A though I do not necessarily agree with everything that Stephen Schneider (2001) has said here, I am very pleased he has made the effort to constructively engage my analysis, Construction of Global Warming and the Politics of Science (Demeritt 2001b). I am pleased also that we seem to agree broadly about the importance of more open public discussion to understanding better the values of climate change science. As a social practice, science involves not only certain normative values, judgments, and social relations among its practitioners, but also more obviously political questions about the instrumental values (and material provision necessary to achieve those objectives) that its applications are to serve in society. Unlike some scientists, Schneider is quick to acknowledge that science is value-laden in both these senses. Though Schneider and I disagree somewhat about the precise philosophical and political implications of that realization, we do agree that it demands a more nuanced understanding of scientific knowledge than the simplistic oppositions between fact and value, science and politics, objective and subjective, upstream and downstream, that have long structured public debate about global warming and other environmental problems. I am hopeful that this exchange will help spark the kind of public dialogue about science and the understanding of science that we agree is long overdue. By publishing in the Annals, my aim was to foster such a discussion. In these days of academic specialization, the Annals provides a rare forum for addressing, not just a broad spectrum of human geographers and others working in the human sciences, but also physical geographers and other natural scientists. Interdisciplinary communication is notoriously difficult. Different disciplinary communities have their own working assumptions and expectations, technical terminologies and shorthands, methods of research, styles of presentation, and standards of evaluation. Without an insider's experience, outsiders often find it difficult to understand and evaluate specialist debates. In my article, I tried to call attention to some of the informal working assumptions and practices of climate modelers by retracing the history of several scientific controversies. The problems at issue in debates about the construction and validation of climate models are much better recognized upstream among climate modelers themselves than downstream by other research scientists, impact-assessment experts, science advisors, policy makers, and political interests interested in predicting and managing global warming. I tried to explore the political implications both of modelers' own modeling practices and of the ways in which those practices are understood, and to some extent shaped, by the interactions of modelers and their models within this wider epistemic community of scientists and policy advisors. As Schneider's response rightly observes, my own analysis was itself reliant on a repertoire of specialist terms and theories drawn from recent work in critical human geography and in particular from science and technology studies (STS). This somewhat heterogeneous body of work has challenged the self-image of science as an epistemologically objective and value-free study of the self-evident (once discovered) facts of a real and ontologically objective world. This umbrella of metaphysical beliefs is common among practicing scientists, as Schneider himself notes. Scientists have often reacted with hostility to the claims made about science by STS. One of the key terms in those debates about science is construction. Along with its various cognatesconstructed, constructionism, deconstruction, etc.construction is often applied, opposed, or both to science, objectivity, facts, nature, and reality. These terms describe difficult concepts with long and complex histories (Daston 1992; Demeritt 1998, 2001b), but they have been thrown around in some quite different and often very imprecise ways. Ian Hacking (1999, 22-23) calls such as fact, truth, and elevator words because they tend to work at a number of different levels in philosophical debates. Facts, truths, and reality are not simply things in the world like rocks or trees. They are also used abstractly to describe the nature of the world itself and to characterize our knowledge of it. Thus, the correspondence theory of truth holds that true propositions reflect the world as it really is (Rorty 1979). Notice the circularity that this definition involves. One