The analysis of speech

What constitutes legitimate analysis of speech?

This question has been brought into sharp focus by the most recent position of the journal Frontiers that they put out last Friday. This statement claimed that our paper Recursive Fury (, had been retracted because it “did not sufficiently protect the rights of the studied subjects.” This seemingly stands in contrast to the contractually-agreed retraction statement, signed by legal representatives of the journal and the authors, that Frontiers “…did not identify any issues with the academic and ethical aspects of the study.”

It is helpful that the Frontiers affair involved two contrasting ways in which speech was being used by the various participants. Let us therefore analyze those two ways in turn.

The complainant(s).

Although we have destroyed all correspondence and documents involving the allegations against us at the request of Frontiers, and although now, a year later, our recollection of those events is minimal, Graham Readfearn has put something about the allegations into the public domain that has received little attention to date.

Readfearn states that the complaints against us alleged “malice” on the part of the authors in various ways. As far as I understand it, malice is a legal term meaning an improper motive in making a statement; and, if proved in Court, removes some defenses to charges of defamation.

In the present context, it is most relevant that the accusations of malice against John Cook, one of the authors of Recursive Fury, were based on his apparent sanctioning of “vile commentary” against the complainant and other bloggers.

Indeed, the material cited in support contains irate statements that none of the authors of Recursive Fury would countenance.

None of the authors made those statements.

One will fail to find anything like those comments on Cook’s blog, None of the more than 88,000 public comments posted there to date contain anything that could be remotely construed as vitriolic or polemical—that’s because 7000 comments were deleted by moderators owing to their inflammatory content.

So where did the “vile commentary” come from and how did John Cook “sanction” it?

The vile commentary was made by third parties unconnected to Recursive Fury on a private forum that was password-protected, and whose purpose was to permit open and completely uncensored discussion among a small group of collaborators. Those comments were posted in the expectation of privacy, and they became public only through a criminal act—a hack attack on Skepticalscience that has been explored in great forensic depth.

John Cook neither wrote those comments, nor could he be reasonably expected to moderate them. They were made in private and became public by an illegal act by parties unknown.

What John did was to host a private forum on which other people vented their anger. If that is malice, then so would be inopportune comments by your friends at an illegally wire-tapped dinner party. You better censor what your guests say in case your next party is bugged, lest you be accused of malice.

The complainant’s conduct follows a common pattern in the Subterranean War on Science: Use of private correspondence obtained by an illegal act to construct allegations against scientists. Except that in this case, to allege malice against John Cook, hackers trolled through two years of his private conversations and found exactly nothing.

Zip. Zilch. Bupkis.

All the hackers and trolls could find were other parties expressing anger in the expectation of privacy. I cannot think of clearer evidence for the absence of malice in John Cook’s conduct.

I nonetheless think there might be evidence of malice here.

Maybe some readers can spot it.

The authors of Recursive Fury.

Recursive Fury was conducted with ethics approval (of course!) and Frontiers entered into a contractual agreement for the retraction that noted that their review “…did not identify any issues with the academic and ethical aspects of the study”.

And what did Recursive Fury do? It presented a narrative analysis of public discourse in the blogosphere in the aggregate. We did not categorize anyone into anything, we categorized statements.

That’s all.

This is the difference between saying “Joe is a racist” and saying “When Joe and Fred get together in a bar at night their discourse contains racist elements based on application of the following scholarly criteria.” Now, we could have withheld the sources of all those statements, thereby anonymizing the analysis and protecting the identity of those who feel that their public statements are too fragile to survive scholarly scrutiny.

However, we considered this unwise in light of the pervasive allegations against (climate) scientists that they are “hiding data.”

Folks, we did not hide the data.

We made them all available. And they are still here:

By the way, there are ample precedents for this kind of work, including other hot-button issues such as anti-Semitism. Yes, there is a scholarly paper out there that analyzes the public speeches of contemporary Austrian politicians for their anti-Semitic undertones. (I am not linking to that study here, lest the researcher be caught up in the turmoil of requests for his/her data, or requests to destroy the data, or requests to provide ethics approval, or his/her entire email correspondence during the last 13 years.)

Here then is the crucial question about the analysis of speech that arises from the Frontiers affair:

Are public statements by people who knowingly made them in public, subject to scholarly analysis? Or is it only stolen correspondence by third parties made in the expectation of privacy that can be used to allege malice on the part of someone who never said anything malicious himself?

In Whose Hands the Future?