119 Comments

1. There is a lot of depressing stuff here, and I am not sure which part of it is worst. Possibly the cave in to a social media storm, showing as it does an apparent failure of integrity.

   LikeLike

2. The problem seems to be that any area that has traditionally attracted more male than female practitioners seems to be automatically labelled as misogynistic. The uproar following Alessandro Strumia’s public observations made it clear enough how woke academic establishments have now become. There seems to be more energy expended on gender and diversity issues than there is on the science, though I hope this is just the impression given.

   LikeLike

3. >”Rather the focus must be upon the prediction of the probability distribution of the system’s future possible states by the generation of ensembles of model solutions.”

   There’s your main groupthink problem right there. The range of possible future states generated by ensembles of model solutions does not form a probability distribution. The fact that some ill-educated individual in charge at the IPCC seems to think it does should not be a reason why the rest of them follow suit. This is a basic error and it troubles me that it lies at the heart of climate science.

   LikeLiked by 1 person

4. John when I was a young science sprogling I believed that finding an error of the magnitude you suggest lies at the heart of climate science predictions would be considered as equivalent to finding the holy grail. It would be the making of any scientist who found it and explained its significance. Why is this “discovery “ swept beneath a carpet? Is it just too horrific to contemplate? Are vested interests too much to overcome?

   LikeLike

5. Silly me, I thought smoke, even smoking native plants, produced
   CO2, methane and other nasty particulates!

   LikeLike

6. Alan,

   The following quote illustrates that there has been a lot of pressure placed upon climate scientists to do what they know to be wrong. This makes it even more of a scandal in my opinion:

   >”That pressures to provide subjective probability judgements were being exerted upon the experts is revealed by the following statement made by a participant modeller: ’What they were very keen for us to do at IPCC [1990], and modellers refused and we didn’t do it, was to say we’ve got this range 1.5 – 4.5°C, what are the probability limits of that? You can’t do it. It’s not the same as experimental error. The range is nothing to do with probability – it is not a normal distribution or a skewed distribution. Who knows what it is ?’”

   LikeLike

7. P.S. They may have been holding out back in 1990 but clearly they have since fallen in line. Everyone treats it as a probability distribution even though many know that it isn’t one. As Trenberth put it, what they do has “no epistemic validity”. And yet that’s what they do.

   LikeLiked by 1 person

8. Another quote that I’ve provided before on this subject is the following from Gavin Schmidt. I think it sums up the situation quite nicely:

   “Collections of the data from the different groups, called multi-model ensembles, have some interesting properties. Most notably the average of all the models is frequently closer to the observations than any individual model. But does this mean that the average of all the model projections into the future is in fact the best projection? And does the variability in the model projections truly measure the uncertainty? These are unanswerable questions…”

   “..Model agreements (or spreads) are therefore not equivalent to probability statements. Since we cannot hope to span the full range of possible models (including all possible parameterizations) or to assess the uncertainty of physics about which we so far have no knowledge, hope that any ensemble range can ever be used as a surrogate for a full probability density function of future climate is futile…”

   “..Yet demands from policy makers for scientific-looking probability distributions for regional climate changes are mounting, and while there are a number of ways to provide them, all, in my opinion, are equally unverifiable. Therefore, while it is seductive to attempt to corner our ignorance with the seeming certainty of 95-percent confidence intervals, the comfort it gives is likely to be an illusion.”

   I think I disagree when he says “These are unanswerable questions”. They are unanswered at present but reducing epistemic uncertainty will help enormously. Unfortunately, however, the IPCC insists on ploughing on with the idea that the ensembles generate a probability distribution.

   LikeLiked by 1 person

9. John R, I came here from Quadrant Online (where this essay by the wonderful Tony Thomas was also published), as Tony republished two of these comments of yours. They caught my attention greatly, in a way most climate change debunking usually does not (I’m an agnostic on this issue, but do think a Kuhnian paradigm is very possible as having captured science here). I’m wondering, would you be able to write up an essay of your own, to really explain more fully how you distinguish the fallacy you talk about here, from valid probability sets. So far I wouldn’t feel confident repeating your points in my own arguments. I need a “probability density functions for dummies”, specifically applied to this issue. It’s a smoking hot idea, but needs building out.

   LikeLike

10. Ianalexs,

    I would be happy to oblige but not in the near future. It’s not the sort of thing that can be thrown together easily and I am very short of time right now. Nevertheless, I am gratified by your interest. I do believe that it is an intriguing aspect of the climate change debate that does not receive the attention it deserves. I have some ideas as to why this should be so, and perhaps I will air them if and when I get around to it.

    In the meantime, read everything you can regarding the difference between aleatory and epistemic uncertainty and why one should not treat the latter as though it were the former. That is essentially what they are doing wrong.

    LikeLike

11. Thanks John. Yes I’ve now read a few articles on the difference between aleatory variability (e.g. rolling dice) and epistemic uncertainty (the
    model has knowledge gaps). Theories of hazard modelling of various types as well as actuarial and project management risk modelling distinguish between these. Earthquake modelling for instance is instructive because it doesn’t have the moral pressure of climate modelling. In one such article, I found this interesting quote:

    “In seismic hazard analyses, it is common to use logic trees to handle the epistemic uncertainty …The weights on the branches of logic
    trees are often called probabilities, but they are better characterized as weights that reflect the current scientific judgments in the relative
    merit in the alternative models. Calling these weights “probabilities” implies a mathematical basis that does not exist. Epistemic uncertainty
    is due to limited data (often very limited). In seismic hazard analyses, evaluating the alternative models involves considering alternative
    simplified physical models, data from analogous regions, and empirical observations. These are subjective. In some cases, uncertainties are
    developed from statistical evaluations, but that is not usually the case. … “Logic trees were (sic) used to allow multiple models to be
    considered with weights that reflected the degree of belief of the scientific community (or at least the seismic hazard analyst) in the
    alternative models.” (www.ce.memphis.edu/7137/PDFs/Abrahamson/C05.pdf). Note that the evaluative weightings among different model outcomes are based on “degree of belief”.

    I once had to model the agricultural potential of a particular crop within the province of a certain developing country. I inherited a GIS
    script that purported to be such a model. When I read the code, I was shocked at how exceptionally crude the assumptions were, usually just 2-
    or 3-option if-then lookups. We’re talking about variables like soils and rainfall. I did my best to work these over with actual equations,
    but these were still crude linear assumptions and without good empirical constants: the science just wasn’t there. Anyway, I ran the model
    repeatedly, tweaking it until it gave results that reflected what I knew to be the case in some wards of the province, and just thinking “ohhh
    boy”. I now know this was a kind of Bayesian probability evaluation. Was it of any value? Well, policy required knowing how much to push crop
    planting in the wards that we didn’t know about. It could have been done without the fancy GIS modelling, because it was just good old
    prognostication in new bottles. But at least, it did match the situation in the wards I could observe, and the assumptions were now explicit and
    consistently applied. The policy decision makers got a colourful, detailed *computer modelled*! map that looked really scientific. But yeah, ohhh boy. Don’t watch how the sausage was made. The real moral of this story is that most people haven’t built a computer model, so
    don’t realise it’s kind of smoke and mirrors. If it was taught in school, maybe a lot more people would be a bit more probing. But also,
    at the end of the day, I got paid for it and so did it anyway. Also also, any map-making is a representation of the world, always limited by
    some lack of data, and you can’t endlessly hedge and qualify. You just take the best punt you can. We just don’t draw spouting whales in the
    corners and write “there be dragons” anymore while sketching in our terra incognita.

    Anyhow, John, I still think this topic (“No, there is no probabilistic distribution in climate models”) would make a great article — for
    Quillette — i.e. for people who like to keep an independent mind, while still kicking wokeness up the derriere of course — that’s the sizzle and spice. When you have time.

    LikeLiked by 2 people

12. Ianalexs,

    That’s an excellent quote you have found there. Two articles that I’ve written here on Cliscep that may interest you are ‘Gleick, what’s not to like’ and ‘No one does wrong quite like Lewandowsky’. If you read the comments you can see how Dr Ken Rice of ATTP fame seemed to think I was making a pedantic technical point. I tried but failed to convince him otherwise.

    LikeLike

13. Ok, yes I’ve now read your two articles where you are careful to distinguish epistemic uncertainty from aleatory variability. I’m only more convinced that this distinction is essential to learn for anyone wishing to reflect on the IPCC modelling. If you don’t know it (as I didn’t until this week) you are, in this field, illiterate.

    I find it dismaying that the Gleick fellow you were speaking of, used devices of political rhetoric in his debate against Schellenberger instead of rational argument. From my experience in the seminar rooms, when academics resort to ad hominem, it means they’re out of ammo and they’ve lost the debate. Although I think you were suggesting Gleick was really trying a “look, squirrel” tactic to distract from the existence of epistemic uncertainty.

    As to the mountebank Lewandowsky and the false authority of the rotten Phil Trans Roy Soc special issue papers: this is why scholars of uncertainty theory in fields other than climate science are interesting, because they are more candid in their assessments, unburdened by the need to misrepresent theory in the service of ideological righteousness. They provide an alternative source of legitimate authority to build out the concepts.

    In your comments, your quote from J.P. van der Sluijs 1997 thesis: “Being the product of deterministic models, the 1.5°C to 4.5°C range is not a probability distribution. There have, nevertheless, been attempts to provide a ‘best guess’ from the range. This has been regarded as a further useful simplification for policy-makers. However, non-specialists – such as policy-makers, journalists and other scientists – may have interpreted the range of climate sensitivity values as a virtual simulacrum of a probability distribution, the ‘best guess’ becoming the ‘most likely’ value.”. Get that: “virtual simulacrum of a probability distribution”. Boom. Again, an unguarded gloss, from the days before ideological fealty to The Science became so intense.

    You’ve really got to get this out to the wider world; it’s a matter of basic literacy. Gotchas like the Barrier Reef not actually dying and mangrove dieback not actually climate change do indeed mock the hubris, but this is top-drawer. I read a lot of sites. Quillette is the one where this fits.

    LikeLiked by 3 people

14. Ianalexs,

    >”…this is why scholars of uncertainty theory in fields other than climate science are interesting, because they are more candid in their assessments, unburdened by the need to misrepresent theory in the service of ideological righteousness. They provide an alternative source of legitimate authority to build out the concepts.”

    Well said. This is not a climate science issue but a much broader one that has implications for the climate modelling fraternity. It never ceases to amaze me how relatively uncontroversial observations regarding the nature of uncertainty nevertheless attract very negative responses when people attempt to make them within climate science. The lack of regard for the distinction between epistemic and aleatory uncertainty may be partly due to ignorance (I came across it relatively late in my career, despite my scientific training) but I also suspect that it doesn’t suit net zero advocates to dwell upon it. The need to reduce epistemic uncertainty in order to gain a better perspective on risk is accepted best practice outside of climate science. However, placed in the context of the climate debate, it becomes a ‘scam’, advocated by bad actors, supposedly unaware of what ‘scientific’ uncertainty really is. We are supposedly ignorant, anti-scientific people who are pushing a discourse of delay that suits our own purposes. Well, the net zero advocates can say what they want, but what they cannot say with any credibility is that it is they who understand uncertainty best – not whilst they continue to treat model ensemble outputs as if they were data measurements displaying aleatory variability.

    Incidentally, here are two more of my old articles that you may find interesting: “What do you think to the show so far?” and “The Confidence of Living in the Matrix”.

    LikeLiked by 1 person

15. Ianalexs,

    You refer to ‘the Gleick fellow’, but you have to understand that he is a highly credentialed science communicator. This is his list of awards according to Wikipedia:

    1999 Elected Academician of the International Water Academy, Oslo, Norway
    2001 Named by the BBC as a “Visionary on the Environment” in its Essential Guide to the 21st Century
    2001 Appointed to Water Science and Technology Board of the National Academy of Sciences, Washington, D.C.
    2003 MacArthur Fellow “Genius Award”
    2005 Elected Fellow of the International Water Resources Association
    2006 Elected Fellow of the American Association for the Advancement of Science
    2006 Elected Member of the United States National Academy of Sciences
    2008 Selected to Present the Abel Wolman Distinguished Lecture at the United States National Academy of Sciences, April 23, 2008, Washington, D.C.
    2008 Named by Wired Magazine’s Smart List as one of “15 people the next President should listen to”
    2009 Keynote Lecturer at the Nobel Conference at Gustavus Adolphus College in St. Peter, Minnesota
    2010 Named “Visionary: A Catalyst for an Enlightened Future” in the Los Angeles Times Magazine, January 3, 2010
    2011 Winner, along with the Pacific Institute of the first U.S. Water Prize
    2011 Winner of the IWRA Ven Te Chow Memorial Award
    2012 Nominee for the Rockefeller Foundation Next Century Innovators Award.
    2012 Named one of 25 “Water Heroes” by Xylem.
    2013 Honored with a Lifetime Achievement Award by the Silicon Valley Water Conservation Awards, March 21, 2013
    2014 Named one of world’s “Top 10 Water Tweeters” by the Guardian.
    2015 Received the Leadership and Achievement Award from the Council of Scientific Society Presidents.
    2015 Received the Carla Bard Environmental Education Award from the Bay Institute.
    2018 Carl Sagan Prize for Science Popularization
    2019 Awarded the Boris Mints Institute Prize.

    Given the kudos that comes with such accolades, it is all the more important that his understanding on the subject should be clear and correct. And yet he was capable of writing the following:

    “Shellenberger misunderstands the concept of ‘uncertainty’ in science, making the classic mistake of thinking about uncertainty in the colloquial sense of ‘We don’t know’ rather than the way scientists use it to present ‘a range of possibilities’.”

