Re: The Green and Gold Roads to Open Access

From: Stevan Harnad <>
Date: Sun, 12 Dec 2004 19:10:22 +0000

On Sun, 12 Dec 2004, Matthew Cockerill wrote:

> On 12 Dec 2004, at 14:43, Stevan Harnad wrote:
> > Just to see where your mathematical logic leads, Rick, and then work
> > backwards:
> > are you suggesting that all 2.5 million articles currently published
> > in the planet's 24,000 peer-reviewed journals could be rechanneled to
> > just 1 of those 24,000 journals? No? Then let's do some mathematical
> > induction: 2? 2%? 5%?
> >
> > No, 95% of the annual peer-reviewed literature cannot be squeezed
> > through the keyhole of the 1383 OA journals known to exist today:
> > Moreover, there is a *function* being performed
> > by that large and diverse array of existing journals, not only in
> > covering all fields, but in covering, hierarchically, each level of
> > each field. That's what "choosing a suitable journal" means (above):
> > choosing the journal that covers the subject matter, and at the highest
> > quality level (peer-review standard) that the paper can manage to
> > successfully meet. This means (among other things) that journals must
> > be selective -- in some cases (the top) *very* selective. It is not
> > just a matter of squeezing all candidates into the tiny arbitrary subset
> > of them that happen to share a certain cost-recovery model today!
> Imagine that back in 1983, I tried to persuade you that mobile phone
> cellular networks were going to transform how people communicate.
> "Ah no", you would have counter-argued, "mobile phone networks
> only have the capacity to handle 5% of the calls that are being made, so
> clearly that can't be the way forwards. But meanwhile, everyone has a
> fixed line phone. And you know what? If you carry around a really long
> extension cord with you whenever you leave the house, that gives
> you pretty much the same benefits. So instead of grappling with these
> new-fangled, unfamiliar mobile phone thingies (which can only handle 5%
> of calls), we could *all* have mobile communication now, if only everyone
> would plug in a really long extension cable and carry it with them at
> all times. Frankly I am utterly baffled by people's inability to
> recognize that mobile communication is in their grasp. And why do people
> keep conflating mobile communication with mobile phones?!"

This is not an argument from logic, it is an argument from analogy, and both logic
and evidence quite naturally (and gently) show how the analogy does not even come
close to fitting either the facts of the logic of OA!

Your analogy is that OA (gold) journals are like cell-phones (when there
were still few) and that OA (green) self-archiving is like making the
cord of an ordinary phone longer. And that scepticism about the speed and
likelihood of growth of OA journals when they are at 5% in 2004
is like skepticism about the speed and likelihood of growth of cell phones
when they were at 5% in 1983.

Now let me count the ways in which the reality of researchers' needs and
journal publishing goes against the analogy with cell-phones (or diesel
engines, or motorcars, or computers, or TVs, or the web, or whatever piece
of technology you choose in your sanguine projections -- though there
will be something more to say about the analogy with the web in a moment):

(1) OA journals are not a new piece of either hardware or software: they are
merely a different cost-recovery model, and one that has not yet been tested and
shown to be viable, sustainable and scalable. (I am not saying it will not; I am
saying it has not yet been shown.)

(2) Until the viability, sustainability and scalability of the OA journal
cost-recovery model has been tested and confirmed, it represents an undeniable
risk for publishers.

(3) As a consequence of this risk, very few publishers have dared to adopt the
OA journal cost-recovery model to date. (This is not to say that the brave new
publishers like BMJ or PLoS were wrong to try, nor that they are bound to fail;
just that few have tried, and it is clear why not.)

(4) Now, because only about 5% of the total 24,000 peer-reviewed journals have
taken the risk of trying the OA cost-recovery model today, it follows that
only about 5% of articles can be published in an OA journal today,
even if the author, undeterred by the author-institution publication cost
(as, I agree, he should not be, if the journal is otherwise suitable)
wishes to publish in an OA journal.

(5) First pause: There is no counterpart for this in the growth of
cell-phone manufacture and usage: Providers were quite happy to have a go,
and users were quite willing to buy and use. There was no counterpart
of the uncertainty and risk about the cost-recovery model (which was
much the same as with the conventional phone): just ordinary innovation,
competition, and market forces.

