Re: Cologne Summit on Open Access Publishing

From: Stevan Harnad <>
Date: Fri, 10 Dec 2004 13:38:25 +0000

Claudia Koltzenburg has provided an excellent summary of the Cologne
OA Publishing Summit at:

I will only clarify a few points that were not quite captured as
Claudia heroically keyed in as much of the discussion as she could!


When Mark McCabe presented his hypothetical economic models for OA and
non-OA publishing -- -- he included
a component for the market effect on acceptance rate and quality for OA
journal articles (predicting that author-pays may drive the number of
articles accepted up and hence the journal quality standards down).

I asked Mark whether he thought that peer-reviewers' acceptance criteria
were bound by his economic model. He replied that if not, then those
of the editor were. I suggested that he may have misunderstood the
peer-review process...

This question has been much-discussed in the AmSci Forum over the years,
for example:

    "The quality of (and hence the author-demand for) peer-reviewed
    journals is based on the rigour and selectivity of the peer-review
    they provide. i.e., on the journal's quality standards. You can't buy
    your way into a peer-reviewed journal. Acceptance depends on the peers
    (the qualified experts, who review for the journals for free!). If a
    journal relaxes its quality standards and lowers its rejection-rate
    to get buy-in dollars, it simply loses its quality track-record and
    citation impact (and peer-reviewers), and then authors no longer
    want to be published in it! (Peer-review has its own "supply/demand"
    criteria, having nothing to do with money.)"

In short, conventional supply/demand modeling of peer-reviewed publication
involves some incorrect assumptions, not taking into account either the
give-away nature of researchers' writings, the give-away nature of peer-reviewers'
services, and the fact that impact itself translates into research
and career income that is incomensurable with the trade transactions,
whether they are for a product to the client user-institution (the
peer-reviewed text) or a service (peer-review and certification) to the
client author-institution. Mark McCabe's models completely miss this distinction,
yet it is absolutely fundamental to this highly anomalous form of publication.
McCabe's authors might as well be journalists selling their words, rather
than researchers, advertising their quality-certified work free for all would-be


In her summary of my own presentation
Claudia writes:

    "...It is a pity researchers do as yet not care enough for OA because
    with 92% of the journals being green, and 20% both green and gold,
    a big step has been made."

What I said was that researchers do not yet know about or understand the
causal connection between access and impact, and that even when they do,
not enough of them (only 20%) provide open access (5% by publishing in
OA, journals, 15% by self-archiving) unless they are "coerced" (by their
employers and research funders) to do so. This is not because they do not
care, but because researchers are a peculiar breed! They even had to be
"coerced" to publish at all (publish or perish") -- by making
their salaries and research funding contingent on publishing their
findings. Otherwise many would just be putting their findings into a
desk-drawer and moving on to do the next piece of research!

    "Harnad concludes that maybe this is a precondition since scientists
    must be 'bunkers' because otherwise they would be making money
    instead of doing research. Harnad sees the Berlin Declaration as
    nothing much more than a pious hope and points out that it is the
    researchers themselves who should act."

The precondition for 100% OA is 100% OA provision by authors (either by publishing
in OA journals or by self-archiving). Although collecting and publicising the
evidence that OA substantially enhances research impact will help OA
grow, what is needed is that research employers and funders should make
OA self-archiving an explicit condition on career advancement and research funding,
just as publishing itself already is. Signing abstract Declarations of
Principle in favour of OA is fine, but we also need to sign and implement concrete
Declarations of Commitment to putting the OA Principle into Practice!

    "In an author survey, Swan & Brown (2004a, 2004b) report that the
    vast majority of their author sample indicated that [they do *not*
    self-archive now, but] they would self-archive *willingly* -- if
    their employer (or funding body) required them to do so!"

    Swan, A. & Brown, S.N. (2004a) JISC/OSI Journal Authors Survey Report.

    Swan, A. & Brown, S.N. (2004b) Authors and open access
    publishing. Learned Publishing 2004:17(3) 219-224.

Harnad, S., Brody, T., Vallieres, F., Carr, L., Hitchcock, S., Gingras,
Y, Oppenheim, C., Stamerjohanns, H., & Hilf, E. (2004) The Access/Impact
Problem and the Green and Gold Roads to Open Access. Serials Review 30.

(3) Claudia quotes:

    McCabe (from the floor): if everyone self-archives and libraries cancel their
    subscriptions, how will financing journals work?

    Harnad: IF everything is OA ...

