“Bring me the broomstick of the Wicked Witch of the West”, the Wizard of Oz booms. It is an iconic phrase in one of the most-watched movies of all time. The phrase is also both poignant and baffling. Of course, Dorothy and her companions are crestfallen to hear this “very small task” the Wizard has in mind for them. “If we do that”, the Tin Man stammers, “we would have to kill her.”

It is impossible not to side with Dorothy and the Tin Man here. The Wizard’s task – which requires that the four floundering protagonists gain entry into a well-fortified castle and kill a powerful witch who is bent on vengeance — seems disproportionate to the companions’ requests of the Wizard. In addition, three of the four requests our heroes make — a brain, a heart, and courage – are arguably possessions needed before the task, not afterwards. The task seems all the more unfair because it is arbitrary. Plucking a broomstick from the clutches of a dead witch and presenting it to the Wizard in no way helps him provide the companions with a heart, a brain, courage, or a trip to Kansas.

So it seems also with proposals to the National Science Foundation. Getting a proposal funded by NSF requires generating a central scientific question, crafting a clever and engaging thesis, producing a set of testable and bulletproof predictions, and making the case that the research will both engage undergraduate students and edify the public. Don’t misunderstand me here; these are all worthy goals. I am glad that the NSF insists on these strict standards. Yet scientists seeking funds from NSF often feel that, like the Tin Man, they must “prove themselves worthy” of funding by completing a task both disproportionate and disconnected to the scientific work they propose to do.

Like the broomstick challenge faced by Dorothy and her companions, an NSF proposal is a formidable undertaking. While I am sure some of my colleagues come to the task more easily than I do, I estimate that preparing a grant proposal to NSF from scratch takes about as much time as writing two scientific papers. Therefore, it is often not clear whether I should spend my time writing a grant proposal that stands little chance of funding but would permit us to continue our work or publishing two papers. An NSF proposal, as you can imagine, is never published. Bits and pieces of a proposal might find their way into later journal articles, but the proposal itself is a document read only by a panel of colleagues who sit in judgment. Considering the time spent in preparation and the fact that the funding rate has fallen by about half in the past ten years (to 1/6 of all ecology proposals funded over a three-year period, including initial submissions and re-submissions), submitting an NSF proposal has become a high stakes gamble.

Proposal-writing, moreover, is not like doing scientific work. Most scientists work within a certain conceptual framework, to be sure, but their day-to-day efforts to answer scientific questions within this framework force them to follow a tortuous path comprising many steps. Here is the first disconnect between scientific research and a grant proposal. A successful proposal must center around a single, unifying question. Scientific research rarely addresses a single, coherent, all-encompassing question. Instead, most science is particulate, consisting of a long set of meandering steps, each clearly related only to the preceding one.

On the Loon Project, I first learned that loons engage in territorial battles. Thus, I initially aimed to describe territorial contests and their purpose. During that work, I learned that males fight more dangerously than females. This finding led me to examine differences between males and females that might lead to the difference in fighting, which, in turn, led to the discovery that males choose the nest location and have a greater stake in remaining on a familiar territory than females. Even after three logical steps, my journey had taken me far afield — from territoriality to nesting behavior. Discrete steps are essential, by the way, because scientists must publish their findings routinely in scientific journals in order to justify further research, gain tenure, and have a chance to attract extramural funding. So, like most other scientists, I published my work incrementally through short papers focused on narrow topics, not in a book or monograph addressing a broad question. Most importantly, while the logical path I followed makes sense in retrospect, I could not have written a grant proposal that anticipated it.

Another difference between the practice of science and the task of attracting funding for science relates to the role of serendipity and chance discovery. At its best, science is exciting, because the outcome of any experiment or set of observations is not known ahead of time. Some years ago, we expanded our study area to include a larger sample of marked loons for territorial study, chiefly to learn if young adult prebreeders establish “footholds” in certain lakes where they intruded often in order to increase their chance of later competing successfully to settle on those lakes. We did ultimately answer this question — young loons do not use footholds – but while surveying new lakes of various shapes and sizes, we blundered upon an unsuspected pattern. Young loons, we learned, settle to breed on lakes that closely resemble their natal one in size and pH. (The finding stands as a rare case in which an animal seems to learn a preference early in life that is disadvantageous to it later.) If we had not veered from our normal path to describe natal habitat imprinting, we would have been ignoring a crucial finding of great value to other scientists. Yet publishing a paper on natal habitat imprinting cost us precious time and energy that might have been spent solely studying territorial behavior. The new finding took us into the field of habitat selection, a subdiscipline that our proposal had not anticipated. So neither writing a grant proposal nor conducting funded work leaves room for an unexpected discovery that leads to a new line of investigation.

Why am I suddenly so critical of the procedure for acquiring research funding from the National Science Foundation? You guessed it: sour grapes. I just learned that our proposal to NSF, which I spent most of my spring semester working on, was not recommended for funding. Two reviewers loved it, two hated it, and several others were on the fence. While I am in mourning now and shall be for some time, this is not a complete train-wreck; the reviewers were quite specific and helpful in their criticisms. So resubmission of a greatly-revised version of the proposal that addresses reviewers’ concerns might meet with a better outcome. At the moment, though, I am feeling like the Tin Man did after the Wizard’s request!