The group was mainly composed of people with publishing responsibilities but most of them were scientists and many had affiliations with learned societies. The group concentrated its discussions on the costs of primary publishing in print form in serials, on the costs incurred in making the print journal electronically available and on the savings resulting from the same process.
Secondary publishing, publishing in non-serials and e-journals (journals with no print equivalent) with their special costs were only discussed tangentially.
The group did not like the term 'costs' and preferred investment or economics as the circumstances of the discussion allowed but this preference was not consistently adhered to.
Two publishers, Blackwells Science (Mr. Campbell) and the American Physical Society (Dr. Lustig) shared with the group their rather different analyses of the costs of producing a journals. Mr. Campbell's paper is to be published in LOGOS (Robert Campbell and David Russon : Access to a Journal, LOGOS, NYP) and Dr. Lustig has kindly agreed that some of his conclusions should be presented as an appendix to this report.
What became clear was that many of the publishers present did not accept the validity of either of the ways in which costs were presented and many did not agree that it was possible to separate out the costs in this way because they represented only an arbitrary slice at a particular point in time of the investment needed to transform the whole publishing process in line with the sort of developments which were discussed at the Conference. Some also felt strongly that it was impossible to compare print and electronic publishing in this way when their thinking was in terms of creating a common electronic platform which would enable product to be run off in a number of different formats.
The only area of agreement was the high percentage of fixed costs (in the range of 70 to 80%) which are incurred before any copies of any journal are printed or distributed.
Because assumptions about costs are so central to the debate about the impact of electronic publishing particularly on the libraries, resolution III.3 was submitted.
It was easier to come to a general consensus about the costs of authors and the costs of libraries though discussion about libraries was handicapped by the lack of librarians in the group for much of the time and thus was of necessity sketchy.
It was recognised that the economics of the whole STM publishing process is underpinned by the willingness of the academic community to work for less than consultancy fees and often for nothing. These hidden costs were not explored. There is no reason to suppose this the situation as existing will change with e-publishing.
There was a long discussion about direct costs to authors of papers particularly because of the model presented by Professor Lederberg among others under the terms of which the author pays for the cost of presenting a paper while the reader receives the paper free. This model did not find favour with the group.
The perception was that the culture of authors paying to publish through page charges or submission charges is neither in good shape in the USA where it has been common and has never been the preferred way of operating outside that country. In addition there was a strong resistance to what was perceived as a tendency for publishers to push back to the authors to costs of formatting electronic files.
Nevertheless it was felt that, as the dissemination of the results of research is an intrinsic part of the research itself, a resolution recommending the extension throughout the world of the practice common among government and non-governmental funding agencies in the USA of taking into account publication costs in research funding could appropriately be put forward.[III.1]
It was also the view of the group that investment by the publishing community in electronic dissemination was pointless if the advantages of electronic access cannot be realised because the journals cannot be brought to the desk of the individual researcher due to a lack of internal networking and training. It was felt that funding should be earmarked specially for the building up of an appropriate infrastructure of this sort within the context of library provision.[III.2]
The group was also concerned with the infrastructure in a wider sense with particular concentration on concerns that the scientific community will not have available sufficient band-width at a lower price for communication over the Internet to be made easier and more effective rather than becoming restricted and more expensive. The concerns of this group are partly reflected by a resolution derived from the deliberations of another group.[IV.2]
New electronic technologies are continuing to alter the way science is done, the way science is taught and the way science is applied to the various aspects of our lives. There is little surprise, therefore, that these same technologies are continuing to alter how we document the record of scientific achievement and how we give access to that record.
The Working Group devoted a considerable amount of time to a number of perspectives on the electronic archive. While there is general agreement on the need to insure that a permanent record of scientific achievement will be maintained, there were varying levels of agreement on the nature of the electronic archive; what its purposes are; how accessible it should be; who is responsible for maintaining it and who will finance it?
There is some difference of opinion as to whether an electronic archive should be a "static" collection (or network of collections) as distinct from an active record with assurances of the preservation of the file. There was even less agreement as to who should be responsible for ensuring the preservation of the scientific record. In some instances the scientific society is most likely to assume responsibility. In others, certain publishers will assume it. However, at some point, especially in developing countries, government responsibility for assuring that provision is made seems appropriate.
There was general agreement that authors, scientific publishers, learned societies and governments share some responsibility for financing maintenance of the record. The specific provisions are likely to vary among the countries.
There are some strong national interests in maintaining the record of science, but the long term interests are clearly international whether viewed by scientists, users, publishers, societies or governments.
