Brian Cox took a bit of flack because of a Guardian article about a speech he’d given in which he appeared to be suggesting that – when discussing climate science – scientists should sound more certain than they are. I don’t think this is what he was suggesting. What I think he was arguing was that we should be careful not to sound less certain than we actually are; almost the opposite of what some have interpreted him as saying.
Brian has clarified what he meant in a post on his own blog. He argues that the term uncertainty is often misunderstood and misused. It doesn’t mean we’re uncertain. If anything, it’s the opposite. It represents a confidence interval; it tells us how confident we are about a particular result. He went to say something that really resonated with me. He suggested that there is something that we can say with certainty, which is :
The consensus scientific view is the best we can do at any given time, given the available data and our understanding of it. It is not legitimate and certainly of no scientific value (although there may be political value) to attack a prediction because you don’t like the consequences, or you don’t like the sort of people who are happy with the prediction, or you don’t like the people who made the prediction, or you don’t like the sort of policy responses that prediction might suggest or encourage, or even if you simply see yourself as a challenger of consensus views in the name of some ideal or other.
This probably describes my reason for starting this blog. We have an immense amount of information about climate science, and all of the information is presented with suitable confidence intervals. All of this information is available to be used by our policy makers to decide what we should, or should not, do with regards to climate change. The confidence intervals include the possibility that we will not warm much and, hence, that the impact of climate change will be minimal. However, our understanding at the moment is that this is very unlikely. Similarly, it is possible that warming will be extremely high and the impacts will be severe even if we do reduce our emissions but, again, this is unlikely.
Essentially, the best evidence we have today is the consensus scientific view which represents our best understanding. It could, and will, change as we gather more data and improve our models and theories. It could even up being very wrong, but it’s still the best we have today. All of those who like to point out that consensus views have been wrong in the past, should recognise two very obvious truths : there are also very many consensus views that have turned out to be correct, and – even though some consensus views did turn out to be wrong – those views were still be the best evidence of their day. It makes no sense to argue against a consensus position simply because it could be wrong. That’s an argument for ignoring evidence and basing policy decisions on ideology alone.
Many also try to argue against the consensus position on the basis of it having no significance with respect to science itself. This sounds good, but really doesn’t make much sense. If you’re an ecologist who would like to study the possible impact of climate change on some ecological system, you need to understand the consensus position. You can’t be expected to re-invent the wheel and redo all the climate modelling so as to inform your own research. You use what others have already done. It’s true that the consensus position should not define our understanding indefinitely, in the sense that someone could make a discovery that overturns it, but that doesn’t mean that the consensus position has no relevance whatsoever.
Generally speaking, what Brian Cox seems to be saying is very much what I’ve been trying to say here. The consensus position represents our best understanding today, and includes confidence intervals that tell us something of the likelihood of the different possible outcomes. Sensible policy should be based on this position. If your policy preference requires arguing against this consensus position, then that would suggest that your policy preference is weak if the consensus position turns out to be right. You may turn out to be right, but that would be more by luck than design, and I see no reason why this makes any sense whatsoever. Of course, our understanding will change with time, but that doesn’t mean that the consensus position is not the best evidence we have today.
I may have said this before, but I’d be really interested to see if those who typically argue against mainstream climate science (Andrew Montford being a prime example) can actually construct an argument for their preferred policy option that doesn’t require claiming that there is something fundamentally wrong with the consensus position. I don’t think they can, but it would interesting to see them try. Of course, individuals are welcome to believe whatever they like, but I see no reason why our policy makers should base their decisions on anything other than the best evidence we have today.
I thought I might add a comment about this Telegraph article arguing that Brian Cox is wrong to say that Knowledge should not be controversial. This would seem to illustrate more that the author (Brendan O’Neill) has not understood what Brian Cox was getting at (or is intentionally mis-representing him). To me, what Brian Cox is suggesting is self-evidently true. Knowledge itself should not be controversial. If we know something then that knowledge itself should not be controversial. What we do, given that knowledge, could well be controversial, but the knowledge itself is not.
I assume you understand that as long a a single person disagrees with something, it can be labeled controversial.
