Changing the Atmosphere: Expert Knowledge and Environmental Governance (review)

Changing the Atmosphere is essential reading for anyone interested in environmental issues, climate change and global warming in particular. Inspired by a science-studies approach, it treats scientific knowledge, and the people and institutions that produce it, not as value-neutral, consensually established inputs to policy debates, but rather as key components actually embedded in the formulation and legitimation of such debates and their outcomes. The essays offer "detailed, empirically grounded case studies of settings in which people make and interpret knowledge about the earth's climate and link that knowledge to political decisions" (p. 15). As such they are not merely intended as an "academic" exercise. They are fine examples of how intellectuals can be politically engaged without sacrificing analytical rigor, precision of language, or clarity of exposition.

An extremely well-crafted introduction by the editors situates the argument, summarizes the individual contributions, and shows how they are interconnected. It is followed by nine chapters that introduce readers to the science and technology of weather prediction, locating these in the broader moral, institutional, and political context of environmental governance. The concept of climate has been profoundly modified over the past fifty years. In 1941 the U.S. Department of Agriculture asserted that "the climate of a place is merely a build-up of all the weather from day to day" (p. 7). This locally bound, additive concept has now been replaced by a global, systems concept. The need for expert knowledge has emerged along with the evolution of this "scientific" concept of climate. That knowledge, in turn, empowers "us" to intervene, and to manage the climate locally (hence the idea of climate management as a weapon during the cold war) and globally (from whence, more recently, the less bellicose aim of responsible environmental governance).

Technology has played a crucial role in this process. If it is possible to study climate change on a global level it is because we now have at our disposal a worldwide network of local data-collection points on land and sea (balloons, buoys, and the like), linked together in real time, along with satellite images from space. The masses of information that these provide would be useless without extremely high-powered computers that can be used both to interpret the data and to handle models that represent climate at a global level. As coeditor Paul N. Edwards stresses, these computer-based models make "inaccurate, incomplete, inconsistent, poorly calibrated, and temporarily brief data function as global by correcting, interpolating, completing and gridding them" (p. 62). They literally build empirically based images of the world. The complexity of the phenomena they are modeling, and the intrinsic limitations of the data and of the models themselves, mean, however, that their findings are necessarily contestable. One line of attack is simply to assert that since they produce uncertain knowledge they are not even scientific—a positivistic reproach vigorously rebutted by Stephen Norton and Frederick Suppe in their contribution to the book. More fundamentally, the inherent uncertainty in computer models of the climate leaves scope for differing interpretations of the findings and for lively disputes among experts. This inevitable absence of scientific consensus opens a space for the politicization of the debate over the extent and causes of climate change, of global warming, and of environmental degradation.

Climate change is a global phenomenon with life-and-death implications for millions of people, notably in some of the poorest countries on the globe. Knowledge of how, and why, the climate is changing is, however, produced by a handful of people concentrated in a few centers in the rich, industrialized world. The resources at their disposal, the questions they ask, the techniques and skills they deploy to answer them, the main causes of weather-change that they identify, their related policy proposals—all are the prerogative of privilege. Frequently their "universalistic" science is...