‘Predatory’ is a loaded term. Because it is commonly used to refer to animals that eat other animals, as in ‘birds of prey’, it has the effect of dehumanizing a person who is called a predator. In many such uses, including in the definition of a ‘sexual predator’ from various state laws, the word carries with it the idea that the person is violent, and even that they suffer from a ‘mental abnormality’. There is, of course, a somewhat more benign sense of predatory; the Oxford English Dictionary, for example, includes ‘ruthlessly exploitative’ as part of its definition in relation to persons. As we consider how ‘predatory’ has come to be applied to some publishing practices, and especially to open access (OA) publishers, it is helpful to keep this context in mind, since it helps us understand both the agenda behind the usage and its inapplicability in a great many cases.
In the discussions that have followed the decision to remove Jeffrey Beall’s list of ‘Potential, possible, or probable predatory scholarly open-access publishers’ from its internet home at the University of Colorado, Denver – it can still be found at other locations, but is not being maintained – only a few have examined or challenged the appellation ‘predatory’. Yet the word itself is polemical in this context, and, since Beall himself is a strident opponent of almost all open access, that polemic seems intentional. Worse, the use of ‘predatory’ as an umbrella term for all kinds of abuses hides the difference between practices that really are ‘ruthlessly exploitative’ and those that may well grow out of mere inexperience or lack of competence.
To understand how thoroughly this idea of predatory open access has led to conflations and confusions, one need only look at the article in The New York Times about the problem of less than scrupulous OA publishers that featured in Beall. The article begins with a cautionary tale about scholarly conferences that use confusing names, charge presenters a high fee for the privilege, and do not exercise much review over the presentations, and then quotes a professor of medicine and journal editor as calling ‘this phenomenon … the dark side of open access’. Unfortunately, the story the article tells has nothing at all to do with open access; it involves, as mentioned, a conference to which academics were invited, and at which they could, in fact, present, if they paid the fee and did not mind that the event was much less prestigious than suggested in the publicity. There is no mention of publication in the Times’ anecdote, whether open or not. This disconnect illustrates, I think, how completely the phrase ‘predatory open access’ has come to conflate all kinds of practices related to scholarship that are considered, by the person using the term, undesirable. It is this lazy usage of a loaded term in a loose and undisciplined way that rendered Beall’s list, as well as much of the discussion of that list, unhelpful.
Beall’s list, of course, has been challenged before. The point that I particularly want to discuss – that Beall’s list is both overinclusive and underinclusive – is suggested in a column by Monica Berger and Jill Cirasella in College and Research Libraries News called ‘Beyond Beall’s List: Better understanding predatory publishers’. Although their article continues to use the unfortunate terminology of predation, Berger and Cirasella actually introduce two important caveats about Beall’s categorization. They recognize, first, that low-quality, as well as predatory, publishing practices are ‘not unique to OA journals’. That is, Beall’s category is underinclusive; it does not fully encompass the problem we want to understand. And then, they note, citing Walt Crawford, that some of the publishing practices Beall condemns do not deserve to be called predatory; in this, his list is overinclusive, because it lumps together practices that are not predatory, but which may result from inexperience or amateurishness, with those that are truly exploitative. The demise of Beall’s list offers an opportunity for us to think more broadly, and with greater nuance, about a range of problematic practices in academic publishing, and to consider more thoughtful approaches to protecting scholarly authors. I want to suggest that a crowdsourced model, rather than lists created by individuals or small groups, is the best approach we can take at this time.
Thinking about business models
To give it its best (and most important) justification, the concern over predatory publishing – I will refer from here on to ‘highly exploitative publishing practices’ – is coincident with the development of article processing charges (APCs) as a way to finance OA publishing. APCs are not, themselves, an exploitative practice. They were developed initially by early OA publishers, notably the Public Library of Science, or PLOS, and quickly adopted by commercial publishers as a way to monetize the growing demand for open access. But the close tie between APCs and the growing awareness of publishing abuses makes it imperative for us to examine the range of business models for distributing scholarship.
