Recently British journalist Richard Poynder, a longtime observer of open access and scholarly communication, interviewed me. We discussed the history of Biomedical Digital Libraries, an open access journal I once edited, as well as my views on open access today. The interview is here.
My most cited quote is the following: "I must say that the framing of open access as a means of saving money has been and remains a serious strategic error.”
Yesterday I posted the interview to Facebook. Scott Plutchak and I began a dialogue about it in the comments, and I'll continue that dialogue with this post. But first I'll insert a quote from myself, and then Scott, from the comment stream:
Me: "There's no doubt that publishers of all fields are making significant contributions to, and investments in, scholarly communications. This is why I have no interest in driving publishers out of business. That said, publishers can make all of these investments because they have copyright control over the articles--the articles are their IP. And in the case of publicly funded research that's a troubling form of IP. In my perfect world STEM articles would be simply produced, [Andrew] Odlyzko style [Odlyzko argues that scholarly publishing can be much cheaper]. The real money for publishers would come from licensing tools that help mine and filter the literature. I don't doubt that the staff of BMJ, NEJM et al make significant efforts under the current model. I'm saying that model could be different."
Scott: "But a high quality article isn't "simply produced." It's a collaborative project that involves the authors, the reviewers, the editorial staff. Who doublechecks the statistics to make sure they hold up? Who makes sure the images haven't been manipulated? Who works with the authors on the figures to make sure they're as clear and correct as they need to be, sometimes adding multimedia? Look at what eLife is doing with moving the peer review process beyond just "open" into a collaborative production. Look at what the BMJ does in managing conflicts of interest and requirements for transparency of data, particularly with clinical trials, These publishers ARE changing the model, they're embedding the article in a rich contextual environment that links to the rest of the literature and that increasingly connects to contextualized datasets, while also investing significantly in tools and processes to improve the integrity of the ecosystem. This is a complex and very expensive process but it seems to me to be the direction that we ought to embrace. A world of "simply produced" traditional format articles associated with some data mining & filtering tools looks like a step backward to me."
And now, to continue. I realized after reading Scott's reply that "simply produced" established a strawman which I did not mean to create. Andrew Odlyzko is an academic (a brilliant one), and for his purposes some light copy-editing and straightforward article posting fits the bill. But that is insufficient for stewardship of the scholarly record, particularly for ensuring its preservation and establishing contextual linkages between the articles and other artifacts in the bioscience literature. Publishers do add value. (Yes, here I have seamlessly moved from all of scholarship to the biosciences specifically--because the biosciences are what I know best, and is also the content that has been the sharpest focus of calls for open access.)
Scott cites two bioscience publishers in his comments--eLife and the BMJ. It is important to note that both of these publications are open access. So the careful curation and bold experimentation that Scott describes occurs in an ecosystem in which all of the content is available to all readers. This is not cheap; as I said, the focus on cost reduction as the prime motivation for open access is a "serious strategic error." But it does show that open access can be compatible with thriving publications. Open access presents no inherent threat to the viability of publishers, even if it does challenge traditional funding streams.
Many publishers have taken a middle ground thus far--not fully open access, not fully subscription-based. One such publication is the venerable NEJM. Perusing their table of contents today I discovered one article which is open access and another which is only available to subscribers. There are many more articles than these two, of course; most are subscriber only, but a substantial portion are open access. I am focusing on these two for the purpose of illustration.
The NEJM's subscriber-only article is funded by the NIH, as shown in this statement: "Supported by grants from the National Institutes of Allergy and Infectious Diseases (1U01 HHSN272200700031C, to Dr. Chambers) and the National Center for Research Resources (UL1RR033176, now at the National Center for Advancing Translational Sciences, UL1TR000124)."
Hmm. That statement contains lot of rich metadata which could link this article back to the grant submission as well as any other reports required by NIH. (At the moment there are no hyperlinks from the grant numbers to any other content.) A subscriber-only article which reports on publicly funded research should provide easy links to the public records that describe and present that research.
The CHORUS system developed by publishers routes readers to the version of record for an article, but that version will often be under embargo. I'm proposing links from the article to official grant records, which would be available immediately to all. On the Scholarly Kitchen David Wojick has advocated the same thing.
While proactive linkages between articles and grant records will always be useful, they will eventually be less essential as a means of providing access. In the long run immediate open access to the version of record for articles in the biosciences--no more embargoes--will be the norm. In such a world publishers like eLife and BMJ and others will continue to thrive. They won't be able to control access, but they will be able to develop and license tools (data mining is just one example) that help users make the most of their content.