Claims of Claims of Certainty

The argument about excessive certainty from partisans has come up frequently, and has recurred at Curry’s. I think the argument is without merit. I have never seen excessive confidence from the scientific sector about climate. Everybody admits that there is much we don’t know. The possibility of “unpleasant surprises in the greenhouse” has long been acknowledged.

Let me summarize once again the “uncertainty is not your friend” argument.

Despite popular perception, there is no debate in physical climatology about first order processes.

You don’t find Lindzen or Christie or Spencer or even Michaels questioning the accumulation of CO2 or the existence of a CO2 forcing. You have to sink even lower than Singer to find someone saying something like that.

What is going on in physical climatology that bears any resemblance at all to the public vision of it is arguments over an estimate of a quantity that must exist (the sensitivity) and its confidence bounds. (Typically, the sensitivity is expressed as degrees C equilibrium warming per doubling of CO2.)

A soi-disant “skeptic” climate change naysayer can be relied upon to implicitly or explicitly make the following case about the physics:

  1. the mainstream greatly overestimates the sensitivity
  2. uncertainty is large
  3. policy should therefore act as if the sensitivity is small and the uncertainty is small

It’s my belief that the case from their 2 to their 3 is so incoherent that the argument can be defeated there without much reference to science.

(Recall Judith Curry’s claim that the temperature sensitivity to CO2 doubling is between zero and ten C.It seems to me that Dr. Curry does not act coherently with a belief that 10 C is a plausible sensitivity! This puts us at enormous risk. Suppose the middle were excluded, and we had merely a 50/50 chance of a sensitivity of 0 and a sensitivity of 10 C. Wouldn’t the 10 C outcome dominate our rational behavior? Nevertheless it’s fair to say that Curry’s fan base embraces her fascination with uncertainty and somehow associates it with a no-policy stance.)

It is possible to argue endlessly over point 1. However, the so-called skeptics habitually make the policy case themselves by advocating point 2. Insofar as stopping the futile nitpicking over science in the public sphere and moving on to reasonable policy is concerned, they have already given us the argument we need.

The position to take in the policy sphere is that either the mainstream is right OR the uncertainty is large. A modest understanding of risk weighting can carry the day from there.

Quoting myself from 1999:

  • It is clear enough that a change of 20 C would be cataclysmic, whether that change is a warming or a cooling. It is also clear enough that a change of 0.2 C is of little consequence, and may be on net beneficial. There is no reason to believe that this function is linear. The contemplated changes (~2 C) are large enough that we can not have total confidence that the impacts will not be catastrophic.
  • There is no plausible argument that any particular climate change will have a beneficial impact comparable to the worst plausible case negative impact.
  • The risk-weighted cost of unrestrained anthropogenic perturbation must therefore be dominated by the fact that the plausible worst cases have more cost than the plausible best cases have benefit.

If we think about the situation fairly, Curry does us (those of us who want a sane policy) a great favor by playing up uncertainty.

It is not necessary to defend the science to get to a sane policy. It is enough to say “if you argue that the science is weak, you must admit that the uncertainty is large” and that “there must be some number that represents the sensitivity”. Then proceed directly to risk arguments.

Yet so many people get this backwards. In order to defend status quo/laisser faire emissions policy, you must advocate for very low uncertainty. You must be absolutely confident in a scientific position that is not in fact in line with most experts. Otherwise the risk remains high.

I’ll have more to say about Steven Mosher’s recent article at Curry’s, and I suspect some of the more postnormally-oriented may have their own responses, but I wanted to thank Mosher for acknowledging that I made the “uncertainty is not your friend” argument long ago. Being the slacker that I am, I am not especially attached to getting credit for things, but this may have been my most important single contribution. I appreciate Eli’s efforts in working to establish this, and Mosher’s recognition of it.

Here I just wanted to repeat the main thrust of it, and reassert my claim that it is not mainstream science that has been making claims of excessive certainty, implicitly or explicitly. All we are claiming is that it is very hard (and getting harder) to exclude a very large risk. If there is excessive confidence, it is coming from the critics.

