Page:Sm all cc.pdf/196

 ==Big Science versus Little Science == As the geometric growth in number of scientists collides with the linear growth in available science funding, debate is inevitable about where the scarce resources should go. Much of this debate has centered on the issue of big versus little science. More accurately, since there is a continuum between the two, the question is: what are the optimum proportions between large and small projects within a discipline, in order to maximize the scientific payoff per dollar expended?

Big science can be in the form of a single multi-investigator project or research thrust, or a large facility that is used by many researchers for their individual small-science projects. Multibillion dollar examples of the former are the Human Genome Project, (cancelled) supercollider, and space station, though even within these projects there are many moderate-scale subprojects. Examples of large facilities for small-scale projects are telescopes, oceanographic ships, supercomputers, and Antarctic research stations.

Proponents of small science point out that most major discoveries have been a product of small research groups working with modest funding. Such projects are very cost-effective, because most of the money goes to scientists rather than to the equipment and technicians that generally consume most big-science dollars. Advocates of large science accept these arguments, but they claim that the waves of small science have merely washed around some key problems that were too expensive to tackle. Now these problems are the most critical issues remaining to be solved; they can no longer be bypassed.

Most scientists do small science. If science were democratic, many of the big science projects could not fly. Thus the proponents of the largest projects seek a different constituency; they also solicit line-item funding that does not obviously reduce small-science funding. Successful proponents of big science tend to be well known senior scientists who already head large groups and who are on committees charged with outlining new directions for a field. Younger and less famous scientists feel left out. This week my closest colleague at Columbia University won the largest grant that Columbia had ever received. Yet most of my friends there are less successful ‘softmoney’ researchers, who doubt that they will be able to write enough successful proposals to provide their own salaries next year. Also this week, cosmologists are ecstatic over the results of the big-science COBE satellite: the big bang theory has received remarkably strong confirmation, through a mapping of the original subtle heterogeneity of its radiation. Is there a more fundamental scientific question that the origin of the universe, the mother of all singularities?

Debate over the Human Genome Project was often personal. James Wyngaarden, who was head of the National Institutes for Health when NIH started funding of the project, said “Most knowledgeable people and most eminent scientists are solidly behind [the project]. The ones who are critical are journeymen biochemists who may be having a hard time competing themselves.” James Watson, Nobel laureate and previous head of the program at NIH, said, “It’s essentially immoral not to get it done as fast as possible.” [Angier, 1992] Polarization and alienation are hazards of the battle. The big-science projects generate another hazard: grand expectations. Virtually all funded proposals make confident predictions of valuable results; optimism and a modest amount of oversell are almost prerequisites for funding. Most of these projects will be somewhat fruitful, partly because the investigators are free to react and refocus their research to avoid obstacles and exploit discoveries, but most projects will also deliver less than the proposals promised. Small-science projects can get away with this because there is safety in