The Science Police

Keith Kloor has an article in Issues in Science and Technology called The Science Police. I became aware of it because he linked to one of my posts so as to highlight a comment by Michael Tobis. I tried to leave a comment, but it hasn’t appeared, so thought I’d write this instead.

The general gist of these science police type arguments is that there is some group of people who police what it is acceptable to say and who do their utmost to prevent those, who don’t toe the line, from speaking publicly. It will typically involve examples of supposedly biased reviews (of papers or grant proposals), difficulty getting a place to speak at meetings, and criticism that supposedly crosses some line.

Some of these may well be examples of poor practice, but sometimes (often?) not. Sometimes reviews can be justifiably negative and sometimes you just don’t get selected to speak at meetings. It doesn’t necessarily imply some kind of insidious attempt to silence those who might present alternative views; it could simply be that the person’s work isn’t very interesting/good, rather than because it challenges the orthodoxy (I would argue that the former is much more likely than the latter).

However, a big part of Keith Kloor’s article covered the conflict between Realclimate and Roger Pielke Jr. As an example, he highlights a comment by Roger Pielke Jr on this post in which Roger says:

My objection with RC is not that you guys act politically, but that you act politically but claim not to be. This mismatch is what I have argued is a factor that contributes to the politicization of science.

It seems Keith Kloor might have a point, but not in the way he intended. There do appear to be people who try to police science, but they’re people like Roger Pielke Jr who think that they get to define what is acceptable public behaviour by scientists. The above not only accuses the Realclimate contributors of being dishonest (not acknowledging their political activities) but that their activities contribute to the politicization of science; their activities are damaging science and, hence, they should behave differently.

Roger’s overall argument appears to be that as soon as a scientific topic becomes politically relevant, any public engagement related to that topic is immediately political. Most scientists would argue, however, that informing the public about a scientific topic is not inherently political, at least not in the sense of it being a form of advocacy; that would require having an explicit policy preference. In fact, most scientists would argue that it’s crucial that scientists engage publicly so that the public and policy makers can be suitably informed about a topic.

Of course, scientists who engage publicly should be clear about the role that they regard themselves as playing. There’s nothing wrong with a scientist advocating for something specific as long as they make clear that they’re presenting their own views about a topic, rather than presenting some kind of summary of our scientific understanding.

So, my general view of these science police type arguments is that they almost do what they claim others are doing; rather than engaging with one’s critics, it’s an attempt to deligitimise them by suggesting that they’re trying to close down/control the discussion. It’s unfortunate, because some of those being criticised do cover topics that are worth discussing, and do have some interesting views. However, rather than considering what their critics are saying, they seem to prefer spending their time complaining about how they’re being treated, while appearing to avoid considering that maybe the responsibility lies mostly with them. If your critics are having success, yet you regard what you’re saying as consistent with the evidence, then maybe you need to find a way of saying it that is still consistent with the evidence, but that is harder to criticise?

This entry was posted in Climate change, ClimateBall, Research, Roger Pielke Jr, Science, Scientists, The philosophy of science, The scientific method and tagged , , , , , . Bookmark the permalink.

61 Responses to The Science Police

  1. The part of the article that I found most remarkable was:

    Pielke is among the most cited and published academics on climate change and severe weather. Yet he says he has been told by a National Science Foundation (NSF) officer: “Don’t even bother submitting an NSF proposal, because we won’t be able to find a reviewer who will give you a positive score.”

    If I discovered that it would be extremely difficult to find a reviewer who would give my work a positive score, I don’t think I would assume that everyone else was biased in some way. I think I would probably be much more concerned that maybe my research was so poor that noone was likely to rate it positively. In my experience, there are many who would rate someone’s work positively if it was indeed good, even if they had issues with the person presenting the work.

  2. RICKA says:


    What if the work is good – but not published because of who wrote it?

    If that were the case, I think you would agree that this would be bad.

  3. Rick,

    What if the work is good – but not published because of who wrote it?

    If that were the case, I think you would agree that this would be bad.

    Indeed, but in my experience good work gets published (although maybe not always where you might like it to be published) and gets noticed (even if it takes longer than it maybe should).

  4. Joshua says:

    I’m actually impressed that Keith can market basically the same article, over and over, for so many years.