    Not only is this a remarkably imprecise statement to have been made by a much-lauded science communicator, but the best interpretation I could make of it calls into question whether Gleick properly understands the basic dualism that lies at the heart of probability and uncertainty theory. He may still do so, but I have tried and failed to find anything in his writings that directly addresses the issue. The above quote is all I have to go on and, from someone with so many plaudits, that’s just not good enough. The only thing I can say for sure is that Gleick thinks that there is a scientific concept of uncertainty that transcends the colloquial, and that’s just nonsense.

    LikeLiked by 1 person

16. Ian, there is something missing from John’s list in 2012:
    https://wattsupwiththat.com/2012/02/20/breaking-gleick-confesses/

    Andrew Revkin at the time:

    Now, Gleick has admitted to an act that leaves his reputation in ruins*

    … the * an edit leading to a reassessment two days later:

    Gleick’s reputation and credibility are seriously damaged, not necessarily in ruins or destroyed.

    Look at all those lovely awards he obtained. Several of them seem to be subsequent to 2012. Nothing to see here. Move along. There’s actually no need to understand uncertainty so long as you’re certain about the important things. And have unimpeachable integrity. -_-

    LikeLiked by 2 people

17. I wrote in 2019:
    If you’re a climate scientist you can blot your copybook horribly but the prizes keep coming. You might not have heard of California’s Dr Peter H. Gleick, but read on. He’s been creaming it with prizes lately, $US100,000 from Israel’s Boris Mints Institute in April for the “Strategic Global Challenge of Fresh Water” and the Carl Sagan Prize last year for “researchers who have contributed mightily to the public understanding and appreciation of science.”  He’s scored more than 30 honors and awards all-up including a $US500,000 MacArthur “Genius” award for 2003.

    Nice work, Gleick, but you’re the same man who in 2012 raided e-documents from the minor sceptic thinktank Heartland Institute.  Its CEO Joe Bast said that Gleick “impersonated a board member of the Heartland Institute, stole his identity by creating a fake email address, and proceeded to use that fake email address to steal documents that were prepared for a board meeting. He read those documents, concluded that there was no smoking gun in them, and then forged a two-page memo.” Gleick denied forging the document. The forgery, among other fabrications, showed Heartland receiving  $US200,000 from the Koch brothers’ Foundation, when the reality was a mere $US25,000, and even that sum was actually for a health-care study.

    Gleick confessed he committed the thefts because he believed Heartland was preventing a “rational debate” on global warming, even though he had refused a Heartland invitation to a fee-paid after-dinner debate shortly before he stole the documents.  Gleick said

    “in a serious lapse of my own professional judgment and ethics, I solicited and received … materials directly from the Heartland Institute under someone else’s name…I forwarded, anonymously, the documents I had received to a set of journalists and experts working on climate issues…My judgment was blinded by my frustration with the ongoing efforts — often anonymous, well-funded, and coordinated — to attack climate science and scientists .., and by the lack of transparency of the organizations involved. Nevertheless I deeply regret my own actions in this case. I offer my personal apologies to all those affected.”

    LikeLiked by 1 person

18. Tony, Jit,

    All perfectly valid points, but if the scientific community can find its way to awarding the Nobel Prize to someone who took part in the Holocaust, then accommodating Gleick’s lapse of ‘professional judgment and ethics’ would be child’s play:

    How to Become an Überdogg

    Never mind, there is always the Guardian to provide us all with the right moral compass:

    https://www.theguardian.com/environment/2012/feb/27/peter-gleick-heartland-institute-lie

    You keep forgetting, we are the bad guys.

    LikeLiked by 3 people

19. Yes, well, I suppose as a born sceptic (or at least, an long-ago teenage cynic) I personally don’t really care about how many awards Gleick has. For one, that only demonstrates how much he is admired by an in-group, no matter how powerful. Secondly, from a Kuhnian perspective, authority heuristics (where people just believe what authorities say rather than evaluating the evidence for themselves) is a kind of flywheel effect that keeps epistemic paradigms in power. But both these things are aspects of psychology of power in societies: people for whom belonging is important are susceptible to it, outsiders not so much. You can’t change it, the majority will always have an authority bias, that’s the inward pull of the herd itself. Reason is insufficient to convince. I’m not saying anything new here, am I?

    Interesting Tony, about Gleick’s fraud. It’s a case, isn’t it, of the “moral lie”. Gosh, don’t we know it around the case of Bruce Pascoe? “He might be lying, but those are the lies we need for social justice to happen.”. Gleick’s authority is a kind of capital that was drawn down (he felt he had to apologise) but he’s so powerful, it didn’t destroy him. Shades of shooting someone on Fifth Avenue. — update — wow, JIT has just linked to a Guardian article as I write this — and it’s all right there in the article title: “Peter Gleick lied, but was it justified by the wider good?”

    LikeLiked by 1 person

20. I actually wanted to pick up on an aspect of Gleick’s reasoning in John R’s comment above: Gleick saying that there’s a difference between high-culture uncertainty theory, which is always aleatory variability such as you get when rolling dice, as opposed to low-culture uncertainty which is ignorant people confessing they aren’t very educated and don’t know very much: and that’s what epistemic uncertainty is.

    But again, outside climate change’s authority-worshipping herd, there’s theory on this that blows past the grand poobah’s proclamation. Here’s a pertinent quote from “Epistemic vs. Aleatory uncertainty” at http://apppm.man.dtu.dk

    “Hutchins as cited in Fox and Ülkümen (2011)[7] has identified that natural languages reflect the intuitive distinguishing of cognitive concepts from individuals. Epistemic and aleatory uncertainty incorporation in natural language was anticipated and empirically validated (Teigen and Fox, Üklümen and Malle as cited in Fox and Ülkümen (2011)[7]). For example, phrases “I am 70% sure that…” and “I think there is a 75% chance that…” express epistemic and aleatory uncertainty respectively.”

    LikeLiked by 1 person

21. Apologies: the Guardian quote I mentioned was in John R’s comment. The Denmark Tech Uni quote should have “(sic)” applied on the word “distinguish” (distinguishment?), and more importantly on the word “change” (sic), which seems intended to be “chance”. And, while I’m housekeeping here, I’m still annoyed that my second post here (above) is full of hard carriage returns that destroy the visual flow of my argument (it happened because I cut-pasted), but I have no edit privilege to correct it. Can a mod do it for me? Can I get edit privileges to edit my own?

    LikeLike

22. Ian,

    I have done the edits you requested. If you ask nicely there is always a moderator on hand to oblige you, but I’m afraid the power to edit your own comments on Cliscep comes only after years of meditation in a Tibetan monastery 🙂

    LikeLiked by 2 people

23. I feel I should apologise to Tony for having turned the discussion on his article into a technical one regarding the nature of uncertainty, rather than one addressing the problems encountered when scientific establishments go woke. At least it offered an opportunity for some to remind the rest what a self-confessed fraud Gleick is. However, I wish to indulge myself further here by reflecting upon his criticism of Shellenberger. I’ve been trying to give Gleick the benefit of the doubt regarding his remark, since it is open to interpretation. However, it is difficult to reconcile it with the following quote I found in another Royal Society paper:

    “Epistemic uncertainty generally, but not always, concerns past or present phenomena that we currently don’t know but could, at least in theory, know or establish. Such epistemic uncertainty is an integral part of every stage of the scientific process: from the assumptions we have, the observations we note, to the extrapolations and the generalizations that we make.”

    So much for “..the colloquial sense of ‘We don’t know’ rather than the way scientists use it…”

    https://royalsocietypublishing.org/doi/pdf/10.1098/rsos.181870

    Incidentally, if you look at figure 7 in the above paper you will see a classic example of the error that I’ve been pointing out here, in which the uncertainties relating to ensemble-based projections are treated in exactly the same way as they are for uncertainties relating to measurements of the past. Furthermore, the confidence matrix used by the IPCC is replicated uncritically. At least one of the authors (Spiegelhalter) is a pretty switched-on guy for whom I have a great deal of respect; so I have to wonder what is going on here.

    LikeLiked by 2 people

24. John, I looked at that paper. Their research question, paraphrased: “what is the appropriate level of bluster for how we present the public with uncertain conclusions from gappy models”. Figure 7: isn’t it wonderful that when you are epistemically uncertain — as in, don’t know if your model actually works (but you and your mates’ Bayesian gut feeling is that you’re *pretty sure* it does), — as opposed to “I can graph the probability density distribution of pinholes around the dart board down at the pub, although I couldn’t say exactly where the next dart will land”: isn’t it wonderful that you can still say that for 80 years hence, you’ve got it pegged within ±1°C (within a *quantified 90% probability*) what the effect of certain macro variables on global rise in temperature will be. Hmmm. Maybe that’s where the crisis of public trust comes from.

    I noticed that the article did not cover the range of options for dealing with epistemic uncertainty that the slightly scrappy source I cited above did (Denmark Technical University, http://apppm.man.dtu.dk):

    Epistemic uncertainty quantification methods:
    – Bayesian probability,
    – Evidence theory,
    – Interval analysis or probability boxes,
    – Probability theory,
    – Possibility theory,
    – Fuzzy set theory,
    – Generalized information theory.

    *not*
    – Probability density function,
    – Cumulative distribution function.

    Now, ok, their paper was not actually about all the different ways you can quantify the risk that your ignorance presents, it was about appropriate level of pretending certainty when communicating to the plebs. But still, academics do like to show they’ve covered all the bases. And here they didn’t, leaving open the question: do they know this stuff? Or maybe not?

    LikeLiked by 1 person

25. Ian,

    For me the alarm bells started sounding when I read this about aleatory uncertainty in the paper’s introduction:

    “This generally relates to future events, which we can’t know for certain. This form of uncertainty is an essential part of the assessment, communication and management of both quantifiable and unquantifiable future risks, and prominent examples include uncertain economic forecasts, climate change models and actuarial survival curves.”

    That sounds very much like they think that climate model uncertainty is basically aleatory!

    Just because they are modelling a system that is inherently variable, that doesn’t mean that the spread of projections in a climate model ensemble is a reflection of that aleatory nature. In the case of the climate models, the spread is a reflection of the epistemic uncertainties undermining the models (the aleatory uncertainty, if anything, integrates out over time). Consequently, it is a spread that cannot be treated using the techniques employed for the aleatory.

    Ian, it is not just in the Spiegelhalter paper that one fails to find your list of techniques for dealing with epistemic uncertainty; try finding them in any of the IPCC’s reports.

    LikeLiked by 2 people

26. Spiegelhalter, by the way, was one of the three mathematicians who appeared on the BBC documentary ‘Climate Change by Numbers’. Another was Norman Fenton, Professor of Risk Information Management at Queen Mary London University, who became a climate change sceptic as a result of his experience making the programme. Here is one of the things he had to say:

    “Although I obviously have a bias, my enduring impression from working on the programme is that the scientific discussion about the statistics of climate change would benefit from a more extensive Bayesian approach. Recently some researchers have started to do this, but it is an area where I feel causal Bayesian network models could shed further light and this is something that I would strongly recommend.”

    So, Ian, you can see that you and I are in very good company.

    LikeLiked by 2 people

27. I was looking around for more literature on aleatory and epistemic uncertainty in general circulation climate models. The google results were sparse (mostly it returned hits about ethics, values, policy and risk response, not theory of the mode of uncertainty itself). But I did find one paper that was interesting in that it did raise difficult issues for the modelling:
    E. Winsburg 2012, Values and Uncertainties in the Predictions of Global Climate Models. Kennedy Institute of Ethics Journal 22(12). (http://www.winsberg.net/uploads/7/4/4/3/74439291/22.2.winsberg.pdf)
    Maybe you have already discussed this paper here in Cliscep. Apologies if this is old ground, there’s a lot of stuff here and I’m only slowly reading it.

    Just to preface: I don’t know much about climate modelling. I’m agnostic about the results of it, and of the overall risk of climate change, and what causes it. I don’t deny, I don’t uncritically accept. But I do strongly believe that established protocols around assessment of probabilities and risk should be used. I would very much like to “leave it all to the experts”, but I’m not sure the experts are beyond reproach in this regard. I think there’s a lot of hype — and even a ruling class totalitarian demand — to make people bow down before a kind of idolatry, and that makes me wary and resistant. So those are my personal motivations.

    As to the modelling: from what I have read, all types of climate models are based on physical principles encoded as equations, and many parameters that modify those equations. Not much difference from engineering, which reduces design to relationships between relevant variables, with liberal use of constants to calibrate to known observations. In engineering design, there’s no stochastic spread of results, there’s only different outcomes depending how you adjust your parameters. A cynic might call it the “GIGO” approach, which it could be if there’s significant unknown variables and not enough empirical basis. There’s that, and then also, as far as I know, there are no rules-based behavioural-type climate models that set off unpredictable simulated behaviours and do then result, after a series of runs, in a statistically distributed outcome. So all I can see is that the “probabilities” that are cited for global temperature rise are a kind of calculation from differing results from an ensemble of different climate models (i.e., different computer programs developed by different labs). So that’s where Winsburg comes in. There’s no real pithy one-liner in the paper, but a taste of his concerns are apparent in this quote:

    “Ensemble methods assume that, in some relevant respect, the set of available models represent something like a sample of independent draws from the space of possible model structures.
    This is surely the greatest problem with ensemble statistical methods. The average and standard deviation of a set of trials is only meaningful if those trials represent a random sample of independent draws from the
    relevant space—in this case the space of possible model structures. Many commentators have noted that this assumption is not met by the set of climate models on the market. … Perhaps we are meant to assume, instead, that the existing models are randomly distributed around the ideal model, in some kind of normal distribution, on analogy to measurement theory. But modeling isn’t measurement, and so there is very little reason to think this assumption holds.”