Further discussion focused on what material should be eligible for the electronic archive and whether legal deposit carries some assurance that intellectual property rights will be respected. There was moderate agreement that only "published" material should be eligible for an electronic archive. However, this view held, primarily due to lack of a clear concept of what types of "unpublished" material might be significant additions to the record.
The concern for restricting access to the record established by legal deposit was challenged by the need for scientists to have relatively free access to the record. The scientists also insist that the electronic record be operable to enable them to navigate between their current work and the electronic record quickly and easily.
There was consensus that "one" central record of scientific achievement was not feasible, but that a linked system of many repositories is more likely.
What was clear from the initial discussions in the Group was that our concept of an electronic archive is based primarily on our current experience rather than a more ambitious vision of what scientists and those who benefit from scientific achievements will need from an electronic record.
At this stage the discussion has focused on the promulgation of standards as the key strategy for stimulating the development of electronic archiving of science. Standards for eligibility, standards for maintenance and standards for access were all discussed at length, along with standards for the structure and content of electronic archives.
What little discussion of costs occurred was overlaid with considerable scepticism that governments would be prepared to shoulder much of the burden.
Twenty-six people participated in this working group. Included were academics and publishers from three developing regions of the world - Asia (India and Nepal), Latin America (Brazil and Mexico), the Middle East (Lebanon and Turkey), and Sub-Saharan Africa (Cote d'Ivoire, Malawi, and Uganda). In addition, three European publishers attended the session - one from Elsevier, one from Chapman and Hall, and one from Oxford University Press. And there were representatives of several international organisations, universities, and government agencies.
We were a diverse group, to say the least, and the same can be said for the term the "developing world". Before even beginning to discuss the six basic questions laid out below, the group decided that developing countries cannot be lumped together as a single entity because political and infrastructure conditions vary within regions and also from region to region. In addition, circumstances can diverge considerably within countries.
Participants accepted the following definition for "electronic publishing": Electronic publishing is understood here in the broader sense (comprising production of scientific literature, data banks, indices and other information relevant to science and technology). During discussion early on, we also agreed to extend this definition to include electronic communications because e-mail is often the first form of Internet activity in developing countries.
The six questions that the group addressed were as follows:
1. Who is responsible for producing electronic publications in the developing world?
2. How can electronic publishing be used to improve the visibility of developing countries' science?
3. Who is responsible for extending networks in developing countries?
4. What technologies are advisable for electronic publishing in developing countries?
5. Will electronic publications in the long run be less costly (more accessible) for the end-user in these countries than print publications and what strategies can be adopted to this end?
6. What is the role of international cooperation?
The first half of the discussion concentrated on the first three questions because they were interrelated in nature. The responsibility for producing scientific literature (print or electronic) rests in the hands of the local scientific communities and institutions.
It was pointed out with concern, however, that scientific publishing in many developing countries is of a very fragile nature. Scientists often prefer to publish in North American or European journals rather than in journals of their own country or region. Often academic and research institutions in developing countries value these journals more highly, as do many scientists. There is a perception that "local" journals are not as good. As a result, changing the medium-from print to online-will not improve the situation of endogenous publishing unless the scientific communities of developing countries re-evaluate and strengthen their own publications. Governments can play a role, as well. In both Mexico and Brazil, for instance, the Councils for Science and Technology have set up committees to establish criteria and evaluate local journals.
A related problem is the fact that scientific research in developing countries is often unknown to the rest of the world and sometimes not even circulated outside of the institution where it is being carried out. It does not appear in the international indices nor are there adequate local indexing and abstraction services to fill the gaps. So what's to be done? One task is to use the new technologies to the maximum extent possible to produce appropriate indices and databases of developing country research information, and also take advantage of specialised international indices, such as Medline and CAB Abstracts.
And, just as developing countries need to create their own databases and electronic products, they also need to take more control over the technology itself. Successful software development in India, Mexico, and Brazil was mentioned as an example in this regard. Thus, there was consensus among the participants that electronic publishing can ameliorate many of the problems sketched out above-for those publishers who already have the technology at hand, the capacity, and the material to publish, and for those users who are already plugged in. Making information available online is a process, we agreed, which starts informally with e-mail communications among colleagues.
Governments play different roles in different countries. Everyone agreed that they have major responsibilities for infrastructure strengthening as well as promulgating regulations and tariffs that are favourable to connectivity. Although some governments are more sensitive to the need to implement Internet connectivity, others place Internet at a lower priority. It is up to the local communities-the users and producers of information-to push for an enabling environment. This includes the private sector. Moreover, this "enabling environment" is more than laying lines or installing computers, it also encompasses training end-users how to use the technology to its full extent.