Hopefully, Brendan O’Neill can apply his thinking for example, to controversy over nicotine …just recently, one commenter assured us of nicotine::
” Judith Olney
In reply to john r walker
Exactly John, and a drug that is no more dangerous, when used in personal vapourisers, than caffeine in coffee, tea, or soft drinks. A drug which has also been studied extensively, and has been deemed safe for long term use in NRTs, which is the same nicotine that is used in personal vapourisers.
The other ingredients in vape liquid have also been studied extensively, over many decades, and are regarded as safe. ”
Hopefully, Brendan O’ Neill will educate any younger relatives in controversy over nicotine, and smoking , making sure that medical knowledge is always open to challenge and controversy.
==> “All of those who like to point out that consensus views have been wrong in the past,”
Poor form, but I’d like to re-post two comments I made at Climate etc. Please delete them if you think it amounts to spamming.
Seems to me that despite thousands? of posts about the “consensus” in the climate wars blogosphere, and hundreds of thousands? of comments on those posts, a reasonable conversation between non-tribalists about the “consensus” would be relatively uncomplicated. It would go something like the following:
“Realist” (meaning someone who is relatively concerned about the potential for ACO2 to affect the climate negatively)..
“Skeptic” (meaning someone who is relatively unconcerned about the potential for ACO2 to affect the climate negatively).
After getting a few amusing responses from “skeptics” (same as it ever were), I thought I’d post a slightly different version of the “consensus” discussion.
“Realist” (meaning someone who is relatively concerned about the potential for ACO2 to affect the climate negatively)..
“Skeptic” (meaning someone who is relatively unconcerned about the potential for ACO2 to affect the climate negatively).
“Well, yes that is true, but the existence of that overwhelming prevalence of view among published experts does help to inform us non-experts about probabilities involved. Surely, when you seek out medical advice, while you understand that the “consensus” might be wrong, you use the prevalence of informed opinion to help guide you about probabilities yourself on a regular basis. While not infallible, there is a reason why people regularly apply that logic when assessing risk in the face of uncertainty on matters that are highly technical and complicated and “wicked ”
Your former dialogue is what you would hope would happen (the consensus exists, so let’s accept that and move one), the latter is indeed what seems to typically happens (consensus views have been wrong before, groupthink, Lysenkoism, ….).
Of course Lysenkoism wasn’t scientific consensus but a politically applied pseudo-consensus. Given the freedom and the time to do so, Lysenko was easily proven wrong but he had the backing of the Communist leadership in the Soviet Union and prospered as a result. The consensus of scientists in the rest of the world turned out to be correct.
The reason Lysenkoism is often bandied in discussions about climate change is that it is easier to proclaim a conspiracy than it is to deal with the mountain of evidence that current climate change is being forced by human activities.
Indeed. It’s the “this has happened before, therefore it’s happening now” argument.
Science has no formal ways for deciding, what should be considered as well established truth. The only measure of that is the extent of consensus. Formation of consensus is an integral and essential part of the scientific process.
Nothing in the above has been formalized, but all is widely accepted in clear cases.
What remains are questions like:
– How do we know that there’s a consensus and how strong that is?
– When we observe the presence of clear consensus, we may still lack knowledge on it’s precise content.
These points are highly relevant for the climate discussion. Even most skeptics agree on the presence of some consensus and accept that some statements are virtually certainly true, but more or less everyone has a somewhat different view of the limits of well established knowledge. A majority of active climate scientists are likely to include quite a lot in knowledge that’s likely true, but might question whether all that is really well established scientific knowledge.
It seems obvious that Brendan O’Neill has not understood at all, what Brian Cox was talking about or what science is about, but I don’t think that the issues are as easily solvable as Brian Cox seems to be implying.
What’s the field of research that tries to find practical answers for assessment of the significance of consensus in cases, where it’s not straightforward, but an answer is needed? That’s STS.
Well, given that we have the IPCC reports which – from what I’ve seen – are a fair representation of or current understanding of the science, what Brian seems to be suggesting seems quite straightforward.
Maybe, but from what I’ve seen I’m not convinced they’re particularly good at this. Some understanding of basic science would seem to be a prerequisite.
Anders, I would accept the IPCC reports as a fair representation of current understanding. However, how does the IPCC quantification of climate sensitivity (ECS) fit into the Cox model? That quantification is a likely range of 1.5-4.5 C, 5% confidence limit of 1 C (“extremely unlikely to be below”) and a 90% upper limit of 6 C (“very unlikely greater than 6 C”; not 90% upper limit, not upper limit of the 90% range). What is more, the likely range in the 5% lower limit are known with “high confidence”, whereas the 90% upper limit is known with only “medium confidence”.
