Digital Humanities and the Graduate Research Methods Class

Increasingly, humanities research methods are digital, and using digital resources is not the opposite of traditional humanities research.1 In 2019 Ted Underwood noted, “It seems increasingly taken for granted that digital media and computational methods can play a role in the humanities” (“Digital Humanities,” 96). He concluded his brief reflection by saying, “Now it is just a matter of doing the work and teaching others how to do it” (97). The reality of our current digital research practices needs to be clearly reflected in how we design graduate research methods classes. I contend that graduate students need to learn to “read” digital projects and search interfaces—ideally, early in their graduate programs. “Reading” interfaces and projects involves learning their scope and boundaries, determining how they mediate the knowledge of the present, and establishing how to utilize them ethically and efficiently. Tanya Clement posits that as digital humanists, we need to be explicit about our research methodologies and why and how they matter. Graduate students must learn to apply their rigorous critical skills and information literacy to digital research methodologies, or they will be ineffective at humanities research (“Where Is Methodology in Digital Humanities?”).

This chapter is not the first to argue that all humanities students need to learn digital research methods.2 Indeed, Christine L. Borgman maintains that “graduate school is rather late” to “learn how to use and how to evaluate digital cultural materials” (para. 66). This chapter offers concrete examples of how graduate humanities students can be encouraged to use and assess digital resources. The examples here are taken from an English-department research methods course structured around small weekly assignments, but many of the principles extend beyond literary studies and could be used in other disciplines.3 In this brief piece, I offer suggested readings and assignments that lead to discussions around three interrelated topics: library catalogs and search engines; databases and bibliographies; and digital humanities projects.

So often when a digital humanities component is included in a course, the question becomes, “But what do you cut to fit in the digital?” In a research methods class, however, the question becomes, “Are we teaching current research practice if we exclude the digital?” The three topics outlined here were incorporated into a literary studies research methods class, but they could apply to other disciplines too. In this class, we covered other nondigital topics such as descriptive and analytic bibliography, textual studies, and manuscript studies; we also covered digital topics such as digital editions, text analysis, and data visualization. The goal of this class was not to have students create digital resources but to learn to use them to undertake different kinds of literary and cultural analyses.

Basic library searching skills are fundamental to humanities research at all levels and need to be taught in undergraduate classes, yet it is important not to assume that graduate students possess these research skills. One way I address the unevenness of students’ familiarity and skill with library research is by assigning chapters from Thomas Mann’s Oxford Guide to Library Research throughout the semester. Texas A&M University, like many institutions, had both an online library catalog and the single search bar library discovery system.4 Candace Benefiel, the English subject librarian, would join us to discuss the difference between the two searches.5 Students then wrote a short response comparing the results of two searches about a lesser-known author in their field.

Once students learn some basic skills for how to navigate the library catalog (or unified search solution, as the case may be)—using, for instance, Boolean operators or faceted searching—we turn to the results that are returned from these searches. Students are often surprised to learn that algorithmic biases filter the results they seek, even in their home institution’s library catalog. Safiya Noble’s Algorithms of Oppression and Matthew Reidsma’s post on bias in library algorithms both offer clearly written examples of how algorithms can offer problematic search results. Reidsma, for instance, shows how ProQuest previously suggested “hearsay in United States law” when a user looked for “rape in the United States” (“Algorithmic Bias in Library Discovery Systems”). Noble’s examples of googling “Black girls” and “Asian girls” continue to offer pornographic and fetishizing results on the first page. To avoid the false “algorithms were problematic but then we fixed them” narrative, I bring in a recent story of algorithmic bias from the news and find a problematic example using Google autocomplete or Google Images to discuss in class.6 In July 2020, for instance, a Google Image search for “professor” yielded results where the first twenty-five images returned—that is, an entire screen of results—were white males (including one white boy playacting professor); the twenty-sixth image was a woman and the thirty-fifth image was a Black man. Because search results change—and can change depending on who searches them, but that is another teachable moment7—it is important to find a current example before each class. Critically assessing search results and their weighting is a fundamental skill for humanities researchers. Having to sift through biased results puts pressure on the researcher; assuming that digital results are impartial leads to biased research.

Many databases and bibliographies give students the impression that they have searched everything, yet determining a database’s limitations is pivotal to making sound claims about existing scholarly discourse and crafting arguments that enter confidently into that discourse. An undergraduate might think that searching JSTOR is an adequate research method, but, as Lisa Gitelman points out, “If students or scholars stop with JSTOR—without seeking additional sources or pondering its contents and their limits . . . it becomes the system governing statements” (78). I offer paired sessions on databases and bibliographies of secondary sources:8 before the first class, students complete a small assignment where they search and compare results from two databases (such as Project MUSE and Literature Criticism Online). Before the second class, students undertake the same search using field-specific bibliographies after reading Graziano’s “Retrieval Performance and Indexing Differences in ABELL and MLAIB” as a model for the importance of comparative queries.

