On Thu, 5 May 2005, Michael Carroll wrote:
> I have a question that I'm sure you've covered in a prior
> posting, but it would help to get the executive summary version
> of your answer. I hear you to be arguing:
>> 1. 100% OA is desirable from society's perspective.
> 2. 100% OA also is in authors' and their employing institutions' self
> interest, respectively.
> 3. 100% OA does not require changes in business models or allocations
> of copyright.
> 4. Technological hurdles to 100% OA are trivial.
> 5. Therefore the only real obstacle to 100% OA (at least in STM
> literature) is authors' unwillingness to step over the trivial
> technological hurdle.
>> My question to you as a scientist is why? What explains the failure of
> so many authors to act in their self interest?
Swan & Brown's survey results answer this question completely.
Swan, Alma and Brown, Sheridan (2004) Authors and open
access publishing. Learned Publishing 17(3):pp. 219-224.
Swan, Alma and Brown, Sheridan (2004) JISC/OSI JOURNAL AUTHORS SURVEY
Report. In JISC Report
The one-word summary is KEYSTROKES. It is merely keystroke inertia that stands in
our way (and has been for 10+ years), nothing else.
Here is an executive summary of the reasons the keystrokes are not being
performed, in order of priority:
(1) Author unawareness of the impact-enhancing effects of OA
(2) Author unawareness of the possibility and the means of doing
(3) Among the (minority of) authors who are aware of OA and its
benefits (1 and 2), the incorrect assumption that it is time-consuming
to do the keystrokes (it actually 6-10 minutes per paper, i.e., 40 minutes
per year for the average researcher):
Carr, L. and Harnad, S. (2005) Keystroke Economy: A Study of
the Time and Effort Involved in Self-Archiving.
(4) Author reports, from Swan & Brown's two international,
cross-disciplinary surveys, that they are busy, overloaded, and
will only self-archive if/when they are required to do so by the
employers and/or funders, but if they are so required, 79% report
they will do it WILLINGLY, 17% that they will do it reluctantly,
and only 4% that they will not do it.
So what is clearly needed for immediate 96% self-archiving is
(a) a mandatory self-archiving policy for universities and
research-funders (as recommended by Berlin 3 and the UK Select
(b) an information campaign to inform researchers,
their institutions and their funders about (i) how OA self-archiving
enhances research impact
as well as (ii) how to go about self-archiving
> To put it another way, your consistent response to various posts is
> that they miss the point - the only thing that has to happen is to get
> scientists to use the rights they have under their copyright agreements
> to self-archive. If we all were to agree with you, what action(s)
> has/have to occur to bring about that change in behavior?
> One answer could be that institutions have to mandate self-archiving
> (and I know you have many posts along these lines), and the studies show
> that many scientists would welcome such a mandate. But that response
> actually indicates that scientists either (a) don't know that they
> already can self archive; or (b) don't think the benefits to them of
> self-archiving are worth the costs of doing it unless they're ordered
Not quite: What it indicates is that that researchers (and their
institutions) (a) don't know that they already can self archive (as you
AND they (b) don't know the impact-enhancing benefits of self-archiving,
AND (c) they don't know how to self-archive, nor what time/effort it involves,
AND consequently, being busy, (d) they will not do it until/unless it is required.
One might note in passing that many of them would not publish either, unless
it was required ("publish or perish")!
So what is needed is carrots and sticks: Information on the benefits and on the
ease of the means, and a policy requiring it. The rewards (for publication and
citation impact, which hiring, promotion and even funding committees are already
rewarding) are already in place. It is merely the causal connection that needs
to be made explicit for researchers.
(For 32 other causes of Zeno's Keystroke Paralysis, see:
> If your answer is along one of these lines, what specific actions need
> to be taken - and by whom - to alter authors' understanding or
> cost-benefit calculation?
Maybe only a crass translation of all of this into salary and research
dollars will do the trick (and I have urged doing that explicit
translation many times before!).
