RE: going online only... (and JSTOR)

[Kevin Guthrie <KG@jstor.org>]

Dear all,

David Goodman recently posted a message to lib-license expressing concerns
about two paragraphs in JSTOR's Library License Agreement.  I would
normally respond directly and only to David, but I have decided to respond
to the full list because the paragraphs in question pertain to archiving,
an issue I consider to be extremely important.  JSTOR's mission is to
serve as a trusted archive of journal back issues for the community, and
as such has been focused on the issue for its entire existence.  So please
forgive me for the length of this message.  If any of you have specific
questions or concerns about your own JSTOR license agreements, please feel
free to contact me directly (kg@jstor.org) and I will make sure that you
get prompt answers.

When evaluating JSTOR's provisions pertaining to archiving, it is
essential that one understand JSTOR's mission and how if fits into the
system-wide context of archiving for the community.  David compares
JSTOR's license to that of Elsevier.  JSTOR and Elsevier have completely
different governance structures, missions and objectives, and it does not
follow that the provisions for archiving that might be appropriate for one
are necessarily the same that should be in place for the other.

Why is the mission and governance structure of an organization so
important? Because as a community we need to develop a robust and
distributed set of archives in order to protect today's electronic
scholarship for tomorrow's scholars, researchers and students.  We need to
have trusted entities take responsibility for specific domains if we are
going to be able to protect digital resouces at the scale that is
necessary.  JSTOR's provisions should be assessed in the context of
JSTOR's broader archival mission and how that mission fits within the
overall system of scholarly communication.

David makes a fair point with his suggestion that distributing CD-ROMs to
participating institutions in the event JSTOR ceases to exist is a
provision that is out of date. Circumstances have made the original JSTOR
provision less practically meaningful than when we first offered it to
libraries in 1997.  At that time, the JSTOR archive was a relatively small
resource and there were a small number of libraries participating.  
Distributing the resource to participating institutions in the event that
JSTOR ceased to exist was actually possible.  But with 10 million pages
now in the archive and over 1,400 participating institutions worldwide, it
is not practical to imagine distributing the archive in that fashion.
Further, even if we were to choose to distribute the archive, most, if not
all, of the 1,450 institutions would not want to invest the resources
locally that would be required to mount and maintain it.

David asserts in his message that the best current practice is for
suppliers to provide material to an alternate source, such as a national
library, so the material can be made available in the event that the
supplier ceases to exist.  The point I want to make here is that JSTOR
itself is such an alternate source; it receives content from publishers to
be archived. We would be willing to consider exporting the JSTOR data to a
national library for archival purposes, but given JSTOR's archival
mission, its focus, and the assets it already has dedicated to the
archiving task, would that be the best use of community resources?  
National libraries are going to play an extremely important role in the
effort to archive electronic materials, but they are not going to be able
to do it alone.  They have an enormous portfolio of problems to tackle,
many of which extend far beyond the realm of scholarly journal literature.  
We believe there is a need for not-for-profit, non-governmental
organizations focused on archiving to take some of this burden off
national libraries.

Perhaps it will be reassuring to the community to know that the actual
database itself is stored, maintained and operated at three separate
universities in two different countries (the University of Michigan,
Princeton University, and the University of Manchester [UK]).  Were JSTOR
to cease to exist, those universities would have the data in-hand and
could consider entering into relationships with JSTOR participating
publishers to provide archival services.  But I do not think we should
adopt an approach to this provision determined by what might be regarded
as standard practice for commercial enterprises.  And I think a key
question is whether JSTOR is going to be trusted as an archive in its own
right.

We understand that it is not enough to declare oneself an archive.  JSTOR
must earn the trust of the community.  And it is essential that we have
both the resources and the plans in place to give the community confidence
that JSTOR will be able to fulfill its promises.  JSTOR offers access to
the archive on terms that are acceptable to publishers, in a way that is
attractive to libraries and users, and that offers a business model that
enables JSTOR to sustain its primary function, which is preservation.
JSTOR's fee structure includes both an archive capital fee and an annual
access fee.  The archive capital fee is funding an endowment that provides
a base of resources for the future data and software migrations that will
be necessary to keep the archival database current with evolving
technologies. JSTOR's annual access fees cover the ongoing expenses
associated with our regular operating activities.

The second point David makes concerns the question of what happens if an
institution stops "subscribing".  David expresses his view that the best
practice in such a situation is for the supplier to offer the library the
choice of obtaining CDs, obtaining tape, or obtaining continued access to
the purchased portion at an appropriate price.

