Experiments in high energy physics have continuously increased in size, complexity and cost. The problem of proposing, approving, and scheduling experiments has increased even more dramatically. The questions for good experiments used to be''when and how''. It is about to become ''whether''. Startup of the new accelerator at Batavia forces an answer to the question soon. This note is a first attempt at suggesting a formal structure by which proposed experiments can be expeditiously and yet carefully evaluated with some built-in mechanisms to insure objective and fair treatment. A reasonably efficient system has evolved over the years for assigning beam time at the major US accelerators. The system works roughly as follows: Proposals for experiments are submitted in confidence to the laboratory director or his designated officer. The director consults a panel of advisors chosen by him who review the proposals, meet with the proposers, and offer their assessments to the director. The director then announces his program decisions in the form of approved beam hours and priorities. Until recently, nearly all reasonably sound experiments could expect to be approved and the principle competition was for beam hours and priorities. Rejected experiments were generally agreed by mutual consent of the advisory panel and the proposers to contain flaws in technique or calculation which appeared in debate and which rendered them unfeasible. The advantage of this system was that the advisors relieved the director of being sole arbiter and judge of the physics. The confidential nature of the proceedings avoided unnecessary embarrassment to experimenters when correctable errors were uncovered in the review of fundamentally sound experiments. Very casual proposals were accepted on the reputations of the men responsible. The system was informal, adequate and effective. Two things have combined in the last few years which threaten to invalidate this system: (1) more good experiments are being seriously proposed than can be accommodated with existing facilities; (2) lead times to fund, staff, and prepare experiments have increased so greatly that major policy decisions must be made prior to formal approval in many cases. The result has been that experiments which ultimately receive approval are often mistimed for the experiment's staff schedule. Those which are completely rejected result in months and even years of wasted effort. It will soon get much worse. How then might we move to accommodate these changes and head off some of the worst problems? Consider the diagram in Figure 1. It is a sort of flow chart to map the progress of an experimental proposal as it passes through the NAL decision-making process. Clearly, since all questions are posed in the yes-no format, it is somewhat simplified over real life. The rectangular boxes are the heart of the system and are probably more important than the flow paths. It differs from the present system in three important ways: (1) almost every part of the process is public; (2) criticism and discussion are invited between competitors in the presence of the advisory panel; and (3) a formal appeal procedure is suggested.
