Skepticalscience – Rewriting History

“…resist the temptation to reply to [trolls].
Instead, do what the troll hates most — simply remove the comment.”
— John Cook

The recent censorship episode at the skepticalscience.com brings an often overlooked aspect to the forefront. The target of deletion Prof Roger Pielke Sr, runs a blog. The actions of Skepticalscience were revealed because he posted them there.

What if a scientist or a lay person, interacted with websites like Skepticalscience and did not have a blog?

Consider what Skepticalscience did in reader Paul and AnthonySG1′s cases. In 2007, the website had an article explaining Antarctica’s cooling —a thorn in the pitch for a clean story about global warming— as an “uniquely” regional phenomenon. It talked of how ‘Antarctica was overall losing ice’, citing a peer-reviewed paper Velicogna et al 2003 for support.

The response in the comments section from Cook’s readers was simple: ‘Antarctic ice is increasing. You cannot take a paper that has three years worth of data and conclude that the continent was losing ice’. They cited references that Skepticalscience neglected – which showed an overall increase in Antarctic sea ice.

The rewriting that John Cook undertook is now recounted at Bishop Hill. In the first step Cook changed the entire article, taking off from the criticisms. Next, he deleted his original ‘responses’, and added new ones that made it appear as though these commenters did not know what they were talking about.

The rewriting of Skepticalscience history

After this was openly revealed, John Cook offered explanations for his actions. It went something like this: ‘I accidentally mistook my readers to have responded to my updated article. Thinking that was indeed the case, their comments sounded silly to me. So I ended up adding responses to guide new readers’

A closer examination of the threads on Skepticalscience, reveals a different picture. Let us begin by examining a few examples to get a sense of what these might be.

Let us start with the thread “Climate models are unreliable”. As is known, the website portrays skeptical arguments as such simple statements and offers rebuttals. The article was published sometime late 2007.

In July 2008, ’poptech’ left a comment which questioned assertions made in the article. He quoted scientists at the Realclimate consensus blog:

Comment from reader 'poptech' - deleted in 2011

From mid-2008, Poptech’s comment remained intact on the thread till as recently as Feb 2011 . At some point afterward, the comment was deleted. Another of poptech’s comments upthread, to which three commenters responded (example) was deleted, leaving the responses hanging mid-air.

Consider the thread “Solar activity & climate: is the sun causing global warming?” In Sept 2007 reader Ben Lankamp left a comment supporting Cook’s assertions on the thread. It contained a graph of total solar irradiance. The comment remained intact for close to three years till Aug 2010.

Inexplicably, at some point afterward the comment was deleted. The next snapshot shows the article with Lankamp’s comment missing and changed graphs.

Was Cook deleting comments that didn’t play well with his changes? It is hard to say. Even innocuous ones seem to get the boot. For instance, observe once again Antarctic ice thread of 2007.  Reader ‘irkantska’s comment from Aug 2009 is still there in Sept 2009:

Look at the Jan 2010 entry next. The comment is gone. By this time, Cook has rearranged the conversation and added admonishing responses to readers who had provided links. He did not have any replies for irkantska. The comment was simply bumped off.

Is Cook merely deleting comments from his readers alone? Let us look at the Hurricane Katrina weblog entry titled ‘Did Global Warming cause Hurricane Katrina’. The changes made by Cook start getting complex.

Cook’s article on Hurricane Katrina came out in 2007. It claimed that global warming caused an increase in hurricane intensity. By November 2007 reader ‘Wondering Aloud’ (WA) had offered his comment,

WA, as can be seen, joked about the article’s title and said that climate science theory predicted a decrease of hurricane frequency and intensity with global warming. Cook asked if WA ‘knew of any papers’ that showed this.

WA offered a citation in the next post.

‘Wondering Aloud’ noted the irony of Kerry Emmanuel himself being the prime proponent of a theory of increasing intensity of hurricanes, when he was a co-author of the very Free et al 2004 that predicted no change in intensity.

You can look at the thread today. Comments 2 and 4 are gone. Comment 3 has been left behind and its content doesn’t make any sense (it was a response to Cook’s comment in 2). And presumably since comment 4 was deleted comments 5, 6, 7, and 8 by readers responding to WA’s point in #4, have all been deleted. They were present until Sept 2009. They were gone by Dec 2009.

If a reader stumbles on the Katrina thread at Skepticalscience today, he or she would not know any of the above. Instead, what one sees is a comment, with a response from Cook. The changes make it look as though ‘the science’ changed from 2004 to 2005 which Cook is helpfully pointing out.

The whole conversation has been turned on its head.

What’s more, Cook has gotten rid of his own comments in the process. In 2007, a more thoughtful Cook began his response to WA:  “Statements such as “Katrina was caused by GW or GW causes more hurricanes” are on shaky ground”.

These remarks are wiped clean from the historical record.

In a recent letter to a local newspaper editor, global warming professor Michael Mann guided its readers in Colorado to skepticalscience.com as an authentic source of information. He asserted that the true story about his hockey-stick graph was told there. Let us examine what skepticalscience.com did with their hockey-stick threads.

