Sunday, April 10, 2005

Roller Coaster Ride

In Part One of this two-part interview Dr. Alma Swan, co-founder and director of UK-based scholarly publishing consultancy Key Perspectives (KPL), discussed the findings of two recent KPL surveys on Open Access (OA). In this second part she explains why self-archiving should not be viewed as a threat to subscription-based journals, and discusses how scholarly publishing is likely to evolve in the future. Publishers, she concludes, should prepare themselves for a roller coaster ride.

RP: You've researched OA; you've thought a lot about it; and you've concluded that OA is optimal and inevitable [See Part One of the interview]. Does this mean that you expect to see all research made freely available on the Web?

AS: I do. At least, whatever research its creators wish to give away.

RP: As you explained, there are two ways in which researchers can embrace OA: they can use the so-called Gold Road, where they publish their papers in new-style OA journals, or the Green Road, where they continue publishing in traditional subscription-based journals, but then self-archive their papers. Which, in your view, is the optimal way of providing OA: green or gold?

AS: You ask this as though it’s an either/or situation, yet both methods co-exist at the moment, both are growing, and both have their place. For instance, some journals have already converted to gold, some are in transition, and there are now plenty of born-gold journals. But it will not be possible to successfully convert all existing journals to a gold model, at least not without major disruptions to the businesses of some publishers, and not in a hurry. For this reason alone the Green Road is also important.

RP: Going gold is a choice that publishers can make. Self-archiving, by contrast, is a choice researchers make. While 92% of publishers have given permission to self-archive, many seem to be increasingly concerned that if too many academics do self-archive their papers on the Web then libraries will stop subscribing to journals.

AS: True. Many publishers, particularly learned society publishers, express a great deal of concern about how self-archiving will detrimentally affect their business. Indeed, so great is their concern that they are imposing embargoes on self-archiving by their authors — trying to prevent them from self-archiving articles until six months after publication, or in some cases even 12 months. But these concerns are all based on speculative anxieties. I call it self-terrorising.

RP: In your most recent survey you specifically examined these speculative anxieties didn't you?

AS: Actually, it was not part of a survey. It was a separate exercise I carried out. But yes, I wanted to establish whether publishers' concerns over self-archiving have any foundation. To that end I asked learned societies in physics — specifically the American Physical Society (APS), and the Institute of Physics Publishing(IoPP) — about their experiences living alongside the physics e-print arXiv.

RP: Why physics?

AS: Because self-archiving has been taking place in physics since 1991, when the physics eprint repository arXiv was created. Indeed archiving in arXiv is now a long-established way of working for physicists in certain areas — high energy physics, condensed matter and astrophysics for example. For 14 years they have been depositing their preprints and, increasingly their postprints in arXiv, which now has around 300,000 articles in it. This means that each year about one third of all the published articles in physics are archived in arXiv.

RP: What does 14 years of arXiv tell us about the impact of self-archiving on traditional journals? Can the two things co-exist?

AS: It appears so. I asked both APS and IoPP what their experiences have been in carrying out their business activities alongside this massively successful example of a self-archiving repository. They both said that they had not noticed any greater downturn in subscriptions to those journals that cover the same sort of physics as arXiv than for any of their journals covering other areas of physics.

RP: So despite the fact that physicists are making their papers freely available on the Web, libraries are not cancelling their subscriptions to physics journals?

AS: Exactly. APS and IoPP could not ascribe any subscription cancellations to arXiv, and — importantly — neither of them said they considered arXiv to be a threat to their business. In fact, the APS has even helped to establish one of the arXiv mirror sites at the Brookhaven National Laboratory and both societies appear to consider arXiv a positive help to their publishing business. At the very least this suggests that present anxiety from other quarters is ill-founded.

RP: But can we conclude that the physics model is applicable to all other subjects, and that the impact of self-archiving will be equally benign elsewhere? Perhaps physics is a special case?

AS: Why should it not be as benign in other areas? Why assume something more complex than necessary when there is no evidence for the assumption? It is possible that physics is special, but we should not assume that it is unless there is good reason, and nobody’s made a convincing one to me yet.

RP: Maybe, but this is a serious matter for publishers: their businesses depend on libraries continuing to subscribe to their journals?

