In the latest American Prospect magazine, I’ve got a review of a fascinating new book called The Polluters: The Making of Our Chemically Altered Environment, by Benjamin Ross and Stephen Amter.
What’s important about this book, as I write, is that it’s a history of “the American struggle for environmental protection before the triumphal 1970s, when Congress passed the Clean Air Act and many other landmark environmental laws.”
If you think corporations behave badly today when it comes to seeking to avoid regulation…well, you have no idea how far we’ve come, and what model citizens they are compared to, say, what they were the 1920s. That’s the invaluable historical perspective that The Polluters provides.
But as I write, while things have changed, they’ve also stayed the same:
…polluters have continued to use the same basic techniques to undermine regulation. Ross and Amter label the most effective strategy “spill, study, and stall”: If you don’t want to stop polluting, just insist that the science is uncertain and there’s no basis for action. Cook up a few questionable studies that reanalyze the data, divert attention to other possible culprits, or call for new research. The tobacco industry didn’t invent these gambits; as Ross and Amter show, the chemical industry used the same techniques to fight the regulation of tetraethyl lead in the early 20th century and the regulation of air pollution in the 1940s. The agenda is the same in the current climate debate.
In sum, we have nearly a century of evidence that the same science game is played over and over and over again. Isn’t that enough evidence to warrant concerted steps to stop it? I end my review thusly:
There’s no doubt from this saga that we still need strong government regulation: 100 years of experience shows that companies cannot be trusted to regulate themselves. But we can go further. We probably also need more explicit sanctions to prevent science from being cynically used to stall public policy — the research equivalent of filing frivolous motions in a courtroom. The prostitution of science is much too easy. It happens far too often. And at this point, the evidence is overwhelming that it’s a systematic strategy that industry will continue to employ unless there are penalties to be paid.
You can read the full review here, and order Ross and Amter’s The Polluters here.
September 4th, 2010 at 1:44 pm
“The prostitution of science is much too easy. It happens far too often.”
Really? So what you’re saying is – we shouldn’t trust science?
Ha! I’ve got a few more examples for you.
Vaccinations make a lot of money for pharmaceutical companies. When their safety has been called into question, science has ‘cooked up’ studies to re-analyse the data and claim that there is no basis for action.
Power lines carry electricity through inhabited areas for the electricity companies. When the safety of this was questioned (“Currents of Death”) the electricity industry funded research to test it (at great expense), and found no connection.
When the radiation from cell phones was connected to cancer and neurological complaints, the cell phone industry commissioned research, and found there was negligible evidence for it. It was all in their imagination.
When the organic food brigade connected pesticides and GM crops to cancer and long term health risks, the agricultural industry did research to show that there was no significant risk, and the safety testing was rigorous.
In the 50s and 60s, old people said this new “rock-and-roll” music would lead to a corruption of public morals, and a breakdown in society’s behavioural standards. Anarchist and nihilist lyrics would cause crime, mental illness, and even suicide. (You may remember some famous court cases regarding certain ‘heavy metal’ acts.) The music industry commissioned studies to show that it wouldn’t.
And so on.
The problem is that you can’t tell a priori which side is is the “real” science and which the industry-funded delayer. This always happens. A new industry starts up and starts to change things, and dozens of people will crawl out of the woodwork to protest about its dangers. The industry will object that the protesters have no evidence. Sometimes they’re right, and sometimes they’re wrong. But if you “take concerted steps to stop it”, if you assume a priori that because they’re industry and have a financial interest they’re automatically in the wrong, you will enable people to shut down any industry they don’t like, for any reason, without evidence. And they will.
Social and technological progress will come to an end. The forces of Conservatism will be able to maintain the status quo for all eternity.
There is a class of logical fallacy known as the ‘ad hominem’ argument. Contrary to popular belief, this is not merely an argument that insults the opposition, (the ‘ad hominem abusive’). Ad hominem means that you argue the truth or falsity of an argument or conclusion on the basis of characteristics of the person making the argument.
Thus, arguing on the basis of conflicts of interest, funding, political beliefs, employers, or the other unrelated beliefs held constitute ‘ad hominem’ argument. Likewise, arguing that a scientist is ‘expert’, independent of industry, or similar and should therefore be believed. Any judgement should be based solely on the content of the argument and evidence – and while conflict of interest is good reason to be especially careful in checking those arguments, it is in no way reason to reject them. Nor to punish said scientists afterwards should they turn out to have been honestly mistaken.
A lot of the success of pseudo-scientific mistrust of science has been built on this widespread acceptance of ad hominem dismissals of industrial science – for all the advances and quality of life improvements it has brought us. And yet the absolute authority of science that conforms to ones prejudices is accepted without question – in seeming defiance of the earlier stated belief that “The prostitution of science is much too easy. It happens far too often.”
Yes. It does. The wrong side has won far too often, too. That’s precisely why we always have to examine all science suspiciously, examine the arguments and evidence carefully, and try to find its flaws. Only through the failure of a well-motivated, well-resourced attempt to find any flaws, in circumstances where we would expect them to show up if the hypothesis was in fact false, can we gain any confidence in the conclusion.
The same goes for claims made both for and against industry, both sides using the same sort of tactics. The adversarial system helps us in finding the truth. (Or whatever approximation to it that science is capable of.) Regulation needs to be based on solid and unassailable evidence, not on anecdote, pseudo-scientific fads, and cenophobia.
September 4th, 2010 at 3:05 pm
Hello Nullius in Verba,
very well said!
We had a model case for science “prostitution” in Germany.
It was the war between butter (farmers, diary industry)
versus oleomargarine (colonies, copra, Unilever and competitors, whaling)
which lasted from about 1900 to the 1980ties.
Each side hired professors to write books or papers
on which is “better”, “more natural” or blaming the others
product to be “artificial” and so on.
