Alma Swan's concise justification of open access in support of her
'progress' article came by way of a chance discussion on the
liblicense list from 2 April with the subject lines 'The Value of Open Access' and 'The value of open access & update to the Dramatic Growth of Open Access'. This thread was initiated by Peter Banks, former medical publisher and apparently the scourge of OA supporters, and highlighted how even some pro-OA cases can distract from the core motivations driving OA, i.e. access, connected information, and impact, which have to be kept at the forefront and understood if OA is to make faster progress. Here are some edited extracts from the thread.
Peter Banks:
There is no question that OA offers potentially significant benefits to society. All other things being equal, free public access to scientific information is clearly a good thing. Rick Anderson, Open access - clear benefits, hidden costs, Learned Publishing, 20, 83-84
"I think that this common assumption merits a far more critical examination than it has received. ... can access to scientific information accelerate research
"In talking with researchers at major research institutions, I have yet to meet a single one who felt that access to information was a limiting factor in research.
"As for the public and patients, there has been too little examination of how lay people use and misuse Internet information."
John Houghton, 4 April: "the marginal benefit of OA depends, in part, on the size of the margin. Perhaps we should begin with some evidence.
1. In a survey of more than 5,500 senior researchers,
Rowlands and Nicholas (2005, p23) found that almost 74% thought that high prices made it difficult to access the journal literature.
2.
Sparks (2005, pp26-28) appeared to report that almost half of the 750 researchers she surveyed reported having problems gaining access to the resources they needed for their research, with more than half in medical and biological sciences (52.5%) and arts and humanities (53.4%) reporting difficulties.
3. "All researchers appear to have similar levels of access to the journal materials they need. The issue of ease of access to journals shows little meaningful variation by discipline around 50% of all researchers, regardless of discipline, experience problems."
EPS and Oppenheim (2006)
"In regard to costs and benefits, there are a number of people looking into the issue of the relative costs and benefits of toll versus open access. Our own work in Australia is one example.
Houghton, J.W. Steele, C. & Sheehan, P.J. (2006)
"By way of triangulation, a 50% to 75% reported access difficulty correlates pretty well with the additional level of citation of OA articles reported by many studies (e.g. the open citation project, at
http://opcit.eprints.org/oacitation-biblio.html)."
Michael Kurtz, 4 April: "If, being conservative, I estimate that 33% of all desired article access do not occur due to access problems then:
Kurtz, et al, discussed the value of the interlinked system of electronic journals and indices in astronomy (which approximate a fully interlinked, local OA system) and found that it amounts to an efficiency increase in 6% in total research output.
"A one third reduction of this benefit, due to closed access, would be 2% of total research. If the total R&D budget of the world is a tera-dollar then 2% is $20 billion. For comparison the entire direct cost of publishing is normally estimated at a little less than 1%, or $10 billion.
"This says that the value of OA is twice the cost of the STM publishing industry."
Peter Banks, 4 April: "the question I am asking is not whether problems with access ever exist. Certainly they do. The question is, were those barriers to be removed, would we see a sudden surge of research, an improvement in clinical practice, or a rising tide of patient and public understanding?
"My sense is that the benefits of OA (often described as "vast" or "overwhelming") have been wildly exaggerated and the costs trivialized.
"In part, I look at this from a personal viewpoint as a person with a serious disease. In the US, much of the movement for open access on Capitol Hill (see the background on FRPAA, for example) has been couched in terms of benefits to patients"
Charles W. Bailey, Jr., 3 April: "Regarding laypersons' use of medical information ... Sharon Terry recounts her struggle to gain access to medical literature that might help her two children who suffer from pseudoxanthoma elasticum (PXE):
We spent hours copying articles from bound journals. But fees gate the research libraries of private medical schools. These fees became too costly for us to manage, and we needed to gain access to the material without paying for entry into the library each time. Sharon Terry, In the Public Interest: Open Access, College & Research Libraries News 66, no. 7 (2005): 522.
Peter Banks, 5 April: "I have nothing but the highest respect for Sharon Terry, who took extraordinary steps to understand and find treatment for her children's condition.
"Unlike Ms. Terry, many will never be able to read the primary literature even if it were freely accessible.
"I would think that rather than proposing to help patients by promoting OA, the NIH could do far more good by creating a patient-centered database of genetic disorders, with the latest information on incidence, etiology, diagnosis, and treatment--and, perhaps more importantly, a list of any clinical trials and leading investigators for the disease.