AS: Of course publishers must look to the future of their businesses, but it’s terribly unsatisfactory to plan a business on the basis of imagined scenarios. Yes, business schools teach scenario planning as a legitimate technique, and used in conjunction with other strategic planning techniques it has its place, but not as a primary methodology, and especially not when used alone.

RP: So they should quit the self-terrorising habit!

AS: They should. I simply can’t see the attraction of surmising that there is a worse case scenario without evidence suggesting that there might be. If evidence starts coming to light, we’ll monitor it, assess it and report it, but at the moment it’s just not there.

The point is that, OA aside, there are going to be huge changes in scholarly communications. This will be driven by digitisation rather than OA per se. It is likely that one consequence of these changes, for instance, is that the primary research journal as an entity may no longer be the answer to user needs.

RP: What other changes do you see going forward?

AS: As a result of the Internet, scholarly communication is entering an extremely exciting phase, with avenues opening up that have never been available before. In the print age, with toll-access [subscriptions-based access] dominating the scholarly communication arena, scholars were limited to communicating with each other in a one-to-some, non-responsive, severely-delayed way. Now we’re seeing new, additional opportunities for one-to-few or one-to-all (both options available), immediate, response-integrated communications.

RP: In other words much more complex, and presumably richer, forms of scholarly communication?

AS: Absolutely. In print, only textual or graphical material can be communicated. In the future, any digital output form can be accommodated. Scholars will be communicating in many different ways in the course of a week’s work — talking one-to-one, talking to the world via webcasts, talking to a selected audience via computer-conferencing, outputting material in digital text form, in digital graphical form, in digital audio or video form, in the form of terabyte-sized datasets or as one, critically-important statistic. They’ll be responding to one another’s output in real time, or as near as damn it, collaborating as never before, sharing data and views in ways not previously possible.

And I haven’t even started on all the opportunities that come from changing patterns of research globally, new research from new producers, the increasing scientific-literacy of the developing world and the way digital technologies are providing enhanced communications channels for its output.

RP: This will bring benefits right?

AS: Yes. Research will be the primary beneficiary, with increased efficiency, shortened cycles, less duplication and going over old ground, far less lost for all time in obscure, small-circulation publications. It will be possible to collate and integrate research output in ways we’ve never seen yet, and data mining and allied techniques will reveal new, unrealised relationships and connections. In time we will all experience the benefits.

RP: In what sort of timescale do you envisage these changes to scholarly communication taking place?

AS: I think we will see very dramatic changes, and probably pretty fast. It’s exciting. You need, by the way, to differentiate between scholarly communication and scholarly publishing. There are going to be all sorts of outputs from scholarly endeavour that form raw material to which a publisher can add value and then persuade someone else it is worth paying for. There is likely to be a plethora of new formats for research output, for instance, and new requirements from researchers to satisfy.

RP: In distinguishing between scholarly publishing and scholarly communication you are also signalling that many of these outputs could by-pass the publishers?

AS: Some forms of scholarly communication will be just that — scholar-to-scholar, without a third party involved. But others will require additional, expert, input.

RP: How can publishers and learned societies ensure they have a stable future?

AS: What’s a stable future? None of us have the luxury of that, do we? I don’t see why either commercial publishers or learned societies should not have a future, but stability is probably asking rather a lot unless we’re all going to move to command economies, which we’re not, Richard. In short, we’re in for a roller-coaster ride, and everything’s going to be in a state of flux for a long time yet.

But the good news is that all these changes will mean loads of opportunities for scholarly publishing too. It’s going to be an interesting time, the next couple of decades, and I believe there will continue to be a role for publishers.

RP: So it's going to be a scary new world — all the scarier since scholarly communication is currently in such a mess. Indeed, many argue that publishers brought it all on themselves. Why do you think scholarly publishing is in such a mess?

AS: Scary? Exciting and challenging, yes. Not too scary, though. There’s a bit of a mess at the moment because the whole system is completely bizarre. First, it’s not a perfect market, so no journal or article can ever be a true substitute for another.

Second, the users of the items purchased are not the purchasers, so the people who have to make the purchasing decisions are constantly trying to match these decisions with the interests of the ultimate users, which is not an easy task.

