SUMMARY: At a time when the immediate problem for Open Access (OA) Institutional Repositories (IRs) is not withdrawal but deposit (IRs are still mostly empty), there is no need for an institutional or departmental mediator/moderator/approver phase in the self-archiving process. Authorised institutional authors should all be able to deposit/approve and delete/approve their own papers, instantaneously. If it is felt that there is a need for vetting deposits, let the deposits be monitored only after they have been successfully deposited and are visible in the IR. A distinction also needs to be made between (i) unrefereed preprints and (ii) refereed postprints of published articles. (i) If you want authors to be willing to deposit their unrefereed preprints at all, you must allow them to remove them at will, instantaneously. (Discourage removal, by all means, but don't disallow it.) (ii) For refereed postprints, 99% of the time authors will never want to remove them. They are published. The postprint is merely a supplement to the published version, for those would-be users who cannot afford access to the published version. The published version (at the publisher's website) cannot be withdrawn; so withdrawing the access-supplement in the author's own IR is in general pointless and counterproductive.
Technically speaking, the
GNU EPrints software allows deposited papers to be removed instantly by the author/depositor. But whether it is the author/depositor who can remove the paper or a mediator/moderator/approver is a matter of individual
IR (Institutional Repository)
policy, not of software capability. It is merely a permissions parameter-setting on the software.
Given that IRs' principal problem today is not withdrawal but deposit (IRs are still mostly empty), I strongly recommend that departments and institutions
drop the foolish and unnecessary mediator/moderator/approver phase, and set the parameter so that authorised institutional users (i.e., all employed researcher/authors) can deposit/approve and delete/approve their own papers, instantaneously.
There is no need to make the simple deposit process seem complicated or threatening by interposing a moderator into either the deposit or the withdrawal procedure.
(If it is felt that there is a need for vetting deposits, let the deposits be
monitored subsequently, not
moderated antecedently, and let the deposits [not the removals] be over-ruled by the monitor, as and when needed (plagiarism, libel, quackery),
after the deposit by the authorised institutional researcher/author or proxy has been made, not before [when it would needlessly hold up the deposit and frustrate authors, who need
encouragement today -- not the opposite, with foolish, arbitrary rules and delays].)
As to the worry about withdrawal in general: We are talking here about (i) unrefereed preprints and (ii) refereed postprints of published articles. This distinction needs to be borne clearly in mind, in setting IR policy:
UNREFEREED PREPRINTS: If you want authors to be willing to deposit their unrefereed preprints at all, you must allow them to remove them at will, instantaneously. (Discourage removal, by all means, but don't disallow it.)
(It is a good and useful author practice to self-archive preprints: it establishes priority, it elicits corrective peer feedback, it creates a historic record of stages of development of a work, it accelerates and increases research impact and progress. But if the institution imposes a foolishly draconian removal policy, authors will simply be discouraged from taking the useful step of depositing their unrefereed preprints in the first place).
PEER-REVIEWED POSTPRINTS: Here the fact is that 99% of the time authors will never want to remove postprints. They are published. The postprint is merely a supplement to the published version, for those would-be users who cannot afford access to the published version. The published version (at the publisher's website) cannot be withdrawn; so withdrawing the access-supplement in the author's own IR is in general not only counterproductive but pointless.
The (two) rare cases when it might be necessary to remove a postprint are these (the second turns out to be bogus too): (1) if an error is discovered in the author's self-archived version, and the author does not wish to just add an updated version, to be tagged as the latest version, but to remove the erroneous version too: this is reasonable, and can and should be facilitated by the IR policy and parameter-settings, not deterred; the original URL will become null, with a "redirect" to the new version;
(2) if the author has a (probably unnecessary) panic-attack about copyright after deposit: but here the removal should be strongly discouraged; instead, downgrading access -- from Open Access to Closed Access -- is the sensible option for such a case. Then the metadata remain visible, the full-text remains in the archive (but invisible) and the author can individually authorise eprints on a case by case basis to be emailed by the IR software to would-be users who request it semi-automatically via the EMAIL-EPRINT button (already implemented on both the GNU EPrints and the DSpace IR software).
Stevan Harnad
American Scientist Open Access Forum