Truth was very simple: butter was much more expensive,
so it was a matter of income first. (At least during
the first 60 years of that war).
Regards
Georg
September 4th, 2010 at 7:49 pm
The problem is that you can’t tell a priori which side is is the “real” science and which the industry-funded delayer.
There are *some* cases where you can’t–no one argues there aren’t grey areas. But it’s easy to find ones that are black and white. Shake the trees in the tobacco industry and you find tons of indiscriminating memos that showed they knew exactly what they were doing. You don’t lose a bunch of multi-billion dollar law suits for no reason.
September 5th, 2010 at 11:20 am
Maybe there should be a well publicized saw to the effect, Extraordinarily well subsidized claims require far more than extraordinary skepticism.
The record of industry and commerce is that they will tell highly financed lies and sell them to the public through the media just about every time. They do it every time, that record makes the promotion of ideas by industry, commerce and their proxies unbelievable without the most extraordinary and unsubsidized evidence.
In a week where Hawking’s declaration has gotten massive attention, this is, by far, the more worthwhile science related story.
September 5th, 2010 at 11:56 am
#3,
Ah, yes, the memos. Much like Climategate, yes?
The tobacco one is quite interesting, because I’ve seen quite a lot of evidence that both sides practised (and still practise) such methods – not so much a case of ‘black and white’, as ‘black and black’. I don’t condone any deception in research.
But even with tobacco, you couldn’t decide a priori that they were being dishonest just because they were an industry being faced with prohibition – you needed to examine the arguments and evidence (and memos) to find out. Good science relies on industries funding work to defend technological change and prevent unnecessary regulation, just as it relies on anti-industry advocacy groups funding work to challenge industry and bring regulation about. You need both. You have to work to keep both honest.
The problem comes when the advocacy is presented as if it were impartial science. The problem comes too when you pre-judge which side is ‘real science’ by which conclusion or body you advocate. Examine the evidence.
September 5th, 2010 at 1:12 pm
Much like Climategate, yes?
No, not much like Climategate. With climategate, angry things were said by researchers in private emails (you’ve never said anything angry in a private email?) but accusations of twisting research results were shown as false, by the media when they bothered to research, and by inquiries by the research institutions that climate scientists worked at.
But even with tobacco, you couldn’t decide a priori that they were being dishonest just because they were an industry being faced with prohibition – you needed to examine the arguments and evidence (and memos) to find out.
For the people who were victims of the tobacco company’s deception, once we learned what was going on, it was too late, right? And as Chris is saying above, this is a kind of deception that is being used again and again. There is a discernible pattern. The only pattern discernible with climate researchers is that very small problems (in some cases non-problems) are given intense scrutiny, but the overall body of evidence keeps piling up–as the IPCC reported levels of certainty show, and as the body of research itself shows.
September 5th, 2010 at 1:43 pm
BTW tobacco “industry”
is tobacco grown in plants or plantations?
I cannot speak for USA, but in Germany the big
money with smokering is made by the state,
“earning” about 90 % of the cigerettes price
as tobacco tax.
Georg
September 5th, 2010 at 2:42 pm
Georg (@7)
Here in America we have mostly forgotten that “big business” is really just very large collections of our neighbors. If anything it’s about the same as “big government” – ethics become secondary to current needs as seen by the participants (despite what people think). You can see quite a lot of the farce when you look at local\state financing and realize they are dependent on taxing the hell out of all sorts of things they want everyone to “quite” – except if everyone actually did quite they would be in a world of financial hurt… Thus they are very often on the same side in private – the argument is about where the money goes…
Nullius in Verba,
Well said, as always.
September 5th, 2010 at 3:35 pm
Here in America we have mostly forgotten that “big business” is really just very large collections of our neighbors.
Of course “big business” isn’t inherently bad. But my neighbors don’t exist to maximize profits for legions of anonymous shareholders. The structure of accountability is a bit different, between my neighbor and a large corporation (which tends to have large, consequential operations and lawyers, consultants, and lobbyists on retainer).
If the truth seems to oppose something a corporation wants to do, it’s not because the truth has an anti-corporate agenda. It’s just the way the world works (or should work). Sorry to make an obvious point, but sometimes you have to be constrained in what you do if there are going to be bad consequences for human beings.
September 5th, 2010 at 3:58 pm
“With climategate, angry things were said by researchers in private emails (you’ve never said anything angry in a private email?) but accusations of twisting research results were shown as false, by the media when they bothered to research, and by inquiries by the research institutions that climate scientists worked at.”
If it was just a matter of angry, it wouldn’t have had anything like the impact it did. The media’s approach to research was to ask the scientists what they meant by a particular statement, and take the answer at face value. And none of the inquiries have investigated the science (and they say so).
But I’m genuinely interested in how people have come to the conclusions they have about Climategate, so I’d appreciate it if you told me some more about your views.
First – have you actually read many of the emails in question? Have you ever read the Harry_read_me file? And where did you get your information from to come to the conclusion that you did?
Now, just as a thought-experiment, I’d like you to try to imagine that the following comments were found in the tobacco industry memos. If that were the case, would you conclude that they were evidence of no more than being ‘angry’?
Regarding some data that showed the residuals from a temperature reconstruction were “red” – meaning they had a strong low frequency component, a classic sign that an incorrect statistical model has been used to fit the data, and the result of the calculation is therefore likely to be inaccurate:
“p.s. I know I probably don’t need to mention this, but just to insure absolutely clarify on this, I’m providing these for your own personal use, since you’re a trusted colleague. So please don’t pass this along to others without checking w/ me first. This is the sort of “dirty laundry” one doesn’t want to fall into the hands of those who might potentially try to distort things…”
Regarding data about which stations were used to form a reconstruction, necessary to check the maths – note, this was sent long before any FOIA requests were submitted. Note also, deleting information subject to a request is illegal.
