In America’s competitive scientific arena, different areas contend for limited funds, as their benefits to society—cures for disease, national defense, cultural value—are weighed. The pure and the applied, the theoretical and the experimental, also vie for dollars and recognition.
Two other categories, big science and little science, fracture the lines in yet another direction. What are they? Do their differences matter?
Should we as scientists, or the United States as a nation, care? My research experience in condensed matter physics (CMP), as it changes from “little” to “big,” has made me confront these questions.
When I was a graduate physics student in the 1960s, “big” meant high-energy elementary particle physics. Economics and technology dictated that only large organizations could deal with huge machines like particle accelerators. As a result, the high-energy style is to cluster numerous experimental stations around an accelerator, supported by many millions of federal dollars. Each station sustains at least one large team that combines different talents to produce data. Despite joke about high-energy research papers with page-long lists of authors, ( there seems to be no other way.
For reasons that lay deep within me, I knew as a graduate student that I didn’t want to be absorbed into a large team in a large laboratory. I turned instead to CMP, where spectrometers and other tabletop equipment produced first-rate results. Several such facilities, each costing between $10,000 and $50,000, nicely furnished a 20-foot by 30-foot laboratory. Each could be operated by one or two people. This seemed just right to me, and I happily began my dissertation research, mostly alone. Of course I learned from my adviser and others in our group, but the whole operation included only six to eight people on average.
Today, I and many others still do tabletop CMP research in small groups. Important science still emerges in such a setting, such as the discovery of high-temperature superconductivity by a two-person team. But there are clear signs that CMP research—and research in general—is changing from scattered cottage industry to centralized cartel.
One omen is the growth of large facilities, as a synchrotron radiation source replaces laboratory spectrometers, or a molecular beam epitaxy laboratory supplants small crystal growth equipment. As in high-energy physics, these core installations feed many large teams. Brookhaven’s National Synchrotron Light Source, for instance, supports about 100 stations for X-ray, ultraviolet, and infrared research in solids.
Another sign is the major corporate and governmental presence in important segments of CMP, like materials science. Goal-oriented research and a traditional belief in team methods favor large, centralized approaches in many industrial and federal laboratories. (Small groups can also flourish in these environments—the two discoverers of high-temperature superconductivity are industrial scientists.)
The trend to bigness isn’t limited to CMP. Large research machines are everywhere. Chemistry extensively uses expensive, centralized nuclear magnetic resonance equipment. Biophysics and biomedicine use the same enormous synchrotrons that serve CMP. Even mathematics, the classic small-group enterprise, now uses large computers. In fact, computers are part of the all-too-efficient networks that further jeopardize creative solitude. We instantly trade data and theories via BITNET, DARPANET and Fax machines, encouraging feverish analysis and sometimes rapid consensus.
Bigger groups are inevitable as we use bigger machines to more thoroughly understand nature. Certainly a large group is desirable—for instance, when it unites people from different disciplines to attack a research problem.
On the debit side, scientists trained in large groups are probably more specialized than those scientists who learn varied abilities in small settings.
But I’m mainly troubled by the astonishing loss of individual research. Although large teams do well in meeting defined goals, as in World War II’s Manhattan Project, most original scientific ideas bubble up from deep within a single psyche. Almost by definition, and because new ideas need long-simmering in solitude before they face reality, big groups rarely overthrow conventional wisdom.
For this reason alone, it’s harmful to encourage further decline in small groups, as is now done by preferential federal funding for big science, And even apart from any special creativity that goes with small science, we need its distinct perspective to keep big operations honest.
America’s Vietnam experience was a stunning example of how the momentum of a big organization generates its own justification to make a tightening spiral. Diverse views help us steer clear of such whirlpools, which also occur in science, especially where it intersects government.
I believe, too, that for the same reasons that artists use different environments or new media to stimulate creativity, scientists need different ways to do science. This will help ensure that research careers attract good minds of every sort—those that flourish in isolation, and those that thrive on interaction.
Ideally, the wise scientist uses different scales in groups and machines to enrich the research experience. Just as Picasso moved through Blue, Cubist, and other periods, each with its own style, I’ve moved, from a lengthy Small Group-Tabletop Period through an intense Large Synchrotron Period to my new Solitary Period, during which I interact mostly with my microcomputer. And, just as major artistic movements come and go, research will also change. Big and little science might be joined by minuscule science, based on the tiny, blood-cell-size machines and sensors now being made from silicon, if these devices have computational power or even intelligence, a large research group may-consist of a few humans with millions of silicon partners.
Sidney Perkowitz is Charles Howard Candler Professor of Physics at Emory University in Atlanta.