] can volunteer themselves from anywhere in the world and work in a variety of languages really levels the playing field in terms of say, a qualified scholar who only works in Chinese having the same degree of input as her colleague at Harvard. Individual entries can be commented upon by other experts, creating potentially long conversations (and a string of alternate answers) that we think will give a much more accurate picture of the state of the field than any individual, single-authored (or even multiple-authored) encyclopedia entry. 

Another great strength of the DRH over a standard online encyclopedia is that it is fundamentally a database. Although entries are organized around a given unit (currently, “religious group”) with various tags attached to it, the answers themselves are grounded ultimately in space and time. This means that users can explore the DRH in ways that completely ignore the tags or labels assigned by experts, opting instead to coordinate the information on religion in the DRH with all sorts of other geo- and temporally-located data. So, although you can use the DRH like an encyclopedia in browse-mode, it’s actually a whole lot more. We’ll soon be rolling out powerful visualization and analyze tools in addition to our current browse and search functions. 

In addition to the comment boxes and ability to link to text and images, which resembles what people expect from an online encyclopedia, the DRH is built around a questionnaire that asks experts to answer "yes," "no," or "field doesn't know." These choices create an entire quantitative dimension not possessed by any encyclopedia, and they allow all sorts of powerful visualization and analytic functions. Especially given the rate at which academia is expanding and the difficulty of keeping up with all the relevant qualitative material out there, this quantitative aspect allows one to get an instant snapshot of the state of opinion on a topic. 

Aside from the collection of data, is there a bigger question that the DRH is trying to answer?

The central research question that led to the DRH was the hypothesis that the rise of large-scale societies has been associated with certain forms of “prosocial” religious belief and practices, such as belief in powerful, moralistic and punishing god(s), costly displays, supernatural surveillance, rituals that involve synchrony and arousal, and moral realism, to name a few. Once the DRH grows large enough, we still hope to be able to pit this hypothesis against competing ones—for instance, that it was primarily technological, agricultural, or legal innovations that can account for large-scale societies, or simply a matter of appropriate climates combined with luck of the draw in terms of local species types. Other research teams will no doubt draw on DRH to answer all sorts of questions about the dynamics of human religiosity.

On this point, The DRH is built with the comparative model of religious studies in mind.

Yes, the very raison d’être is comparative work, and not just between “religions” (groups, users of a temple, inhabitants of a certain landscape), but between religious belief and behavior and other phenomena, such as politics, law, technological, economics, social norms, even climate. I've personally written extensively why comparative work should be (and really still is) at the heart of the academic study of religion, and we’re hoping that the DRH as a tool can help to advance the comparative project.

What do you say to critics who advise against the comparative methodology in religious studies?

I think they are deeply mistaken. In my broader work, I’ve been arguing for a long time that, in our reaction against naive, culturally myopic grand theories developed in the early decades of our discipline, we’ve wildly overcorrected, fetishizing difference and remaining trapped in a model of strong cultural constructivism within which comparative work makes no sense. Thick description and cultural nuance are obviously important, but it’s hard to see how a process of endless thick description ever leads to an interesting or coherent research agenda. I think that embodied cognition gives us a coherent, empirically plausible framework for doing comparative work, and that comparative religion really should be the bread and butter of religious studies as a discipline. 

In terms of defining “religion” (typically one of stumbling blocks cited by those suspicious of comparative work), we don’t. We have a questionnaire, and if a prospective expert finds that they can answer some or most of these questions with regard to the group (or place or text or ritual) that they have in mind, then it’s “religious” for our purposes. If it quacks like a duck…

More narrowly though, the DRH emerged from academics focusing on the cognitive study of religion (CSR).

The DRH got its start as an initiative of the international research network known as the Cultural Evolution of Religion Research Consortium (CERC), and its original purpose was to gather data about the religious historical record that could be coordinated with other data on polities, population movements, climate, et cetera, in order to test hypotheses about cultural evolution against the historical record.