“The two MMs have been after the CRU station data for years. If they ever hear there is a Freedom of Information Act now in the UK, I think I’ll delete the file rather than send to anyone.”
Similarly with this one.
“I wouldn’t worry about the code. If FOIA does ever get used by anyone, there is also IPR to consider as well. Data is covered by all the agreements we sign with people, so I will be hiding behind them.”
Regarding a peer-review of a paper that showed both theoretically and practically (by Monte Carlo methods) that the methods used by dendroclimatologists to reconstruct temperature were biased and incorrect:
“It won’t be easy to dismiss out of hand as the math appears to be correct theoretically, …”
Concerning an allegation of research fraud made by a mathematician called Doug Keenan against Phil Jones and Wei-Chyung Wang. They claimed in a paper, which the IPCC relied upon to demonstrate that UHI did not invalidate their results, that a set of Chinese weather stations had been selected having relatively few, if any, changes in instrumentation, location, or observation times. Doug Keenan showed that most of them had no records existing to act as a basis for such a selection (not surprising given China’s history), and of those that did most of them did indeed have such changes in location. Tom Wigley, director of CRU at the time of the events, said:
“Seems to me that Keenan has a valid point. The statements in the papers that he quotes seem to be incorrect statements, and that someone (WCW at the very least) must have known at the time that they were incorrect.”
Regarding the question of why Wang claimed for over a year to have documents supporting his case, but when asked if we could see them proved unable to produce them:
“The next puzzle is why Wei-Chyung didn’t make the hard copy information available. Either it does not exist, or he thought it was too much trouble to access and copy. My guess is that it does not exist — if it did then why was it not in the DOE report? In support of this, it seems that there are other papers from 1991 and 1997 that show that the data do not exist. What are these papers? Do they really show this?”
Wang was cleared of research misconduct by his own university, but by an enquiry that met in secret, published no evidence, and whose report the complainant was not permitted to see. Tom Wigley commented: “It also seems to me that the University at Albany has screwed up. To accept a complaint from Keenan and not refer directly to the complaint and the complainant in its report really is asking for trouble.”
And here’s a comment from a CRU programmer, trying desperately to fix a badly written, badly corrupted climate database that nevertheless passed peer-review and is included in the IPCC reports. This particular document is solid gold from end to end, but here’s just one comment from it as an example.
What do you think? If this was tobacco science, would you be satisfied with the explanations we’ve received so far? Even if you don’t agree, do you at least see why some other people might see it as rather more significant than people writing some “angry” emails?
“I am seriously worried that our flagship gridded data product is produced by
Delaunay triangulation – apparently linear as well. As far as I can see, this renders the station counts totally meaningless. It also means that we cannot say exactly how the gridded data is arrived at from a statistical perspective – since we’re using an off-the-shelf product that isn’t documented sufficiently to say that. Why this wasn’t coded up in Fortran I don’t know – time pressures perhaps? Was too much effort expended on homogenisation, that there wasn’t enough time to write a gridding procedure? Of course, it’s too late for me to fix it too. Meh.”
September 5th, 2010 at 10:37 pm
I’m not going to argue with you about those comments (that’s all been done elsewhere) except to say that 1) if you looked at 13 years worth of my emails, I’m sure you could find some unflattering things said about my projects, products and/or employers, but we fulfilled expectations 99.99% of the time (in fact, it says something that this is all the email thieves could find), 2) some of these appear ripped out of context, 3) they are only talking about a small number of data sets, AGW doesn’t depend on depend on the small number datasets you could nitpick here, 4) they were under attack from politically motivated people who have shown bad faith–I forgive them for not wanting to release data that could be nitpicked to death with small or non-existent problems blown out of proportion, and 5) the researchers’ institutions investigated them and cleared them of wrongdoing.
September 6th, 2010 at 12:48 pm
Nullius, I have to agree with Jon about pointing to email conversations as proof of conspiracy or fraud. I could certainly be indicted by some of the comments and mistakes I’ve made in via email.
I think you’re missing this: Good scientists are extremely careful about being accurate when they publish. However,
when scientists talk with each other, they don’t sound at all like Mr Spock – in conversation, most of the time they don’t carefully consider everything they say and qualify every statement to make sure its absolutely correct. Rather, they bounce ideas off of each other, think out loud, shoot from the hip, change positions, present strawmen for attack, often argue like the lawyers out of the courtroom on “Law and Order”. Its a common practice to state some position as a challenge for another to find the flaw in it. Through the process of arguing, they learn. Understanding ensues. Its a messy process.
By extension, e-mail communications ARE A PART OF THIS PROCESS. They are NOT publications, they’re scientific conversations frozen into electronic records. And because the conservations can be continued face-to-face, at conferences or over the phone, you cannot count on the emails being complete records.
Also, in your Harry’s README example, how do you know the author didn’t write this comment in a fit of exasperation, and then later changed his mind. Again, that README file is more a personal log or record than a README that would be included in a GNU software package intended for a general audience. As such, it falls into the more informal category of the emails. I’ve got dozens, possibly hundreds of files called README in project directories, mostly for my own use as a record. Many are old and probably have errors which I’ve never gone back and corrected.
and this comment
Why this wasn’t coded up in Fortran I don’t know –
makes laugh out loud – amongst the scientists I’ve worked with, “which programming language is better” arguments are more heated and personal than politics or religion. If that’s not a personal opinion, I don’t know what is.
I worked for a guy who was desperate to see a signal in a spectrum. Almost every scan would show some structure at the 1 to 2 sigma level. He would make us go back and integrate ten times as much, and the structure would integrate out, but there would be some new possible “peak” (2 sigma or less) at a different position and he would make us go and integrate on that 10 times longer still. And that one would go away. It seemed to never end (nothing was ever found).
