PART VI ][ Chapter 25

The Resistance to Digital Humanities

DAVID GREETHAM

This essay is a perhaps foreseeable follow-up to an earlier piece on “The Resistance to Philology” (Greetham),¹ published in the collection The Margins of the Text. That volume dealt not just with those parts of a text that typically were relegated to the bibliographical margins (titles, annotations, marginalia, etc.) but also with those features of textual discourse (race, gender, sexual orientation, class, among others) that had been marginalized in discussions of textual scholarship. The collection had been prompted by the discovery that in some otherwise highly regarded academic institutions, a scholarly edition, bibliography, or textual study counted as only one half of a “real” book in promotion and tenure decisions. The critical hardback monograph was the gold standard by which scholarly and intellectual achievement was to be measured.

Now, a decade and more later, it is unclear whether that institutional prejudice against bibliographical and editorial work has been overcome or whether it has been compounded by a newer dismissive attitude, this time toward digital and electronic “publications.” Since the great majority of new scholarly editions established in the last twenty years and more have some prominent digital component (electronic text, hyperlinks and hypermedia, and so on), if the institutional marginalization of text has been joined by a similar prejudice (or, at best, an equivocal attitude) toward medium, then those of us working with electronic text are confronted with a double whammy in an increasingly competitive academic atmosphere. Put in a related form: is the very concept of digital humanities (DH) seen in some quarters as an oxymoron, the passing off of technē as if it were critique; and, a fortiori, if this oxymoronic DH is concerned with the production of textual or bibliographical resources, are those scholars engaging in such pursuits under a two-fold suspicion by the general community of “humanists”?

In a short essay describing the potential problem, it is unlikely that we will find a smoking gun or that the usual cloak of confidentiality regarding tenure and promotion will be sufficiently lifted to provide a clear view of academic predispositions. Nonetheless, there is little sign of the old prejudices against textual study having been lifted. In part, this continued disdain may be related to the positivist and antihermeneutical postures of the more technical (and less critical) claims of some textuists (e.g., that textual scholarship is a “science,” with demonstrable proofs), a self-characterization that only feeds the suspicions of some humanities scholars that bibliographical and textual research belongs in current humanities departments only as a “service” activity, not fully integrated in or related to the loftier philosophical aspirations of postformalist humanities. In the original “Resistance” essay, I argued that textuists should embrace hermeneutics rather than science to become “dangerous” again; and Hans Walter Gabler (2010) has recently promoted a recognition of textual criticism as the sine qua non of editorial and bibliographical activity.

It is probably true that, because digital work has at least acquired a veneer of the “sexy” and the “new,” while there may be some chartable unease about the quantification and the technical coding (SGML, HTML, XML) aspects of electronic work, the proliferation of recent print publications (and even movies) on the Internet, information (see Gleick), online social networks, and so on has given a public prominence to digital humanities, whatever that is. The ongoing series of New York Times articles on “Humanities 2.0”; the citing of blogs as evidence in the popular press; the use of Twitter and digital phones in recent political movements (Iran, Egypt, Tunisia, Yemen); and the continued, perhaps exacerbated, concerns of government with the control of copyright in digital environments—all of these features of what Bourdieu calls “fields” of “cultural production”—show that the electronic environment is a persistent and well-traversed area of our common discourse. Whether that discourse will admit digital bibliographical scholarship, digital editing, and digital textuality as academic credentials is another matter.