One of Cook’s earliest pages on the hockey stick came out sometime late-2007. There were hardly any reader responses, and by November 2009 the thread had accumulated a grand total of eleven comments. In Nov 2007 commenter ‘nomann’ had disagreed with Cook’s contentions, with Cook’s response following close behind. Commenter ‘Will Nitschke’ posted a series of comments pointing out aspects of the hockey-stick issue that were left out, and again Cook responded. Then notably, ‘saluki’ left a comment that drew attention to researcher Linah Ababneh’s dissertation, the proxy weightings used by Mann, and the problem with the stripbark phenomenon that affected Mann’s work. In early 2009, reader ‘sjkhayes’ posted a numbered summary of the issues with Mann’s work and inquired if there were any proxy reconstructions that were free from these problems.

In Jan 2010, all these comments were gone.

The page changed as well. McIntyre and McKitrick’s graph of a corrected-hockey stick, which Cook had used from a Michael Crichton page, was gone.

A fresh round of commenting began on the same thread.

By May 2010, in line was commenter poptech again, who posted a list of papers and reports which refuted, in part or in whole, the Mannian hockey-stick.

John Cook was obliging. He posted a response:

Poptech’s comment survived on the page till as recently as May 2011. On the current version of the page, it is gone. Again, the responses to poptech were left hanging in the air.

So, the problem is not confined to just a handful of comments, here and there. Whole batches of them are deleted at times. Is a strange computer glitch wiping out comments in bulk on the skepticalscience.com server?

Let us look at another topic on Skepticalscience that relates to Michael Mann’s paleoclimate work – the Medieval Warm Period.

The page came out sometime before October 2007 and this is how it looked in Sept 2009. Compare that with the current version which is here.

Cook has deleted virtually dozens and dozens of comments on this thread. None of them appear to be abusive, or offensive, or ‘ad-hominem’. The changes are simply too many to be adequately documented with screen captures. The comments are from late 2007 – early 2008 and the deletions occur somewhere between Sept 2009 and Dec 2009.

Take the exchange between ‘Adamski’ and ‘chris’ (comments 36, 37, 38, 39 originally):

Comments from 'Adamski' and 'chris' as they appeared in Sept 2009

Nov 2009 - the Adamski-chris conversation moves up due to bulk deletions! Comment #37 from chris goes missing

Feb 2010 - Comment #37 makes a comeback but chris has lost his name.

Sept 2011 - the comments as they are, in their final position

What is more: as can be seen from the screen captures above, Cook goes into the comments and deletes commenters’ references to each others’ posts. This is no computer glitch and it demonstrates he knew what he was doing.  Nor does this square with the explanations Cook provided at Bishop Hill. . Again, as before, parts of a conversation are deleted and altered in such a way, the end result looks like something that never happened.

Let us look at yet another page that relates to Michael Mann’s paleoclimate work – the ‘trick’ to ‘hide the decline’.

Scientist Phil Jones’ email about his using ‘Mike’s Nature trick’ in order to ‘hide the decline’ is famous. However, even the ‘trick it seemed could be defended by skepticalscience. Their initial page, published sometime before Mar 2010 was customarily simple and suggested that the whole thing was a non issue.

By Dec 2010, author James Wight had extensively updated the page. This is how he explained Jones’ actions:

Reader ‘JeanS’ now added the first comment, referring to the key claim in the last sentence:

John James Wight clearly understood his reader’s critique because he took its message to heart.

The main article was rewritten deleting all previous references to what exactly the “trick” was. Instead of the claim that the hiding involved the truncating of ‘unreliable’ tree ring data – which was the basis for claiming that no malfeasance occured, one now sees simply a proclamation from the Muir-Russell review to the same effect. The highlighted claim is missing from all three versions of the article.

Just as in instances before, JeanS’ comment has left hanging in the air referring to a statement that doesn’t exist.

Why does John Cook do this?

The deletions carried out by Cook don’t make sense as an exercise in moderation. They seem driven by an ardent need to present a clean and neat view of global warming. Of a need to reassure that no intelligent discussions exist, and all possible questions have (long) been answered.

The structure of Cook’s website appears to push things in his direction. In the beginning, pages are born as undemanding and easy arguments. Cook then seems to realize that the skeptical arguments are more involved and complex than the simplistic picture he presents. He updates the same pages with more detail. But messy comments have accumulated below the line, sticking out like sore thumbs. The ‘broad picture’ that Cook so wants to convey is sullied.

In the meantime fresh readers, oblivious to the confusing mish-mash of claim and counter-claim, arrive in greater numbers on the shores of the global warming debate. Journalists, policy-makers and other influential opinion-makers land up everyday at skepticalscience, looking for a quick grasp on the consensus position in climate issues. How does one protect these newcomers?

Cook’s solution: the inconvenient comments go flying out the window.

One clearly sees that the mission of the website underwent a change ~end of 2009. In the earlier years, Cook seems welcoming to comments. His interest it seemed was to point out findings from scientific papers, that he thought contradicted climate skeptics’ claims. By November 2009, Cook had arrived at a dramatically different viewpoint. He saw ‘global warming skepticism’ as a sort of a mental illness or a psychiatric condition, with the afflicted being beyond any hope. Psychologic diagnoses permeates his thinking from that point on.

Cook voices his thoughts on the shift in a post in November 2009. It is hard to fathom, why, anybody who ran a website and worked hard at attracting and nurturing an online community, would commit the most fundamental of indiscretions with his readers’ comments – deleting and moulding them at his own whim.

As seen in his response above, Cook viewed the comments section of his website topics as a resource, to be used for ‘educating’ the public.

From there on, editing, deleting and moulding the historical record probably did not seem any wrong to Cook.