Nullius, I think you’re looking for smoking guns and finding 1 or 2 sigma blips when you don’t understand the background levels.
September 6th, 2010 at 1:38 pm
I haven’t read the “Harry_read_me” file, but I have read the so-called “smoking gun” memos such as the one mentioning “Mike’s Nature trick.” I’ve also read a lot of commentary on them. This confirmed my initial impression that the CRU e-mails are not evidence of serious misconduct — except, possibly, an effort to evade British FOIA requests. So you can accuse me of confirmation bias, but I would dispute that.
I’ll also point out that the CRU staff are not the only ones doing climate science, so even if the investigations had shown them guilty of scientific misconduct, this would not invalidate work done elsewhere.
Now let’s look at your examples. In the first, someone wants to keep “dirty laundry” away from those he thinks might distort it. The quoted phrase might indicate guilt, except that the quotes were placed by the original writer. That suggests he didn’t really regard it as what we think of as dirty laundry, in the sense of the Don Henley song. And if I were working on climate science data, I’d want to keep it away from known distorters too.
Regarding the comment that it will be hard to dismiss the criticism of the dendroclimatology statistics, this to me sounds like science working properly. If, in the end, valid criticism was dismissed out of hand, that would be wrong. Do you have proof that this happened?
Doug Keenan’s work on the chinese weather stations looks valid to me, based only on what you present. Someone else agrees. Who? Why, Tom Wigley, director of CRO at the time. Are you suggesting that there was widespread fraud among the staff, but he did not know about it?
The programmer’s comment again seems to raise a valid concern. But what is its context? Was there no later chance to revisit this analysis with better methods? Logic suggests this is not the case, since the world is now aware of his concerns.
To summarize, this whole episode (which you call ClimateGate and I call SwiftHack) reminds me very much of the wrangles over the Hockey Stick in that both have errors but give essentially correct results, yet certain individuals keep on raising long-debunked allegations.
This stands in stark contrast to the earlier situation in the tobacco industry, where internal memos from their own scientists acknowledged the validity of the science linking smoking and lung cancer. Oreskes and Conway have a must-read chapter on this in Merchants of Doubt.
September 6th, 2010 at 3:31 pm
Jon,
Thank you. Much as I suspected. Remember, I wasn’t asking you what you thought of Climategate, per se. I was asking what you would have thought of it had these been tobacco industry memos. The answer, it seems, is that you would defend them as normal, understandable, unimportant, and not a reliable representation of what they really thought. I believe the tobacco lobby said much the same thing…
Sean,
“Nullius, I have to agree with Jon about pointing to email conversations as proof of conspiracy or fraud.”
Do you mean when Jon cited private tobacco industry memos as proof of conspiracy or fraud, or when he dismissed climate industry emails as not being proof of anything at all?
Yes, I agree that scientists can be careless in private informal talk. You may have noticed that I’m not always at journal-level standards in my blog comments here. But I cannot conceive of any circumstances where I would say “It won’t be easy to dismiss out of hand as the math appears to be correct theoretically, …” Nor would I ever propose deleting data to keep it out of somebody’s hands – not simply because I am more careful about scientific accuracy in casual conversation, but because there would be no acceptable motive to do so.
“Also, in your Harry’s README example, how do you know the author didn’t write this comment in a fit of exasperation, and then later changed his mind.”
How do you know the author didn’t mean every word, and wouldn’t stand by it to this day?
How, for that matter, do you know that those private tobacco industry memos weren’t written in a fit of exasperation, that the executives and scientists would now back away from? Isn’t it the same question?
The point is, we have prima facie evidence to say this particular (peer-reviewed and published) database is an unreliable and non-replicable mess (I didn’t even quote the worst bit), and absolutely no evidence that it’s not. None has been presented. No context or explanations or even excuses have been given. The question hasn’t even been asked, officially. Don’t you want to know? Wouldn’t you think it was important that not only was the data correct, but it could be shown to be correct? Isn’t that what this inquest process is supposed to be all about?
But still we have people saying there’s nothing there, nothing has been found, there’s no proof, they’ve all been totally exonerated, and the science – Harry’s database included – stands solid. Oh yeah.
If the tobacco industry, or any of those other polluters Chris was talking about, ever tried such a thing, what would people say?
Yes, I agree that Harry’s read-me was personal opinion, not meant for anybody else’s eyes. Clearly so, given some of the things he says about his former colleagues. I suspect, for that reason, it’s probably more honest than usual. I’m not taking it as gospel, but I do think it is worth taking seriously.
“Why this wasn’t coded up in Fortran I don’t know”
Yes, that is funny! Especially considering some of the things the software people have said about Fortran! It’s got GOTOs in it…
If I understand what you’re saying about blips and background levels, I think you’re saying that all climate science is done at this standard, and these practices are endemic. If so, I agree. In fact, I’ve long believed that far from being any sort of conspiracy or intent to deceive, the main problem the scientists have had with the sceptics is plain and simple embarrassment.
They write amateurish code, their data is a corrupted and undocumented mess, they make basic mistakes – but they manage to put a professional gloss on the whole operation when they come to publish and present to the outside world. Insiders know and accept how things are, which is why they’ll share data and code with colleagues. But the appearance of these sceptics proved a shock. They asked politely for the data, and then published for all the world to see all the errors and flaws! All the “dirty laundry”.
That’s why they closed up like a clam against the sceptics’ demands for data. That’s why Phil Jones would rather delete the file than send it to anybody – because it would reveal what a mess he’d made of it. They’re just a bunch of guys trying to make an honest living – and yeah, in real life things aren’t as slick and polished as they make it appear in public. Doesn’t everybody do that? But that’s worked for science for the past two hundred years, and airplanes don’t fall out of the sky, nuclear reactors don’t blow up, and bridges don’t fall down, so what are you worried about? It’s all just a politically motivated attack, one that puts people’s careers at risk!
