By my definition 24 of the 13,950 articles, 0.17% or 1 in 581, clearly reject human-caused global warming or endorse a cause other than CO2 emissions for observed warming. The articles have a total of 33,690 individual authors (rounded to 33,700 in the figure). The 24 rejecting papers have a total of 34 authors, about 1 in 1,000.
What can we conclude from this study?
1. In the scientific literature, there is virtually no disagreement that humans are causing global warming.
2. The authors of the handful of papers that reject human-caused global warming tend not to agree with, or even to cite, each other's work.
3. Other than those authors themselves, only a handful of other scientists cite the few rejecting articles. Those who do cite them do not themselves necessarily reject human-caused global warming.
4. The rejecting authors have no alternative theory that satisfactorily explains the observed warming.
5. Some of the most vocal global warming deniers (for example, Lindzen, Monckton, Singer, and Watts) have never written a scientific paper in which they say explicitly that human-caused global warming is false. Why? Because if they did, they would have to come up with the evidence to back up the claim, and there is no such evidence. If there were, surely they would have told us by now.
6. The vast majority of climate scientists accept the theory that human emissions of greenhouse gases are causing the observed global warming. Here is how I arrive at this deduction.
When a new scientific theory is first proposed, scientists often go out of their way to state explicitly that they reject it, or that they accept it. This was the case with continental drift in the 1920s, with plate tectonics in the 1960s, and with the Alvarez theory of dinosaur extinction in the 1980s. One reading the literature in these fields can usually tell from the title of an article alone whether an author rejects the new theory. But after a theory achieves maturity and becomes the ruling paradigm, scientists no longer see any point in stating explicitly that they accept the now-no-longer-new theory. They take it as a given, often as an observational fact—like the measured movement of tectonic plates and the measured global temperature rise. To explicitly endorse the ruling theory would have the counter-effect of suggesting that the theory is in doubt. It is obvious that global warming has achieved the status of the ruling paradigm of climate science. Thus it is reasonable to assume that publishing scientists who today reject human-caused global warming would make it clear that they do so, while publishing scientists who accept the theory would not feel the need to say so explicitly. As a practical matter, virtually all of the global warming papers that Oreskes and I separately reviewed can be classified as about effects, mitigation, adaptation, methods of detecting, climate modeling, and paleoclimatology. Authors of these papers would hardly be likely to deny the existence of the very thing they are writing about. It is theoretically possible that a paper on paleoclimatology could be the exception, dealing with the lack of evidence for CO2-driven global warming in the geologic past, say, leading the author to question the seriousness of modern, human-caused global warming. I did not find such papers. What we know for a fact is that among the authors of peer-reviewed articles, only a tiny fraction, which I estimate as about 1 author in 1,000, rejects human-caused global warming. In my opinion, based on my understanding of the history of science, it is reasonable to conclude that nearly all publishing climate scientists accept that human activities are causing the Earth to warm. Polls of scientists reinforce this conclusion, but polls are no substitute for the primary, peer-reviewed literature, the ground truth of science.