New SPARC/ARL/ACRL Brochure on Open Access

From: Stevan Harnad <>
Date: Mon, 28 Jun 2004 21:05:04 +0100

    Prior Topic Thread:
    "Re: A Role for SPARC in Freeing the Refereed Literature"

SPARC, ARL, and ACRL have published a brochure on open access.

    From press release:
        "the new brochure presents the benefits of open access to authors,
        readers, teachers, scholars, and scientists."

This is a useful document, but contains a few errors and one major
omission (at the end):

> Access scientific and scholarly research online free of charge
> [and] free of licensing restrictions

The research in question is peer-reviewed journal articles. And the
access is to the full-text, downloadable for free from the Web. The
licensing issue is not an essential or central one as a precondition
for providing OA to an author's own articles via self-archiving. It
would be incorrect to suggest that copyright or licensing agreements
need to be renegotiated by authors with their journals in order to make
their articles Open Access (OA).

The most important counterexamples are the more than 80% of journals
that have already given their green light to author self-archiving:

This means that for that 80% of journals at least, there are no
copyright/licensing restrictions on self-archiving and there is no need
for negotiations with the publisher in order to make any of the articles
in those journals OA by self-archiving them.

Even for the remaining 20% of journals, many will give their green light
to self-archiving if asked on an individual article basis.

It is always a good idea for an author to retain copyright, and to
selectively license rights whenever possible. But it is not a precondition
for providing OA via self-archiving, and to suggest that it is would be
to put needless handicaps on the growth of OA via self-archiving,
sending authors into unnecessary conflicts with their publishers,
with uncertain outcomes.

> You know first-hand that scientists and scholars are not
> paid for their journal articles. In most cases, you must transfer
> copyright to a journal before it will publish your work. While you
> might receive royalties for your books and software, your compensation
> for journal articles is more abstract: your field advances and your
> career develops.

This is all correct. In a nut-shell, peer-reviewed journal article
authors, rather than selling their articles for royalty revenues from
their sale, give away their articles, in order to maximize their usage
and impact, which generates revenue indirectly for the author and his

> If you're giving up your royalties and intellectual property rights,
> shouldn't readers be the beneficiaries? By removing price and permission
> barriers, open access makes your work easier to use. Open access serves
> your interests as the author and the interests of all potential readers.

The primary beneficiaries of making an author's article OA are the
author, his institution, his research funder, and research itself; the
secondary beneficiary is the research user at other institutions. The
author benefits because increased usage means increased research
impact. Maximizing access maximizes usage, which maximizes impact.

> In the age of print, open access was physically and economically
> impossible. But thanks to the Internet, it's an emerging reality. Now,
> the tradition of producing journal articles without expectation of
> payment combined with electronic publishing offers an unprecedented
> public good: the free online availability of peer-reviewed scientific
> and scholarly journal articles.

The online era has made it possible to maximize user access and thereby
maximize research impact in ways that were physically and economically
impossible in the paper era.

> Think about what this kind of distribution will mean for the
> enlargement of your audience, the widespread sharing of knowledge,
> and the acceleration of research. Open-access archives and journals
> are both practical and lawful. Implementations around the world are
> proving that they surpass traditional subscription-based journals in
> their cost-effectiveness and service to science and scholarship.

This is a somewhat muddled comparison: There are two ways to provide OA
to an article: (1) to publish it in an OA journal or (2) to publish it
in a TA (i.e., subscription-toll-based) journal *and* to self-archive
it in an OA archive, as a supplement to the TA version, so all would-be
users can access it, whether or not their institutions can afford to
pay for access to the TA journal version.

Hence for OA that is provided via (2) -- OA = TA + SA -- it would make no
sense to say that OA impact is greater than TA impact! Of course OA impact
is greater than TA impact, since OA impact = TA impact plus SA impact.

There is also no evidence yet for saying that impact from OA journals
(1) is greater than impact from TA journals. The only published study
so far comparing OA and TA journal impact finds no difference:

    Pringle, J. (2004) Do open access journals have impact? Nature (Web

It is also hard to do controlled comparisons between OA and TA journals,
to ensure that the only relevant difference is that one journal is OA
and the other TA. One can control for irrelevant journal differences by
comparing OA vs. TA *articles* within the same journal (and year), the
OA articles being self-archived TA articles. It then stands to reason
that the only way for impact to go is up, with
OA impact = TA impact + SA impact.

And impact does go up, substantially:

    Harnad, S. & Brody, T. (2004) Comparing the Impact of Open Access
    (OA) vs. Non-OA Articles in the Same Journals. D-Lib Magazine.

It is not clear what "cost-effectiveness" means here (and it would
probably be best not to mix the journal affordability problem with the
access/impact problem). User institutions pay for access to the
(peer-reviewed) output of other institutions. Author institutions
only pay for their own output if/when it is published in OA journals.
Cost-effectiveness is too vague to be reckoned here, except to note that
OA via TA + SA costs the author institution next to nothing, and
maximizes its research impact.

