Embrace uncertainty, and tell us about it
Today sees the publication of the last of three inquiries into the famous release of emails from the (CRU), late last year. (The previous inquiries were by the and a Science Assessment Panel led by Lord Oxburgh ().)
The a former Scottish civil servant, will surely include a further call for scientists to be more open and transparent about their methods, by publishing source code for the computer software that they use, as well as the original data. This much has already emerged as a theme in previous reports.
As Darrell Ince, professor of Computing at the Open University, told Newsnight, there's no reason these days why all of this cannot be stored in an easily accessible form on the web, for anyone to read and make sense of. This would help to make it straight forward for outsiders to verify research, or not. And that, after all, has been the way that the best science has worked for a very long time. Why should climate science be an exception?
There will undoubtedly be a call too for greater acknowledgement of uncertainty, of nuance in findings - and for that to be clearly expressed. I know I keep banging on about this, but it's an emerging mantra for the increasing number of areas where science rubs up against policy making (see ). And I suspect that the conclusions of the third inquiry into "climategate" will sound a note along these lines.
Comment number 1.
At 7th Jul 2010, barriesingleton wrote:PROPER SCIENCE HAS NO TRUTH ONLY RELATIVE PROBABILITIES
Meanwhile news media do not seek truth - only sensation. In this 'climate' reality is lost to the mass, and only those outside the world of funded science, yet possessed of scientific reasoning, have any chance of drawing valid conclusions.
The email argument is froth.
Complain about this comment (Comment number 1)
Comment number 2.
At 7th Jul 2010, vagueofgodalming wrote:there's no reason these days why all of this cannot be stored in an easily accessible form on the web
No doubt that's what will happen now, but it won't end there, and it won't end well. People will 'discover' that the code is not written to the sort of standards that would be expected of, say, the software that operates a nuclear power station, and start to nitpick all sorts of aspects (I suspect, particularly, they'll start demanding documentation that doesn't exist).
The scientists will say "sort it out yourselves", the critics will say "but we're not scientists (funny how they're 'citizen scientists' when they're criticising, but 'just regular folks asking questions' when any real work is required), we're just auditing for the sake of good science", and the scientists will point out that the point of releasing information is so that other scientists can repeat the work as a whole, not duplicate every step individually.
None of which would matter (after all, the scientists might benefit from some of the nitpicks), except there'll be another bunch of critics pointing at the nitpickers and saying "See! This proves climate science is bunk!"
There is no good faith in the criticisms - the aim is not to improve climate science, but to stop it influencing policy.
Complain about this comment (Comment number 2)
Comment number 3.
At 7th Jul 2010, Neil Robertson wrote:"There will undoubtedly be a call too for greater acknowledgement of uncertainty, of nuance in findings - and for that to be clearly expressed. I know I keep banging on about this, but it's an emerging mantra for the increasing number of areas where science rubs up against policy making (see swine flu story from last week). And I suspect that the conclusions of the third inquiry into "climategate" will sound a note along these lines." (Susan Watts)
Keep banging on about this Susan .... the other recent example where 'nuance' might well have led to a different decision was
in the decision to release Megrahi given that medical opinion
was subsequently revised by one of the doctors to estimate his
life expectancy at anything between 3 months and TWENTY YEARS -
having previously persuaded Ministers that 3 months was likely.
Of course in that case, the other real doubt related to forensic
evidence that convicted him in the first place which may now not
be tested again
And keep banging on about the need to label graphs properly and give sources for data .....
Complain about this comment (Comment number 3)
Comment number 4.
At 7th Jul 2010, jauntycyclist wrote:science be 'open source'? like linux? outrageous. can't make money out of that idea.
Complain about this comment (Comment number 4)