Appendix 3:

I C M FORMAT

The Executive Committee has been discussing ways, specially the ones presented by Prof. David Mumford, in which the organization and structure of the International Congresses might evolve in order to better serve the needs of the mathematical community. For many years, the Congresses have been organized in the same way: the Executive Committee appoints a Program Committee, which decides on the sections and appoints subcommittees for each section. These recommend section speakers to the Program Committee, which makes a final selection of both plenary and sectional speakers. To prevent any pressure being exerted on the organizations, these committees have all been secret, the Program Committee being announced at the Congress itself. The Executive Committee has been considering whether there are ways to involve a larger group of people in the organization and to offer talks with different aims from the plenary and sectional talks. The purpose of this agenda item is to get feedback from the entire membership of the IMU on wether such changes should be explored. Our proposal is not to make definitive decisions, but to generate a set of suggestions and recommendations to give the next Program Committte for their consideration.

The specific ideas which we have considered include these:

1. Make the name of the chairman of the Program Committee public so that any mathematician may write to this person with names of proposed speakers or topics of proposed talks. Such input might even be solicited by announcements in the newsletters of various professional journals.

2. In order to involve more effectively mathematicians from applied areas (including computer science, mathematical physics, statistics as well as industrial and applied mathematics), some sessions can be sponsored and organized by a joint subcommittee between the IMU and other professional organizations. Rather than attempting to systematically cover all major developments in the previous 4 years in applied areas (which would be a major enlargement of the Congress), such sessions might take specific 'hot' topics and present a set of talks on such topics which give mathematicians in unrelated fields a chance to get an overview of this development. The joint committee sponsoring these sessions might also give feedback to the Program Committee for talks in other areas which could attract more participation from mathematicians in their applied area.

3. The program might include some less formal talks of broad interest possibly in the evenings. Ideas here range from something like the Gibbs lectures given at the AMS to non-research topics like a survey of mathematical software now available for research or teaching. Getting names for such speakers may especially need suggestions from the community at large.

We hope that all National Committees and their delegations will consider these questions prior to the General Assem bly and will feel free to raise any other idea in this discussion.