> In 1997 1998, 85 percent of the most highly cited articles were open
> access. Articles with lower citation impact were more likely to be
> restricted access.

This datum comes only from computer science. It has since been replicated
in physics, and is now being tested in other disciplines too. It looks as
if the impact-maximizing effect will be true across all fields, though
both the proportion of OA articles and the size of the OA/TA advantage
may vary.

> Steve Lawrence, a scientist at NEC Research Institute, analyzed nearly
> 120,000 computer science articles cited in a standard disciplinary
> bibliography. When he looked at articles with successively higher levels
> of impact or citations, he found successively higher percentages of
> open-access articles, and vice versa. He found the strength of this
> correlation steadily increased over a decade. source: Steve Lawrence,
> Online or Invisible? Nature , Vol. 411, No. 6837, p. 521, 2001.
> http: / / www. neci. nec. com/ ~ lawrence/ papers/ online-nature01/

Here are further confirmations of the effect, in other fields:

    Brody, T., Stamerjohanns, H., Harnad, S. Gingras, Y. Vallieres,
    F. & Oppenheim, C. (2004) The effect of Open Access on Citation
    Impact. Presented at: National Policies on Open Access (OA)
    Provision for University Research Output: an International
    meeting. Southampton University, Southampton UK. 19 February 2004.

    Kurtz, M.J. (2004) Restrictive access policies cut readership
    of electronic research journal articles by a factor of two,
    Harvard-Smithsonian Centre for Astrophysics, Cambridge, MA

    Odlyzko, A.M. (2002) The rapid evolution of
    scholarly communication." Learned Publishing 15: 7-19

> Open access increases research impact.

> As an author You will have a larger potential audience
> than any subscriber-restricted journal can give you, even the most
> prestigious. Open access can increase the impact of your work, shorten
> the delay between acceptance and publication, and make your articles
> more effective by making them easier to find and use. Your work will
> be visible to every search and retrieval tool.

This is all correct, but it does make it sound as if the way to get
this OA is by not publishing in a "subscriber-restricted" (TA) journal
-- whereas the fact is that the most OA comes from publishing in a TA
journal and going on to self-archive the article.

> As a reader You will have free online access to the literature necessary
> for your research.

> As a teacher You will know that your students have convenient access
> to the information they need. With open- access literature, the author
> or copyright holder has given permission in advance for making and
> distributing copies. No more delays, doubts, or fees. No more fair-use
> judgment calls, fear of liability, and painful decisions to err on
> the side of caution and non-use.

All of this copyright/permission information is irrelevant for
self-archived OA articles. They are on the web, like everything else on
the web. The library is not even involved in their use. The user accesses
the article just as they access anything else that is available toll-free
on the Web. Copyrights, permissions, delays, doubt, fair-use judgment
calls, liability fears and painful decisions are not involved.

> As a scientist or scholar You will be coming to grips with a journal
> publishing system that is no longer sustainable. Despite the opportunity
> for expanded global sharing of knowledge brought by the Internet,
> prices of many journals have spiralled out of control and libraries have
> had no choice but to cancel subscriptions, defer new subscriptions,
> and cut into their book budgets. There are many potential solutions to
> this crisis, but open access is the most effective. And it is within
> the reach of scientists and scholars.

100% OA would be a certain solution to the access/impact problem and
*might* also bring eventual relief to the affordability problem too, but
authors will never be induced to provide OA (either by publishing in OA
journals or by self-archiving the articles they publish in TA journals)
for the sake of solving the affordability problems of libraries or
sustainability problems of publishing.

It would be much more helpful if the access and affordability problems
were disentangled, because OA will come from the motivation of authors
(and their universities and funders) to solve the access problem, for
the sake of impact, not from the motivation of libraries to solve the
affordability problem for the sake of access!

> Why should you care about open access? How can you provide open access
> to your work? Submit your work to open-access journals. There are over
> 1000 peer-reviewed journals listed in the Directory of Open Access
> Journals ( www. doaj. org) , and new journals appear online every month.

> Deposit your preprints in an open digital archive hosted by your
> institution or discipline. For a list, see www. arl. org/ sparc/ repos.
> If your publisher permits it, deposit your postprint ( the revised
> version, as published in the journal) in an open archive.

This is excellent advice, but would be strengthened by pointing out two
highly pertinent pieces of data: 5% of journals are OA journals, whereas
over 80% of journals have given their green light to self-archiving. That
would give authors a far clearer idea of what the real priorities and
probabilities are.

> If you submit your work to a subscription-based journal, retain your
> rights and the rights of your readers by attaching the SPARC author s
> addendum ( to the publisher s copyright
> form. If the journal will not consent to this, be persistent. The
> discussion will help the publisher understand what matters to authors.
> Make sure you can put the postprint on your personal website or
> better yet, in an open digital archive hosted by your institution
> or discipline.

