Thursday, February 11, 2016

Are ‘predatory’ journals completely negative, or also a sign of something positive? *

* This post was first published on the SciELO in Perspective blog, on February 2nd, 2016 (and translations in Portuguese and Spanish are available on SciELO in Perspective as well).

It’s not nice for parents to find that their toddler child is telling lies – or at least trying to tell lies. But is it so bad? It seems that the ability to tell lies is a sign of a well-developing theory of mind. And a well-developed theory of mind is very helpful in social interaction and things like having empathy would be very difficult without it. So if your child is beginning to tell lies, it might be cause for celebration rather than regret! [1] Of course, it doesn’t mean that lying should be unconditionally encouraged, but a well-developed theory of mind may actually help in developing proper morals as well when the child grows up.

The emergence of so-called ‘predatory journals’ (see Beall’s list of predatory publishers) could be seen in a similar light. Predatory journals are not desirable, it goes without saying, but the fact that they come about is a sign of a developing market, and a true market in scientific publishing services is a good thing, in my view. A true market offers choice to the party who pays. In the classical subscription model of scholarly journal literature, the party who pays is typically the institutional library, and they have little or no choice as to which journal to subscribe to and which not, if the journals are considered to be relevant to the research being carried out at their institution. And if they are forced to make a choice, for instance because of lack of funds, it is an agonising decision which choice to make.

In some countries – in Latin America, but also The Netherlands, for instance – the government is playing a direct role in negotiations with publishers to provide access to scientific literature. In principle, this results in a single buyer, countering the effect of a publisher monopoly by being able to bring much more negotiating clout to the table, and as a result quite possibly wider access (as is indeed being reported from Latin America [2]). But it doesn’t amount to a functioning market. The absence of a real choice is not remedied. Even a government cannot just choose to provide access to one publisher’s content at the exclusion of others, simply because no single publisher provides all the content the research community needs. And a monopoly (single provider)/monopsony (single buyer) system has drawbacks as well. It has a tendency to lock in a certain balance of powers (or an imbalance; Stockholm Syndrome has been mentioned to me in that regard), and in doing so, presents an impediment to progress in the quest for increasing efficiency at decreasing cost. It removes, or at least diminishes, the incentives of individual or small players to decide to buy or sell at a range of prices and service levels. For authors, for instance, it removes any economic motivation to decide to publish in low-APC journals or to communicate their research results via preprint servers. And it makes it very difficult for new publishing initiatives to gain a foothold.

A true functioning market is not just ‘peaches and cream’, though, since any true market also attracts rogues. And predatory journals certainly are. (By the way, I remain to be convinced that all publishers and journals on Beall’s list are indeed genuinely predatory; it is my impression that some of his criteria are rather shaky.) The existence of rogues is an inevitability of human nature, I’m afraid, but to authors, the attraction, insofar as there is one, of potentially predatory journals comes in part from their usually low charges and quick publication.

Yet, even with the drawback of being polluted by predatory journals, a functioning market is preferable to a quasi-market, completely dominated by monopolies or monopoly-like players. A system of subscriptions, in which the party who pays – the institutional library – has practically no meaningful choice of what to buy, differs from one of article processing charges (APCs, which make open access possible), in that the party who pays ­– the author – is the party who does have a meaningful choice of where to submit and publish. So ‘flipping’ the system from subscriptions to APCs does deliver something much more akin to a functioning market, and ‘caveat emptor’, ‘buyer beware’, applies to all markets.

It is not strictly true, of course, that in an APC system the author pays. It is most often the funder who does, by including the means to pay in the authors’ grants. This leads me to think that the subscription system could also be made to work as a quasi-market, if authors (the ones having the real choice) were made aware of the cost of the choices they make. This would be the case if authors had to find or request money in their grants for articles published in subscription journals as well, and not just for articles published in APC-supported open access ones. Imagine what will happen if authors were presented with a bill – of, say, $5000, a reasonable estimate of the collective cost per article of subscription journals; more for so-called ‘prestige’ journals – by their institution for every article they publish in a subscription journal. The likelihood is that they will more often choose APC-supported open venues, especially when it is slowly dawning upon the scientific community that openness in itself is an essential part of the quality of a published article [3].

Making researchers responsible for the financial aspects of their publishing decisions also fits into the general logic of public policies, which put the responsibility of how grants are used on the researchers, after all. Besides, responsibilities that focus on the per-article publishing costs incurred are naturally more scalable with the growth of research output.

This may even help solving the underlying problem with the current journal publishing system: the conflation of scientific communication and career-advancing reputation management [4]. This goes beyond what publishers do: in many cases, the same institutions in charge of purchase decisions of journals are also in charge of research promotion and evaluation, which is mostly – if unjustifiably – based on journal prestige, which may well constitute a conflict of interest. The consequence of separating the two may be that more articles are posted as so-called ‘preprints’ (for communication) and that not all of those are subsequently published in journals (for career purposes), given the cost of the latter. Global preprint repositories such as arΧiv, bioRχiv, and others, would acquire critical importance in that scenario. I am not dismissing the need for reputation building and career progression, but should that really stand in the way of communication for the sake of scientific progress? Is it right that the results of research – particularly of research carried out supported by public funds – should be used primarily for the purpose of authors’ reputation stratification (establishing a ‘pecking order’) via journals, many of which are not openly accessible?

The publishers of journals will say that establishing a pecking order is not their only task, or even their most important one. Most of them maintain that their gate-keeping function, via the peer review they commission, is their true raison d’être. This is a strange argument, as they are only in control of the gate to their tiny little patch of the vast forest. Virtually every article can find a journal in which to be published and so be added to the scholarly literature. What is presented as gate-keeping is in effect just sorting and ranking articles according to vague criteria such as ‘quality’ or ‘relevance’ and therefore not distinguishable from establishing a pecking order. It may help some scientists, but it doesn’t help science. On the contrary: scientific communication is being held hostage by pecking order concerns. And what’s more, these efforts are costing the scientific community a great deal. In time, money, and impediments to sharing knowledge.

It bears repeating: scientific communication and reputation management should not be combined in the same system.

Apart from the obvious costs mentioned, this kind of ‘gate-keeping’ is abusing peer review. Not only does it cause ‘cascading’, many articles being rejected, resubmitted elsewhere, being rejected again, et cetera, until they are eventually accepted. At each stage needing peer review again, and so being a tremendous burden on peer reviewers. Peer review should be aimed at helping authors to improve their articles, by questioning assertions, methods, and the like, and only as a last resort rejecting for publication (in the way of some open access journals such as PLOS-One). In this light, the emergence of ‘easy’ journals, even ‘predatory’ ones, is not miraculous. It takes some getting used to, I presume, but I see great potential in an approach like this: author-arranged open peer review [5].

Jan Velterop

References
1. PAN DING, X; et al. Theory-of-Mind Training Causes Honest Young Children to Lie. Psychological Science. 2015, vol. 26 nº 11, pp. 1812-1821. DOI: 10.1177/0956797615604628
2. Personal communication.
3. VELTEROP, J. Openness and quality of a published article. SciELO in Perspective. [viewed 27 January 2016]. Available from: http://blog.scielo.org/en/2015/12/16/openness-and-quality-of-a-published-article/
4. VELTEROP, J. Science (which needs communication) first, careers (which need selectivity) later. SciELO in Perspective. [viewed 27 January 2016]. Available from: http://blog.scielo.org/en/2015/10/29/science-which-needs-communication-first-careers-which-need-selectivity-later/
5. Peer Review by Endorsement (PRE). ScienceOpen. Available from: http://about.scienceopen.com/peer-review-by-endorsement-pre/