Literary Research: Costs and Impact


One of the standard labor practices of research universities is to hire, pay, and promote faculty members on the basis of the research they produce. In the humanities, professors write books and articles and universities reward them accordingly. The system amounts to a considerable expenditure for the institution and a significant portion of faculty time and energy. Is the outcome worth the investment? In the humanities, is research publication the best use of university resources and faculty talent?

This paper examines the system of research productivity in literary studies—its policies and expectations, and its costs and outputs. To illustrate the system, we review English departments at SUNY-Buffalo, University of Georgia, University of Illinois and University of Vermont. We focus on departments in order to add empirical evidence to the rising debate over the standing and costs of the humanities in research universities, and departments are the places in which personnel decisions primarily happen. We select English departments because English remains the most prominent literary research unit. English departments are also large enough to provide more reliable averages of faculty research activity than foreign language departments. We aim to measure the system of research productivity by departmental “wholes,” not to single out any particular book, essay, or faculty member for evaluation.

We choose public universities because of the availability of financial data, and we select these four departments because three of them represent different points in the rankings of English departments—none of them in the top 20—while the fourth isn’t ranked at all because it doesn’t have a doctoral program, though it still requires that faculty members conduct research. The spread demonstrates how deeply the research mandate has permeated public universities across the country and is no longer a feature of only a few elite institutions.1

Our review reveals that

  • Universities make substantial investments in faculty research through direct compensation—for example, in 2008–09 the University of Illinois paid its 57 regular English department faculty members $1.34 million dollars to conduct research.
  • Faculty members respond to this support by producing ample numbers of scholarly books and articles— for example, from 2004 to 2009, University of Georgia English professors published 22 authored or coauthored books, 15 edited or co-edited books, and 200 research articles.
  • Once those books and essays are published, the vast majority of them attract meager attention from other scholars—for example, of 16 research articles published by University of Vermont professors in 2004, 11 of them received 0–2 citations, three received 3–6 citations, one received seven citations, and one 11. Books receive more citations on average, but not enough to justify the labor that went into their making.

There is a glaring mismatch between the resources these universities and faculty members invest and the impact of most published scholarship. Despite the scant attention paid to this scholarship, a faculty member’s promotion and annual review depends heavily on the professor’s published work. A university’s resources and human capital is thereby squandered as highly-trained and intelligent professionals toil over projects that have little consequence.

Literary Research 

Many professors enjoy the work, finding it rewarding and helpful to their other professional duties, but if their books and essays do not find readers sufficient to justify the effort, the publication mandate falls short of its rationale, namely, to promote scholarly communication and the advancement of knowledge. To put it bluntly, universities ask English professors to labor upon projects of little value to others, incurring significant opportunity costs. Every hour a professor aims toward research is an hour not aimed toward other duties, duties which might more effectively support other parts of the university’s mission, such as mentoring undergraduates outside of class.


By the simple yardstick of output, in the last 50 years the enterprise of literary research has been an astounding success. In the fields of English and foreign languages and literatures, the number of annual scholarly publications (books, essays, reviews, dissertations, editions, notes, etc.) climbed from 13,757 in 1959 to around 70,000 in recent years.2 The Modern Language Association’s count of the number of periodicals that publish literary research and criticism rose from 1,139 in 1959 to 4,686 active today.

The explosion in scholarly production stems from the expansion of the professorate—but only in part. According to the MLA, not only have more individuals entered the fields, but more literature and language departments have asked them to produce more published work: “The demands placed on candidates for tenure, especially demands for publication, have been expanding in kind and increasing in quantity.”3 The percentage of departments that consider research more important than teaching has doubled since the 1960s.4 Of those departments, one-third expect development of a second book before they award tenure, the percentage of doctorate-granting departments alone raising that rate to 49.8 percent.5

As a result, an energetic cohort of language and literature scholars has arisen, junior and senior professors, adjuncts and lecturers, post-docs and graduate students laboring on research projects as a primary activity. As they do so, they must take into account the future of those projects as published items, for the university rewards only literary research that results in peer-reviewed books and essays in a timely fashion, a demand that influences the topics scholars select and the methods they employ. The works they produce take years of study, long hours spent in libraries working through dense and scattered materials. The cumulative effort claims an enormous share of time, resources, money, and intelligence. Individuals are trained, hired, promoted, and paid to publish; the competition for jobs and status raises the output needed for each person to keep pace. The MLA counts 700+ departments across the country demanding that faculty members issue books and articles, indicating that the old publish-or-perish formula which used to apply to a small elite group of schools has become a national policy steering more than 50,000 graduate student, lecturer, adjunct, tenure-track, and tenured language and literature practitioners and aspirants.

