November 2005 / Vol. 55 No. 11 • BioScience 961Special Roundtable SectionThe science of understanding climate change wasinstigated in the 19th century by a desire to understandthe comings and goings of ice ages, given new relevance by thecontroversy over the effects of supersonic transport in the1970s, and driven in recent decades by compelling data show-ing an increase in atmospheric carbon dioxide. The fieldcoalesced under the strong leadership of a few key scientistswho saw the broad problem, organized other scientists, andcommunicated effectively to international leaders and thepublic at large. Today the community of climate scientists in-cludes participants from many different disciplines and fromall around the world. While multiple formal scientific mod-els play important roles in the understanding of climate sci-entists, no model is sufficiently representative of the wholeclimate system to adequately describe the full process and con-sequences of global warming. Rather, the knowledge em-bodied in multiple models, as well as knowledge that does notfit into formal models that make up climate science, is un-derstood as a whole through a collective process of learningand understanding among scientists with diverse trainingand expertise.After addressing how a collective learning process getsstarted, we argue that there are four ways through whichdispersed disciplinary knowledges are being brought togetherin the understanding of climate change. These are (1) build-ing large global circulation models starting from basic phys-ical principles, (2) integrating models that were originally builtto understand economic and ecological systems separately,(3) preparing assessment reports on the state of climate sci-ence for scientists and policymakers, and (4) working withina distributed learning network whereby individual scholarsadjust their own research design and interpretation in responseto what they learn from other scholars. These four differentapproaches complement each other to inform the collectiveunderstanding of climate change among the community ofscientists engaged in the process. Our concern here is primarilywith the interaction among scientists, rather than the relatedinteraction between scientists, policymakers, and broadersocial institutions.Early visions of newly critical systems In a world where most scientists advance by reducing prob-lems into parts and tackling pieces, how do systemic questionsand temporary answers arise and attract sufficient attentionby other scientists to become a new field of inquiry? SvanteArrhenius calculated in 1896 that a doubling of carbon diox-ide in the atmosphere would lead to an increase in global tem-perature of 5 to 6 degrees Celsius. Although he was primarilytrying to understand whether a decrease in carbon dioxidecould bring on an ice age, he did note that by burning fossilfuels, people had become a driving force of climate. A few sci-entists elaborated further on Arrhenius’s argument over thenext half-century, but global warming did not become a sig-nificant collective area of work until Charles Keeling beganRichard B. Norgaard (e-mail: [email protected]) works in the Energy and Resources Group, University of California, Berkeley, CA 94720. Paul Baer (e-mail: [email protected]) works in the Center for Environmental Scienceand Policy, Stanford University, Stanford, CA 94305. © 2005 American Instituteof Biological Sciences.Collectively Seeing ClimateChange: The Limits of FormalModels RICHARD B. NORGAARD AND PAUL BAERUnderstanding the risks posed by anthropogenic climate change and the possible societal responses to those risks has generated a prototypical example of the challenge of “collectively seeing complex systems.” After briefly examining the ways in which problems like climate change reach thescientific and public agenda, we look at four different ways in which scientists collectively address the problem: general circulation models, integratedassessment models, formal assessments (e.g., the Intergovernmental Panel on Climate Change), and distributed learning networks. We examine the strengths and limitations of each of these methods, and suggest ways in which a greater self-consciousness of the need for plural approaches couldimprove the basis for learning and decisionmaking.Keywords: assessments, climate change, epistemology, sociology of sciencemeasuring carbon dioxide concentrations on Mauna Loa inHawaii in 1958. By 1963, Keeling was concerned that carbondioxide concentrations were increasing considerably faster thanheretofore thought. Then there was a period during the mid-1960s through the mid-1970s in which the proposed fleet ofsupersonic transport planes and the use of aerosols in spraycans seemed to threaten cooling, offsetting the greenhouse ef-fect. Stephen Schneider argued that, whether the climate waswarming or cooling, the uncertainty justified paying attentionand taking corrective actions, given the tremendous impor-tance of climate and the risks to agriculture (Schneider andMesirow 1976). The potentially cooling driving forces wereaverted because supersonic transport proved expensive, yetKeeling’s data showed carbon dioxide in the atmosphereconsistently rising. Thus scientists’concern returned to warm-ing, marked by National Academy of Sciences (NAS) reportsin 1974 and 1977 arguing largely for more research funding(Weart 2003).Our central point here is that a new conception of what con-stitutes a critical system of interactions is necessary, eventhough it might only be a crude description in the beginning.Somehow the boundaries, key components, important in-teractions, and insights into the possible consequences of a dy-namic between people and their environment must becomesufficiently established to rally scientists to begin to work to-gether to further understand the potential problem. The hap-hazard way in which critical questions have arisen andscientists have rallied to work collectively on them inthe past is not comforting. Could we have averted cli-mate change more successfully had scientists pur-sued the implications of Arrhenius’s calculationsooner? How many equally critical questions havebeen raised but are languishing now? And could sci-ence be better organized to look for and respond toareas in which collectively trying to understand acomplex system might be fruitful? There are clear parallels with the difficulties, indeedfailures, of national intelligence agencies trying tocollect