Here again it is important to distinguish JSTOR from a publishing house.
Unlike a publisher, JSTOR does not generate income from subsciptions to
current issues of journal literature, with a trailing archive of past
issues.  JSTOR *is* past issues.  And it is all of the past issues.  So on
the day a library chooses to participate in JSTOR, they gain access to
every single bit of content that JSTOR has digitized and is authorized to
provide, from Volume 1, Issue 1, right up to the moving wall.  So to use
David's words, "to provide continuing access to the purchased portion" is
to provide continuing access to JSTOR.  And "to provide continuing access
to JSTOR at an appropriate fee" is to continue to participate in JSTOR.  
This is not a license provision added on to address archiving.  This is
JSTOR.

JSTOR has proven to be a very useful database, in fact, it is more heavily
used than any of us might have imagined it would be.  But that does not
make JSTOR a commercial enterprise, and it doesn't alter our mission or
the reasons we were founded. We don't regard JSTOR participation as a
subscription but rather as a contribution to support a centralized
archive. The question really is whether libraries can regard JSTOR as
"their" archive, in effect outsourcing this important function for a
defined group of materials.  The cost of JSTOR participation is far, far
less than the ongoing cost of maintaining the paper materials on the
shelf.  If a college or university decides to withdraw from JSTOR, that
decision is more like deaccessioning the volumes than it is like
cancelling a subscription. Fortunately, it is a reversible decision.  
Libraries need only begin paying JSTOR's annual access fee to restore
their collections.  They don't need to reacquire thousands of volumes of
paper journals.  They don't even have to pay JSTOR's archive capital fee;
they have already paid that one-time only fee.

JSTOR is a totally new kind of organization, a totally new kind of
resource. It simply could not have existed prior to the broad availability
and access to computer and networking technologies that make up the World
Wide Web. These technologies make it possible to store materials centrally
that once needed to be held and protected in a distributed fashion.  
Working together to support a centralized archive promotes economies of
scale that not only reduce system-wide costs but also allow the costs to
be spread over many institutions.

I am grateful for this opportunity to explain some of the fundamental
principles that are the foundation of JSTOR and its mission and appreciate
your patience in reading through this rather long posting.

Sincerely, 

Kevin

*********************
Kevin M. Guthrie
President
JSTOR
120 Fifth Avenue
New York, NY  10011
212.229.3700

-----Original Message-----
From: David Goodman [mailto:dgoodman@phoenix.Princeton.EDU]
Sent: Sunday, September 22, 2002 2:38 PM
To: liblicense-l@lists.yale.edu
Subject: Re: going online only... (and JSTOR)

Back years of Annual Reviews are available through JSTOR.
JSTOR's current contract terms provide:

"6.3 In the event JSTOR ceases to exist and the archive services
contemplated by this Agreement cease to be offered, JSTOR shall provide
Licensee with one (1) complete set of CD-ROMs (or their equivalent at
JSTOR's option) containing the digitized images of the journals contained
in the Archive as of such date. JSTOR hereby grants to Licensee a
nonexclusive, royalty-free, perpetual license to use such images ...

6.4 Upon termination of this Agreement all online access to the Archive by
Licensee and Authorized Users shall be terminated. Print copies of
Materials may be retained by Licensee or Authorized Users and used..."

In my opinion, neither of these provisions meets the best current
practice.

The best current practice for situations like 6.3, where the supplier
ceases to exist, is for the supplier to provide material to an alternative
source such as a national library, where the material will be archived and
supplied if needed.

The best current practice for situations like 6.4, where the institution
stops subscribing, is for the supplier to offer the library the choice of
obtaining CDs, obtaining tape, of obtaining continued access to the
purchased portion at an appropriate fee.

Elsevier and some other commercial suppliers now offer the equivalent of
both provisions (not all of them do--read carefully). It is time for
JStor, which was once a leader, both technically and otherwise, to update
its contracts. It has begun at last to be open url compliant, and the
modernization of its contracts ought to follow.

For contracts with suitable permanency provisions, I do not regard it
essential for even research libraries to maintain print outside their
primary collecting areas. I think it would be desirable to keep existing
print at least in storage, because it may prove preferable for some users,
and we ought to serve all our users if we can, not just the vast majority
who prefer an electronic format.

For contracts without such provisions, no research library should discard
print, or cease a print subscription if the title is considered important.

In the case of libraries that do not consider themselves to be research
libraries, a policy of providing only current direct access to some
materials, and relying on document delivery or ILL for the rest, might
well be justifiable. In any case, all libraries, research or otherwise,
rely upon document delivery and ILL for a great many titles, including
both current and older ones.