Six Flavors of Open Access:
Successes and Possibilities for STM Journals
June 2004
Liber Conference
Ann Okerson
ann.okerson@yale.edu
Introduction/Background
A broad
commitment among participants in the scientific and scholarly publishing
community to achieving the widest possible access to the literature has begun
to bear fruit. Many of the players have
had this kind of commitment all along, for example, through sharing around of
preprints in physical sciences disciplines.
As everyone knows, Paul Ginsparg's e-print server, arXiv, started
in 1991, when he automated the high energy physics community's pre-electronic
article-sharing practices, creating a mechanism for these articles to be
submitted and mounted on a server. In
turn, this innovation led the way for other disciplines; it is estimated today
that perhaps 140 e-print servers from small to large-scale exist. (It is worth nothing that other fields, such
as the biomedical sciences, did not have a preprint tradition and its
scientists are not much using new technologies to create aggregated article
distribution mechanisms.)
Libraries have
also demonstrated an energetic access commitment. This shows in various ways: through the
tradition of most libraries to provide on-site access to their collections; to
circulation of materials; and to books and articles being delivered through
interlibrary loan. In the United States,
these practices have been codified in the Copyright Act, Section 108, with
limitations being placed via both law and negotiated best practices.
The success of
the Internet for communications and distribution of information challenges all
of our prior assumptions about the way access will work in the future. For
example:
A. Do the definitions
of "widest possible access" need to remain as they have been in
print?
B. Do the costs of
publishing journals, if new technologies are astutely exploited, need to remain
at today's levels?
C. Should libraries
and library budgets have a future role as the aggregators of resources on
behalf of the research and educational communities?
D. Should the
value-adding features of journal publication continue as we have known them?
A number of
voices would answer more than one of the above questions with a resounding
"NO." These days, in most
conferences and meetings, emotions run high. People on panels such as this are
chosen perhaps because they have the "right" point of view, or else
one so annoying as to cause heated debate and awaken the audience! Intelligent and well-meaning people can be
found on different sides and continuums of the questions above.
I spend a lot
of time thinking about all these matters and what the role of my university and
profession should be in these exciting times, so full of potential and also so
full of pitfalls. Should things
change? That is a silly question -- of
course they should and they are. Is
change inevitably chaotic or is there any way to direct some parts of it and
learn from it?
The Six Flavors[1]
As soon as we
begin to talk about Open Access, however, we discover that it is a treat that
comes in many flavors, like ice cream. I
love ice cream, especially in the summertime, so perhaps that makes me a good
person to review the different flavors and talk about their merits and
possibilities.
For, to speak
seriously, the term "open access" has taken on multiple mutations or
flavors and when people argue about whether OA is "good" or
"bad," it often turns out that they are arguing about some different
things. Also, it becomes possible to
argue whether or which variants are real OA or some new, sneaky publisher
practice that sounds good but is intended to leave us all worse off than
before. For example, has an STM journal
publisher really had a change of heart when he issues a press release
trumpeting a commitment to some flavor of open access, or is said publisher
engaging in the time-honored business practice of giving away as much as
plausible in order to sell as much as possible?
Trying to guess at, or even to understand, others' motivations is
probably a dead end.
What I will do
here, then, rather than select a position and argue for it, is to review the
successes and possibilities of the present moment. To start, let us assume an ideal well beyond
anything we can achieve now: that ideal
is full online access to all research materials everywhere, materials of the
highest quality , from the moment of publication, 24 hours a day, 7 days a
week, 365 days a year, in perpetuity.
Measured against that ideal, I will identify (1) initiatives that have
shown success (such as: partial open
access, open access for developing nations, and delayed open access) and (2)
those that are actively objects of experiments (author-pays, institutional
repositories, and self-archiving). Then
I will conclude with some predictions and some cautions for the long-range
future.
1. Partial Open Access
This model is
in some ways very old, for publishers have long practiced giving away some
kinds of access to their materials while insisting on recovering income for
others. (As noted above, the protected
status of libraries in the U.S. Copyright Act indeed requires publishers to cooperate in that particular form of open
access, and the economic system of "big science" publishing has grown
up around the assumption that this extension of access will be part of the
system.) Today, many publishers
differentiate forms of their publication or the ways they allow access. We see this on the web all the time in
periodical sites (for example, Salon
magazine) that offer some content freely to all but then charge for 'premium
content.'