What I actually replied was that "IF everything is OA" is already
a tall order; and if/when it is met, the entire landscape changes,
in many ways. So let us first actually get there (100% OA), rather
than worrying instead about doomsday speculations regarding its
hypothetical sequelae. We've already wasted a decade doing that.

Meanwhile, speculations can be answered by counterspeculations:

IF 100% OA generates cancellation pressure there will first be
cost-cutting and downsizing, and IF/WHEN cancellations ever become
so great that the remaining essential costs cannot be covered from
toll-access-revenues, THEN there will be eo ipso the basis for a natural
transition to the OA Publishing cost-recovery model, publishers' essential
costs now being covered by the author-institution, for its own outgoing
publications, out of its own cancellation *annual windfall savings*
(ex hypothesi, under the cancellation doomsday scenario!), formerly paid
by the user-institution to buy in the outgoing publications of other
institutions -- but now cancelled, hence saved.

But it is far better to act on the empirical evidence -- which is that
(i) there is no cancellation pressure from self-archiving to date (some
of it going on for over 10 years now, and having reached 100% in some
subfields years ago), (ii) 92% of journals already have a green policy
on author-self-archiving, and (iii) the actual demonstrated benefits of
self-archiving for for research, researchers, their employers and funders
vastly outweigh any hypothetical risks to their publishers -- rather than
to just keep modeling and speculating: Hypotheses non Fingo.

    McCabe: Would it be correct to say that authors who are self-archiving
    are the ones who are better?

    Harnad: The usage impact is reflected in the download figure but the
    relation of reads and citation is as yet undeterminable. However,
    scientists should be in a position that they can pick the best
    stuff to quote, not just what your institution happens to have a
    subscription for.

I'm afraid Claudia's notes got this exchange a bit garbled (or perhaps it
*was* a bit garbled!). McCabe was asking whether the OA impact advantage
was merely an artifact of the fact that higher-quality authors are
selectively self-archiving their higher-quality work. I replied that
this self-selection quality-bias was most definitely *one* of the (at
least) 6 factors underlying the OA advantage, as I had listed them in
my talk:

But quality bias is certainly not the only factor, nor the biggest (though the
studies quantifying the relative size of the contributions of the 6 factors
remain to be done, and some of them are methodologically tricky to do): There is
already evidence, however, that the OA advantage cannot be just a self-selection
quality bias, because (i) the OA advantage persists even as the percentage of
self-archiving increases and (ii) even at 100% OA (as in astrophysics),
there remains an OA advantage in the form of 3 times as many downloads
per article, compared to pre-OA.

Also, given the fact (true for every single one of the 2.5 million
articles published in the world's 24,000 peer-reviewed journals) that each
of those articles is today inaccessible to many of its potential users
(because their institutions cannot afford toll-access to the journal in
which it appears), it is rather far-fetched to imagine that fulfilling this
*necessary* (even if not sufficient) condition for research usage and
impact does not actually generate usage and impact! We already know it
substantially increases usage in the form of downloads, and we already
know that downloads correlate significantly with citations. The rest of
the details will unfold with time.

This point was further elaborated in connection with Jan Velterop's
(friendly) question:

    Jan Velterop (from the floor): do the best authors self-archive?

    Harnad: The best scientists seem to self-archive first, so for the
    time being there is the advantage of a certain quality bias (if this
    correlation holds true); [but[ by the time 100% is OA, there will
    no [longer be] a quality bias [by definition!].

(4) Following Tim Brody's talk :

    Koch (from the floor): The UK Institute of Physics (IOP) says the
    number of downloads is about the same for OA journals as for their
    non-OA journals. Question: Does a high quality journal show the
    same effect as a low-quality journal?

    Brody: Good question, something to look into.

ISI compared OA and non-OA journals (equating for field) and found no impact
differences. This is comparing apples with oranges (as no two journals have the
same a-priori impact).

The right way to make this comparison is by comparing OA and non-OA articles
within the *same* journal, same issue/year. Tim did, and found a significant OA
impact advantage.

    Koch: Is the Latency inherent to OA or to technology use at large?

    Brody: in Physics, e.g., there is a strong pre-print
    culture, i.e. hits are generated through an alerting function rather
    than by browsing behaviour.

This refers to the fact that the peak of citations is getting earlier
and earlier across the years for OA content (in ArXiv): It is getting
cited earlier and earlier, because both the cited work and the citing
work are available earlier and earlier. This is factor 1 -- "Early
Advantage" -- of the 6 components of the OA advantage. But it is not
just for preprints! A self-archived postprint of a published article
also has this early-advantage over non-self-archived postprints in the
same journal/issue. And, as with preprints, the advantage is not just a
phase-advancing of the same citation curve (everything simply happening
earlier), but a true and permanent overall increase in the area under
the curve (i.e., the total number of citations).

