Seeing the Forest from the Trees: A Reflection on the Challenges of Digital Analysis in Buddhist Studies

The first time I attempted to integrate digital tools into my research, I had a common experience: I had my dataset, entered the data into spreadsheets and uploaded to the platform (at that time, Palladio), and produced a crude network analysis of key terms shared across a 100-chapter Japanese Buddhist text. Unfortunately, I had produced a visualized version of something I had already demonstrated elsewhere through analog research methods. My first experience with research of this type revealed to me a challenge inherent to research using digital tools in Buddhist Studies, namely knowing where and when to deploy them and, perhaps more importantly, how to read data in ways that are beneficial to, rather than duplicative of, interdisciplinary research.

The field of Buddhist Studies has become increasingly interdisciplinary insofar as scholars must engage—often simultaneously—historical, sociological, linguistic, material, and philosophical methodologies within and across Buddhist cultures. With such a range of possible angles on a given research project, the problem of how to apply digital tools and interpret the data through several methodological registers becomes even more daunting. To best leverage digital tools in this context, how should Buddhist Studies scholars think about platforms like GIS, Palladio, Cytoscape, and Gephi in relation to their research projects? Which, if any, is the right tool for the job? How should scholars interpret digital data differently, if at all, in reaching their conclusions?

To answer these and other questions, I reached out to Chinese Buddhist Studies scholar Jason Protass, William A. Dyer, Jr. Assistant Professor of the Humanities and Assistant Professor of Religious Studies at Brown University. Dr. Protass has published research on the history of Chan Buddhism in the Northern Song era, with an emphases on poetry, textual cultures, lineages, and geospatial relationships. In his most recent work, he has leveraged GIS to help visualize patterns across hagiographic anthologies to indicate the spatial location of Chan masters and their networks. This has allowed his source material to transform from texts into maps, which involves very different interpretive strategies.

In his view, linguistic and archival practices that have emerged in East Asian Buddhist textual cultures, especially, have supported the integration of digital research tools in Buddhist Studies:

Buddhist Studies is a large and heterogenous field. I can speak to some natural strengths in the study of texts related to East Asian Buddhisms. First, the Sinograph, or Chinese character, has been a relatively stable writing system for millennia, facilitating the digitization of premodern texts. Second, the use of particles, rather than declension or conjugation, facilitates analysis of digital text. Third, stakeholder communities, especially in Taiwan and Japan, invested in creating reliable digital editions of Buddhist texts and freely circulated them. These factors will allow some scholars to jump into digital analysis with relatively fewer barriers.

Indeed, efforts to build textual databases such as CBETA, SAT, and, more recently, BuddhaNexus have widened the scope of accessibility for those interested in conducting digital analysis. These and other databases offer organized, accurate, and flexible bodies of data that can be wielded in several digital contexts. And yet, despite these forms of support, scholars still encounter several challenges when approaching textual corpora using digital tools because, ultimately, a dataset that is simply made easier to wield does not necessarily allow scholars a clearer view of conclusions within it. According to Dr. Protass, two general pitfalls await scholars:

First, most Buddhist Studies research questions will not be enhanced by digital humanities methods, and the questions that are best answered with digital methods may not be the most important questions to ask. That leads to the next point. Second, a visualization only makes visible patterns already in a text—the tools do not explain why those patterns are there, how they are significant (for whom), nor indicate what they exclude. A DH visualization is not valuable without interpretation and critique.

This is a critical point that I think is not stated often enough, especially for those who have no prior experience with digital analysis: visualizations offer neither answers nor explanations of the data. Just as I had done when I produced my first network analysis using Palladio and found that the visualization simply duplicated my prior findings, many scholars may find that digital analysis does not offer the monumental discoveries they had originally expected. While encountering dead ends is part and parcel of both digital and analog research, there is often an expectation that the processing power of digital research tools will inevitably clear a path to previously unseen conclusions. As Dr. Protass describes, however, we ought to apply the same critical rigor in our use of digital tools that we do in analog research:

Scholars of Buddhist Studies habitually reflect on disparities between idealized textual representations and the history of lived religion. We understand we work with truth claims embedded in religious narratives. Buddhist Studies has a developed repertoire of critical thought that we can bring to digital humanities, and I think Buddhist Studies scholars will continue to be innovative in our uses of digital humanities. For example, I will be interested to see how others create and analyze maps of imagined geographies. Buddhist Studies scholars new to this work will benefit from engaging recent scholarship in the various subfields of geography. And, when beginning to think about interpretation and critique of DH visualizations, I would look for analogies in traditional source critique: Who created this text and when? What are the biases, exclusions, or perspectives assumed by the author(s)?