(6) Now let's continue, with the "long-cord" story: Not only is there
already a viable alternative to the untested and risky OA cost-recovery
model (in which I believe, by the way -- but I also believe it is
premature), but, unlike the far-fetched "long-cord" analogy, which
clearly does not have even an infinitesimal portion of the functionality
of cell-phones, the self-archiving alternative provides 100% of the
functionality of OA, just as OA publishing does: If the author has
any benefits from OA at all, the benefits are equal whether the OA is
achieved via gold or green. Ditto for the user.

(7) Nor does self-archiving have any of the "long-cord" disadvantages
dictated by the analogy: Distributed institutional self-archiving is
simple, easy, and highly desirable in its own right, over and above its
power to confer 100% OA. (It could even eventually take over both the
access-provision and archiving burden from *all* journals, making them all
"wireless" peer-review service-providers instead of tying them down with
having to provide and distribute and store paper and PDF products! But
here I speculate only in order to show how unapt the long-cord analogy

(8) Yes, the self-archiving option requires that a few extra keystrokes
be performed (by the author or by some other designated party), but those
are one-time keystrokes per article (hardly comparable to walking around
with a mile-long cord!) and a lot cheaper (for somebody) than paying
the OA journal costs.

(9) The same is true of archive creation and maintenance. It costs a
little (very little) to an institution, to provide OA archives for all
its authors article output, but it costs incomparably less that creating
and maintaining a new journal, and has none of the attendant risks.

(10) Now back to innovative technology and the web. For if you really
want an apt analogy with cell-phone uptake, it is probably web uptake:
They both happened so fast, and conferred such huge benefits. Are OA
journals their counterparts here, or is it rather online self-archiving

(11) I would say that if anything is analogous to scepticism about the
functionality and desirability of cell-phones in 1983, and preference for just
making the conventional telephone line longer, it is authors' failure to
self-archive and reach 100% OA already a decade ago.

(12) Since then, OAI-interoperability has come along to sweeten the
package. So have free OAI-compliant software, citation-counting research
engines, green lights from publishers, and growing empirical evidence
of the impact-enhancing power of OA.

(13) Yet it still looks as if it will require stronger incentives from
researchers' employers and funders (in the form of requiring OA provision
as a condition of funding and advancement, by way of an extension of
publish-or-perish, in the interest of maximizing the research impact of
those publications) in order to get them to take advantage of the new medium
at last.

(14) OA-provision mandates from employers and funders cannot and will not dictate
where researchers may publish. (So they cannot be mandates for publishing in OA

(15) But OA self-archiving can be mandated it. And that will bring 100% OA
(wireless) with 100% probability.

(16) It may or may not also eventually usher in the era of OA publishing.

(17) But that is not something that authors are or should be concerned about.
Nor does it matter remotely as much as reaching 100% itself.

(18) For 100% OA (wireless) will already satisfy all authors' needs --
at least all those needs that would have inclined them toward OA at
all, in either its gold or its green varieties!

Stevan Harnad

P.S. A much more interesting critique of my argument that you can't
squeeze 95% of the current literature into 5% of its current journals
would be something along the lines of the Parmenides Paradox or the Ship
of Theseus (by way of a critique of my induction on 1, 2, 2%, 5%,...). But
the answer to this has been given already: We need to create/convert
roughly 23,000 more OA journals to scale up to 100% OA whereas we could
already do so virtually overnight via OA self-archiving. But alas it is
easier to squeeze a camel through the eye of a needle than to get most
publishers to adopt the OA cost-recovery model today. It is hence
incomparably easier and more sensible to squeeze researchers just a
little, so they self-archive: After all, the OA is all for their benefit anyway!

And researchers have already *told* us what they will and will not do: They will
not self-archive if they are not required to do so -- but if they *are* required
to do so, the vast majority say they *will* do so, and do so *willingly* (just as
they already comply willingly with the requirement to "publish or perish").
So if OA is worthwhile at all, it is worthwhile mandating that researchers
do their part to make it happen.

    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.
Received on Sun Dec 12 2004 - 19:10:22 GMT

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