There is nothing magical about climate. Like anywhere else, more uncertainty means more risk.

Image: Judith Curry at the keyboard via UCAR



  1. This from Mosher summarizes why it's a complete waste of time to try to interact with him:

    "While the uncertainty about facts of sensitivity are accepted in this path of argument the certainty is shifted to certainty about values and certainty about impacts. In short, the argument becomes that while we are uncertain about sensitivity the certainty we have about large impacts and trans-generational obligations necessitates action."

    Yep, he's very happy to argue with you about risk. Endlessly.

    And in the same thread Ravetz got a chance to show what a piece of work he is:

    As to the corruptions, they are quite clear. When the political imperatives dominate over the internal criteria of quality, then the enterprise becomes corrupt. This is partly because under those conditions the scientists become what Roger Pielke Jr. has called ‘stealth advocates’ – that is deceivers of the public. The example of Lysenko is a good one here. In all his campaigns he needed to fake the field statistics along with the test results. But that was standard practice under Stalin, when everyone had to report a large annual percentage increase in production, or face the consequences. The same sorts of corruptions occur whenever the balance between internal and external criteria of quality is tilted too far in one direction. Those criteria may be industrial, military, policy, ideology, etc.; and the corruption in the other direction occurs when the scientists will not or can not deliver on the promises they made to obtain external support.

    As to methods, I would belatedly start the discussion by saying that in PN situations, the scientists need to be more aware than under normal conditions. For example, in ordinary research there are standard techniques and parameters in statistical testing; these are part of the paradigm, adherence to them is necessary for getting results published; and they are rarely reconsidered. As it happens, the tests are interpreted by the ‘alpha’, which is designed to avoid the error of excess sensitivity; that ensures that tests will exclude possibly spurious results. (How well this quality-criterion works in practice is another issue altogether). Now, in situations where an early-warning would be appropriate, the relevant parameter is ‘beta’, which is designed to avoid the error of excess selectivity. That way the community will not fail to be alerted to possible harms.

    This may be obvious to all the readers of this blog. But among many scientists, even in environmental or policy-related fields, their training in ‘normal science’ has excluded reflection on such issues. They solve the puzzles that they were taught to solve at school, and issues of relevance to real problems are totally foreign to their awareness. It seems to me just possible that some of the climate scientists were wading into those extremely difficult fields of data management with little or now awareness of just how hazardous they are. Then it was too late!

    When I read those excellent discussions on WUWT, where blunders and pitfalls of data management are revealed, I reflect that this indeed is what PNScience is about. There is an extended peer community, whose members cannot be ignored because of their lack of formal qualifications, and who are exercising the best sort of quality assurance on publications in their field. They do not restrict their attention to ‘warmistas’ – the recent discussion of Anthony Watts’s study shows that this community is as rigorous about its friends as about its opponents. What more could one want?

    Sheesh. I'm no psychoanalyst, but I'd say this guy's resentment of climate scientists is palpable.

  2. Your analysis of the implications of large uncertainty is on point, but you are arguing that, because of this, the necessity of certain policy changes is clear. If not, then who cares? But there is risk in these changes as well, and not only to the entrenched "powers that be" (NOT The Powers That Be band, btw).

    It's pretty clear both in my comments here and in my own little blogospherical domain where I stand on the issue of climate change (despite my lifestyle) but it's flawed to fail to address the risk of making whatever policy changes you might implement.

    Take an extreme example: suppose that the only way to have a significant impact on the trajectory of climate would be to reduce the population to, say, 600,000,000 and that those people would need to live a pre-agricultural lifestyle. Then, come what may, almost everyone would say "we'll cope with what comes rather than make such a sacrifice." Then take the other extreme: suppose that other than a big hit on shareholders and executives of some businesses whose income derives from fossil fuel burning, no one would be hurt at all by the transition to a zero net carbon emission economy and the transition was clearly demonstrated to be frictionless. Then I imagine that most everyone would say "let's do it."