In scientific
publishing, perhaps the best example is the report publications of the U. S.
National Academy of Sciences, which has for several years experimented with
mixing free access and paid access. For
example, for some time the National Academy Press has posted the full text of
their reports for free access to all, charging for the traditional print books
themselves. Now, according to an April
2004 press release, the NAP appears to be charging not only for the printed
books but also for printable PDF files, while continuing to allow free and open
access to HTML files; developing nations are excepted from all PDF
payment. NAS follows the history of
their publications closely and will monitor carefully the impact on revenue and
use of this revised arrangement. Given
that NAP is starting to charge readers for something they used to receive for
free, this arrangement has attracted negative comment -- but it must be
admitted that if the present arrangement had come about replacing a prior
policy of charging users for all forms of access, it would be hailed as a major
step forward. On some level, this
partial access is open access: a version
free for all everywhere, as likely perpetual as the National Academy itself may
be.
By the way,
according to announcements on its web site, the British Medical Journal (BMJ),
after 10 years, is re-instituting subscription fee to its online customers as
of 2005; the exception being for developing nations, whose users will receive
immediate free access. After an
unspecified period of time, free access will continue to be made available to
all readers. The reason is cited as
rapidly declining library subscriptions and loss of revenue.
2. Delayed Open Access
Another way in
which publishers have traditionally segmented their market is by delay. The hardcover book that sells for $29.95 when
published today may become available – with exactly the same content – in a
paperback edition selling for $6.99 a few months later. The notion of differentiation by delay
underlies contemporary STM journal experiments in which publishers – here the
journals published by HighWire Press are probably the most consistent example –
commit to making their publications freely available on the web after a
specified period of time, typically 6 or 12 months. Customers who pay for the journals on this
model are those for whom the currency of new information is of high value. Here again, given ubiquity and perpetuity of
access, we come very close to the ideal of open access, bearing only the
limitation of some delay. The argument
in favor of this form is that it allows a proven business model to continue
functioning, while going a very long way towards facilitating access. The argument by some against this model, of
course, is that it does not go far enough.
To resolve that disagreement, one would need a way of thinking about the
value of information in its most current form, a value that would probably vary
sharply from one discipline to another and from one user to another.
3. Open Access for Developing Nations
Over the last
handful of years, a wide variety of arrangements have been announced and
implemented by scientific and scholarly publishers to the advantage of
residents of developing nations. (We
have catalogued some of the leading examples on a website:
<http://www.library.yale.edu/~llicense/develop.shtml>) This practice typically involves making
journals that would otherwise be electronically available only to subscribers
freely and openly available to residents of a defined set of nation
states. The definitions of
"developing nations" can vary (mostly according to published lists or
per capita annual income) but all emphasize the most economically challenged
nations of Africa, Latin America, and Asia.
The arguments
in favor of such a practice mix the noble and the ignoble, depending on who is
arguing. The noble argument holds that,
just as we seek ways now to make prescription medicines available at affordable
prices to the neediest peoples, so we can and should make scientific
information available to those who would otherwise be unable to afford the
prices paid by more affluent institutions.
The ignoble argument holds that in many countries the revenue-generating
market is trivially small by comparison to what it is in "first
world" countries. It makes sense,
therefore, to do away with the cost and bother of managing subscriptions to
clients in those countries. Given the
near-zero costs of electronic distribution, the loss to net revenue for a big
publisher is arguably trivial – and some suspect that perhaps publishers have
actually been losing money in those markets anyhow. For technical reasons, people in many
developing countries are still unable to gain access, but the increase in scope
of availability is nonetheless significant.
4. Author-pays Open Access
Author-pays is
the most ardently advanced form of OA for journals that continue to rely on the
model of the established publisher and the long-existing journal style. That is, author-pays advocates do not much
imagine a change in players or journal value-adding; they argue that only one
change needs to be made: i.e., the
journals need to be funded in essentially a different way. Advocates reason (1) that authors rather than
readers are the primary beneficiaries of article publication; and (2) that much
research is funded by government grants that could more fully support article
publication. Models that depend on users
to fund the publication should be abolished in favor of reliance on some form
of payment from authors who publish in the journals.