    This is a different take on uncertainty, that is saying “not aleatory variability” without, as far as I can see, actually implying “epistemic uncertainty”. If I get his overall argument correctly, he is saying that all things considered, even a Bayesian probability approach can’t work here. Anyway, what it does do is provide a strong argument that, when we see assertions like “90% probability of a x°C rise in global temperature by year 2100”, and if this is a result of statistical analysis of results from an ensemble of climate models, then it is an invalid, even meaningless, statement.

    LikeLiked by 1 person

28. Ian,

    I am familiar with the Winsburg paper and I seem to remember discussing it in the dim and distant past, but not necessarily here at Cliscep. The quote you provide nicely sums up the problem.

    Another paper you may be interested to read is “Uncertainty in regional climate modelling: A review”, by A.M. Foley of the National University of Ireland, Maynooth, Republic of Ireland. Foley has a good go at trying to justify the use of probability distributions for measuring uncertainty in climate model outputs but he acknowledges so many caveats that he only succeeds in convincing me otherwise. For example, there is this statement:

    “An objective approach was used by Jones (2000a), which relied on properties of classic probability distributions. If the uncertainties associated with various sources are taken to be uniform and independent, then when multiplied together they will yield a peaked probability distribution for key climatic variables.”

    Yes, of course they would…if.

    At one point he argues that all one has to do is think of the probabilities as Bayesian, i.e.:

    “The probabilities used by climate change researchers are not classical frequentist probabilities. They would be better defined as Bayesian probabilities.”

    Precisely. And that is exactly why one shouldn’t be using techniques that were developed for frequentist probability.

    In the end, the author seems to fall back on the good old precautionary principle to justify the acceptance of all the modelling caveats – he says we just can’t afford to wait until the uncertainties are tamed. Besides which, he says that no matter how dodgy the model predictions are they are still useful within a Robust Decision Making (RDM) context:

    “Fortunately, uncertainty in RCMs can be minimized, quantified and communicated effectively, and, in spite of their uncertainties, regional climate models can provide valuable information for the robust decision-making process.”

    If you have been reading my back catalogue, you will know that I’m all for RDM. Unfortunately, there is nothing remotely robust about net zero.

    LikeLiked by 1 person

29. Ian,

    Just to change the subject slightly, the statistical shenanigans that come with probability density functions and climate model ensembles are only one point of concern; the other, of course, is the black art of tuning. This peer-reviewed paper casts some light on the nature of the problem:

    “The Art and Science of Climate Model Tuning”

    https://journals.ametsoc.org/view/journals/bams/98/3/bams-d-15-00135.1.xml

    The quotes I could take from it are many, but here are just a few. The first alludes to a lack of transparency:

    “Although the need for parameter tuning was recognized in pioneering modeling work (e.g., Manabe and Wetherald 1975) and discussed as an important aspect in epistemological studies of climate modeling (Edwards 2001), the importance of tuning is probably not advertised as it should be. It is often ignored when discussing the performances of climate models in multimodel analyses. In fact, the tuning strategy was not even part of the required documentation of the CMIP phase 5 (CMIP5) simulations.”

    The following explanation is offered:

    “Why such a lack of transparency? This may be because tuning is often seen as an unavoidable but dirty part of climate modeling, more engineering than science, an act of tinkering that does not merit recording in the scientific literature. There may also be some concern that explaining that models are tuned may strengthen the arguments of those claiming to question the validity of climate change projections. Tuning may be seen indeed as an unspeakable way to compensate for model errors.”

    Well they are right about that. Then there is a comment regarding the use of the term ‘calibration’:

    “Defined this way, tuning is usually called calibration in other application areas of complex numerical models (Kennedy and O’Hagan 2001). Some climate modelers are reluctant to use this term, however, since they know that by adjusting parameters they also compensate, intentionally or not, for some (often unknown) deficiencies in the model formulation itself.”

    Yep, that’s why sceptics are concerned. And for those that are concerned with the a posteriori nature of tuning, there is little comfort to be found here:

    “Restricting adjustment to the process level may also be a good way to avoid compensating model structural errors in the tuning procedure. However, because of the multiapplication nature of climate models, because of consistency issues across the model and its components, because of the limitations of process studies metrics (sampling issues, lack of energy constraints), and also simply because the climate system itself is not observed with sufficient fidelity to fully constrain models, an a posteriori adjustment will probably remain necessary for a while.”

    And finally, a candid reflection upon the risks associated with overtuning:

    “Either reducing the number of models or overtuning, especially if an explicit or implicit consensus emerges in the community on a particular combination of metrics, would artificially reduce the dispersion of climate simulations. It would not reduce the uncertainty but only hide it.”

    That is another good reason why the dispersion cannot be treated as reflecting an objective, aleatoric variability from which uncertainty can be quantified. Amongst other things, it is a function of tuning processes that are subjective and essentially epistemic in their purpose.

    LikeLiked by 2 people

30. John Ridgway says,

    ”There may also be some concern that explaining that models are tuned may strengthen the arguments of those claiming to question the validity of climate change projections. Tuning may be seen indeed as an unspeakable way to compensate for model errors.”

    Well yes, isn’t that a kind of theory inoculation?

    LikeLike

31. It’s interesting how the sceptic’s viewpoint on mathematical modelling is often portrayed. Take, for example, this from Dr Ken Rice. In answer to the following comment:

    “The question for those who don’t like mathematical modelling is what are they going to replace it with and whether that is amenable to proper scrutiny?”

    He replies:

    “Indeed, it’s easy to criticise, especially when something almost certainly has flaws. The real issue is what would you do instead?”

    Then there is this from his moderator, Willard:

    “Of course contrarians may take away that scientists themselves admit that the Modulz are Stoopid, but then we already know that contrarians always win.”

    https://andthentheresphysics.wordpress.com/2022/09/18/the-role-of-mathematical-modelling/

    I think it is a shame that these people do not come on Cliscep more often to challenge the specific concerns raised, rather than spend time ridiculing the sceptics that they have created in their imaginings for their own entertainment.

    Also, you will note in the quote repeated by Beth above, that the author does not speak of those questioning models but ‘those claiming to question’. Even when they hear what we say, they presume it to be disingenuous.

    LikeLiked by 2 people

32. Let’s face it, mathematical models are the only way of understanding how the climate works and might change and so provide predictions. What also should be acknowledged is that such predictions cannot be relied upon. There is much that is not known about the climate that doesn’t get into the models and the complexities are too great to be adequately modelled. These caveats commonly are not acknowledged and belief in the accuracy of model predictions commonly is over rated.

    LikeLiked by 1 person

33. Alan,

    Yes, of course. The real issue is not what we would replace them with or what we would do instead. That’s never been the issue. I would welcome a debate regarding the real issues but there are those who find greater satisfaction in portraying us as smplistic dullards, even when all we do is quote what the modelling community is admitting amongst themselves.

    LikeLike

34. “Let’s face it, mathematical models are the only way of understanding how the climate works and might change and so provide predictions.” Let’s not forget that in other departments of the new ideology, numerical models are a symptom of white supremacy, and we should actually turn to indigenous ways of knowing. Indigenes around the world have known for eons that doublings of atm CO2 would spell disaster. Well, not exactly, but Koyaanisqatsi, you know. It’s a feeling, a lived experience, a personal truth.

    “It’s interesting how the sceptic’s viewpoint on mathematical modelling is often portrayed.”. I just read Judith Curry’s “Climate Models for the Layman”, following on from the discussion we had above. A nice overview of the mechanics, and the appropriate cautions were discussed. But in searching for her work, Google top-billed her Wikipedia entry. Wikipedia: “Curry has become known as a contrarian scientist hosting a blog which is part of the climate change denial blogosphere.”.

    I went into the Wikipedia ‘talk’ page for the entry, just to see how many people had tried and failed to bring the article to a more neutral point of view. A few. More interesting though were the ‘talk’ comments of the people who favoured calling her a denialist. They were arguing that ‘doubt’ — or ‘skepticism’ as they had it — was a synonym of ‘denial’. There were hundreds of words in the ‘talk’ page to justify this conflation, coming at it any which way. My observations were: a) there is an assumption that there is an established, a priori truth that has to be defended against ‘denial’ (a form of immoral conduct); b) the people making these arguments seemed genuinely unable to contemplate any difference between doubt and denial: i.e. a Manichean (black and white, binary) mindset; and c) there were repeated mentions of the ‘tobacco playbook’, i.e., that any doubt is necessarily cynically motivated and probably driven by an unseen powerful cabal.

    On this last principle, even school debates would be fraught: the nay team, by introducing doubt, is clearly running the ‘tobacco playbook’. I wonder too, if we ought to similarly remove ‘doubt’ from legal proceedings: get the criminals off the streets with prosecution-only trials (you don’t agree with that proposal? ah, so you admit it then, you enthusiastically support child rape!). For the new ideology, reasonable doubt is denial, after all. And indeed, on that line of reason, all truth is what our lords and betters say it is. I do believe I was reading today that New Zealand’s Prime Minister was making exactly this argument to the UN today: there must be a *global* ban on anyone doubting (denying!) the narratives of the hegemony. Reasonable doubt is disinformation, is the tobacco playbook, is immoral; and the global state should enforce its removal.

    LikeLiked by 3 people

35. Ian,

    There is a lot of stuff written about how ‘denier’ is just an innocent and accurate description of what we are. Others mock the fact that we object to the term because it is designed to put us in the same moral camp as Holocaust deniers. Apparently, we are just playing the victim card here since no such slur has ever actually been made. Which means that the following list of quotes must have been made up:

    http://www.populartechnology.net/2014/02/skeptics-smeared-as-holocaust-deniers.html

    I particularly liked Prof Bill McGuire’s contribution:

    “We have Holocaust deniers; we have climate change deniers. And to be honest, I don’t think there’s a great deal of difference.”

    This from a guy who is convinced that global warming will cause earthquakes!

    Also, if you have not already done so, you may be interested in reading what the dictionaries make of the semantic conundrum:

    https://cliscep.com/2021/05/25/dictionary-corner/

    LikeLiked by 2 people

36. John, ok, looked at that. In a way, the name-calling and wilful misrepresentation of climate scepticism is not too bad, when you compare with how the new ideology is defended on its other battlefields. When you read dissenting online zines like Spiked, Quillette, The Critic, or UnHerd, you see just how nasty the new ideologists can get — and as you’d be well aware in the UK, in some cases have the police, hence state repressive power, quite in their own pocket. In those fields, such as transgender ideology and racialist ideology, the hatred, outright slander and thuggish mob behaviour on the part of the new ideologists — which regularly leads to people being thrown out of their career — is frightening. And yet, with great personal cost incurred by the key moral objectors, the battle lines are moving, especially in the case of transgender ideology.

    My punt is that the hold that the new establishment has on climate science will eventually give way as well. Yes, they have on their side both an invested old guard who have been there since the 1980s, and then their acolytes who have been trained only to follow the doctrine. But this isn’t new. Kuhn’s 1962 “Structure of Scientific Revolutions” is still the roadmap. The “scientific consensus” is just a social club, heavily invested in its own survival against competition. The current repressive situation is a reminder that the self-image of science as a smooth, rational, cumulative voyage toward the telos of ultimate truth is just a post-war liberal-technocratic trope. Actually, it’s as human as everything else: a contested, bitter and often ugly process. And currently, it looks like the good ship Climate Science has chugged itself into a corner. Their enemy now is better ideas.

    Aphorisms from a brighter age:

    “Whenever a theory appears to you as the only possible one, take this as a sign that you have neither understood the theory nor the problem which it was intended to solve.” Karl Popper

    “I’d rather have questions that can’t be answered than answers that can’t be questioned.” Richard Feynman

    LikeLiked by 3 people

37. Ian,

    I wish I could share your optimism that a paradigm shift may be in the wings. The idea of an existential threat from climate change is so culturally entrenched that our leader of the opposition felt safe to go on our TVs last night and declare in a party political broadcast that we can solve all of our energy supply problems with ‘British wind’ (apparently there is ‘no reason why not’). All I can say is thank God for Brexit. European wind is nowhere near as reliable.

    Anyway, I heard no guffaws of laughter ringing down our street so I think that puts us in deep do-do.

    LikeLike

38. Ian, John,

    you just have to watch BBC news on Hurricane Ian 🙂 today to see the nudge factor in action.

    any weather event worldwide is now given top billing as another example/proof for MMCC.

    This media hype has been ingrained in most people as they think this is “the news” & a fact.

    LikeLike

39. May I as a simple lurker here on this thread express my utter delight at the privilege of being able to read the interchange of ideas between ianalexs and John. I feel sure that both of you are enjoying yourselves. May it long continue.

    LikeLiked by 3 people

40. Alan,

    I can confirm that interacting with Ian has been a pleasure. It’s always gratifying to find someone who gets your point and is able to make intelligent and relevant contributions of their own.

    LikeLike

41. > I think it is a shame that these people do not come on Cliscep

    U sure?

    LikeLike

42. John –

    It seems curious to me that you’d want to hear from people whose agenda is to lower our living standards, destabilize our electricity grid, and destroy 50 years of our progress.

    What kinds of things could someone like that say that would be materially of interest to you?

    I suppose discussions with Hitler might be interesting. Right?

    LikeLike

43. John –

    > Apparently, we are just playing the victim card here since no such slur has ever actually been made.

    I feel you, man. If only those alarmists would stop calling us names, we could get on with more productive lives.

