Naomi Oreskes and Post-Modern Science

Post-modernism is many things and its exact meaning is subject to argument, but I think most would agree that it explicitly rejects things like formalism and realism in favor of socially constructed narratives.  In that sense, what I mean by "post-modern science" is not necessarily a rejection of scientific evidence, but a prioritization where support for the favored narrative is more important than the details of scientific evidence.  We have seen this for quite a while in climate science, where alarmists, when they talk among themselves, discuss how it is more important for them to support the narrative (catastrophic global warming and, tied with this, an increasing strain of anti-capitalism ala Naomi Klein) than to be true to the facts all the time.  As a result, many climate scientists would argue (and have) that accurately expressing the uncertainties in their analysis or documenting counter-veiling evidence is wrong, because it dilutes the narrative.

I think this is the context in which Naomi Oreskes' recent NY Times article should be read.  It is telling she uses the issue of secondhand tobacco smoke as an example, because that is one of the best examples I can think of when we let the narrative and our preferred social policy (e.g. banning smoking) to trump the actual scientific evidence.  The work used to justify second hand smoke bans is some of the worst science I can think of, and this is what she is holding up as the example she wants to emulate in climate.  I have had arguments on second hand smoke where I point out the weakness and in some cases the absurdity of the evidence.  When cornered, defenders of bans will say, "well, its something we should do anyway."  That is post-modern science -- narrative over rigid adherence to facts.

I have written before on post-modern science here and here.

If you want post-modern science in a nutshell, think of the term "fake but accurate".  It is one of the most post-modern phrases I can imagine.  It means that certain data, or an analysis, or experiment was somehow wrong or corrupted or failed typical standards of scientific rigor, but was none-the-less "accurate".  How can that be?  Because accuracy is not defined as logical conformance to observations.  It has been redefined as "consistent with the narrative."  She actually argues that our standard of evidence should be reduced for things we already "know".  But know do we "know" it if we have not checked the evidence?  Because for Oreskes, and probably for an unfortunately large portion of modern academia, we "know" things because they are part of the narrative constructed by these self-same academic elites.


  1. JW:

    Also, known as Lysenkoism.

    It's not about the truth or divining the right answer, but control. It's always control.

  2. HenryBowman419:

    Naomi Oreskes is a serious problem for science. I am a scientist, and I have met other scientists who think she is wonderful, largely because she espouses views that are consistent with their political ideology. She is, of course, not a scientist, but one who writes about science without actually knowing very much about such. She is little more than an ill-informed journalist (redundancy alert!), though she publishes in journals—not scientific journals (Science long ago being such).

    I think she is truly horrible, but there are many very much like her. Each time I read anyone referencing anything she has done, I cringe.

  3. JBK:

    I'm a chemist. I find her argument completely confusing. To deal with her claim that 'the science is well understood'. all one has to do is point at the fact that the models diverge completely from the data. The models having the temperatures going up and up but the real temperatures haven't statistically increased in 18 to 20 years.

    And if you do attain the state of having a good understanding of a system, you don't back away from the requirements of getting a good fit with the data and the model. You do what the physicists do, you tighten up the limits. 95% confidence is 3 standard deviations, Physicists demand 5 standard deviations for someone to claim that they have shown a claim to be plausible. That's better than 99%.

    Let me mention what 95% means. If you are 95% confident that your model will predict future temperatures. And that is the claim of these Global Climate Models. Then measured temperatures should 19 times out of 20 (on average) conform to what the model predicted. But we know that is not true, right now 20 out of 20 measurements fail to fall within the models predictions.

  4. Michael Wilson:

    Naomi Oreskes and Erik Conway came out with a truly disturbing book recently. Devastating review of it here:
    (discovered by way of Roger Pielke Jr.,

  5. Richard Harrington:

    Read Jonathan Haidt's book, The Righteous Mind,, for an explanation of this behavior. One of his points is that people make decisions at an intuitive level first and then fumble around for a rational basis for backing up their intuitions.

  6. Matthew Slyfield:

    Climate models don't make predictions.

  7. JBK:

    Oh really? Where do those trend lines that the climatanistas wave around come from? I can google: 'climate model predictions" and find pages with 90 model trend lines rather easily.

  8. Matthew Slyfield:

    From the models, but if you try to claim that the model predictions have failed, those same climatanistas (to use your term) will turn around and claim the the models only make projections, not predictions and will act all shocked that you want to compare their projections to real observations.