Proposals for
author-pays vary in their approach. Most
are not dissimilar to the long-familiar "page charges" – wherein
authors pay for part or all of the costs of publication, typically on a
per-page-published basis. In the past,
this funding approach was common among not-for-profit learned society
publishers and infrequent among for-profit commercial publishers. In recent years, learned societies have
tended to move away from page charges, arguing that by so charging, they put
their journals at a disadvantage in attracting the best submissions.
Note that there
is now developing a variant on the author-pays form of article funding, namely
"institution-pays." In this
variant, large research institutions (or their libraries), which house many
scientists who would publish in top quality journals, are offered a "bulk
rate" or "membership" – a flat fee that allows all their
scientists to publish in the journal at no additional charge or at a discounted
charge.
It is only fair
to say that the author-pays model must be listed among the maybes rather than
the successes. Several changes must take
place for this model to succeed.
(a) First, some pilot journals must show sustained success in
attracting top quality articles and in maintaining themselves in business on
the revenue stream generated by their model.
One of those pilots at the moment is functioning on a substantial
grant. provided by sponsors. This gives the journals a small number of
years in which to prove their concept's worth.
(b) Second, authors and institutions must find solutions to the
question of organizing payment. Will
funding agencies be willing to support publication through the grants that
support research (without reducing the funds available for direct support of
research)? Will research institutions be
able to identify indirect cost recovery dollars from research grants that can
be redirected to support scientists? Can
libraries take current subscription dollars and use them to support
institution-pays or author-pays science?
Can the overall system sustain the potential loss of revenues that are
now derived from private sector customers who may or may not directly engage in
research but consume the output of the world's science enthusiastically and at
high prices?
(To speak only to the library question, which impacts our library's budget: I am concerned
that these new model journals will, while they compete with traditional model
journals, have the effect of increasing the overall cost of access. We might have some difficulty justifying to
the university's leaders a decision to devote additional funds to unproven new
model journals on the hope that the
cost of traditional model journals will begin to go down by a comparable or
greater amount sometime in the future.)
(c) It is early days to understand the longer-term costs of
maintaining quality journals and so not easy to know what to charge authors
(those who can and will pay). Will one
up front payment sustain an article and surrounding apparatus over time, given
that no further revenue will be collected in any way, because every use will be
fair use?
5. Creating Institutional Repositories
To what extent,
some ask, are publishers and journals themselves necessary? Can research institutions eliminate middle
men and become their own publishers?
Some of the most committed enthusiasts for open access believe this can
and must happen. The view is that, whether
placed in traditional journals or not, articles by scientists ought now to be
posted to online, indexed, searchable repositories managed by their
institutions. The underlying notion is
appealing one and, in a way, reverts to the old model of the University Press,
where a small shop produced and distributed the books and journals of the
university's faculty. (Note that today
nearly all university presses are in deep trouble, with cutbacks everywhere.)
Many advocates
of this solution probably hold that in the end, ubiquitous access to research
results so published would allow users to think about suspending subscriptions
to traditional publications, which would then either disappear or diminish in
number and cost. Ideally, for-profit
journals would be put out of business altogether; the larger and most costly
society journals might also fade.
Whatever one thinks of this point of view and its objectives, here
again, there are challenges from funding the new while continuing to pay for
the old. Interestingly, what had once
been a hotly controversial preliminary question – whether traditional journal
publishers would allow authors to house their published articles in
institutional repositories as well – seems to have been resolved with an
official policy decision of Reed Elsevier to allow its authors to self-archive
both pre-refereed articles and refereed postprints, a practice Elsevier Science
had (mostly) quietly permitted earlier.
Though some critics hold that the limitations Elsevier has placed on this
right could be crippling, some other of the leading advocates of open access
have hailed the Elsevier decision as a great step forward.
6. Self or eprint archiving
The oldest
model in use of electronic open access is also the most open to continuing question. As noted earlier, the archive created at Los
Alamos by Paul Ginsparg in 1991 (winner of a prestigious MacArthur Fellowship,
Ginsparg is now at Cornell), has been a
success in its and other disciplines.
The archive itself cannot necessarily be a success for OTHER
disciplines, but its model could be. But
the eprint model has still not scaled globally and there are important pockets
of resistance. Eprint archiving seems to
raise the touchstone question that distinguishes believers in ideal open access
from a more pragmatic group. To explain
this view, I must move to my concluding remarks.
Conclusion
In conclusion,
we can lift up our eyes from the current scene and look out ten years. What do we really think might happen in that
time? Views differ – of course!
