Re: FOS Newsletter Excerpts

From: Stevan Harnad <>
Date: Fri, 7 Sep 2001 18:37:45 +0100

Peter Suber's Free Online Scholarship Newsletter

is such a valuable resource that I have bounced my own copy to the
Forum (at the risk of having my subscription unsubscribed by anyone who
gets it!).

[Peter: is there any way to subscribe the AmSci Forum itself to the FOS
Newsletter in such a way that it does not receive the "UNSUB" code with
every Newsletter?]

On Thu, 6 Sep 2001, Peter Suber wrote:

> However, the most interesting and controversial source of revenue will be
> author fees. ... BMC has not yet adopted
> the policy, and will not do so until 2002 at the earliest. The idea is to
> charge authors about $500 per article. Jan estimates that this will cover
> the full cost of electronic publication (peer review, mark-up, hosting, and
> preservation). He also estimates that it is roughly one-tenth the cost of
> print publication, at least in the STM fields. The fee would be waived for
> authors from developing countries and in some other circumstances.

May I just add that the only ESSENTIAL cost here is the peer review
(implementation) cost (because peers referee for free)? The rest is
just an OPTION, an ADD-ON, if the work is self-archived by its author
in his institution's OAI compliant Eprint Archive.

It is for this reason a misnomer to call the essential cost of
peer-review (implementation) a "page charge," because that unfortunately
perpetuates the Gutenberg-era practise of holding the essentials (peer
review) hostage to the optional add-ons (mark-up, hosting, preservation).
Distributed, institution-based OAI-compliant Eprint Archives and
services can do the hosting and preservation, and the author can do the
minimal "mark-up" called for by OAI-compliance.

This frees the refereed literature. Then the optional publisher add-ons
can be paid for in any of a number of ways (Subscription, Site-License,
Pay-Per-View [S/L/P], etc.), Author-Page-Charges being only one of the

This PostGutenberg partitioning into the essentials and the add-ons
is not just a pedantic exercise. It is the only way to get the real
factors into focus. Then we can make an informed judgement, instead
of wrapping it all together into the one Gutenberg option, as before.

[There is some debate over how much the essential service (peer review)
actually cost, per paper. I think it averages $200 per paper, across
the entire corpus of 20K journals. APS Editors have reported in this
Forum that their essential peer-review costs are $500 per paper,
without counting in the add-ons (mark-up, hosting, preservation). These
are empirical questions, but insofar as BMC is concerned, the premise
here is that the $500 includes the essentials AND the add-ons, whereas
what we should be talking about is ONLY the portion that covers the
essentials (which I think will prove to be closer to $200 per paper for
the average journal).]

> What do you think of author fees?
> I'll be frank: I have mixed feelings about author fees. On the one hand,
> author fees give readers free online access to the literature and they give
> journals the revenue they need to make it happen. On the other hand, many
> authors won't be able to afford them. While I admit that journals
> providing free online access need some revenue, it remains the case (1)
> that journals needn't get their revenue from authors and (2) that we can
> achieve free online scholarship without getting it from journals.

Not only is it a misnomer to refer to page charges, but it is a misnomer
to refer to author fees. Some forms of payment (e.g., author page
charges) are indeed author fees. But the neutral way to describe the
costs of the peer-review service that provides the quality-control and
certification of a research institution's (e.g., a university's)
research output is as research peer-reviewing costs. Like all the other
costs associated with doing research, those costs are not the
author/researcher's but the research institution's.

So, to put it simply, it is misleading and prejudicial to present the
options as a choice between either (1) an S/L/P-fee-based, non-free
refereed literature or (2) an author-fee based free refereed
literature! Those are not the true alternatives.

The true alternatives are: (i) wrapping the essential peer-review costs
in inextricably with the inessential add-ons, and then charging for the
joint PRODUCT (via S/L/P, author fees, or what have you), as we do now,
or (ii) freeing the refereed draft (through institutional
self-archiving) and paying for the essential peer review SERVICE (if
and when the publisher's institutional S/L/P revenue is no longer
enough to pay for it) out of the (institutional!) S/L/P savings (not
the author's pocket!).

> First, journals needn't get their revenue from authors. The costs of
> online journals could be borne by universities, learned societies,
> foundations, governments, or endowments.

True, but we are not talking about the costs of online journals (or
their "pages")! We are talking about the costs of (E) the essential peer
review service and the costs of (O) the optional add-ons, separately.

One could just as well have exhorted the "universities, learned
societies, foundations, governments, or endowments" to pay the S/L/P
costs, and that would have freed the literature too, but alas there
is no guarantee that any of these worthy sources of funds will volunteer
to take on the expense! Whereas for peer-review service charges it is
clear precisely what it is that they are part of the cost OF (namely,
the refereed research itself) and it is also clear where the funds
to pay them will come from come from, if and when they are needed: the
need for them will come if/when declining publisher S/L/P revenue (for
the online page PRODUCT) is no longer enough to cover the essentials
(for the peer review SERVICE): they can then be covered from (only a
small portion of) the corresponding annual windfall institutional S/L/P
revenue savings (from cancellations of the S/L/P PRODUCT).

