Anysley Kellow is a professor of political science at the University of Tasmania. In an interview five years ago, he said, about global warming, “we’ve got a much broader range of choice to respond to a problem that is much more uncertain than certain people who are pushing the issue would have us believe.”
As in a protection racket, the people trying to scare you benefit from you being scared:
I did a study of electricity planning, including here in Tasmania, the good old Hydro Electric Commission in the old days—and the logic was much the same; they would produce forecasts of [hard-to-meet] future demand which were then taken as immutable, and then they would try and justify particular policy responses to those. In the case here it was with hydro dam construction.
I learned about Professor Kellow’s work from a comment about status-trading among scientists. I wrote to him to ask what work of his was being referred to. He replied:
I think the reference is just to my 2007 book (Science and Public Policy: The Virtuous Corruption of Virtual Environmental Science), where I write about the shrinking size of groups which possess expertise, the effect of the communications revolution in establishing close networks of cooperation, and the effect of this on quality-assurance processes like peer review. The prevailing paradigm then becomes a ’club good’ from the defense of which all members benefit (in status, grant success and career advancement).
The problem is exacerbated by some of the circumstances revealed by Climategate: not just pressure on editors, and influence in being IPCC lead authors, but peer review in climate journals where submitting authors nominate reviewers, the identity of authors is known to those approached to referee papers, and so on. I am so accustomed to double-blind peer review that I found it hard to believe that this was a common practice.
When we add this to the lack of disclosure of raw data and code, we have serious reliability problems underlying science upon which we are basing very costly policy. We know in social science research the potential for subjective factors to obtrude into data manipulation even when researchers do not consciously mean for this to happen, so we often see data preparation and analysis performed by independent teams, and emphasise transparency, disclosure of methods, double-blind peer review, and so on.
That’s a good point about single-blind peer review. I agree, it should all be double-blind, no exceptions. In psychology authors don’t know the name of reviewers but reviewers know the names of authors. You can request double-blind review but then your paper enters the review process with a “paranoid” label attached.
“In psychology authors don’t know the name of reviewers but reviewers know the names of authors.”
I don’t think this is true WRT developmental journals and, I thought, APA journals in general.
Of course, patterns of self-citations and/or citations of unpublished work often make it reasonably clear who the authors might be, but that’s a different problem and harder to solve.
Seth – what do you personally think peer review actually does? Not what it ought to do in a perfect world, but what it actually does.
This would refer to (something like) what the average peer review in the average journal does; or maybe to the minimal peer review in a low ranked but Thomson/ ISS listed journal.
David L Hull (Science as a process, Chicago UP, 1988) – who probably knows more than anyone else about how science functioned up until a couple of decades ago (because he is so smart, knowledegable, and well-connected – and also had done more empirical study and analysis than anyone else) – believed that peer review mostly functioned so that research groupings could prevent the publication of low quality/ dissident work from among their own research group; and this was why peer review needed to be anonymous: i.e. because by PR scientists were actually excluding their friends and colleagues from publications which might embarras or descredit the research program – not because by PR they were excluding their enemies.
What we call science nowadays isn’t really science at all – so modern peer review doubtless has a different function from the golden age.
The reaction to Climategate shows us this – there are two views: the cynical view that Climategate revelations are just normal behaviour for scientists – so that half of scientists are worse than the Climategate mob; and the other view (which is supposedly the moralizing view) that Climategate behaviour is about one standard deviation worse than average – which means that about 16 percent of scientists behave worse than that.
Well, Climategate behaviour is at least two SDs worse than science was when I began working in it 30 years ago (ie only about 2 percent of scientists behaved as badly as the Climategate mob) – which is such a large change as to be qualitative.
And the response to Climategate shows why. Since prestige in modern ‘science’ comes from funding, the mainstream view is that whatever highly funded scientists do is, by definition, OK. (The idea that science is tested by checking its predictions against the real world is left out.)
For me, Climategate is a confirmation that science – mainstream science – is broken; it isn’t really science at all; and if we want real science again, we will need to rebuild it almost from the ground up.
Effects of peer review: 1. Helps editors decide what papers to reject. 2. Increases quality of what’s published (relative to random choice). 3. Provides many suggestions for improvement. 4. Reduces unorthodoxy. 5. Used by reviewers to protect their own reputations. 6. Burdens authors with stupid comments they must answer.
I believe it’s a lot better than random selection in most cases but its effect on stuff that is extremely good may be negative.