    Although at least in the past he was willing to make a passing reference to the question of whether RPJr engages in behaviors that contribute to his hostile reception. I wonder why in this case he decided to go along with the whole RPJr. is a poor innocent victim line of rhetoric.

    Someone call Haidt; he is so concerned about the impact of “victimhood culture,” I’d think he’d want to focus some attention of the Heterodox Academy on RPJr.’s drama-queening.

    That said,, I think it’s a shame that Roger isn’t more accountable for ;his own contributions to the vitriol. Not that his contributes justify the vitriol directed his way – I don’t think that it does; but I do think that if he were more accountable and proactive, it would open up the possibility for some improvement. I do think he offers consequential analyses to the policy debate and I do think it’s a shame that he doesn’t do what he can to increase receptivity to his work. or in fact, actively cultivates an approach that insures hostility. It isn’t as if the kind of reaction he gets isn’t entirely predicable by virtue his condescending and antagonistic approach.

  5. Chris says:

    What if the work is good – but not published because of who wrote it?

    That would be rather astonishing Rick. Any old rubbish can be published somewhere – the idea that something important/worthwhile can’t be published is just silly.

    Of course one might not get one’s precious study published where one would like. And there are many instances of seminal work failing publication at the first attempt (see lots of examples; e.g.: ). However this work was eventually published and got the recognition it deserved.

    I think that’s pretty much the bottom line. Anything can get published, and everything that gets published gets the recognition and has the influence it deserves. The idea that important work isn’t published because (for example) “of who wrote it” sounds a little like the whining that can happen when someone feels he (it’s pretty much always a “he”!) isn’t getting the recognition he feels he’s entitled to 🙂

  6. JCH says:

    RickA, despite all the attempts to prevent Karl from publishing, he was able to publish, so I think you can relax on that one. Then they attempted destroy him after the fact and failed. Things are looking up.

  7. Chris says:

    quite relevant possibly though in a peripheral sense – a rather perceptive discussion of scientific publishing :

  8. > What if the work is good – but not published because

    You’ve missed the point. RP is complaining about grant applications, not publications.

  9. You’ve missed the point. RP is complaining about grant applications, not publications.

    Indeed, but the Kloor article also refers to a reviewer of someone else’s paper being concerned about the policy implications.

  10. Arthur Smith says:

    I belief Keith Kloor spent a year or more in Colorado with Pielke’s, from which this habit of taking Roger Jr’s side in everything seems to have originated. Did he ask for any documents to verify the anecdote about grant funding? Who knows what truth there is in it? And, ATTP, you have it exactly right – these cries of victimization are clearly attempts to “police” others, telling them what they can or cannot say. Fundamentally we have basic disagreements in the world about right and wrong and these kind of discussions can go nowhere because of it, I tell you what you are doing is wrong, you tell me that my saying that is wrong, and round and round it goes. The liberal mind tries to break out of this through a more tolerant live-and-let-live attitude, which of course then leaves the last accusation of wrongdoing to the other side, in general. We simply hope that onlookers can see for themselves who is behaving appropriate and otherwise. But it’s not always obvious.

  11. Arthur,
    I came across one of your posts about Kloor, today.

  12. Francis says:

    Tone policing. The last cry of the truly desperate. “It’s not that RealClimate is wrong, but they’re so mean.”

    Frankly it’s pathetic. Kloor and Pielke Jr. had nice gigs going as some of the biggest tone policers around (The Honest Broker, really?) but now that the data is so irrefutably against them they’ve either moved on (Pielke) or reduced to whining that they’re seeing their own past conduct now being used against their “side” (Kloor).

    American science and American politics are chock full of highly ambitious people who know how to throw sharp elbows. Competition is inherent in both systems, especially when they are zero-sum games (like grant awards and elected offices). If you’ve got something to say, pull up your socks and keep at it.

  13. russellseitz says:

    “The liberal mind tries to break out of this through a more tolerant live-and-let-live attitude, which of course then leaves the last accusation of wrongdoing to the other side, in general. We simply hope that onlookers can see for themselves who is behaving appropriate and otherwise. But it’s not always obvious.”

    Especially at AAAS headquarters- A 97% consensus about a physically obvious scientific truth is not much of a concern. A nominally nonpartisan national organization whose apparat consists almost exclusively of members in good standing of one party, with one of its recently retired Representatives is more troubling.

