We are currently in the midst of a kairotic moment in the history of writing studies, when we have both the living memory of the field’s development and the technological memory to gather and query large amounts of information. Twenty years ago, we could not easily chart the complex web of interpersonal and interscholastic connections that emerged since the founding of our field’s major institutions, and as time moves on we risk losing those rich histories. Given recent advances in digital tools and methods for data analysis, we ask, what can a data-driven approach to academic genealogy contribute to understandings of the complex influences and growth of our field?
In order to address this question, we created the Writing Studies Tree (WST), a crowdsourced, online, open-access, interactive database of individual scholars, educational institutions, and the disciplinary movements that connect them.1One of our first important decisions was what to call the site. In choosing “writing studies” as opposed to “rhetoric” or “composition” or one of the combinations of punctuation (debated, e.g., in Enculturation 5:1 and 5:2), we are following the lead of Susan Miller (2002), who argued that the term “writing studies” “assumes interest in all available writers and all available writings” (p. 45) and thus “can create the greatest success and shared integrity for our field [. . .] offering a way to describe the cultural work undertaken in any act of writing” (p. 41). Combining a fixed data structure with open editing privileges, the WST aims to document and democratize our knowledge of the relationships among scholars and institutions. It does so by rendering such relationships as data2Christine L. Borgman (2015) argued that such a rendering-into-data is in itself a scholarly act of transformation (pp. xvii). that can then be visualized and interacted with across several scales: from one person, to one institution, to the full network. At the time of this writing, the database contains roughly 1,500 people, 450 institutions, and 3,900 relationships, all navigable by browsing, searching, or filtering a graph of the full network.
When we say that the WST allows for "a data-driven approach to academic genealogy," what we are asking is this: if ideas are propagated through relationships as well as through publications, how many confluences of writing studies theory, teaching philosophy, and other disciplinary trends can we uncover if we trace who studied where, and with whom, or who taught where, and with whom?
Genealogy in Academia
The Writing Studies Tree is a collaborative academic genealogy, but this project has, itself, a lineage. Genealogies of mentor/student relationships have a long history in academia, stretching at least as far back as the links tying Socrates to Plato and through Aristotle to Alexander the Great. The Mathematics Genealogy Project (MGP), one of the oldest online databases of such relationships, has data as far back as the 14th century and as recent as this year’s graduate students.
Genealogies are appealing for a number of reasons. Apart from the emotional satisfaction of being able to trace a direct line from some influential figure to one’s own present studies, the uncovering of such interpersonal links constitutes an inquiry into the forces, choices, and chances that shape the questions and methods of a field over time. As Gesa Kirsch and Liz Rohan (2008) have noted, “how a researcher chooses a subject is a subject unto itself” (p. 1), and academic genealogy is one means of studying that meta-subject. It can also serve to generate new lines of research by connecting members of “families” who may not have previously realized they had mentors or collaborators in common. (For a fuller discussion of the family tree structure—including its drawback and limitations—see “Design Principles.”) In Kirsch and Rohan’s words, “the connections woven through so many of our scholarly and personal lives that remain largely unarticulated” can nonetheless play a major role in the processes of “finding and sustaining a research project” (ibid.). Academic genealogies offer a methodology for tracing these interweavings and making them visible for and usable to future generations of scholars.
Most existing online academic genealogical databases, such as the MGP and NeuroTree, have been developed predominantly for STEM fields (science, technology, engineering, and mathematics), in which graduate students and post-doctoral fellows tend to work within the lab of one principal investigator at a time: relationships are typically represented as hierarchical and unambiguous, allowing lines of descent to be clearly delineated. Mentoring relationships in the humanities are often more nebulous, with ideas from coursework intertwining with suggestions from multiple dissertation committee members to yield research projects separate from, though shaped by, the agendas of these various advisors.3It seems, however, that the influence of the Digital Humanities (and emerging humanities labs) may alter the nature of academic genealogies within the humanities.
The difficulty of discovering and representing the complex nature of relationships in our field has contributed to a major distinguishing design feature of the WST, namely, its crowdsourced approach to data gathering and curation. The MGP allows site visitors to suggest changes, but requires that all information be vetted by an internal administrative staff before becoming part of the official record, causing significant delays between participation and impact. Much like the MGP, the MPACT Project (see below) depends on a centralized team of researchers involved in laborious data entry and verification. NeuroTree allows any registered user to post information, but only that user and administrators can then update that information. The WST, by contrast, directly involves its audience in the process of gathering and curating information, and thus interpellates participants as active shapers of the WST’s representation of the discipline’s history. Although this policy opens the site to potential inaccuracies and incomplete data, it also allows for users to directly address these issues. (For a more complete explanation of our decision to use crowdsourcing, see both the “Rapid Input” and “Participatory Design: Trust in Users” sections under “Design Principles.”)