It can still be used for this purpose, and many of the questions in the Religious Group Questionnaire reflect the concerns of scholars approaching religion from the perspective of CSR. It’s since evolved into something for much more general use, and the modifications made to the Religious Group Questionnaire, as well as the introduction of new questionnaires, reflect this expansion in purpose and usefulness.

DRH contributors are intended to submit entries based on their judgment, and citations are not required. Tell me about where in someone’s research process you imagine the DRH fitting in.

The idea is that a DRH entry is like an encyclopedia entry, where an expert in the field gives us their summary of what they think the state of the art of scholarship is. Citations are not required, but encouraged; in the end, the expert him or herself is the authority for the claims being made in the entry.

We now have a very powerful and intuitive “Affirm/Challenge/Comment” function, whereby an expert browsing another expert’s entry can “re-answer” any given question, affirming with another “yes,” for instance, challenging with a “no” or “field doesn’t know,” or simply commenting on it. Comments can also be appended to affirmations and challenges. Others browsing the entry later can see that there have been additions to the question and open a window showing them the entire answer history.

A mockup of what an extensive series of challenges, affirmations, and comments will look like (not real data).

Once we have a large enough contributor base, it will become clear which views are outliers and which represent the mainstream, or on what topics there is a high degree of consensus or almost no consensus. We’ve sometimes joked that the DRH could be used as a dissertation topic finder, allowing students to see which areas of their field really remain unsettled.

And who do you imagine uses the DRH?

Anyone interested in the variety and history of human religious belief and behavior. Experts will use it for a form of lit review before working on a new project or for substantiating generalizations they want to make about the historical record. Scientists will use it to study, for instance, the cultural evolution of religious beliefs and practices. Newly hired assistant professors will use it when they are suddenly being required to teach courses outside their grad training and need to get quickly up to speed on a new area. Teachers will use it in the classroom, and the general public will turn to it as a free, convenient, yet scholarly-rigorous source of knowledge. The religious studies department at the University of Miami is planning to use the DRH in the classroom as a pedagogical tool for training undergraduate students how to make generalizations and what sort of research is necessary to answer “yes” or “no” to a given question. A group of scholars I spoke to about the project at NYU Shanghai, who were dubious about our intended uses for the DRH, think it would be a great way to trace intellectual lineages in religious studies. People will use it for all sorts of things we can’t possibly imagine right now. 

The DRH entries are crowdsourced from experts. Who does the vetting, and given that even some experts hold minority or divergent views, is there a way to indicate which areas have reached scholarly consensus?

Experts are vetted by a member of our editorial team, someone with expertise in their area. Our general rule of thumb is ABD or above in a relevant field at a respectable university or college qualifies one to be an expert. We don’t want editors imposing their intellectual views too strongly on potential contributors. That said, entries are also reviewed by the editors, and in at least one case so far we deemed an entry, as formulated, not up to the DRH standards. (It was subsequently radically reframed, improved, and published.)

The degree of scholarly consensus can be viewed at a glance on any given question regarding any given religious group, or time period and geographic region, however people want to slice up the historical record. The new visualization function will not only display answers to particular questions on a map-time continuum, but also have a visual indication of the degree of agreement on that answer, so you can immediately visually see where scholarly consensus and disagreement exist.

In a field so decidedly in agreement about nuance—and generally, more-exposition-the-better—explain the A/B questioning. In some entries, I was desperate to hear more than “yes”, or even more enticing, “field doesn’t know.”

Experts hate this. Clicking a “Yes” / “No” / “Field Doesn’t Know” box goes against all of our training, which involves, as you say, exploring nuance and complicating categories. At the end of the day, though, we often have an intuition, or even a strongly formed opinion, about how to answer most of the questions in the DRH questionnaire, and the process of going through the questionnaire actually tends to help people clarify their opinions. Comment boxes allow space for the nuances (“I answered yes, but this is a very complicated question, because…”) and fill in the qualitative back story. We also plan down the line to have some automated way to scan comments boxes for certain phrases that would indicate experts didn’t like the question, or how it was formulated, this will help us to identify parts of the questionnaire that need to be improved or simply dropped.

