from a Proposal for a Conference on``The Future of Mathematical Communication'' Joe Christy Head of Mathematical Computing David Hoffman Head of the Scientific Graphics Initiative Mathematical Sciences Research Institute _______________________________________________________________________________________________________ This fall, MSRI will host a special conference on The Future of Mathematical Communication. The workshop is being organized by Joe Christy, John Gage, David Hoffman, Stewart Loken, Andrew Odlyzko, and Richard Palais. Rationale for the Conference The Current Revolution in Electronic Communication In the last two years, there has been a huge increase in the research use of the Internet, including multimedia electronic mail (using the MIME standard), and networked information servers and browsers for easy access to text, animation and executable programs (for example World Wide Web / Mosaic). Preprint servers have been established in several areas of mathematics, at MSRI and elsewhere. The AMS has set up a server covering all of mathematics. At least five electronic journals in mathematics are now in operation, namely the Electronic Journal of Combinatorics , the Electronic Journal of Differential Equations , the Electronic Transactions on Numerical Analysis , the New York Journal of Mathematics , and the Ulam Quarterly . Mathematical societies in the United States and Europe are seriously exploring the impact of various sorts of electronic archiving and distribution mathematical papers. The possibility of the easy creation of an e-journal, consisting of papers read and evaluated by a recognized group of mathematicians, holds many opportunities for greatly increasing the speed and quality of mathematical communication. There are also serious problems and open questions: ¥ Will there be a role for professional societies (AMS, SIAM, MAA) in managing this change? Can they adapt fast enough? How will these Societies cooperate with their counterparts in other countries? ¥ Will there be a role for traditional mathematics publishers? What would it be? ¥ Will there be a role for university libraries in providing access to and in archiving mathematical information? How can they adapt? ¥ How will the process of evaluation and certification be handled? ¥ What will be the economics of this evolving system; i.e. who is really paying for it now and who is going to pay for it in the future? ¥ Should we use the democratizing potential of the Internet to de-stratify mathematics? How will it alter communication among researchers, teachers and students of mathematics, and the scientifically aware public? The situation is changing rapidly. The ACM (Association for Computing Machinery) has outlined a comprehensive plan for changing the nature of its publication methods, proposing print-on-demand, pay-per-view, and profiling to automatically inform a member about the existence of an article that coincides with his/her professional interests. SIAM is planning analogous, although more limited services. The AMS Committee on Publication is currently deliberating many issues including electronic data exchange, the transition to paper-less journals and the need for authoring tools for more complex forms of mathematical ``papers,''that will include animation, program execution and hypertext-style references. The DMV (Deutsche Mathematische-Vereinigung) focused its last national meeting at the end of June. 1994 on ``New Perspectives of a Distributed Electronic Information System." Among the topics discussed were: distributed preprints and software, information systems in mathematics, electronic reviewing, electronic journals, and electronic access to historical documents which currently only exist on paper. There have already been some notable experiments. The European Mathematical Trust sponsored the ambitious Euromath project to develop a unified mathematical computing environment; the first version of the Euromath System was released this spring. The CORE (Chemistry Online Retrieval Experiment) Project, a collaborative effort of American Chemical Society, Bellcore, Chemical Abstracts Service, Cornell University (Mann Library), and OCLC, created an electronic database of chemistry literature available on the desktop computers of scholars. The core of CORE is a large and coherent collection of machine-readable text, consisting of the full-text (complete with the original phototypesetting markup) of 20 American Chemical Society journals, spanning from the present back to 1975. A number of monographs from Springer-Verlag will also be included. Elsevier Publishers is experimenting in its TULIP project with on-demand electronic delivery of its journals in material science to 15 universities. The San Francisco Campus of the University of California, in collaboration with Bell Labs and Springer Verlag has recently started the RedSage project, using Bell Labs's RightPages Software, to perform a similar experiment in the medical field. Quite recently, the University of California has entered into an agreement with the IEEE to make available most of its publications since 1989 (including 90 journals, 200 conference proceedings, 100 technical standards and 75 books per annum) online to the nine UC campuses, beginning in 1995. These will be accessible using MELVYL, the University's computerized library information network. Some of the moves made by professions other than mathematics may not be appropriate for our discipline. Some of the more controversial issues, potential problems and wonderful opportunities for mathematics have been described in two articles which we believe define the debate. On the one hand, there is the cautionary ``Roadkill on the electronic highway? The threat to the mathematical literature'', by Frank Quinn. On the other, there is the optimistic ``Tragic loss or good riddance? The impending demise of traditional