APCs, of course, are one subcategory of so-called gold open access, which generally refers to articles, and usually journals, being distributed in an open manner from the moment of first publication. It is important to recognize that there are several business models encompassed in the idea of gold OA; in their recent article on the transition to open access, Martin Eve, Saskia de Vries and Johan Rooryck distinguish three types of open access under the ‘gold’ umbrella:
- hybrid publishing, where selected articles are made open upon payment of an APC in an otherwise subscription journal
- immediate (gold) OA that requires the payment of an author-side APC
- gold OA that does not require an author-side fee, often due to society or foundation support for the publication; this is sometimes called ‘platinum’ open access.
These three models differ somewhat in their potential for abuse. APCs, of course, have given rise to the highly exploitative practices that have recently received so much attention. Those forms of exploitation, as well as other potential abuses such as failing to make an APC-supported article actually openly accessible, are also inherent in the hybrid model. Platinum OA, however, is less subject to these abuses, because, as Eve, de Vries and Rooryck note, it does not invite the same concentration of costs on authors that defines APCs. In subscription models, of course, the costs are distributed to a group of subscribers, often educational institutions, whereas with APCs, costs are shifted to a group of individuals, which naturally invites abuse. The more distributed the support for a particular business model, or the more divorced that support is from market forces, the less potential there is for exploitation and abusive practice.
One important benefit of using APCs is that doing so lowers the economic barrier to entry into the publishing business. Organizations and even individuals are able to take advantage of the affordances of the internet, which significantly lowers the costs of publishing, and can finance their operations without significant capital by charging per article to cover their costs. This has been an important opportunity for many in the developing world, whose voices are sometimes excluded from traditional publishing, and many OA journals that have been condemned by Beall really only fell foul of his criteria because they were not well capitalized. This does not mean, however, that APCs have been an unmixed blessing for scholars in the developing world.
In fact, an ironic effect of the transition from subscriptions to APCs is that it does not eliminate the access problem, but merely reverses it; instead of readers, it is authors who are potentially excluded. Whereas access to scholarship published under a subscription model has been a problem for those without the means to pay the steep subscription costs, with APCs it is the authors who encounter an access barrier. Libraries, in fact, have begun creating funds to help faculty researchers pay these author-side fees, which is an attempt to even the playing field for publishing and reflects the ironic shift of the access difficulty. Researchers in the humanities and social sciences, where grant dollars to support research and publication are less common, are one group that suffers from the access problem created by APCs. But most damaging is the effect on researchers in the developing world, where the opportunity to publish in well-known journals based in Western Europe or North America can be closed off by APCs.
Green open access, where the author or author’s institution makes a copy of an article available through an OA repository, does not suffer from the same potential for abuse, but it is a piecemeal approach to openness, and still relies, very often, on the terms of a copyright transfer agreement. Commercial publishers continue to use these agreements to discourage such ‘self-archiving’, despite their proclamations of support for increased openness.
Finally, it is worth noting that Eve, de Vries and Rooryck propose a different business model to support open access, one that is founded on scholar-controlled and discipline-specific consortia that would fund specific sets of journals. As they report, this model is already beginning to be put into practice, and it could significantly reduce the potential for abusive publication practices. If academic institutions were to get behind this approach to scholarly communications, and transition their library budgets to this form of support for distributing scholarship, the transition to open access could be much smoother and less susceptible to undesirable business practices.
Focusing on practices, not labels
To move beyond the misapplication and prejudice to which the language of predatory publishing has been subject, it is useful not only to distinguish different business models, but also to recognize that there is a continuum of undesirable, and sometimes exploitative, practices that plague scholarly publishing. The reasons behind these practices are themselves found on a continuum, ranging from the outright desire to exploit others for profit, through greed, carelessness or incompetence, to inexperience and underfunding.
