Re: Online Self-Archiving: Distinguishing the Optimal from the Optional

From: Stevan Harnad <>
Date: Wed, 11 Dec 2002 22:24:48 +0000

On Thu, 5 Dec 2002, Arthur P. Smith (APS) wrote:

> I assume you don't think journals, right now, as a whole, are unprofitable.
> So costs are definitely being covered right now. Libraries may be having
> difficulty paying for what they want, but they are at least paying for what
> they're currently paying for, right now. Extending access for any one
> journal electronically from those who currently pay to those who don't,
> for free or a very nominal fee (as is being done with many consortium
> agreements) makes access "affordable for all would-be user/researchers
> and their institutions" with essentially no change in payments at all -
> no need for tolls to "shrink to just cost" or worse "shrink to peer-review
> service provision" if current subscribers keep paying the same amount.
> So it could be done right now - and in fact it is already happening,
> through differential pricing and these consortium agreements: publishers
> for the most part do see the benefits to everybody from expanding access
> and have been very openly pursuing these options.

There are about 20,000 refereed journals a year, planet-wide, publishing
about 2,000,000 articles. I have no idea how many would-be
researcher/users there are, but an *extremely* conservative estimate
would be if we assume that at least the authors of those articles are
potential users of one another's articles, and that they produce and
average of 4 articles per year, which means 500,000 would-be users,
planet-wide, for those annual 2,000,000 articles.

Now, what you say above only makes sense if your idea -- which is to
extend consortial agreements [i.e., huge collective site-licenses] to
more institutions -- will result in (near) 100% of the planet's 500,000
would-be users having full-text access to (near) 100% of the planet's
2,000,000 annual articles.

The following is a wild guess on my part (but, considering how
conservative was my estimate of the number of would-be users, I think
it's a well-buffered guess): I doubt that even 10% the planet's would-be
users have access to even 10% of that annual corpus today. I also don't
believe it would approach anywhere near 100% access to 100% of it if
it were all made available to their institutions at cost. The reason
is that "at cost" would still be the cost of a product, namely a text,
and that product is simply no longer necessary. (Those who think it
is necessary should test that by trying to sell it as a separate
"value-added" product, in competition with the self-archived,
peer-reviewed, open-access version.)

All that is still necessary from peer-reviewed journal publishers is
a *service,* namely, peer review, and we already know how much that
service alone would cost the planet: At the conservative estimate of
$500 per article, it would cost $10,000,000 annually [this error later
corrected: should have been $1 billion]. Now what do you
think the planet is paying now, annually, for those 20,000 journals
(collectively, in subscriptions, license, and pay-per-view, by those
institutions that can currently afford it)? [Hint: about four times
that much, i.e., about $4 billion) And how much would the total cost
for that text-product be even if it were sold at cost? (Still far
closer to 40M than 10M)

Those, I think, are the relevant ball-park figures. And they are to be
recokoned for the peer-reviewed research literature as a whole, and not
just from the vantage point of one or a few journals.

But I don't think any of this has anything whatsoever to do with journal
publishers -- and certainly not with APS, which is the most benign
and progressive publisher of them all! It concerns the researchers
themselves, both when they wear their author-hats (seeking impact on
would-be users) and when they wear their would-be user hats: If
researchers don't have open-access now, because they have not done what
the authors of 200,000 papers in physics, 500,000 papers in computer
science, and who knows how many other authors have already done -- namely,
to provide immediate open access to their own peer-reviewed research
output by self-archiving it -- then they have only themselves, and
definitely not their publishers, to blame.

> What I just said applies to "right now" but the bigger worry has been about
> the future. Assuming, however, that my earlier statement was accurate, that
> the primary recent problem has been from new countries entering research
> activity with initially large activity/publication expense ratios, it seems
> like the future looks very promising too, and we need only wait for
> publication spending in these countries to catch up, as it appears to
> be starting to.

I think the future looks very promising too, but not because I imagine
that more countries and more libraries paying more money for the 20K
journals is going to provide open access. It's because the self-archiving
token is bound to drop for the rest of the research community -- and many
of us are working hard to ensure that that happens sooner rather than
later, by demonstrating to them, quantitatively, the causal connection
between reseach access and research impact, and what they can do to
maximize it, right now.

> The long term promise of electronic delivery is this ability for expanded
> access where communication costs grow only linearly with worldwide
> research output, not quadratically. We're making that transition now,
> and access to published materials (still subject to access tolls) is
> now expanding, reversing the shrinkage of the last 4 decades. The
> time when self-archiving could make a real difference has come, and gone,
> already.

As long as access is to a toll-based *product*, it will not be
openly accessible. Fortunately, the product is no longer needed (for
peer-reviewed research); only the service is needed -- and there is
enough to pay for that service several times over out of the annual
windfall savings on the access-tolls for the product, if and when it
ever becomes necessary. But meanwhile, researchers should forget about
all that and just self-archive!

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 September Forum (98 & 99 & 00 & 01 & 02):

Discussion can be posted to:

See also the Budapest Open Access Initiative:

the Free Online Scholarship Movement:

the OAI site:

and the free OAI institutional archiving software site:
Received on Wed Dec 11 2002 - 22:24:48 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:46:46 GMT