Michael Eisen's point is fundamental enough to be worth considering very
explicitly and with considerable attentiveness. I hope many voices will
make themselves heard on this, because what is at issue goes to the
heart of open access provision itself and what can be done to provide
maximum open access right now. (Please note the "right now" because it
is at the heart of the issue, open access being already long overdue.)

There is nothing hard to understand here, but there are several things
that need to be kept clearly and explicitly in mind, in order to avoid
needless misunderstandings (and the lost opportunities for open access
provision that results from them).

In what follows,

    OA = Open Access

    TA = Toll Access

and OA means:

    FREE, IMMEDIATE, PERMANENT ACCESS TO REFEREED-ARTICLE FULL-TEXTS ONLINE

http://www.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/3223.html
http://www.earlham.edu/~peters/fos/boaifaq.htm#openaccess

In approximate numbers, we are taking about how to provide OA, in the above sense,
to the yearly 2,500,000 articles that appear in the planet's 24,000 peer-reviewed
journals (across all disciplines and languages).

Before I proceed to a point-by-point commentary on Mike's posting, I will
reproduce it in full. But before that I will provide a succinct summary
of my reply. Here it is:

    There is only one, unified OA provision strategy: "Publish your
    article in an OA journal if a suitable one exists, otherwise
    publish your article in a suitable TA journal and also self-archive
    it." There is no competition between these two components; they
    are complementary. The discussion below is only about what is
    the immediate scope for each component today. All are agreed that both
    components are underutilized. The only disagreement is about *how
    much* each component is underutilized. The disagreement would
    be immediately mooted if the advocates of each component always
    explicitly advocated their own component as only one part of the
    unified OA provision strategy: "Publish your article in an OA journal
    if a suitable one exists otherwise publish your article in a suitable
    TA journal and also self-archive it."

First, here is Mike's comment in full:

On Thu, 11 Dec 2003, Michael Eisen wrote:

> I think Sally is absolutely correct that less than 2.5% of published content
> is published in open access journals, but that doesn't count the large
> amount of material that is made freely available by fee-for-access
> publishers through their own websites or through PubMed Central. I, of
> course, don't count this later class as being truly open access, but it is
> as available as self-archived content and should be given its proper due.
>
> I would also like to object, once again, to Stevan's continued use of this
> 5% open access / 95% self-archiving number. It's grossly unfair to contrast
> reality (<5% of articles currently published in open access journals) on
> one side with potential (that 95% - or more accurately something like 50% -
> of articles COULD be self-archived). With BMC's diverse collection of
> journals, PLoS, and the many other open-access publishers in DOAJ (including
> high-end journals like PLoS Biology, J. Biol, JCI, BMJ) virtually any
> biomedical research article could be published in an open-access journal
> today.
>
> Thus, most authors - many, many more than the 5% you imply - who want to
> make their work freely available have a choice - they can publish it in a
> "green" fee-for-access journal and self-archive it, or they can publish in
> an open access "gold" journal. They may have reasons to choose the former
> route, and there is certainly a lot of work that needs to be done to make
> open access journals more appealing, but let's stop implying that the open
> access journal option wasn't available.

I now reply point by point:

> I think Sally is absolutely correct that less than 2.5% of published content
> is published in open access journals, but that doesn't count the large
> amount of material that is made freely available by fee-for-access
> publishers through their own websites or through PubMed Central. I, of
> course, don't count this latter class as being truly open access, but it is
> as available as self-archived content and should be given its proper due.

I completely agree with Mike that all freely-accessible full-text journal
articles should be counted, but I don't think it is giving them their
proper due to decline to count them as "truly" OA! Unless, of course,
they fail to meet the full OA definition:

    FREE, IMMEDIATE, PERMANENT ACCESS TO REFEREED-ARTICLE FULL-TEXTS ONLINE

    "Is there any need for a universal Open Access label?"
    http://www.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/3223.html

They could fail to meet that definition not only by failing to be free, or
failing to be refereed journal articles, or failing to be full-texts online.
They could also fail by not being immediate or by not being permanent. (A
journal that makes its contents free online after a delayed embargo period of 6
months to 2 years or more is certainly no OA journal. Nor is a journal that
temporarily makes its contents free online as a from of advertisement, but then
removes them.)

    Harnad, S. (2001) AAAS's Response: Too Little,
    Too Late. Science dEbates [online] 2 April 2001.
    http://www.sciencemag.org/cgi/eletters/291/5512/2318b

Mike and I are agreed on this. We do disagree, however, on the free/open
distinction (which I consider completely spurious):

    "Free Access vs. Open Access"
    http://www.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/2956.html

But there is also an important logical point which Mike seems to have overlooked:
If a journal provides the following for *all* of its articles:

    FREE, IMMEDIATE, PERMANENT ACCESS TO REFEREED-ARTICLE FULL-TEXTS ONLINE

on its own journal website or PubMed Central's, then that is OA journal
publishing ("gold"), not OA self-archiving ("green")! The "self" in
"self-archiving" is the author (and the author's institution). It
does not help us provide clarity and understanding to conflate the two
components of the unified OA provision strategy by failing to distinguish
OA provision by the journal (OA publishing, gold) from OA provision by the
author/institution (OA-self-archiving, green).

