Re: Author Publication Charge Debate

From: Stevan Harnad <>
Date: Mon, 9 Feb 2004 16:21:18 +0000

Before replying, I would like to summarize the reasoning of A.R.
Suhail as I understand it:

    (1) Many researchers have to pay for access to toll-access (TA)
    journals and articles out of their own pockets. (Their research
    institution, if any, is not the one that pays.)

    (2) If they in addition had to pay to publish in Open Access (OA)
    Journals out of their own pockets, they would not be able to afford
    to publish.

    (3) Therefore (i) OA journals are not a good idea.

    (4) Nor is (ii) OA self-archiving a good idea (because it might lead
    to OA journals and hence author payment).

    (5) Hence only embargoed access -- i.e., free online access after a
    one year delay, as offered now by a few TA journals -- should be
    sought, not OA.

Suhail also suggests that he is writing only about and against OA
journals, not OA self-archiving (but of course if OA self-archiving could
eventually lead to OA journals, and publishing in OA journals would be
unaffordable for such authors, then his reasoning, if correct, is also
against OA self-archiving, indeed against OA by any means).

I will let Jan Velterop of BMC and Mike Eisen of PLoS reply about the
question of how individual authors who cannot afford to pay OA publishing
costs are currently accommodated by BMC and PLoS. I believe there are
mechanisms in place to ensure that no author who cannot pay need pay.

As an advocate of OA self-archiving rather than OA publishing at this
time, I do not need to answer this question, because for my approach (the
"green" road) it is merely a hypothetical question ("What if the eventual
outcome were to be X?") whereas for the gold road of OA publishing the
question is an actual and immediate one.

However, I do have an answer -- equally hypothetical -- to the
hypothetical question of how OA journal publication costs would be
covered if OA self-archiving were eventually to induce a transition from
the TA cost-recovery model to the OA cost-recovery model:

    'The Green Road to Open Access: A Leveraged Transition'

    "(1) The average total revenue per article jointly paid today by
    those institutions that can afford the tolls (for one of the 2.5
    million articles published annually in one of the planet's 24,000
    peer-reviewed journals) is about $1500-$2000.

    "(2) The high-end estimate of the cost per accepted article for
    peer-review alone is $500 (or about 33% of the total revenue per

    "(3) If and when all institutions no longer spent their current
    serials budgets to pay those tolls for access to the incoming
    peer-reviewed articles from other institutions -- i.e., if and when
    there were 100% annual windfall savings from toll cancellations --
    there would be enough money to pay the peer-review service costs
    for their own *outgoing* research three times over."

Note that this is all reckoned at the institutional level, where
the lion's share of the current toll-based cost-recovery also occurs

Suhail is somewhat careful about his conditional probabilities: Pointing
out that he is writing from the standpoint of those authors who have
to pay for access to TA journal articles out of their own pockets,
Suhail estimates that that corresponds to 2/3 of authors publishing in
TA journals today -- but that that 2/3 does not apply to the number of
*articles* appearing in TA journals today.

I do not know where the 2/3 figure comes from, but clearly if the actual
proportion of *articles* published by those authors were far smaller --
say 10% of the total number published in the top half of the journal
hierarchy -- then the natural solution so that the entire research
community worldwide should have the benefits of OA would be to cover
all OA costs only from the 90% of articles published by institutionally
affiliated authors (out of their institutions' annual windfall 100%
toll-savings), and add the cost of the 10% unaffiliated authors to
the annual cost per article.

As far as I know, this is *already* what BMC and PLoS are doing -- so
far informally and pre-emptively -- even now, when there has not yet
been a penny of institutional windfall toll-saving (because there is
not remotely enough OA yet to allow the institutional cancellation of
TA journals!): They are not charging authors who cannot afford to pay.

Suhail paid, once. Perhaps that was a mistake. He should have requested
not to have to pay. That would be the better strategy next time, rather
than to suggest that OA should come to a halt worldwide -- which is
rather like suggesting that public transport or medicare should come to
a halt because I have once paid a bill needlessly.

Now some specific quote/comments:

On Sun, 8 Feb 2004, Suhail A. R. wrote:

>sh> I think A.R. Suhail continues to misunderstand a rather fundamental point
>sh> here. His suggestion is that OA journals are not a good idea because
>sh> they are unaffordable for some authors. My reply was that there is a
>sh> free alternative for any author who either cannot afford to publish in
>sh> an OA journal or for whom a suitable OA journal does not exist: Publish
>sh> in a conventional toll-access (TA) journal and make your article OA by
>sh> self-archiving it.
>sh> Suhail unfortunately continues to pass over this point in silence,
> I think that the point of contention here is that I do not think that OA
> journals & OA self archiving should be lumped together as one and the same
> idea about OA. They are entirely different and right now I am discussing OA
> journals only. I don't understand how self archiving my TA article makes OA
> journals a good idea? By self archiving my TA article, I am actually
> furthering the concept that OA journals are not a good idea. We have to
> either support OA self archiving or support OA journals, not both.

