Re: EPrints, DSpace or ESpace?

[Stevan Harnad <harnad@ecs.soton.ac.uk>]

On Sun, 9 Nov 2003, Neil Beagrie wrote:

> the final version of the report by Maggie Jones from the e-journal
> archiving study is now available in pdf on the jisc website at
> 
> http://www.jisc.ac.uk/uploaded_documents/ejournalsfinal.pdf

This report has useful information for those who are interested in
licensed toll-access journal content and in the preservation (archiving)
of such licensed content, but it unfortunately misunderstands the relation
between the archiving concerns for such licensed (i.e., bought-in)
toll-access content and the concerns and purposes of the self-archiving of
researchers' own institutional research output (mostly also published in
toll-access journals).

The report is right that this is a *parallel* form of archiving, but it is
in error about what is actually paralleling what! The relevant passages
are:

>   "E-Print Repositories: The rapid escalation of e-print repositories has
>   been regarded by some of its champions as a potential replacement
>   for more traditional scholarly communication provided by licensed
>   e-journals."

All kinds of things have no doubt been said by all kinds of champions and
challengers, but the standpoint of the Budapest Open Access Initiative
(BOAI) on self-archiving is quite clear: Self-archiving in eprint archives
is an alternative way of providing access -- *open access* -- to the *very
same articles*, i.e., the articles that authors have published in
toll-access journals. Hence eprint archives are not *replacements*
(substitutes) for the journals but *supplements* to them, intended to
allow authors to provide access to their articles for all those would-be
users whose institutions cannot afford the tolls for the toll-access
version.

http://www.soros.org/openaccess/read.shtml
http://www.eprints.org/self-faq/

>   "The emphasis to date has been on encouraging scholars to
>   deposit content into the archives, rather than on preservation
>   requirements. Indeed some proponents of e-print repositories argue
>   against considering preservation requirements at this stage as being
>   largely irrelevant for the time being and something which can be
>   considered later (if at all)."

Once it is at last understood that the self-archived versions in the
eprint archives are supplements to rather than substitutes for the
toll-access versions, it will perhaps also be at last understood that the
primary preservation burden is not on those supplements, self-provided
only to maximise access and impact. The primary preservation burden is on
the proprietary toll-access versions. Those are the ones that publishers
sell and those are the ones that libraries buy; and as long as that buying
and selling goes on, those are also the parties responsible for the
permanence of the proprietary product, *not* the authors (and their
institutions), who are only providing a supplementary version in order to
provide access for those would-be users who cannot afford the proprietary
version.

Yes, the self-archivers are interested in providing open-access to their
work not only today, but tomorrow, and after-tomorrow. And they are doing
so. The work self-archived in the Physics Arxiv in 1991, for example, is
still alive and well, fully useable and used, in 2003, thank you very
much, and was even successfully retro-fitted for OAI-compliance in 1999.
And all this self-archived work will continue to be kept openly accessible
by researchers and their institutions. And some day, possibly, if and when
the access-tolls are no longer being paid at all, and all archiving is
offloaded on the network of OAI-compliant eprint archives, *then* the
eprint archives can take over the primary burden of archiving too.

But for now, they are only a parallel form of *access-provision* to the
very same literature, and they are not the ones that have, or should worry
about, the primary preservation burden. Nor are they alternatives to the
journals; they are just alternative forms of access. If/when the "golden"
option prevails, and all journals convert to open-access, covering costs
from author/institution submission fees, per paper, instead of
reader/institution toll-access fees, per journal, then these archives will
be poised to assume the preservation burden. But at no time will this mean
that eprint archives "replace" journals: It merely means that open-access
journals will become essentially peer-review service providers and
certifiers rather than the providers of a paper or online text.

Short form: Access-provision will be come unbundled from publication (for
the refereed research literature). Publication will mean have been
accepted as meeting the established peer-review quality standards of a
journals. Authors provide the research and the text. The journal provides
the peer-review (and editing) service. And the network of
OAI-interoperable institutional eprint archives provide the access.

>   "As these repositories will be expected to contain valuable scholarly
>   resources, it is to be hoped that their preservation will be taken
>   into consideration and a related JISC funded study is currently
>   investigating this [24]."

It is no doubt the needless proliferation of synonyms, near-synonyms, and
euphemisms that keeps us so confused about the purposes of eprint
archives. Can I propose that we reserve the term "institutional
repository" for all the *other* things an institution may wish to archive,
manage and preserve: bought-in toll-access digital content, institutional
input *other* than peer-reviewed articles (pre- and post-peer-review),
such as courseware, etc., and any e-publishing ventures institutions may
be contemplating to increase their revenue streams.

That way we can reserve "eprint archives" for the specific content
targeted by the BOAI, namely, institutional refereed-research output,
self-archived in order to maximize its impact by maximizing access to it.

http://www.ecs.soton.ac.uk/~harnad/Temp/self-archiving_files/Slide0045.gif

>   "In the meantime, the immediate concern of libraries is whether they
>   can rely on continued access to material they have paid for through
>   licences. While there will be overlapping issues to be addressed in
>   preserving licensed e-journals and e-prints, at this stage they need
>   to be considered separately. The development of e-print repositories
>   should be watched with interest and offer the potential for more
>   rapid access to scholarly research. However, this study recommends
>   they should not be regarded as a complete substitute for licensed
>   e-journals but as a parallel development."

As long as the *nature* of this parallelism is clearly understood, the
above seems like a very reasonable recommendation!

Stevan Harnad