As in the case of the earlier “Resistance” essay, this problem can be seen as one of rhetoric (though recognizing that this usage is not meant to undervalue the very real and practical concerns of scholars, particularly younger scholars, facing the career-determining decisions of review committees). Thus, in the first issue of Digital Humanities Quarterly (DHQ, whose very title shows an indebtedness to, or desire to connect with, an established “print” mode of production), Joseph Raben’s apologia (“Tenure, Promotion and Digital Publication”) makes much of the fact that DHQ is a “a totally online scholarly publication,” while at the same time noting that “the absence of parallel publications in other sectors of humanities research is a measure of the distance still to be traveled before computer publishing is regarded as fully equal to the book and the print journal.” Raben puts the issue very starkly when he recognizes that some potential contributors to DHQ may be all too aware that “appearance in electronic media is not as highly regarded by the gatekeepers of tenure and promotion as the traditional hard-bound book and the article offprint, at least in the humanities.” The enthusiastic embrace of electronic work by such senior scholars as Jerome J. McGann (whose hypermedia archive of Dante Gabriel Rossetti and such foundational critical writings as the “Rationale of HyperText” have provided a paradigm for DH) has to be measured against the reluctance of other textual critics to grant DH, and specifically electronic editing, the status of supersessionist medium. So G. Thomas Tanselle, who is generally regarded as the exemplary contemporary figure in the continuity of the print-based Greg-Bowers school of “copytext” editing, was (unexpectedly) brought in (co-opted) by the Modern Language Association (MLA) to write a preface to the volume on Electronic Textual Editing, in which he discounted the “hyperbolic writing and speaking about the computer age, as if the computer age were basically discontinuous with what went before” and warned that “when the excitement leads to the idea that the computer alters the ontology of texts and makes possible new kinds of reading and analysis, it has gone too far” (Tanselle). I share Tanselle’s view that much of what Paul Duguid has pointedly referred to as early “liberation technology” by computer proselytes placed the rhetoric of DH in a dangerously vertiginous position. And I also agree that, if DH is to be accorded a recognizable and stable position in academic writing, it must not cut itself loose from the textual and critical history that precedes it. As I have remarked elsewhere, I would maintain that the rhetorical stance of Matthew G. Kirschenbaum’s Mechanisms: New Media and the Forensic Imagination (in which digital forensics is linked to the bibliographical principles of W. W. Greg, Fredson Bowers, and G. Thomas Tanselle) is not just an appropriate ideological tactic but may also serve to head off the potential ghettoization of DH and electronic editing with it.

But how far has the ghettoization already gone; and, in the absence of the smoking gun, how do we measure the “resistance” of my title? In his brief introduction to the scholarly mandate of DHQ, Raben cites some disturbing statistics from a 2006–7 MLA Report of the MLA Task Force on Evaluating Scholarship for Tenure and Promotion. Raben concentrates on the findings that “40.8% of departments in doctorate-granting institutions report no experience in evaluating refereed articles in electronic format, and 65.7% report no experience in evaluating monographs in electronic format,” although 88.9 percent of doctorate-granting institutions rate the publication of a monograph as “very important” or “important” for tenure. These statistics are telling enough, but the report itself contains much more that should be of continuing concern to DH practitioners. It is hardly promising that across the board, only 28.1 percent of departments considered monographs in electronic format “important” to the evaluation of scholarship, and overall “[r]efereed articles in digital media count for tenure and promotion in less than half as many departments as refereed articles in print,” and “[m]onographs in electronic formats have a place in the evaluation of scholarship for tenure and promotion in only about one-third as many departments as print monographs” (MLA Report of the Task Force, 43–44). There is clearly a suspicion that somehow the “open access” principles of a good deal of electronic scholarship produce a less than rigorous evaluation procedure than in traditional print contexts. This suspicion was one of the reasons that Hoyt Duggan founded SEENET (Society for Early English and Norse Texts)—to create a forum for peer review that would match that usually associated with print publication, which was also a founding principle of NINES (Networked Infrastructure for Nineteenth-Century Electronic Scholarship). The MLA report finds that such rigorous scholarly procedures are now evident in electronic journals, which are “increasingly run by editorial boards committed to peer review” (44). The institutional paradox is that while “digital forms of scholarship increasingly pervade academic life, work in this area has not yet received proper recognition when candidates are evaluated for promotion and tenure” (44). It is impossible to estimate how many younger scholars may hesitate to begin work on digital projects, knowing (or sensing) that their labors will count for less during the important career-making moments, a challenge that I think may be compounded when the electronic scholarship is editorial or bibliographical. So although I can testify from thirty years involvement with the interdisciplinary Society for Textual Scholarship that the proportion of younger (and, I believe, of female) participants in recent conferences has greatly increased in the last decade or so, we must wonder how many of these younger people will make it into even the midranks of the profession.

What is this scholarly environment as it promotes (or fails to promote) digital and digital textual work? The MLA report was published only five years ago; and while at several points its authors look toward a more collegial institutional atmosphere in the near future, we should not expect that moment to have yet arrived. Nonetheless, in part as a response to the grim findings of the report, the MLA has been in the forefront of an attempt to make literature and language scholars and administrators more familiar with digital work and its demands.

In a very helpful e-mail correspondence, Stephen Olsen (associate director of research and manager of digital services at the MLA) points to a number of such “outreach” programs by and within the MLA, including a series of electronic roundtables at the annual conventions, together with presentations of digital projects at individual computer stations. Workshops on evaluating digital scholarship are also organized at the MLA, as they are at Association of Departments of English (ADE) and Association of Departments of Foreign Languages (ADFL; especially notable being the summer seminars for administrators). The MLA Guidelines for Evaluating Work with Digital Media in the Modern Languages together with several important publications by the ADE (see Fitzpatrick, “Planned Obsolescence”; Hayles, “How We Read”; and Kirschenbaum, “What Is Digital Humanities?”) have emerged from the summer seminars and are now available on the ADE website. Such forceful programs from the leading U.S. professional organizations may indeed lead to an administrative adjustment of perspective, but reports from other quarters suggest that there is a good deal of proselytizing still to be done. The real test will be to see whether there is substantial change in the statistics when the MLA next conducts a follow-up report to the earlier one.

In the series of e-mail, personal, and phone interviews (which must clearly remain impressionistic rather than statistical) I conducted for this article, those scholars who primarily occupied a position in a research or archival wing of an institution tended to be more hopeful about the recognition of electronic work. Thus Bethany Nowviskie, who serves as director of digital research and scholarship at the University of Virginia Library and sits on the MLA Committee on Information Technology, has drafted an article on the “preconditions” to the evaluation of collaborative digital scholarship, to be published in a cluster in Profession, edited by Laura Mandell and Susan Schreibman. This will build upon the 2006–7 report and provide what may become a template for future evaluations. Similarly, Michael Bérubé, director of the Institute for Arts and Humanities and Paterno family professor in literature at Pennsylvania State University and coauthor of the MLA report, finds that “forms of scholarship other than the monograph are absurdly devalued in some quarters” but that the work of Kirschenbaum and others is “beginning to change things …however gradually.” A less hopeful diagnosis (and prognosis) is offered by Stephen Ramsay, associate professor in English at the University of Nebraska–Lincoln and a fellow in the Center for Digital Research in the Humanities, in an e-mail and phone conversation about recent case histories. Ramsay noted that such cases “provide a sad commentary on the state of DH acceptance. Just when I think we’ve arrived, [these cases] remind me that in some quarters things are precisely as they were when I started (fifteen years or so ago).” Ramsay also feels that there may be an institutional “fear” of some of the more technical (and less “humanistic”) aspects of DH, especially quantification and data mining, but that, for institutions like his own University of Nebraska, DH offers an opportunity to create an academic and intellectual “presence” that would not have been available in predigital days.

The problem may be, as both Ramsay and Bérubé suggest, that during a period of transition it may be particularly difficult to determine where we are in the unfolding of that transition (and even more difficult to predict when and if DH will have been fully accepted as a “respectable” element in humanities research). As Morris Eaves (also a coauthor of the MLA report) observes, in characteristically pointed language,

Eaves’s recognition that a resistance to DH may be another form of the resistance I charted earlier to “philology” is particularly telling, for it means that textual scholars working in electronic environments (i.e., pretty much everyone) have to parry a two-pronged hostility: that textual study is not critical; and that, if presented digitally, is a mere technē. So the “hopeful signs” that Eaves looks for may depend on what “flavor” of DH is being offered.

There can be no doubt that the sheer range of DH projects, institutes, and scholarly investment has increased enormously in the last decade or so. The survey of DH activities in Matthew G. Kirschenbaum’s “What Is Digital Humanities and What Is It Doing in English Departments?” (2010) sees a “robust professional apparatus that is probably more rooted in English than any other departmental home” (1) but then cites examples that are challenging if not perplexing or professionally disturbing (the book series Topics in the Digital Humanities from the University of Illinois Press, like this current volume, in codex format, and the well-circulated story of Brain Croxall, a recent Emory PhD whose 2009 MLA paper “The Absent Present: Today’s Faculty” was read in absentia because he could not get travel funds to attend the convention in Philadelphia). That Kirschenbaum chooses a particularly dystopian citation from Cynthia Selfe as his epigraph (“People who say that the last battles of the computer revolution in English departments have been fought and won don’t know what they’re talking about”) enfolds the account of apparent progress within a rhetoric of immobility, even regression.