Third, the ultimate consumers are also the raw material suppliers, and they are extremely unusual in any capitalist market situation in that they happily supply their raw material for free — and almost as happily pay to have it back again to use later.

And fourth, — to confound things just a bit more – these same raw material suppliers/consumers also carry out one of the stages of production (peer review). Add to these ingredients the fact that it is a global enterprise, so there is little room for tailoring to local economic conditions, and you have all the ingredients of a mess.

RP: What then should publishers do about OA?

AS: Publishers will need to define their roles with a great deal of thought. They have vast opportunities here. Their roles will change, I’ve no doubt. There will be blurring of the boundaries between traditional publishing models — between, for instance, primary publishing, secondary publishing, and tertiary publishing (where massive value is added in the form of additional content types).

RP: I'm conscious that most discussion about the threats that OA poses is focused on primary publishers, but there are presumably implications for secondary publishers like ISI too, particularly in a world where OA and Google Scholar look set to become dominant. Is ISI under threat for instance?

AS: You said ‘implications’ and then changed to ‘threat’. There certainly are implications, but I would think in an OA world they should be viewed mainly as opportunities not threats. OA material needs to be searched for and retrieved just as effectively as non-OA material so there’s plenty for ISI and the like to be getting their teeth into if there’s something there that will enhance their business.

RP: But what about Google Scholar? Effectively this will be offering what ISI does, but for free?

AS: Google Scholar doesn’t do anything like ISI or Scopus do. It’s just a technology exploiting the ‘good enough’ tendency of the average researcher. At the moment it’s pretty poor, judging from my own little experiments using it, although that will no doubt change. I’m also prepared to buy the argument that these new kinds of search technologies will grow ever more sophisticated, but it will be a long time before they have the functionality of one of today’s good indexing services.

RP: And Elsevier has spent its investment dollars wisely on Scopus has it?

AS: Only Elsevier’s management knows how that question can be answered thus far. At first I was surprised at it, but after some reflection I saw the logic, and a lot of dollars have gone into it, so Elsevier is clearly serious about it. There are some good business reasons for having such a service and it will be interesting to watch how it is employed strategically. It’s certainly created a new battlefront in the secondary publishing theatre of war.

RP: We've talked about primary publishers, we've talked about secondary publishers. But what learned societies? Many believe that it is learned societies who are most threatened by OA. Do you agree?

AS: You differentiate between commercial and society publishers. Commercial publishers will continue to do what every business needs to do — analyse where the value lies in the business and concentrate on developing that. And since many large society publishers operate in ways that are to all intents and purposes indistinguishable from their commercial counterparts they will do the same. They have the same sort of structures, operational procedures and overheads.

RP: How do you mean?

AS: A commercial publisher may call us in, tell us what they want us to try to achieve and say, “But remember, whatever you end up recommending we must have a bottom-line figure of 25%”. A society publisher may call us in and say, “We must have a 25% surplus from our publishing operation.” Where exactly is the difference? They both need a strategy that produces 25% on the bottom line. Both are working under the same imperative — a financial return for the owners of the business — and both will have to adopt the same sort of tactics in the marketplace to achieve the result.

RP: So there is no difference between a commercial publisher and a society in terms of publishing?

AS: Well, some society publishers operate under quite different sets of requirements and aims. In a sense they are freed up to do much more creative and far-reaching things. These are the ones that can take a look at their publishing operations and say “how can we go about facilitating the best scholarly communication within the subject area we represent?”

The answer to that so far has been to publish journals, but there may be different answers in the future. There is room for immense creativity in this regard. Peer review will always be with us, for instance, and who better to mediate it than a learned society? As the spokesman for the American Physical Society said to me, “As long as peer review is valued by the community, we will be doing peer review.”

RP: This assumes that peer review will continue to be valued, but does not the future you paint have implications for peer review too?

AS: No. There will always be a need for a peer review system and where there is a cost involved this will need to be paid for each article.

RP: I find it interesting that people in the OA movement will argue about most things, but (with a few exceptions) most seem to agree on the need for traditional peer review to be preserved. At the same time everyone agrees that peer review is far from perfect. Could we not hope that the coming changes will improve on peer review too?

