Editor’s note: This is a guest posting by Karin Wulf, Director of the Omohundro Institute of Early American History and Culture and Professor of History at the College of William & Mary.

As a regular reader of The Scholarly Kitchen, I was delighted to meet Rick Anderson at the annual conference of the American Historical Association earlier this year. I was startled, however, by Rick’s description of the hostility towards open access (OA) that he perceived in the conversations he witnessed there — conversations that, as he put it, ‘(used) the kind of language to describe OA advocates that I’m used to hearing OA advocates use to describe commercial publishers’. Rick’s comment was just a brief response to Joe Esposito’s post “Making a Case for Open Access,” but I want to be clear: historians definitely aren’t hostile to access, or even necessarily to open access. In my experience, however, and certainly this is the case for the Institute I direct and the journal we publish, the William and Mary Quarterly, many historians are beginning to look more critically at the problems that open access proposes to solve, the monolithic solutions that are offered, and the potential consequences for scholars and their scholarship. What we find is troubling, even while there is tremendous enthusiasm among historians, as there has been for decades, for seeking ways to disseminate historical scholarship more widely.

The Scholarly Kitchen is an incredible resource about scholarly publishing; at the Omohundro Institute of Early American History and Culture, it’s even a home page for some of us. But open access discussions on the The Scholarly Kitchen, like almost all discussions, analyses, and policies concerning OA, focus on STEM (Science, Technology, Engineering and Mathematics). It’s not as if we don’t have ample evidence now that OA has been primarily driven by the very real problem with for-profit publishers of primarily STEM publications. But we also now have ample evidence of the great divide between the circumstances in STEM publishing that have resulted in the emergence of OA policies and mandates from funders, governments, and universities and the realities of humanities and social sciences (HSS) scholarly publishing. (A coherent and compelling statement about the very different impact and implications of open access in the humanities is the British Academy’s submission of evidence to the review of RCUK, summarizing and extending their report on Open Access Journals in Humanities and Social Sciences. David Crotty interviewed the report’s primary author, Chris Wickham, for the Kitchen.)

Yet it has been disconcerting to hear how slowly groups promoting and producing OA, including librarians, are incorporating that information. Sometimes it is clear that groups and individuals simply have different visions for scholarly communication generally, and OA is part of that. That’s fine, and we can debate the virtues and advantages—or not—of such visions. But what we can’t do is operate without a fundamental understanding of the virtues and advantages—or not—of scholarly communication as they have developed within STEM and HSS and within their particular disciplinary or field ecosystems.

There needs to be more and regular attention to the importance of heterogeneous models of access, dissemination, and production. Scholarship is developed in very different ways, within very distinctive research and publication ecosystems. No one would suggest that biologists and film scholars organize, finance, and undertake their research along similar lines. And we know very well that the resulting scholarship is not consumed in the same way. Why, then, should we assume that the results of that research–published scholarship—can be produced and disseminated in the same way? Especially when analyses now suggest that this is an unlikely—perhaps even undesirable—outcome.

It worries me deeply that most conversations flatten out these differences, and that few mention, let alone prioritize, the importance of the scholarship itself. Financing circulation is obviously important, but creating scholarship is a more complex and collaborative process than OA advocates and policies recognize or accommodate, involving many layers of skill and labor. I’m going to confine my remarks in this post to the way that humanities scholarship, and more specifically scholarship within my own discipline of history, is developed and produced within scholarly journals.

In the course of exploring OA issues, the Omohundro Institute staff and I have used the William and Mary Quarterly as a case study in a variety of ways, including as the basis for a paper on OA. We’ve now taken this a step further to document the editorial and production labor required to produce an article in our journal. Ours is the leading journal in early American history, an expansive, often multidisciplinary field that stretches across four-plus centuries, from the fifteenth through the early nineteenth, across the North American continent and around the Atlantic basin.

We used the occasion of an author’s recent exemption-from-copyright-restrictions request to research the production of the essay in question. We don’t have time to vigorously police the web for PDFs of WMQ articles, but we know plenty of them are out there, and when we notice them we do ask either the individual or the site to take them down and replace them with a stable link to JSTOR. The author who received a takedown notice from us via Academia.edu asked if we could please make an exception and allow him to keep his article posted there. (Among the interesting aspects of this was the subject line Academia.edu uses to inform authors of takedown notices, which includes a faux-cozy frowny face.)

In response to the author’s request, our staff perused the full 182-page file, including manuscripts, correspondence with readers and editors, and fact-checking, from original submission to final publication. Leaving aside the considerable time spent by four expert referees, a mapmaker, and our typesetter and printer, in-house time exceeded 130 hours. The results didn’t surprise me, or the staff, and they likely didn’t surprise the author either, given the amount of feedback and contact they’d had. But it does document exactly how the production of high-quality humanities scholarship is intensely collaborative—among scholars, and then among authors and editors.

