Comments on Prometheus

RP wrote:

In the discussion motivated by Oreskes' Essay, I have seen one claim made that there are more than 11,000 articles on "climate change" in the ISI database and suggestions that about 10% somehow contradict the IPCC consensus position.

This claim is so wacky I couldn't even place it. RP didn't seem to either. It turns out (see Peisers reply to my comment (side note: is it weird that Peiser is supplying the references for Pielkes articles?)) that this is ref'd to Timo Hämeranta, another std.septic, and it isn't even correct: even TH only claims to have 4,000 abstracts, and I would guess that his categorising ability is no higher than Peisers. What exactly is the point in giving prominence to stupid claims like this one, except to give them credibility?

OTOH, RP *does* say:

Like Oreskes, I am happy to take the IPCC as the best assessment of state of climate science, and its conclusions as an accurate measure of the central tendency of views among the climate science community. The work of the IPCC, including its certainties and uncertainties, is plenty good enough for the development and promulgation of a steady stream of policy options on climate

which is confirmation of his non-skeptic status, should one ever doubt it (ahem).

But RP then continues:

But so what? If that number is 1% or 40%, it does not make any difference whatsoever from the standpoint of policy action.

This is wrong. If 40% of papers doubted the consensus, there would be no consensus.


Anonymous said...

William- Thanks for these comments. A "consensus" is a measure of central tendency; of agreement. We could very easily quantify such agreement and say that only the 5% tails of the distribution disagree with the consensus. Or we could say that every perspective outside of one sigma represents disagreement. The choice of what constitutes "disagreement" is entirely arbitrary, and the IPCC has done this e.g., in the creation of its range of temperature sensitivity. There are scientifically sound studies that would suggest numbers outside of this range (e.g., see the SRES stroylines), but the "consensus" was created as a subset of this range. My point is simply that policy ought to consider the full range of scientific views, and not a subset. Hence how one choses to define the "consensus" is less important then the characteristics of the complete distribution of views.

William M. Connolley said...

I partly agree. But "agreement with the consensus" isn't obviously a thing with a metric in this fashion, so I don't think it can very easily be quantified. All normal distributions have a midpoint, but we're not talking about such a thing in this case. It is most certainly *not* true that what constitutes disagreement is entirely arbitrary - read Peisers categorisation of abstracts, & you'll find that he is just wrong in a number of cases (and admitting ambiguity in some others).

If 40% of things disagreed enough to be in the "disagree with consensus" position, I still think there would be no consensus.

I very much doubt that policy is capable of dealing with the full range of views, unless those get weighted by their probability/support, in which case we're back at the consensus.

Anonymous said...

William- The consensus perspective of IPCC contributors, like any collection of expert judgments, can of course be quantified. For an climate change example see:

Morgan, G.M., L.F. Pitelka, and E. Shevliakova, 2001. Elicitation of Expert Judgments of Climate Change Impacts on Forest Ecosystems, Climatic Change, 49:279-307.

The methods recommended by Morgan et al. are decidedly not the same as what Oreskes and Peiser have done (neither of which give us much useful information for policy, but plenty of grist for political wrangling).

Policy making is quite able to deal with a distribution of expert opinion. See e.g.:



Lempert, RJ and Schlesinger, ME 2000, Robust Strategies for Abating Climate Change, Climatic Change 45:387–401.

More thoughts along these lines here:

Pielke, Jr., R. A., 2001: Room for doubt. Nature, 410:151.