But it’s not the blip I’m pointing to, it’s the noise level. It’s not just one error in the IPCC report, one dodgy database, one badly done reconstruction. It’s not just one thermometer sited right next to the barbecue. It’s the general issue of quality assurance. This isn’t just another unworldly theoretical academic study, pushing the boundaries of arcane knowledge back a few inches – this has an impact bigger than all the flight avionics and nuclear reactors and those other things safety engineers take very, very seriously. This is about saving the world.
So all those things you’d want industry to do before they release their latest new “totally non-toxic, honest!” chemical into the world, that’s what we want to see from climate science.
September 7th, 2010 at 12:33 am
Nullius,
I don’t think you can make a serious case equating tobacco companies with climate science. For starters, the total US 2011 allocation for climate science is $2.6 billion (which includes all NASA satellites/launches and NASA and NOAA’s extensive networks – the actual NSF research slice is about 500 million, DOE’s maybe half that). While that may seem large, it pales next to the approximately $100 billion in annual cigarette sales. Tobacco companies easily spend $10 billion a year just in advertising. The mission of tobacco companies is to make money by selling tobacco. The mission of climate research is to understand the Earth’s climate. If there really were money-hungry climate scientists out there, they’re in the wrong business. Any so inclined would have or should have moved to the financial sector or some other more lucrative sector by now. Sorry, but trying to claim that the two institutions are similar just doesn’t fly.
Regarding the memos, before I judged I’d have to see if they were informal conversations or more official and report-like. However, in the tobacco story, one or more whistleblowers (at least one was highly placed) came forward and testified about the shenanigans. Has this happened with the CRU? Has Harry come forward? That would help your case.
Your point that embarrassment is a more likely motivator than conspiracy is reasonable and may even be true in some cases. However, I think the mistake you’re making in general is that you expect scientists to know what they’re doing. They don’t, always. If they did, it wouldn’t be science, it would be engineering. The good scientists are always trying to be on the cutting edge of their fields. That often means constantly being on the learning curve with new techniques and ideas in order to push on past the edge of the frontier. Yes, they’re not always professional programmers nor statisticians, but I do take exception to the attitudes towards data. Most of the scientists I’ve worked with had high ethical values when it came to correctness and validity of data.
And at the end of the day, yes there could be problems with the CRU data, as with any scientific endeavor. But there’s nothing stopping other efforts to reproduce or even surpass it. Reproducibility is a critical part of the self-correcting aspect of modern science. If there are problems, thats the way to handle it, not subpoenaing email conversations. Or at least, thats the first step to expose it. The subpoenas can come later, if its found there’s a problem. In fact, there’s even institutional procedures before it gets to supoenas, as demonstrated in the recent Hauser case at Harvard. (That in itself resulted from several whistleblowers and failure to reproduce the data BTW)
This has an impact bigger than all the flight avionics and nuclear reactors and those other things safety engineers take very, very seriously. This is about saving the world.
If you are suggesting that climate research be funded at levels of nuclear and aerospace industries, I’m all for it! (better yet, fund it at a sizeable fraction of the tobacco industry budget)
September 7th, 2010 at 10:39 am
Comparing research scientists to tobacco companies is a false equivalence if there ever was one. The incentive for research scientists is to be standards-adhering professionals and get tenure, etc. The incentive for tobacco companies is to protect company profits. That’s what their shareholders demand and that’s what companies are pros at delivering.
The efforts of tobacco companies were plainly PR, even at the expense of distortion of the truth for the public. The efforts of climate scientists might have been in some sense PR (“let’s not give the raw data to the people who’ve shown bad faith”), but also quality control (“hey, that data collection effort doesn’t look like it’s working right.”) Again, to claim equivalence between the two borders on the ridiculous.
September 7th, 2010 at 3:26 pm
#15,
There are other motivations besides money. I’m not going to speculate on what motivates them (and arguing from motivation is also fallacious reasoning) but I’ll just note that some people go to lifestyle extremes even for such trivial matters as world-wide fame and reputation.
The question of whistleblowers is an interesting one. There has been a lot of speculation that the Climategate emails were not hacked, they were leaked. The theory goes that the emails were collected in response to the FOIA requests that were going through, and that the university was having difficulty deciding whether to respond to, and that when the university finally decided not to comply, the next day the emails appeared on the web. The theory goes that it was an insider with a conscience unimpressed by their obstructionism. ‘Harry’ has even been suggested as a possibility, although he seems to me to be an extremely unlikely suspect.
But the police say enquiries are proceeding, and they can’t comment yet.
“However, I think the mistake you’re making in general is that you expect scientists to know what they’re doing. They don’t, always.”
But seriously, the extent to which scientists don’t know what they’re doing is not quite that straightforward. I don’t expect scientists to know what the outcome of the research will be, or even how they’ll go about achieving it. But I do expect them to know the basics. I do expect them to have some common sense. Not knowing exactly what they’re doing in the proper scientific sense is not embarrassing – not knowing things they ought to know is. It doesn’t need a huge amount of money, just a bit of care.
You’re quite right that there’s nothing to stop others reproducing the work. Quite a few have – that’s how we found out about a lot of the problems, and that was long before Climategate. But we don’t see much sign of that self-correction yet.
Take one example – one of Steve McIntyre’s complaints had been that the tree ring people were using old data surveyed decades ago, and hadn’t brought the series up to date. (The suspicion being that they didn’t because they knew if they did their most dramatic results would probably disappear or diverge.) The response from the scientists was something along the lines that it was expensive and difficult to do such field work transporting heavy equipment into remote areas. So Steve came up with what he sarcastically termed “The Starbucks Hypothesis”, which was that it was possible to have a latte at Starbucks in the morning and update the tree ring series the same afternoon. He picked the Greybill and Idso Bristlecone series at Almagre, and with a group of friends, took some group photos at Starbucks in the morning, and re-sampled many of the same bristlecones that formed such an essential component of the MBH98 Hockeystick later that day.