    LikeLike

44. John, a hypothetical question :

    Had you asked the late Saul Kripke :

    “Is ‘Climate Skeptic ‘ a rigid designator in the set of real intellectual positions?”

    how would he have answered ?

    LikeLike

45. “..Model agreements (or spreads) are therefore not equivalent to probability statements. ”

    sitting here in florida i think gives one an interesting perspective on model spread

    https://weather.com/science/weather-explainers/news/spaghetti-models-tropics-tropical-storm-hurricane

    the spread in the spagetti will be due to epistemic and aleatory uncertainty. although i would prefer the terms structural and aleatory.

    with aleatory we can narrow the results by running the model multiple times. if the uncertainty is aleatory

    it does even out. if your a good modeler you can point to the exact point where divergences occur.

    this variable or that drives the uncertainty.

    in any case the point isnt how we sort this out.

    the point is this. I was on the east side of the state. uncertainty in where the thing would hit the west side

    was moot.

    the best science, with gaping holes in understanding, predicted hits from tampa to fort meyers.

    and the best science said definately the west side and not the east.

    was east possible? yes logically possible. we dont know everything

    in spite of the track uncertainty the information was still reliable.

    trucks needed to be staged
    sand bags filled

    nobody knew with cerataity
    nobody knew the sources of uncertainty.

    yet nobody rejects the models

    in the absence of models what do you use?

    where do you preposotion the water, the ulility truck?
    who evacuates first? why?

    LikeLike

46. Had you asked the late Saul Kripke :

    “Is ‘Climate Skeptic ‘ a rigid designator in the set of real intellectual positions?”

    how would he have answered ?

    he would tell you to read his work harder.

    LikeLike

47. ATTP Pest control measures urgently required. Ken expected shortly.

    Strange words (cerataity, preposotion) emanating from Florida, understandable perhaps.

    Are climate models and those predicting hurricane tracks and strengths comparable. I doubt it if only because hurricane predictive models have many prior examples to work with. Previous climate changes were not directly observable by science.

    LikeLike

48. Steven,

    Thank you for commenting.

    It is important to me that you appreciate that I am not arguing for the abandonment of models or suggesting that ensembles are a bad idea. Clearly, for some purposes models are all we have and they will always have some utility. Such utility will depend upon the questions being asked and the importance of getting a reliable answer. In your example, for the question you asked, the utility was high. But you said it yourself, the aleatory uncertainty in your example can be isolated. My concern is whether the utility can then be misjudged if the remaining uncertainty is treated as aleatory when it is in fact epistemic. I can envisage a situation where such a distinction might matter and I invite a debate as to whether we are in that position with regard to climate models and the questions that are being answered by relying upon them.

    LikeLiked by 1 person

49. Alan, the visitors seem to have arrived due to a(n) (dis)honourable mention in a comment on a thread at aTTP’s place. So far as I am concerned, they are welcome, but I would prefer it if they had something useful and intelligible to say.

    Of all the comments, only Steven Mosher’s add anything and say something of interest or relevance. Your response to the decent point he makes is the one I would have made.

    LikeLike

50. Willard,

    >”U sure?”

    Yes I is. But, as you know, the invitation was ‘to challenge the specific concerns raised’, which in this case are concerns regarding the assessment of a model’s utility and how delineating aleatory and epistemic uncertainty is important in this regard. Steven has made a welcome contribution to this debate. Obviously, I would be less happy if someone were to just step forward and belch into the microphone.

    As for wondering why your comments sometimes disappear here on Cliscep, I wonder why you wonder why. Surely you must understand that it happens for exactly the same reasons why my comments sometimes disappear on ATTP.

    LikeLiked by 1 person

51. Willard demonstrates why we neither like him nor understand him. It’s not my thread, but if it was, I would leave his comments for everyone to see. People like him do climate scepticism many favours.

    LikeLike

52. Mark,

    JR mentions my name, and quotes a remark that refers to this post:

    https://andthentheresphysics.wordpress.com/2018/01/11/can-contrarians-lose/

    This exchange does nothing to refute the point made in that post, and that post has little to do with da stoopid modulz.

    I stumbled upon JR’s summoning only because I was looking for a mention of Nic’s latest paper in various contrarian outlets. I know that I am not welcome here. JR knows it too.

    So if he could cut the pretense, that’d be great.

    Your own likeability is charming every untainted soul in the universe. One day we’ll all become contrarians like you. You know why? Because contrarians can’t lose.

    You got this.

    LikeLike

53. Mark,

    It’s not my thread either, but my vote would also be to leave Willard’s comment up. Of all the people I’ve had to deal with on the internet, he must be the most easily annoyed. I hesitate to say anything further lest it result in a further outburst. I hear what you say about doing climate scepticism favours, but there will come a point when we have no choice but to moderate, and I fear that is what Willard may be aiming for.

    In the meantime, I’m having serious second thoughts about sending that Christmas card I bought for him.

    LikeLike

54. Doubling down doubles the impression you create, Willard. Good work.

    LikeLike

55. I would think it ought to quadruple it by now, Mark.

    Be seeing you.

    LikeLike

56. Willard,

    To belatedly address Joshua’s point, Hitler would be welcome on this thread if he had anything useful to say about the validity or otherwise of stating 95% confidence intervals for the spread of climate model outputs. It’s not that you are unwelcome here, it’s just that I found Steven Mosher’s considered response more welcome than being called an entitled twat. I’m just funny that way.

    LikeLiked by 1 person

57. Steven (and John) –

    I thought this

    https://cliscep.com/2022/09/14/posturing-and-piffle-at-the-academy-of-science/#comment-129368

    was an excellent comment and addresses why (imo) almost all the arguments regarding the number of modeling angels dancing on the head of a pin (seen in the climate-o-ohere) boil down to ideology-based aggression and defense.

    This is about mitigation of high damage, even low probability, risk. That’s a complex topic with much nuance and complexity and conditional probability. That problem construct is like a siren call for ideology-based antipathy and motivated reasoning. COVID is a fantastic parallel, that by no coincidence lines up is long an deological dimension in an almost exactly congruent manner.with knsfe change modeling. Nic’s COVID modeling is a work of art.

    LikeLike

58. John –

    > more welcome than being called an entitled twat. I’m just funny that way.

    Which comment are you referring to?

    And, indeed.

    So then could you explain the difference in your view between alarmists (an obviously welcome descriptor) and entitled twats? I can think of many dimensions of difference but other, I think important and probably in practice most operational, dimensions of similarity.

    LikeLike

59. Joshua,

    Apologies, your last but one post was in spam and I have just released it. I don’t know why it ended up there while your last comment got through OK.

    LikeLike

60. And John –

    I almost forgot.

    Aa I said – it seems curious to me that you’d want to hear from people whose agenda is to lower our living standards, destabilize our electricity grid, and destroy 50 years of our progress.

    Doesn’t that description sound very much like an entitled twat?

    I’ll be less indirect.

    If I were someone who advocated for certain policies that the author of the OP is clearly in disagreement with (be careful about assuming I am), seems to me there’s a clear assumption that description above would apply to me. As such, then if I were to comment here it would be under the condition of being called an entitled twat (as indeed, that description of “agenda” would fit comfortably under an umbrella held up by entitled twats).

    You said:

    > you must understand that it happens for exactly the same reasons why my comments sometimes disappear on ATTP.

    You implied the reason comments removed here for the condition of (effectively) calling people entitled twats. So then are you saying your comments at ATTP are calling people entitled twats?

    I’d suggest a model to you where indeed, the mechanisms here are not a different animal than those at ATTP.

    Hence my reference to alarmists and agendas. Perhaps you could rethink those concepts and really evaluate why they remain here – in fact firm the backbone of a top line post – and caling you an entitled twat would be deleted?

    LikeLike

61. Get the pest control quickest, the contamination has spread.
    But leave the evidence of their fouling for all to see.

    LikeLike

62. Mark –

    Not a problem. I sometimes remember to copy my comments before posting, and when I do inadvertent trips to the moderation bin aren’t terribly troublesome.

    LikeLike

63. Alan –

    Thank you for making my point.

    And now I get the pleasure of watching your comment deleted.

    LikeLike

64. Joshua,

    For what I would hope are obvious reasons, I will not be elaborating upon why I think being called an entitled twat is unhelpful within the context of a discussion regarding uncertainty in models.

    I will return to your take on Steven’s comment when I have more time. I might even wait to see if Steven responds first.

    LikeLike

65. John –

    > For what I would hope are obvious reasons, I will not be elaborating upon why I think being called an entitled twat is unhelpful within the context of a discussion regarding uncertainty in models.

    I tried to be more direct but apparently it didn’t work. I wasn’t asking you to comment on why you think someone calling you an entitled would be helpful. Not in the least. So the question remains, was I really that unclear are are you unwilling to address what I was actually asking you to comment on?

    And there’s crossover here. You are using a model to understand comments and the associated moderation. Seems to me that I’m your model you are CLEARLY gliding right past important uncertainties as if they didn’t even exist.

    As to the root of the reason why you’re doing that (say, if it’s for epistemic reasons), I actually find that less important than that you can see that it’s happening. But the reason why is certainly a second-order issue. So for that reason I challenge you to return to the topic (one that you willingly started to engage in earlier) of whether or not that problem exists. If you just simply insist it doesn’t, then obviously the root cause could never be interpreted.

    LikeLike

66. John –

    And just to be clear.

    I’m not ultimately interested in talking about why people’s comments are or aren’t moderated (least of all individual comments), but more to discuss the issue of how people use models, such as the ones people use to evaluate moderation, as a window into interrogated some broader questions about modeling.

    LikeLike

67. Joshua,

    Since you value directness, I will be more direct with you. I’m simply not interested in deciphering your comments. I do not think it is clear enough where you are going with them (possibly my fault). As I said, I welcome comments that respond to the issues raised regarding propagation of uncertainty in climate models but it is not at all clear that that is what you are doing. I will be getting back to you regarding you comment about angels and pins, however.

    LikeLike

68. John –

    Thanks for the follow up comment.

    > As I said, I welcome comments that respond to the issues raised regarding propagation of uncertainty in climate models but it is not at all clear that that is what you are doing. I

    Summarizing my point with respect to moderation: I think that climate models, like all models, should be viewed in context – at least some of the time. So I was trying to draw the parallels between propagation of uncertainty in moderation models and propagation of uncertainty in climate models. I think there’s much to be said about the parallels, in real time, right here on this very blog. As such, the first place to start would be examining the uncertainty in moderation models.

    Again, it’s not an end in itself.

    Talking about moderation models isn’t necessary for talking about climate models, obviously. But I’d suggest that an inability (or perhaps unwillingness) to comprehensively assess the uncertainty in something like moderation models – *where people are clearly identity-invested* – suggests a problematic interrogation of the (identity-based) uncertainty in climate models. Of course, a lack of desire to do it doesn’t necessarily translate into inability or unwillingness. But if argue of you (or more importantly we) don’t get to the roots where the identity-based context is obvious, how likely is it that you (we) can to it in places where the connection is more obscure?

    This relates to my methodological bias as a teacher: you demonstrate mastery of a concept on a more concrete domain before moving into mastering the concept in a more abstract domain. Not a necessary sequence but often helpful.

    LikeLike

69. Sorry – upon rereading, I realize that likely came across as condescending. Not intended. I don’t see myself as the teacher here. But I am pushing to engage at a more concrete level as part of a sequence.

    LikeLike

70. Joshua, I appreciate that you are trying to engage more usefully and less offensively than Willard does. However, I’m not sure that talking about moderation models in the context of Cliscep contributes much, if anything at all.

    I don’t know the basis of moderation at aTTP’s place, but over here, we all just moderate comments on our own articles, and there is no party line. For myself, I welcome comments that are relevant, I discourage rudeness, and I will delete (or edit) comments that are defamatory. I also appreciate it when people write in clear and readily-understandable English, rather than leaving me scratching my head trying to decipher their ramblings. That’s about it, I think.

    LikeLiked by 1 person

71. Thanks Mark –

    So you’ll moderate for rudeness. But I should stop babbling like an illiterate. Is that it?

    LikeLike

72. Joshua,

    I too appreciate your sincere attempts to engage, and I will get back to you (watch this space, as they say). But my main focus will remain the topic of uncertainty analysis in climate modelling and the appropriate techniques to use. I will seek to persuade you that there are concerns of practical importance that are still worth considering despite what Steven has said above.

    LikeLike

73. Joshua, actually my comments weren’t aimed at you, though it might be helpful to explain how moderation works here. Your comments are pretty clear and intelligible. 😉

    LikeLike

74. Mark –

    Apologies for taking that wrongly. I don’t think Willard will take offense if I say sometimes I find his comments hard to parse (although usually I find worth the effort when I do).

    John –

    I’ll circle back. I don’t think I need convincing that there are concerns of practical importance. I’d be surprised if Steven did either. For me, anyways, it’s a question of priorities.

    LikeLike

75. Joshua – you lost me with –
    “Talking about moderation models isn’t necessary for talking about climate models, obviously. But I’d suggest that an inability (or perhaps unwillingness) to comprehensively assess the uncertainty in something like moderation models – *where people are clearly identity-invested* – suggests a problematic interrogation of the (identity-based) uncertainty in climate models.”

    English may not your native language, but this statement makes no sense.

    LikeLike

76. Joshua – after a reread, what I think you are trying to say is –
    this site/post/comments have “an inability (or perhaps unwillingness) to comprehensively assess the uncertainty in something like moderation models– *where people are clearly identity-invested*”
    waffle – that’s no answer to John the entitled question.

    ps – your comment above – “I suppose discussions with Hitler might be interesting. Right?”

    what a stupid comment, are you really a teacher?

