The Global Warming Consensus

Environmentalists keep telling me that there is a consensus of scientists about global warming, its causes, and its possible severity. I've pointed out in the past that a large majority isn't the same as a consensus, and that even a large majority can be defeated by one scientist who shows that the majority is wrong. When 100 prominent German physicists signed a declaration denying the validity of "Jewish physics," Einstein observed that it would have only required one of them to prove that he wrong--if that were the case. As this article from Physics Today some years ago pointed out:

Philosophers of science say that measurements are "theory laden," and they are. But good experimenters are irredeemable skeptics who thoroughly enjoy refuting the more speculative ideas of their theoretical colleagues. Through experience, they know how to exclude bias and make valid judgments that withstand the tests of time.3 Hypotheses that run this harrowing gauntlet and survive acquire a certain hardness--or reality--that mere fashions never achieve. This quality is what distinguishes science from the arts.

But many of today's practicing theorists seem to be unconcerned that their hypotheses should eventually confront objective, real-world observations. In a recent colloquium I attended, one young theorist presented a talk on his ideas about what had transpired before the Big Bang. When asked what observable consequences might obtain, he answered that there weren't any, for inflation washes away almost all preexisting features. Young theorists are encouraged in such reasoning by their senior colleagues, some of whom have recently become enamored of the possibility of operating time machines near cosmic strings or wormholes. Even granting the existence of cosmic strings, which is dubious, I have a difficult time imagining how anyone could ever mount an expedition to test those ideas.

I like to call this way of theorizing "Platonic physics," because implicit within it is Plato's famous admonition that the mathematical forms of experience are somehow more real than the fuzzy shadows they cast on the walls of our dingy material caves. And, in reaction to the seemingly insuperable problems of making measurements to test the increasingly abstract theories of today, some people have even begun to suggest that we relax our criteria for establishing scientific fact. Perhaps mathematical beauty, naturalness, or rigidity--that Nature couldn't possibly choose any other alternative--should suffice. Or maybe "computer experiments," as Stephen Wolfram intimated last year in A New Kind of Science, can replace measurements. According to a leading science historian, such a subtle but ultimately sweeping philosophical shift in theory justification may already be underway.

Still, it does sound pretty impressive when you hear that most scientific papers accept that global warming is happening, and is largely man-made. Or is it still so impressive when you read something like this?

In 2004, history professor Naomi Oreskes performed a survey of research papers on climate change. Examining peer-reviewed papers published on the ISI Web of Science database from 1993 to 2003, she found a majority supported the "consensus view," defined as humans were having at least some effect on global climate change. Oreskes' work has been repeatedly cited, but as some of its data is now nearly 15 years old, its conclusions are becoming somewhat dated.

Medical researcher Dr. Klaus-Martin Schulte recently updated this research. Using the same database and search terms as Oreskes, he examined all papers published from 2004 to February 2007. The results have been submitted to the journal Energy and Environment, of which DailyTech has obtained a pre-publication copy. The figures are surprising.

Of 528 total papers on climate change, only 38 (7%) gave an explicit endorsement of the consensus. If one considers "implicit" endorsement (accepting the consensus without explicit statement), the figure rises to 45%. However, while only 32 papers (6%) reject the consensus outright, the largest category (48%) are neutral papers, refusing to either accept or reject the hypothesis. This is no "consensus."

The figures are even more shocking when one remembers the watered-down definition of consensus here. Not only does it not require supporting that man is the "primary" cause of warming, but it doesn't require any belief or support for "catastrophic" global warming. In fact of all papers published in this period (2004 to February 2007), only a single one makes any reference to climate change leading to catastrophic results. [emphasis in original]

I can see some potential problems with this analysis. Papers that are neutral might not bother to take a position because they don't feel a need to do so. I think I can safely assume that a survey of papers published in astrophysics journals would find that they are overwhelmingly neutral about the shape of the Earth; but this doesn't imply an absence of consensus that the Earth is flat. (Well, yes, there is the Flat Earth Society web page, and it doesn't seem to be a parody, but I doubt that its members are publishing papers in astrophysics journals.)

Still, there is some controversy about the extent (if any) of anthropogenic contributions to global warming, as the papers that I mentioned here demonstrate. It does seem that there is a genuine question about whether this scientific consensus still exists. If Schulte truly used the same methodology as Oreskes--and the consensus Oreskes found no longer exists, then I can see two real possibilities:

1. If Oreskes' finding of a consensus was valid, and Schulte uses the same methods today and finds the consensus isn't there, then perhaps the consensus has evaporated.

2. If Schulte's use of Oreskes' methodology is defective for determining consensus or its absence, then perhaps Oreskes finding of consensus was defective, too.

UPDATE: More about the history of the dispute here.