The system deserves some scrutiny. Does the research productivity model best serve the universities which implement it? More pointedly: Do the products of literary research justify the labor that goes into their making?

The question applies not to individuals, but to the system as a whole. Of course these publications are worth it to people who need to publish if they wish to obtain a job and keep it. Universities, however, must think in larger, policy- and cost-oriented terms, judging not the research record of a particular professor, but rather the research record of the entire department, weighing the impact of all its products against the financing of them. Hence, in reviewing its tenure and promotion policies and the financing of them, it cannot highlight only the successful books and essays of the past. The unsuccessful ones must be considered, too.

The following sections outline the issue in three parts with illustrations from English Departments at University of Georgia, SUNY-Buffalo, University of Vermont, and University of Illinois. The first section calculates the investments made in research activities of regular faculty members. We consider only one means of support, the budgeted portion of salary devoted to research. We ignore grants, fellowships, course releases, sabbaticals, travel reimbursements, and other subsidies for two reasons. One, they are so inconsistent that compiling reliable data on them over several years’ time is extremely difficult. Two, we wish to estimate the cost of research support as a regular policy of universities, not as a function of outside funders such as the National Endowment for the Humanities. We note, too, that English and foreign language professors on average receive little support apart from salary relative to the sciences. According to the Humanities Indicator Project, in 2008 academic spending on humanities research and development was only 0.49 percent of spending on science and engineering R & D.6 Still, in ignoring other inside and outside means of support, we should regard our cost estimates as conservative ones.

In the second section, we determine the production of research items by regular faculty members, those professors paid to produce research and evaluated on how much they produce. We use a simple tally: the number of books and articles generated in recent years. We take results for each department as a whole, emphasizing the system at the departmental level, not at the individual level. If some professors have multiple publications in recent years and others have few publications, we average them out and make no judgments about those individuals.

Finally, in the third section, we turn to the consumption of those works, our gauge being the number of times they have been cited by other scholars in subsequent books and articles. Here we invoke the ultimate purpose of scholarly publications, namely, assimilation of them by other scholars. If research is only intended to improve a professor’s teaching, then there is no reason for that professor to publish it. Those studies that do not issue in publication, however, are not considered in the evaluation of the professor at occasions of promotion or annual reviews. The system doesn’t require any kind of research at all—it demands published research, that is, research intended for others’ viewing. Citation, therefore, is a crucial indicator of value.


I n light of the many complaints that the humanities are underfunded, we should keep one basic fact in mind. It is that professors are hired and paid to do three things, research, teaching, and service. According to customary employment at research universities, then, regular English professors are supposed to divide their time between research, teaching, and service. Duties are outlined in a “Faculty Handbook,” with statements such as this from the University of Illinois:

For most faculty members, the primary basis for promotion and tenure will be evidence of high quality in both teaching and research, with consideration also being given to evidence of valuable public service or service to the University and to professional communities.7

And this from the University of Vermont handbook

In considering candidates for reappointment, promotion, and/or tenure, each candidate will be judged with respect to the proposed rank, status, and duties, considering his or her record of performance in teaching, scholarly research or other creative work, and service.8

Universities generally refrain from setting exact proportions for the time and energy faculty must devote to each pursuit, but at research institutions, research exertions always amount to at least one-third of a faculty member’s evaluation. Accordingly, at the very minimum, professors should devote one-third of their work to research, and we may regard one-third of a professor’s compensation as payment for research. A university that pays an associate professor $75,000 in annual salary regards at least $25,000 of that amount as research pay—once again, a conservative estimate.