Inherent to the successful integration of digital tools in research, then, is the recognition of its limitations. Just as one would question the perspectives through which geographical maps are constructed, one should make similar considerations in assessing visualizations produced through digital analysis. The interdisciplinary nature of Buddhist Studies—which often demands an engagement with disciplines for which scholars lack formal training—provides opportunities to develop unique analytical approaches particular to those fields outside of Buddhist Studies. In this way, interactions with digitally rendered forms of data not only persist long after the moment of production, but demand dynamic and responsible methods of interpretation to best leverage them in research. Dr. Protass’ own research has demanded such methods:

As for this project’s process, I began with a question that arose during my own traditional reading of primary sources: did geography matter in the decline of the Yunmen lineage c. 1130? Although I realized that Chan masters taught in monasteries across a wide landscape, I could not keep track of where thousands of teachers were located just in my head. When I made a list on paper the spatial relationships between all locations were not evident. I created a GIS dataset so I could visualize the spatial patterns within the hagiographic anthologies known as Lamp Records. The resulting maps showed me the locations of hundreds or thousands of people at a single glance. And yet, the maps did not answer my question of how geography mattered in the decline of the Yunmen lineage; the maps are not explanations in and of themselves, even though they were critical for my process and for the published argument. After I recognized some key geographic patterns visible in the maps, the final step for my research was grounded in the more traditional reading of sources (especially literati miscellanies and local gazetteers that recorded the destruction of monasteries). The entire process involved moving between perspectives, which we might describe as ‘close’ and ‘distant’ forms of reading. […] The metaphor of ‘distant reading’ emphasizes interpretation in the use of digital tools to analyze texts—it implies that creating and viewing a map can be an act of reading. To change the metaphor, we might imagine viewing a pointillist painting. Only when we step back, the patterns of the image come into view. At the same time, one sees something significant only when viewing the same image closely: the pressure of brushstrokes, the depth of the pigment, gradations of color. Neither a distant nor close viewing can wholly account for the pointillist image. One sees different aspects depending on the perspective. So too with hagiographic and other texts. Moving between perspectives allowed me to create a finely textured geographic history of the decline of the Yunmen Chan lineage.

Map of Chan Lineages (900–960), from Jason Protass, “A Geographic History of Song Dynasty Chan Buddhism,” p. 157.

Whether in a digital or analog context, the challenge of interpretation confronts the researcher with many possible avenues. Like in analog research, the best approach in digital research tends to be many and varied; diverse and sometimes contradictory views of the material at hand can yield exciting and unexpected conclusions. In some cases, like in the work of Dr. Protass, what is necessary is the privileging of analytical methods that critically engage issues like social or institutional imperatives, and unseen linkages, exclusions, or exceptions at play underneath the data being visualized. Just as datasets are prone to the misapplication of digital tools, visualizations are prone to underinterpretation or misinterpretation. The interdisciplinary demands within the field of Buddhist Studies means that looking closely, from afar, and from a variety of angles is even more important.

While I can recall my discouragement during my initial attempts at network analysis many years ago, it has been tempered over time as I have learned more about how scholars in Buddhist Studies have integrated digital tools within their own research. Dr. Protass speaks to a very common experience within the field, especially for those just beginning to explore the potential of DH within Buddhist Studies. While that potential is alluring, it can be blinding, too. Understanding the limitations, suitability, and degrees of applicability of digital tools is the first step in avoiding the pitfalls described above. And like analog research, interpretative strategies in digital research, especially in Buddhist Studies, are not “one-size-fits-all.” The methodological demands of the field create new demands on these strategies, and as the field continues to evolve, scholars should meet those demands by establishing critical and investigatory views of their own visualizations.

Banner Photo: Song-era stele, taken by Jason Protass (Zhenjiang, 2019)

One thought on “Seeing the Forest from the Trees: A Reflection on the Challenges of Digital Analysis in Buddhist Studies

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s