Publishing embargoes.

No Gravatar

Publishing embargoes seem a relatively new phenomenon, probably starting in areas of science when the data produced for a scientific article was considered more valuable than the narrative of that article. However, the concept of the embargo seems to be spreading to cover other aspects of publishing, and I came across one recently which appears to take such embargoes into new and uncharted territory.

One example (there are many others) of embargoes continuing to operate in the era of open science and open data relates to crystallographically derived coordinates for macromolecules. Biomolecular structures are allowed to be embargoed for a maximum of one year before becoming openly available or "released" (considered a friendlier term than embargo). A more recent phenomenon is of embargoes on press releases which may be prepared by authors and or publishers to accompany the appearance of any article considered especially newsworthy. The publisher will then request that the press release is only released to coincide with the actual publication time and date of the article itself. Both of these types of embargo are more or less accepted by both parties. But in the last five years or so, new types of embargo have been introduced and it is these I want to discuss here.

  1. The self-archive or "green open access" version of an article, in the form of the last author version of an accepted manuscript prior to copy-editing and other operations by a publisher. Such Green OA versions are now a mandatory requirement from funders (in the UK), arising from the need to conduct a "REF" or research excellence framework assessment of all (UK) universities every seven years or so. In order to allow assessors and funding councils unencumbered access to these research outputs, the authors must self-archive their publications in a suitable institutional repository. In general therefore, there should always exist two versions of any scientific paper authored within these guidelines, the AV (author version) and VoR (Version of Record, held by the publisher, and carrying the guarantee of peer review). Publishers now embargo author versions until the VoR version has been published, and sometimes even up to 18 months beyond this period.
  2. The "supporting information" or SI embargo. This is closely related to the crystallographic data embargo noted above, but it applies in general to most other data and information associated with an article. Until very recently, most SI was in fact handled by the publisher themselves, and so it was released at the same time as the article. Since it is becoming more common to deposit data and SI in a separate repository, some publishers mandate that the release dates of this material must not precede the article itself. Deposition of such data has also become a mandatory requirement from (UK) funders since May 2015, and I have blogged about such "research data management" often here. In effect, both the scientific article and the data supporting it achieve their own DOIs or persistent digital identifiers, allowing easy and independent access to either the article OR its data. In fact, assigning such a DOI has a more subtle effect; creating a DOI means that metadata describing the object is also created and then aggregated by the agency issuing the DOI such as CrossRef and DataCite. Importantly, one should note that SI which is handled purely by the publisher will not have its own separate DOI and it will not have its own metadata. The data metadata for example can include the DOI for the article, and vice versa. I have shown examples of the utility of such metadata for data in an earlier post.
  3. So now we come to the most recent embargo, which has surfaced since around May 2015, as increasingly data has become a first class object in its own right with its own DOI and importantly its own metadata. There is now evidence that some publishers are requesting that this very metadata about data is also subjected to an embargo, not to be released before the article which makes use of that data is itself released. So data can be deposited in "dark form" prior to a publication, but the metadata (which carries the date stamp and provenance for the deposition) may have to be "dark" or embargoed. Actually, this is not yet very common; for example I asked the Royal Society of Chemistry what their policy was, with the reply "the Royal Society of Chemistry wouldn’t require metadata about the data files to be embargoed".

We live in an era where the very careers of reseachers can be determined by their claim to priority about scientific discoveries. The date stamps for priority continue to be largely controlled and issued by publishers and some may decide that it will be in their business interests to extend their control to data. Perhaps they may even wish to control all aspects of publication including the data and its metadata, acting as self-proclaimed research facilitators.

At this moment, this has not happened; both data and its metadata can remain open and FAIR. Which is where I think we should go in the future in the interests of open science itself.

Tags: , , , , , , ,

Leave a Reply