Why do we need journals anymore?

There was a recent email on PLANET Google Groups asking about what the ‘top journals’ were in the field. I wrote a response, while not directly answering the question. This post is an elaboration/edit of that post.

Do we need journals any more? I say this as/has been an associate editor for journals and appreciating the full scope of publication infrastructure that publishers provide. I also appreciate the need for cognitive shortcuts to quickly evaluate the value of research by looking to the reputation of the journal. Perhaps these shortcuts are useful to promotion and tenure committees.

But I wonder what the point of a journal is, from the perspective of the reader. If we take a journal to be a thematically coherent collection of published research that validates and fixes particular approaches, methods, questions and evidentiary standards for a field, then a quick peek at the some of recent issues of the ‘top’ journals invalidates this view (save for some special issues). The methodological and theoretical diversity is quite astounding and should rightly be celebrated. Perhaps, coherence is overrated.

One might argue that this diversity is precisely the point. The diversity in the bundle/collection allows for serendipitous discoveries, that the readers would not otherwise find when they are engaged in ‘project-directed’ reading. However, it has been a minute since I have engaged in a cover-cover reading/skimming of a journal. It is almost never the case, when the journal is largely available through electronic databases through the library and not a paper copy that arrives in my mailbox (forget about going to the stacks). And it is certainly never the case, when I use search engines such as Google Scholar or ResearchGate. In fact, it is because of the imprecise nature of the search results that I found the most useful things in the unlikeliest places, not because of curated collections.

We are in an age of the great unbundling. We no longer have to buy cable package subscriptions; we consume individual clips on YouTube. Yet the publishers still insist on selling bundles of bundles to libraries. But the dam is breaking. Universities are requiring self-archiving (largely open access) of journal articles (see http://roarmap.eprints.org/ for a list of policy adopters) . Many authors publish their works on sites such as ResearchGate and Mendeley. Funders are now prohibiting publications in journals that do not provide open access. Incentives are aligned as well; authors want wide distribution of their articles rather than bundled behind restricted access. Web, preprint servers (arXiv, SSRN etc.), social networks (e.g. ResearchGate) coupled with reasonably good search engines solve the ‘distribution’ problem. These new(ish) modes of distribution also solve the matching problem; matching authors to readers. We no longer have to rely on small number of contributors/content creators that published in particular venues (because disciplinary traditions, peer networks etc.). Nor do we have to rely on small audiences within the field, who subscribe to these journals. Journal as a distribution channel is largely passe.

So, why publish in journals at all? This is a different question than why publish journals. Ultimately, this circles back to Tej’s inquiry about the reputation of the journal. Journals/editors act as gate keepers, for better or for worse. Just the process of writing for a journal, often sharpens arguments. Revisions demanded by editors and reviewers often make the manuscript better. Being published in a journal, through algorithmic magic, places a work in a more visible position in the search results. These advantages are invisible to the reader. If this is the primary role for a journal then there are number of other models that can be adopted. See peer voting in StackOverflow, or SciRate, for example. Of course, I am not advocating unvetted comments section that is open to all.

Ultimately, we publish in journals not to reach the right audiences, but because that is what our professional standards, funding bodies and hiring committees demand. It is telling that both the lists Andrew forwarded, contain no open access journals (maybe this has to do with publication dates for these articles). It is also telling that these are rankings by Western academics with access to resources to access these journal bundles. These demands have become ends in and of themselves, rather than facilitating knowledge creation and standing on each other’s shoulders. It is time to rethink some of these.