The WST’s closest predecessor within the humanities is the Phylosophy Project (Phylo), another academic genealogy founded at The Graduate Center, CUNY.4We worked closely with one of Phylo’s founders, Chris Alen Sula, throughout the early stages of the WST’s development process. Sula and the present authors have all completed the Graduate Center’s Interactive Technology and Pedagogy Doctoral Certificate Program, which has proved a fertile ground for thinking through work on digital academic genealogies. However, Phylo is not crowdsourced, and has shifted to focus on tracking job market activity in the field, which are both marked differences from the WST.5For job-market specific research in writing studies, refer to the work of Jim Ridolfo, Claire Lauer, and Nathan Johnson; we hope to work with these scholars to find ways to effectively join our work for greater accessibility and usability in the future. Another influence on our methods has been the cross-disciplinary MPACT Project (n.d.), which is devoted to “defining and assessing mentoring as a scholarly activity, examining the emergence and interaction of disciplines, and identifying patterns of knowledge diffusion.” This project, based in Library and Information Sciences but encompassing a broad array of fields (though not writing studies), has recently extended its data-gathering beyond PhD dissertation titles, author names, and names of dissertation advisors (chairs). The formulas developed by this project to measure impact have served as a springboard for our ongoing analysis of the WST network.
Genealogy in Writing Studies
We believe that writing studies is a particularly interesting test case for studying academic genealogies because of the myriad contexts in which writing is taught: in addition to the usual professorial influences, scholars in our field receive training from (and as) administrators of writing programs and writing centers, as well as editors of scholarly journals. Many of these mentoring roles are not documented in the most common modes of scholarly self-presentation, such as the curriculum vitae, professional profile, educational transcript, or citation trail. This makes writing studies both a challenging and rewarding network to build and visualize.
Joseph Harris’s (1997) A Teaching Subject: Composition Since 1966 offers a glimpse into both the benefits and challenges of exploring these connections. Midway through the text, Harris traced the threads of a particular approach to “voice” in writing back to a tangled weave of influences among scholars and teachers at the University of Pittsburgh, Rutgers, NYU, Amherst, and Harvard (p. 34-43). “A distinctive feature of what I have been calling here the Amherst School,” Harris noted, “is the way its ideas and concerns have been formed and passed along as much through classroom teaching as scholarly books or articles” (p. 127)—and not only direct teaching, professor to student, but also professors and graduate students teaching alongside one another.
For example, Harris pointed out the influence of Theodore Baird on “William Coles, Walker Gibson, and Roger Sale—each of whom taught near the start of his career at Amherst” (p. 34) and Coles’ influence on David Bartholomae, who “taught with Coles at the University of Pittsburgh” and who “freely and often admits the influence Coles has had on his thinking” (p. 38). Then again, perhaps the affinity would not have been so pronounced had Bartholomae not previously “studied with Richard Poirier at Rutgers, who had himself done his undergraduate work at Amherst and had then taught for a time at Harvard with Reuben Brower, a former colleague of Baird’s” (p. 127, note 7). In the Amherst School, at least, multiple lines of influence converge in a dizzying array of who-knows-whom that can be difficult to parse out, much less systematically map; see Figure 2 below for one hand-drawn attempt. Once mapped, though, new insights are possible, such as the connection between Bartholomae’s work with basic writers (discussed in The Study of Error, for example) and the close attention to language employed by the literary critics just mentioned.6Note that this argues in favor of including Poirier in the WST (as he currently is): even if Poirier wouldn’t have seen himself as working within writing studies, his approach has still shaped the field in important and documentable ways.
In Harris’s case, two personal connections to the Amherst School made his discovery more possible: at the time he was researching A Teaching Subject, he was teaching alongside Bartholomae at Pittsburgh (Harris, p. xii); before that, he had studied with Gordon Pradl, a student of Coles (p. 127). It seems significant to us that Harris listed these connections in his acknowledgments and in an endnote, respectively. In fact, this kind of information about personal connections to people and places is often relegated to footnotes, acknowledgements, or other marginal spaces. We believe, however, that these footnotes of history are crucial to our understanding of our discipline, and that the WST’s foregrounding of influential relationships is an important and necessary intervention. Having an open database can broaden access to otherwise privileged information, while simultaneously increasing the number of connections to be made. The following screencapture demonstrates the expansive potential of putting the "Amherst School" network described above into the WST's database.