It would be helpful to point out that 53% of journals have already given
their green light to postprint self-archiving, and that even for
the remaining 47%, copyright retention is not a prerequisite for
self-archiving, hence it is not a necessary precondition (though it is
always desirable).

What is needed is not persistence in trying to retain copyright but
persistence in self-archiving!

> If you submit your work to a subscription-based journal, offer to pay
> the costs of providing open online access. A growing number of journals
> accept this as a way to experiment with the methods and economics of
> open-access publishing. Other journals are waiting to be asked.

Why recommend this blanket offer to pay journals to self-archive for
you, when over 80% have already given you the green light to
self-archive for yourself? Wouldn't it be more helpful to mention
what percentage of journals offer paid self-archiving (and are not also

> How can you help the cause of open access? If your
> institution or discipline doesn t already host an open digital archive,
> help launch one. Open- source software exists to help you create and
> maintain them.

> Help launch an open-access journal in your field. Open-source journal
> management software can automate clerical tasks and keep costs down.

> Serve on the editorial board or referee papers for an open-access
> journal.

Why favor OA journals over green journals? Green already generates at
least 3 times as much OA, and could produce 100% immediate OA virtually
overnight. Creating and converting OA journals is a far slower and less
certain option:

> When sitting on grant-review panels or hiring, tenure, or promotion
> committees, give due weight to peer-reviewed publications regardless of
> their price or medium. And don't rely only on prestige or impact factor
> this discriminates against new journals that may be of high quality.

What has this to do with OA? Over 80% of journals are green, only 5%
are OA journals ("gold"). Who suggested that promotion or funding panels
ever weigh journal price one way or the other? And virtually all journals
are hybrid now (with both a paper and electronic version), so what has
that to with promotion and funding panels?

And why is an appeal to OA as a means of maximizing impact being coupled
with an appeal to ignore impact?

> Help your professional associations understand open access. Serve on
> their committees and governing boards, and write opinion pieces for
> their newsletters. Nudge them into adopting open access for their own
> journals and endorsing open access for other journals in the field.

But make sure you understand what you mean, and what you are nudging
journals into: Why not also nudge them to become green, rather than just
gold (OA Journals)?

> If you are a journal editor, encourage your publisher to adopt an
> open-access business model. If the publisher is unwilling and pursues
> policies that restrict access, consider following the example of
> journals in disciplines such as biology and mathematics by declaring
> independence. Along with the rest of your editorial board, resign from
> the journal and launch a new, open-access journal to serve the same
> niche. Organizations such as SPARC can aid you in this transition.

Why "declare independence" rather than go green? (As with the author
copyright recommendations, these suggestions are setting the goal post
needlessly -- and counterproductively -- high.)

> Help your library make intelligent decisions about subscriptions
> and cancellations by having a discussion about the real value of
> scholarly journals. Librarians often feel pressured to take actions that
> perpetuate the pricing crisis by subscribing to journals whose price
> may not be a true reflection of their size, quality, impact, or usage.

The amount of increased access that comes from publisher pricing and
library selection decisions is minor compared to the amount that would
come from OA: The affordability problem needs to be disentangled from
the access/impact problem.

> Educate colleagues and the next generation of scientists and scholars.
> You can prevent damaging myths or alarmist claims about open access
> from circulating without challenge. Open access is compatible with
> peer review, copyright, and career advancement.

All true, But this brochure itself is still perpetuating myths too!

> Open Access Astronomy researchers who made their Astrophysical Journal
> article open access using the arXiv. org e-print server doubled the
> citation rate of their articles. source: Stevens-Rayburn, Sarah. Account
> of 2003 AAS Publication Board meeting email to PAM electronic discussion
> list, November 13, 2003.

See the much stronger data above, based not on one journal, but on many
articles, in many journals, in many fields.

> Scientists who chose the open access option when they published in
> Limnology and Oceanography had approximately three times more downloads
> of their articles from that journal s website. source: American Society
> of Limnology and Oceanography.

Again the vastly disproportionate on the 5% ("golden") road to OA:
OA journal creation/conversion.

Yet among the suggestions made here, the most important one of all
is missing! The suggestion to authors' institutions and funders that,
in order to maximize the impact of their research output, they should
adopt a policy to encourage (or better, mandate) OA provision for all
institutional research output -- by whichever means is suitable for each
particular article: publishing it in a suitable OA journal or publishing
it in a suitable TA journal and self-archiving it!

Stevan Harnad

NOTE: A complete archive of the ongoing discussion of providing open
access to the peer-reviewed research literature online (1998-2004)
is available at the American Scientist Open Access Forum:
        To join the Forum:
        Post discussion to:
        Hypermail Archive:

Unified Dual Open-Access-Provision Policy:
    BOAI-2 ("gold"): Publish your article in a suitable open-access
            journal whenever one exists.
    BOAI-1 ("green"): Otherwise, publish your article in a suitable
            toll-access journal and also self-archive it.
Received on Mon Jun 28 2004 - 21:05:04 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:29 GMT