But there is no denying that the early-advantage is most dramatic for preprints
(and it should be interpreted as yet another incentive to self-archive preprints
too!). The primary focus of the OA movement, however, continues to be the
peer-reviewed, published postprint, not the preprint.

    Harnad summary (from the floor): if an article is freely accessible
    online, it has a higher impact. Sceptical about comparing journals,
    rather compare same author's results re citation ranking.

This rather cryptic paraphrase has now been amplified in the discussion above.


    Frank Gannon, European Molecular Biology Organization (EMBO)...
    should governments and agencies interfere (e.g. NIH, Wellcome
    Trust)? (No scholar wants constrictions here, and the marketing
    of one's career should stay a personal freedom.)

    Prosser (from the floor): Telling researchers where to publish is
    not being practised; however, there is a proposal re funding bodies
    to claim to have a copy of the paper.

David Prosser's correct rebuttal did not come through sufficiently load and clear,
nor did mine:

    Harnad (from the floor): No one is proposing mandating OAP. The
    purpose of self-archiving is to make research available, and it is
    irrevelevant where it is put if it is put on an OAI compatible location.

The point is that none of the governmental, institutional and research-funder
OA mandates is a mandate to publish in Open Access journals! How could it be?
How can anyone tell researchers where to publish (other than that they should
aim for the highest-quality/highest-impact venue they can, as they do already)?
*All* the mandates, without exception, are OA *self-archiving* mandates, not OA
publishing mandates! Yet their critic keep arguing against them (irrelevantly,
hence ineffectually) as if they were mandates to publish in OA journals!

    "Guide for the Perplexed: Re: UK Select Committee Inquiry"

    "Critique of PSP/AAP Critique of NIH Proposal"

    "Critique of STM Critique of NIH Proposal"

    "Critique of APS Critique of NIH Proposal"

    "The UK report, press coverage, and the Green and Gold Roads to Open Access

    "AAU misinterprets House Appropriations Committee Recommendation"

    Gannon: What is the aim of the OA movement?
    (a) Reduce the power of big publishing companies?
    (b) Move away from impact factor driven decisions made
    by authors and selection committees?
    (c) Reduce (excessive) profits made from publication?
    (d) Make research results more available?
    (Gannon: this last aim is highly desirable);

The *only* aim of the OA movement is (or should be) (d): OA is
intended to solve the article access/impact problem, not the journal
affordability/pricing problem.

It is not the OA movement's business to (a) reduce the power of big
publishing companies (though 100% OA itself might eventually do that).

It is not the OA movement's business to (b) move anyone away from "impact
factor driven" decisions (though 100% OA will certainly give them a much
richer and stronger repertoire of objective impact indicators to use).

It is not the OA movement's business to (c) reduce anyone's profits
(though 100% OA itself might eventually do that).

The sole business of the OA movement is to make peer-reviewed research articles
100% OA so as to maximise their research impact by maximising potential users'
access to them webwide.


    Re: Jan Velterop's presentation:

Because Jan had said during his talk that he thought OA self-archiving
was merely treating the symptom (head-ache) and not curing the cause,
I asked:

    Harnad (from the floor): What is the purpose of the OA movement?

    Velterop: Can only answer for BMC: BMC believes OA is good, and that
    it can be done on a commercial basis.

And one can only add, for other than BMC, that OA is not only good,
but it is the purpose of the OA movement! And OA can also be done on
a non-commercial basis, by author/institution self-archiving!

    Harnad: Non-Open Access is the pain.

    Velterop: BMC is also killing the cause of the pain.

The cause of the pain is needless impact-loss from Non-Open Access and
100% OA self-archiving cures that pain, completely. The rest is purely
speculation about the *consequences* of 100% OA self-archiving, not
causes. (Hypotheses non Fingo.)

Stevan Harnad

A complete Hypermail archive of the ongoing discussion of providing
open access to the peer-reviewed research literature online (1998-2004)
is available at:
        To join or leave the Forum or change your subscription address:
        Post discussion to:

UNIVERSITIES: If you have adopted or plan to adopt an institutional
policy of providing Open Access to your own research article output,
please describe your policy at:

    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 Fri Dec 10 2004 - 13:38:25 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:43 GMT