As part of his basic procedure he took several cores from each tree, and found the result that cores taken from the exact same tree could be radically different. A core on one side shows a fairly flat wiggle, one taken at an angle from it shows a huge peak ramping up at the end. The rings were asymmetric! His theory goes that these are what are known as “stripbark” trees – all the bark has been peeled off one side at some point in the past, and the growth of the bit that’s left spurts to compensate. You get huge 6-sigma jumps in ring width at the end of the record in the 20th century. Core the trees at a different angle, and the growth spurts disappear.
So – fieldwork replicated, analysis replicated and corrected with the new up-to-date data, data published on the web and free for anyone to use. We knew before that the trees in question were not responding to temperature – Greybill and Idso reported that in their original paper when they first published the data, before Mann used it. We have another theory as to what they are responding to, and at the least, we know now that because of the asymmetry, any reconstruction relying on stripbark bristlecones needs to be withdrawn and re-done; or at least, not republished endlessly to ‘prove’ unprecedented global warming.
Have they? No of course not. The issue “doesn’t matter” because the original reconstructions got the ‘right answer’.
So what, precisely, is it that you expect the sceptics to do? Polite requests for data don’t work. Replication doesn’t work. Publishing in peer-reviewed journals doesn’t work – even when you can get it past the gatekeepers. Congressional enquiries don’t work. The Freedom of Information Act doesn’t really work. Making a fuss on the web has had only limited success. So what’s left?
Next time you see that famous picture of Michael Mann leaning on a sawn tree trunk, just take a close look at the rings…
September 7th, 2010 at 5:37 pm
“Arguing from motivation is also fallacious reasoning.” It isn’t in a court of law. In the legal profession you ask “cui bono” all the time. When you’re accusing someone of something, you normally need a motive.
And what you’re talking about with bristlecones, etc. is a moot point. You can make lots of hockeysticks with all sorts of proxies that don’t involve tree rings:
http://www.skepticalscience.com/Hockey-stick-without-tree-rings.html
You’re basically doing what I described before. You’re taking little outrages that amount to nothing in the overall scheme (all this stuff about problematic tree ring data was *published*, it wasn’t hidden -hardly an outrage) and blowing these “outrages” up as if they called every data set into question, which is patently ridiculous. It’s not science–it’s conspiracy mongering (“I found this ‘problem,’ therefore all data sets are suspect, even when done completely independently of the data set I’m criticizing.”)
September 7th, 2010 at 7:14 pm
Jon,
Ah, yes! The marvellous Mann 08! The one where he pasted in the Tiljander series upside-down, yes?
(In case you don’t know, although you ought to, Mia Tiljander took some lake bed mud cores and reconstructed local temperature from the early parts – but the later parts were corrupted by run-off from modern era ditch digging. This was reported in the paper. Mann pushed it into his algorithm anyway, and because his algorithm simply looks for a 20th century up-spike, the bogus 20th century down-spike in the data got picked up, turned upside-down to make it match modern temperatures, and then given a strong weighting in the result. Because Mann is not totally without shame, a large part of the modern period from this proxy was truncated, but the series was still left in upside-down, and the medieval warm period that Tiljander had reported became a negative cold period, pulling the reconstructed temperature down.)
But all of that is besides the point. This is a tactic I call the cups-and-balls trick. Whichever cup you pick, the ball is always under another cup. Whichever reconstruction we show to be bogus, you just shift attention to all the other reconstructions.
We’re talking about the bristlecones, and all the reconstructions that rely on them. Please don’t change the subject. When the climate science community have actually conceded the point and removed all the papers relying on bristlecones from consideration (and there are quite a few – not just MBH98) then we can move on to deal with Yamal, the old version of Polar Urals, the Gaspe cedars, and upside-down Tiljander. But the fact that you won’t actually throw away the bristlecones is still a problem.
Saying an error “doesn’t matter” because you have many more calculations of the same result is missing the point. It’s the principle – if there’s a problem with a paper, then you remove it, and make your case on what remains. You don’t keep on citing the original paper as if nothing had happened. You don’t keep on using calculations that you know are wrong. It’s unscientific.
When you can show me an IPCC report that does not contain any results relying on the bristlecones, then we can move on to discuss the others. Because if such a paper passes review, even though it is well known that they’re not to be relied upon, then the review is thereby shown to be ineffective, and the output has no credibility. If even bristlecone papers can still pass muster, what on Earth makes you think the rest of the papers are any good?
September 8th, 2010 at 10:03 am
Nullius: We’re talking about the bristlecones, and all the reconstructions that rely on them. Please don’t change the subject.
Um, no. *You* changed the subject to bristlecone pines. Bristlecone pines are *not* what this thread is about. This thread, or at least the pos,t is about how certain interests manipulate the public discussion about science. (And by the way, proxies aren’t even the major evidence for AGW.)
But the fact that you won’t actually throw away the bristlecones is still a problem.
Look at the link I posted above about proxy studies. It “throws away” not only bristlecone pines, but tree ring proxies altogether. You still get the same results.
Saying an error “doesn’t matter” because you have many more calculations of the same result is missing the point. It’s the principle…
Point #1. No one ever said the problem with the dataset “didn’t matter.” (It’s not an “error”, it’s a recognized problem–a limitation of the specific dataset.) Indeed, they said it *did* matter, and in fact they published papers on the problem.