Although there have been a handful of prior attempts to gather information on academic genealogies within writing studies, as best we can determine, none of this information has been made available for use, scaling, and aggregation. One notable project is the Technical Communication Genealogy Project (n.d.), which aims to gather data into “a ‘family tree’ of scholars and practitioners and a ‘map’ of research topics and the universities from which the graduate work was completed”; after receiving IRB approval some data was gathered, but—perhaps, in part, owing to the constraints of that approval—much of the information remains confidential and not publicly released.
There have also been less formalized attempts to learn of the mentoring relationships that have had formative influences on teaching and researching within writing studies. In 2005, for instance, Susan McLeod started a thread on WPA-L (the highly active listserv for Writing Program Administrators) with the subject line “re: wpa genealogy,” generating around 65 replies. Citing the example of the MGP, McLeod asked,
Is there a TA supervisor in your past (like Joyce in mine) to whom you could point as a WPA parent? I’d like to construct, just for fun, a genealogy that would show the influence of some of these early exemplars in the ‘pre-professional’ days of writing program administration. […] Unless this is of interest to the non-geezers, you can respond to me off-list.These are important questions and histories. However, framing the matter as “just for fun” and a matter only for “geezers”—not even worthy, necessarily, of on-list attention—downplayed the potential of the dataset it might produce. To our knowledge, the many stories told and names mentioned in the ensuing discussion were left as individual curios, without being aggregated or analyzed in the context of other connections. Only now, a decade later, are we gathering them in the WST for further exploration.
One respondent on WPA-L, Roxane Kirkwood (2005, January 13), mentioned that she had been researching a similar topic for a year at that point with collaborator Morgan Gresham, with plans to present at CCCC (the Conference on College Composition and Communication). Kirkwood and Gresham’s research model again forestalled scalability: “So far,” Kirkwood wrote, “we have done longer interviews with a small group, but have begun gathering stories/memories/moments from a wider selection. If you would like to share your family tree with us, please visit http://www.surveymonkey.com/s.asp?u=29430616397. The survey will close at the end of January.”
While we are encouraged by Kirkwood’s conceptualization of collecting colleague’s “family tree(s),”7For a discussion of the good and the bad in the family tree metaphor, see the “Family Tree View” section under “Design Principles.” here again we see problems for the development of a (re-)usable, rich dataset. First, the closure of the survey cuts off incoming data streams. Second, the request for “stories/memories/moments” makes the process labor intensive: the information takes a lot of time to compose and contribute initially, and a lot of time to find, read, and interpret later, especially as the number of entries increases.
As a corrective, the WST combines structured data with spaces for unstructured text in a way that prior academic genealogies have not, and in so doing makes histories of mentorship and collaboration more visible. In addition to the required elements of our data structure, such as names of people or institutions and labels for each kind of relationship, each entry form includes an open, or free, text field where users can contribute biographical information, references to websites and other sources of further material, and any other data not captured by the WST’s formal data structure. In this way, we hope to be as inclusive as possible of whatever users value and wish to share: if Kirkwood and Gresham or their research informants want to share stories, memories, or moments, the architecture of the WST can accommodate that information. But it does so in a way that is already data-linked.
Before the WST went live in 2011, our team used the list of CCCC Chairs, dating back to 1949, to pre-populate the site; in some cases, these scholars have already passed, and obituaries and memorials honoring their work were the most useful texts available for learning about their academic history and connections to other scholars working in the field. These documents were filled with eloquent, rich histories our team wanted to embed alongside the formal relationships they identify. As with Living Rhetoric and Composition: Stories of the Discipline (1999), in which editors Duane Roen, Stuart Brown, and Theresa Enos gathered the stories of how nineteen major disciplinary figures entered the field, we believe that a central and far-reaching interactive archive such as the WST can “help to situate the scholars, their work, and, importantly, the development of the profession” (Roen et al., 1999, p. xv). Maureen Daly Goggin (2000) similarly argued in Authoring a Discipline—after tracing a similar line of inter-generational continuities among editors of the major journals in writing studies—that “what is significant about these [connections, or ‘begats,’ as she puts it earlier] is that they help us understand the genealogy of the discipline, the strands of social and disciplinary influences that comprise the fabric of rhetoric and composition” (p. 156). The WST offers an interactive, navigable genealogy that helps tie those strands together, and enables researchers to search through the tangled web to extract meaning.
Genealogy as Disciplinary Metadata
The WST broadens conventional means of examining the intellectual and social history of a discipline. Though academic genealogy has traditionally been understood as “the quantitative study of intellectual heritage operationalized through chains of students and their advisors” (Sugimoto, 2014, p. 1), in the WST, we extend the student-advisor relationship to include colleagues with whom one has worked collaboratively, the capacities in which that work has occurred, and the schools and institutions that have fostered these experiences of mutual mentorship. Looking through the lens of academic genealogy privileges relationships amid the variety of outputs and impact factors of our field. This methodology provides an alternative measurement system of impact, an altmetric based on the influence of mentoring relationships.
In privileging this altmetric, we see the WST as participating in a surge of interest in the disciplinary metadata of writing studies. Recent studies have examined changes in the number of cited sources (Mueller, 2012a) and in the frequency with which particular sources are cited (Lucas and Loewe, 2011). Such bibliometric studies present a valuable view of the public conversations and debates that shape the discipline, allowing us to grasp trends that elude individuals’ limited perception. Increasingly, there are also a number of data mining efforts that have sifted the text of job listings (Lauer, 2013) and published speeches (Mueller, 2012b) for the evolving frequency of keywords over time; explored the notion of “invisible colleges” in writing studies through the training of faculty in the field (Johnson, 2015a); and mapped geographic distribution of job listings and doctoral programs (Ridolfo, n.d.), journals (Tirrell, 2012), and digital technology (Tirrell, n.d.) in our field. We agree with Nathan Johnson (2015b) that “instead of interpreting bibliometric models as a measurement of scholarship, they could be interpreted as models of the powerful dynamics influencing academic work” (p. 97). At the same time, such models are necessarily partial, as the authors working with them have been careful to state. The WST contributes another partial view, one that captures the behind-the-scenes relationships that form between colleagues in teaching seminars, in graduate courses, in advising meetings, and through the editorial process.
Because they often exist outside the publication record, such behind-the-scenes relationships could remain local or personal knowledge, disaggregated from other comparable histories and patterns, or they could even be lost entirely. Though successful genealogical research in the field (cf. Harris, 1997; Goggin, 2000) demonstrates that a database is not strictly necessary to address these concerns, digital tools make it possible to more rapidly pool many small knowledge-bases into a collaborative effort of disciplinary self-awareness—one reason, perhaps, for the aforementioned uptick in scholarly metadata analysis. And while we are sensitive to Wendy Chun’s caution that obsessive capture of data would “create, rather than solve, archival nightmares” (Chun, 2008, p. 149) as storage outpaces our ability to search through it, we aim with the WST to mitigate the risks and maximize the possibility of discovery; we provide fixed data structures and encourage that new pages be entered via links to existing pages, promoting a pool of data that is self-contained and interconnected.
[ + ]
|1.||↑||One of our first important decisions was what to call the site. In choosing “writing studies” as opposed to “rhetoric” or “composition” or one of the combinations of punctuation (debated, e.g., in Enculturation 5:1 and 5:2), we are following the lead of Susan Miller (2002), who argued that the term “writing studies” “assumes interest in all available writers and all available writings” (p. 45) and thus “can create the greatest success and shared integrity for our field [. . .] offering a way to describe the cultural work undertaken in any act of writing” (p. 41).|
|2.||↑||Christine L. Borgman (2015) argued that such a rendering-into-data is in itself a scholarly act of transformation (p. xvii).|
|3.||↑||It seems, however, that the influence of the Digital Humanities (and emerging humanities labs) may alter the nature of academic genealogies within the humanities.|
|4.||↑||We worked closely with one of Phylo’s founders, Chris Alen Sula, throughout the early stages of the WST’s development process. Sula and the present authors have all completed the Graduate Center’s Interactive Technology and Pedagogy Doctoral Certificate Program, which has proved a fertile ground for thinking through work on digital academic genealogies.|
|5.||↑||For job-market specific research in writing studies, refer to the work of Jim Ridolfo, Claire Lauer, and Nathan Johnson; we hope to work with these scholars to find ways to effectively join our work for greater accessibility and usability in the future.|
|6.||↑||Note that this argues in favor of including Poirier in the WST (as he currently is): even if Poirier wouldn’t have seen himself as working within writing studies, his approach has still shaped the field in important and documentable ways.|
|7.||↑||For a discussion of the good and the bad in the family tree metaphor, see the “Family Tree View” section under “Design Principles.”|