Going a Step Further Than Open Access and Open Source: COVE and the Promise of Open Assembly

Abstract

This articles asks if the principles of open source and open access are sufficient to safeguard our intellectual labor and to guard against the predatory logic of a world dominated by capitalist systems of production and dissemination. Both open source and open access face a similar problem, as it happens: neglect and obsolescence, as well as the most pernicious Achilles’ heel of the vast majority of digital humanities initiatives: long-term sustainability. COVE offers an alternative to both long-term sustainability and the collective sharing of content.

Two principles currently govern the work of the digital humanities: open access, the notion that content should be freely accessible to all without paywalls or other restrictions; and open source, software whose underlying source code is made freely available for reuse and modification. COVE, which stands for Collaborative Organization for Virtual Education at covecollective.org, subscribes to both principles: we support an open-access publication platform, COVE Editions, where we publish material such as scholarly editions after they are put through peer review, revision, and copyediting; we also support the open-source movement by using and modifying open-source tools like TimelineJS, Open Layers, Drupal, and Annotation Studio, and sharing our code through a GitHub repository.

However, COVE seeks to go a little further by asking if the principles of open source and open access are sufficient to safeguard our intellectual labor and to guard against the predatory logic of a world dominated by capitalist systems of production and dissemination. Both open source and open access face a similar problem, as it happens: neglect and obsolescence, as well as the most pernicious Achilles’ heel of the vast majority of digital humanities initiatives, long-term sustainability. COVE offers an alternative to both long-term sustainability and the collective sharing of content.

The Problem

There are so many examples of the problem of sustainability that I hardly have to list them. One need mention only the University of Maryland’s decision in 2018 to stop funding Romantic Circles, despite its impressive usage stats and the importance of the site as a trailblazing early digital humanities initiative. The site was forced to move to the University of Colorado, as a result. Even sites with massive amounts of investment are not safe. NINES at nines.org, with which I was involved from its inception, is an excellent example of what I mean. One would be hard-pressed to think of a humanities project with more investment tied to one institution, the University of Virginia: $2.8 million of support from Mellon to Jerome McGann there; two years of summer institute funding to Andrew Stauffer at the university from the National Endowment for the Humanities; $300,000 per year for five years directly from the University of Virginia; and a dedicated tenure line for NINES. And yet none of that ensured that the project could rely on funding from the university after a change of leadership there. The project has gone from over $4 million of support to almost nothing.

Similarly, a great deal of money has been spent on attempts to make open-source tools more sustainable and easily available across networks. In Europe, the Methods Network (2005–8) “was an AHRC-funded, multi-disciplinary partnership . . . providing a national forum for the exchange and dissemination of expertise in the use of Information and Communication Technologies.”¹ Around the same time, the Arts and Humanities Data Service (AHDS) aimed to “collect, preserve and promote the electronic resources which result from research and teaching in the arts and humanities.”² As these efforts were ending in Europe, Project Bamboo began in the United States. Funded generously by the Mellon Foundation from 2008 to 2012, Project Bamboo was a “multi-institutional, interdisciplinary, and inter-organizational effort” to “collectively tackle the question: How can we enhance arts and humanities research through the development of shared technology services?”³ Despite these massive investments, amounting to millions of euros (almost $3 million for Project Bamboo alone), almost nothing remains today of these initiatives. Project Bamboo was largely reduced to a listing of open-source tools dubbed the DiRT Directory—a collection that was perhaps most striking for the number of tools in the list that were broken, obsolete, or no longer maintained. Now, the DiRT site itself is no longer maintained and cannot be accessed as a live site.

A similar problem exists for open access, of course. Huge amounts of material have been digitized and made readily available to the public through projects like the HathiTrust and Project Gutenberg, but even if such resources have support that can ensure the persistence of the links (something that is usually not guaranteed), the individual texts sit statically on these sites requiring you to hunt for and collect them yourself. It is surprisingly difficult to use these and analogous resources to assemble sets of reliable texts to support integrated access and analysis—a point that applies to scholars working individually, with other scholars, or with students. Print anthologies, by contrast, are expensive and provide minimal scope for customization, collaborative analysis, or persistent annotation that can be consulted in different academic contexts.

