Lewandowsky's Logic

It’s Lewandowsky time, again. ‘Are you a poor logician? Logically, you might never know‘, he observes with Richard Pancost over at the Conversation Nonversation. More about that logic shortly…

At the Nonversation, of course, comments from Lewandowsky’s critics have been removed. Meanwhile, Andy West has a three part series (one, two, three) over at WUWT examining Lewandowsky’s work in considerable detail. Lewandowsky and Pancost’s (‘L&P’ hereafter) new logic, the deleted comments and the exhaustive critique of his work reflect in microcosm the climate change debate.

It has been argued here before that the amazing thing about Lew is not simply the low quality of his work, but the failure of academe to act as a check on it. The professor of psychology makes bold claims. He believes that he understands the entire world’s relationship to the natural world. He believes he understands the natural world, and professes expertise in climate science. And he believes he knows how society should be organised. Surely he is a true Renaissance Man… A polymath… A Renaissance Polymath… Or he is an epic blowhard.

The task of taking Lewandowsky et al to task has fallen to bloggers (and latterly José Duarte). It is bloggers who have identified the problems with Lew’s claims, and tried to bring them to the attention of a largely indifferent academic community, who don’t seem at all willing to hold an academic to account.

Perhaps this is understandable. The climate debate is hostile. And in spite of claims that bloggers are the agents of this hostility, Academics can be found making statements — such as the idea that influential climate sceptics ought to face the death penalty — which don’t exactly serve to cool the atmosphere, much less shed any light on the matters of debate.

But the trouble for those academics is the fact that, if the academy’s standard is so low as is necessary to admit Lewandowsky’s work, it says something about the standards of every discipline that Lew comments on, from psychology outwards, to climate science. If the academy cannot check itself, its population have no business acting as a check on society, and on power in particular. The academy is redundant.

This not to say that academe’s rightful role is to be the check on society, or to supply policy-makers with the closest possible approximation of the the Truth. But those roles are what the academy seems to increasingly sold itself as in recent years.

Of course, someone might say that the academy — for the first time since scholasticism was eschewed — really had settled on a consensus. But a more likely explanation is that a political settlement — dogma — was established, either spontaneously or by design. The most political arguments made in the climate debate by academics are made by those pronouncing on the ‘psychology’ of climate change ‘denial’, just as psychologists were able, per the political orthodoxy of the era, to deal with political dissidents in Soviet Russia. Whether or not a scientific consensus on climate change exists, and whatever the substance of that consensus is, academics have been slow to realise that ‘climate change’ has an ideological form, no matter how well grounded in science it is. Head-shrinking your political opponents is as political act as any form of apartheid. It delimits the putative subject’s political and civil rights, and confers to the psychologist political authority over that subject.

Academic resistance to that observation results in hostility towards not only those who might dare utter it, but towards the public in general. Lewandowsky, amongst others, set about overcoming the impasse by belittling sceptics, primarily, but also the faculties of the wider public, and thereby to elevate academics. The public must STFU, said Lewandowsky, or challenge his work through the ‘proper channels’ — i.e. academic publishing. When that happened, and a paper he had authored was eventually withdrawn, he accused those he had previously accused of ‘conspiracy ideation’, of organising a conspiracy against him.

Because I value freedom of speech and academic freedom, I oppose and resist the bullying and intimidation employed by some opponents who refuse to engage in scientific debate by avoiding peer review. My thoughts and experiences are summarized in an article on the Subterranean War on Science.


In no way do my values suggest that debate should be curtailed: I merely insist that a scientific debate should take place in the scientific literature and that the public be put in a position where it can make an informed judgment about the voices that are opposing mainstream science on crucial issues ranging from climate change to vaccination.

The problem for Lewandowsky is that if the observation that academia and its institutions have been colonised by political environmentalism is true to any extent, there would exist a barrier against dissent passing peer review. And there is good evidence that this is the case. There are entire academic institutions and university departments given over to a particular view of climate change, and of promoting that view in the public sphere. There is almost no possibility of substantive criticism emerging from these institutions, or through academic publishers, the editors and peer reviewers of which hail from those organisations, and whose editorial policies are equivalent to a lobbying organisation’s. Lewandowsky was calling for debate to be curtailed. That is what it means to create ‘ethics’ which preclude the unwashed masses from the sphere of public debate, and that is what is meant by his studies into the ‘psychology’ of named people he disagrees with and those who have the audacity to disagree with him. That is exactly what Lewandowsky did when he used dodgy statistical techniques to attack his critics, and what he did when he hid his shameful political ideas behind dodgy mathematics

Lewandowsky and Pancost’s (L&P) new argument is that…

It is an unfortunate paradox: if you’re bad at something, you probably also lack the skills to assess your own performance. And if you don’t know much about a topic, you’re unlikely to be aware of the scope of your own ignorance.

Is this true? I cannot play the banjo, which as anyone who has seen this scene in Deliverance knows, it looks like it would be fun to be able to play. Yet I am very confident indeed in my ability to assess my own performance. If I were to attempt a performance, it would be as obvious to my own ears as it would be to any expert’s. Yet L&P counter:

Ignorance is associated with exaggerated confidence in one’s abilities, whereas experts are unduly tentative about their performance. This basic finding has been replicated numerous times in many different circumstances. There is very little doubt about its status as a fundamental aspect of human behaviour.