Using that formula, if we add up the funds for research professors by income alone, disregarding various forms of administrative and supplemental pay and excluding lecturers, instructors, and creative writers, we get the following sums.


Each state university invests at least $23,000 per regular faculty member in research activity each year, with SUNY-Buffalo standing highest at nearly $29,000 per professor—a substantial part of the overall budget for the department. While the numbers above only apply to a single year, the average is unlikely to change very much over a few years’ time.

These payments direct English professors to allot 700 hours per year to research activities (one-third of a 40-hour per week work schedule), and a corresponding part of departments’ review, promotion, and pay is based upon the research publications that ensue. As a contractual matter, the university expects a return from its research employees in the form of published scholarship. If a professor does not yield sufficient publications in peer-reviewed forums, university policies make promotion virtually impossible.


I n the ladder of research publications in English fields, an authored book from a recognized scholarly press stands at the top. The MLA terms it “the holy grail for achieving tenure.”13 Articles in peerreviewed journals or in collections edited by notable figures and published by top presses come next.

At the English Departments listed above, over the last several years, professors have amply fulfilled their responsibilities, generating an impressive number of writings that evince toilsome years of inquiry and scholarly exchange.


The English Department at each state university issues thousands of pages of literary research every year. During the years covered, 2004–2009, Vermont produced more books than it had research faculty members, while Georgia produced five times as many essays as it had faculty members. Keep in mind, too, the tremendous labor it takes to research and write a literary monograph.

Many of these books and articles are of superb scholarly merit, having passed through diligent peer review processes. They demonstrate that these departments are dynamic centers of research in English. Professors have also published book reviews, encyclopedia entries, interviews, and occasional writings for popular periodicals and Web sites. They have presented research at conferences and colloquia, as well as executing confidential evaluations of others’ scholarship. Put all the research-related activities together and you get a portrait of eminently industrious research groups. The image of the tenured professor idling away in perfect job security, enjoying tenure without making further contributions to his or her field, doesn’t stand.

The prevailing pattern that we find in examining the curriculum vitae of faculty members is not that of a drop-off in publication once the job security of tenure arrives, but rather a continuation of publication at the same pace as before. Once they have earned tenure, most professors maintain their focus on research, perhaps because they have absorbed the research identity after so many years as a graduate student working on a dissertation and as an assistant professor compiling enough publications to win promotion, or because annual reviews continue after tenure and publication affects raises and further promotions. Either way, judged by material output alone, investment in literary research at these campuses yields abundant returns. Research universities consider the annual yield of books and articles by English professors to be, precisely, the goods they have paid for, those items enhancing the prestige of the institution and demonstrating the professional vitality of the department.


One: Essays Cited in Other Essays

The justification of a book or article doesn’t end with its publication. It proves itself by its reception among other scholars. Writings in literary studies rarely reach lay readers, and most of the sales of them come from library orders.

There is no practical way to register every time a teacher brings an idea from a book into a classroom, much less tally the episodes in which it causes an adjustment in how a reader regards the subject matter. We can begin, however, by following the sciences and counting citations. Citation counts are a crude indicator of value, to be sure, and existing search engines fall short of comprehensive for literary studies. The main instrument in the sciences, Thomson Reuters, picks up only a portion of humanities journals and no humanities monographs. For the humanities the best instrument is Google Scholar, which searches most recognized journals in literary studies, including those in Project Muse and in JSTOR (excluding book reviews). It includes only a fraction of monographs, though.

In spite of these drawbacks, the tracking of citations is a helpful start in determining what happens to books and articles once they appear. We start with the most reliable use of Google Scholar, citations of essays that appear in recognized scholarly journals (and some books).

  • Of 23 research articles published by University of Georgia professors in 2004, 16 of them received 0–2 citations, four of them garnered 3–6 citations, one garnered eight, one 11, and one 16.
  • Of 13 research articles published in 2004 by SUNY-Buffalo professors, 11 of them received 0–2 citations, one garnered five citations, one 12.
  • Of 16 research articles published by University of Vermont professors in 2004, 11 of them received 0–2 citations, three garnered 3-6 citations, one garnered seven, one 11.
  • Of 17 research articles published by University of Illinois English professors in 2004, 11 of them received 0–2 citations in subsequent years through 2010, four garnered 3–6 citations, one garnered 20, one 27.