  14. Susan Anderson says:

    After years of observing and putting up with RPJr and his machinations, the words “bad faith” come immediately to mind. Unskeptical “skeptics” are often caught red-handed turning meaning inside out; taking their assertions at face value gives them an advantage they are not willing to give to the people they accuse. The idea that scientists are extraordinarily honest, curious, and interested in finding out seems a foreign concept to them. They are always looking for an angle.

    Going a step further, I’d say RPJr makes quite a good living doing this, and is a favorite to provide dubious accusations on behalf of the Republican side of things. And these Republicans are not my mother’s Republicans, with whom it was possible to have an honest disagreement. That kind have been driven out or joined Democrats. These are out for the main chance, power and wealth for their masters.

  15. Susan Anderson says:

    RPJr had a thing going with Revkin for quite a while there too. It was frustrating to watch. He’s got a way with him.

  16. russellseitz says:

    I think Andy judged RPJr’s book by its cover:
    Cave hominibus cumque liber solo.

  17. Steven Mosher says:

    Science has a border patrol.
    Green lines you cannot cross.

    Don’t make me list them all.

  18. Steven Mosher says:

    You want tone police?
    Read the mails.
    There is always policing of meaning and understanding.
    There is always social enforcing of the boundaries.
    Regardless of the discipline.
    This is not to say it is all a social construct.

  19. Susan Anderson says:

    Russell: you might also says, birds of a feather flock together (I have no Latin, or I’d find something clever for you).

    SM, not sure if that was intended for me, but if it was I’ve had my tone policed in the service of lies for years and if anything it helps me focus. Scientists tend to think more facts and more logic is the cure; unfortunately, people destroyed societies and civilizations for their beliefs and prejudices throughout recorded time. Surprisingly many just want somebody to blame, not somebody to help them work on setting things to rights. It’s getting late.

  20. Willard says:

    Why not quote MT’s comment in full:

    A bit taken aback at the summary phrase “implicated in making Michael Tobis famous” as if that were intrinsically a bad thing…

    But [Junior] arguably had a hand in promoting the idea (in my opinion incorrect) that I am unreasonable and politically dogmatic, which may in turn have contributed to killing my formal scientific career. I certainly never got any federal grants after that event, and not for want of trying. Of course there’s no proof that this false controversy raised about me affected my career, but I’m pretty confident it didn’t help.

    That said, though Eli among others is convinced it’s [Junior]’s fault for calling it to Morano’s attention, I somehow have trouble attaching more blame for my troubles to Roger than to Morano.

    [Junior] is, in my opinion, a problematic figure, who is quick to criticize while being quick to take offense. He’s often right and often wrong, which can be a useful role in itself, but he ought to be able to take as well as he gives if he actually wants the net of his contribution to be constructive.

    Among all the comments in that Poor Junior thread, Keith chooses one highlighting the time Junior pinged Morano and smeared MT. In an article about science police.

    You can’t make this up.


    Would KeithK be able to name published academics on climate change and severe weather?

  21. Roger Jones says:

    I saw the Kloor article and was thinking of posting something, because these examples are all tilted one way. ATTP has pointed this out.

    Have I been science-policed by an anonymous reviewer who had the bearing and tone of one of the subjects of this post? Absolutely. The only motive for rejection that we could figure out was jealousy.
    It’s a pity there was no time (work pressures) to resubmit elsewhere – it would have been a highly cited paper on adaptation and risk (The framework was published by Cambridge in an adaptation guide a few years later).

    Have I been science-policed by a science blogger with an open mind? Absolutely. In this case our thesis pointed to higher, rather than lesser risk, but it was non-orthodox. Frankly, as a person submitting papers on controversial subjects, this is to be expected. The solution is not to call the waaahmbulance, but to build a bigger mousetrap.

    Science is a picnic compared to the economics police. If you ain’t singing their tune, you’ll get nowhere.

  22. dikranmarsupial says:

    Ricka wrote “What if the work is good – but not published because of who wrote it?”