And in conversation at the Society for Textual Scholarship conference at Penn State in March 2011, Kirschenbaum (who is both author of the highly regarded Mechanisms and the forthcoming Track Changes and associate professor in the Department of English at the University of Maryland as well as associate director of the Maryland Institute for Technology in the Humanities and thus of the “double appointment” model of a number of successful DHers) agreed that the print monograph still overwhelmingly influences tenure and promotion decisions and this at a time when print publication is becoming more difficult to sustain economically. It may be that promotion and tenure committees will be reasonably tolerant of a candidate’s involvement with DH but that this tolerance will not necessarily extend to a faculty member whose primary work has been digital and who has yet to produce the traditional print monograph. This problem is in part addressed by the MLA wiki “Short Guide to Evaluation of Digital Work,” which poses a series of questions that institutions should consider in trying to evaluate nonprint work. These questions include the following: Is it accessible to the community of study? Have there been expert consultations? Has this been shown to others for expert opinion? Has the work been reviewed? Can it be submitted for peer review? Has the work been presented at conferences? (a question that in a qualification, “It should, however, be recognized that many candidates don’t have the funding to travel to international conferences,” recognizes the practical issues raised by the Croxall case). The wiki takes the funding issue as indicative of the phenomenological divide between print and digital production: “Digital work is hard to review once it is done and published online as our peer review mechanisms are typically connected to publication decisions. For this reason competitive funding decisions like the allocation of a grant should be considered as an alternative form of review. While what is reviewed is not the finished work so much as the project and track record of the principal investigators, a history of getting grants is a good indication that the candidate is submitting her research potential for review where there is real competition.”

A broader view of the status of DH was undertaken during the workshop session of Project Bamboo, held in Princeton in 2008. Among the concerns raised were those similar to the MLA report, for example, that while “new faculty are brave and will try new stuff,” in terms of “career incentives,” they are “worried about what they’re judged on” because “[h]umanities are rewarded by publishing a book” (Project Bamboo, “Exercise 6A Scribe Notes”). The same (unidentified) speaker went on to describe such a bias as confirming current emphasis on “the glory of the book” although conceding that “[w]e are at the cusp of a reward system changing for humanities,” which might reflect a shift whereby the “distinction between technologists and people who use technology [is] disappearing …and that “people think through technology” (Project Bamboo, “Exercise 6A Scribe Notes”).

The anxieties surrounding DH that emerged at the conference were wide ranging and included a greater concern about getting published early in graduate school than in previous generations; observations that digital work had actually failed to produce academic advancement; a worry that campuses don’t know how to credit online work; a nostalgia—among certain Bamboo participants—for older, more “reliable” analog technologies; and a fear of information overload, particularly as it affects younger students learning to organize their research and citations for the first time. Despite all of these reservations, the conference also made a tentative distinction between “standing faculty [who] are more about just writing books” and incoming hires who will “incorporate technology in the classroom. They use blackboard, can digitize clips, work with library, Instructional Technology. I look to my new colleagues for inspiration on how to use technology” (Project Bamboo, “Exercise 6A Scribe Notes”). Nowhere else was the paradox between the anxieties over DH and the expectation that professional academics are utilizing and advancing digital scholarship more apparent.

Indeed, these anxieties might be increasing, with not only an institutional expectation of two published books in humanities combined with the widespread failure of digital publication to produce advancement. One solution suggested “targeting select senior faculty who are not dinosaurs and developing their projects…. These are the people who can influence tenure decisions” (26), a view reinforced later that “rather than a parallel peer review system, what we should do is look for senior established colleagues who can take risks and would do something different—persuade them to publish something in a digital form. By taking that step it makes it more respectable for junior people to do” (33). But a step like this would have to address entrenched institution-wide predispositions that would not be easily overcome and that this could have a cumulative effect: “if they don’t get tenure at our place, it will be hard for them elsewhere” (26).

Unlike the almost contemporary and statistically based MLA report, these comments from the initial Bamboo conference are clearly largely impressionistic, albeit informed by many specific examples of the problems faced by DH. And these examples must be recognized as issuing from a self-selected group of international institutions that were all, to one extent or another, supportive of and involved in the new digital media, unlike the MLA respondents who were not so selected but instead represented language and literature programs throughout North America. If the Bamboo conferees could register fairly widespread unease about the status of DH, how widespread must the problem be in the academy at large? Of course, it may be that those at Bamboo, precisely because they are self-selected, are more aware of the problems and barriers facing digital scholarship than are the more comprehensive MLA respondents.

