Mandate Institutional Deposit -- Then Harvest Where You Please

1. The only substantive issue is how to get peer-reviewed journal articles to be made Open Access (OA), today.

2. Twenty years of evidence shows that -- except in the very few subfields that self-archive spontaneously, unmandated -- the only way to get those articles to be made OA is to mandate (require) that they be made OA.

3. Institutions are the source of all peer-reviewed journal articles, in all fields, funded and unfunded.

4. Authors who do not self-archive spontaneously, unmandated, can only be mandated to do it once (not multiple times, in multiple places).

5. The only ones that can systematically monitor and ensure that all of their research output, in all fields, funded and unfunded, is self-archived, in compliance with self-archiving mandates are authors' own institutions.

6. The only way institutions can systematically monitor and ensure that all of their research output is self-archived is if it is deposited, convergently, in their own institutional repository -- not if it is deposited, divergently, here and there, institution-externally. (Institutional back-harvesting of its own institution-external content is so unrealistic as to be hardly worthy of discussion.)

7. The metadata of institutionally deposited articles can be -- and are being -- harvested institution-externally by many harvesters (foremost among them being google and google scholar).

8. The full-texts of institutional deposits are being harvested too (by google and google scholar for sure) -- although for most purposes users only need a link to the full-text in the institutional repository.

9. The power and functionality of OA harvesters can and will be enhanced dramatically -- but not until much, much more of their target content is OA than is OA today.

10. Till then it's simply not worth most people's time to enhance functionality over such sparse content.

11. Which brings us straight back to the need for effective OA self-archiving mandates, systematically (hence institutionally) monitored to ensure compliance.

12. Arxiv's functionality does not come from the fact that its authors deposit directly in Arxiv: it comes from the fact that they deposit, and deposit reliably (near 100%), unmandated.

13. Ditto for those who share protein or crystallographic data centrally, unmandated.

14. The real problem is all of that vast majority of OA's target content that is not being deposited -- either institutionally or institution-externally -- because deposit has not yet been mandated.

15. Immediate deposit of all peer-reviewed research output can be mandated by both institutions and funders.

16. Immediate Open Access to the deposit would be desirable, but access to deposits can be embargoed, if there is a wish to comply with publisher embargoes on OA .

17. This compromise can and should be made, if necessary, in the interest of hastening and facilitating the universal adoption of immediate-deposit mandates by all institutions and funders.

18. (Institutional repositories' email-eprint-request Button is there to tide over user needs during embargoes.)

19. The other compromise that can and should be made, because it is indeed necessary, is not to insist prematurely on further rights -- over and above free online access -- that publishers are not yet willing to allow, such as text-mining, re-mix and re-publication rights.

20. First things first: Don't fail to grasp what's already within reach by over-reaching for what's not yet within reach: Don't let the perfect be the enemy of the good.

21. Mandate institutional deposit -- and let harvesters harvest where and when they please.