According to these results, slightly less than one in eight research essays attract significant attention in scholarly periodicals in the six years after they are published.“Significant attention” is a fuzzy marker, of course, but a fair threshold may be set at an average of more than one citation per year. That is, if in the six years after publication an article received six or fewer citations, we consider that “insignificant attention,” while more than six counts as “significant.” Whether it takes place in a crowded field such as Shakespeare or feminist studies, or in a relatively small one such as John Dryden or medieval drama, notice taken in only one venue per year marks most cases as under-appreciated. We understand the many variables involved in a book’s or article’s reception, such as the trendiness of the topic, but we also maintain that a meager citation count for an article doesn’t justify the research behind it and the production of it no matter what the situation.

Because of the limits of Google Scholar, a search engine that included more humanities journals and related social science journals would raise the final tally of citations that appear in subsequent essays, but, I suspect, not significantly so. Most of the journal essays above appeared originally in a journal picked up by Google Scholar, and the hundreds of humanities and related social science journals from which the tool draws make up the vital center of the discipline. It is unlikely that the essays profiled here gained notable attention in journals outside it.

Two: Citation of Essays in Books—Estimates by Case Study

The citation of essays in subsequent books is another matter. Because of the limits of search engines with regards to scholarly books, registering the citation of essays in books published later on the same topic must be a piecemeal endeavor. Here, we select essays published several years ago, then review books published in the same field more recently for citations of those prior works. We use the MLA International Bibliography to gather the list of recent books, for instance, typing “Margaret Cavendish” into the subject box, then searching the books that come up for notices of an essay on Cavendish that appeared earlier.

  • An essay on Piers Plowman published in a 2004 collection of essays on the Middle Ages received one citation in 13 books published on Piers Plowman between 2007 and 2010. Google Scholar pulled up two citations from 2004 onward.
  • A 2002 essay on Margaret Cavendish in a prominent literary quarterly received one citation in the seven books published on the author between 2007 and 2010. Google Scholar pulled up two citations from 2004 onward.
  • An essay on Rudyard Kipling in a 2005 collection of scholarly essays on late-19th-century poetry received no citations in seven relevant books on the author published between 2007 and 2010. Google Scholar pulled up two citations from 2005 onward.
  • A 2000 essay on Henry James in a prominent literary journal received one citation in 23 relevant books on the author published between 2007 and 2010. Google Scholar pulled up seven citations from 2000 onward.
  • An essay on Emily Dickinson that appeared in a 2004 collection garnered one citation in 12 relevant books on the poet that appeared between 2007 and 2010. Google Scholar pulled up four citations from 2004 onward. Three other essays on Dickinson published 2003–05, two by the same professor, did not receive any citations.
  • One essay on Shakespeare that appeared in a 2001 collection on Ovid and Renaissance literature collected one citation in the 111 relevant books published on the poet in 2008 and 2009. Four other essays on Shakespeare by three other professors did not receive a single citation.

These are disappointing outcomes, with essays typically drawing one or zero citations in subsequent books. If they are representative, then the vast majority of the thousands of essays that are published every year in scholarly literary journals and in essay collections are virtually overlooked by monographs published on the same subjects. It is tempting to judge the success of literary research as a whole by its most successful examples, those essays that have acquired prominence and piled up high citation counts. In light of the findings above, however, singling out the few highly-cited essays misrepresents the nature and workings of the entire system of literary research production. In terms of a university’s expenditures, two associate professors with the same salary might have sharply different research outputs, but they still pose the same costs to the school. A star professor publishing an article in an elite journal that attracts many active researchers’ notice only forms one small part of a department’s budget and profile. The work of other professors make up the rest, and the inquiries we have outlined above indicate that most of their essay publications fall into the “insignificant attention” category. Their performances should be averaged together with the celebrated performances.