In his blog post on ‘Should We Retire the Term “Predatory Publishing”?’, Rick Anderson provides a helpful list of seven types of practice that reflect what he calls ‘scholarly bad faith’. Anderson’s list is useful, and we can correlate it nicely to our discussion of business models.
The first three types of bad faith that Anderson lists all seem to fall into the category of exploitation. Deceiving authors into paying for non-existent services, selling fake credentials, or using such credentials to gain advantage in the competitive world of scholarship all clearly involve a malicious intent and they are all practices that cut across the different types of publishing models. Deception to convince authors to pay a fee, which is the fundamental practice for which Beall condemned some OA publishers, is a version of Anderson’s first category, but it does not exhaust the practice.
In fact, Anderson’s next item in his list of bad faith practices – presenting a journal or book as rigorous and scholarly when it really is not – is a version of the same type of exploitation, and it has been happening in the subscription world for many years. The difference is simply that librarians have been the ones to identify and try to avoid this abuse, whereas in the APC environment, authors must do so. One could add that the big deals that most commercial publishers now offer are simply a way of hiding this practice and compelling institutions to pay for shoddy journals in order to gain access to quality ones.
Problems with peer review, which are certainly not limited to OA publications, deserve some special consideration when talking about the practice of selling something that is not as rigorous as it claims to be. Anderson’s list focuses on ‘bad faith’, but when we look at the history of poor reviewing of scholarly works, it is clear that the problem has many sources and diverse causes. The existence of ‘peer review rings’ that co-operate to review each other’s works positively and cite them often represents bad faith by the authors who participate, but probably just carelessness or incompetence on the part of the publisher who allows it to happen. Similarly, the recent controversy over a book on school desegregation that was sent for review to a person identified by some as a white supremacist shows a shocking inattention and lack of knowledge of the field, especially for a scholarly society and a university press, but was unlikely to be malicious. Finally, many problems with peer review seem to stem from publishing organizations that operate at high volume and lack either the staff or the expertise to really supervise the process. This difficulty is ubiquitous in scholarly publishing and probably is most common at the largest commercial publishing houses, although it also occurs, presumably, at small start-up publishers, many of whom are now open access and APC based.
Two of the practices that Anderson condemns as bad faith are based directly on pricing. The first, ‘leveraging monopoly power excessively to exact maximum revenues from academic customers’, sounds like a definition of the business model of Elsevier or the Nature Publishing Group. It is not restricted, however, to subscription prices; the pricing of APCs for hybrid or open journals from commercial publishers is also subject to this abuse. The $5,200 per article APC for publishing in Nature Communications is surely an example of an abuse of market power, as well as pricing that is exclusive of all but those with significant grant funding. This form of undesirable publishing practice clearly cuts across a variety of business models; it can be found, obviously, in APC-based journals, and especially, it seems, with hybrid publishing, as well as in subscription-based publications. The other pricing-base abuse that Anderson identifies, the packaging of low-quality journals with more desirable publications through publisher big deals, is obviously a subscription-based practice.
Finally, there are undesirable publication practices that do not obviously involve bad faith, but result from carelessness or incompetence. One complaint about Beall’s list, as has been noted, is that he conflated such practices, at least when found in OA journals, with those that were based on bad faith. Some of these practices, such as irregular or infrequent publication, are also found across the spectrum of publishing models. Librarians have been aware for years that certain publishers had difficulty publishing issues on a regular schedule; back when libraries had to check in each issue of thousands of subscription journals and claim missing issues, such publishers were widely known and discussed. This problem has not disappeared, and we should recognize that, in our new environment, when delayed publication occurs with an OA journal, the problem may be inadequate funding and few staff. Thus, the very feature that lowers the barrier to entry for many publishers – lower costs – can also create the situation where the publisher may have trouble getting issues ready for publication.
Once we accept that undesirable publication practices take many forms, stem from multiple causes, from bad faith to inexperience, and can be found in all of the different business models for scholarly publication, it is clear why so-called ‘blacklists’ like the one Beall maintained are not an adequate solution to the problem. So we must ask what other ways we might find to address the issue of negative publication practices, which is both complex and widespread.
