I think that many people believe that a cause and effect relationship between human activity and elevations in global temperatures over time has been unequivocally demonstrated. That is not true. Inferring a cause and effect relationship on the basis of statistical data requires controlled experimentation and, of course, controlled experimentation to infer a cause and effect relationship in the referenced case is not possible. It should be noted that the IPCC concedes that point in its 2007 Working Group 1 Report, The Physical Science Basis of Climate Change. I'd post a link to the report but I don't have 15 posts yet and this site doesn't allow posting links until a person gets to that point. But you can find it at the IPCC web site. So find it, pull up the pdf version of Chapter 9, Understanding and Attributing Climate Change,, then go to page 660 (the sixth page of that chapter). Or you can pull up the web page version and do a search for the word "experiment." In the pdf version, you'll find the relevant language starting at the bottom of the page in the left column. It goes like this: "Attribution of causes of climate change is the process of establishing the most likely causes for the detected change with some defined level of confidence (see Glossary). As noted in the SAR (IPCC, 1996) and the TAR (IPCC, 2001), unequivocal attribution would require controlled experimentation with the climate system. Since that is not possible..." After the word "possible" the authors go on to explain what they did given the situation. But that is not important to the point. The situation is that unequivocal "attribution" is not possible. Otherwise, I don't agree with their use of the terminology "level of confidence." The concept of a confidence level with respect to inferring cause and effect is intimately tied to the concept of a controlled experiment. You can't really have a credible confidence level with respect to cause and effect without the experimental control, and I think the IPCC is misleading the public in representing probability statements it makes as confidence levels. I think it gives the statements an aura of quantitative credibility that is not really justified. There are other references in the report to controlled experimentation being needed to infer cause and effect, but the one I quoted is the most direct and unambiguous instance of conceding the principle that I could find. Finally, note that there are a lot of references to "experiments" in the report that are not experiments on anything real. They are experiments on models. They tell the investigators what doing different things will make the models say. They're great for infering cause and effect with respect to what the models do. But they can't be viewed as legitimately inferring cause and effect with respect to the actual climate system.