    Given the number of obviously incorrect journal papers (Essenhigh, Humlum et al, Loehle/Scafetta model, Douglass et al on model/observation comparison, Gerlich & Tscheuschner etc.) written by climate skeptic scientists that have made it into print, it seems unlikely that there are good papers that are not published because of who wrote them. Likewise grants, I read “The Chilling Stars” about Svensmark’s cosmic ray theory, and despite his grumpling about it he seems to have been rather better funded than I have been (note also CLOUD project, which was rather well funded). On the other hand, there are good reasons why Salby has not published his theory (namely that it is obviously incorrect), so peoples views on how good a piece of work actually is may be, shall we say “somewhat idiosyncratic”.

  23. Willard,
    Indeed, I’d missed the irony of Keith Kloor quoting from that comment – one in which MT suggests that his career may have been harmed by something RPJr might have done.


    Frankly, as a person submitting papers on controversial subjects, this is to be expected.

    Indeed, it’s not meant to be easy.

    Yes, given the kind of papers that do get published, the idea that someone will be unable to publish a good paper is a little bizarre.

  24. Steven Mosher says:

    I think that’s pretty much the bottom line. Anything can get published, and everything that gets published gets the recognition and has the influence it deserves. The idea that important work isn’t published because (for example) “of who wrote it” sounds a little like the whining that can happen when someone feels he (it’s pretty much always a “he”!) isn’t getting the recognition he feels he’s entitled to 🙂


    Don’t tempt me to publish some reviewer comments that were clearly personal. Maybe after he dies.

  25. dikranmarsupial says:

    ” Anything can get published, and everything that gets published gets the recognition and has the influence it deserves. ”

    I don’t agree with this bit. Essenhigh’s paper has had far more influence than it deserved (which was none as it was incorrect and the argument known to be incorrect before it was published – there is an explicit warning in the first IPCC WG1 report against making exactly the mistake Essenhigh made). Likewise the paper on the body mass of dinosaurs on which I also submitted a peer-reviewed comment (the original paper got press coverage, my paper pointing out it was wrong didn’t). Quite often bad papers have far more influence and recognition than they deserve, at least in the public (mis-)understanding of science, more than in the scientific community itself.

    I advise my students to write papers in the same way a good player plays chess (i.e. assume “best play” from your “opponent” and make the move for which your opponent has the weakest optimal reply, rather than the strongest move available to you – i.e. make the strongest argument the reviewer cannot refute). I suspect I have given reviews of papers that their authors viewed as personal. They were wrong though. It is human nature to perceive criticism as a personal sleight, especially when given in written form without the non-verbal cues we get face-to-face. Apply Halnon’s razor (or more temperately worded variations thereof) and get on with it is my advice.

  26. chris says:

    Yes fair enough dikran – I meant everything published gets the recognition and influence it deserves scientifically speaking. Obviously the whole point of sneaking self-serving stuff into the literature is to have an influence politically-speaking but from a scientific point of view this stuff (i.e. self serving rubbish!) is recognised as such and doesn’t really influence scientists (they may choose to publish a rebuttal) or the progression of a scientific field

  27. dikranmarsupial says:

    chris, indeed. I have come to the conclusion that it is only really worth the effort (given the lack of reward and the amount of hassle involved) of writing comments for papers that are likely to seriously mislead the public as in science generally bad papers are quickly recognized as being bad.

  28. Marco says:

    “Don’t tempt me to publish some reviewer comments that were clearly personal.”

    Yup, that stuff happens. All the time. To just about everyone. Weirdly, it appears scientists are humans, and thereby subject to the same failings as other human beings!

    Talking about personal:
    “There is no such thing as quasicrystals, only quasi-scientists.”

  29. russellseitz says:

    Susan, I first heard Andy Revkin use the expression ‘honest broker’ a decade ago; beware of the man with just one book.

  30. Eli Rabett says:

    Russell, as Eli has points out often, people who use “honest broker” have not a clue about what a broker does.

  31. mt says:

    It’s timely for me to re-enter the conversation beyond my hovering on Twitter, but I don’t particularly relish it being in the role as having an expert opinion on Roger Pielke Jr.

    I think that Roger occasionally asks a good question. The Honest Broker proposes to ask what the relationship between pure science and policy ought to be. I plowed through the book, and his suggestions as to what to do about this are, to my reading, obviously not just flawed but poorly thought out. (In general, I think that summarizes Roger Jr. Good questions, muddled answers.)

    But I continue to insist that he is asking a good question in Honest Broker.

    See on this.

