Fixing the few flaws in the RCUK self-archiving mandates by pinning down WHEN and WHERE to deposit

---

SUMMARY: The three recent RCUK self-archiving mandates (ESRC, BBSRC, MRC) are extremely timely and welcome, but they still have two serious -- though easily remedied -- flaws. They are vague about both (1) WHEN and (2) WHERE research should be self-archived:
     (1) WHEN: It should be specified that the author's final, peer-reviewed, accepted draft should be deposited immediately upon acceptance for publication. Any allowable delay should pertain only to the date at which access to the deposited text is set as Open Access, not to the date at which the text is deposited. It should be strongly recommended to set access as Open Access immediately, but for articles from the 6% of journals that do not yet endorse immediate Open Access self-archiving, access can be set as Closed Access (for a maximum of 6 months). The semi-automatic EMAIL-EPRINT-REQUEST feature of the Institutional Repository software will allow the author to fulfill individual eprint requests from fellow-researchers during a Closed Access embargo interval.
     (2) WHERE: It should be specified that the deposit should be preferentially in the author's own institutional repository. Central repositories may harvest from the institutional repository if they wish, but the optimal practice, and the one that will scale to cover all researchers at all institutions, is to deposit locally; only if the institution does not yet have a repository should the author deposit directly in a central repository.

---

The three RCUK self-archiving mandates (ESRC, BBSRC, MRC), though extremely timely and welcome per se, are still needlessly wishy-washy about one important thing: When the deposit should take place. Some (ESRC and BBSRC) say, vaguely, "at the earliest opportunity." Others (BBSRC) say "within six months of publication". And there is also hedging with: "depends upon publishers' agreements with their author".

None of this is specific enough for a clear, effective mandate. These are open-ended loopholes needlessly inviting non-compliance or opt-out. But there is an extremely simple way to fix them, and I and others will be pushing very hard for the fix to be implemented and announced formally, so they can serve as emulable policy models for the rest of the research world.

(1) Deposit of the full text and metadata should take place immediately upon the article's acceptance for publication.

(2) The author, however, has a choice as to whether to set the deposited full-text immediately to Open Access or temporarily to Closed Access (for a maximum of 6 months, if necessary).

(3) Closed Access means that only the metadata (author, title, journal, date, etc.) are visible on the web. The full-text is not openly accessible.

(4) However, into the two most widely used institutional repository softwares, we have already built the semi-automatic EMAIL-EPRINT request button. Whenever a Closed Access text is needed, the would-be user pastes in his email address and clicks; that automatically sends an email request to the author, which contains a URL on which the author can click in order to have the full-text eprint emailed to the requester. (Individual researcher-to-researcher distribution of the researcher's own article by the researcher to individual requesters falls under Fair Use.)

(5) The permissible Closed Access interval should be capped at 6 months at the most.

So with immediate deposit mandated, but immediate OA-setting only strongly recommended where possible (94% of journals have already given immediate author OA self-archiving their green light), the EMAIL-EPRINT button will tide over any embargo period (max. 6 months) for the 6% of articles that are not immediately OA.

But the immediate-deposit requirement itself has to be made crystal clear, lest the mandate turn out to be so fuzzy and open-ended as to be no mandate at all!

Also, all councils should mandate deposit preferentially in the author's own institutional repository (IR). Central repositories can harvest from there, if they wish. Direct deposit in a central repository is unnecessary, nonoptimal, and does not scale; it should only be done in cases where an author's institution does not yet have an IR.

Stevan Harnad
American Scientist Open Access Forum