I think it is Mike's spurious free/open distinction that allows him to
fail to make this absolutely fundamental distinction between the two
complementary components of the unified OA strategy.

To repeat: if the free access is delayed or temporary, I agree that that is not
OA. But otherwise it is OA, and should be counted as OA. Moreover, if it is
provided by the journal rather than the author, it should be credited to OA
journal-publishing and not to OA self-archiving.

> I would also like to object, once again, to Stevan's continued use of this
> 5% open access / 95% self-archiving number. It's grossly unfair to contrast
> reality (<5% of articles currently published in open access journals) on
> one side with potential (that 95% - or more accurately something like 50% -
> of articles COULD be self-archived). With BMC's diverse collection of
> journals, PLoS, and the many other open-access publishers in DOAJ (including
> high-end journals like PLoS Biology, J. Biol, JCI, BMJ) virtually any
> biomedical research article could be published in an open-access journal
> today.

I do not contrast reality with potential. I contrast reality with reality,
and potential with potential. The reality is that of the 2,500,000 articles
published yearly in the 24,000 peer-reviewed journals, about 10% of them
are OA today: about 2.5% for having been published in an OA journal, and about
7.5% for having been published in a TA journal and self-archived by its author.

http://www.ecs.soton.ac.uk/~harnad/Temp/self-archiving_files/Slide0024.gif

So the reality/reality contrast is 7.5%/2.5% or about three
times as much OA provision via OA self-archiving, today, as via
OA journal-publishing, today. These are estimates, and could be
inaccurate, though it is far more likely that the OA self-archiving
figure is an underestimate than that the OA journal article figure is
an underestimate because there is no way yet to tally how many articles
have been made OA by self-archiving on the author's arbitrary website
(only the OAI-compliant archives are being systematically harvested, e.g.,
by OAIster http://oaister.umdl.umich.edu/o/oaister/ -- but more general
harvesters such as Citeseer http://citeseer.nj.nec.com/cs in computer
science suggest that there might be a good deal more self-archived OA
content out there) whereas we are closer to having a total tally of
OA journals, and hence their contents, via the Directory of Open Access
Journals in Lund http://www.doaj.org/ -- though, as this is growing daily,
mostly with OA journals that have not yet made their existence known,
the true total there could be higher than the current 600. It is for
that reason that I have doubled the OA journal share estimate to 5%,
to be conservative.

But that was the (estimate of) the reality/reality contrast. Now what
about the potential/potential contrast? Let us first agree that both
components of the unified OA provision strategy are underutilized:
There is definitely room for more articles in the existing c. 600 OA
journals. And there is also definitely room for more OA articles in the
existing OA archives. How much room? That is the potential/potential
question.

But let us agree that we are talking about realistic potential, and
not pipe-dreams. It does not count as "potential" to suppose that all
the remaining 23,400 TA journals have the "potential" to be persuaded to
convert to OA! That is a logical possibility, but not much more than that,
at this moment. I wish it were otherwise, but right now, the established
TA journals are not inclined to convert to OA, and I am not inclined to
wait for them to be persuaded to do so -- nor, I hope, is the rest of the
research community inclined to wait for OA to come via that route alone.
That is why we have the unified OA provision strategy: Because there
is another route too, and it does not depend on the research community
persuading the TA journal publishing community to provide OA for them,
but on persuading the research community to provide OA for itself --
by self-archiving its own TA articles.

And I want to suggest that the potential for the research community to
persuade *itself* to provide OA is far higher than the potential for the
research community (or anyone else) to persuade the publishing community
to provide it for them. I would say that the relative size of those two
potentials is realistically reflected by the figures I suggested: 95%/5%
(at this time: remember we are talking about the potential for *immediate*
OA).

http://www.ecs.soton.ac.uk/~harnad/Temp/self-archiving_files/Slide0026.gif
http://www.ecs.soton.ac.uk/~harnad/Temp/self-archiving_files/Slide0028.gif

But there may be another sense of "potential" Mike and others have in
mind: There is definitely room for more articles in the existing OA
journals. But does anyone really believe that there is as much room in
OA journals today, for today's daily/weekly/monthly article-output as
there is in institutional archives, today? I would say that the difference
is still an order of magnitude. But in any case, if that is the case --
if 23,000 journals-worth of TA articles can already today be redirected to
the 600 OA journals that exist today, so be it! That potential is fully covered
by our unified OA provision strategy:

    "Publish your article in an OA journal if a suitable one exists,
    otherwise publish your article in a suitable TA journal and also
    self-archive it."