This is obviously bullshit, or no progress would have ever been made in any field in which it is possible to be an expert, for the simple reason that, if confidence is inversely proportional to ignorance, nobody would ever have developed the inclination to advance their understanding.

L&P refer to the Dunning-Kruger effect, which claims that people who do less well in tests of their knowledge over-estimate their performance. Say L&P:

People whose logical ability was in the bottom 12% (so that 88 out of 100 people performed better than they did) judged their own performance to be among the top third of the distribution. Conversely, the outstanding logicians who outperformed 86% of their peers judged themselves to be merely in the top quarter (roughly) of the distribution, thereby underestimating their performance.

Ignorance is associated with exaggerated confidence in one’s abilities, whereas experts are unduly tentative about their performance. This basic finding has been replicated numerous times in many different circumstances. There is very little doubt about its status as a fundamental aspect of human behaviour.

The claim, of course, is that sceptical climate change experts’ confidence belies their expertise. However, L&P’s own argument defeats them. Kruger and Dunning’s observation applies to the first and second quantile of their study’s participants, not to people with knowledge. Even the unnamed subject of L&P’s broadside — Anthony Watts — has extensive knowledge of the concepts in the climate debate, whether or not he counts, on their view, as a an expert. One could not, for example, devise a study such as Watt’s Surface Stations project and formulate a hypothesis about the recording of errors in the temperature record without such knowledge. And the same holds for individuals with established expertise in climate science, like Richard Lindzen, Patrick Michaels and John Christie, who are routinely vilified by Lewandowsky’s colleagues.

L&P extend the point. The problem, they claim, is that a putative expert’s confidence is fundamental in our evaluation of their level of expertise.

Does this mean that the poorest-performing — and hence most over-confident — expert is believed more than the top performer whose displayed confidence may be a little more tentative? This rather discomforting possibility cannot be ruled out on the basis of existing data.

If this is true, it has more worrying implications for L&P’s own argument than it has for climate sceptics…

there is an overwhelming scientific consensus that greenhouse gas emissions from our economic activities are altering the Earth’s climate. This consensus is expressed in more than 95% of the scientific literature and it is shared by a similar fraction — 97-98% – of publishing experts in the area. In the present context, it is relevant that research has found that the “relative climate expertise and scientific prominence” of the few dissenting researchers “are substantially below that of the convinced researchers”.

The “overwhelming consensus” might well be, on L&P’s admission, nothing more than the product of so many self-deceived experts’ over-estimation of themselves, except for their caveat that experts are “those who publish in the peer-reviewed literature in their area of expertise”. It’s a surprising thing for a Chair of Cognitive Psychology at University of Bristol and a Professor of Biogeochemistry, Director of the Cabot Institute at University of Bristol to admit about debates about climate science, in a non peer-reviewed journal such as the Nonversation. L&P still defeat themselves.

The more substantive problem with the argument, if we take it at face value, is it’s own inability to understand the terms of the climate debate. L&P serve, again, as an object lesson in ‘the consensus without an object‘. As is discussed in the previous post, journalists tripped over their own ignorance of the debate. The idea of consensus preceded their knowledge of the consensus. So when their knowledge of the arguments in currency did advance, it appeared to them as a change of argument in the climate debate. The facts were plain: the journalists didn’t know what the consensus was, nor what the argument of the sceptics was, and so they didn’t notice that the putative sceptics’ arguments were not in fact outside of the consensus at all.

L&P play a similar game. Equally there is no object — no substance — to the consensus they propose. It means nothing to say “there is an overwhelming scientific consensus that greenhouse gas emissions from our economic activities are altering the Earth’s climate”, because the statement is not quantified. This is the point also discussed in the previous post. If L&P quantified the consensus, they would be forced to notice that the putative sceptics’ claims, also quantified, do not exist outside of the consensus. The consensus is broad.

L&P’s logic is a binary logic applied to a debate to which binary axioms do not apply. Even at its most simple, the debate about global warming is a debate about a property of the Earth, which is a question of degree. The ‘consensus’ position encompasses a range of estimates, produced by research, none of which is true (or closer to the truth) by virtue of its proximity to others. To reduce the debate to a matter of binary logic is make statements equivalent to the claim that ‘the economy is true’, or ‘Wednesday is purple’.

L&P’s omission of nuance is startling, given their claim to be concerned about misinformation by ignorant non-experts. I don’t think they’re ignorant. I think they do it deliberately. If I have over-estimated their intelligence, I apologise to them.

8 thoughts on “Lewandowsky's Logic”

  1. A smelling error: “The “overwhelming consensus” might well be, on L&P’s admission, nothing more than the product of so many self-deceived exerts’ over-estimation of themselves,”

    I’ve also reproduced my own article: “http://scottishsceptic.co.uk/2014/11/06/bush-blair-lied-about-wmd-so-lewandowsky-lied-about-skeptics-the-motivated-lies-attacks-intended-to-discredit-honest-people/”.

    Basically it seems to be that Lewandowsky reasoning went “because sceptics are conservatives and because conservatives lied about WMD, it is therefore legitimate for Lewandowsky to lie about sceptics … and because people remember the lie and ignore any FORCED corrections … Lewandowsky incorrectly reasoned that people would stop listening to sceptics.

    So, e.g. his great strategy would mean an end to Republican rule in the US and the end of UKIP in the UK.

Leave a Reply

Your email address will not be published. Required fields are marked *