An important point that was raised at this juncture is that dependence on aid programs is neither feasible nor desirable. Developing countries must take on these responsibilities for themselves. Large sums of money are not required. Implementing and using electronic connectivity can be accomplished without huge outlays of funding, but only if there is sufficient organisation and planning. Sharing costs among institutions for high-speed links may be cheaper and more efficient, for example, than for each person to have his own dial-up links.
Moving on to question 4, participants agreed that no one medium is sufficient unto itself. A mix of technologies - print, CD-ROM, e-mail, and interactive Internet - should be employed, depending on the need of the user, the nature of the product, and local infra-structure conditions. Local publishers and information sources should begin digitising their information as quickly as possible, in whatever form is most appropriate, but taking into consideration the need to observe standards and to ensure convertibility of databases. In addition, we hope that electronic publishers in the North will be sensitive to the needs of developing country scientists who cannot access online journals and consider production of their digitised information on compact disc as well. We want to emphasise that CD-ROM is not a dying technology; it is used in academic libraries in North America and Europe, and is very appropriate to developing countries world-wide. Mexico, Brazil, Zimbabwe, India and South Africa, to name only five countries, are active CD-ROM producers.
Participants expressed concern over the cost of electronic publications-if indeed they will be more expensive than the print versions. It's too early to draw any conclusions.
Before closing, participants briefly discussed the importance of regional cooperation to strengthen capacity and promote sustainability. Along these lines, we all agreed that local expertise was more desirable than bringing in outside consultants. In addition, participants reviewed one or two areas that might be appropriate for intervention by international organisations - among them persuading more publishers to institute discounted prices for developing countries.
The first matter discussed was the extent of the diversity of information-gathering methods of scientists? Each scientific discipline has its particular information gathering and dissemination methods, and the needs of scientists will also vary according to other factors such as level of responsibility, place of work or language. A range of basic skills common to most scientists can nevertheless be identified. In particular good authoring skills, adapted to the electronic environment, will be required to rapid and efficient dissemination of information through electronic publishing techniques. In addition, supporting staff such as technicians, network specialists and managers should receive special training so that they can contribute to the development of scientific electronic publishing.
This was followed by considering what skills would be required for scientists to take full advantage of electronic publishing? The skills for accessing and effectively using electronic publications should be seen as part of a broader required competency in information retrieval techniques and computer and network literacy. Additional skills are needed for preparation of electronic articles and for the organisation of mechanisms for their dissemination.
All scientists should receive training in the above areas. Since students can learn these skills more quickly and naturally than scientists who have already acquired work habits, training in information resources and library use should be a part of all scientific education, if possible as early as the undergraduate level. This training should stress characterisation of and access to scientific information including theory and practice of retrieval techniques, as well as general authorship skills including proper structuring of information for publication. Basic training should also aim at providing an understanding of the principles, main activities and tools (web servers, compression techniques, etc.) required for the electronic production and dissemination of information, but need not include detailed practice in this area.
All institutions of higher learning in science and all major employers of scientists should ensure that their students (and employees) receive appropriate training in use of information resources. Such training should take place so far as possible in a discipline or employment specific context, and should take full account of existing training opportunities including commercial offers. Information training activity should be co-ordinated at the level of each organisation, and libraries are often seen as the most appropriate institutional setting for this function.
The group then considered what are likely new developments and will they require training. It is clear that developments in electronic publishing will increasingly influence the work of scientists. The field of electronic information services, and of electronic publishing in particular, is, however, evolving so rapidly that it is probably not fruitful to try to anticipate specific developments. The use of specific tools and techniques can be taught as required, or even self-learned, once scientists have acquired a basic information culture.
The question was asked: Should libraries provide dedicated technical support for aspects of electronic information? The information resources required in and for science are becoming ever more numerous and complex, with the result that intermediaries will probably always be needed even as working scientists take more responsibility for gathering and dissemination of information. Other types of technical support are also needed to ensure effective access to and use of information - for example advice on the development and interconnection of computer networks.
Operational and methodological support in the handling of electronic information should be organised for individual scientists and scientific institutions. Libraries, which are increasingly assuming gateway and publishing functions in addition to their traditional roles in classifying, storing and disseminating documentation. should help fulfil this role. This implies a new type of training for information professionals to be able to perform this function.