It is clearly appropriate to consider some hard cases. Both examples to follow come from my own experience, and both recognize the “resistance” that is my subject and (the first more than the second) some signs of overcoming this problem. I cite first the testimony of Dr. Stephen Brier, senior academic technology officer at my institution, the CUNY Graduate Center.

The two most striking elements of this testimony from an administrator (rather than the faculty and students recorded in the Bamboo conversations) are (1) the recognition that the impetus for institutional and scholarly advances in DH comes primarily from graduate students and only secondarily from doctoral faculty, and (2) there will be no easy or quick transformation toward DH within the “conventional academic arenas.” This latter prognosis reflects the uncertainties expressed by Ramsay and others about our current position in the progress of DH. The fact that the support of DH within the Graduate Center comes from two levels—the top administration and the present and future students—does reinforce that frequent complaint within Bamboo that the problems lie with entrenched senior faculty and with the departments they represent, and this condition is doubtless widespread in the current academy. There is, however, one further part of Brier’s institutional description that may make movement within CUNY easier than at other graduate and doctoral programs. Unlike the norm elsewhere, CUNY Graduate Center students typically teach (as independent instructors rather than as TAs) their own courses throughout the undergraduate (and sometimes graduate) levels. They thus get an opportunity to test and improve DH techniques in pedagogical as well as research circumstances and of course to react with (and learn from) the next generation of potential scholars in electronic environments. Even so, Brier is not “sanguine” (any more than MLA or Bamboo) that we should expect major reappraisals of the scholarly validity of DH research throughout the profession.

And that caveat leads to my second example—of a former graduate student of mine who had done a dissertation on the relations between electronic and print media and who then got a tenure-track position in a branch of a state university, continued to work on digital issues, and was then denied tenure and promotion, in a complex case that exemplifies several of the issues in this survey. For example, like the MLA’s reporting that all too few institutions had experience in evaluating DH scholarship, a senior administrator admitted that “I would be more confident of recommending promotion if I knew better how to assess [the candidate’s] achievement in digital scholarship” but lay the responsibility for providing this information on the specific department: “[i]t would help considerably if the Department [which had denied tenure] would clarify what it expects in the way of scholarship that justifies promotion,” a demonstration of (at best) an ambivalence to DH at midlevel and (at worst) an outright hostility. This same administrator wrote that “when I compare [the candidate’s] scholarly and professional activities to some elsewhere in the college who have achieved promotion to Associate Professor, I cannot see that [the candidate] has been less successful than them in showing a persistent, dedicated engagement with getting articles published and with involvement in a variety of professional activities.” It is, of course, very unlikely that a department denying tenure and promotion will be overtly dismissive of all forms of digital scholarship; but, when such a denial invokes the hoary distinction of “form” (i.e., medium) over “substance,” it seems fairly clear that this is indeed another “resistance” to the new medium: the negative evaluation from the department includes the comment that “[t]he evidence presented for the pedagogical value of [the candidate’s] work in digital humanities appears to pride form over substance.” There are those who have apparently still not been able to accept the dicta of Marshall McLuhan and for whom “medium” continues to be a nonsubject. When that medium is digital, substance is no longer visible to those who will not see. When this sort of prejudice occurs in promotion and tenure decisions, it raises the wider issue of whether such decisions should be based on form, substance, or both and what should be the relationship between the two.

And what one sees is obviously dependent on what one is looking for and the valuation one places on these data. A particularly apt demonstration of this question of fact and significance occurred in one of the “Humanities 2.0” series of articles in the New York Times. If we examine the content, methodology, results of, and response to “In 500 Billion Words, New Window on Culture” (Cohen), we can chart a range of opinions on the utility and critical value of the data. In brief, the report on the digitization of the five hundred billion words makes clear that the research should illuminate such fields as “lexicography, the evolution of grammar, collective memory, the adoption of technology, the pursuit of fame, censorship, and historical epidemiology” and that this new “culturomics …extends the boundaries of rigorous quantitative inquiry to a wide array of new phenomena spanning the social sciences and the humanities” (Michel).