We should strongly encourage OA journal advocates to formally and
explicitly promote this unified strategy, exactly as we encourage
university administrators and research-funders to do so: The accent
is on OA provision, today, by whichever way suits!

http://www.ecs.soton.ac.uk/~harnad/Temp/self-archiving_files/Slide0022.gif

> 95% - or more accurately something like 50% -
> of articles COULD be self-archived

This is incorrect. What is correct is that 55% of journals sampled by Romeo
are either "blue/green" or "gold" -- i.e. already formally support author
self-archiving. Many of the remaining 45% "white" journals will agree too, if
asked.

http://www.lboro.ac.uk/departments/ls/disresearch/romeo/Romeo%20Publisher%20Policies.htm

But even that increased percentage does not fully cover the "COULD" --
as the 250,000 papers self-archived by physicists since 1991 attests:
All those papers were self-archived *without asking* -- and in the 12 years
only 4 or so of the 250,000 papers has been removed for copyright reasons!

   "Re: Copyright: Form, Content, and Prepublication Incarnations"
   http://www.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/3188.html

> Thus, most authors - many, many more than the 5% you imply - who want to
> make their work freely available have a choice - they can publish it in a
> "green" fee-for-access journal and self-archive it, or they can publish in
> an open access "gold" journal. They may have reasons to choose the former
> route, and there is certainly a lot of work that needs to be done to make
> open access journals more appealing, but let's stop implying that the open
> access journal option wasn't available.

Dear Mike, I can hardly be described as implying that the open access
journals option is not available, when it is trailed on every one of
my postings, like a mantra, as a component of the unified gold/green
OA-provision strategy! If the odds prove to be better than 5/95 of
finding a suitable OA journal, that's fine with me! My figures merely
represent the 600/23,400 ratio of OA/TA journals today. And the reason
I keep pointing them out is that too many researchers still think the
golden road is the *only* way to provide OA today -- whereas the truth
is much closer to the opposite!

Stevan Harnad

NOTE: A complete archive of the ongoing discussion of providing open
access to the peer-reviewed research literature online is available at
the American Scientist Open Access Forum (98 & 99 & 00 & 01 & 02 & 03):
    http://amsci-forum.amsci.org/archives/september98-forum.html
    http://www.cogsci.soton.ac.uk/~harnad/Hypermail/Amsci/index.html
    Post discussion to: [log in to unmask]

Dual Open-Access-Provision Policy:
    BOAI-2 ("gold"): Publish your article in a suitable open-access
            journal whenever one exists.
            http://www.earlham.edu/~peters/fos/boaifaq.htm#journals
    BOAI-1 ("green"): Otherwise, publish your article in a suitable
            toll-access journal and also self-archive it.
            http://www.eprints.org/self-faq/
    http://www.soros.org/openaccess/read.shtml
    http://www.ecs.soton.ac.uk/~harnad/Temp/berlin.htm

> ----- Original Message -----
> From: "Stevan Harnad" <[log in to unmask]>
> To: <[log in to unmask]>
> Sent: Thursday, December 11, 2003 1:29 PM
> Subject: Re: Journals > Peer-Reviewed Journals > Open-Access Journals < Open
> Access
>
>
> > On Thu, 11 Dec 2003, Sally Morris wrote:
> >
> > > I would question Stevan's estimate that 2.5% of articles are published
> in OA
> > > journals.  While it does indeed look as if 2 - 2.5% of peer reviewed
> > > journals are OA (that is, if all those listed by Lund et al are peer
> > > reviewed), I very much doubt that they carry as many articles as the
> rest.
> > > This is because OA journals are, almost without exception, relatively
> new
> > > and extremely long-established journals tend to be far, far, bigger in
> terms
> > > of issues and articles published per year.
> >
> > I don't disagree with Sally's suggestion that 2.5% of journals does
> > not necessarily mean 2.5% of articles published in journals. I was
> > very deliberately using a very conservative, high-end estimate (sometimes
> > I even use 5%) merely to illustrate how minuscule is the amount of OA that
> > can currently be provided via the OA journal route ("gold") and hence
> > how important it is to supplement it via the OA self-archiving route
> > ("green"), today.
> > http://www.ecs.soton.ac.uk/~harnad/Temp/self-archiving_files/Slide0024.gif
> >
> > Stevan Harnad
> >
> > NOTE: A complete archive of the ongoing discussion of providing open
> > access to the peer-reviewed research literature online is available at
> > the American Scientist Open Access Forum (98 & 99 & 00 & 01 & 02 & 03):
> >     http://amsci-forum.amsci.org/archives/september98-forum.html
> >     http://www.cogsci.soton.ac.uk/~harnad/Hypermail/Amsci/index.html
> >     Post discussion to: [log in to unmask]
> >
> > Dual Open-Access-Provision Policy:
> >     BOAI-2 ("gold"): Publish your article in a suitable open-access
> >             journal whenever one exists.
> >             http://www.earlham.edu/~peters/fos/boaifaq.htm#journals
> >     BOAI-1 ("green"): Otherwise, publish your article in a suitable
> >             toll-access journal and also self-archive it.
> >             http://www.eprints.org/self-faq/
> >     http://www.soros.org/openaccess/read.shtml
> >     http://www.ecs.soton.ac.uk/~harnad/Temp/berlin.htm
> >
>