Writing the First Proposal

You must convince the referees that the project is so far along that it would be a mistake to stop it. Put another way:

Every first proposal should read as a renewal proposal.

An Abstract should be supplied even if the agency does not request one. Write it last. Often this succinct sales pitch for the proposal is the only thing read by the last person with decision power over your grant. Furthermore, sometimes referees will structure their report on the basis of your abstract.

1. The Introduction explains the general relevance of your research in a broader context. This section:
   1. shows the granting agency how your research fits in with other areas it funds and

   2. demonstrates that you understand much more physics then you are proposing to do and hence if the opportunity arose could move quickly into developing areas.

2. The Review of Previous Research persuades the reviewer that you are already a productive member in the area of your proposal. If you are fresh faculty member writing your first proposal, this may seem difficult to do. But if you are really proposing to work in an area in which you have never worked before, it is extremely unlikely you will get funded. While only old farts with a track record of research can get grants in brand new areas, most old farts are not so stupid as to try. The usual procedure is to use another grant to get started in a new area so that those results form Part II of the grant proposal.
   • This section of the proposal should contain both a review of the field and what you have done in it. The end of section should, if at all possible, leave the reviewer with a clear view of important problems you are already on the way to solving.

3. The Proposed Research describes what you plan to do. There is a terrible tendency to put in lots of equations (even if you are an experimentalist). To the contrary, the best proposals contain no equations at all! If you feel the need of a bunch of equations, try making a figure or table that indicates the procedure. Self-explanatory figures demonstrate you know what you are doing. (Any experienced referee recognizes it is hard to construct good figures and nearly impossible to construct good tables.)
   • Break this section up into subsections (and sometimes unnumbered but labelled sub-subsections). The hardest job for the referee is figuring out what the proposer wants to do. Clarity is a premium. Put the most important part of the proposal first. The referee is most likely to read this. If it is clear, he will forgive less clear subsequent subsections. But if the first subsection is unclear, or, worse yet, wrong(!), the referee will quite properly conclude you are incompetent and downgrade your proposal.

   • There is a natural tendency to propose too much. What you want to demonstrate is that you have clearly identified the next problem to do (in a developing field) and that you have a sensible (if not brilliant) way to proceed. Further if possible it is wise to indicate what are the fallback positions if your mainline of attack should fail. What you want to avoid is giving the referee a chance to say: `This idea can't work for the following clear reason'. Also to be avoided is proposals that evoke responses such as: `While this scheme might work, it critically depends on the following miracle occurring.' Now in the case of experimentalists proposing very audacious projects, this is a hard to avoid. You should clearly indicate that you have a thorough command of the difficulties and, at least in some cases, have thought of alternate strategies -- i.e., that you are a real physicist. Which brings me back to the start of the paragraph: a real physicist, while thinking far in the future, doesn't reveal her preliminary thoughts to a referee. The general maxim is: don't expose areas you are not prepared to defend.

   • Note well: the proposal, while a natural renewal of the previous research, should not appear as a routine one -- i.e., as just a continuation of old work (or even worse, of one's thesis). The proposal should be new, exciting and novel while not seeming crazy, far-out, or impossible so that the reviewers can exhibit real enthusiasm for it.

4. The Summary clearly marshals the arguments for your proposal. If you do this well, the referee may just copy some of your sentences. Keep it short and number the points.

5. Budget. This is more difficult for the experimentalist, since it must contain a capital budget. In any case you should not be terrible concerned if the budget is too large. The agency will generally not be disturbed by referee complaints that the budget is too large, it is quite prepared to negotiate with you once it is convinced that you can do something it views as appropriate. On the other hand a too small budget is a mistake, since if you don't ask for it, the agency won't give it to you. (Note the one exception to this rule: some granting agencies -- Research Corporation, Petroleum Research Fund-ACS, etc. -- have strict rules on the size of the budget; in those cases overasking can hurt since it indicates you are not smart enough to read the rules.)
   • One obvious don't: young investigators can't expect support for postdocs (there is a presumption that young investigators are not experienced eough to supervise individuals that close in age and experience). The budget should contain requests for: graduate students (no more than two), summer salary, several expendable areas -- travel(enough for the relevant APS/Society meeting and one summer conference, plus some funds for at least finishing students to attend a meeting); publications (drafting charges, publications charges and reprints); computing (this is very tricky with all the supercomputing being funded separately; but some funds should be included because computing is never free in my experience); misc (includes telephone costs for you and students, expendables [paper and pencils at my institution], xeroxing [including ample funds for new students who always have a xeroxing binge]). Finally another item that increasingly finds its way into proposals is requests for workstations/terminals. (Note: when asking for such hardware you should also include the cost of service contracts, where appropriate.)
     
     

Signal to Noise.

Local Advice.