(1) OA self-archiving and OA journal-publishing are not "entirely
different" as both are means of providing OA.

(2) It is not the case that "We have to either support OA self archiving
or support OA journals, not both": We can and should support both.

(3) OA self-archiving in particular was recommended to Suhail because he
said he could not afford OA journal publishing in order to provide OA
to your articles.

(4) The causal connection between the two was implicitly made by Suhail,
because he took a stand against *OA* itself -- which is the outcome of
both OA strategies -- on the grounds that OA would lead to OA journals
and hence to unaffordable publication costs for authors who currently
pay their TA costs out of their pockets.

(5) Suhail argued that OA was accordingly a bad idea, and that we should
instead advocate embargoed access: TA followed by free online access
after a one-year delay.

This seems to me either a case of throwing out the baby with the
bathwater, or the tail wagging the dog, or both.

>sh> returning repeatedly to his own point that if an author spends his money
>sh> on publishing in an OA journal (as Suhail has done, once), he has less
>sh> money to purchase access to TA articles.
> This is precisely the point I am trying to make

Then let the author (i) continue publishing in TA journals, (ii) continue
spending his money to purchase access to TA articles (by other authors)
*and* (iii) provide OA to his own articles (for other authors) by
self-archiving them. If those other authors do the same, that just might
start saving this author some of the money he is spending to purchase
TA articles. Forget about the spectre of unaffordable OA publication
charges: They are counterfactual conjectures at this time, and there
are good reasons to believe they will never happen.

>sh> Suhail also repeats that (because of the OA journal affordability
>sh> problem), delayed (embargoed) access a year after publication should be
>sh> the goal, not OA! (Once again, passing over OA self-archiving in silence.)
> Self archiving may be good, but then I consider OA self archiving a topic
> distinct from OA journals, and my main concern is the threat posed by OA
> journals to authors like us. Why the emphasis on OA self archiving?

Because they are a means of providing OA, and this is all about the benefits and
feasibility of providing OA. If OA journal publishing is unaffordable for
you, *just don't do it*! Self-archive instead.

>sh> Over 95% of journals today are TA and fewer than 5% are OA. Consequently,
>sh> the OA journal option for providing immediate OA to an author's own articles
>sh> is a 5% option even if *every* author could afford it. The only OA option
>sh> for the remaining 95% of articles is already self-archiving. Wouldn't
>sh> the rational application of Suhail's observation that not all authors
>sh> can afford to publish in an OA journal be that those authors' articles
>sh> (whatever fraction they actually represent of the 5% for which there
>sh> exists a suitable OA journal today) should simply be added to the 95%
>sh> of articles for which the author provides OA by self-archiving them --
>sh> rather than to take them as an argument against providing OA through OA
>sh> publishing if and when it is available, and affordable?
> Authors providing OA is not the issue here. The issue is OA journals
> becoming more popular thus closing the doors for authors without
> money. How will self archiving stop this threat? Will there be a need
> for OA self archiving if TA journals convert to OA?

There is no "threat" from OA journals: Most journals today are
TA. Immediate OA can be provided by self-archiving. OA itself is not a
"threat" but an immense (attainable, and long-overdue) benefit.

"Will there be a need for OA self archiving if TA journals convert to OA?"
Yes, because part of that (hypothetical) conversion would involve
cost-cutting and downsizing from the present TA publication
model to the OA model: scaling down to only the essentials and
their costs. Peer-review may well prove to be the only essential
(hence the only cost to be recovered). Text-generation as well as
access-provision and archiving would be offloaded onto the worldwide
network of institutional OA eprint archives (e.g.., OAIster, writ large: ).

But far more than any speculative role it might play once the golden age
of universal OA is upon us, OA self-archiving is crucial in order to get
us there from here. It is the 95% solution, right now. Arguing for turning
the 5% solution into a still-birth -- on the grounds (1) that one has
oneself needlessly paid an unaffordable OA publishing charge one time,
and (2) that *if* this were generalized to all journals today, with
no adaptive changes at all, it would make all publishing affordable for
such authors -- seems to me as counterproductive as it is counterfactual.

