Mere speculation reported as established scientific fact is the overarching flaw in the wildly alarmist Climate Science Special Report (CSSR) of the US Global Change Research Program. This flaw could also be called the Fallacy of False Confidence.
This fallacy is very simple and the CSSR does it uniformly. Take purely speculative results found in a scholarly journal article and report them as established scientific facts. Even better, supplement the factual looking statement with a parenthetical confidence level — very high, high, medium or low — which is also stated as a fact. This confidence level approach raises a host of issues and pernicious confusions. These are what my old professor would call “weasel words.” They deliberately convey a false meaning.
Here we consider just a few of the issues:
1. Whose confidence? When a seemingly factual statement is given a confidence level, exactly whose confidence is being reported here? There is a great range of possibilities, which carry very different weights, especially in a policy context like this.
It might be as small as just the author or authors of that specific statement. Or it could be all of the authors or just those who finalize the report. In these cases the weight is quite small, just a few people who have a certain confidence in the statement. “Who cares?” one might ask, especially since the CSSP authors are clearly extreme alarmists.
Or these confidence judgments might be attributable to the USGCRP Coordination Office, which is running the project. Or even to the people in the 13 USGCRP member agencies that reviewed or approved the report. Or to these federal science agencies themselves, or even to the federal government per se. Are these confidence judgments federal policies or just the opinion of a few people? It makes a very big difference, especially if students are taught that this is the official government word.
2. Dissenting opinions? That everyone has the same level of confidence for the hundreds of claims and statements that are ranked in the report is not credible. In reality many of these statements are highly contentious, so we would expect a lot of disagreement over the confidence levels. The more people involved the greater the disagreements.
There is no hint of these disagreements in the CSSR. Instead the confidence levels appear magically out of the mist, as though they were some kind of objective measurement of something, like the weight of a ham.
3. No confidence in speculation. When it comes to science, speculation is not something to put a level of confidence on. Hypotheses are meant to be explored and tested, not put to a vote. Polling a handful of people is not science, so the result is not a meaningful measure of anything except what they believe. This is especially true if there is widespread disagreement, which is certainly the case with most of the CSSR.
4. What does medium confidence mean? A great many statements are labeled “medium confidence.” Does medium confidence mean there is a 50-50 tossup chance that the statement is true (according to those polled)? Is it like a 50% chance of rain? If so then surely the proper thing to say in these cases is that “We do not know.” Instead the CSSR states this unknown as known, but with just medium parenthetical confidence.
This really makes no sense at all. If it is maybe, maybe not, then all we are really being told is that we do not know. But it reads like a statement of fact, not of ignorance. This is bound to mislead people, especially students who do not know about the debate.
5. Missing language of science. In particular, the report fails to use the established scientific language for expressing uncertainty. Scientific writing is typically very careful about not overstating certainty. This is done by using terms like suggests, is possible, might be the case, is plausible to think that, maybe, hypothesize that, etc.
The lack of this scientific uncertainty language makes the report easy to exaggerate when quoted or paraphrased. Dropping the parenthetical expression of limited confidence turns the sentence into an absolute truth claim. The statement “A is B (medium confidence),” which means “A might be B,” instead becomes “A is B,” which is a much stronger claim and not what is actually said.
For that matter terms like “high confidence” and “very high confidence” are merely reports of some author or group’s opinion. In the case of the CSSR these people are clearly extreme alarmists, which make their opinions scientifically irrelevant.
The whole report is junk of course but these weasel words paint the junk with camouflage.