scholarly journals'' by Andrew Odlyzko. The time is ripe for the mathematical community to examine the issues, set goals, and coordinate its efforts. Tools and standards for the creation and exchange of electronic data In the last decade, the acceptance by the worldwide mathematical community of TeX as a standard for the preparation and distribution of mathematical documents has provided enormous benefits for the interchange of textual mathematical information. During the same period, the capability of computers to aid mathematical communication and research by the creation mathematical images and animations, and by allowing the examination of mathematical examples previously beyond reach, also has resulted in many significant advances. The current and expected advances in the next decade in computer power, networking speed and ease of use will benefit mathematicians greatly, provided there is a similar convergence on accepted standards among mathematicians for the exchange of visual mathematical information. This includes a variety of different sorts of items, among them: ¥ image storage formats (TIFF, GIFF, various proprietary formats, etc); ¥ image or page description and markup languages (PostScript, HTML); ¥ compression and animation protocols (mpeg, QuickTime); ¥ conventions for the description and exchange of geometric data among various viewers and mathematical applications. Separate from the issue of what form mathematical journals may take and how they will be distributed is the question of what will constitute a mathematical article or communication in the future. It will be possible to include audio, animation and even pointers to executable programs. This is now done in conventional paper journals by reference to an ftp source for either code, images or animation. At the present time, it is not easy to produce and distribute a mathematical text that contains animation, audio or executable programs as an integral part of its exposition. It will be necessary to have appropriate authoring tools available to mathematicians, as well as some agreement on standard data formats. An effort is already underway to do this via the mechanism of TeX `` specials.'' Whether this is the right approach or not is now being debated. In the short run this will allow a TeX document to be ``marked up'' in a way compatible with HTML and Mosaic. Scientists are beginning to take full advantage of these tools for communicating directly with teachers, students and the public: the K-12 Internet access project which has grown out of MSRI's ``Conversations between Teachers and Researchers''; MSRI's involvement in the proposed NSF Institute for Science Education; LBL's interactive virtual Frog dissection; the NSF-funded Geometry Forum, based at Swarthmore College; and the Geometry Center's Gallery of Interactive Online Geometry. Mathematical Teleconferencing In the next few years, it will be possible to use the Internet in an integrated fashion for simultaneous voice, image and ``whiteboard sessions'' between collaborating researchers. We hope to be able to demonstrate this capability at the conference. Presently it is possible to broadcast seminars to individuals with appropriate software and network connections. It is likely that this will become an important way to extend the utility of major mathematical centers such as MSRI. We need to address those issues particular to mathematics, as well as general technical and economic questions. Conference Organization We plan to keep the focus upon mathematics, emphasizing its special needs as well as new opportunities for the discipline. We propose six discussions, each led by a handful of experts and centered on a particular topic chosen by the organizing committee. We also plan to have six to eight main lectures by leaders in their respective fields. We will tie these lectures to the discussions. Where appropriate, the discussion leaders will prepare short statements or presentations to set the stage for the conversations. There will be opening and closing presentations by two individuals who will frame the conference. We will have available demonstrations of a variety of networked information servers, preprint servers and mathematical authoring tools for hypertext. The Computer Science Division of LBL, under the direction of Stu Loken, will provide us with the necessary hardware, software and personnel to multicast the conference over an MBone channel of the Internet. We are now in the process of finding out which sites in the US and Europe can receive this. We expect that there will be full audio and approximately one frame per second of video. We will offer a ``starter kit'' to mathematics departments and mathematics institutes that have appropriate Internet access and sufficient technical expertise to successfully install the software necessary to receive this broadcast. This not only has the potential of greatly increasing the audience for the conference, but it will also be a demonstration of a new (to mathematics) use of the Internet for communication. If possible, we will also demonstrate the newly available (from LBL) desktop video-conferencing software with two-way whiteboard capability. Discussion Topics The discussions will be organized in order to cover the following topics: ¥ The various forms of electronic journals and how they are distinct from preprint servers; ¥ Archiving, correcting and commenting on electronic documents; ¥ The role of professional societies and traditional publishers in the transition to paperless journals; ¥ New contents and multiple paths through a document; ¥ Making authoring software easier to use for the creation of these new ``papers;'' ¥ Verification, authentication and copyright issues; ¥ TeX, hypertext and Mosaic; ¥ Building bridges between researchers and elementary teachers and students; ¥ Real-time mathematical collaboration over the Internet.