This scaling down to the essentials is completely masked if we think of
it as journal page-charges for the online version of the old Gutenberg
PRODUCT, namely, journal pages!

The (peer-reviewed) "pages" will instead be provided, free (like the
research itself), by the institutional Eprint Archives, and the journal
publishers will provide only the essentials: the peer review service
(paid for out of the institutional savings from cancellation of the
options) certifying them as refereed and accepted as having met that
journal's established quality standards.

> In my own scale of values, we
> should rely on reader payments last, author payments second to last, and
> advertising third to last. When readers have to pay, then readership is
> limited to those who can afford to pay. This hinders both research and
> education. When authors have to pay, then publication is limited to those
> who can afford to pay. This also hinders both research and
> education. When advertisers have to pay, then either objectivity or the
> appearance of objectivity is compromised. Even if advertising does not
> distort editorial policy, readers shouldn't have to wonder about whether it
> does. That leaves universities, learned societies, foundations,
> governments, endowments, and creative new ideas for generating
> revenue. Let's try these diligently before we conclude that they cannot
> work and that we must retreat to advertisers or authors.

By all means. But meanwhile, may I patiently suggest that we also go
ahead and self-archive our own respective portions of it all (all the
refereed papers of all the 20K refereed journals), rather than waiting
to first find volunteers to pay the putative costs if and when it
should become necessary?

> On the other hand, I acknowledge that we're not very close to
> institutionalizing the practice of supporting electronic publication
> through fees or contributions by universities, learned societies,
> foundations, governments, or endowments. For example, universities give
> disk space on their servers freely to faculty, but they are not as free
> with funds for copy editors or peer review facilitators. Most foundations
> will not even consider giving a grant to build an endowment for an
> electronic journal or other scholarly resource. Are author fees acceptable
> as an interim solution while we work on making them unnecessary? If they
> make literature free for readers, are they at least better than systems
> that charge readers?

It would be obscene to offload any of the costs on the
author/researchers' pockets for a "product" that is a hybrid of
essentials and options. Let authors just self-archive and let others
worry about how to cover the essential costs, should it eventually
become necessary.

> In the natural sciences more than the social sciences or humanities,
> research is funded, and it's very reasonable to ask funding agencies to
> subsidize publication. But how soon can we make it commonplace for
> foundations to provide for publication costs when making research
> grants? (How soon can we make it commonplace to require free online
> publication as a condition of research grants?) Even if the model doesn't
> transfer well to other fields, it might be made workable in the most funded
> disciplines or for the funded research within any discipline.

Grants should never be forced to pay for inessential options. But it
costs nothing (and would be an excellent idea) if grants mandated not
only that the research findings must be reported in refereed journals,
as they already do now, but that the refereed reports of this publicly
funded research must also be made publicly accessible free online.

> Second, we can have free online scholarship without getting it from online
> journals. The best way to do so is through what Stevan Harnad calls
> self-archiving. It works like this. Authors put unrefereed preprints
> online in institutional archives. Then they submit their articles to
> refereed journals. If the articles are accepted, and if the publisher
> allows, then authors put the refereed postprints in the same institutional
> archives. If a publisher does not consent to this, then the author puts
> the "corrigenda" (the differences between the final version and the
> preprint version) in the archive. Harnad and others have written free
> software for creating interoperable archives for just this
> purpose. Institutions can host these archives at no cost to them beyond
> the disk space they occupy.
> Having said that, I'd like to see self-archiving practised alongside a
> thriving system of free online journals. If this is to happen, then we
> still need a way to subsidize the costs of the online journals. What do
> you think about author fees as a solution to this problem? Please share
> your thoughts on our discussion forum.

The survival of the peer reviewed journals is essential to the system
just described, for they perform the essential service of peer review.
But there will never be any need for authors to reach into their
pockets to pay those essential costs: If and when it is ever needed,
because they are no longer being covered by institutional payments to
publishers for the add-on product, there will be more than enough to
cover the essential costs out of the annual institutional windfall
savings from the cancellation of the add-on product.

> BioMed Central debate on author fees
> FOS discussion forum

Stevan Harnad
Professor of Cognitive Science
Department of Electronics and phone: +44 23-80 592-582
             Computer Science fax: +44 23-80 592-865
University of Southampton
Highfield, Southampton

NOTE: A complete archive of the ongoing discussion of providing free
access to the refereed journal literature online is available at the
American Scientist September Forum (98 & 99 & 00 & 01):

You may join the list at the site above.

Discussion can be posted to:
Received on Fri Sep 07 2001 - 19:17:43 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:46:15 GMT