>sh> (In addition, it needs to be pointed out that the OA/TA distinction mainly
>sh> concerns institutional tolls, rather than individual-user tolls. Users
>sh> have access to the TA journals for which their own institutions can
>sh> afford to pay the tolls. They may occasionally take out an individual TA
>sh> subscription, or may pay TA for an individual article on an individual
>sh> pay-to-view basis. But the lion's share of worldwide TA expenditure -- and
>sh> of the TA access-denial/impact-denial problem -- is in the institutional
>sh> tolls; individual subscription tolls have long ceased to be a significant
>sh> means of accessing the journal literature in the online age. Moreover,
>sh> at most institutions, even the user's individual pay-to-view and
>sh> interlibrary-loan tolls are paid by the user's institution, rather than
>sh> by the user. Once again, one cannot and should not generalise from the
>sh> fraction of cases where a user may still be spending money from his own
>sh> pocket to access the journal literature to the vast majority in which
>sh> it is the user's institution that is bearing this burden. And certainly
>sh> not as a rationale for arguing against either OA journals or OA.)
> I do not agree to this either. It is a fraction of users in USA & UK, but
> for the vast majority of countries worldwide, research is done by the user
> spending money from his own pocket to access journal literature. While
> libraries exist even in our countries to help get access to an article,
> there is no system in place to help for paying OA author costs at all.
> Therfore while paying for access is bearable, paying for publication is not.

So don't pay! Self-archive!

>sh> If OA journals are unaffordable for an author's own articles because of the
>sh> author's individual TA expenditures for access to TA articles by others, let the
>sh> author not publish in an OA journal but make his own TA articles OA by
>sh> self-archiving them -- and use all his available money to pay for the TA articles
>sh> of others as before.
> That is a good idea, and is precisely what I am saying. However I am also
> adding that OA journals should not be allowed to take over from TA, because
> then we wouldn't publish and then will not have anything to self archive.
> That is logical I hope?

I didn't hear you say anything about doing self-archiving, just about
*not* publishing in OA journals! What would be logical, and helpful,
would be to "allow" (indeed encourage) authors to provide OA by whichever
of the two means is suitable for them, and not to formulate counterfactual
conjectures (particularly ones that are so easily and naturally countered
by counter-counterfactuals)!

The reality today is that we have mostly TA, and consequently a great deal
of research access and impact and productivity and progress is being lost,
needlessly. OA provision is now feasible, by whichever of the two means
is suitable to the author. Authors should provide OA, now, by whatever
means is suitable to them. That is all they need to know, or do.

The rest is all just a prolongation of at least a decade already lost
in Zeno's Paralysis:

>sh> Correct. Which is precisely the reason all authors should provide
>sh> immediate OA to their own articles -- whether by publishing them in an OA
>sh> journal (5%) or by self-archiving them (95%). What author wants a 1-year
>sh> embargo before those would-be users who cannot afford the access-tolls
>sh> can use his work! How does that help develop their careers?
> It is not a one year embargo on access, its a one year embargo on *free*
> access. Comparing this to author charges is like comparing apples to
> oranges.

Of course that is what I meant too! TA followed by toll-free access after
a year. (Who on earth would propose suppressing all access whatsoever
for a year!)

    "Shulenburger on open access: so NEAR and yet so far"

>sh> (1) "OA" does not equal "OA journal-publishing" (5%). (There is OA
>sh> self-archiving too (95%)).
> They are both completely distinct issues

They are complementary means of providing OA, and even have some direct
causal connections between them. They are most definitely not "completely
distinct issues."

>sh> (2) OA is beneficial to the author as well as to the user.
> I disagree in terms of OA journals. Self archiving is a different issue

You disagree about the benefits of OA journals to authors who cannot afford to
publish that way. You do not (I hope) disagree about the benefits of OA journals
for the users of those journals. Self-archiving is not a different issue. It is a
means of providing those benefits (5%) far more quickly and far more widely (95%).

>sh> (3) OA's benefits (or feasibility) have nothing to do with whether or
>sh> not research is "institutionalised."
> For OA journals, yes. For self-archiving, no.

For both: If you cannot afford to publish in an OA journal today,
don't. But you still have (like all other users) the benefits of the
(toll-free) *use* of articles in those OA journals. And for your own
articles you can provide the benefits of that use (to their would-be
users) and the benefits of the resulting research impact (to you and
your research) by self-archiving them.

Both the benefits and the feasibility, with both (complementary)
means of providing OA.

>sh> (4) The feasibility of providing OA does not depend on whether or not
>sh> the author can afford OA publishing charges.
> For OA journals it does. For OA self archiving it does not.

That would only be true if self-archiving were not in fact universally
feasible to all authors, in the same way that taking a bus to work is
feasible to all passengers: It is as if I say "The feasibility of getting
to work does not depend on whether or not the passenger can afford taxi
charges" and you reply: "For taxis it does. For busses it does not."

