June 27, 2013
Jefferson Lab is a user facility; we conduct experiments, dominantly, but not exclusively, on the CEBAF (Continuous Electron Beam Accelerator Facility). Which experiments are run, and for what periods, result from decisions made by the Director. However, there is a lot of input to the process. In fact, the process an experiment might follow - from idea to publication - is laid out in a detailed flow chart that can be found on the Jefferson Lab web page. One important omission is how to get approval for the foreign travel to Stockholm for the Nobel ceremony, but apart from that, it’s pretty complete.
The process, largely orchestrated by the Deputy Director for Science and Technology, Bob McKeown, and the AD for Experimental Physics, Rolf Ent, is sociologically very interesting. A key piece is the activity of the Program Advisory Committee, which met June 17-20. Under consideration by the PAC were proposals from potential users of the facility. The committee has 14 members, all physicists of considerable renown and together, covering all the significant aspects of the Jefferson Lab program. Currently, five, including the Chair of the Users Group Board of Directors, are from U.S. institutions, and another nine are from across the world; none are Jefferson Lab staff.
Depending on the need, the PAC may meet more than once a year, but at Jefferson Lab, as in most laboratories, there is usually at least one big meeting a year. For a summer PAC, we are expected to invite proposals in say March and to receive the proposals about seven weeks before the date of the PAC meeting. The deadline needs to be specified to the hour, including time zone. Even then, there are requests for special treatment which must be resisted.
Long before the meeting, the PAC Chair designates a primary and secondary reader for an individual proposal. These readers interact with each other and with the proponents. In parallel, internal Technical Advisory Committees look for, and provide to, the PAC, input or feedback on issues that they perceive; similarly, our theorists look at the proposal from the point of view of understanding that the goals of the experiment could actually be achieved. Before a public presentation by the proponent, the readers brief the rest of the PAC on their take on the proposal, what they see as issues, how it fits into the potential program, what is its relationship to previous proposals accepted and deferred.
Then comes the public presentation in the auditorium. Make or break is the sense of the proponents. In my experience there is considerable nervousness. The speaker may be a bright young post-doc or a seasoned veteran, but anyone can fail to respond to a good question. And the speaker is representing the team, so (s)he has to carry a lot of responsibility.
Discussion in the PAC resumes, perhaps not immediately, but certainly before the presentation has faded from their minds. Rarely is the discussion limited to one or two members; the designated readers take precedence only in the initial phase. Chairpersons are special and it’s quite amazing how we find them, but we do, and they always provoke and cajole, and strong feelings emerge.
At some point not at the beginning, but after discussion of perhaps a suite of proposals with similar subject, the PAC chair will push subtly towards an approval recommendation, sometimes with a vote needed to reach a conclusion. It can be quite surprising what might be the vehicle for a conclusion. For example, each proposal asks for a certain amount of beam time. Usually that is a composite of some days with one target, some with another; some at low momentum transfer squared, some at high. These divide into longer and shorter periods of beam. An individual PAC member may say that (s)he would cut the high momentum transfer measurement because it is one point in 10 and half the time. Another might argue that that is where the physics is to be found. Then someone suggests that without that point, the whole experiment is weak. Then someone agrees and the whole proposal can be on skids, or vice versa.
But that is not the end. Once all the proposals have been considered for approval, we come around again to grade the proposals and to assign the beam time that goes along with the grade. Often the discussions are re-opened. Some experiments develop a consensus in favor and a high grade, others have a divisive effect. All the PAC members are heard, all have opinions; they learn each other’s foibles and predilections. When something is unclear, the original proposal, or even a previous proposal is opened up and the relevant knowledge found, and opinion is confirmed or not by checking on what the PAC said previously. Always, the PAC manages to stand behind its collective opinion.
After the PAC meets and writes its report, the lab completes the process by publishing the report with a cover letter from the Director indicating acceptance of the recommendations.
When described like this the whole process seems a little arcane, but it has amazingly good acceptance by the community and resistance to adjustments seems to be the rule rather than the exception. There are also jokes about such and such an experiment having had the maximum impact of any ever run at the lab, but having been rated only as a B+ by the PAC; nevertheless, no one wants a grade that is not at least A-. We should also not forget that the eventual scheduling of a mutually compatible set of experiments is a matter of more than letter grades; it takes a lot of ingenuity from the experimental physics division and from the accelerator division. This process deserves a discussion of its own.
Finally, a quirk which I only appreciated this year goes as follows: At one U.S. lab, until recently, the main meeting was held in an attractive town in the Rockies which is named after a tree. At Jefferson Lab, we take the PAC process so seriously that we cloister the committee in a room (L102) in CEBAF Center with no windows and let them out periodically to listen to presentations and to sleep. We can only claim that our physics is worth it!