Point #2. Principles inhere in *people,* not the natural world. Let’s say for the sake of argument that Michael Mann is “unprincipled.” (His research institution cleared him of wrongdoing, but nevermind.) Michael Mann isn’t the only one doing the research, and neither is his institution! There are plenty of other researchers, institutions, and completely different proxies being researched, and they are producing the same result. Does the lack of principles inhere in the profession of science itself? If you make that argument, you fit very well into the cast of characters Chris Mooney describes in *The Republican War on Science* (and let me underline the word “characters,” because they really are loons).
If even bristlecone papers can still pass muster…
I’m sure the IPCC calls out the problems with the bristlecone pines in its assessment. But if I’ve got a puzzle that shows a picture of an elephant, but I’m missing the piece showing its tail, do I say it’s not an elephant?
…what on Earth makes you think the rest of the papers are any good?
Exactly. You argue that one small problem blows up an entire dataset… a researcher’s entire body of work… the researcher’s institution… *all* scientific institutions, etc. etc. etc. It’s an ever increasing ring of conspiracy–the old Paranoid Style of Politics applied to science.
Doubt is your product.
September 8th, 2010 at 2:04 pm
“This thread, or at least the post is about how certain interests manipulate the public discussion about science.”
Quite so. And in this case, we were using bristlecones as an example of such manipulation.
“Look at the link I posted above about proxy studies. It “throws away” not only bristlecone pines, but tree ring proxies altogether. You still get the same results.”
If that link listed the only proxy studies used to make the case, then you might have a point. But they don’t.
I’ll try an analogy. Say you have performed five different calculations.
1+1 = 23
5+6 = 23
17+2 = 23
-3 + 6 = 23
14*3 = 23
The answer is 23, you say. All the studies confirm it.
Somebody goes to great lengths to show from the fundamental axioms of addition that 1+1 is not actually 23. “You can’t add up,” they say. And you reply, “but look at the last one on the list – I’m not using addition, and I still get the same answer! So my addition is clearly not the problem!”
And then you go back to using all five results as mutually confirming instances to prove that the answer is 23.
What I want to do is first get you to concede that 1+1 is not 23, agree that it shouldn’t be there, and take it off the list. The fact that 14*3 = 23 does not use addition does not make 1+1 = 23 correct. When we’ve got 1+1 off the list, then we’ll talk about 14*3.
But whether 14*3 = 23 or not, you’ve still got the IPCC saying the equivalent of 1+1 = 23, even though they knew at the time that it wasn’t.
This throwing up of alternatives to distract attention from the errors being pointed out is a classic example of “how certain interests manipulate the public discussion about science.” Science says that once somebody has falsified the claim 1+1 = 23, you stop using it. Anything that continues to use a falsified claim is not science and is not to be relied upon. So long as the IPCC report contains any reconstructions using stripbark bristlecones, it is not science, and it is not to be relied upon. Period.
“No one ever said the problem with the dataset “didn’t matter.””
Actually, yes they did. But that’s another story.
“There are plenty of other researchers, institutions, and completely different proxies being researched, and they are producing the same result.”
Most of the research is closely related, and a lot of them use permutations on the same small set of proxies. And where it isn’t and they don’t, they do quite often get different results. Loehle 07, for instance. Or the divergence problem. Or Mann’s ‘censored’ directory even.
But this is just the 23 argument again. It doesn’t matter if 23 is the right answer or not, what matters is that we stop using results we know to be wrong.
“I’m sure the IPCC calls out the problems with the bristlecone pines in its assessment.”
So if it knows they’re wrong (and actually all it admits to is that the situation is unclear), why does it still use them? Why not just use the evidence that hasn’t been falsified?
“You argue that one small problem blows up an entire dataset”
No. I argue that if a review process passes a result that it knows to be wrong, and which it has been told is wrong, then it blows up the scientific credibility of the review. And that of all the people pushing it.
We can discuss the rest of the dataset once we’ve got the ground rules about eliminating falsified methods sorted.
September 8th, 2010 at 5:36 pm
Let’s see here. Hmmm:
*Click!* http://www.skepticalscience.com/Hockey-stick-without-tree-rings.html
No bristlecone pines. No false equations equaling 23 either.
See, I could get into a one way hash argument with you on the use of bristlecone pines in paleoclimate research, when it’s justified, when it’s not, like you want me to. But I don’t even need to do that.
September 8th, 2010 at 8:22 pm
Nullius,
Jon has presented an excellent example of the scientific reasoning mindset which I’ve been trying to get across in many comment threads here.
The more experienced (the better) the scientist, the more they are apt to be suspicious of the results obtained by one method, and the more they will want confirmation by as many different methods as possible. The link Jon gives lists three methods: glacier length, stalagmites and borehole temperatures, that confirm the basic hockey stick shape. None of these suffer from the stripbark systematic that you’re claiming for the bristlecones, and thats an example of what gives them their power in the argument.
The scientific mindset is to be cautious about systematics, known or unknown, in any one of the methods. But if several independent methods point to the same result, one can have more confidence, because of the shrinking likelihood that the systematics – different in each case – would all produce the same error. What systematic effects in stalagmite data could alter the curve to agree with the different systematic effects in bristlecone pines? And then what systematic effects could alter the glacial lengths to agree BOTH with the different systematics in the bristlecone pines AND the stalagmite systematics? And then what kind of systematics could occur in the borehole data which cause it to agree with the systematics in the bristlecone pine AND the stalagmite AND the glacial length systematics, which are all different. As independent methods accumulate and agree, the case becomes stronger: the simplest and most likely explanation is that they’re all basically measuring the same reality, and they confirm each other.
This kind of argument is very powerful and is highly valued. I think as scientists mature, they gravitate towards these kinds of arguments, and they worry less about programming finesse and fancy statistical methods, and worry more about the reliance on one method only because of unforeseen “gotchas”, and therefore look for other methods to confirm their results. I say things off the cuff like scientists don’t know how to program and they don’t know what they’re doing a lot of the time, but in this sense they are wise and I have a great deal of respect for them.