Some pessimists
hold that if we do not succeed in the fight for full open access, giant
traditional publishers will once again swallow us all in their maw and we will
live in a world much like the present one, only more expensive. One reason that I am not counted among the
believers in the most ideal form of open access is that I am not a pessimist of
this kind.
Another kind of
prophet (pessimist or optimist?) thinks that the existing system of scientific,
technical, and medical publishing will wither away, to be replaced by a world
in which information becomes free and where something like a mix of
institutional repositories and self-archiving, with some certification
overlays, will suffice to serve the needs of scholars, scientists, and students
at a level of quality as high as that we now enjoy. For this to come to pass, the possibilities
outlined above all must succeed at a high level, and, in particular, need to
achieve a near-universality of success that attracts the vast majority of
authors to use their new models. We can
welcome experiments that explore that set of possibilities to see what is and
is not possible.
But if, as is
probable, neither extreme emerges the winner, then the likeliest outcome will
be something in the diverse middle.
Imagine the possibility of a world in which many of today's existing
publishers (joined by new entrants of many sorts) survive, thrive, and redefine
their business plans. They eliminate or
move to OA the marginal (or very highly specialized) journals and concentrate
on providing new services around their best products – sophisticated indexing,
searching, knowledge management, and other possibilities we do not now fully
imagine. That is, an appreciable number
of publishers, each exploring their own particular markets and disciplines –
and publishers all have two important markets, authors and readers; they must
attract the best of both – take steps that provide integrated access and
services, for a cost, to their and to OA journals in ways equal to the best we
now imagine and perhaps better still.
At the same
time, let us assume universities and research centers could take more interest
in organizing the output of their scholars, moving towards the institutional
repository model, gradually, at different speeds and for different types of
materials in different institutions, depending on local needs and uptake.
Given the rich
array of options before us, at the end of a ten-year future history we are most
likely to find ourselves in a place more diverse and muddled as today's, and
doubtless with its own controversies. I
hope that in outlining such a chaotic future, I am suggest ways in which the
future for users of scholarly information will be appreciably better than the
present. Most of us do share the desire
to see the broadest possible access to the best possible information. If we do not achieve utopia, we will still
take pleasure in reaching a better place.
A few last
remarks: The biggest variable in
prophecy in this area is not connected to the elaboration of specific
experiments or business models, but rather to the vulnerability of today's
large, expensive, and contentious journal enterprise to substantial exogenous
events. What if either economic
circumstances in the global economy or unanticipated challenges to the business
model of the traditional commercial publisher should drive one or two of the
giants unexpectedly out of the STM journal business? Or, what if unfortunate circumstances were to
divert further funding from research and social programs to defense and battles
against terrorism? Would wars and rumors
of wars so far constrain funding for institutional research and science that we
would be challenged simply to maintain current models, accept the loss of many
journal titles, but labor hard to maintain the best we can? There is not much that any of us can do to
affect such possibilities, and we would doubtless all share in painful
consequences of realignment, should the negative unpredictable come to pass.
Even putting aside
the intrusion of world events, there is always the wild card of the talented
entrepreneur. Someone, and probably a
single individual, not a committee, may see an opportunity to restructure the
entire environment, as Bill Gates did in computing.
At this point,
I am leaving behind the domain of the invited paper or ice cream critic and
beginning to practice an alternate career as a political journalist or thriller
writer, so I will restrict myself to saying that the uncertainties that finally
surround the landscape in which we work are more substantial than we might
imagine, and that we cannot depend on wishes or prophecy to select our next course of action for us. Instead, we are constrained to think about
steps we can take, with existing resources, which responsibly and successfully
can move a large system collectively forward in multiple ways: improving the quality of the intellectual
product by supporting good science; improving the quality of the intellectual
product by supporting technological innovation in forms of presentation and
dissemination of information; and improving the quality of the intellectual experience of the product by bringing
the greatest benefits to the widest possible audience.
If we had only one task, our lives might be simpler: we have many tasks, and so we are very unlikely to know what "simple" means – at least in our lifetimes!
[1] For the
metaphor, thanks to John Willinsky, whose article I discovered quite
fortuitously and accidentally after having finished this text: "The Nine Flavours of Open Access Scholarly Publishing,"
in the Journal of Post Graduate Medicine (JPGM), Vol. 49, No. 3
(2003): 263 - 267. Also available on the Web via an easy Google
search and link.