Humanities scholarship is not a reporting of research results, but evidence-based argument developed through narrative and analysis. Published scholarship, either in a journal article or in a monograph (and often the former appears in revised form as part of the latter), usually begins as a conference or seminar paper, shared and read and commented on in informal exchange before entering the formal publication review process. That process includes evaluation by external peer reviewers (at our journal, 4-6 readers for each essay) and a substantive response from the Editor. The time that the Editor spent in soliciting and corresponding with reviewers as well as in crafting a ten-page response to our Academia.edu author’s original submission was time spent on an eventual publication. In fact, though, much of the Editor’s work is devoted to improving scholarship that will never be published in our journal. At the WMQ only about 8-10% of the essays that the Editor reviews and responds to move on to another round of revision and then to publication. Scholarship that is not published with us often goes on, improved by the process, to other journals. This intensive and collaborative process brings hundreds of scholars into conversation and critical exchange with every issue of the journal that is published. It is a vital piece of the scholarly ecosystem in our discipline.

The articles that do move into production and publication in the WMQ are copyedited by the Managing Editor, who has a graduate degree in early American history and more than ten years’ experience in the field. The copyediting, as well as the fact-checking of every citation (of primary or secondary references) by a team of William & Mary graduate students trained in editorial work, offers another set of exchanges with authors, refining their arguments—and often catching key errors, from major to minor, in the transcription or analysis of their source materials. Even the most careful scholars make mistakes; this is why scholarship is fundamentally collaborative, even for single-author publications (as is the case for most humanities work).

We hear a lot of talk about quality content. But the processes that produce it require time, money, and expertise. Charles Seife’s chilling recent account of failures in FDA research regulation and in subsequent peer-reviewed literature is a STEM example of the impact of sloppy work. And while bad history, unlike bad food or drugs, usually won’t kill you, is it really so unimportant to get things right—or to make them as strong and as accurate and as well-argued as they might be? Does it matter whether black soldiers fought for the Confederacy and that, until one of my colleagues vigorously pursued the issue, Virginia’s primary school textbooks stated that they did? Does it matter if an author misstates a figure, mistranscribes or mistranslates a passage, mischaracterizes secondary literature, or simply tangles their important historical argument so that it’s not likely to be as well understood as it should be? We think it does. And we think for our discipline it takes the kind of process described here, including editorial work, to produce high quality scholarship.

The current business model of charging subscription fees and holding copyright might make historical journals seem old-fashioned to some, but a long-standing practice of very modest subscription rates has both supported this system of scholarly development and been a key way to get historical journals disseminated as widely as possible. Historical journals have long been keen to enhance electronic circulation. The late lamented History Cooperative was a nonprofit, collective effort to make historical scholarship more accessible. Many journals on JSTOR now participate, as we do, in Register and Read, which offers free access to up to 78 articles a year. And we accept reductions in our subscription price for all sorts of programs in developing countries and elsewhere.

Let’s look at what kind of modest pricing we’re talking about. Our journal is online both through JSTOR’s current scholarship program and on Project Muse. Libraries can subscribe to the electronic journal for $93 a year ($99 if they want a print copy too). Of course, subscription prices are actually less important now for library budgets; Cost Per Use (CPU) is the key metric for assessing the value of a journal. I should note that we are lucky to be housed within the College of William & Mary’s own Swem Library, and we spend a lot of time talking with our library colleagues about the issues posed here. For Swem, serials have to meet the $50 CPU mark or they are likely to be cancelled–unless a department can make a persuasive case for retaining a journal despite a higher cost. Our journal, by sharp contrast, has a CPU of about ten cents. Among the libraries I’ve spoken with, large public and smaller private universities, I could not find an example of our CPU rising above twelve cents.

While historical journals in the United States are dominated by nonprofit publishers, and by an unspoken low-subscription-price pact, other humanities disciplines—and the journals within them—have other business models. The Modern Language Association is prominent in its advocacy of OA, for example, and its principal journal, Proceedings of the Modern Language Association, is indeed open access. To submit to the journal, however, one must be a (paying) member of the organization. And the MLA Bibliography is available by subscription only; our library could subscribe to it for about $8,000. As I pointed out to our erstwhile author, Academia.edu (the domain name was acquired from a subsidiary that registered it before the advent of restrictions allotting that designation to postsecondary institutions) has its own business model for sustainability, aiming to gain revenue from selling analytics.

A common refrain is that “society publishers” use subscription revenue to cover other organizational activities. I hesitate to invoke the old canard about historians, that our research often leads us to conclude that “things are complicated,” but it is true that this too is complicated. Society publishers are as variable across disciplines as are research and publication practices. The revenue from our very modest subscriptions, for example, supports about 30% of the direct cost of the journal, while the Omohundro Institute covers the rest. Even for journals that use freelance rather than in-house copyediting or that publish through a university press rather than manage their own publication, subscription revenues are vital to basic operations.

There are any number of serious topics worth continuing to explore when we think about open access and the humanities—or, for that matter, OA and any disciplinary, interdisciplinary, or multidisciplinary scholarship. “Prestige” as a measure of a journal’s value, the feasibility or desirability of mega-journals or production cooperatives, the distinction between for-profit and non-profit publishers in the ecologies of scholarship—these are just a few of the many OA-related issues worth mulling over. As we consider these topics, though, we must do so in the context of not simply the circulation of scholarship but also in its production.