This morning the U.N. Intergovernmental Panel on Climate Change released its "Fourth Assessment Report," but just in the form of a 12-page "Summary for Policymakers." The report itself, about 1,600 pages, will be available only in May. The IPCC explains it needs time to "adjust" the scientific report to make it consistent with its summary.

The summary actually is a semipolitical document negotiated by delegates from 150 governments. Evidently, the IPCC, which prides itself on being strictly scientific and policy-neutral, wants to make its report politically correct.

This raises legitimate doubts about the scientific credibility of the IPCC's conclusions. The "cleansing" of the report — and the attendant delay in publication — is also feeding wild speculation about climate catastrophes, with many leaks to compliant newspapers.

Compared to earlier reports, the "Fourth Assessment" is really quite sober, perhaps because a real scientist less given to ideology heads the effort. The summary projects slightly lower temperature increases than previous reports, for example. Also, the last report, in 2001, featured the Hockeystick, a graph that purportedly illustrated that the 20th century was "unusually warm." Its underlying science was flawed by incorrect statistics, and apparently the IPCC now implicitly agrees, for the Hockeystick does not appear in the summary.

The IPCC's estimates for sea-level rise are about half of previous values given. The IPCC is under attack by extremist scientists who think it is too optimistic and that the numbers should be more catastrophic. NASA scientist Jim Hansen's sea-level value is about 20 times higher than that of the IPCC. I suppose that makes him, as well as Al Gore, a climate "contrarian."

Notwithstanding these more restrained points, the IPCC fails to provide any real support for its key conclusion: "It is very likely that anthropogenic greenhouse-gas increases caused most of the observed increase in globally averaged temperatures since the mid-20th century." The IPCC ignores contrary evidence.

The whole question of anthropogenic, or human-caused, global warming is central to setting any policy of climate mitigation and therefore warrants closer examination.

A commonly cited "proof" for human-caused global warming claims there is a scientific consensus. This claim is based mainly on a flawed essay by Naomi Oreskes of the University of California, San Diego, which appeared in the journal Science in December 2004. But even if a majority of scientists had voted for human-caused global warming, that's not how science works. Unlike in politics, the majority does not rule. Rather, every advance in science has come from a minority that found that observed facts contradicted the prevailing hypothesis. Sometimes it took only one scientist; think of Galileo or Einstein.

Another so-called "proof" offered for human-caused global warming is that glaciers are melting and Arctic sea-ice is disappearing. But this is a necessary consequence of warming and says nothing about its cause. Any warming — whether man-made or natural — will melt ice. Confusing cause and effect is faulty logic, not proof.

Some cite the fact that the climate is currently warming and the level of carbon dioxide in the atmosphere is increasing. This is true, but correlation is never proof of causation. In Europe, the birth rate is decreasing and so is the number of storks. Does this correlation prove that storks bring babies? Besides, the climate cooled for much of the 20th century, between 1940 and 1975, even while carbon dioxide was increasing rapidly.

Well, what about some 20 greenhouse climate models, all predicting warming — all the way to 11.5 C from as low as 1.4 C, for a doubling of atmospheric carbon dioxide? Yet no one can tell us which of these models is correct — if any. And none of these models can explain why the climate cooled between 1940 and 1975 — without special assumptions. In any case, model results are never evidence. Only actual observations count.

Crucially, greenhouse models cannot explain the observed patterns of warming — temperature trends at different latitudes and altitudes. These data, published in a U.S. government scientific report in May 2006, lead us to conclude that the human contribution is not significant. Most of current warming must therefore stem from natural causes. It may well be part of an unstoppable solar-driven 1,500-year cycle of warming and cooling that's been documented in ice cores, ocean sediments, stalagmites, and so forth — going back a million years.

If indeed most of current warming is natural rather than from greenhouse gases, there is little point in reducing carbon-dioxide emissions. Further, carbon dioxide is not an atmospheric pollutant. Programs and policies for carbon dioxide control should therefore be scrapped — including uneconomic alternative energy sources, carbon-sequestration efforts, and costly emission-trading schemes. All of these waste money and squander scarce resources, without in any way affecting the atmosphere or climate. Humans have adapted to major climate changes in the past, and we should have no problem doing so in the future.

Mr. Singer, professor emeritus of environmental sciences at the University of Virginia and former director of the U.S. Weather Satellite Service, is the author with Dennis Avery of "Unstoppable Global Warming: Every 1,500 Years."