    LikeLike

77. dfhunter –

    Thanks. I now know to eliminate you from the list of folks here I might potentially have a fruitful engagement with.

    LikeLike

78. Joshua – no probs – people who are clearly identity-invested are not my type, anyway, have a banana & be fruitful.

    LikeLike

79. I don’t know what this is all about. Recent comments lack context. I’m not an old hand here, I don’t have an historical understanding of the rancour.

    Intelligibility of comments: I have to say a lot of recent comments here are bordering on unintelligible. Here’s the one that gets me the most. There are two links recent commenters have made, that go back up to a supposedly “excellent” comment posted in the recent swathe of stuff. I could hardly make sense of it. Spelling errors left because whoever wrote it couldn’t give a rat’s (or perhaps has dyslexia problem, in which case, ok), incoherent grammar, poor development of argument, and with changes in topic (= rambling). Here’s a concept: “parsing effort”. Sure, write a dashed-off first draft and post it. It’s incoherent, but hey, let the reader spend minutes reading over and over to try and figure out (parsing) what you’re writing. Because hey, why respect people by spending time working it up before you post, so they have a chance of reading it straight off? No dude, if you write consistently incomprehensible tripe, I’m under no obligation to read it. I’m also busy. If you don’t add to my day, you’re off the list.

    As to the substantive aspect of what I think that “excellent” post was possibly trying to say (can’t be bothered actually to go back and read it again — quid pro quo — you don’t care, I don’t care). If you think there’s something off in the reasoning so far made about why supposedly computed probabilities of temperature rise are a basic math mistake (mistaking epistemic for aleatory uncertainty), then provide *your* reasoning (preferably with a source, as I did for my posts, so I can read your source too). I for one will read it, if it’s reasonable and readable. But if it sounds like drunken rubbish, I won’t.

    Is this not all just common sense and courtesy, and indicate a wish to communicate productively? Or are we back to the main problem with the new ideology — “debate is hate”, “anyone who I don’t agree with is a faaaasciiiist”. (Remember the sappy hippie Neil in BBC2’s “The Young Ones”? Yeah, that’s the voice.) In which case, how much more imbecilic can you get.

    LikeLiked by 1 person

80. ianalexs –

    > No dude, if you write consistently incomprehensible tripe, I’m under no obligation to read it.

    Are you under the impression that Steven thinks you’re obligated to read his comment? I doubt it.

    That’s how Steven writes comments. I think it’s true for a variety of reasons. Sometimes I find it annoying. Sometimes I think it’s strategic in a way I don’t like. Sometimes I think it’s strategic in a way that I find useful. Sometimes it’s probably pretty random.

    Here’s my suggestion to you. If you think it’s worth your time to try to figure something out, do so. Ask clarifying questions if there’s a part you don’t understand. If you don’t think it’s worth your time, don’t spend your time that way. It’s entirely up to you.

    > Because hey, why respect people by spending time working it up before you post,

    I suggest that you not take it personally. It’s probably not about respect towards you or anyone else.
    It’s not really about common sense nor courtesy. It’s probably not about a lack of desire to communicate productively. Think about possibilities that lie beyond your incredulity, as to why someone might write a comment in that manner. What reasons might exist for writing something that way, where rudeness, lack of respect, or a lack of desire to communicate productively aren’t the reason? Are you really willing to rule all other possibilities out?

    Anyway, complaining about it, seems to me, is REALLY the waste of time – except maybe if you gain some value from venting in ways that won’t materially affect anything.

    So that’s my two cents. But then again, I often have typos and misspellings (I’m not dyslexic), or fairly incoherent or not terribly well worked out thoughts. So there’s that.

    LikeLike

81. ianalexs,

    I appreciate that you are a recent (and welcome) visitor, so recent developments on this thread are likely to be somewhat bewildering. What has happened is that another website (“and Then There’s Physics”) posted an article and one of the comments on that article linked to the discussion taking place on this thread. That drove a lot of traffic here. What you make of that traffic depends on your point of view.

    There is a bit of “history” between Cliscep and aTTP. Trying to be objective, let’s say we find each other’s positions on many issues to be mutually incomprehensible. Regrettably, interactions, when they occur, tend to be unpleasant, even though I for one, and I know others here (and I hope some there) would prefer it to be otherwise.

    FWIW I share your bewilderment at the weird language used in some of the comments by the recent visitors. Notwithstanding Joshua’s attempts to justify them, I am firmly in your camp. I take the view that garbled comments are a sign of a garbled mind. Inability to express oneself clearly, demonstrates either confusion or an unwillingness to make the effort to ensure that one is understood, so far as I am concerned. A more sceptical view might be that ambiguity enables one to take cover there in the event that one’s argument (assuming it can be discerned) is demonstrated to be wrong.

    I hope this little spat doesn’t prevent you from visiting us and contributing here. As I commented above, I think some of our visitors’ comments go a long way towards making our case for us and undermining their own. Nastiness, arrogance, and lack of clarity of thought tend not to win debates in my experience. Again, just my view – I am sure that they would disagree, but then we almost always disagree.

    LikeLiked by 1 person

82. Joshua,

    In his comment Steven said, “In any case the point isn’t how we sort this out”.

    My position is that it is precisely the point. He went on to suggest that the real point is that the models, albeit subject to uncertainty, are still sufficiently skilful when viewed as an ensemble. But I ask how one can make that judgement reliably whilst using mathematical techniques for assessing ensemble skill that lack ‘epistemic validity’ (as Trenberth would say). I am not overstating a philosophical point here and I am not using it to justify “ideology-based aggression”. I am making the point because I believe it to be central and fundamental, and I see nothing in Steven’s comment that acknowledges or agrees with that. So, whilst I appreciated his engagement, I didn’t find his comment to be excellent.

    LikeLiked by 1 person

83. Ian,

    It is seemingly impossible to have a discussion on climate change with someone who does not respect your position without that morphing into a lack of personal respect. I am not pointing fingers, but there it is. It helps not to get too much into it but we are all human. That said, things get awfully difficult once someone has convinced themselves that everything you say is disingenuous.

    LikeLiked by 1 person

84. I find myself, almost reluctantly, in agreement with Joshua about Steve’s writing ‘style’. As long as I have read his posts, they have been the same -full of uncorrected spelling errors, sometimes tortuous logic yet sometimes containing pearls if not of wisdom then ideas to ponder and digest. It is commonly worthwhile to unravel his prose, even if eventually you reject his premise.

    I will also acknowledge that I made an error of judgement about you Joshua, although I am still suspicious about your intent. For example, what is all this about modelling moderation? Is this merely a distraction from John’s analysis of problems inherent in assemblies of climate change model results?

    LikeLike

85. Ok, I read Steven’s initial comment again. It’s actually about weather modelling (I think?), and not about the multidecadal global climate modelling we’d been focussing on. But anyway, it was kind of relevant to discussing distinctions between aleatory vs epistemic (or structural, as in terminology from the papers in Phil. Trans. R. Soc. A (2007)?). Steven didn’t have anything substantive to say on that, e.g. “Ian, John, you’ve got it wrong on probabilistic modelling, here’s the reason why”. And look, if that was the case, and someone comes along and really provides the pathway to a better understanding, then that’s good. I’ve been wrong before, I’ll be wrong again. But this didn’t happen in the current exchange. Nobody said anything sensible. BTW, I’m still reading old papers (like that Phil. Trans. R. Soc. A stuff), with an open mind that I might have missed something important. But it’s heavy reading and I’m getting a lot of interruptions. Actually, I’d love a gloss of it.

    It’s not much of a critique, but as far as Steven’s point goes — if I have it right (?) — it’s this: No matter the models are imperfect or even in some respects illogical (from the probabilistic angle, anyway, as we’d been discussing), we’re better off with them than not. I agree. Reasoning what the future may hold is a good idea. But let’s not lose sight that, I’d wager, humans have done this since culture began. Sometimes prophesy, sometimes ritual revelation through omens and oracles, we’ve been at it forever. So it’s not much of an argument to say “well, any prediction is better than nothing”. I think I mentioned above, I’ve built computational models too. With lots of flaws. Best I could do. I wrote policy recommendations. But I also asked for more research to refine the model (didn’t happen). What we want is a method that utilises the best of our considerable knowledge, but without fudging over the knowledge gaps. Thus, I think, this forum (or the bits I’ve read anyway): it’s a way to form an opinion on just how much faith to put in current orthodoxy. Annoying as it may be, but there’s always a tendency among some people to see things differently.

    We’d had a thread about probability theory here in prior comments above, just one aspect of seeing things differently. It’s arcane and nerdy. I don’t see why it was bombed by, apparently, another tribe who live in the next valley. So, other tribe, let us discuss our nerdy stuff around our own campfire. If you have wise words, sit and join. If you’re just about bombing the thread, why bother?

    Lastly. I saw something about whether to delete posts in this thread or not. Some saying no, don’t delete. It’s your prerogative, whoever it is who mods here. But I’d point out, if this is in part a commitment to free speech, then here’s a distinction: Free speech is a principle about viewpoint diversity. It’s about allowing dissenting or even apparently whacky views. But what it isn’t, is a carte blanche for poor behaviour, or substandard contributions. If someone comes drunk to a seminar and starts blathering, it’s fine to have them leave. If someone writes a paper that isn’t up to standard in a journal, its fine to reject it. Dear mods, it’s your private space, you can delete posts that don’t meet your implicit “terms and conditions” with regard to behavioural expectations around speech. As long as, I would say, you don’t delete good faith arguments that are minority opinions.

    Sorry, this post is really too long.

    LikeLiked by 3 people

86. Ian,

    I notice that the offending post has now been deleted. I don’t know by whom and that doesn’t matter. The poster who used the foul language in an attempt to deeply offend shouldn’t be surprised by this.

    I agree that the substantive points you and I have raised above have not been addressed by the recent contributors but I suspect that these points are either not thought of as being substantive or they are assumed to have little practical importance. This is not an uncommon reaction and it is not restricted to the climate modelling debate. I first became concerned with the inappropriate use of uncertainty analysis techniques many moons ago when I witnessed the widespread use of Monte Carlo to model project risk in the workplace. Same mistake, different setting. My objections were universally derided by my work colleagues.

    LikeLiked by 1 person

87. John –

    I’m in danger of very quickly getting in over my mathematical (let alone intelligence) head, but be that as it may….

    > But I ask how one can make that judgement reliably whilst using mathematical techniques for assessing ensemble skill that lack ‘epistemic validity’ (as Trenberth would say).

    Ok, so the first thing I had to do was look up “epistemic validity.” Having done so, what I see it’s a complex concept that I’m sure I have only limited mastery over – but to the extent that I correctly understand, I think the basic question to proceed further is to ask how you determine a ‘lack of epistemic validity.’ Seems a very complex calculation to me – and one that would necessarily tread into subjective-adjacent domains, if not outright into subjective territory. From where I sit, without delving further into that question – your determination looks to me like circular reasoning, or affirming the consequent. What Wikipedia tells me is things like “Epistemic validity uses both psychological and political dynamics when investigating social phenomena.” Psychological and political dynamics: Are there two areas which are less connected to subjectivity? So then, I need to know how you determine a lack of epistemic validity.

    So I actually wrote quite a bit more related to Steven’s comment after initially indicating that I think we can’t really proceed further without a clarification of terms. Not a good idea. And now I’ve pulled that part back out. I often find that because of the limitations of this medium, conversations get too defuse because there are so many different threads and online you can kind of jump around like “random access” whereas in normal conversations there’s more of an imposed “analog” sequence that makes the discussion more coherent. So I’ll just quit here, except:

    > I am not overstating a philosophical point here and I am not using it to justify “ideology-based aggression”.

    A few words about that. I get that you don’t want to focus on that issue. I do (for reasons I tried to explain above), but I don’t have to insist on it. I hope that we could return to it later. But I will just say, for now, for your consideration only but not for a response – that de-coupling this discussion from ideology-based aggression (or defense) is, IMO, a very demanding task – especially for people who are heavily identified along an ideological/identity vector (with respect to climate change if not necessarily along a related, distinct political vector), as we both are. In theory I would imagine it’s possible but I don’t think I’ve seen it as long as I’ve been observing the climate-o-sphere. I could even ground what I’m saying in your post about the floods in Pakistan, if that might help. That said (sorry for the long foray), I’ll let it go at least for now

    LikeLike

88. Alan –

    > For example, what is all this about modelling moderation? Is this merely a distraction from John’s analysis of problems inherent in assemblies of climate change model results?

    No.

    This notion of “distraction” in online discussions has always struck me as perplexing. Why would I have an intent of “distracting” from John’s analysis? Is the thinking that I’m lurking, with a deep fear of the outcomes if John’s discussion with folks at cliscep is able to continue unfettered? And so I feel a need to “distract?” I can assure you that’s not the case. Not only do I not fear such outcomes, neither would I think that my comments would materially alter the course of what goes on here. As I see it, what is going on here is part of a much, much larger complex of phenomena, that have played out over a very long period of time, over a wide variety of arenas and domains, and even trying my very hardest to “distract,” I’d have essentially zero effect in changing trajectories.

    No, I’m exploring ideas and understanding. I’m satisfying instinctive and addictive drives. I’m seeking validation. I’m challenging people who disagree with me. I’m criticizing. But no, I’m not offering “merely a distraction.” That isn’t to say that you might not see it as a distraction from your interests. In which case I suggest you not read what I write. If you’re distracted, that’s because you’ve decided to be distracted.