    What’s more interesting about RPJr is how he’s called upon by the footdraggers in the US Congress to testify on what is really a modest null result (about cost of disasters normalized by wealth not showing a clear trend). Null results are not normally highly prized in science. They are really a dime a dozen. That Roger feels that somehow inserting this non-result into IPCC is not enough, and that somehow his ideas are being suppressed by the Idea Police, would be risible if it did not also fit so neatly into the naysayer narrative.

    But this leads me into a mad recursive tangle… if one shouldn’t speak of the Idea Police because of its political implications, isn’t that an Idea Police constraint itself? Doesn’t disapproval of the claim somehow prove the claim? If Roger shouldn’t say that people say Roger shouldn’t say X, doesn’t that demonstrate a value of X that I am saying Roger shouldn’t say? Aaargh…

    Nevertheless, I think what Roger is doing nowadays in consorting with conspiracy theorists and climate policy opponents is ethically wrong. It’s sort of a trap. How can I say this without proving his point?

    The answer comes back to Roger’s original question. There are multiple roles in the interface between science and policy. He is pretending to being an aggrieved scientist with a suppressed result (so suppressed, in fact, that it made it into an IPCC report, but never mind that!) But his result is trivial and of marginal importance; further, he is a political scientist and not a climate scientist. Null results are normally not of great prominence; how this particular one became a cause celebre presumably has little to do with its scientific value and much to do with its polemic value. Roger constantly criticizes scientists acting as stealth advocates. This is a perfect example of him dishing it out and not being willing to take it. Here he is being an obvious advocate while pretending to offer something of scientific value. It’s exactly the conflation of roles that he is so critical of in others that he is indulging in to excess himself.

    I think there are important ethical issues Roger raises. But he’s hardly a paragon of ethical precision in his own approaches to them. Ultimately there are people putting him in the spotlight that don’t have the public interest at heart. He should ask himself why he is so popular with them.

  32. Joshua says:

    =={ He should ask himself why he is so popular with them. }==

    I’m not sure why he should be asking himself that question, as he actively cultivates his popularity with them. That, IMO, suggests he already has an answer to that question, and it is an answer he’s quite content with. IMO, he wouldn’t so consistently and confidently go out of his way to explicitly provoke and accuse, (let alone his favored practice of implicitly provoking and accusing under a cloak of plausible deniability) , if he didn’t have an answer with which he is content.

    The positive response he gets from those who seek to undercut the science he days he supports, is always entirely foreseeable. The positive reinforcement he proves to them is likewise, foreseeable.

    All of wbich, IMO, in the end, unfortunately undermines the value of his good questions, let alone devalue even further his less than particularly valuable answers.

    That said, I do wish that there were fewer he responded to him in a manner that only serves to vindicate his healthy sense of victimhood. I wonder what would happen if more people just thanked holding for his concerns . Maybe he would then see his way clearer to be more positive?

  33. MT,
    Interesting comment, thanks.

    Ultimately there are people putting him in the spotlight that don’t have the public interest at heart. He should ask himself why he is so popular with them.

    As you may know, Roger is coming to London to give a talk hosted by the Global Warming Policy Foundation. According to this post, it is being partly funded by his blog’s tip jar. So, it appears that he is part funding a trip to give a talk hosted by an organisation widely regarded as spreading misinformation. Maybe someone should tell him?

  34. Marco says:

    “Maybe someone should tell him?”

    Tell him what? He knows that fully well.
    and even more clear:
    Roger Pielke Jr., an environmental scientist at the University of Colorado and professor long critical of the politicization of the climate debate, says the group uses science to cloak its political agenda. Pielke emphasizes, however, that as a lobbying group GWPF “has every right to advance whatever arguments it wants. It often focuses on stealth advocacy — hiding its politics in science — a strategy common across the climate issue, found on all ‘sides,’ and is pretty common across many issues.”

  35. Roger Jones says:

    I have written and junked a number of posts about these null results for measuring extreme events through normalised damages, because it gets complex if you want to be the honest broker about what those tests mean. It is almost impossible to get a positive result from this procedure as you know. The events are too sporadic, disaster management has improved a lot in most places, reporting rates have changed etc. These are complex enough to make p values meaningless either way (null or signal). To get a signal that way will take from 2020 to 2050 or longer depending on the variable. That is not good risk management.