One can forgive users for giving up and turning to for-profit databases like Eighteenth-Century Collections Online, Early English Books Online, or ProQuest instead, but there are serious sustainability issues with these resources as well, this time on the side of libraries, given their cost. The same goes for online teaching tools. Although many universities have hyped the importance of active learning or a “flipped classroom” (i.e., an approach in which students conduct hands-on, experiential, and often collaborative projects), scholars lack easy access to open-source platforms that will help them manage users and workflow. Scholars have generally been forced to choose proprietary, for-profit platforms like Perusall or to become full-blown coders so they can incorporate the most robust open-source tools on their own. As a result, an ever greater divide has emerged between digital humanities practitioners and those rooted in more traditional analytical methodologies. Universities, in the meantime, have been paying more every year for commercial databases and teaching tools, while prices for these items have increased beyond the rate of inflation. The situation was already at a breaking point well before the COVID-19 pandemic completely upended library and university budgets.

Robert Darnton put the problem well when, in 2012, Harvard Libraries stopped many of its subscriptions to databases from commercial providers. He was head of Harvard Libraries when he wrote,

And yet, as I have been laying out, open access and open source have their own problems that have allowed for-profit databases and teaching tools to maintain their stranglehold on university budgets.

Obviously, we need better, more sustainable solutions.

COVE’s Approach to Sustainability

COVE offers a possible way forward; as I explain at the end of this article and as I argue elsewhere, I am not proposing that COVE is a heroic solution that swoops in to fix all problems. Rather, it is a mechanism for the collective sharing of not only tools and content but also time and money. Before I explain, I want to underscore the challenge here; it is supremely difficult to maintain a platform such as COVE given the costs of development, platform upgrades, and good programmers. The biggest challenge is simply ensuring that there is enough money coming in each month to cover the costs of the infrastructure. COVE is still in what business people call the Valley of Death, a point between early adoption and widespread acceptance, between precarity and robust money flow. As a nonprofit, COVE also does not have access to the deep pockets of commercial ventures, which is what has allowed many such ventures to succeed despite being unprofitable for years. It took Facebook roughly five years before it could report its first profit; Amazon ten; Tesla thirteen.

An example of how the current system is clearly rigged is the federally funded Small Business Innovation Research and Small Business Technology Transfer (SBIR/STTR) program, which helps to fund the entrepreneurial ventures of STEM faculty at American universities. The federal government provides over $3.4 billion of investment every year to help university faculty draw angel investors to their projects, after which, of course, they gain even more investment. Naturally, when one has such deep pockets, it is easier to succeed, especially since the investments are designed precisely to help get companies over the negative cash flow of expected initial failure.

Nonprofits are not allowed to apply.

If a nonprofit with a vision for the collective sharing of goods somehow manages to succeed and challenge a for-profit company—as, for example, happened when Zotero threatened to steal Endnote’s market share—the for-profit company can use the court system to slap massively expensive, if baseless lawsuits onto the nonprofit, a mechanism designed to shut them down or buy them out. The nonprofit usually does not have the money to fend off the attack. Zotero succeeded only because it had the backing of George Mason University, where the individuals who created the software were employed.

Wary of such challenges, COVE aims to learn from the history of NINES. Rather than rely on a single institution for support, COVE has sought to bring together a variety of international funders while exploring a series of nonprofit cost-recovery mechanisms, including the following.

The COVE Consortium

A number of universities (Purdue, the University of Chicago, Rutgers, the University of Nebraska–Lincoln, the University of Florida, the University of Delaware, Texas Christian, Vanderbilt, Toronto Metropolitan University, Carleton, the University of Alberta, the University of Exeter, the University of Birmingham, Birkbeck, and Queen Mary) have agreed to provide financial support to pay their own graduate students in COVE training and work (text encoding, copyediting, GIS, social media, database management, and so on). Students are fairly paid and acquire skills that they can use in their own diverse career tracks (sometimes called “alt-ac”). We professionalize our research assistants, illustrating how to turn the skills students acquire at COVE into marketable résumés.

Scholarly Organizations

So far, COVE is supported by four scholarly organizations: the North American Victorian Studies Association, the British Association for Victorian Studies, the Australasian Victorian Studies Association, and the North American Society for the Study of Romanticism. Each organization not only provides yearly financial support but also helps to advertise COVE tools and content to their members. This arrangement also provides oversight and ensures that COVE has a clear succession plan, which is laid out in our constitution. Liaison officers from each organization sit on the COVE Advisory Board and provide oversight as full voting members of the board.