"OA seems neither an efficient not an effective way to help parents in situations like Ms. Terry's."
Tony McSean, 7 April: "The best estimate we have of OA costs' impact on medical research funding is that of Mark Wolpert of the Wellcome Trust who estimated that their programme would cost between one and two percent of grant funding. ... As Rick Anderson said, there is serious money involved and we need to have an evidence base that it produces the best value."
Alma Swan, 5 April:
> can access to scientific information accelerate research
"Yes, it can. Open access is essential for the optimal progress of research ...
more"
Joe Esposito, 5 April: "There is a fundamental misunderstanding of how the Internet works in this post. OA is contrasted with hardcopy and Web 1.0 applications. Everything that is listed here for OA can be done (and done better) with proprietary services."
Alma Swan, 7 April: "There is no misunderstanding, fundamental or otherwise. OA is contrasted with CA (closed access) on the Web (I must confess I hadn't given much thought to print: a rather quaint concept in the context of obtaining scientific information). Authors cannot cite articles they don't know about, they certainly cannot cite early any articles they don't know about, semantic technologies cannot get at closed access articles to work on them, proprietory services do not construct one research space, and researchers whose work requires them to reach out into other fields cannot find articles that are not in their library (and the concept of 'new fields' frequently means that their library does not provide the materials they need). Indeed, this whole area of servicing the demands of collaborative and pooled research is a major issue that research libraries are now having to start facing up to"
Joe Esposito, 7 April: "This is just wrong. The proposition that is being put forth here is between fully open access material and fully closed access material, a contrast that does not exist on the Internet, where materials are found through Web-exposed indexes, published keywords, linking arrangements, abstracts and summaries, hot-linked citations, and various (and evolving) algorithmic methods.
"This industry also has serious and ethical players (who are expensive), who redesign content for maximum Web "findability" and can greatly and legitimately increase the prominence of research materials. This costs money: it requires well-designed Web presences, superior marketing personnel, and extensive attention to Web analytics. ... commercial and well-run not-for-profit research publishers have always provided, continue to provide, and will provide in the future the best vehicle for researchers who publish.
"The basic problem here is the insistence that Web findability and access are somehow one and the same. It just ain't true. You can put up anything you want on a Web server and you may even get lucky and have Google and some other search engines index it. And this is where advocates of OA start and end the discussion."
Comment: Swan, on the one hand, arguing for an open access layer of research information, and Esposito and Banks on the other, promoting the case for better education and improved information through value-added publishing, can both be right. The latter does not obviate the case for the former; simply the former should now be seen as the fundamental basis on which to build the latter.
Imre Simon, 4 April: "We can build today computer programs to mine the literature and discover a lot of otherwise hidden structure and hidden clustering in the whole of the literature. This can only be done by computers, feeding them the full text of all articles. With today's fragmented access policies the said very rich researcher can access, as a human, the papers he or she wishes to read, but can't feed all the paper's full text in a computer program to discover the hidden structures and connections. Even he or she looses a wealth of information.
"Hence, even the richest researcher, working at the richest institution, needs universal access but no one has such a thing today. Open Access would surely be the best way to get universal access for everyone."
Concluding comment: The justification for the value of OA to
science, as used by Swan, is an important distinction, and quite different from the justifications typically offered, as we see here, by librarians (solving the 'serials crisis'), politicians (public, or taxpayer, access to e.g. medical information) or publishers (the economic case, from those publishers that offer some case for OA). This is besides the wider arguments for OA of democracy and equality of access, and the benefits this will bring for developing countries, etc., that are not really touched on here. These cases have their own validity, and variously inform (or mis-inform) this list thread, but won't appeal to scientists or researchers. The difference essentially lies in imagining or anticipating papers beyond the familiar article or journal packages and additionally placing them in the context of a linked information continuum, as Simon suggests. That this is massively important to the process of science and research is best exemplified in a 2001 viewpoint (overlooking the focus on central rather than institutional repositories, understandable for the time) that included a number of Nobel laureates among the author list: (Roberts
et al.,
Building A "GenBank" of the Published Literature,
Science) "the material that is freely accessible, on a controlled basis, one paper at a time, at a journal's Web site differs from material that is freely accessible in a ... comprehensive collection." Only OA can do this, and alone makes the case for OA irrefutable. The value contention cannot really be made without reference to institutional repositories and green OA; on the basis presented here the value case against OA is at worst specious and at best moot.