  9. MJ:

    95% is 2 standard deviations, assuming you're talking about a normal distribution. 3 standard deviations would be 99%+.

  10. MJ:

    Yes, she's pretty awful, but it sounds like the real problem for science is the other scientists you've met. Some people just drink the kool-aid. Others actually make it.

  11. JBK:

    Right when you cannot argue data argue semantics.

  12. CapitalistRoader:

    Lysenkoism had a 30+ year run. If you pick 1988 as the start date of global warming hysteria then we have a few more years of nonsense to look forward to..

  13. JBK:

    Heh, that's right 2 sigma not three.

  14. chelmer:

    As the King says in the King and I, "That tho' a man may be in doubt of what he know. Very quickly he will fight. He'll fight to prove that what he does not know is so, is so."

  15. chelmer:

    Mere fact, my friend, mere fact. But I feel that it's 3.

  16. Matthew Slyfield:

    There is an old legal saying: If the law is against you, pound on the facts. If the facts are against you, pound on the law. If both the facts and the law are against you, pound on the table.

    The alarmists have long been reduced to pounding on the table.

  17. TeeJaw:

    That’s what lawyers do, find evidence to support the result they want and then use clever tools to exclude the other guy’s evidence. That’s OK for them, it’s supposed to be an adversarial process. Science is supposed to be more collaborative, and it all goes wrong when “scientists” act too much like lawyers presenting a case. The government is paying them to be this way by putting them on government grant welfare. So-called journalists like Oreskes should be exposing this instead of drinking the kook-aid.

  18. FelineCannonball:

    Are you actually looking at the error envelope from 1000+ model runs or the mean model result? The latter is not a prediction in the sense you seem to expect.

    There's a basic difference between predicting weather and year by year global temperature and predicting long term changes in mean temperature. If you want a high degree of statistical significance in the models you have to go out a few decades and look at a multi-year mean.

  19. JBK:

    We have gone out two decades with no statistical warming. That more than matches the less than two decade climb in temperatures that the climanistas are so excited about.

  20. FelineCannonball:


    Even if that was true, it doesn't mean much. If you look at individual model runs of any modern GCM, they are riddled with contemporary one to two decade global temperature "hiati." That is a scale at which internal climate variability dominates. At longer time periods they become less relevant. Predicting oscillations (or weather) is comparatively difficult in comparison to predicting the sign and order of magnitude of long term changes in global mean temperature driven by shifts in the radiation budget. Equilibrium shifts are even easier than transient shifts, but that's an even longer time frame.

    Here's an example model run:

    Short term trends are basically much ado about nothing unless you're complaining about crappy El Niño predictions. You are right. They are crappy, but there is no attempt to even incorporate them in GCMs or predict short term trends reliably in the curves shown in IPCC reports.

  21. FelineCannonball:

    BTW, recent temperature shifts are basically the least convincing evidence for a CO2 driven greenhouse effect. The prediction is still barely outside the noise from internal climate oscillations. Any trend lines you try to develop from the last 40 years of data have huge error bars and are hardly of any use ground-truthing models. Come back in 2050 for the real answer when models start diverging from the background.

  22. JBK:

    2050 is by no means far enough out. I have seen statements from statisticians who claim that you need 9 cycles of a periodic variation to have a good enough understanding of the data to detect variations from the expected.

    The PDO which is turning out to be a major driver of climate variation has up to a 30 year cycle. We need 270 years of good instrumental data for that one climate driver. And other cycles over a 100 years long are being discovered.

    The modern instrumental record started around 1890 or so for temperature data. Meaning that we have 125 years or so of good data. Thus we are just using the anal extraction method to make projections or predictions (whichever you prefer) on that short baseline.

  23. FelineCannonball:

    Depends on the answer you want. 2050 projections diverge significantly from past trends. It's where the data starts to get useful instead of forcing you to squint.

    If you want to define the new equilibrium, I suggest starting around the year 12000.

  24. robert:

    while 'second-hand' smoke is mostly bs imo, [and i hate smokers and smoking] side-stream smoke [which is what comes off a lit cigarette and is not moderated by contaminating the smoker's lungs... ;-] is a much more pernicious animal.

