Harold Varmus on the NIH Green Open Access Self-Archiving Mandate

---

SUMMARY: Harold Varmus thinks the NIH Green OA self-archiving mandate isn't enough because (1) it doesn't provide enough usage rights, (2) it is subject to embargoes, (3) it only covers research from mandating funders, and (4) it doesn't reform copyright transfer. This is a miscalculation of practical priorities and an underestimation of the technical power of Green OA, which resides in self-mandates by institutions (such as Harvard's), rather than just funder mandates like NIH's. Institutions are the producers of all research output, and their Green OA self-mandates ensure the self-archiving of all their own published article output, in all disciplines, funded or unfunded, in their own Institutional Repositories (IRs). Self-archiving provides for all the immediate access and usage needs of all individual researchers, webwide. Access to most deposited articles can already be set as Open Access immediately. For the rest, IRs' semi-automatic "email eprint request" Button provides almost-immediate access. Access embargoes will die under the growing pressure of universal Green OA's power and benefits. Institutions' own IRs are also the natural locus for mandating direct deposit by both institutional and funder mandates. Copyright retention is not necessary as a precondition for mandating Green OA and puts the adoption of Green OA mandates at risk by demanding too much. Once Green OA mandates generate universal Green OA, copyright retention will follow naturally of its own accord.

---

Varmus, Harold (2008) Progress toward Public Access to Science. PLoS Biol 6(4): e101 doi:10.1371/journal.pbio.0060101

Harold Varmus welcomes the NIH Green OA self-archiving mandate and the increased access it brings, but says it isn't enough because (1) it doesn't provide enough usage rights, (2) it is subject to embargoes, (3) it only covers research from mandating funders, (4) it doesn't reform copyright transfer. He instead stresses his preference for publishing in Gold OA journals, which provide (1) - (4). (He has said elsewhere that he does not consider Green OA to be OA.)

I believe Professor Varmus is mistaken on all four counts because of a miscalculation of practical priorities and an underestimation of the technical power of Green OA. Green OA can be mandated, whereas Gold OA cannot (and need not be). What is urgently needed by research and researchers today is OA, and that is provided immediately by Green OA. (Green OA is also likely to lead eventually to a transition to Gold OA.)

Professor Varmus also speaks of Green OA self-archiving as if it were a matter of central "public libraries" (like PubMed Central, which he co-founded, along with co-founding PLoS) that are "inherently archival" and provide only embargoed access rather than immediate access.

In reality the greatest power of Green OA self-archiving mandates resides mostly in self-mandates by institutions (such as Harvard's), rather than just funder mandates like NIH's. Institutions are the producers of all research output, and their Green OA self-mandates ensure the self-archiving of all their own published article output, in all disciplines, funded or unfunded, in their own Institutional Repositories (IRs). Those IRs are neither libraries nor archives. They are providers of immediate research access for would-be users worldwide, and they also provide an interim solution for usage needs during embargoes:

(1) USAGE RIGHTS. Self-archiving the author's final refereed draft (the "postprint") makes it possible for any user, webwide, to access, link, read, download, store, print-off, and data-mine the full text, as well as for search engines like google to harvest and invert it, for Google Scholar and OAIster to make it jointly searchable, for Citebase and Citeseer to provide download, citation and other ranking metrics, and for "public libraries" like PubMed Central to harvest them into archival central collections.

This provides for all the immediate access and usage needs of all individual researchers. Certain 3rd-party database, data-mining, and republication rights are still uncertain, but once Green OA mandates generate universal Green OA, these enhanced uses will follow naturally under the growing pressure generated by OA's demonstrated power and benefits to the worldwide research community. Over-reaching for Gold now risks losing the Green that is within our immediate grasp.

(2) ACCESS EMBARGOES. Access to 62% of deposited articles can already be set as Open Access immediately. For the remaining 38% of articles, they can be deposited immediately with access set as Closed Access during the embargo. IRs can all implement the semi-automatic "email eprint request" Button, which provides almost-immediate access even to these embargoed deposits. When individual users reach a Closed Access item, they paste their email addresses in a box provided by the IR, click, and the author receives an instant email request for the eprint. With one click, the author authorizes fulfilling the eprint request, and the IR automatically emails the eprint to the requester.

This provides for all the immediate access and usage needs of all individual researchers during any access embargo. Once Green OA mandates generate universal Green OA, access embargoes will die their well-deserved natural deaths of their own accord under the growing pressure generated by OA's demonstrated power and benefits to the worldwide research community. Again, grasp what is within reach first.

(3) UNFUNDED RESEARCH. Funder mandates only cover funded research, but they also encourage, complement and reinforce institutional mandates, which cover all research output, in all disciplines. Institutions' own IRs are also the natural, convergent locus for mandating direct deposit by both institutional and funder mandates. All IRs are OAI-compliant and interoperable, so their contents can be exported to funder repositories such as PubMed Central, and institutions can help monitor and ensure compliance with funder mandates as well as with their own institutional mandates -- but only if direct deposit itself is systematically convergent rather than diverging to multiple, arbitrary, institution-external deposit sites.

(4) COPYRIGHT RETENTION. Copyright retention is always welcome, but it is not only not necessary for providing Green OA, but, in asking for more than necessary, it risks making authors feel that it may put acceptance by their journal of choice at risk. Consequently, Harvard, for example, has found it necessary to add an opt-out clause to its copyright-retention mandate, which not only means that it is not really a mandate, but that it is not ensured of providing OA for all of its research output.

An immediate deposit IR mandate without opt-out (and with the Button) provides for all the immediate access and usage needs of all individual researchers, and once Green OA mandates generate universal Green OA, copyright retention will follow naturally of its own accord. First things first.

Stevan Harnad
American Scientist Open Access Forum