Re: Response to J. Kleiner - Institutional Repository Idea

["David Groenewegen" <d.groenewegen@ballarat.edu.au>]

Joe,

This is a most interesting idea, but it occurs to me that your estimate on
the number of DRs you would need is way too high. While there may be
24,000 journals in the world there are not 24,000 areas of study - many
existing journals are either be in direct competition for papers in the
same area (Tetrahedron Letters, Organics Letters and SynLett for instance
cover much the same turf) , or reflect regional societies that cover the
same area - many countries have learned societies in Physics, Chemistry
etc etc, all of them producing a journal. Duplicating this structure is
pointless, and would only make the job harder.

As arXiv.org has shown, the old divisions (which were based on specialised
audience and the need to keep print titles relevant and affordable) need
not and indeed should not apply. Any physics article can find a home
there, regardless of which specialised print journal it was published in.
Better to have fewer DRs (which is also far more manageable), but more
mirrors to share the load. This also allows those without the resources to
manage a DR the ability to host a mirror, which is less work, but still
critical. It also makes for more copies for preservation purposes.

David Groenewegen
Information Resources Management Librarian
Information Services
University of Ballarat
Ballarat VIC 3353
AUSTRALIA
email: d.groenewegen@ballarat.edu.au 

>>> espositoj@gmail.com 10/19/05 8:19 am >>>

Janellyn Kleiner's recent post about what positive steps libraries could
be taking prompts me to share an idea I have been nursing for some time
about institutional repositories. I hasten to add that I am not a
librarian myself and that this long post may be even screwier than my
usual.

Institutional repositories (IRs) align themselves, understandably, with
their parent institutions.  Since most institutions at least in part serve
undergraduates, for whom the goal of creating "the well-rounded person"
has not been entirely abandoned, IRs set out to cover everything--to put
the universe into the university.  Let's call this the vertical axis:

the self-contained institution, with the IR that reflects the
institution's goals and constituencies.  Researchers, on the other hand,
tend to align themselves with other researchers in their fields.  The
expert on the use of microalgae for CO2 mitigation happens to reside at
Tulane, but his or her intellectual colleagues may sit at the U. of Hawaii
or in Tokyo.  Research thus is horizontal, straddling multiple
institutions.  This is the world of professional societies and academic
fields (which are reflected in journals publishing).  There is a tension
here:  libraries and IRs are being asked to face in two directions,
vertically and horizontally, straining resources.

Not surprisingly, the actual use of IRs is less than many had hoped for,
and much of the use is for such things as students' papers.  Nothing wrong
with that, but it is not in keeping with the often-declared goal of
"capturing the intellectual output of the university."  What I propose is
that in addition to IRs (which ultimately are simply going to be
extensions of course-management systems, so why not just hand off this
function to Blackboard and be done with it?), libraries organize
disciplinary repositories or DRs.  These would be horizontal, not
vertical, and reflect the actual research activities of the global
intellectual community.  There are 24,000 journals today, which is a
starting point for the number of DRs we will need.  With about 10,000
colleges and universities in the world today, with allowances for
different ways of counting, that comes to about 2.4 DRs per institution,
though naturally one would expect Harvard and the University of Chicago to
do more than Middlesex Community College or an emerging institution in
sub-Saharan Africa.

What I propose is that these DRs be assembled on a consortial basis, with
institutions sharing access to DRs and each institution taking
charge--exclusive charge, so as to avoid redundancy--of a certain number
of topics.  How to assign who does what will not be easy, but it simply
makes no sense for there to be competing DRs for some segment of
nanotechnology or Keats research.  Universities can save buckets of money
by recognizing that in some cases, there is no need to be universal.

How would this work?  Progressively, I would hope. The larger institutions
would take over the curation of more disciplines, but even the smallest
would have to contribute something in order to get access to all the rest.  
The definitive DR on stem-cell research may be curated at Hopkins and the
history of Silicon Valley at San Jose State--not really comparable, to be
sure--but Hopkins and SJS would each have access to the other's DR.  To
each according to his means.  To join the consortium, an institution would
have to propose to the governing board what DRs the prospective member
plans to sponsor and curate.  The stern gaze of the board would prevent
free riders or "cheap riders":  Carry your weight in curation or be an
outcast.

As for independent scholars without institutional affiliation, I propose
that they would gain access by doing the equivalent of purchasing a
library card from a member institution.  For $50 you get everything.

This plan solves a number of problems.  It aligns repositories with the
research community--horizontally, in DRs.  It saves money by negating the
need for institutions to try to cover everything, a pointless and
unnecessary endeavor in the world of the Internet.  For those
uncomfortable with commercial organizations operating within the academic
community, it provides a purely consortial arrangement among similar
not-for-profits.  It is progressive, enabling the participation of Third
World scholars on the same level of access as their lucky counterparts in
Oxford and Palo Alto.  It provides a good ROI for major institutions, and
a fabulous ROI for small ones.  It eliminates the free-rider problem by
mandating some level of curation, however small (but scaled to an
institution's resources), and thus provides an incentive for all
institutions to get involved.  And it captures the output of academic
institutions in such a way as to provide significant incentives for
researchers to participate (which is the problem with IRs:  little
researcher participation).

Open Access purists will note that this plan falls short of full OA.  
That is correct:  this is Almost Open Access, as it requires institutional
affiliation (which you can get for the cost of a library card).  The
virtue of AOA as opposed to OA is that AOA is sufficiently suasive to
ensure economic commitment and participation.  Traditional publishers (for
whom there is absolutely nothing in this plan) will remark that AOA is
what they have advocated all along.  That is also correct.  But publishers
will never grow comfortable with pure OA, as their business training will
not permit them to expend 100% of their effort to satisfy 1% of demand.
 
But they are not needed for this plan, so their comfort is besides the
point.

Joe Esposito