    A simple way is to measure the indicators for disasters directly. In the main, they show that many hazards are increasing (e.g., wildfires, large floods but not flood frequency). If people quote those, figures then show graphs of insurance costs to show damages, Roger goes ballistic. This happened to the Climate Commission in Australia – my findings, showed a 38% increase in forest fire danger index in Victoria in 1998-1999 over the previous period 1972-1997 (it was a shift). They had Ross Bradstock and colleague’s work who show that fire weather is the biggest driver of fire severity. RPJr and colleagues at Macquarie Uni had done a study on normalised fire damages and deaths 1925-2009 finding no signal (Compton et al., Wea. Clim. Soc, 2, 300-310, 2010). For deaths, only three events mattered in their analysis – one each in the 1920s and 1930s and Black Saturday in 2009, the rest were much lower level. They also normalised insurance losses, finding no trends.

    There was no climate signal in SE Australia until 1968. There were big fires in 1967-68, 1983 and 1994. The next big change in fire weather was in 98-99 as I said (this is data driven, not my spin). There were large fires with property losses in SE Australia 2001-2 (NSW), 2003 (Vic, Canberra), 2006/7 (Vic), 2009 (Vic, SA – Black Saturday), 2013 (Tasmania), 2013-14 (NSW) and 2015 (SA).

    Yet John McAnaney in 2013 on The Conversation (RPJr’s colleague – RPJr backed this up on his blog) said the lack of a trend in normalised losses meant that land-use planning was the worst culprit. That frequency and severity of wildfires were not the issue, but exposure. That is dishonest. Without discounting exposure (it is a serious issue), normalised data can say nothing about the event frequency and magnitude. To say “we’ve always had fires” and there is no evidence of change is misdirection at best. But have a look at those dates above – they are not 15 to 30 years apart. You will not find one firey in Australia who doesn’t think what they are fighting hasn’t changed. It’s their fantastic work that has prevented those normalised losses going up.

  36. Roger Jones says:

    sorry, wrong brackets [Mod: fixed]

  37. Patrick says:

    For many years, Keith Kloor played cipher to Roger Pielke Jr. He can’t seem to let some of that go. He’s now moved from climate change contrarian to GMO contrarian, as has the Breakthrough Institute and Jon Entine.

  38. Marco says:

    Patrick, Kloor is not a GMO contrarian.

  39. Marco says:

    I think Roger (Jones) hits the nail on the head regarding the main criticism of Roger (Pielke Jr) and how he frames his research. That normalised damages, which often do not include mitigation(!) costs, do not increase, does not mean the actual events have not increased.

  40. Marco,

    That normalised damages, which often do not include mitigation(!) costs, do not increase, does not mean the actual events have not increased.

    Indeed, this has always been my issue. I’ve seen arguments as to why Roger’s analysis isn’t quite correct (as illustrated by Roger Jones above) but even if he is, a lack of a trend in normalised damage, does not mean there is no trend in the actual events themselves. However, what Roger Pielke Jr says publicly is often interpreted as the latter. One might argue that this is not his fault as he’s only responsible for what he says, not for how it’s interpreted. However, if someone is aware that they are regularly misinterpreted, you might expect them to try and find a way of presenting your information that is less easy to misinterpret.

  41. mt says:

    Roger J., thanks for that perspective.

    This brings me to a broader point about how we think about things. Some people are a bit overattached to statistics. There are other ways to think about things.

    In the event of a weird accident… hmm… say a nitrous oxide tank explodes in a dentist’s office and detritus smashes a window and injures a passing pedestrian for instance… the event may be too rare to do any statistical analysis. This doesn’t mean the exploding tank didn’t cause the injury. Statistics for N = 1 events are never significant, but sometimes we know things just by observing what is happening.

    If one is looking for absence of evidence of causality (rather then evidence of absence) it is a trivial matter to reduce the sample size until significance goes away. This will always be possible. Generally people doing statistical reasoning look for the largest possible sample size for this reason.

    The peculiar statistics of landfalling hurricanes in the US is a fine example. RPJr. talks about this a lot. But these are rare events embedded in a larger picture of tropical storm trends. Looking at the smaller subset of course weakens the statistics.