Also, we must add an important observation. These differing degrees of attention do not neatly correspond to scholarly quality. Many of these essays are not ignored because of poor scholarship. In fact, several of them are of excellent quality. They meet all the requirements of intelligent analysis, and they observe the protocols of academic research, with copious footnotes, high field awareness, and cogent arguments. Their topics are central to literary studies, too. In other words, there is nothing intrinsic to the essays that explains their neglect. Instead, we draw a systematic conclusion: in literary studies, the essay is a low-impact medium of scholarly communication. This is to be expected in a system that demands a book for promotion.

Three: Books Cited in Essays

That preference for books is reflected in their performance on the citation metric. If we apply Google Scholar to a sample of authored books published from 2002 to 2005 by current members of each department, we get the following results for citations in periodicals. 



  • Of the eight authored books published by University of Georgia professors between 2002 and 2005, six of them come in at 0–10, one at 19, and one at 59.
  • Of the 13 authored books published between 2002 and 2005 by current SUNY-Buffalo English professors, six of them received 0–10 citations, three received 11–20, two received 21–30, one 38, and one 110.
  • Of the eight authored books published by University of Vermont professors between 2002 and 2005, four of them received 0-10 citations and four of them received 11–20 (four of the top five books are film studies).
  • Of the 12 authored books published by University of Illinois professors between 2002 and 2005, four of them received 0–10 citations, one 11–20, two 21–30, two 31–40, one 41, one 57, and one 82.

Previously, we set the “significant attention” bar for essays at more than one per year. For books, which take much longer to produce, we should raise the threshold proportionately. We should also consider the size of the fields that the books fall into, expecting in smaller fields more than two citations per year, and in larger fields more than four citations per year. This makes the “10 or fewer” category always insignificant, the “more than 20” category always significant, and the “10–20” category uncertain. Given the fact that many citations are of the perfunctory kind, taking the form of “For more on such-and-such, see . . .,” we regard the threshold of significance (more than two citations a year) as a generous one.

For the books listed above, the breakdown is: 20 receiving insignificant attention, 12 significant attention, and nine uncertain.

Four: Books Cited in Books—Estimates by Case Study

To measure the citation of books in subsequent books, we followed the same course, selecting certain books on single-author or single-topic subjects and then searching relevant books published years later for notice of the original ones.

We examined five books, chosen because they have identifiable subject matters and can be tracked well using MLA bibliography tools, Google Scholar, and our own library investigations. We came up with the following results:

  • A 2000 book on Gerard Manley Hopkins garnered four citations in eight relevant books published on the poet in 2007–2010.
  • A 2000 book on William Wycherley was not cited in the one book published on the playwright in 2007–2010.
  • A 2003 book on Thomas Hardy garnered one citation in 16 relevant books on the writer published in 2007–2010.
  • A 2001 book on Bram Stoker garnered one citation in six relevant books on the novelist published in 2007–2010.
  • A 2000 book on representations of the Holocaust garnered six citations in 26 relevant books published on the topic in 2007–2010.

The Hopkins book and, to a lesser degree, the Holocaust book, earned significant notice, but others did not.

Five: Book Chapters Cited in Books—Estimates by Case Study

Finally, we did the same kind of examination of books with chapters on different authors and writings. These are books by scholars on a particular theme or topic or period, but with a long chapter on an author who is focused upon entirely by subsequent books. These volumes performed worst of all, yielding hardly any citations in subsequent books on those same authors and writings.

  • A 2000 book on sentimentality in American literature that contains a 35-page chapter on Mark Twain did not garner any citations in 25 relevant books published on Twain in 2007–2010.
  • A 2002 book that has a 36-page chapter on Andrew Marvell and a 37-page one on Aphra Behn garnered no citations in three relevant books on Marvell published 2007–2010 and no citations in ten relevant books on Behn published in 2007–2010.
  • A 2000 book with a 31-page chapter on Christopher Marlowe garnered one citation in eight relevant books published on Marlowe in 2007–2010.