Despite the evidence of the report itself (e.g., that the Nazi repression of references to Marc Chagall is very clearly documentable by the word frequency statistics, with only a single mention in German from 1933 through 1944), in the Times coverage of the research the well-known critic Louis Menand found the value of the statistics limited and claims for its relevance to humanities overblown. While acknowledging that the Google Books NGram Viewer could be particularly useful for linguists, he suggested that “obviously some of the claims [about cultural history] are a little exaggerated” (Menand). He was also troubled that, among the paper’s thirteen named authors, there was not a single humanist involved. “There’s not even a historian of the book connected to the project,” Menand noted. I share Menand’s unease that there appear to have been no humanists on board (unless one counts linguists in this group) and certainly nobody representing history of the book. In subsequent e-mail correspondence, Menand elaborated on his concerns: “What I had in mind was the various things the authors of the paper claimed to have proved by counting changes in the number of words or phrases. I thought it was a completely superficial way to do cultural history. It’s not that the data might not be relevant, but you’d have to do a lot more work before you could make some of the claims they were making—hence my noting that they did not have a humanist on their team. They seemed to think this could be a purely empirical exercise.”

Menand’s demurrals are well taken, and doubtless the involvement of a representative group of humanists would have affected the interpretative positions that the report covers. But I am less certain that quantitative analyses are a “superficial” way to provide information for “cultural history.” If that were so, then the research of the annaliste school of French history (challenging the previously dominant “great-figure” focus) would be vitiated in relegating its data and, yes, its quantification to a form of nonhumanistic “counting.” It does indeed require a humanistic intervention to try to characterize the raw evidence of the so-called Ngrams that the Google-powered research recorded and analyzed (and admittedly there has been some complaint that the Ngrams may be based on “dirty” data). But when the authors of the report on the perhaps unfortunately named “culturomics” note that the research should illuminate various “humanistic” fields and that this new “culturomics …extends the boundaries of rigorous quantitative inquiry to a wide array of new phenomena” spanning the social sciences and the humanities, I am not sure that these claims are necessarily “exaggerated” (Michel). In the Chagall example, the quantifiable “disappearance” in German-language publications during the Nazi years has to be set against the continued presence of the artist in languages other than German throughout the Nazi hegemony; and, even in German, the references pick up again after 1945. This quantification is “superficial” only in the sense that it provides a numerical series of raw facts; but the moment that we place these mere facts against the template of mid-twentieth-century political history, an interpretive strategy is forced upon us, a strategy that can then be checked and if necessary modified by running other Ngrams involving other (Jewish and non-Jewish) painters, other art forms, other places and types of publication, and other historical periods. Menand is right that the Ngram in and of itself does not prove anything beyond the limited values encoded in the statistics; but I suspect that much of the “resistance” that the traditional culture critic finds in culturomics is derived from the combination of the quantification itself and its being presented in a digital (i.e., machine-produced and machine-readable) format.

In other words, the addition of digitization to quantification makes the data seem even less humanistic and critical. The resistance is simply a more recent form of that distrust of “detail” that Julia Flanders finds in stylometrics and related “pedantic” studies. She cites a 1989 review of John Burrows, Computation into Criticism: “From his pages of tables, graphs, and figures he proves that each of the characters under consideration [from Jane Austen’s novels] has his or her own idiolect …While there seems to be no reason—or at least none that a non-statistician can see—to doubt the validity of his findings, the thought does occur as to whether five years of work …by Professor Burrows to tell his readers this was really necessary” (La Faye, 429). Flanders notes that La Faye has misunderstood Burrows’s argument; but, in addition, the relevance of the review is its resistance to (and devaluation of) the pedantry required to produce a quantifiable result—much the same position of the antiphilological critics I cited in the predecessor to this article (e.g., that “the prestige of fiddling with minute variants and bibliographical details should be low” [Sisson, 616]).

In the various forms of “resistance” encountered in this essay, I think we can see that the critical dismissal of Ngrams and culturomics, the devaluing of digital-based research, the institutional unwillingness to regard work conducted in an electronic medium as on a par with print, and the related continuation of the “gold standard” of the monograph in tenure and promotion decisions are all symptoms of the fact that, to cite again Morris Eave’s colorful figure, the rat has not yet been digested by the python.