>sh> (5) Supporting cost-recovery for OA journals is not a mistake but
>sh> a necessity, if there are to be OA journals at all; but neither
>sh> publishing in OA journals nor paying its costs is a necessity in order
>sh> to provide OA to an author's articles: There is also the 95% alternative:
>sh> self-archiving.
> It is self defeating to support OA journals if you can not publish in them

Support the two roads to OA and take the road that is suitable to you.

>sh> (6) If all articles for which (i) there exists a suitable OA journal
>sh> today (c. 5%) and (ii) their author can afford it were to be published
>sh> in those OA journals, that would be very good for OA, and would help OA
>sh> journals grow beyond 5%, but it would come *no where close* to "closing
>sh> down" the remaining TA journals (95%)!
> I hope we never reach close to that while we poor authors are around

There are many hypothetically possible outcomes if/when OA prevails:

Let us not help hold back OA -- already so long overdue and so beneficial
in its own right -- on the strength of *one* of those many hypothetical
possibilities (and not even the most probable one!).

>sh> (7) Even if all the remaining articles (95%) were made OA by their authors
>sh> by self-archiving them, it is not at all clear that that that would close
>sh> down or convert the remaining TA journals (95%) into OA journals. That
>sh> is a possibility, but by no means a certainty; if it did, though,
>sh> it would also generate institutional windfall savings that would
>sh> be more than enough to pay for all institutional author OA journal
>sh> publishing charges without requiring a penny from the author's pocket
> Assuming that an institutional system of author assistance exists - the
> truth is that majority of countries worldwide don't have such a system

It is not an institutional system of author assistance that is missing. What is
missing is OA. No one needs any assistance in order to provide OA to his own
articles today. That is the only relevant fact. The rest is all just passive
(indeed paretic) speculation.

>ars> Finally, why should an OA journal decide who or who not to give waivers to?.
>ars> I think one way around OA is to make waiver rules that are not in the
>ars> individual journals control. In other words, if an OA journal is to start,
>ars> then it has to follow an international waiver standard by law. These waivers
>ars> would then automatically apply to non funded research regardless of where it
>ars> originates from. The journal then makes its living from funded research
>ars> work. Will this work....I don't really know.
>sh> I suggest that these recommendations be directed to the OA journal
>sh> publishers. They are of very little immediate relevance or interest to
>sh> those whose immediate interest is in Open Access, Now. (Why an
>sh> author for whom OA journal publishing is unaffordable today would prefer
>sh> pondering OA journal "waiver" policy rather than immediately providing
>sh> OA to his articles by self-archiving them is another of Zeno's Koans for
>sh> which I have no answer, and can only leave to future historians of the
>sh> research community's needlessly sluggish trajectory toward the optimal
>sh> and inevitable outcome for research, researchers, and the society that
>sh> funds and benefits from the research.)
> These recommendations are what make open access viable. If these
> recommendations are not taken seriously then viable open access boils down
> to self archiving. Now will self archiving be required after the demise of
> TA journals?. The answer is no. Regardless of how strongly proponents of OA
> feel about its benefits the undeniable reality is that open access journals
> are not for everyone. They are for the privileged few with resources.

(1) These recommendations (about ensuring that provisions are made for
all authors who cannot afford to publish in today's OA journals, today)
are *not* what makes OA viable. What makes OA viable is whatever provides
OA. There is still very little OA being provided today, but three times
as much of it is being provided via OA self-archiving (under 15%) as via
OA journal-publishing (under 5%) -- but the green road of self-archiving
is the more underused one, for its immediate capacity is already 55%
at the absolute minimum, and in reality 100%!

(2) If the recommendation to ensure that provisions are made for all
authors who cannot afford to publish in today's OA journals, today, are
*not* taken seriously then this merely means that that many more authors
will need to provide OA via self-archiving rather than via OA publishing.

(3) Will OA lead to the conversion of all TA journals into OA
journals? (No one knows. I rather think it eventually will, but I also
think that many natural adaptations will first occur along the way,
mainly a downsizing of journal publication to just the essentials,
thereby lowering substantially).

(4) Will self-archiving still be needed if/when all TA journals convert
to OA? Yes, because the access-provision and archiving functions will have
been offloaded onto the network of institutional OA eprint archives,
thereby allowing much of the downsizing and cost-reduction in the
transition from TA to OA.

Stevan Harnad

NOTE: A complete archive of the ongoing discussion of providing open
access to the peer-reviewed research literature online (1998-2004)
is available at the American Scientist Open Access Forum:
        To join the Forum:
        Post discussion to:
        Hypermail Archive:

Unified Dual Open-Access-Provision Policy:
    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 Mon Feb 09 2004 - 16:21:18 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:20 GMT