    The more severe an event, by definition, the more rare its background distribution. The reasons to expect more severe events are not, predominantly, statistical, but physical. The further from background climate statistics the system wanders, all else equal, the more the distribution of extreme events will change. Not every place will have the good fortune to have this matched by a decline in variance; at least that’s not a good bet. Consequently, unprecedented events will emerge.

    For me, cogent examples were the Montreal ice storm of 1998 and the Texas drought of 2011. Because these events were so extreme, either they are too rare to have historical precedent, or they were actually unattainable in the baseline climate. Either way, thinking about them statistically is like thinking about whether it is dangerous to walk under dentist’s windows. It’s better to think about manufacturing standards for gas tanks than statistics of the sort of accident that we hope remains an outlier.

    Anyway, RPJr demonstrates a peculiarity of politicized climatology – that getting a null result in this field can be celebrated even though it is notoriously hard to even get a surprising null result published in other fields. It’s really a very easy strategy, looking for data which lack sufficient statistical power to prove anything. The idea that proving nothing proves something is the peculiar part of this all.

  42. MT,

    The idea that proving nothing proves something is the peculiar part of this all.

    Yes, exactly.

    FWIW, I discovered that Roger has provided a response to my post.

  43. Willard says:

    > I wonder what would happen if more people just thanked holding for his concerns .

    Junior’s Twitter blocklist would grow bigger.

  44. Roger Jones says:

    Michael, and the fact that this sort of analysis gets published while the same people are complaining about climate police?

  45. Roger Jones says:

    Mind you, the so-called ‘climate police’ are using similar statistical methods to RPJr, especially with respect to the so-called pause. They’re not correct either. Climate is not a system that produces meaningful p values when measured as a trend change.

  46. mt says:

    “Climate is not a system that produces meaningful p values when measured as a trend change.”

    I agree with this; I am constantly surprised by how many people try to get statistical significance of climate trends. I don’t think frequentist thinking applies, because I don’t see how it’s possible to construct a meaningful null hypothesis.

    I think people are taking the null hypothesis to be that the time series is pure white noise.

    Disproving that is easy since it isn’t true even in the undisturbed quasi-stationary Holocene; saying that the trend is therefore meaningful doesn’t follow from statistical reasoning. All we know is that it isn’t uncorrelated samples, and we knew that already.

  47. Willard says:

    Junior may not be the stat guru Freedom Fighters portray:

    [Junior] (here, here, here, and here) and James Annan (here, here, and here) have been arguing over what it would take to verify the IPCC’s predictions. It’s been fun to watch, but not especially intellectually engaging because I already (think I) know the answer and because I haven’t bothered to look up the word “aleatory”.

    In the weather forecasting world, probabilistic predictions (such as the IPCC’s “very likely” or the weatherman’s “20% chance of rain”) get evaluated using something called the Brier skill score. This is a summary measure that combines two dimensions of quality: “reliability” and “resolution”. A set of forecasts is perfectly reliable if the stated probability corresponds to the actual probability. For example, if the weatherman states a “20% chance of rain”) for fifty days and it rains on ten of those days, the forecast is reliable, or well-calibrated. A set of forecasts is highly resolved if the stated probabilities depart a lot from the overall probabilities. For example, if on average it rains one day out of five, fifty forecasts of “20% chance of rain” are not very useful compared to forty forecasts of 0% chance of rain and ten forecasts of 100% chance of rain, assuming the forecaster picks the right days — resolution without reliability is not useful, just as reliability without resolution is not useful.

    Most of what the IPCC predicts is the sign of a trend. A priori, anything could go up or down with equal probability, so the IPCC could make perfectly reliable but useless predictions by saying for all weather and climate variables have equal chances of going up or down. In a few decades, we’ll find out whether the IPCC’s departures from 50/50 chances are reliable, and meanwhile we can presume that, as understanding and ability to predict increases, future forecasts will tend to have more resolution.

    NG’s 50/50 claim should be the null. Some would rather say “was.”

    That Junior’s not a betting man is a Good Thing.

  48. The Very Reverend Jebediah Hypotenuse says:

    When making an argument from ignorance, just assume logical equipartion of a proposition and its contradiction…

    One fish, two fish, red team, blue team.

    Basically, Perry is saying he believes nature has a larger role than humans in recent warming. I, too, believe the oceans might well be a primary driver of climate change, but whether the human/nature ratio is 50/50, or less, or more than that is up for debate. We simply don’t know.