    I tried to explain why I think the discussion of moderation parallels discussions of modeling. Apparently I haven’t succeeded. I’d be happy to try again. At some point, if I keep trying and no one gets it, I have to accept that maybe what I’m going for just isn’t coherent (no matter if it seems coherent to me). I’d like to know if that’s the case. But I can’t get anywhere in finding that out if my interlocutor is assuming bad faith on my part – where they think that my goal is “mere distraction.”

    LikeLiked by 1 person

89. Joshua,

    Thank you for persisting. As I said in a comment earlier today, I would hope that some visitors here arriving via aTTP’s site would like interactions not to be unpleasant. You are doing a good job of persuading me that you are such a person, and are going some way towards restoring my faith in human nature.

    LikeLike

90. Mark –

    Thanks. Let’s see what happens.

    I’ll add another piece. I’m not saying that “trolls” don’t exist. They most certainly do. And I’ve been called a “troll” many times. In my own personal nomenclature – “troll” means someone who comes into a discussion and makes provocative statements with no intention of actually getting involved in a discussion to support those statements, but only to annoy and get under the skin of those who hold a differing viewpoint – usually being the “consensus” viewpoint at a given website.

    I’m here to discuss.

    ‘What I find is that “troll” typically gets applied merely as a means of pejorative labeling. It’s usually subjective. I’ve been called a troll many times even though (at least most of the time) I’m not merely jumping in to provoke, but I’m there to challenge the prevailing view with an intent of good faith interrogation. I suppose at times I’m “trolling” but that’s only if I’m sure that there’s no one around who’s interested in good faith discussion with someone who has my views. I have no particular reason to make that assumption here.

    LikeLiked by 1 person

91. Joshua,

    It’s interesting how we differed in our reaction to the Trenberth quote. Your first reaction was to assume ‘epistemic validity’ was a defined term and went looking for its definition on-line. My first reaction was to assume it wasn’t and sought to interpret it within the context of use. I’m not saying which was the better approach, but I will admit that, having now googled it, I wish I hadn’t. What wiki is saying about it casts doubt in my mind that Trenberth was using the term as wiki-defined. At the very least, I agree with what you say regarding the rabbit hole that talking more about ‘epistemic validity’ is likely to take us down.

    So perhaps the better thing for me to do is to concentrate instead on the Gavin Schmidt quote when he says:

    “Since we cannot hope to span the full range of possible models (including all possible parameterizations) or to assess the uncertainty of physics about which we so far have no knowledge, hope that any ensemble range can ever be used as a surrogate for a full probability density function of future climate is futile…”

    The way I might put this is that an ensemble range can never be used as a surrogate for a full probability density function because such a function would not be a valid expression of the epistemic uncertainty; hence there is a lack of epistemic validity.

    As for the lack of interest in discussing ideology-based aggression, it isn’t because I don’t recognise its ubiquity, nor do I deny there are those who may take my points and apply them in such a way. It’s just that I fear it may distract (that word again) from the technicality of the argument I am making. In fact, please believe me that having a climate-relevant ideology is not a required basis for my concerns. I could argue my point regarding the importance of delineating aleatory and epistemic uncertainty in models without once mentioning climate change. In fact, that’s exactly what I did in a professional capacity for a number of years (within the context of project risk and safety risk analyses) long before I started looking to see if the same problem was recognised within the climate science community. So if we are to discuss the perils of ideology-based aggression I would wish to do so another time.

    LikeLike

92. John –

    > It’s interesting how we differed in our reaction to the Trenberth quote. Your first reaction was to assume ‘epistemic validity’ was a defined term and went looking for its definition on-line.

    Not quite. I didn’t know to what extent it was a defined term, and not having context to work out a likely meaning (other than your basic reference to it), had no where else to go but investigate a bit

    > The way I might put this is that an ensemble range can never be used as a surrogate for a full probability density function because such a function would not be a valid expression of the epistemic uncertainty; hence there is a lack of epistemic validity.

    Well, that does seem circular to me. It seems to me like saying you think there’s a lack of epistemic validity because you think using an incomplete probability density function isn’t valid. Seems to just go back to how “valid” is defined. And I’d argue that’s going to be subjective, at least to some degree. It’s going to be, at least to a meaningful extent, based on values. Or even more so (as I think we tend to underestimate the extent to which values are shared, when we reverse engineer from differing views to conclude that’s because of differing values [I imagine I’d have to unpack that a bit, but I’m hoping it’s meaning is reasonably apparent without my having done so]), based on positions which are staked out based on a variety of input signals.

    So let’s see if that makes any sense to you before we go any further. I’m not suggesting that you’d agree – but you think that what I’m saying makes enough sense to you that you could paraphrase it in a way I’d agree captures the gist of it. *

    > As for the lack of interest in discussing ideology-based aggression, it isn’t because I don’t recognise its ubiquity, nor do I deny there are those who may take my points and apply them in such a way.

    I think it’s important to note that I’ve been clear to state that I’m talking about identity-defense as well as identity-aggression. Not to be a stickler, but you stated it once only in the aggressive component and I repeated the addition of the defense part. So I’m being explicit this time. I think it’s important. And I’ll point out that in this very thread, at the very beginning of this branch, there was a framing of a reason to defend “skeptic” views on mathematical modeling from an (presumably aggressive) intrusion via mischaracterization, or perhaps misunderstanding (willful or malicious or otherwise).

    > It’s just that I fear it may distract (that word again) from the technicality of the argument I am making. In fact,

    I get that.

    > please believe me that having a climate-relevant ideology is not a required basis for my concerns.

    Sorry, but I have to dig in here. I want to stress that it’s not a matter of disbelief, per se, if I find that testimony insufficient.

    If I’m not capable of fully understanding the issues at hand, I hope to be enlightened – but as I understand it, what we’re discussing here is the question of modeling where there are some basic views which exist in tension with one another along a spectrum of valid views, but which are in opposition as well as are both reasonable basic, logical, viewpoints. A short version would be something like:: mathematical modeling is tough. All mathematical models are wrong, but some are useful. Within that imperfect world, sometimes models are too uncertain to be of practical value (but maybe get used anyway). And sometimes it’s beneficial to integrate their results despite the degree of embedded uncertainty (but they get dismissed because they fail to reach an unreasonable standard of certainty). When we enter that dynamic of tension, we have a basic foundation of experience and knowledge and history and, well, ideology. I’m not saying that our ideology defines our view, but that ideology and history and knowledge act like a moderator variable on the relationship between our foundational understanding and our ultimate viewpoint in a given context. I don’t see that process as actually avoidable, except in rare circumstances. But all that much more rare when our identity is materially shaped by those foundation inputs.

    > I could argue my point regarding the importance of delineating aleatory and epistemic uncertainty in models without once mentioning climate change.

    Sure. I don’t doubt that. But that doesn’t mean that your point on that issue, based on previous knowledge and experience, doesn’t affect how you reason about modeling in the context of climate change

    > In fact, that’s exactly what I did in a professional capacity for a number of years (within the context of project risk and safety risk analyses) long before I started looking to see if the same problem was recognised within the climate science community.

    So, for me, this rather underscores my point.

    > So if we are to discuss the perils of ideology-based aggression I would wish to do so another time.

    Which is absolutely fine. No need to go further into that realm explicitly. But I felt a response in line was in order.

    * Sorry for being verbose and perhaps unclear. I have a lot of views on these issues and it’s hard for me to (1) cut parts out because many of them feel critical to me and, (2) get my points across without even there taking some shortcuts despite how lengthly it turns out. I’ll fully understand if further engagement requires more of an commitment than what you;’re interested in. Perhaps the way to go is to limit the focus, at least for now, to the point of the asterisk above.

    LikeLike

93. Joshua – changed my mind – you really are a teacher.

    LikeLike

94. dfhunter –

    Is your goal to provide an object lesson for my definition of trolling? If so you’re on the right track.

    LikeLike

95. Joshua,

    We are talking about mathematical validity here. The term is being used here in the same sense that one might say it is mathematically invalid to ignore negative solutions to a quadratic equation. Probability can be used to express variability and incertitude and, depending upon which is the case, different statistical approaches apply and different conclusions are mathematically justified. In many cases, it is not obvious to what extent variability or incertitude are being represented by the probability. It is not valid in such circumstances to proceed as if only variability applies, thereby using the branch of mathematics that applies to variability. When you ask a set of experts with differing views, each lacking a full understating, you are not taking a measurement, so it is inappropriate to analyse the spread of views using measurement theory. The same applies to their models. If you have used mathematical techniques that are not designed to capture the epistemic uncertainty then claiming that you have properly analysed the uncertainty would be unjustified, unless you can demonstrate that the uncertainty is purely aleatory. That is the sense in which epistemic validity is absent, though the real point is that it is not mathematically valid.

    Everything I have said regarding aggression applies equally to defense. I was just being lazy by not referencing both. My mentioning of encountering the problem outside of climate science was an attempt to emphasise that a mathematical universal is at issue here, rather than a validity that might be domain specific.

    LikeLike

96. Joshua. I could be wrong but I would have taken dfhunter’s late recognition of you as really being a teacher as a real positive. It suggests to me that henceforth he will give your writings more credence.
    I judge that now you have at least two people here who have positively changed their minds about you. That may not bother you but it is at least unusual.

    LikeLike

97. Joshua,

    I should add that I am not disputing what you say regarding value judgements. All I am saying is that when value judgements are made under uncertainty, the situation is not helped when the mathematical basis used to assess the uncertainty is unsound. As I used to say many times in my capacity as a functional safety analyst, it is not just the gap between what you know and what you need to know that is important, it is also the gap between what you know and what you think you know.

    LikeLike

98. John –

    I’ll start with your follow on comment first…

    > I should add that I am not disputing what you say regarding value judgements. All I am saying is that when value judgements are made under uncertainty, the situation is not helped when the mathematical basis used to assess the uncertainty is unsound. As I used to say many times in my capacity as a functional safety analyst, it is not just the gap between what you know and what you need to know that is important, it is also the gap between what you know and what you think you know.

    As that fairly well addresses what my response to your first comment was going to be. (and then I’ll ramble quite a bit).

    The problem for me is that I can’t assess the mathematical soundness on my own, My assumption that it’s not (usually) practical to consider it a dichotomous state of soundness versus unsoundness, but that instead there are varying degrees of soundness and unsoundness, as it were, centering around some decision point. At some stage, we have to stake a claim towards one direction from that centering point. And how we do that mixes in many elements. Certainly mathematical skill would be an important one. But values (which again, I personally mostly reduce to ideological orientation with values as a moderator) is a rather critical and inescapable one, in my experience.

    I’m going to assume that there are people with mathematical skill who would differ with you on where the lines get drawn here. So the idea of “mathematical validity” seems like a black box for me, with a variety of outputs. I’m not inclined to just take anyone’s word for it, but I try to assess probabilities from a bird’s eye view.

    I appreciate, if I understand you properly, your point about the importance of not conflating variability and uncertainty. (Even if I’m wrong that’s what you were getting to, it’s something I’m going to ponder a bit. It brings to mind some of what Kahneman talks about when discussing the relationship between variability and noise).

    As for now, when you say this:

    > It is not valid in such circumstances to proceed as if only variability applies, thereby using the branch of mathematics that applies to variability

    It’s still a bit too abstract, as someone without technical skills here or an overabundance of intelligence, for me to put my head around. It’s not clear to me how climate modeling operates on proceeding as if only variability applies. So here are some rather inchoate thoughts.

    When Gavin says this, “Collections of the data from the different groups, called multi-model ensembles, have some interesting properties. Most notably the average of all the models is frequently closer to the observations than any individual model. But does this mean that the average of all the model projections into the future is in fact the best projection? And does the variability in the model projections truly measure the uncertainty? These are unanswerable questions…”

    I think of a question I’ve had for quite a while about whether “ensembles” are averaging among a variety of (mostly independent) projections – which never quite made sense to me. I think of the image I had when I read Steven’s post. I see a bunch of projected lines of the Hurricane. None (or hardly any) of them go towards the east (I think it shouldn’t be capitalized there) side of Florida. There’s a chance the east side will be hit, but living there I’d feel comfortable enough not to board up my windows, say. But I’d check to make sure the gas tanks for my generator weren’t near empty. If I were living on the west side, if I looked spatially at the projections, I wouldn’t average between the various projection east to west and assume the probability of getting hit is higher in the center of the range. I’d use other factors to try to assess probabilities

    But I”m not getting to where you’re saying, with such certitude, that what the models do “is not valid.” Or how they proceed as if only variability applies – in particular because there are confidence intervals attached to the probability assessments. And I’m still stuck on the notion these issues existing on a spectrum of when something is more or less recommendable, depending on a variety of elements, and the different people have a different take on how various criteria for judgement.

    And without “appealing to authority,” I find it hard to believe that if this issue were so black and white, there would be such a large number of qualified people who are stumbling around blind to this question. Of course, there could be a uniform biasing influence among them. It’s certainly not out of the range of possibility. But, IMO, that potential has to run in different directions.

    I’ll look at what you wrote a bit more.

    As for this:

    > Everything I have said regarding aggression applies equally to defense. I was just being lazy by not referencing both. My mentioning of encountering the problem outside of climate science was an attempt to emphasise that a mathematical universal is at issue here, rather than a validity that might be domain specific.

    I get that you’re talking about universal issues, that aren’t domain specific, and that you’ve interrogated across contexts. My point in that regard is that you would apply that generalized experience differently in different contexts, and that I don’t see how you could avoid the full range of influences when doing so, including a variety of values- (or ideology-) associated cognitive and psychological mechanisms. I have heard many, many people say that they’re doing that with a full conviction that they are – but when there’s so much complexity involved, and so much uncertainty and potentially subjective evaluations, mixed in with strong identity orientation – I can’t recall observing a convincing example (even if I can’t always rule it out. Contrary-wise, I have seen many situations where when people were fully convinced that was the case, from my observations it was obviously not the case. The ability to fool oneself in that regard is quite robust.