Overall, in spite of the scattered examples of plentiful citation here and there in each of the five categories, the impact looks meager. We may admire those books and articles that inspired and influenced others, and we may admire, too, many of those works that did not. But this is to focus on discrete products, not on the system as a whole. If, in pursuing our own research and teaching, we borrow from this book and that essay, judging them sound and helpful studies, we vindicate the labor behind them. But if we incorporate into our judgment all the books and essays published in the field that we did not look at (or that, having looked at them, we didn’t invoke), even though they were relevant to the topic, then the system looks altogether different. One highly-cited book published in 2006 on X does not necessarily sustain the five other books published on X that year that passed into virtual oblivion. The labor of the unnoticed books should be added to the labor of the noticed book in our calculation of the viability of the research productivity mandate. If each book required four years of reading, research, and writing, and only one of those six books attracted significant attention, in our cost-benefit evaluation we should set that one book on the benefit side and 24 years of labor on the cost side.

We cannot expect every item of scholarship to collect large citation counts, of course, but if the system of scholarly research functions properly, the majority of them will. As a policy matter, the question becomes, “What rate of success should universities require out of literature and language departments?” Certainly, given the time-consuming nature of literary research and the abundant faculty talents dedicated to it, the portion of books and essays of scholarly merit that attract significant attention should surpass 50 percent. I consider this a low threshold, but it actually exceeds the current success rate by a considerable margin. If the clear majority of literary critical essays and books published by faculty members fail to make an impact, their universities must reconsider their commitment to the research enterprise.


At a time when higher education is facing budget cuts and criticism over rising tuition costs, decisionmakers at universities have an opportunity to review the purpose of their humanities departments. If English department publication yields benefits far less than the costs require, then a change in policy is in order. By lowering research demands on professors, universities may be able to steer them toward more productive and meaningful practices. Campus leaders may, in fact, find a grateful constituency among the faculty. I do not know of any English professor who is happy with the pace of publication that prevails at research institutions. They may enjoy the process of research, but the rising demand for publication output, for instance, the requirement that they have several essays and one book already published plus another project under way in order to earn tenure, places their research agendas on a hasty schedule. Projects that won’t fit the deadlines are avoided. Lines of inquiry that have no quick prospect of finding a publisher are avoided. Research becomes less exploratory and provisional, more aligned with prevailing trends and interests. Most importantly, scholars slip into a nerve-wracking schedule that is contrary to humanistic study (and emotional well-being). The best way to restore it is to solicit fewer printed pages from them.

Notes and References:

  1.  In the 2009 U.S. News & World Report rankings of English doctoral programs, Illinois stands at #22, SUNYBuffalo at #41, and Georgia at #58. See top-humanities-schools/english-rankings
  2. The Modern Language Association International Bibliography provides annual sums of publications in all areas of language and literary studies.
  3.  “Report of the MLA Task Force on Evaluating Scholarship for Tenure and Promotion,” Profession (2007): pp. 9-71; quotation on p. 10.
  4.  “Report of the MLA Task Force on Evaluating Scholarship for Tenure and Promotion,” p. 10.
  5.  “Report of the MLA Task Force on Evaluating Scholarship for Tenure and Promotion,” p. 10
  6.  Humanities Indicators, Humanities Resource Center Online, Section D, “Support for Humanities Research” (—accessed July 22, 2011).
  7.  University of Illinois Academic Staff Handbook,
  8.  University of Vermont, University Manual, Appendix B.1, Non-Unionized Faculty Handbook, p. 16
  9.  Information directly supplied by University of Georgia. I thank administrators there for their prompt response to requests for faculty data.
  10.  Information collected from the New York State Payroll Database, January 2009, located at
  11.  Information collected from University of Vermont Annual Lists of Base Pay, located at ~isis/?Page=bpay0.html.
  12.  Information collected from the University of Illinois Salary List 2008–09, located at details/UniversityOfIllinoisSalaryList2008-2009.
  13.  See MLA Ad Hoc Committee on the Future of Scholarly Publishing, “The Future of Scholarly Publishing,” Profession (2002): 177.
  14.  Information collected from faculty CVs provided directly by University of Georgia English Department.
  15.  Information collected from the SUNY-Buffalo English Department Web site ( faculty_staff/) and from searches on Google Scholar.
  16.  Information collected from CVs provided at University of Vermont English Department Web site (http://, plus Google Scholar.
  17.  Information collected from University of Illinois English Department Web site (http://www.english., plus Google Scholar.