In considering what additional factors can facilitate access to and dissemination of electronic publications, it was agreed that adequate and reasonably priced network access is essential for scientific work today. Although cheap, high capacity data communication is predicted for the future, there is a concern that available band-width will be insufficient in the medium term to take care of scientific information needs, particularly in the context of competition from non-scientific network applications. One, technically feasible, solution would be for national authorities to ensure priority network access for bona fide scientific usage.
The introduction of a common system of meta-information to identify scientific information resources would not only facilitate access and retrieval, but would also reduce demands on network and server capacity caused by repetitive, "brute force" inventorying of electronic information resources. Cooperative research by the scientific and library/information communities will be essential in finding appropriate cooperative solutions to the organisation of electronic scientific information.
If current trends continue, it may also become necessary for scientific communities to go further by limiting access to their servers by general network probing programs.
The final topic addressed was the steps that should be taken by the international community to foster improved use of resources. The various scientific communities will increasingly be called upon to take more responsibility in the information transfer process, including some functions which have been traditionally the responsibility of publishers. The international scientific community should facilitate exchange of experience between scientists and scientific disciplines concerning electronic publishing applications, with a view to identifying best practice and encouraging cooperative research and development. This work could be facilitated by establishment of an international committee on electronic publishing by ICSU or one of its bodies.
A useful first step would be the establishment of an information network to refer users to information resources, tools and standards relating to electronic publishing in science. Top level access to this information should be incorporated into the ICSU World-wide Web site.
There was general acceptance of the desirability of peer review - largely on the basis of the print-on-paper model, although the point was freely made that full advantage should be taken of the facilities for the presentation of sound, film and computer modelling. It was accepted that developing countries would be disadvantaged mutually, but expected that they could leap-frog into the future.
The ways of establishing priority of publication were touched upon, and it seems that the requirements differ from science to science. In high-energy physics, priority may be established by a comment in a public conference (Ginsparg), but in other sciences, where interpersonal competition may be greater, stricter criteria prevail.
Quality control was regarded as crucial. This covered the need for scientific validation, good copy editing and the deposition of the definitive version of a paper. Here there could be a bonus in the electronic world by tracking of the way in which thoughts were expressed and concepts established.
It was accepted that theories and the interpretation of results can change, but the raw data should be reliable and
unquestioned. The tracking can be done by adding tags or links to subsequent revisions. It was even suggested that the changes made during revision of a manuscript should be flagged and made available to the public, with Paul Ginsparg arguing that a lot of useful material was generated in the path between referee and author and that it was a pity to lose it.
Sydney Hall noted that an audit trail had been included in the CIF crystallography data base. This tracked any changes made to the information stored and had proved extremely valuable. It was also pointed out that some means of electronically date-stamping electronic manuscripts was needed, although it would be difficult to do this in a way that could not be subsequently altered.
The quality control and validation of some databases also requires attention, for it is desirable to check the validity of the material deposited within it, and to prevent subsequent manipulation of the information. It was clear that it is harder to avoid the doctoring or amendment of digital material than with printed documents.
Another suggestion (Kircz) (not universally welcomed) was that papers could be broken up into electronic modules, and each module or component being independently refereed. This could stiffen the refereeing process (making it more rigorous), but would greatly increase the burden of the refereeing process.
Discussion then veered into an exploration of the definition of "publication". Could presentation of information or observation on web home pages be regarded as publication? Yes.
The fact that some journals refuse to consider papers describing findings previously released to the New York Times was alluded to, and led to consideration of what constituted public dissemination of information. It was decided to make no recommendation whether or not such newspaper disclosure should disqualify scientific material from journal publication?
St John of John Wiley pointed to the possible difficulty in assessing papers based on data-sets that were not accessible to referees. How should they react?
The value of moderated discussions also came up. Should they be preserved? Ought they to be reviewed and edited? Who might do this? And could the possible legal complications for the publishers (who would be accorded responsibility for libellous statements) be avoided. Again, there was no agreement on these matters.
Finally, one of our Japanese colleagues enquired how the quality of papers in an electronic database can be assessed by one not familiar with the journals represented. One way could be for learned societies to attach their logo to the journal title as a mark of authority. Another could be to use an R in a circle, rather like the copyright symbol, to indicate that the paper has been properly refereed. This was thought to be a useful suggestion.
Listing by Author's Name, in alphabetical order
Return to the ICSU Press/UNESCO Conference Programme Homepage
University of Illinois at Urbana-Champaign
The Library of the University of Illinois at Urbana-Champaign
Comments to: Tim Cole