    LikeLike

99. Alan –

    Let’s see if dhunter clarifies. I’d be more than happy to apologize for my snark if I’m wrong.

    LikeLike

100. Joshua,

     I don’t know what I can say at this point that won’t sound arrogant or patronising, but the truth is that this is an area that I have studied for several years now, both for professional reasons and, since retirement, out of sheer curiosity. I wouldn’t expect someone who hasn’t shared that obsession or focus to necessarily appreciate everything I am getting at. If I may, I would like to respond primarily to the following:

     ‘And without “appealing to authority,” I find it hard to believe that if this issue were so black and white, there would be such a large number of qualified people who are stumbling around blind to this question. Of course, there could be a uniform biasing influence among them. It’s certainly not out of the range of possibility. But, IMO, that potential has to run in different directions.’

     Welcome to my world 🙂

     The fascinating thing is that there are in fact so many ‘qualified people who are stumbling around blind to this question’, and not just in the field of climate science. If I were the sort who was impressed by authority I would still be amongst them, never having questioned the orthodox view on this subject. Some personal history may help here:

     All of this first started for me when I was advised to use Monte Carlo simulation to improve the detail and reliability of project outturn predictions for the software development company I worked for. I won’t go into detail but, because of my physics background, I was already familiar with using Monte Carlo simulation to model system variability but I couldn’t see how it would help in the modelling of project risk (particularly in a software development context). I couldn’t articulate my concerns very well at the time, so I started researching to see if my concerns had any proper mathematical basis. I quickly discovered that they had, and that the basis was well-established and documented, even within the literature of Monte Carlo. Take the following EPA technical guidance, for example:

     https://www.epa.gov/risk/use-monte-carlo-simulation-risk-assessments#:~:text=Monte%20Carlo%20simulation%20also%20has,well%2Ddescribed%20differences%20among%20individuals.

     So, the question I then had was why I was surrounded by so many qualified people who were stumbling around blind to this question. It wasn’t that I was wrong and everyone else was right — I had suspected that something was wrong with the orthodox approach, and I had confirmed it through investigation. Since then, my thinking has been widened to try to understand how orthodox views can become so entrenched, and followed by qualified people, whilst being patently wrong. I won’t go into my detailed thinking on this phenomenon but instead finish by repeating a climate modeller’s quote that I like to use:

     “What they were very keen for us to do at IPCC [1990], and modellers refused and we didn’t do it, was to say we’ve got this range 1.5 – 4.5°C, what are the probability limits of that? You can’t do it. It’s not the same as experimental error. The range is nothing to do with probability – it is not a normal distribution or a skewed distribution. Who knows what it is ?’”

     So here is the question to ask: Why, if they knew it was wrong and refused to do it back in 1990, are they all doing it now? Is it the same reason why everyone is still using Monte Carlo to analyse epistemic project risk when only a little investigation is needed to determine that it is invalid?

     LikeLiked by 2 people

101. John –

     > Since then, my thinking has been widened to try to understand how orthodox views can become so entrenched, and followed by qualified people, whilst being patently wrong.

     I’ve been following Andrew Gelman’s statistics blog pretty closely for going on two years now (despite knowing next to nothing about statistics). A common topic of discussion there is among a subset of very knowledgeable statisticians, who regularly decry the ubiquitous misuse of statistics – such as through an over-reliance on Null Hypothesis Statistics Testing- throughout the research space across many domains. I certainly think that orthodox views are not unimpeachable, by any means.

     That said, at Gelman’s blog, I see a very active discussion among statisticians and researchers who presumably once adopted orthodox views but who have adapted and changed. I see an active discussion and debate about the extent of the problems. I see people who, IMO, over-egg the pudding and see invalid NHST around every corner. There’s one guy – very smart with a lot of domain relevant expertise – who seems quite convinced that as a result of invalid reliance on NHST, medical science in balance causes far more harm than benefit. My point being that there’s a lot of uncertainty there surrounding the implications of the orthodox views.

     So you say…

     > I don’t know what I can say at this point that won’t sound arrogant or patronising,…

     I don’t really think those attributes are the point. I think this isn’t about arrogance or being patronized. I think that’s personalizing the issue at hand, when this is a systems problem. I think it’s a problem involving uncertainty and basic human psychology and cognitive attributes (and biases). I think we don’t have to personalize it because this is a question of modeling more general features that don’t really function at the individual level. I think there’s a basic structural tension in play. Or I should say a number of structural tensions that exist in a number of different vectors or dimensions.

     One dimension might be: to what extent do orthodox views protect against views that are too much a product of idiosyncratic reasoning – perhaps borne out of the uniqueness of an individuals experiences, or a person who for one reason or another has a particular axe to grind, or who’s holding a hammer and looking for nails? And on the other side, to what extent do unorthodox views protect against a calcification or inertia or group think or “capture” (say “elite capture” or “epistemic capture”) among the orthodoxy? Both ends of the model have strong validity, IMO, even though they might seem mutually exclusive.

     I don’t see any way to apply a general rule here. With NHST, I’m seeing enough momentum stemming from (what once was more so) unorthodoxy, to think that the tide is turning against an orthodoxy. But yet that doesn’t mean that there’s a widespread agreement, even at a place like Gelman’s blog, that medical science causes more harm than good – although there is certainly a cohort of aligned critics of practices like NHST who seem quite convinced that’s the case.

     So someone like myself has to wander through the landscape and try to find where the obstacles and signposts are.

     So not to personalize it, but I do see someone with a particular identity orientation, who has a particularly strong view on a particular issue directly related to that identity orientation, with much of a legacy experience in struggling with that issue, applying a very strongly held perspective derived from expertise, in opposition to an orthodoxy. That matches a model construct that I have (as mentioned above) with tensions that I think structurally will come into play. So I look for signposts as I wander the landscape. One, critical signpost, is if someone that I see as almost necessarily operating from within that model, saying that they function outside of that model. Again, I know it will seem hard to believe, but I’m not personalizing this. I see this happening all over the place in all different kinds of domains. And this claim isn’t only being made by the unorthodox. It’s also a claim that’s almost always implicit if not always stated, among the orthodox as well.

     I’m not weighing in on your personal situation. You may, indeed, be residing outside of the structural tension for all I know. It happens. All I can do is consider the probabilities as they match my experience.

     As for the Monte Carlo reference, and the quote from the climate modeler. They seem relevant and on point. But they necessarily lack context. What would that modeler say about your reference? How would other modelers (in disagreement) respond to your application of that reference? What would the author of the EPA warning about the use of the Monte Carlo method, have to say about your specific application to any particular context? What do other references have to say on this issue of distinguishing variability and uncertainty, and in particular with application to the contexts you’re focusing on? That information wouldn’t be dispositive. There are always vested interests in protecting an orthodoxy against an unwanted intrusion, for all manner of reasons – none the least because of identity-related cognition. The information provided by other modelers or other sources would need to be fed into my model.

     In this particular frame, I think of the modeling of Stevens and Palmer. I would imagine the quote you provided might have been stated by one of them. They seem pretty critical of the treatment of uncertainty in the climate modeling community more generally. But thinking back to Steven’s (Mosher) comment that started all of this. It seems to me Stevens and Palmer are pretty clearly aligned with Steven regarding the general sign (positive or negative) of the functional value of climate modeling, despite the uncertainty or embedded problems (which I don’t doubt fit within a model) of how climate models distinguish between variability and uncertainty. Again, I though his use of the hurricane forecast as an analogy worked quite well – as an entry into thinking about, potentially high damage function, perhaps low probability, risk that plays out over long time horizons, where many, many cognitive biases related to risk perception, come into play.

     LikeLike

102. Sorry – Null Hypothesis Significance Testing

     LikeLike

103. we have Satellites that can give a good landfall estimate for hurricane, plus we have historic data (models only give further estimate/spread)

     LikeLiked by 1 person

104. Joshua,

     >”But thinking back to Steven’s (Mosher) comment that started all of this.”

     In point of fact, Steven’s comment didn’t start all of this, it is just where you came in. Actually, ‘all of this’ was started by me when I commented at the head of the thread by saying, “the range of possible future states generated by ensembles of model solutions does not form a probability distribution.” Reacting to this, someone with the username ‘Ianalexs’ requested that I elaborate in a full article. Instead, I suggested that he do some background reading and pointed him in the right direction. Having followed my advice, he was able quickly to return with contributions of his own that demonstrated that he fully understood the issue. What he would have found in his research are countless papers explaining the duality that lies at the heart of the concept of probability and the implications that has for the mathematics of uncertainty analysis. What he would not have found (because it isn’t there) is anything that argues that the concern is misguided or that the criticisms of probability theory are unjustified. That the issue has implications for the ability of a probability distribution to capture epistemic uncertainty is not controversial. There are only those who know about the problems and the mathematical techniques proposed to address it, and those that don’t. Statistics is replete with views in opposition, but this is not a case of that.

     The other point I have tried to get across is that this is a generic issue and that I came to my current level of understanding long before I took any interest in climate science. However, once I did, I looked in the expectation of finding climate scientists who know about the problems and the mathematical techniques proposed to address it, and those that didn’t. What intrigued me, however, is that there was a third group who knew about the problem but ignored it anyway. I don’t think I had seen that in any other domain. I suspect that there may be a number of reasons why this is the case but finding a scientist who is prepared to justify such a stance is rare. That’s why my reference to the A. M. Foley paper was important. He is one of the few who I have found attempting to justify the continued treatment of ensemble spreads as if they were probability distributions when he knew they were not. Ultimately, he seemed to think it was justified because the concern is rendered academic by the application of the precautionary principle. Steven Mosher also attempted to persuade me that the correct mathematical treatment of uncertainty is not the issue, but for different reasons. It strikes me that it is these individuals who are ‘struggling’ with ‘structural tensions that exist in a number of different vectors or dimensions’. All I am doing is insisting that the correct branch of mathematics be applied because the wrong one is prone to lead the analyst towards the wrong conclusions. I don’t need to worry about vectors or dimensions to justify that stance.

     Finally, I’d just like to remind you that I am not arguing for the abandonment of mathematical modelling in climate science, nor against the practice of using model ensembles.

     I sense that this debate has run its course now, since I am in danger of repeating myself. Please feel free to add your own closing statement. If it demonstrates that I have failed to convert you, I’ll not be surprised. A capitulation and transfer to the dark side was never on the cards. You have been frank, sincere and courteous throughout and that is all I can ask for. Thank you for your interest and please feel welcome to return to this site any time.

     LikeLiked by 2 people

105. John –

     Thanks for your comment. Reading back over what I’d written I saw that I was getting repetitious and I agree the convo has most likely approached using up its usefulness, but I will add some more comments.

     First, when I was referring to when this all started I was thinking of my entry point. More specifically, where Steven said:

     the best science, with gaping holes in understanding, predicted hits from tampa to fort meyers.

     and the best science said definately the west side and not the east.

     was east possible? yes logically possible. we dont know everything

     and which I was thinking was the root of our disagreement.

     From reading your last comment, I get the impression you thought I was arguing against the idea that there’s a duality that lies at the heart of the concept of probability and the implications that has for the mathematics of uncertainty analysis. I won’t pretend to have a full understanding of the issue, but at the level I can understand it what you’re saying there makes intuitive sense to me.

     If in all the paragraphs I wrote, I gave the conveyed that I think the concern is misguided or that the criticisms of probability theory are unjustified then it just goes to show that sometimes writing more only compounds a lack of clarity. To the extent that I was suggesting (without actually knowing) that there might be differing views on that issue in the statistics space, it was only that there might be different views about the parameters of, or significance of, that problem in various applied contexts.

     Finally,

     > If it demonstrates that I have failed to convert you, I’ll not be surprised. A capitulation and transfer to the dark side was never on the cards. You have been frank,

     I know that was more or less a throw away line… and I’m sometimes accused by people in my life of being too concrete… but I want to stress that I don’t see light and dark sides here. I don’t see any conversion to take place. I see these as parts of an integrated whole and that the tendency to see them in some kind of binary or zero sum frame comes not from the intrinsic features of the questions, but from how we reshuffle the issues to line up with our cognitive and psychological biases. A benefit of this discussion with you here for me is that it’s helped me to consolidate my awareness of how strongly I view these questions as an integrated whole .

     A suggestion: Gelman often writes posts about issues in statistics, based on letters or emails written to him asking him to address those issues. He has weighed in on statistics issues in climate science in the past. The commentariate there is not terribly partisan on a lot of issues (although there’s clearly a general bias towards accepting the consensus view on issues like climate sensitivity, there isn’t at all uniformity), but it comprises people with a very sophisticated grasp of issues in statistics who tend to engage in quality discussion with very little of the petty bickering typically seen online. To the extent that you do have any interest in seeing that community discuss how you’re overlaying general issues in the mathematics of uncertainty analysis onto the work of climate modelers, you might consider sending him an email. If he did put something up, I, at least, would find the ensuing discussion interesting.

     Thanks again.

     LikeLike

106. Alan –

     I”m thinking that the lack of clarifying comment from dfhunter increases the likelihood that my impression was the correct one.

     Not that that (his?) impression about teachers really matters to me, but it’s an interesting window into the dynamics of social media engagement.

     LikeLike

107. Joshua,

     Thank you for your further comments and clarifications. I’ve taken a look at the Gelman blog and I agree that it is worth tracking. Whether or not I will take up your suggestion any time soon depends upon personal commitments. As I explained to Ianalexs, I’ve got quite a lot on my plate at the moment.