  25. muirgeo:

    Fake but accurate is exactly how we should describe the "true-believers" of free market capitalism. The evidence in support of the dangers of man made climate change is far superior to any evidence supporting the idea of free-market capitalism. It's more than mere coincidence that the free market adherents are super critical of the data supporting climate change.... somebodies ideology has to budge. If only the free marketeers were as critical of their own beliefs as they are of the science of climate change I think their positions would be in the same dust bin as geocentrisim or creationism.

  26. JBK:

    Do tell. How do you explain the fact that carbon dioxide concentrations have continued to increase but temperature has not increased for 20 years? Is that your far superior evidence.

  27. muirgeo:

    The temperatures have increased. Look at the 5 year trend from 20 years ago and compared to the most recent. THAT is how I explain it ... by not denying the facts. Also global oceans heat content is going up a lot as well masking some of the atmospheric warming.

  28. Milton_Hayek:

    I've also seen this referred to as "Post Normal Science" - when the crisis under consideration (e.g. AGW) is alleged to be so compelling and filled with potential calamity, that government policy needs must trump the need for scientific understanding.

    It's simply a new justification for statism under the cloak of scientific authority. This is similar to the 19th century scientistic arguments for Marxism.

  29. William O'Keefe:

    She is on a vendetta to attack and slander anyone who is a climate skeptic. Her book was a tasteless, baseless attack on three dead scientists of national renown who could not defend themselves. See the review on

  30. TruthisaPeskyThing:

    I can understand your conclusion if you do not seek to understand the data. You refer to GISS data which takes actual data through a series of adjustment processes. Without the adjustment processes, the actual data shows no increase for close to 20 years. An honest researcher would ask if the adjustment processes are necessary and accurate? Through hours -- in fact, days -- of research on the subject, I come to conclusion that the adjustment processes are not accurate. It is easy to do computer gymnastics with metadata, but one needs to look at whether the results match verifiable observations. For example, the ice on the Great Lakes is inconsistent with the results of the adjustment processes. The lack of record high temperatures is inconsistent with the adjustment processes. And -- while you can have increasing temperatures without drought -- the documented heat and drought of the 1930s and the 1950s are inconsistent with the adjustment processes. Also, the results of the adjustment processes are inconsistent with balloon and satellite data. Moreover, the adjustment processes are inconsistent with detailed records at individuals stations. Certainly, I can understand what prompts the desire for adjustments, but the evidence shows that adjustments do not produce reliable results. To ignore these issues is denying the facts.

    Regarding ocean heat content, I believe more investigation is justified there. The initial reports from the Argo system showed no increase. Then a change in algorithm was implemented, and now there is an increase. That change may or may not be justified, but the repetition of such adjustments raises red flags to me. Nevertheless, I am quite sure that ocean heat content was up in 2014 even if increases were not there in the previous few years.

  31. JBK:

    And here we have the truth of the matter:

    “You will never read or hear any of this from the scientific and
    political establishments,” Dr. Terry Hughes, professor emeritus of earth
    sciences and climate change at the University of Maine, told The College Fix. “I’m now retired, so I have no scientific career to protect by spreading lies.”

    He said he thinks dire global warming predictions are really all about lassoing federal research funding and votes.

    “Too many (the majority) of climate research scientists are quite willing to
    prostitute their science by giving these politicians what they want,”
    the glaciologist added."

    They are in it for the money and will never admit that they created 'global warming' for the cash it puts in their pockets.

  32. MJ:

    Utter nonsense.

    The difference is that believers in free markets can point to actual data, and, where necessary, empirical analysis to make their case, as opposed to citing dire predictions from long-term forecasting models that are sensitive to initial conditions and parameter assumptions.

  33. KPP:

    A quote by Karl Popper seems to fit here:

    "The game of science is, in principle, without end. He who decides
    one day that scientific statements do not call for any further test,
    and that they can be regarded as finally verified, retires from the

    - Ch. 2 "On the Problem of a Theory of Scientific Method"

  34. herdgadfly:

    The Indiana legislature is readying legislation to ban the sale of e-cigarettes (vapes, whatever) to keep our children safe from getting hooked on nicotine for life. Never mind that e-cigs immediately take away the danger of cancer and death from smoking cigarettes and provide perhaps the easiest path to cigarette smoking cessation.

    BTW, when I quit smoking, the chemical hold that nicotine had on me lasted less than four full days. After that, I had to beat down the infantile sucking syndrome that had reemerged as a result of smoking. Hard candy is a wonderful cigarette substitute but it puts inches on your belly.

  35. herdgadfly:

    So how many people have died from "side-stream" smoke?