It is important to note that even if you are intellectually opposed to idea of imposing categorical answers on the religious historical record, you can still use the questions as essentially encyclopedia section headings organizing the comments, which will allow you to zoom in on topics of interest to you or to pull all of the comments on a particular topic from the database.

Do you find there’s a close relationship between the technological infrastructure and the theoretical grounding of the DRH?

Yes, the reason we allow experts (at least once all of our poll types are complete) to slice up the historical or anthropological record however they want and to use any unit of analysis—a discernable group, a small band of scholars, users of a temple, inhabitants of a given landscape, author(s) of a text, participants in a given ritual, worshipers of a specific deity—to answer the questionnaire, reflects our view that “religion” is not any one thing, and that behavior and beliefs that we would want to deem “religious” can, and have, taken a variety of social and cultural forms throughout human history.

Who’s funding the project and for how long? Is there a plan for self-sufficiency?

Until recently almost all of the funding has come from the Canadian government, in the form of a Social Sciences and Humanities Research Council (SSHRC) Partnership Grant colleagues and I were awarded in 2012. The grant was to study the “Evolution of Religion and Morality,” and it established CERC. The overall grant is $3 million over six years (ending in Spring 2018), and the DRH has ended up becoming one of its central projects, with a decent chunk of grant funds having gone toward DRH development, postdocs, grad student interns, etc. Last month we also received a small grant from the John Templeton Foundation to fund additional initiatives 2017–8, and we are applying for a variety of large, three-year grants to fund us beyond that. The University of British Columbia (UBC), Stanford’s Center for Advanced Studies in the Behavioral Sciences (CASBS), and the Peter Wall Institute for Advanced Studies at UBC have all also contributed to the project. 

Our goal is to, by 2020, become more or less financially self-sufficient. All of the major desired technical features should be designed, implemented and well tested by then, and we should have a large and self-sustaining user base. The University of British Columbia, my university, has committed to becoming the institutional home of the DRH. They’ve pledged funds for a tech intern, in-kind tech support, and paying for cloud hosting costs—in other words, they’ll guarantee that the DRH doesn’t disappear and continues to function well. Starting soon we’ll be creating a way for users to donate to the DRH, and we’re hoping in this way to build up a bit of an endowment to continue to fund contributor honoraria, new feature development, et cetera. We may also submit further grant applications; there is a lot of money out there for digital humanities, but by 2020 we should be more or less where we need to be, and the DRH will continue to grow and expand without the need for additional outside funding.

The primary advantage of print is its resiliency as a platform. With projects such as the DRH, a huge amount of resources might be dedicated to a technology that could become obsolete in a matter of years. Why take the risk?

[Editor's note: We pulled in the DRH's technical director, Michael Muthukrishna, to respond to this question]

Digital archiving is a more general problem that applies to all of the Internet, and we (especially our technical director, Michael Muthukrishna) have thought a lot about this. In the short term, everything is backed up in multiple places, and the platform itself is "future-proof," at least in the medium-term, in that we're using an entirely open-source software stack. Even if the maintainers of the software go out of business or stop maintaining it, we have the code from the bottom up. As long as the project is around, we'll keep to the principle Muthukrishna set up at the beginning—“latest, but most robust open source technologies”—so we'll keep upgrading just behind bleeding edge, waiting for the dust to settle before adopting any technology that's at the core.

Also, the data is separate from the platform and can be read by other software. It would take a bit of work, but even without the source code to the DRH, you could build a system just to read it because the data is exportable in multiple formats, including CSV and a variety of GIS map files and most recently, PDFs. If we ever felt like the project was about to collapse and we were running out of money to maintain the servers, we would export to PDF, print it out or even send it to a publisher to create the first "Database of Religious History: Print Edition". We will always maintain exportability even if PDF is no longer what we swap files in.
 Finally, keep in mind that books and paper get destroyed in fires. You could argue that modern distribution prevents this but (a) how many copies of a monograph actually exist and (b) we also engage in distribution. So, overall, we think our platform is less risky than almost anything else, including print. 

As graduate students and contingent hires rotate through, who shepherds the growth and direction of the database? That is, how can you guarantee the human capital necessary to make the project successful?

Again, through 2018 we have a dedicated DRH postdoc, a PhD tech intern, and several graduate students working fulltime on the project. I also spend an inordinate amount of time directing it, and Michael Muthukrishna, our technical director and creator of the original database and platform, continues to direct the technical aspects from his job at the London School of Economics. We also have a full-time administrative assistant. Our proposed grants include funding for all of these personnel through 2020, after which we’ll see—we may no longer need so much staff once the DRH takes off. This coming fall we’ll be assembling an advisory board who will help guide the future development of the DRH.

All that said, the quality of the DRH very much depends on the commitment of our volunteer editorial team. For the next couple years we have funding to reward editors with small perks (a book budget, travel support, potentially even occasional teaching buyouts), but ultimately the success of the DRH depends upon the continued willingness of these scholars to support the project and its vision.

What research went into preproduction of the project?

We spent about two years in small meetings and workshops conceptualizing the abstract structure of the database and making sure that it was both powerful and flexible enough to meet our envisioned needs but also be easily adaptable to new uses. In terms of the current Religious Group Questionnaire, we had multiple—like, twenty!—iterations of this over the years of development, including soliciting feedback from religious historians at workshops and at the AAR. 

Who’s going to provide tech support? Take “consumer” complaints? The complexity of these technologies means maintenance costs can skyrocket.

Through the end of 2018 these will be handled by our dedicated tech intern and our technical partner, Pieoneers, who (since 2015) has been working with us to turn the DRH from something cooked up by a grad student and some online freelancers into a really professional, easy-to-use platform. After 2018 UBC will take over tech maintenance costs (with help from a federal Canadian program that encourages industry-university collaboration), or provide in-kind services, and if we land one of our proposed grants we can also tap Pieoneers. 

What are the primary obstacles to success?

The single biggest obstacle right now is awareness of the DRH: scholars of religion don’t know about it, or if they’ve heard about it they don’t know all of the things it’s capable of doing. There is also a real resistance to taking contributions to an online, peer-reviewed site as seriously as, say, a similar entry in a printed handbook or edited volume, although the former will arguably eventually reach more people, be able to attract more peer input, et cetera. It’s frustrating to me that my colleagues will still jump at the chance to contribute to an edited volume that will take eight years to come out, cost $250, and be read by about twelve people, but are leery about investing time into something like a DRH entry. But this is just a matter of changing habits and perceptions, and also making sure that tenure and merit committees come to take digital humanities contributions seriously.

Academics are inherently conservative, which is good in many ways. Since I was a student, though, we’ve gradually become comfortable with using word processors instead of typewriters, e-mail instead of snail-mail, and online journals instead of the print versions. We take notes on our computers on PDFs, and access online, searchable corpora instead of slogging through huge print concordances. Each of these technological innovations involves tradeoffs: snail-mail kept our volume of correspondence to a manageable level and ensured that we only heard from people when they really had something to say. But it’s also hard for me now to imagine working as an academic without e-mail or the Internet, although I spent most of my graduate student life without them.

The DRH represents a very new way of presenting scholarly information about religion, and a novel platform for scholarly interaction, research and pedagogy. Like all new things, it will take time to catch on, but we’re hopeful that it will before too long take its place among the valuable sources that scholars of religion turn to first when they are in need of a reference source, want to perform a literature reviews, or simply want to explore human religious diversity across the globe and through time.