Institutional and Central Repositories: Interactions

The JISC/SIRIS “Report of the Subject and Institutional Repositories Interactions Study” (November 2008) “was commissioned by JISC to produce a set of practical recommendations for steps that can be taken to improve the interactions between institutional and subject repositories in the UK” but it fails to make clear the single most important reason why Institutional Repositories’ “desired Â?critical massÂ? of content is far from having been achieved.”

The following has been repeatedly demonstrated (1) in cross-national, cross-disciplinary surveys (by Alma Swan, uncited in the report) on what authors state that they will and won’t do and (2) in outcome studies (by Arthur Sale, likewise uncited in the report) that confirm the survey findings, reporting what authors actually do:

Most authors will not deposit until and unless their universities and/or their funders make deposit mandatory. But if and when deposit is made mandatory, over 80% will deposit, and deposit willingly. (A further 15% will deposit reluctantly, and 5% will not comply with the mandate at all.) In contrast, the spontaneous (unmandated) deposit rate is and remains at about 15%, for years now (and adding incentives and assistance but no mandate only raises this deposit rate to about 30%).

The JISC/SIRIS report merely states: “Whether deposit of content is mandatory is a decision that will be made by each institution,” but it does not even list the necessity of mandating deposit as one of its recommendations, even though it is the crucial determinant of whether or not the institutional repository ever manages to attract its target content.

Nor does the JISC/SIRIS report indicate how institutional and funder mandates reinforce one another, nor how to make both mandates and locus of deposit systematically convergent and complementary (deposit institutionally, harvest centrally) rather than divergent and competitive — though surely that is the essence of “Subject and Institutional Repositories Interactions.”

There are now 58 deposit mandates already adopted worldwide (28 from universties/faculties, including Southampton, Glasgow, Liège, Harvard and Stanford, and 30 from funders, including 6/7 Research Councils UK, European Research Council and the US National Institutes of Health) plus at least 11 known mandate proposals pending (including a unanimous recommendation from the European Universities Association council, for its 791 member universities in 46 countries, plus a recommendation to the European Commission from the European Heads of Research Councils).

It is clear now that mandated OA self-archiving is the way that the world will reach universal OA at long last. Who will lead and who will follow will depend on who grasps this, at long last, and takes the initiative. Otherwise, there’s not much point in giving or taking advice on the interactions of empty repositories…

Swan, A., Needham, P., Probets, S., Muir, A., Oppenheim, C., OÂ?Brien, A., Hardy, R., Rowland, F. and Brown, S. (2005) Developing a model for e-prints and open access journal content in UK further and higher education. Learned Publishing, 18 (1). pp. 25-40.

Swan A, Needham P, Probets P, Muir A, O ‘Brien A, Oppenheim C, Hardy R and Rowland F (2004). Delivery, management and access model for E-prints and open access journals within further and higher education Report of JISC study. pp 1-121.

On 30-Nov-08, at 9:08 AM, Neil Jacobs (JISC) replied in JISC-REPOSITORIES:
“Thanks Stevan,

You’re right, of course, the report does not cover policies. The brief for the work was to look for practical ways that subject/funder and institutional repositories can work together within the constraints of the current policies of their host organisations. There are discussions to be had at the policy level, but we felt that there were also practical things to be done now, without waiting for that.”

Hi Neil,

I was referring to the JISC report’s recommendations, which mention a number of things, but not how to get the repositories filled (despite noting the problem that they are empty).

It seems to me that the practical problems of what to do with — and how to work together with — empty repositories are trumped by the practical problem of how to get the repositories filled.

Moreover, the solution to the practical problem of how the repositories (both institutional and subject/funder) can work together is by no means independent of the practical problem of how to get them filled — including the all-important question of the locus of direct deposit:

The crucial question (for both policy and practice) is whether direct deposit is to be divergent and competitive (as it is now, being sometimes institutional and sometimes central) or convergent and synergistic (as it can and ought to be), by systematically mandating convergent institutional deposit, mutually reinforced by both institutional and funder mandates, followed by central harvesting — rather than divergent, competing mandates requiring deposits willy-nilly, resulting in confusion, understandable resistance to divergent or double deposit, and, most important, the failure to capitalize on funder mandates so as to reinforce institutional mandates.

Institutions, after all, are the producers of all refereed research output, in all subjects, and whether funded or unfunded. Get all those institutions to provide OA to all their own refereed research output, and you have 100% OA (and all the central harvests from it that you like).

As it stands, however, funder and institutional mandates are pulling researchers needlessly in divergent directions. And (many) funder mandates in particular, instead of adding their full weight behind the drive to get all refereed research to be made OA, are thinking, parochially, only of their own funded fiefdom, by arbitrarily insisting on direct deposit in central repositories that could easily harvest instead from the institutional repositories, if convergent institutional deposit were mandated by all — with the bonus that all research, and all institutions, would be targeted by all mandates.