     LikeLike

108. John, I’m just commenting to agree that this topic has run its course, at least for now. It’s been a great run. As you say, we covered some pretty useful stuff about epistemic/aleatory uncertainty up toward the beginning of this thread. You provided some links back into cliscep where the issue had been previously discussed by you. On my part, I think the best I found was E. Winsburg 2012, “Values and Uncertainties in the Predictions of Global Climate Models”, Kennedy Institute of Ethics Journal 22 (12) — which I quoted way up above, with commentary. But for benefit of newer entrants to the thread, to recap:

     “Ensemble methods assume that, in some relevant respect, the set of available models represent something like a sample of independent draws from the space of possible model structures. This is surely the greatest problem with ensemble statistical methods. The average and standard deviation of a set of trials is only meaningful if those trials represent a random sample of independent draws from the relevant space — in this case the space of possible model structures. Many commentators have noted that this assumption is not met by the set of climate models on the market. … Perhaps we are meant to assume, instead, that the existing models are randomly distributed around the ideal model, in some kind of normal distribution, on analogy to measurement theory. But modeling isn’t measurement, and so there is very little reason to think this assumption holds.”

     I too haven’t found much time in the past few days. Certainly not enough to track down more about the issue in hand. Although I did find this: “Monte Carlo Methods in Climate Science”, Math Horizons Nov 2013.
     . Not that it’s sceptical but seem to provide an overview. On the sceptical side, although I haven’t had a chance to dig in, my understanding so far is that to run a Monte Carlo simulation, you have to provide on the *input* side a probability distribution. That is, you’ve predetermined a main characteristic of the output, based on your initial hunch. If you get a chance sometime to talk about that (am I right? I just haven’t had the time to sit down and look at this), I’d be all ears. Regretfully though I will have to close multiple browser tabs where I have papers waiting to be read about this topic, and move onto other pressing things.

     But before I bow out for now, I can’t resist this observation: You say “Since then, my thinking has been widened to try to understand how orthodox views can become so entrenched, and followed by qualified people, whilst being patently wrong.”. I recall back to April, when the story broke that Napier Uni in Scotland had been training midwives as to how “birthing persons” can give birth through their penis. These are professional people, who seem to have grotesquely contorted their reasoning so as to fit in with prevailing group ideology. Goodness knows then, what else groups of experts have come to believe, in less publicly scrutinised fields, based on in-group social dynamics rather than evidence or logic. Certainly a reason to maintain an attitude of scepticism, now more so than ever.

     LikeLiked by 1 person

109. Joshua,
     when I said above “changed my mind – you really are a teacher” – I meant that.
     I just have to re-read your comments with John to confirm that.

     from the start snippets from your comments went like this –
     “I suppose discussions with Hitler might be interesting. Right?”
     “I feel you, man. If only those alarmists would stop calling us names, we could get on with more productive lives.”
     “That problem construct is like a siren call for ideology-based antipathy and motivated reasoning”
     “So then could you explain the difference in your view between alarmists (an obviously welcome descriptor) and entitled twats? I can think of many dimensions of difference but other, I think important and probably in practice most operational, dimensions of similarity.”
     “Hence my reference to alarmists and agendas. Perhaps you could rethink those concepts and really evaluate why they remain here – in fact firm the backbone of a top line post – and caling you an entitled twat would be deleted?”

     and I could go on, but you come across as a teacher using big words/concepts to impress rather than educate.

     ps – sorry Alan.

     LikeLiked by 1 person

110. Ian,

     You have basically got it right regarding Monte Carlo. If you have a mathematical model of a physical system, it may feature uncertain variables that quantify boundary conditions. Rather than select single values for these before running the model to see what answer it gives, one can use Monte Carlo simulation to repeatedly sample from their sets of possible values and then present the spread of answers this gives you. The sampling is based upon a posited probability distribution for each such variable. The spread of answers can be used to evaluate the uncertainty inherent in the model. The problem, however, is that the posited probability distributions from which the sampling takes place are supposed to be an expression of physical variability, i.e. it’s akin to taking a measurement. The problems arise when epistemic issues are at the root of the boundary condition uncertainty, which would mean that the choice of probability distribution from which to sample is itself a decision made under uncertainty. As the EPA put it in their technical guidance:

     “Monte Carlo simulation also has important limitations, which have restrained EPA from accepting it as a preferred risk assessment tool. Available software cannot distinguish between variability and uncertainty. Some factors, such as body weight and tap water ingestion, show well-described differences among individuals. These differences are called ‘variability’. Other factors, such as frequency and duration of trespassing, are simply unknown. This lack of knowledge is called ‘uncertainty’. Current Monte Carlo software treats uncertainty as if it were variability, which may produce misleading results.”

     I don’t like their terminology because I prefer to refer to uncertainty as coming from variability or incertitude. However, what they are saying is perfectly true and it is telling that they felt the need to issue such guidance. I suspect misuse of Monte Carlo simulation is widespread, since Monte Carlo can only be safely used once the aleatory and epistemic uncertainties have been isolated.

     LikeLiked by 2 people

111. Dfhunter,

     I agree that Joshua came on too strong for my taste initially. In the past, I might have been tempted to try a rebuff, but on this occasion I was determined to stick to my technical point. Besides which, experience has taught me that online fisticuffs is rarely satisfying in the longer term. I stand by what I said, Joshua is welcome back any time, although I cannot promise that I will welcome everything he says.

     The same goes for Willard and Hitler.

     Well, Hitler at least 😉

     LikeLiked by 1 person

112. John –

     I’m curious if you might respond on this aspect of dfhunter’s characterization:

     >and I could go on, but you come across as a teacher using big words/concepts to impress rather than educate.

     Not only regarding how I “come across,” but also regarding the characterization of teachers. Notice how it went from I must be a teacher based on my comments to I come across a teacher doing something specific. I’m also curious about this:

     > I agree that Joshua came on too strong for my taste initially.

     As to what that means. What did I do that registers for you as coming on too strong?

     LikeLike

113. Joshua – you said above –
     “This relates to my methodological bias as a teacher: you demonstrate mastery of a concept on a more concrete domain before moving into mastering the concept in a more abstract domain. Not a necessary sequence but often helpful.”

     ok – what subject do you teach with mastery?

     as to your question to John – he will reply if he can be bothered – but repeated reuse of being called a “tw*t” for asking an honest question?. that you can’t see that as “coming on too strong” beggars belief.

     LikeLike

114. ps – nothing against Teachers when they Educate & Inspire.

     LikeLiked by 1 person

115. Dfhunter,

     Thank you. I think you have made my point for me.

     Actually, entitlement has never been what this is about, otherwise I would have just pointed out that Joshua is not entitled to a response (I’m sure he knows this anyway). The negative input started following my observation that engaging with what the other person is actually saying is more likely to result in a mutual understanding than if both sides just talk about the other to a third party. The case in point related to the internal discussion over at ATTP regarding what sceptics think about modelling; a discussion which led me to suspect that they did not actually understand what sceptics think about modelling (at least as far as my own views on the subject were concerned). The results of the invitation to respond directly to my comments on the subject were disappointing to say the least, but they were at least enough to confirm my suspicions. Ken and his moderator are now to be found over at ATTP high fiving each other. Ken admitted that he had been lurking here. I might ask why he did not feel motivated enough to point out where I, Trenberth, Schmidt and Winsburg (to name but four) had gone wrong with our reasoning, or at least why such concerns can be ignored in climate science when they are not in any other domain.

     I might ask, but I won’t. I would not now expect anything constructive to emerge from such an invitation.

     LikeLike

116. John –

     Just fyi, I find it rude that I asked you a question, someone who’s established he’s not interested in any meaningful engagement with me answered my question, and then you went on to discuss with him, his answer to me.

     I’m still hoping that you’ll answer my question.

     You said that I came on too strong. I’m curious if you could break that down a bit. The reason, is because I am wondering if the “too strong” aspect of my comments was distinguishable in meaningful ways from, say, this at the top of the post:

     Despite the Academy’s manifest pieties, my opinion is that it seeks through its climate agenda to

     ♦ Lower our living standards through more expensive energy. European households’ bills, for example, are forecast to rise by $A3 trillion at their peak early next year

     ♦ Destabilise our electricity grid with forced introduction of unreliable, intermittent solar and wind-power

     ♦ Destroy half a century’s progress in fuel, mineral and farm exports through net-zero goals for which no practical technology exists – only fantasies of green hydrogen, mega-batteries, carbon capture and storage, and plucking CO2 from the atmosphere.[1]

     That’s obviously a pejorative description of an “agenda.” As I said, anyone coming to comment here, who might support the climate agenda that’s being described, would be coming in as someone who’s effectively labeled as seeking to lower living standards, destabilize an energy grid, and destroy progress across many different domains. In other words, they’re likely to think they’re identified with seeking to to cause widespread suffering. It’s one thing to believe that the result of someone’s agenda will cause suffering. It’s another to say that someone seeks to, through an agenda, cause suffering.

     IMO, the description there suffers from a lack of “cognitive empathy” (the term can be intuitively misleading. If interested you might look it up, but basically it refers to perspective taking).

     So I’d like to know whether you thought the OP came on too strong. I assume you didn’t – in which case I’d like to know why what it was about what I said that struck you as coming on too strong.

     As for this…

     > engaging with what the other person is actually saying is more likely to result in a mutual understanding than if both sides just talk about the other to a third party.

     I agree.

     LikeLike

117. John –

     I just wanted to add that getting Gelman (and the community of statisticians at his blog) to discuss your view of a fundamental flaw in climate modeling would not take much of a time commitment on your part.

     Just writing a fairly brief overview (taking far less time that you’ve taken in these discussions with me) and shooting it to him in an email would be all that it takes.

     He has never failed to respond to emails I’ve sent to him (I never figured out how he finds the time to do all that he does). Of course, he may not be interested in responding to your email, let alone putting up a related post. But, I think that if your analysis gets to the root of an obvious fundamental mathematical/conceptual flaw in climate science modeling, he would find it interesting and might run with it a bit.

     I would imagine that you, as well as I, would find it interesting if he did.

     LikeLike

118. Joshua,

     I apologise if you found me rude; that was not my intention. You might however reflect upon how rude it would have been for me to speculate upon dfhunter’s own contribution and how inappropriate it may have been for you to request that I do. It is my sincere wish that we drop this whole line of argument.

     I do not wish to open up a fresh front on the debate. If I were to do so, I would have to be prepared to follow it up and I am not sure what that would entail. I have not ruled it out, but the timing is not good at present. Unfortunately, neither you nor Willard are in any position to judge whether I am responding appropriately given the personal demands currently being made upon me. Suffice to say, I know that I have already expended far too much time on this thread and I already feel guilty about that.

     Honestly, you have no idea, and I am damned if I am going to go into personal detail for anyone’s benefit.

     LikeLike

119. John –

     > I apologise if you found me rude; that was not my intention.

     No apology needed. I assumed it wasn’t your intention to be rude towards me. I should have been explicit about that. If I had thought it was your intent, I wouldn’t have bothered bringing it up. What would be the point in pointing out rudeness to someone who was intentionally rude? I realize I should have used a different term than “rude” – which tends to convey intent.

     I look at it as a structural problem that I see frequently in blog discussions, where there’s discussion between people about other people. For me, it even goes to when a response to me is directed towards a putative audience – where I’m having a discussion with someone and they respond with something like “When Joshua said X, what he failed to recognize was Y.”

     I think that’s one of the structural pieces that causes problems with the online medium for discussing these issues. There is an inherent confusion about who the audience is. When you’re arguing it out in a pub about something, there’s much less likely to be such a confusion. I think that confusion has an effect of degrading online discussions.

     > You might however reflect upon how rude it would have been for me to speculate upon dfhunter’s own contribution

     I don’t get why it would be rude for you to give an opinion regarding dfhunter’s comments directed at me when I asked you specifically about his comments. That’s an entirely different configuration than you, instead of responding to me directly about a direct question to you (even a question about djhunter’s comment), discussing my question with dfhunter with no direct response to me.

     > and how inappropriate it may have been for you to request that I do.

     I fail to see why it is inappropriate to ask you to comment on someone else’s comment. Maybe you can help me to understand. Should any of our comments here be immune from interrogation? I don’t think so. I certainly never have such an expectation for any of my comments. I don’t think any of us should be sensitive such that we’d take offense to having our comments discussed.

     > It is my sincere wish that we drop this whole line of argument.

     Then I won’t continue beyond this comment – but I happen to think it’s pretty important. I think it gets to the heart of why it’s difficult to have the kinds of discussions you expressed an interest in having. I think a failure to take on that discussion directly is regrettable. And as an aside, it gets back to why I thought that a discussion of the “model” for moderation was relevant to discussing climate modeling.

     > neither you nor Willard are in any position to judge whether I am responding appropriately given the personal demands currently being made upon me.

     I make no such judgement. I never intended to convey a judgement about whether you’re responding appropriately on any particular front. Obviously, what is and isn’t appropriate is an entirely personal evaluation. How could I possibly assess what’s appropriate for you? Again, I make no such judgement.

     I was merely making it clear that Andrew is very responsive and the time commitment it would take to ask him to respond on point could be minimal. Of course, a minimal effort might not seem satisfactory for you, in which case reaching out to him in a way you would find satisfactory would necessarily require a significant time commitment.

     LikeLike

Leave a Reply

Enter your comment here...

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

Email (required) (Address never made public)

Name (required)

Website

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

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

Cancel

Connecting to %s

Notify me of new comments via email.

Notify me of new posts via email.

Δ

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