Heated discussions

The leaking, or perhaps hacking, of hundreds of emails from the servers of the Climatic Research Unit at the University of East Anglia late last year has thrown the already turbulent world of climatology into turmoil. The significance of the emails is hotly disputed, but sceptics of the so-called consensus position allege that they contain evidence of the undermining of the peer-review process, attempts to pressurise journals, the withholding of data and code from outsiders, and at least one episode of the manipulation of results.

The accusations and denials will fly for months to come. So far, no fewer than five inquiries have been announced into various aspects of what has come to be known as Climategate, and some of these will not report until the middle of the year. However, regardless of the outcome, the affair raises ethical issues that will be of interest far beyond the narrow confines of climate science. Some of the most important concern the world of academic publishing.

Among the most serious allegations to emerge in the wake of the leaked emails is that CRU scientists tried to "nobble" scientific journals that accepted papers from sceptics. There are suggestions in the emails that as many as four different journals may have had their normal procedures interfered with.

One particular series of emails dating back to 2003 is a case in point. The story unfolds in messages exchanged by some of the most prominent names in climatology, including Michael Mann, the author of the famous "hockey stick" paper, Phil Jones, the CRU director who has stood aside in the wake of the Climategate affair, and Mike Hulme, at that point the director of the Tyndall Centre for Climatic Change Research and now a prominent media commentator on climatology and its policy implications.

In the messages, the scientists discuss how to deal with the recent publication of a paper in the journal Climate Research by Willie Soon and Sallie Baliunas, two prominent climate-change sceptics from the Harvard-Smithsonian Center for Astrophysics. Mann and his team speak of encouraging colleagues to stop treating Climate Research as a valid publication for scientific findings, and of "getting rid of" both the paper's editor, Chris de Freitas, and the journal's editor-in-chief, Hans von Storch. Another plan involved trying to precipitate a mass resignation from the editorial board. In the event, four of the journal's editors did in fact resign in protest, including von Storch.

Clearly readers who object strongly to elements of the content of a journal have the right to protest - the readers of, say, the Journal of Evolutionary Biology could not be expected to accept the publication of articles about creationism. Few would see the submission of a letter of complaint to the editor as crossing any ethical boundaries in these circumstances. However, when more aggressive measures are adopted by unhappy readers, the nature of the protest and the circumstances that precipitate it must give journal editors pause for thought.

While intelligent design may be objectionable in a scientific journal, Soon and Baliunas' paper was clearly within the boundaries of the scientific method - a valid contribution to the literature, albeit perhaps a flawed one. In these circumstances, how should a journal editor respond to a letter of protest? A single letter may be shrugged off, but what about a campaign of letter writing? This may appear much more like an attempt to impose an orthodoxy than a valid protest. There are, after all, well-established approaches to dealing with the publication of flawed papers, namely the submission of formal comments and critiques.

There are other ways of registering a protest, too. Readers can simply threaten to take their subscriptions elsewhere, and perhaps their authored contributions, too. There is an interesting hint of this sort of behaviour elsewhere in the Climategate emails from 2007 onwards, where scientists discuss the possibility that the International Journal of Climatology might accede to sceptics' requests and force the disclosure of all research materials, including intermediate results.

A discussion ensues in which there appears to be an agreement that this would be unacceptable and that mainstream scientists should refuse to publish in the journal if such a policy were put in place. This brings us to another somewhat nuanced question: what is an ethical way to deal with a journal?

Clearly scientists are free to read whatever journals they like and to publish wherever they choose, but is there a difference between cancelling a subscription and organising a campaign of threats to do so? If it is valid to choose to publish in a different journal, is it still valid to inform a journal that the most prominent practitioners of the specialism it covers will shun it if it does not toe the line? At what point does valid protest elide into something more sinister?

In the case of the International Journal of Climatology, it appears that while the journal was approached regarding its policy, there is no evidence that any threats were ever made. It is perhaps worth noting, however, that at the time of writing, the journal has still to finalise its policy on data sharing.

Issues of the availability of data and computer code have been a constant bone of contention between sceptics and mainstream climatologists. While newspaper headlines have been dominated by what appear to be attempts by CRU staff and their associates in North America to delete correspondence requested under the Freedom of Information Act, the more important story in terms of the conduct of science in this country concerns the repeated refusals of CRU staff to release the data and code underlying their global-temperature index.

As far back as 2005, Jones rejected a request for the data, telling Warwick Hughes, a sceptic and self-described "freelance earth scientist" with several peer-reviewed publications to his name: "Even if (the World Meteorological Organization) agrees, I will still not pass on the data. We have 25 or so years invested in the work. Why should I make the data available to you, when your aim is to try and find something wrong with it?"

Similar obstruction is revealed in attempts by Steve McIntyre, a sceptic and the editor of the blog ClimateAudit.org, to obtain the data underpinning the famous Yamal tree-ring chronology, which was published by the CRU's Keith Briffa and became a critical ingredient in most of the important global-temperature reconstructions.

Having had a direct approach turned down by Briffa, McIntyre approached Science, the journal in which the series had most recently been published. It excused itself by saying that the chronology had been published earlier in a different article, and suggested that McIntyre approach the author of the earlier publication. With the author being Briffa, continued non-disclosure was a foregone conclusion. When the journal in question, Quaternary Science Reviews, also turned down a materials request, a dead end was reached.

Only in 2008, when a third journal finally enforced its own data policy and made Briffa release his figures, was it revealed that parts of this critical dataset were based on just a handful of trees, raising major questions over the data's reliability and role in important public policy decisions.

Is it valid to refuse to release research materials to opponents? It is known that the CRU temperature dataset was sent to a sympathetic researcher in the US just months before a request for the same data from a sceptic was turned down on the grounds that confidentiality agreements prevented their release. In this light, the CRU's claims that the data are confidential look far-fetched.

Sceptics are universally of the opinion that the scientific method requires all research materials to be released to friend and foe alike, but the Climategate emails suggest paranoia among some mainstream climatologists - a sense that sceptics were on a campaign to do them down. This appears to have enabled them to justify to their consciences a steadfast refusal to provide information to their opponents.

If, as the emails suggest, some scientists are in fact putting illegitimate pressure on journals, either to influence the peer-review process or to prevent the release of data, it is easy to see how editors may find it difficult to respond. In the face of a threat by the most prominent scientists in any specialism to shun a given publication, many would surely capitulate.

But given the centrality of replication to the scientific method and of climatology to political policy decisions, a way must be found to ensure that data and code are universally available. If journals were to present a united front on the issue of the availability of materials, it would be a valuable start.