It is not too late to fix this. It is still early days. There is no need to take the status quo for granted, especially given that most repositories are still empty.

I hope the reply will not be the usual (1) “What about researchers whose institutions still don’t have IRs?“: Let those author’s deposit provisionally in DEPOT for now, from which they can be automatically exported to their IRs as soon as they are created, using the SWORD protocol. With all mandates converging systematically on IRs, you can be sure that this will greatly facilitate and accelerate both IR creation and IR deposit mandate adoption. But with just unfocussed attempts to accommodate to the recent, random, and unreflecting status quo, all that is guaranteed is to perpetuate it.

Nor is the right reply (2) “Since all repositories, institutional and subject/funder, are OAI-interoperable, it doesn’t matter where authors deposit!” Yes, they are interoperable, and yes, it would not matter where authors deposited — if they were indeed all depositing in one or the other. But most authors are not depositing, and that is the point. Moreover, most institutions are not mandating deposit at all yet and that is the other point. Funder mandates can help induce institutions — the universal research providers — to create IRs and to adopt institutional deposit mandates if the funder mandates are convergent on IR deposit. But funder mandates have the opposite effect if they instead insist on central deposit. So the fact that both types of repository are interoperable is beside the point.

Une puce à l’oreille (not to be confused with a gadfly),


The Immediate-Deposit/Optional Access (ID/OA) Mandate: Rationale and Model

Optimizing OA Self-Archiving Mandates: What? Where? When? Why? How?

How To Integrate University and Funder Open Access Mandates

Ian Stuart (IS) and Charles Oppenheim (CO) added, in JISC-REPOSITORIES:

IS: “Is [it] strictly true [that ‘Institutions, after all, are the producers of all refereed research output, in all subjects, and whether funded or unfunded.’]? My understanding was that, particularly on the social sciences arena, a number of academics continue to write & publish, even though they have retired. No? (another example of research output that requires the like of the Depot for an OA deposit service…)”

CO: “Ian is correct, especdially in the humanities. Also, there is a large amount of STM scholarly output that comes from industry, especially the pharmaceutical industry.”

Yes, both “problems” are trivially easily solved, and neither is an exception to the fact that both institutional and funder deposit mandates need to be systematically convergent — on institutional repository (IR) deposit — rather than arbitrarily divergent (on willy-nilly institutional and funder/subject repository deposit):

(1) The drivers of universal OA are institutional and funder mandates. So, by definition, a retired author, institutionally unaffiliated and unfunded, cannot be mandated to deposit! But if they are minded to deposit (as they certainly will be, as OA becomes universal thanks to OA mandates), they can deposit in back-up repositories like the UK’s DEPOT. (So, yes, DEPOT and similar back-up repositories elsewhere are an excellent resource for tiding over authors who do not have institutions, or whose institutions do not yet have IRs.)

(2) Or retired academics can deposit in the IR of their former institution: Universities not only have a long tradition of providing facilities for their retired academics, but they have a special interest, now, in showcasing their past and present research output. Whether officially “emeritus” or not, the research output of their own retired staff certainly falls within that ambit. (At UQAM, for example, retired faculty officially retain the right to deposit their publications in UQAM’s IR.)

(3) Universities also have a tradition of providing facilities, where possible, for unaffiliated academics. So, given that all IRs are interoperable, and given that we are talking about published, peer-reviewed research articles, a university tradition will quite naturally evolve for IRs to host the publications of unaffiliated scholars and scientists in need of a repository to make their published work OA.

(4) The pharmaceutical industry may well be secretive about its unpublished research, but it has exactly the same interest as all other research institutions in providing access to its own published research (that’s why they publish it!) — and in gaining access to the published research of other institutions. Hence it is likely that the R&D wing of pharmaceutical companies (and other industrial research producers) will have IRs and IR mandates of their own, as the movement toward universal OA grows. (And where not, other institutional hosts or DEPOT will provide the natural solution.)

Again, I have to remind us all that the glaring and pressing immediate problem today is gapingly empty IRs where they are fully available to all their institutional authors, who are simply not depositing of their own accord (“Zeno’s Paralysis“). The problem is not a press of authors desperate to deposit, but having no place to do so!

And the solution is institutional and funder deposit mandates, converging on the IRs of the universal providers of all research, in all subjects, funded and unfunded: the institutions (mostly universities, partly independent research institutions, partly the R&D of commercial institutions).

We should not now be fretting about borderline cases and putative exceptions whilst we still have not yet taken the obvious and long overdue steps needed to take care of the mainstay.

Stevan Harnad
American Scientist Open Access Forum