AS: It's true that peer review is not perfect now, but like Winston Churchill said of democracy, it’s the best we’ve come up with so far. Moreover, there is potential for new ways of going about peer review that will improve it.

It may change form for instance. In fact, it’s almost inevitable that it will. Digital communications mean that peer review can now be not just pre-publication but can usefully continue after publication. It will become part of a new system of assessment for research and as such will improve. However, it will play the same role that it has always played, which is to stamp a mark of acceptable ‘quality’ on an article.

RP: I suspect the issue that may prove most controversial is copyright. Today many publishers remain unconvinced about OA and view copyright as a tool for resisting change. Even those who are sympathetic to OA clearly hope that it will enable them to retain control of scholarly communication — allowing them, for instance, to insist on embargoed self-archiving. Given this, I'm surprised that OA advocates frequently dismiss copyright as irrelevant. What's your take on copyright?

AS: OA proponents state that self-archiving the final author-version of an article does not contravene copyright in any way, and this view appears to be supported by the copyright experts. So in most instances where the content of an article is generated solely by the author copyright cannot be used to stop that author from self-archiving.

That said, I do think copyright is an important issue. The fact is that copyright raises its head all the time when authors are asked about OA and it is acting as a deterrent to self-archiving. So it can’t be ignored.

RP: Is it mainly an education issue then?

AS: More than that I think. There is talk now, for instance, of the big research funders stipulating that the authors they fund retain copyright on their published articles themselves. And institutions are taking a great interest in this from their own intellectual property point of view. There is, after all, the longstanding example of industry-funded research, where commercial companies have always insisted that copyright is not signed over to a publisher.

RP: This would have implications for the OA movement wouldn’t it?

AS: Right, because ownership of copyright still remains at issue. There are also many instances where copyright becomes a major concern even without OA entering the scene. People publishing articles on the history of art, for example, need to seek out copyright to publish images they themselves have not generated, such as a painting by Michelangelo — or a painting by a living artist with his own copyright on an image, for that matter.

RP: This has always been an issue in publishing hasn’t it?

AS: Sure, it can be a tortuous procedure even in the traditional publishing process. The added complication in an OA environment is that the additional work required to get all the permissions necessary to ensure a self-archived article could carry such an image adds further complications.

I’m not saying it can’t be done, and I’m not trying to throw a spanner in the works. I’m just pointing out that there are situations where the complexity can be quite marked — and the OA movement needs to take this on board.

RP: Wherever you look, of course, copyright is contentious today!

AS: It is. And we can also see all the signs that copyright in general is about to undergo change. The Creative Commons movement is setting the pace, and the Science Commons has also got off the ground now too.

Coupled with this there is a general shift in attitudes — in western societies at least — towards information sharing and increased and unfettered access to information. All this suggests, to me, that the days of publisher-owned copyright are numbered. That is perhaps the most important point concerning copyright so far as scholarly communication is concerned.

RP: Governments clearly have a role to play in adjusting copyright laws to meet the needs of the networked economy too. But what do you think governments — and indeed other research funders — should be doing to help address the current problems of scholarly publishing? Should they be doing anything?

AS: They should, because there is a big public-interest issue here. Funders have the right and responsibility to ensure that access to the work they fund is available to everyone they think can benefit from it. The same applies to governments, but governments should only get involved in the most minimal way.

RP: How do mean: minimal?

AS: I mean that if research is funded by a government (i.e. the public) then that government has the right and duty to make sure the results from it are available to all. Governments shouldn’t have any role, however, in dictating how commercial publishers operate in a free market economy. And neither party needs to start interfering with a researcher’s right to choose which journal they publish in.

RP: So what should governments do specifically?

AS: If they want the research to be freely available all they need to do is require researchers funded by their money to self-archive. That achieves access for all, whilst allowing the process of formally publishing the work to go on in the way researchers choose.

RP: My final question then: Reed Elsevier's CEO Sir Crispin Davis implied — if not directly stated — in The Guardian recently that the simple answer to solving the scholarly communication crisis is for governments just to pump more money into libraries? Does he have a point?

AS: No.

RP: Thank you for your time.

If you would like to comment on this interview please e-mail your views to me at, or to comment publicly hit the comment button below.

No comments: