For the last reflection before our final research assignment, my Digital Humanities in REL course was asked to re-evaluate what we’ve learned throughout the semester. We focused especially on what it means to ‘do data’ and what that might result in for scholars and students and research participants and everything and everyone in between. Because the concept was a bit broad, we tried to narrow our focus with a working definition of Digital Humanities and together came up with this definition:
Digital humanities combine technology with theory. Working in digital humanities requires the recognition of human error and contribution to what seems “given” when using technological interfaces present everywhere. We must critically examine the digital world, just as we analyze literature, by leaving room for humanistic contribution and not completely trusting what appears at face value. We complicate the “givens” of computational methods because knowledge production is a political act.Dr. Wieringa’s DH in REL Class, Fall 2020.
While far from perfect, you can get the gist of what we think it means to do data in the digital humanities and in religious studies more specifically. For me and my classmates, it was important to point out that knowledge does not exist on its own, but in a context that is situated and dependent on the knowledge producer.
It might seem a bit backward to start with the conclusion, but I find giving our final working definition before our analysis is useful in this instance. Dr. Wieringa started our last Zoom call by asking us two questions: How do different modes of analysis transform data? And what do those transformations reveal and conceal? These prompts led me to other questions, naturally, but with a little help from our class discussion, these tidbits came together to form a more coherent understanding of digital humanities in religious studies.
First, what is data? Then, can it be transformed?
To answer Dr. Wieringa’s first question (How do different modes of analysis transform data?), let’s briefly go back to the issue of defining data. Being the humanist I am — born and raised in a theory-heavy department — I find myself returning again and again to issues of definition, and data has been no exception. In previous posts for this course and my last Reflection, I spend more time exploring the issue — but here, I think the biggest point is that anything can count as data because anyone can name their data. Of course, some namers can accomplish more than others based on their social ranking and authority, but we are all collecting and organizing data all the time, based on what catches our eye.
All of this is to say that it is important to recognize that the scholar has a key role in choosing what counts as data in the first place. This makes answering the first question a bit easier. If data is in the eye of the beholder, then the methods the beholder uses will influence the way the data is presented. In other words, data can be and is transformed by different modes of analysis.
As mentioned in my last reflection, the issue of counting data in religious studies specifically was discussed in depth by scholars on a NAASR Conference panel in 2017. The essays were compiled into a book by Equinox Publishing and have been extremely useful for my work in this course. The main argument being debated: what counts as data in the study of religion? This question of classification is not new to religious studies. Scholars have debated over methods of studying religion for quite some time and while all disciplines debate over method, these conversations seem especially evident in religious studies. Of course, I am more familiar with these debates, so they stand out in my mind, but I think a key part in this emphasis is the fact that defining ‘religion’ results in more conflicting definitions than synonymous ones.
If we cannot name the thing we wish to study — religion — how do we know how to study it? Where can we find ‘religion’? What is the best way to observe ‘religion’ happening? How do we know if or when ‘religion’ is happening in the first place? Questions like these lead to more complex questions like can naming something as ‘religion’ change a group’s social standing for better or worse? If you look at Reynolds v. United States, Wisconsin V. Yoder, West Virginia State Board of Education v. Barnette, or the more recent Masterpiece Cakeshop v. Colorado Civil Rights Commission, it’s obvious that naming an act as ‘religious’ or not has real-world, tangible implications, making the study of religion more political than many scholars realize or would like to admit.
All of this is to say that there is something at stake when we, as knowledge producers and consumers, decide what our data might be and why it is worth studying in the first place.
What do those transformations reveal and conceal?
These transformations, if done traditionally, conceal the role of the scholar in knowledge production. If done with the role of the scholar in mind, then they reveal the ways in which all knowledge is situated in particular contexts. The latter is most useful and should be the method that scholars in every field should apply. Knowledge has context and those contexts matter.
As mentioned, the naming of our data carries weight and more often than not tells us more about the scholar doing the naming than the name itself. In this way, the transformation of data reveals and conceals the humanity of our study, and this is especially true in the digital world.
Technology has a way of uniting and dividing society. You can access more people and ideas but you can also disagree with those ideas without having to make a case for your argument face-to-face. A screen provides an imaginary wall that allows for battleship-like attacks to be made. In this way, the human is physically removed from the communication. As anyone who has gotten in a tiff in the comments of a Facebook status can tell you, it’s oftentimes easy to forget that technology is just a tool used by humans for humans. As Marissa Parham explains it during an interview with Melissa Dinsman at the Los Angeles Review of Books, “people make technologies, even as technologies also produce (or foreclose) new opportunities for personhood.” Not only do digital methods transform data but the influence of technology affects the individuals interpreting and using digital data. It is sometimes convenient to forget that there are human decisions being made and influenced through digital work.
Beyond the confidence that the anonymity of tech can sometimes provide is the authority of computational methods. In the last few months, I have researched the ways in which the American public has a growing distrust of science. As COVID-19 has ransacked our nation (and the globe) the changing recommendations from previously reliable sources have led to a mistrust of scientific methods in the public eye. I will avoid that rabbit hole for now, but in the way that the public previously trusted science, there is a newfound trust in the digital world. There is an assumption that computational methods lack the emotion of human actors that can cloud objectivity. But this could not be further from the truth and ignoring the role of humans in producing knowledge by creating and editing data will lead to a mistrust of the digital world just as the objectivity in science crumbled.
It is easy to think that technology holds all the answers to our questions when you can shout “hey, Google!” from your living room and access information in a matter of seconds. There is even a new fridge made by Samsung that lets users make recipes based on the soon to expire groceries in the fridge. I think it is safe to say that most Americans think technology makes life easier, but with simplification comes the loss of critical analysis. This is because the methods used to transform data (especially in the digital world) conceal information by presenting one simplified answer to oftentimes broad and problematic questions. In other words, one answer is provided without acknowledging the context that that answer exists in.
So, the transformation of data conceals a great deal but those transformations could also reveal a great deal if only the viewer knows where to look with a critical eye.
What are the advantages and disadvantages of computational analysis for the study of religion?
The disadvantages and advantages of using computational analysis in humanities research and religious studies specifically rely on issues of definition and interpretation. Definitions of religion (and their interpretations), as discussed earlier, can invoke or remove authority from particular social groups depending on its definition. This issue can be amplified with computational analysis as digital work can sometimes conceal the human efforts that are key to knowledge production.
On the other hand, computational analysis in the study of religion presents many advantages and could even combat the issue of definitions. While not quite a traditional religious studies example, this digital project on Darwin’s Origin of Species does the sort of contextualizing work I have argued for thus far. Through a digital medium, the years of edits that went into creating a work that is now universally renowned is visualized. This visualization combats the idea that Darwin stumbled upon the Galapagos islands, had a spark of inspiration, and wrote down an entire theory of evolution in one sitting. While that narrative is neat, it excludes a key process of the scientific method: revisions.
Why do revisions matter? They show the process of knowledge production. The theory of evolution was not sitting on an island waiting for Darwin to discover it. It was a series of human decision-making, contextualizing, reflection, and editing that led to the final, compact conclusion. In the scientific method, these revisions are especially important because they help the next scientist decide what to try and what to avoid. Much like coders could save time by sharing mistakes in their code with one another on a forum. The bottom line is that knowledge production in any form (whether it be scholars of religion or Dr. Anthony Fauci, or the coder of a dating app) should be acknowledged as a process and as such, the entire process should be shared so the interpreter can get a step closer to understanding the bigger, contextualized picture.
As I reach the conclusion of this Reflection, I realize that I have hinted at many complex ideas without much expansion on what those claims imply. For this I ask for your patience, as I hope to build on these ideas in the final paper for our course. A blog post does not leave much room for the heavy analysis that usually accompanies this sort of brainstorming, but I hope it has given you some insight to what I have learned this semester in DH in REL and where I see these new thoughts and questions leading me.