Here is the preprint of a short paper I have just submitted to Research
Money Magazine (in Canada):
Canadian Universities Need to Self-Archive Their Research Articles
Online To Maximize Their Research Impact
The research community fills about 24,000 peer-reviewed research
journals across all fields and languages worldwide, publishing about
2.5 million articles per year. The output of one research-active
university might be from 1000 to 10,000 or more articles per year
depending on size and productivity. Researchers are employed,
promoted and salaried -- and their research projects are funded --
to a large extent on the basis of the usefulness and impact of their
research. Research that is used more tends to be cited more. So
citations are counted as a measure of usage and impact.
The dollar value (in salary and grant income) of one citation varies
from field to field, depending on the average number of authors,
papers and citations in the field; the marginal value of one citation
also varies with the citation range (0 to 1 being a bigger increment
than 30 to 31, since 60% of articles are not cited at all, 90%
have 0-5 citations, and very few have more than 30 citations:
A much-cited study estimated the "worth" of one citation
(depending on field and range) in 1986 at $50-$1300:
One of the ways researchers try to maximize the usage and impact of
their research is by submitting them to journals with high "impact
factors" (i.e., average citation counts per article). Journal impact
factors vary as citations do: Most journals hover just below and
above 1 (excluding author self-citations); journals with impact
factors above 30 are rare. Success in getting a paper accepted by
a high impact journal depends on the paper's quality and the rigor
of the standards of the journal's peer review system. In general,
higher impact journals (in the same field) tend to have higher
But now there is a new way to increase every article's research
impact, over and above publishing it in the highest quality journal
whose peer review standards it can meet: The online medium has now
made it possible for authors to supplement the usage and impact
that their research receives from those users whose institutions can
afford to subscribe to the journal in which the article is published
with the usage and impact of all potential users whose institutions
cannot afford to subscribe to the journal in which it is published
-- by self-archiving an online version of the article in their own
institutional web archive, openly accessible to all would-be users
There is now a growing number of studies on research impact
for articles across all fields,in each case comparing the
citation counts (always within the same journal and year)
for articles that have and have not been self-archived by
their authors. With virtually no exceptions the articles that
have self-archived supplements are turning out to have 50%
to over 300% greater research impact than those that do not:
http://opcit.eprints.org/oacitation-biblio.html. Considering that
90% of research articles today have 5 or fewer citations, this is
a dramatic result for research progress itself, even before we try
to translate it into its financial "worth" to researchers and their
institutions in terms of prestige and research income in 2005.
Yet, despite its substantial benefits, self-archiving -- now at
10-20% across fields -- is still growing far too slowly:
There exist at least 200 institutional open-access archives worldwide,
but most are less than 20% full, relative to each institution's annual
output of research articles. Canada, with 27 of those archives, is
fourth in the world in archive number (after the US, UK and Germany)
but its archives are as underfilled as the rest, even though Canada
is also high in proportionate research output
Researchers have been slow to self-archive, partly because
they are not yet aware of its benefits, and partly because
they feel they already have enough to do (unaware that
it takes only 6-10 minutes per article to self-archive it:
http://eprints.ecs.soton.ac.uk/10688/). Publishers are certainly not
at fault for the fact that authors have been so slow to self-archive:
Ninety-two percent of the 8450 journals surveyed to date (including
most of the top journals) have already given their authors an explicit
green light to self-archive: http://romeo.eprints.org/
In two international surveys, researchers have indicated
exactly what needs to be done to get them to self-archive:
Seventy-nine percent of authors replied that they do not now
self-archive, and will not self-archive, until and unless their
employers or funders require them to do so; but if/when they do
require it, they will self-archive, and self-archive willingly:
The remedy is on the way: At the recent international conference at
the University of Southampton UK on formulating a concrete policy for
institutions to adopt in order to implement the Berlin Declaration
on Open Access -- http://www.eprints.org/berlin3/outcomes.html -- the
delegates recommended exactly what the researchers in the two surveys
had indicated was needed in order to motivate them to self-archive:
an institutional self-archiving mandate. And soon afterward, some
of the world's biggest research institutions (including FranceÕs
CNRS and the multinational CERN) led the way by adopting the policy:
It is now time for Canada to follow suit:
to the benefit of Canadian researchers, their institutions, their
funders, their funder's funder (i.e., the Canadian tax-payer) and
to the benefit of (worldwide) research itself.