At the last annual meeting of the Society, there was a suggestion that the Society start an electronic journal. The current article is a mildly edited version of a response, requested by Council, to this suggestion. Council endorsed the major recommendations of the report (see the last section) and an advertisement requesting expressions of interest in the Editorship will appear in a later issue of the Gazette .
There appear to be two major reasons for starting an electronic journal.
The first is to make more space available for the publication of mathematics. Both Series A of the Journal and the Bulletin have more good papers than they can publish; Series B of the Journal appears to be reaching this state. Series A has even been compelled to turn down opportunities for special issues which would undoubtedly have contributed significantly to its overall quality and reputation. It seems that there is a demand for more publication space and that there will be a ready source of papers of good quality.
The second reason for starting an electronic journal is that this is likely to be the way of the future. There is thus a case for establishing ourselves in the `marketplace' early and attempting to get useful experience and to establish a good reputation before many others enter the market.
We do not believe that it will be possible to make income from the journal in the first few years of its operation. It may even be difficult to make income in the longer run. The problems of making income, or surviving without, are clearly going to be faced by all academic journals, even those run primarily for a profit. We are not aware that any of the electronic journals currently being operated by mathematicians are charging anything for their services. It is clearly important, however, to keep the possibility of future charging well to the fore.
It may be possible to charge for paper copies. It seems that the best way to offer the electronic journal, however, will be to make either the Tex source or a graphic source, such as a dvi file, of the paper freely available. There is then little reason for the reader to ask for a paper copy. It is likely that the only substantial market would come from underdeveloped countries where the relevant technology is not freely available. While we would expect such readers, at least, to meet the real cost of delivery to them we would not want to make them finance our journal.
It is feasible that we could charge in a way similar to that used by Mathematical Reviews and Zentralblatt . That is, offer both paper and electronic versions, charging a standard (relatively high) price for one of these. A much smaller price, sufficient to cover the basic cost of the paper/disc, is then charged for the second.
It is possible that we could make only a part of the source available and make the rest only available on paper after suitable payment. This is not in the current spirit of academic electronic journals and we feel it is liable to meet with substantial resistance (and result in the journal being effectively ignored).
Another possibility is to make page charges. It is a mechanism used by some paper journals and the charges made could be quite modest. It could however be quite helpful if there are real copy editing costs.
If it is decided not to make serious attempts to charge for the journal then much will be thrown back on the Editor. There will presumably have to be some discussion as to the level of letter-writing, record keeping, copy-editing, Tex correction and proof reading that the Editor ( and editorial assistants) are prepared to do. Council could set limits on what financial support, if any, it regards as reasonable in the light of what we currently spend.
It is not impossible that the home institution of the Editor might also contribute to the running of the journal in return for some recognition of its contribution.
Since it will be possible to start an electronic journal with few resources, it seems likely that there will be numerous electronic journals in the future. Undoubtedly some of these will be of poor academic standard.
On the other hand there are already electronic journals of good academic quality. The Society has an excellent name for its publications and should use that to take a lead in setting standards. In particular, it must ensure that it maintains standards at least at the level of the recent past on which we built our reputation. We should not allow the ease and cheapness of publication to lower the standard desirable for acceptance.
There is evidence that university administrations are willing to consider academic publications on their merits. Inquiries were made of senior administrators at the University of Melbourne and at ANU. The University of Melbourne is now considering the issue and it appears likely that it will be prepared to accept electronic journals on their academic merits. The Institute for Scientific Information, which publishes the Science Citation Index, is beginning to list electronic journals.
While the existence of electronic journals is likely to increase the problem of judging the quality of publication lists, there will be journals of high academic quality amongst them and these will be taken seriously by the academic community.
Archiving appears to be a serious problem for any electronic medium that aspires to normal academic standards of permanence (perhaps 200 years?). There is no certainty that current media will have a long physical life and a considerable likelihood that current media will become obsolete well within the life-span at which we should be aiming. Thus it seems necessary that any papers accepted for publication should be preserved by a paper copy as well as any electronic archiving that is attempted. This paper copy should be supplied, or at least approved, by the author and at least one copy should be made available to a standard archiving source such as a university library. (This assumes that a library can be persuaded to accept these copies.)
It will not be easy to find an appropriate and willing candidate. The ideal candidate will have both substantial academic `weight' as well as great energy and enthusiasm for all the matters of detail that need to be addressed as well as the minor problems that will need to be solved. If there is no initial attempt to charge for the journal, then it is likely that the Editor will have to do substantial copy-editing and \TeX correction on the submitted documents. A good style file which is firmly administered will help to reduce some of this work. In the initial stages there will be a learning curve which may need to be paid for by the Society.
We suggest that the Council establish some guidelines in conjunction with the chosen candidate but then leave the Editor a large amount of leeway to make day-to-day decisions. There will be many more such decisions at the beginning of this Editorship than the normal beginning Editor faces.
There are a number of aspects of this which will differ from the standard paper journal.
Firstly, we recommend that the journal proceed by establishing a precise style file and insisting that all authors closely follow this style file when submitting their papers. This should substantially reduce the amount of copy-editing.
Distribution should be effected by making the journal available on the Internet, presumably via (a link to) the WWW page of the Society. Distribution could also be done automatically by email on request. A similar system is in effect for the Electronic Journal of Combinatorics.
There is an important question of whether the Council wishes the journal to be made available by publishing the Tex source or, alternatively, by publishing only a graphic source such as a dvi file or a Postscript file. The former has many practical advantages such as easy access and the smaller size of file that needs to be transmitted. It does, however, make possible mischievous tampering with the source and may lead to substantial abuses. While it seems unlikely that we could successfully charge for either system of delivery, there seem to be more options available if we retain the Tex source as confidential.
There will be need for substantial initial advertising. It is unreasonable to expect authors of quality papers to submit them to a journal which is largely unknown. This should be both by widespread electronic advertising and also advertising as widely as possible in standard academic media.
In summary, we believe that the Society should begin the process of starting a new electronic journal. We also recommend the following:
If the major part of these recommendations is adopted, we recommend the appointment of the Editor proceed as soon as possible (but with the unusual importance of the choice firmly in mind).
The appointment, together with an agreed financial allocation, could then be decided by a postal ballot of Council. We also recommend that the Editor be given very substantial freedom to formulate the details of the operation of the Journal. The Editor would, of course, have to return to Council for matters which might involve substantial policy decisions or substantial changes in the agreed expenditure.
University of Melbourne
Australian National University