  36. sean2829:

    "Fitting the narrative" is how I'd describe how editors edit in journalism today. Is Naomi the journalist historian simply applying the standards of her field to the subject she reports on?

  37. jdgalt:

    The free market has been tested many times, from Adam Smith's own era up to 20th century Hong Kong. It's proven to work. The trick is not to let the cronyist "chambers of commerce" misapply the name "free market" to their own demands, which are anything but.

  38. MB:

    Whole days of research, huh? Wow...I'm sure those climate scientists really respect your knowledge and dedication.

  39. TruthisaPeskyThing:

    Actually, they did respect my observations, and many showed appreciation for my questions. A couple of decades ago, I had modeling responsibilities, and a couple of them asked me to join them in publishing a paper. I told them that in good conscience I could not put my name on the paper unless we resolved a couple of the model's weaknesses or explicitly acknowledged them. They decided to proceed without me.
    My science background tells me that the type of questions I raise should be welcomed, not dismissed with scorn and a error-filled appeal to authority.

  40. TruthisaPeskyThing:

    I would encourage reading the works of Douglass North. His work analyzes the impact of institutions that we associate with free on markets on economic progress, reduction of poverty, etc. He was awarded the Nobel Prize in Economics -- not that such a distinction guarantees sounds analysis; but in some circles that award is a ticket that enables one's work to be considered.

  41. robert:

    Hash: SHA1

    i don't have a clue and i have no knowledge of any testing protocols the cigarette companies keep secret.

  42. Swami Cat:

    I am another fan of North's work on institutions and "open access orders." I think he does as well of a job as anyone in explaining the modern breakthrough of prosperity which so many take for granted.

  43. DeeBee9:

    What do they do, then, describe the past? Heck, that's not much of a feat.

  44. Matthew Slyfield:

    "What do they do, then, describe the past?"

    Yes they describe the past, and it's more of a feat than you think to describe climate as is was in the distant past when there were no direct measurements.

    However they also make "projections".

    Any time you get a discussion going between climate scientists from the alarmist camp and CAGW skeptics, about the failure of model predictions, the alarmists will insist that they models don't make predictions, they make projections.

    Projections look just like predictions, but they have the magical property that they can't be falsified by comparison to actual observations.

  45. JBK:

    It's even less of a feat than you may have imagined. The modelers 'massage' the models to make them fit historical data points.

    Normally with honest investigators you can test a model by having it try to generate data by setting the starting point in the past. For example you could set Jan 1st 1900 as the current date in a model and have it generate 115 years of data points. You could then see how closely these generated points align with historical data.

    This is called back casting. If you get a good fit then maybe your model is producing good data. If you get a poor fit then you need to do more work on the model. However the people running these models make this impossible by programming in past data in order to get a fit with historical data. No use in back casting since they already fudge factored the models.

  46. marque2:

    And when they can't fudge factor the models any more, they fudge the historical record.

  47. JBK:

    How true, they even flipped one set of data upside down to make it fit their lies better.

  48. marque2:

    It is not just Climate science. diet science is even worse. The key similarity, is they are both funded by government, and both come up with more and more alarmist facts, because it give politicians more reasons to save us, and come up with programs to expand government. The whole Michelle Obama school lunch program is based on discredited science.

    If you get a government grant you better find a problem, because that will lead to more grants, and politicians trumping your work for their benefit. Do a study and say, hey the XYZ seems to be doing OK, and watch your funding disappear.

    Somehow I think we should go back to the old model where most science was privately funded. That is the way it was before WWII

  49. marque2:

    I am sure if there were even a modicum of truth that "side stream smoke" is different, there would be 100 studies funded by the federal government. I wouldn't blame cigarette companies for "hiding the data"

  50. JBK:

    I think every one can agree that tobacco is deadly, the tobacco companies are loathsome, and only idiots would start smoking.

    But notice one extra despicable feature of this affair. Instead on outright banning tobacco, the states decided to make money off the idiots and entered into business with the companies by collecting those fines and ramping up the taxes to stratospheric levels. What is it $6 a pack in NYC?

    The government actually prefers that smokers die early so that they won't have to pay social security to them.

    If they had banned it I doubt a large smuggling operation would have sprung up. Tobacco is a rather noticeable crop and requires quite a bit of processing to turn it into cigarettes. Plus the addicts wouldn't be happy with going home and smoking a black market cancer stick after work. They require more frequent hits of their drug.