For these sessions on databases and bibliographies of scholarship, I assign Gitelman’s article, “Searching and Thinking about JSTOR,” as well as the entries from James L. Harner’s Literary Research Guide (LRG) on “Digital Archives” (which in the LRG are JSTOR, Project Muse, and IngentaConnect) and “Database Vendors” (including EBSCO, ProQuest, and Gale, among others). Harner drily notes that “none of [these] vendors provides a remotely adequate explanation of the scope or editorial procedures governing the databases they purvey,” which highlights the importance of assessing the scope and content of these sources.9 As Harner explains, “Serial bibliographies, indexes, and abstracts (print and electronic) that are published or updated at regular intervals are important resources . . . since they guide researchers to the most recent scholarship.” Harner criticizes “the unfounded assumption that the presence of such electronic bibliographical behemoths as WorldCat or MLAIB [MLA International Bibliography] and Internet search engines have rendered more specialized bibliographies obsolete.”

The main takeaways from the readings, the comparative searching assignments, and the class discussion are first, the importance of field-specific bibliographies and second, a consideration of how digital bibliographies structure the knowledge they present. Perhaps the most striking part of this class is when we undertake to draw a Venn diagram on the board: what are the overlaps among JSTOR, Project MUSE, ProQuest Dissertations, and the MLA International Bibliography? As we add circles to the Venn diagram, it becomes clear to students that no digital search is all-encompassing. I also ask students to identify the gaps in coverage in the bibliographies: it is often non-English voices from disparate geographical regions that are silenced (Gitelman, 77).

Learning to question what is privileged and what is marginalized in our searches (of libraries, online, and of bibliographies and databases) leads nicely to a critical analysis of digital humanities projects, with a focus on what they cover and how they present information.10 I ask students to do a brief presentation (four minutes plus discussion at the end of class) on a digital resource that is germane to their research. I offer students a list of resources they might consider based on their area of focus but also welcome them to propose others for analysis. The students create presentations and online factsheets to offer their classmates the tl;dr (too long; didn’t read) summary to help make the class aware of research projects within and beyond their areas of specialization. In an ideal world, undergraduates will have already learned to assess digital projects in a basic way, using, for instance, the CRAAP test (Currency, Relevance, Authority, Accuracy, Purpose) (Blakeslee). Graduate students need to go beyond asking whether a source is “scholarly” enough; they need what Tara Brabazon calls “critical literacy,” which is “an intervention” that involves the “interpretation, critique and analysis” of digital projects and search results (Brabazon, 30). Critical literacy when it comes to digital projects (and all media) is important not only for students who will become digital humanists but also for research across disciplines as well as for fostering better critical thinking beyond the academy.

Because I teach in a literature department, I emphasize that assessing digital projects is a form of critical reading. As Janine Solberg asserts, for the study of history, “the digital tools and structures that increasingly support our research efforts have material and epistemological implications for how we discover, access, and make sense of the past” (Solberg, 54). Indeed, as I have claimed with Andie Silva, “Digital projects need to be considered as arguments: they argue for the importance of the material they present, and they shape the ways users conceptualize and research texts and archives” (Estill and Silva, 131). Digital projects do not just present data; they offer ways of knowing and understanding the data they present.

Thinking critically about search results, scholarly databases, and digital projects might sound like a simple proposition, yet actually grappling with these ideas is in no way easy. To identify omissions in a project’s scope, for instance, a graduate student would need to be already adept in a given field of study. Thinking about how a digital project presents materials can mean looking at a site’s metadata systems and considering issues of taxonomy and categorization. All too often, digital humanities projects reify problematic notions of race, gender, and canonicity.11 Neither classification systems, nor metadata, nor quantitative data are neutral.12 Miriam Posner suggests that to truly “interrogate race, gender, and other structures of power . . . would be much more difficult and fascinating than anything we have previously imagined for the future of DH; in fact, it would require dismantling and rebuilding much of the organizing logic that underlies our work” (“What’s Next,” 32). Although dismantling and rebuilding digital humanities is beyond the scope of any one course, starting to grasp and then question the “organizing logic” of digital projects is possible but challenging.

Incorporating digital humanities in the graduate research methods class is not about staying abreast of developments in a given subfield, though often new research is shared digitally, and digital projects enable new research and even, as Susan Brown puts it, “new kinds of knowledge” (Brown, para. 14). Rather, bringing digital humanities to bear in the graduate research methods class is about fostering a mode of critical inquiry. A strong humanities research methods class will offer students an understanding of the field and the skills to undertake research using a variety of available techniques. Ultimately, the best graduate research methods classes equip emergent researchers to navigate future digital projects that have not yet been developed.