Foundations and Granting Agencies

A number of foundations that support nonprofit work have provided yearly or biannual contributions to COVE development, including the Curb Center at Vanderbilt University, the Sidney Stern Memorial Trust, the Druwé Fund at the University of Leuven, and the Painted Bunting Fund of the Tides Foundation in San Francisco. We are also extremely thankful to nonprofit entities providing grant support, especially Purdue University, the University of Nebraska–Lincoln, Research Society for the Study of Victorian Periodicals, and the National Endowment for the Humanities. Agencies such as the Arts and Humanities Research Council of the UK, the Social Sciences and Humanities Research Council of Canada, and the National Endowment for the Humanities, all of which have supported COVE generously, are the closest thing we have to an angel investor for nonprofit work, and such agencies are beginning to recognize the need for (and require) sustainability planning as part of their application process.

Individual Subscription

Although COVE provides material for free at our open-access publication space, COVE Editions, we have sought to support the initiative through micro-payments from students (just $10/student/year) in exchange for use of our teaching and publication tools (annotation, timeline-building, map-building, and gallery-building). Payments occur through NAVSA (processed using Stripe), after which students receive a login/password through OAuth authentication. The income is nonprofit and is directly reinvested in COVE.

Institutional Subscription

Our most promising bid for long-term sustainability is through institutional subscription to COVE Studio. COVE falls between an LMS tool and a static database like EEBO or ECCO, the latter of which has also begun to explore tool use, especially data mining. As we have proven with our current fourteen institutional subscribers (Auburn University, Auburn University at Montgomery, Coastal Carolina University, Columbia College Chicago, Old Dominion University, the Penn State system, Princeton, Purdue, Toronto Metropolitan University, l’Università di Macerata, The University of Alabama, the University of Chicago, the University of Nebraska–Lincoln, and the University of St. Thomas), our nonprofit business model works. We can be competitive in this market because student engagement ensures the cost per use for COVE’s content database—an important statistic for these libraries—is exceedingly low compared to that of commercial databases. We provide SAML authentication so users at subscribing institutions can login to COVE Studio for free using the same login/password used for institutional email accounts or eduroam. Payment is processed through Purdue and is reinvested in the platform, nonprofit. Such subscription to COVE Studio in turn supports our open-access publication space, COVE Editions.

Open Assembly: COVE’s Approach to Collective Editing, Teaching, and Publication

So, what is COVE and how does it work? COVE is logically split in two parts. Our password-protected space, COVE Studio, is where we provide primary works that are available for customized anthologies and group annotation; password authentication is key to protecting student privacy and social annotation in this space. COVE Editions is our open-access publication space where one can publish peer-reviewed work or create open-access resources with students using open-source tools that we have made easily interoperable.

COVE began in the Victorian period under the aegis of the North American Victorian Studies Association and has since expanded to other periods and languages beyond English, necessitating a change in the meaning of our acronym from the Central Online Victorian Educator to Collaborative Organization for Virtual Education. Not surprisingly, the site is still rich in content from the Victorian period, including, notably, collections of material by and about Charles Dickens, the Pre-Raphaelite Brotherhood, and Oscar Wilde. The site also makes use of the 1.25 million words of peer-reviewed scholarly content at BRANCH: Britain, Representation, and Nineteenth-Century History (http://branchcollective.org). So far, most works assembled and put through peer review at COVE Editions have been from the nineteenth century. Recent grants from the Research Society for Victorian Periodicals and the University of Nebraska–Lincoln have been devoted exclusively to recovering content by Black authors from the Victorian period, drawing from the archives of SOAS University of London and Adam Matthew digital databases. We are working on this project in collaboration with One More Voice (https://onemorevoice.org), a new digital humanities initiative (launched in June 2020) that focuses on recovering non-European contributions from nineteenth-century British colonial archives (led by COVE associate editor Adrian S. Wisnicki). COVE has also dedicated the 2021–22 school year to encoding entire anthologies of material from the nineteenth century, including British Women Poets of the Romantic Era (ed. Paula Feldman), Transatlantic Romanticism (ed. Lance Newman, Joel Pace, and Chris Koenig-Woodyard), and Literature and Culture at the Fin de Siècle (ed. Talia Schaffer).

