A European Mathematical Database J.H. Coates Chairman Database Committee European Mathematical Society Europe has been the cradle of mathematical research and publication over the last four hundred years. However, Europe's historical role in both publication and research is being threatened on several fronts. In particular, new electronic technology is currently bringing about a profound revolution in the communication and publication of mathematical research. While it is still impossible to foresee at present the full ramifications of this electronic revolution, it is already clear that there is now both the need and the technology to create a comprehensive database of all mathematical publications, which is centred in Europe and owned by the European mathematical community via the European Mathematical Society. Such a European database would be an invaluable research tool for future generations of mathematicians around the world. In addition, it would provide both competition and different traditions to databases emanating from North American sources. The present article sets out both the background and plans for achieving this goal, via an evolution of Zentralblatt f"ur Mathematik und ihre Grenzgebiete. Why action is needed now. Firstly, the current state of technology now makes it feasible to build comprehensive electronic databases, with efficient accessing procedures. In parallel, there is the fact that many journals are now published with an electronic version, and consequently it is much easier to transfer data about the contents of these journals to a central database. It should also be stressed that the vital commercial interests of both private companies and mathematical societies need in no way be compromised by transferring limited information about the contents of their journals (e.g. tables of contents and summaries of articles) to a central database at the time of publication. It may even be possible to eventually persuade many publishers to provide the database with complete copies of their journals several years after their initial publication. Secondly, it is self evident that the ownership of all major databases in mathematics should be clearly vested in the community of mathematicians which produced much of the mathematics in them in the first place. North American mathematicians already in effect own such a database via the ownership of Mathematical Reviews by the American Mathematical Society. Until such time as the European Mathematical Society succeeds in establishing a comprehensive database of its own, there is a danger of the vacuum in Europe being filled by the American Mathematical Society dominating databases on the world scene. This scenario would ultimately lead to a downgrading of Europe's place in international mathematical research. Finally, all mathematicians are aware of the explosive growth in the numbers of both journals and less formal forms of publications which has come with the electronic revolution. In many ways, the cosy traditional world which existed up until the early 1980's, in which a mathematician could keep track of most publications in his field by scanning the contents of a fairly stable list of journals, many of them of long pedigree, on the shelves of his or her university library, is now being profoundly modified. If future generations are to maintain the great mathematical practice of citing in research papers all closely related earlier literature, it seems that this will only be feasible by the systematic use of a comprehensive database. Short term plans. The European mathematical community is fortunate in already possessing the beginnings of an excellent database. Zentralblatt f"ur Mathematik exists both as a conventional printed reviewing journal, and as the electronic database MATH covering all Zentralblatt back issues until 1931, which is available either on CD-ROM or via WWW-access. Zentralblatt is currently run by the Heidelberger Akademie der Wissenschaften, and published by FIZ Karlsruhe, and Springer Verlag. These bodies have invited the European Mathematical Society to collaborate with them to ensure the future evolution of Zentralblatt into a database second to none in the world, which will become an everyday tool for the working mathematician. The European Mathematical Society has willingly accepted this invitation, with the firm understanding that the ownership of this database should in the long term be placed in the hands of the mathematical community. As a first step in the development of this database, the European Mathematical Society has already taken several concrete steps to encourage a more widespread electronic use of Zentralblatt as part of its electronic information service EMIS. Indeed, EMIS now provides a service by which any user, irrespective of whether or not he or she is a subscriber to MATH, or a member of the European Mathematical Society, can carry out limited searches in MATH. Secondly, the Society has just launched a new facility on EMIS called CAP-EMS (Current Awareness of Mathematical Publications). In this new service, publishers will automatically provide the basic data needed by Zentralblatt from the electronic files used for the publication of their journals. This data will consist of tables of contents, together with abstracts when available (or in some cases the first page of each article). It will be freely available for all to consult in EMIS for one year from its date of deposit. At the same time, the data sent to CAP-EMS will be stored permanently in MATH, and will be subject to the usual editorial and review procedures of Zentralblatt. It is hoped to persuade the vast majority of journals published in Europe, and many from elsewhere, to participate in the CAP-EMS scheme. The ownership of the data submitted to CAP-EMS will already be vested in the European Mathematical Society. It should also be stressed that a number of European initiatives are already under way with the aim of exploiting new technology to develop mathematical publication and documentation. The European Mathematical Society has launched its ``Electronic Library of Mathematics'' in EMIS. It will soon contain 30 journals plus conference proceedings, and is freely accessible to any user from 30 mirror servers around the world. The MathDocCell in Grenoble, which is a joint partnership between the University Joseph Fourier, the CNRS and the French Ministry of Education, has made important contributions to the development of software for making efficient searches in the database MATH of Zentralblatt. More generally, the MathDocCell provides an excellent model for a much broader European involvement in Zentralblatt. French and German mathematicians are also participating in developing an electronic version of ``Jahrbuch ueber der Fortschritte die Mathematik'' from 1868-1940. Longer term plans. It is self evident that the long term success of a major database depends on three key ingredients. Firstly, the contents must be very comprehensive, covering nearly all current publications and as large a part as possible of past mathematical literature. Secondly, it must be widely and frequently used by most working mathematicians, who perceive it as a tool as important as the traditional university library (indeed, one might hope with the development of technology that the full contents of most older journals would become available in the database). Thirdly, it must have a sound financial basis, with subscriptions prices being set at a level just sufficient to meet the true costs of running the database and to provide the financial investment needed for its long term development. The European Mathematical Society plans to take action to determine what facilities and services most mathematicians would like in a database of the future. Key questions here include whether or not such a database should be purely electronic, what status a publication should have to be included in the database, and how much effort should be expended in seeking reviewers for articles listed in the database. At the same time, it will do all in its power to promote the use of Zentralblatt by the mathematical community, and to persuade publishers to provide free data for Zentralblatt. In this way, it hopes to come up with a series of concrete proposals about how it might be best for Zentralblatt to evolve over the next five years. Because of the rapid evolution of electronic technology, there will also be a long term need for regular evaluations of the database by the Society. Once the longer term plans of operation for Zentralblatt as a database have been agreed by all relevant parties, The European Mathematical Society will seek the support of the national mathematical societies to appoint a database officer and to establish a database node in each European country. Any costs attached to running these nodes would eventually have to be met from subscriptions to the database. The role of the node would be to generally oversee the automatic collection of material for the database from journals published in that country. The node would equally oversee the operation and access to the database by mathematicians working in that country. An excellent model for such a node is provided by the MathDocCell in Grenoble. Its existence is a first concrete step towards the development of Zentralblatt on a Europe wide basis. The issue of funding the database both in the short term and in the long term remains a difficult one, which must at all costs be solved for the future health of European mathematical research. The European Mathematical Society does not have the financial resources to make a direct financial contribution towards the running of the database, and it recognizes fully that the same is true for the national mathematical societies. It believes that a major part of the funding required to run the database will always have to come from subscriptions. It also believes that the mathematical community will indeed be willing to pay a fair subscription price for a database which genuinely meets its needs. However, the Society feels it can make important contributions to the problems of funding. Specifically, it will do all in its power to persuade the European Union to provide significant financial support to establish the database as an essential large facility for mathematical research in Europe. It will also work with the Heidelberger Akademie der Wissenschaften, FIZ Karlsruhe, and Springer Verlag to make transparent and public the basic cost of running the database and will then advise these bodies on what it feels are fair subscription levels.