“Whatever heavily-funded scientists do is okay.” There is truth to that but it isn’t the whole truth. Heavily-funded scientists must get jobs for their grad students. If they fail to do this it isn’t okay. If over a long time — say, 10-20 years — the heavy funding produces no good results then there is also a problem. That’s what happened to cognitive science. The bloom that was once on cognitive science has shifted to cognitive neuroscience. In the short run, you need merely convince people with money. In the medium run, you must convince colleagues with hiring power. In the long run, you have to convince your whole profession.
That peer review is better than random selection is obvious, but the historically the alternative to peer review has been editorial review; in which the editor decides the content of the journal (using whatever advice from others he thinks appropriate).
Until recently (I’m not sure about now) the top medical journals of which I had some inside knowledge were strongly influenced by the decisions of their editors, who did not use referees for all submissions – i.e. it was a mixed system of peer review (for the dull solid stuff) and editorial review – which was how the heterodox stuff got publish (most of which, naturally, turned out to be wrong; but some of which made the journa;ls reputation over the long term).
My experience is that editorial review is superior to peer review (all else being equal); not just because it does a better job (more efficient etc.), but because peer review degrades science by destroying personal responsibility all round until we get… exactly what we now have – science by committees (i.e. not science at all).
“In the short run, you need merely convince people with money. In the medium run, you must convince colleagues with hiring power. In the long run, you have to convince your whole profession.”
Yes, that it how it used to be. Is that what you have observed in science? It is not what I have observed. If the long term is a ‘generation’ i.e. about 25 years, then it is clear that things have not worked like that.
You mention cognitive neuroscience. This has been dominated by by the techniques of ‘functional’ brain imaging since the early/ mid 1980s – i.e. for a generation. Much of FBI was justified by the supposed medical benefits – these have been essentially zero. The _scientific_ yield of ?billions of dollars invested in FBI has also been approximately zero (ie. at best FBI has been confirmatory of established knowledge).
But the neuroscience professoriate are now pretty much dominated by people involved in brain imaging – it seems. It does not take much to convince them that the way forward is ever-more functional brain imaging.
So, from observation over the past 25 years – I would re-write your para:
“In the short run, you need merely convince people with money to give you large capital grants. In the medium run, the extra productivity and higher prestige research outputs deriving from large capital grants will get your graduate students hired at the best institutions. In the long run, you will re-populate and thereby re-shape the whole profession.”
Bruce, I don’t think it is working out as well as you think for the cognitive neuroscientists. I am told their graduate students have difficulty getting jobs. They have difficulty getting jobs because scanners are so expensive and research money so hard to get. As undergraduates learn this, the field will become less popular. It’s a familiar story line. It happened in animal learning — after 20 or 30 years of overpromising (from the 1940s to the 1960s), the rest of the field (psychology) lost interest and became enthusiastic about something else. From the 1960s to the 1980s, it was cognitive psychology that overpromised. Jobs in that area dried up. Now it is cognitive neuroscience. It may not look like a bubble — but it’s a bubble.
Seth, you ought to check this out. Climategate means nothing to the coterie! They’re still at it — trying to bully and suppress dissent even publicly!
https://rogerpielkejr.blogspot.com/2009/12/climate-scientist-threatens-boycott-of.html
https://volokh.com/2009/12/08/the-homogenized-data-is-false/ — an analysis of the data. The graph displays are amazing.
Figures 7 and 8 are indeed stunners: “homogenizing” in effect changes slight temperature declines into huge temperature increases.
Some good links too.
It is true that status is a benefit that many clubs use as a binding force to holding the club together. But it is lousy glue in most cases. It is the rare club that exists merely to crate status for it’s members. Rather, their primary purpose is always (?) to generate public or club goods of some kind and member labor is directed principally to that end.
Working on status will displace the primary work only if two things are true. First you need the members, most of them, to be isolated. E.g. they must lack a diversified portfolio of other clubs they draw status from. Secondly you need the exceptional situation that inside of the club there is no class structure, i.e. the members have high equality of status. Absent either of these the members can alway call the membership back to the primary purpose of the club.
The idea that the global warming science community is merely defending the status of the community is perverse and self congratulatory on the part of the deniers. The community feels they have won the argument; their concern is now the planet’s survival not the status of their community. And it is projection, it is the deniers who at this point are on the verge of losing face.
Ben, no offense, but you’re full of hot air.