Naomi Oreskes, of the University of California-San Diego and a leading scholar of the history of science, wrote an excellent article on scientific consensus a few years ago as part of a special issue of Environmental Science & Policy which critiqued the debate over Bjorn Lomborg’s The Skeptical Environmentalist. This is of course the same Naomi Oreskes famous for her short essay reviewing abstracts on “global climate change” in Science (a subject I do not wish to discuss in this thread, thanks!). Below I have reproduced a few lengthy excerpts from Naomi’s paper relevant to recent discussions here, though I encourage you to read the whole paper, especially the three cases that she describes. You can find the entire set of papers in the special issue here.


Oreskes, N., 2004. Science and public policy: what’s proof got to do with it? Environmental Science & Policy, 7:369-383 (PDF).

In recent years it has become common for informed defenders of the status quo to argue that the scientific information pertinent to an environmental claim is uncertain, unreliable, and, fundamentally, unproven. Lack of proof is then used to deny demands for action. But the idea that science ever could provide proof upon which to base policy is a misunderstanding (or misrepresentation) of science, and therefore of the role that science ever could play in policy. In all but the most trivial cases, science does not produce logically indisputable proofs about the natural world. At best it produces a robust consensus based on a process of inquiry that allows for continued scrutiny, re-examination, and revision. . .

Most of us realize that proof—at least in an absolute sense—is a theoretical ideal, available in geometry class but not in real life. Nevertheless, many of us still cling to the idea that some set of facts—some body of knowledge—will resolve our problems and make clear how we should proceed. History suggests otherwise: earlier scientific wisdom has been overturned, earlier generations of experts have made mistakes. This is as true in physics and chemistry as in biology and geology. The criteria that are typically invoked in defense of the reliability of scientific knowledge—quantification, replicability, falsifiability—have proved no guarantee. Moreover, experts do not always agree. Even when there is no transparent political, social, or religious dimension to a debate, honest and intelligent people may come to different conclusions in the face of the “same” evidence, because they have focused their sights on different dimensions of that evidence, emphasizing different elements of the evidentiary landscape. Even when a scientific community reaches consensus on a previously contested issue—as earth scientists did in the 1960s over moving continents—there are always dimensions that remain unexplained. In the future, plate tectonics no doubt will be modified, perhaps overturned entirely. Indeed, there are a handful of scientists today who advocate Earth expansion to explain continental separation, and they are of course eager to detail the limitations of plate tectonics theory (e.g. Shieds, 2003). Nevertheless, for now plate tectonics remains the consensus of most Earth scientists: our best basis for understanding the Earth. . .

Scientific consensus is a complex process—involving a matrix of social, political, economic, historical considerations along with the epistemic—and history shows that its achievement typically requires a long time: years, decades, even centuries. But even when a stable consensus is achieved, scientific uncertainty is not eliminated. Rather, once we have deemed the remaining problems as “minor”—which is to say, insufficiently great as to warrant further concern—we simply live with them (Engelhardt and Caplan, 1987). Moreover, the grounds on which scientific communities have concluded that evidence is “good enough” to warrant living with the uncertainties have varied enormously throughout the course of history. A determined individual may choose to pursue these uncertainties, and that determination may successfully destabilize the prior consensus. In a “purely” scientific debate, that determination would, ideally, arise solely from the demands of empirical evidence, but no debate is ever “purely” scientific, given that, at minimum, credibility, reputation, and, perhaps future funding are at stake.