    There's no reason to assume that this range of possible costs and our reaction to it would be linear either and we certainly have very large uncertainty as to where we stand on this continuum as well. I think your analysis neglects this aspect to the detriment of your implied conclusion.

    Sorry to parachute in sporadically...

  3. Rob, I see the issue of the costs of mitigation as a discussion to have after there's some agreement on the nature of the problem, so it was fair for Michael to grapple with the latter first. Of course many will deny the problem out of fear of the possible expense of dealing with it, but then it becomes difficult to have any rational discussion with them.

    Note also what a poor job we're doing implementing known "no regrets" stuff.

  4. Howdy back, Michael. You picked up Texas real fast, I reckon. I called some friends in CA the other day, said, "Howdy," and got back "Thanks for the Texas greeting" - and I half a century and eight thousand miles away from Sam Houston's city that 19th century maps called "a malarial swamp."

    I know that the certainty issue is central to your reasons for starting this blog, but how do we convince the lukewarmers even in the face of the catastrophic loss of Arctic ice ("but the Antarctic ice is growing!")? Never mind that Antarctic ice area has tanked since 1940 and grown about as much as housing has recovered since the crash.

    A recent clash with Fabius Maximus leaves me even more pessimistic. Claiming to be the last word in national security, they have bought J. Curry's uncertainty principle hook, line and sinker. If I say DoD believes in AGW, they say ,don't believe DoD, they're just grandstanding for money. If these mostly retired military men are representative of advisors to power. . .

    Gavin (the Gavin? I don't know) even asked what it would take for FM to be willing to undertake action and received a mumbly mumble cost mumble uncertainty mumble answer.

    This comment really needs a closing argument, but I can't come up with one at the moment. I want to say that what you are doing is great, and I would like to be one who can help. . .

  5. Lindzen many years ago said the sensitivity was small and the certainty of it being small was very high (apparently he was a genius surrounded by a confederacy of dunces). One of those "dunces", Steve Schneider, documented this.

    Not sure if that's still Lindzen's opinion.

  6. well steve you would be wrong. First you have to understand that I am more interested in detailing the various rhetorical stratgies involved than I am in telling you my position on policy. More of that later. What I note is simply the tactic ( tactics are good) of shifting certainty from one aspect of the argument to another.
    In the end of course the last shift is the shift to moral certainty. Even if sensitivity is uncertain and even if the stakes are uncertain, we are certain that we have a moral obligation to future generations. The point being that while MT does make a cogent argument for the utility of uncertainty in certain aspects of the argument i think that fundamentally his position rests on accepting some aspect of the situation as being certain. Whereas the other side does doubt shifting. Shifting doubt from one aspect of the argument to another.

    That's just my description of what goes on. you cannot and should not try to draw a conclusion about my policy position from this description. C02 is a danger. That danger may be great. we should act on the assumption of that danger and work for policies that are effective and achievable.

    So much for your prediction that talking about risk with me would be fruitless. At some point you should question your mind reading abilities. At some point you should question whether your approach to engagement is working. You know, the planet is at stake. I would think it's important enough to you that you might engage in a little self examination.

  7. Sorry, mosh, as far as I can see the horns and forked tail remain very prominent while it's hard to make out even the start of any wings.

    Metaphor aside, your abandonment of Watts is cute, albeit not very interesting considering what absolute jokes he and his site are, and the Muller gig is at best tangent to constructive work. And last I checked you're still hangin' with McI and have neglected to retract your half of "The CRUtape Letters."

    On top of that, you're a sophist of the worst sort. I'm not bad at it myself, but I've gone to considerable effort to cure myself of the disease. Seeing others engage in it is deeply irritating.

    So no, I'm just not seeing the value relative to the time commitment of interacting with you. Michael can and will do as he pleases, of course.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.