    So, while Sec. Perry goes against the supposed consensus of scientists, what he said was not outlandish, and it wasn’t a denial of a known fact. It was a valid opinion on an uncertain area of science.

    As ‘valid opinions’ go, “We just don’t know.” never gets old.

  49. Vinny Burgoo says:

    Is it time for another one of these?

  50. Vinny,
    I’m not sure it’s ever time for one of those.

  51. Willard says:

    It may be time for a simple example of Junior’s stat fu:

    [Junior] has a new post up asserting that 28% of the IPCC’s findings are incorrect. Although it’s obviously a rather implausible figure, I was expecting this claim to be backed up with some sort of evidence of errors, or at least sloppiness, or something, so I had a look at his paper that he cites to justify the claim.

    It turns out that 100%-28% = 72% is merely the average (lower bound) probability level associated with the statements they made. Such as “It is very likely that hot extremes, heat waves and heavy precipitation events will continue to become more frequent.” Here “very likely” means greater than 90%. So, given 10 such statements, the IPCC is saying that they would expect the “very likely” outcome to occur about 9 times, and not occur about once. And similarly for “likely” (66%). Averaging over all the probabilistic statements, it should be expected that in about 28% of cases, the (probabilistically) preferred outcome will not actually happen.

    And in [Junior]-world, this means that 28% of the statements are “incorrect”. Note, however, that he does not make this silly claim in the paper itself, but only in his blog post.

    To see why this interpretation is nonsensical, consider a single roll of a fair die. I state (accurately) that it is “likely” to lie in the range 1-5. If I roll a 6, then in [Junior]-world, my statement was incorrect. However, it was not incorrect, and Roger is simply wrong to claim so.

  52. Pity the comment thread on Roger’s blog has not survived. If the IPCC were certain all of the projections would come to pass, they wouldn’t have made probabilistic projections! Hardly “honest broker” presentation of the contents of the paper. Worth noting that prof. Curry doesn’t understand the IPCC statements on probabilistic projections either.

  53. The Very Reverend Jebediah Hypotenuse says:

    … Curry doesn’t understand the IPCC statements on probabilistic projections either.

    Many climate status-quotidians confuse uncertainty with ignorance.

    Want ‘settled science’ before we enact climate change mitigation policy?

    Want 5 sigmas before you deign to accept the physics?

    It’s not going to happen.
    And it does not need to happen in order to see the shape of things to come.

    If only some folks could just stop peddling the pretense that it does.

  54. When Perry & the V. Rev. Jeb say they “, believe the oceans might well be a primary driver of climate change,” they confuse the slimate system’s controll and squelch nobs.

    Like it or not, the huge thermal mass of the oceans does slow the atmosphere’s radiative foring response to a comparative crawl- the eventual equilibrium response is not in doubt, but thre’s no use denying the relevance of the half millennium turnover time of the hydrosphere, because the atmosphere weighs quite a bit less than the first ten fathoms of ocean- oceanographers ( and geophysicists) have a lot of catching up to do in hydrodynamic modeling, before they can help turn GCM’s into real global systems models

  55. Mal Adapted says:


    Talking about personal:
    “There is no such thing as quasicrystals, only quasi-scientists.”

    Then there’s: “Punctuated equilibria, a theory of evolution by jerks.”

  56. Brian Dodge says:

    Shouldn’t one normalize losses to things like wildfire and floods by multiplying the loss by the amount spent on things like firefighting, insurance premiums, levees, prevention, and so on? My math skills aren’t great, but if you have the same percentage of loss of exposed property (which will change over time), but your expenses for e.g. firefighting have increased 30 fold, doesn’t that imply fires have gotten a lot worse? Forested area in the US hasn’t increased thirty fold; the Mississippi river hasn’t increased its length or shoreline.

  57. Harry Twinotter says:

    The article “The Science Police” is bizarre. I really liked this line: “largely in the treacherous climate arena, where every utterance can be weaponized for rhetorical and political combat.”


  58. Pingback: The return of KK! – wmconnolley: archive

  59. Pingback: Fact mongering | …and Then There's Physics

  60. Pingback: Scenario use in climate research | …and Then There's Physics

  61. Christi says:

    Pretty good post.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.