STEPHAN LEWANDOWSKY

ABC DRUM

Much confusion and spin infects current public discussion of “peer reviewed” research: first we had Maurice Newman, the Chairman of the ABC, who suggested that “distinguished scientists” challenge the overwhelming scientific consensus on climate change by “peer reviewed research”, although he oddly failed to name such research.

Now we have John McLean, an author of a lone article that was celebrated by some media scribes as overturning the scientific consensus on climate change, cry“censorship” because his response to a devastating deconstruction of his work in the peer reviewed literature was not accepted for publication.

So what exactly is peer reviewed research? How does it work?

To understand the current controversy, one must understand that peer review is egalitarian but not indiscriminate; that it is fallible but self-correcting; and that it exercises quality control but not censorship.

Those three attributes of peer review are brought into sharp focus if we examine the recent article by Mr McLean, and colleagues Chris de Freitas and Bob Carter, in the Journal of Geophysical Research.

Mr McLean is a perfect example of the egalitarianism of peer reviewed science.

Authors of peer reviewed articles normally live in the daylight of scholarship and seek maximum visibility—who wouldn’t want to say “Look here, I published this great article and I am at MIT!” Curiously, Mr McLean has proved rather more elusive.

Mr McLean’s published affiliation is given as “Applied Science Consultants” in Croydon, Victoria.

This entity does not have a web page. This entity does not have a recorded phone number. This entity does not have an ABN and it is not registered as a business in Victoria.

The author’s home page can eventually be traced to Switzerland, where it identifies J D McLean as a “computer consultant and occasional travel photographer”.

Does it matter that the author of a scientific article is an occasional travel photographer?

Yes.

But only because it nullifies any claims that the peer reviewed literature is “elitist” and that it favours the academic establishment and “in-group” thinking over the views of outsiders. The fact that an occasional travel photographer can publish in the Journal of Geophysical Research underscores that in science, what matters most is the strength of one’s evidence and arguments and not one’s professional standing.

According to the Australian National Library, no J D McLean has ever submitted a thesis for a research degree at any institution in Australia, for any program in any discipline.

Does it matter that Mr McLean has no verifiable research credentials from an Australian tertiary institution?

No.

I have frequently published peer reviewed papers with students as co-authors before they obtained a higher degree. One must therefore applaud Mr McLean’s entry into the egalitarian world of peer-reviewed science with this first publication.

However, peer reviewed science is not indiscriminate: not all opinions are equal and one cannot choose what to believe on the basis of whim or ideology. What counts are evidence, logic, and competence.

Peer reviewed science is egalitarian but not indiscriminate.

Science carries with it responsibilities such as accountability and subsequent scrutiny—peer review is a spam filter, which works well but not perfectly. The true value of a peer reviewed article lies in whether or not it survives scrutiny upon publication.

If it does not, then peer reviewed science is self-correcting and eventually cleanses the occasional junk that penetrated the spam filter.

The article by McLean and colleagues is a perfect example of the fallible but self-correcting nature of peer reviewed science.

Although the authors loudly proclaimed to the media that their work shows that “no scientific justification exists for emissions regulation” and that it “leaves little room for any warming driven by human emissions”, these claims have now been shown to be wishful thinking at best, and mendacious propaganda at worst.

The Journal of Geophysical Research is publishing a devastating rebuttal of Mr McLean’s work, authored by a team of nine of the world’s leading climate scientists from Japan, the UK, the US, and New Zealand.

This rebuttal uncovered numerous errors and, most crucially, it unambiguously showed that the paper by McLean and colleagues permitted no conclusions about global warming, let alone the lack thereof.

None.

Peer reviewed science is fallible but self-correcting.

In a recent ABC interview, Mr McLean claimed that he was “censored” because his reply to this rebuttal was rejected and thus will not appear in the peer-reviewed literature.

The rejection of this reply is a perfect example that peer review exercises quality control and not censorship.

To understand the difference between quality control and censorship it helps to consider my own experience as editor. I just ended a three-year term as an Associate Editor of a leading journal in my field. During that term, I made 300 editorial decisions based on around 700 peer reviews.

I accepted only 50 manuscripts. Does this mean I censored the remaining 250?

No.

Because in science, a decision against publication is (almost) always quality control and (almost) never censorship.

How can we be certain that the decision to reject a reply by McLean and colleagues was quality control and not censorship?

By analysing their original, now devastatingly critiqued, article.

Fact is that their article was not about long-term global warming. It was about the association between ocean currents and air temperature—in particular the time lag between the warm El Niño current and the ensuing increase in temperature.

Fact is that the article does not contain the words “climate change” except in a citation of the IPCC.

Fact is that the article’s only connection with climate change arises from the phrase “… and perhaps recent trends in global temperature…” in the final sentence.

Those facts are jarringly at odds with the authors’ grandiose claims in press releases that humans do not cause climate change.

Might there nonetheless be a grain of truth about climate change, or lack thereof, in the authors’ public pronouncements?

No.

This is best explained by an analogy involving daily temperature readings between, say, July and December anywhere in Australia. Suppose temperature is recorded twice daily, at midday and at midnight, for those 6 months. It is obvious what we would find: Most days would be hotter than nights and temperature would rise from winter to summer.

Now suppose we change all monthly readings by subtracting them from those of the following month—we subtract July from August, August from September, and so on. This process is called “linear detrending” and it eliminates all equal increments. Days will still be hotter than nights, but the effects of season have been removed. No matter how hot it gets in summer, this detrended analysis would not and could not detect any linear change in monthly temperature.

Astonishingly, McLean and colleagues applied precisely this detrending to their temperature data. Their public statements are thus equivalent to denying the existence of summer and winter because days are hotter than nights.

It is precisely this detrending that was identified by the devastating rebuttal that has now put to rest the paper by McLean and colleagues. (More on the relevant science can be found here, a superb site run by an Australian.)

Thus, in an exemplary demonstration of the self-correcting nature of science, the peer reviewed literature has spoken loudly and clearly: The paper by McLean and colleagues permits no conclusions about global warming, let alone the lack thereof.

None.

The decision to reject a reply by McLean and colleagues, which was made on the basis of three detailed reviews of their work, was thus quality control and not censorship.

It is not censorship to put an end to illegitimate attempts to interpret detrended data as evidence against global warming.

It is quality control.

It is not censorship to put an end to pronouncements based on a botched figure that splices together two discordant temperature series without correction.

It is quality control.

The peer reviewed literature has spoken loudly and clearly: The paper by McLean and colleagues permits no conclusions about global warming, let alone the lack thereof.

This outcome puts to rest the only peer reviewed article that was purportedly about climate change and claimed to challenge the scientific consensus, to have come out of Australia since the IPCC’s 2007 report.

This single article is no more.

What is left standing instead are, for example, the 110 peer reviewed articles on climate change that were published by scientists at the University of New South Wales’ Climate Change Research Center alone since 2007.

Yes, 110 peer reviewed articles since 2007 from just one Australian research center that add to the overwhelming scientific evidence on climate change and its human causes.

110 peer reviewed articles which in the service of humanity seek ways to manage the problem.

110 to 0.

That is the score of the peer reviewed science between just one Australian university and the Australian “skeptics”.

Care to make it 200 to 0? 300 to 0? Just add in a few more Australian universities.

Care to make it umpteen thousands to virtually none? Just read the peer reviewed literature surveyed by the IPCC.

Stephan Lewandowsky is a Winthrop Professor and an Australian Professorial Fellow at the University of Western Australia.