Typically, the GrantDoctor depends on rumor, innuendo, low-level staffers muttering in smoky back rooms, and widely available public documents in writing this column. But every once in a while, he hits on a topic important enough that he ends up talking to someone important. This week, an e-mail to the National Science Foundation's public information office resulted, unexpectedly, in a conversation with Arden Bement, NSF's director.

What topic provoked NSF to deploy its Top Gun to talk to the lowly GrantDoctor? In recent weeks, I've started to suspect that NSF is getting serious about funding frontier science--the kind of work being referred to these days (especially at NSF) as "transformative" or "transformational." Admittedly, the evidence that this is happening is patchy and anecdotal. Over the last year or so, several established scientists I know have had renewal proposals rejected after years or even decades of NSF funding, despite very good reviews and continued productivity. For some senior scientists, the loss of NSF funding is a serious crisis; in at least one case I know about, it led to the early retirement of a very competent scientist. But that's okay--there isn't enough money for everyone, so some win and some lose. If NSF manages to fund more risky science, that's great news for young scientists, who seem, constitutionally, more inclined to take risks than do their grayer, more established colleagues--and whose shorter track records make it harder to convince reviewers that they can get the work done.

Of course it's easy to blame these late-career meltdowns on NSF's lean science budgets, but program officers I've spoken to tell a different story. Although not commenting on specific cases, they indicate that science that seemed innovative years ago may have gotten stale. It may still be good, solid science, but it's not exciting anymore. "There are programs that have gone on for a long time," Bement said in our interview. "It may be that the frontier has moved on." In previous years, that may not have mattered; competent science proposed by experienced scientists usually got funded. Cutting-edge--but risky--science proposed by younger folk usually didn't.

But the times they are a'changin', or so it seems. In our interview, Bement acknowledged that a "shift in emphasis" was occurring, although he insisted that the changes are not wholesale. "Doing research at the frontier is not new for NSF. This is a reaffirmation of what NSF has been doing since its founding."

Every funding organization aspires to transformational science, but most either fail or have limited success. It isn't an easy thing to do, for lots of reasons. If NSF can figure out why it's hard, they can probably figure out how to solve the problem.


Arden Bement, credit: National Science Foundation

One reason that funding frontier science is difficult, Bement said in our interview, is that the frontiers are hard to locate. "Our primary task must be to tenaciously dog the frontier," Bement said in a 2005 speech at AAAS headquarters. "The frontier is our bull's eye." In order to dog that frontier, or hit that bulls-eye, "the key thing is to find out where the frontier is to begin with," Bement told me in our interview. To address this, NSF modified its strategic planning process this year to incorporate a wider range of opinion, incorporating the opinions of NSF staffers and the broader scientific community. Much more emphasis has been placed on public comment. A less insular approach could well allow NSF to do a better job spotting the frontiers. It's worth trying.

Richard Feynman once said that if you can't explain a bit of science in simple terms, you don't understand it. A corollary might be that if you can't explain your work's significance in simple terms, maybe it doesn't have significance--or maybe you just don't understand its significance, which, for a scientist, is just as bad. Many scientists are not good storytellers, and that makes it harder for them to identify and articulate the central narrative of their work. Many scientists don't even know where they stand in relation to the scientific frontiers, and even if they do, they can't articulate it in their grant proposals. Although program officers can help with this, this isn't a problem that NSF can solve. Solving this one is up to scientists.

In recent years, NSF has begun to emphasize its second criterion, "broader impacts," which many scientists interpret as encompassing mainly the softer, social virtues: public engagement, diversification of the scientific workforce, helping women advance into senior positions, and so on. If you want to improve your chances of getting funded, the conventional wisdom goes, add something about outreach, diversity, or fueling economic growth. Yet it has always been clear from NSF's language that "broader impacts" can also include purely scientific and technological virtues. Although NSF has been--legitimately--interested in a project's social contributions, broader impacts, I've come to believe, were initially intended as a way of getting scientists to think hard about--and to articulate--a deeper, longer-term vision.

Still, NSF gets some of the blame for their failure to fund more risky science. Even idealistic young scientists often find their outlooks transformed by a funding system that seems to penalize audacity. Authors of ambitious proposals often find their work dismissed by reviewers who use that very word--ambitious--as a criticism, as Science’s Jeffrey Mervis observed in a news article late in 2004 (institutional or Science subscription required for access). Used this way, "ambitious" is code for "they will probably fail." Of course, it isn't just NSF; NIH, too, has gained a reputation for funding safe science. Funding science on the frontier is hard.

So why is NSF making these changes now? Because the frontier isn't the only thing that has moved on. NSF's role as facilitator of important basic science is more important than it has ever been, Bement says, because other science funding organizations have shifted away from basic work. "We have to pay attention to the fact that there are some shifts in resources," he said in the interview. The result, as he put it in that 2005 speech, is that "if we at NSF stop short in our pursuit of high-risk endeavors, it seems to me that we leave an absolute vacuum. In a science and technology-based world, to divert our focus from the frontier is to put the nation at peril." If NSF doesn't fund research at the frontiers, no one will.

Another reason NSF is changing, Bement says, is a National Science Board (NSB) study, the result, in part, of a 2004 NSB meeting that was covered by Science in the article above, where 15 outside scientists presented their views about what was wrong with NSF. As Jeffrey Mervis wrote, those scientists "suggested how NSF might become more receptive to the handful of ideas that have the potential to set the scientific establishment on its ear."

So what else is NSF doing to improve the function of its ambitious-science receptors? In his AAAS address, Bement suggested that NSF program officers were a key to making this work--a theme he echoed in our conversation. Program officers can--and should--keep an ear to the ground, which puts them in position to respond to the scientific zeitgeist, Bement says. The key to NSF's success, he said, is to give program officers latitude to fund work that is aligned with a field's most important trends and developments--especially work that's rejected by reviewers as being, in their opinion, too risky.

One way of accomplishing this, Bement told me, is by making better use of NSF's "sugar" (SGER) grants--Small Grants for Exploratory Research. NSF program officers are now empowered--and encouraged--to set aside proposals rejected by reviewers and then to fund them at a lower level via the SGER program. This would allow the researcher to bring the work part way to make the work competitive for more substantial funding. Another thing program officers can do, Bement says, is make it clear to reviewers that NSF wants to fund forward-looking science: "It's partly a matter of expressing your expectations," he says.

Despite the important role that program officers will play, NSF's success will depend, as things scientific always do, on scientists, particularly young scientists with fresh ideas--on their ability to identify important problems, find promising routes to solutions, and articulate their ambitions in compelling and convincing ways. So is it time to throw caution to the wind and pitch your most radical ideas? Despite tight budgets, it is indeed a great time to submit ambitious proposals. But a solid research plan is as important as ever. "There are some proposals that come in that are beyond the frontier," Bement says. Those proposals, he implied, don't get funded.

America's leading basic-science agency is in the process is working hard to replace scientific baby steps with big ideas. It's an exciting time to be a scientist.

Be Well,

The GrantDoctor

The GrantDoctor is a pseudonym.

The GrantDoctor is here to help! Send questions to grantdoctor@aaas.org. Please put 'GDR' in the subject header of your e-mail.

Comments, suggestions? Please send your feedback to our editor.