Can one count a consensus?

One of the most repeated claims about global warming is that there is a consensus between scientists that man is causing the current warming. Some papers tried to prove this claim. The last one was a survey by John Cook et al. I became interested in this survey when it came online and several blogs where discussing the method. It seemed odd to me to prove a consensus of scientists by making a survey to evaluate titles and abstracts of papers. I was looking forward to the moment the paper would be published. That day is today.

First, let’s look at the (predictable) findings of this the paper.

We analyze the evolution of the scientific consensus on anthropogenic global warming (AGW) in the peer-reviewed scientific literature, examining 11 944 climate abstracts from 1991-2011 matching the topics ‘global climate change’ or ‘global warming’. We find that 66.4% of abstracts expressed no position on AGW, 32.6% endorsed AGW, 0.7% rejected AGW and 0.3% were uncertain about the cause of global warming. Among abstracts expressing a position on AGW, 97.1% endorsed the consensus position that humans are causing global warming.

Wow. How did they come to this 97% to begin with? Well, simple.

First they removed the number of abstracts with no position on AGW. Then they calculated the percent from the total of the rest. Et voilà: a majority where there wasn’t one before.

Endorsement level	Abstracts	Original percent	Final percent
No position	7,930	66.4	0
Endorsing	3,896	32.6	97.1
Rejecting	78	0.7	1.9
Uncertain	40	0.3	1.0
Total	4,014	100	100

From their full selection of abstracts there wasn’t a consensus, not even a majority unless one ignore most of the data.

Some random thoughts…

There is a strange twist of definitions in the paper

In fact, they looked at abstracts of scientific papers and evaluated if they endorsed AGW or not. Most abstracts were very short anyway (they selected on abstracts with less than 1000 words) and it was difficult to conclude if there was an endorsement or not.

Abstracts expressing “no position” are not only abstracts that do not endorse or reject AGW, but more probably abstracts “from which the position of AGW could not be determined from the title and the abstract”. Calling it “no position” implies it was clear that the paper didn’t make an endorsement or rejection. Most possible it meant “Not possible to determine”: it is not possible to conclude what the paper meant from the title and abstract.

The same with endorsement. This doesn’t mean the paper endorses AGW, it does mean that the evaluator concludes from the title and abstract that the paper endorses AGW. And also the same with rejection. It doesn’t mean the paper rejects AGW, but that the evaluator concludes from the title and abstract that the paper rejects AGW.

Science doesn’t work by consensus

Politics do. Consensus is a political thing and is used in the decision making process. Science isn’t run by raising hands, thank goodness! I can imagine that many scientist could lean over to the same theory. But in a complex issue as the climate it would be doubtful to get most, let alone all, scientists on the same line.

Consensus is not majority

• Consensus: a decision that can be supported by all the participants, even those that would prefer to support another decision. Consensus is not reached if even one member of the group is unwilling to proceed with a decision that he or she cannot support.
• Majority: the largest group, at least 50%.

In this case, they were not trying to find a consensus, but a majority. Not a majority of all abstracts, but a majority of a subset of all abstracts and only from those from which a position could be determined by the evaluator.

In the end, is there is a consensus among scientists that man is largely responsible for the recent warming? I really doubt it and to my humble opinion this paper will not be able to prove it.

Advertisement