Our goal at Studio and Editions is to take this rich content and to rethink how we approach collective editing, teaching, and publication, thus addressing a number of the problems listed above, including the following.

Our Cultural Heritage

The COVE Studio platform currently comprises over 14 million words of reliable, already-encoded, ADA-compliant primary texts (over 1,525 titles, with proper formatting that includes line numbers for poetry and images where they are significant for the original work) as well as 1.25 million words of peer-reviewed critical literature (over 150 titles). Thanks to these texts and the integrated digital tools it provides, COVE enables detailed study in a diverse range of settings, facilitates the work of multiple types of user groups, and allows users to create, develop, and reuse their custom scholarship.

The Canon

COVE has proposed an approach termed “open assembly”: the ability to draw on a large database of content to make customized collections that can also include self-created content. This approach frees teachers from the constraints of static codex and online databases. Consider the issue of teaching previously marginalized authors, for example, those who are lower-class, female, or Black, Indigenous, or people of color (BIPOC). Print anthologies have attempted to include more of these authors, but such anthologies are necessarily limited in what they can do. If a scholar wishes to teach such texts alongside more canonized literary work, they are now limited by the choices of an anthology’s editor, forced to assign an additional textbook that anthologizes exclusively female or BIPOC authors at great expense to students, or required to collect individual texts themselves.

Publishing

COVE has established a peer-review and copyediting publication pipeline at our open-access publication space, COVE Editions. COVE Editions has already shepherded over ten editions through peer review, revision, and final publication, with another fifteen editions at various stages of completion: https://editions.covecollective.org/editions. COVE Editions provides additional tools to support the publishing of material with other scholars or with students, including a map-builder (OpenLayers), a timeline-builder (TimelineJS), and a gallery-builder (Drupal). Scholars have also created peer-reviewed critical editions as the main project of a graduate seminar.

Teaching

COVE is currently being used for teaching by thousands of scholars and students from around the world: https://editions.covecollective.org/content/users-coves-tools. Over the period May to October 2020, COVE saw a 1,500 percent increase in the number of instructors making use of our tools and content in the classroom. Instructor and student numbers will continue to increase dramatically as new institutional subscriber sites continue to go online. COVE’s tools are designed so that scholars and students can start using them without having to learn any code and so that they can easily add images, audio, and film. The response from both students and teachers has been highly positive. Our approach ensures hands-on student engagement and learning, with the ultimate outcome being active research and discovery rather than rote memorization. The COVE approach, which supports asynchronous student research and annotation, is particularly useful for developing inclusive pedagogy that invites participation from minority, first-generation, and BIPOC or Black, Asian, minority ethnic (BAME) students who have had reason to see the traditional classroom as unsupportive.

Sustainability

COVE has established a viable sustainability mechanism through subscription while keeping content open access and tools affordable for both individual users and institutions. The goal has been to ensure that there will always be funding to support the initiative. Although COVE is nonprofit and 501(c)(3) through both the North American Victorian Studies Association (for OAuth) and Purdue (for institutional subscriptions), we have learned from industry how to reach and keep new users. Individuals can acquire subscriptions to COVE (just $10/year through Stripe payment and OAuth authentication), as can entire institutions (through SAML, the mechanism used by commercial providers selling databases to university libraries). COVE’s rates for such subscriptions are considerably lower than those of commercial providers. Most importantly, due to COVE’s noncommercial nature, all funds generated are reinvested directly in sustaining and continuing the development of the COVE website and its content. Our long-term goal is to develop a process for continued and reliable funding for new COVE work, an objective that distinguishes us from many prior initiatives of a similar nature.

Conclusion

Given the crises faced by the humanities in the present day and given the challenges in establishing alternatives, it is easy to become dispirited or simply to let the commercial giants win. True change will only happen, however, if we work together to make it so. There are no easy solutions. COVE, however, is an effort that aims to tackle directly some of the thorniest problems facing us, with the understanding that it can succeed only so long as it remains a collective effort. Indeed, we need to prepare ourselves, all of us, to work together, collaboratively—across organizations, disciplines, initiatives, institutions, and countries—to solve the crises of the humanities in creative ways.⁵

Notes

Bibliography