Blacklists, whitelists and the value of experience
One suggestion that has been made by a number of people, including Rick Anderson in the blog post discussed above, is that we replace the ‘blacklist’ approach with its opposite, by creating ‘whitelists’ of approved journals that are, presumably, safe for authors (and subscribers?). Recently the publication analytics company Cabell’s International has taken this approach and is marketing both a blacklist and a whitelist. Putting aside the irony and opportunism of selling subscription access to a whitelist of approved journals, I want to suggest that whitelists suffer from most of the same problems that plague blacklists. If blacklists are subjective, only as good as the criteria they use, and in need of continuous updating, those same problems exist with the more positively framed whitelists.
A recent article by science journalist Tracy Vence in The Scientist discusses some of the problems with blacklists and catalogues several attempts to create discipline-specific whitelists. In general, the article notes that both efforts are time-intensive, requiring a lot of curation by academics who already feel overworked. That is certainly a problem that both sorts of lists share, but upon closer examination, there really seems very little difference, in terms of obstacles, between the efforts to create white and blacklists.
A principal complaint about Beall’s list was that it was a solo effort, with one person deciding what the criteria for inclusion on the list would be. Cabell’s promises that it will be transparent about the criteria it uses for its lists, and Vence’s article notes that several of the efforts to develop an approved list of journals have involved advisory boards. Nevertheless, predetermined lists of criteria are inevitably both underinclusive and overinclusive, in the ways discussed above. They cannot anticipate all the possible abuses to which the complex system of scholarly communications might be subject. They are usually the work of a small group of people, who naturally will develop the lists based on their own experience. They will miss some good journals, and include some that, at least at times, behave irresponsibly in one way or another.
Another way to phrase this problem is to ask if there would really be any journals at all on the list of responsible journals. Anderson suggests that big deals that are filled with sub-par journals in order to inflate statistics, and therefore, price, is an undesirable practice. He also considers excessive prices a problem. Would these criteria prevent inclusion of any journal from Elsevier or Taylor & Francis on a list of acceptable journals? If the criteria are broad enough, it might be difficult to populate any list, since the current system seems to be broken in so many different ways. These lists, in the end, are always selective and represent a limited perspective.
In her article for The Scientist, Vence offers the beginning of a path out of this dilemma, when she discusses the creation of discipline-specific lists of approved journals, in smaller fields like urology or emergency medicine. By focusing on a smaller field of study, it is possible to benefit from the perspectives and experiences of a larger percentage of the scholars who publish in that field. The logic of this, however, suggests that we should move beyond lists of publications, either positive or negative, and consider ways to crowdsource the task of gathering the actual experiences of those who publish scholarly work. In short, I want to suggest that we need a form of ‘Yelp’ for scholarly publication, where authors are able to report their specific experiences with a particular publisher and guide others on how best to navigate relationships with that publisher. This suggestion would eliminate the heavy lifting required of an individual or small editorial board, as well as the subjectivity inherent in such projects. It would focus on actual experience, rather than a predetermined list of criteria, so that unanticipated problems, as well as praise, could be included.
Ideally, we need to work to transition scholarly communication to a system that does not invite abuse and exploitation. Such a system would need to eliminate a commercial profit motive, regardless of whether that profit is generated from authors or subscribers. Some form of non-commercial open access, such as described by Eve, de Vries and Rooryck, should be our ultimate goal. In the meantime, however, we should look to authors to provide the perspective needed to evaluate specific publications. It is their experience that is fundamental to any system of scholarly communications. While no transition can be successful if it adds a significant amount of labor for faculty authors, the task of providing short reviews that recount their positive and negative experiences is something with which all faculty, in common with the rest of the consumer world, are increasingly familiar. By crowdsourcing the task of evaluation and tying it to actual experience, we can move beyond the problems that have been created by blacklisting and which are still inherent in whitelisting.