September 9th, 2010 at 9:19 am
Another thing you have to look at is the incentives. If this research were in any way as shoddy as Nullius says it is, all sorts of graduate students and postdocs would be making their careers out of shooting it down, and Richard Lindzen would be speaking up to save his reputation. But that’s not happening–either for good reason (it’s solid scientific work), or it’s all a Big Conspiracy and all these scientists are collectively ignoring tenure in favor of putting their collectivist commie pinko thumbs on the scales, etc.
September 10th, 2010 at 12:57 pm
#23,
My earlier reply to this one doesn’t seem to have got through. Not sure why. Perhaps it was too detailed.
#24,
All sorts of graduates and postdocs are. They’re called the AGW sceptics.
There’s a third possibility you didn’t mention – that a lot of scientists are just taking it on trust. They’ve been told this is what their scientific colleagues say, they haven’t bothered to investigate themselves, they all assume that somebody else must have checked it. Just as you assume that if there was something wrong, somebody would have said something, so does everybody else.
Consider another historical example – that of Chandrasekhar and Eddington. Chandrasekhar was a young graduate who developed equations for black holes. The scientific establishment, led by figures such as Eddington and Einstein, said they were impossible. Einstein published papers showing they were impossible. And the rest of the astrophysics community followed them – the consensus was that black holes couldn’t exist, and that consensus lasted for decades. Were Eddington’s claims solid science? No. It was no more than that black holes offended his intuition. Was it a communist conspiracy? No. Just human nature. Successful careers end up with scientists gaining “scientific authority”, taking control of funding committees, acting as reviewers and expert opinions, control promotion and tenure, and their own personal prejudices and confirmation biases become the orthodoxy. Did Chandrasekhar “make his career” by shooting it down? No. He tried for a few years to get people to listen, made no headway, and gave it up and did something else. Only much later, when others had won his battles for him, did he return to the topic.
Science is like that. It self-corrects eventually, but it quite often goes off the tracks for a bit.
September 11th, 2010 at 12:51 am
Nullius, your fabricated erroneous arithmetic examples in #21 have little or no bearing on the measurements being presented.
For starters, your examples are arithmetic errors and not measurements. When measuring, one is attempting to KNOW reality for the first time by taking the measurement. One doesn’t know the answer beforehand, as your example implies.
Secondly, you’ve FORCED your several examples to all agree on the same erroneous value (never mind them all flying in the face of known arithmetic). I’m asking, whats the likelihood of several independent investigators getting the SAME wrong result if they all use DIFFERENT methods and those measurements suffered DIFFERENT problems? Likely that they would all report different values X, Y, Z, … where X != Y, X!=Z, Y!= Z, …. If they were to get roughly the same value X for the measurement, the more likely explanation is that the common X is close to reality.
Science is like that. It self-corrects eventually, but it quite often goes off the tracks for a bit.
Chandrasekhar and Eddington were a couple of theorists. Don’t try to equate their personal dynamics with that of hundreds or thousands of experimental climate researchers. If any two characters get into a snit, plenty more to jump in and fill the void with results right away. The more that are involved, the faster the mistakes are exposed, the faster the progress.
September 11th, 2010 at 11:21 am
#26,
I’m not quite sure how to answer this, since my more comprehensive explanation has been blocked. (and as usual, without explanation, so it isn’t clear what I have to do to fix it.) I’ll try to abbreviate.
First, the two hypotheses under consideration are
1. A warm Medieval Warm Period 1000 years ago, followed by a cool Little Ice Age 500 years ago, followed by a warm modern period recovering from the LIA to the same or slightly below the MWP.
2. A cold, flat, slowly descending temperature for 1000 years, followed by a sudden and unprecedented warming in the last 50 years.
The three graphs you indicate cover either 400 or 500 years. Both hypotheses predict the same general shape over this period. They therefore provide NO INFORMATION WHATSOEVER to distinguish the hypotheses.
The other point is that these are not measurements. They are reconstructions of the temperature – local or global, it isn’t clear which – derived from measurements of other quantities by a complicated mathematical calculation.
There are also a whole variety of other issues with them and their interpretation, which I had previously covered in some detail.
The 23 analogy was simplistic, but it was trying to get across the idea of ‘independent’ calculations done by different methods, each with its own problems. Part of the problem is that most of them are not actually independent, in one sense or another. Nor do reconstructions always get the same answer – I gave several examples in my detailed post of ones that didn’t.
And I also explained why exactly such a pattern is quite a likely outcome even by random chance. The problem is the times series are autocorrelated, which results in some counter-intuitive properties.
Eddington represented an astrophysics community of hundreds of researchers – he was just the most vocal in expressing what was quite a common opinion. Nobody jumped into the void, because none of them were willing to go against the consensus. (And yes, a number of them privately agreed that Chandrasekhar was right.) Chandrasekhar’s work was neglected and nearly forgotten. It’s just one example from many of a firm scientific consensus that turned out to be wrong. Max Planck said of this: “A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.” Things move faster today, but scientists are still only human.
September 11th, 2010 at 11:49 am
Chandrasekhar and Eddington were a couple of theorists. Don’t try to equate their personal dynamics with that of hundreds or thousands of experimental climate researchers.
Appealing to the work of theorists instead of hard data makes the use of the name “Nullius in Verba” a bit ironic, no?
And to repeat myself, if the work of the paleoclimatologists is so shoddy, and bristlecone pines are so important, why aren’t all the postdocs and graduate students publishing bristlecone pine research? Most likely it’s because there is no incentive to, for the reasons we discussed above. The professional reward is just not there.
September 12th, 2010 at 2:35 pm
Nullius, this is great fun but I think we’re abusing the hospitality of our blog hosts with these seemingly endless back-and-forths. Perhaps there’s another forum available for an ad infinitum exchanges on climate change.