The problems start in that the oft quoted range (1.5-4.5) is only the likely range, and therefore cannot be reasonably taken as showing the limits in which we “know” the ECS to lie. Indeed, no limits reasonably so interpreted are stated. We do not know the 99% or 95% confidence intervals on ECS
Further, the confidence is a subjective assessment of the robustness of the conclusion – ie, the extent to which either further evidence or new analysis (in the case of currently conflicting evidence) is likely to result in a change in the stated likelihood. Because it is subjective the interpretation cannot be formally quantified, but with high confidence, it seems unlikely that limits will change by more than plus or minus 0.5C (ie, the likely range in a future report given what we know now may lie in the range 1-5 C, with a lower limit not greater than 2 C, and an upper limit not less than 4 C); but at medium confidence the 90% upper limit may vary by plus or minus 1.5 C. Again, the ranges are subjective so my quantifications of possible variations are not official, and are intended to be illustrative without being to far of a reasonable estimate.
We can avoid the problems relating to confidence by stating the likelihoods as being the “best estimates given current evidence” rather than being “what we know”. That, however, is a weaker narrative than that from Cox. If we want a stronger narrative re ECS, we need to turn subjective qualifiers into quantitative estimates and will as a result misrepresent what is actually known about ECS.
While I do think that the WG1 reports have been good, referring to any specific report brings us outside the realm of the nonformalized concept of consensus that’s so fundamental for science.
> Science has no formal ways for deciding, what should be considered as well established truth.
Funerals could be formal if you dress like you mean it.
@ Joshua: plate tectonics is actually a poor example for pseudo-skeptics to cite. In fact the initial hypotheses of continental drift or mobilism by Wegener and other geoscientists received a somewhat favourable reception in Europe and Latin America from geologists who recognized the geological and biological similarities across the Atlantic. It went into stasis for several decades, largely because of a lack of a known physical mechanism and critical hard data, such as magnetic surveys over the ocean basins and measurements of the cm to dm per year relative movements of continents. Wegener’s position as something of an outsider (as a meteorologist, not a geologist) and his premature death in Greenland may also have played a role.
Cambridge University Press recently published a staggeringly exhaustive, ~2400 page four-volume treatise on the history of plate tectonics (The Continental Drift Controversy). Amazingly the entire work is the product of a single author, scientific historian Henry Frankel.
Ironically, Wegener’s ideas received their most hostile reception from American petroleum geologists. Plus ça change…
What those arguing against the consensus — on principle — always seem to forget, is that the consensus view is only the majority position at this moment. If they are right that the current consensus is wrong and as a consequence it changes, then their pet theory would end up as the consensus position… and they could suddenly find themselves undermined by their own arguments. Of course the truth is that they will always prefer the anti-establishment position: that’s why they concoct, and are so prone to believing, conspiracies.
Tom, that’s why I like ESS. The question of timing is important in terms of adaptation measures, but I would prefer it if the IPCC placed emphasis on the dangers of committing to a certain degree of climate change, regardless of whether it takes 50 years or 200 years to reach it. As I’ve noted before, ECS also has the drawback of being notional (due to slow feedbacks kicking in). The flip side of that coin is the even larger problem of known and unknown unknowns kicking in soon enough to render ECS-based projections a dead letter. Recent observations have shown the GCMs to be behind the curve re feedbacks relating to ice sheets, Arctic sea ice, permafrost and vegetation (especially tropical rain forests), meaning that current model projections *must* be too low. AFAICT the IPCC doesn’t want to go there since it’s hard to do so without seeming to be trashing the models, which are greatly needed.
Magma on Wegner: yes, but for a shorter history:
This example gets used often, but people should read Naomi Oreskes’ excellent The Rejection of Continental Drift: Theory and Method in American Earth Science (1999)
Wegener had the right general idea, but no good explanation for the causal mechanism, so there was a long period of competing hypotheses, with no real consensus, and a strong geographic split between schools of thought, which is the main focus of the book.
American earth science mostly rejected Wegener, but there was much more support for Wegener in Europe, including the eminent UK geoscientist Arthur Holmes
Naomi went to Imperial College London (school of mines) as an undergrad, and her Preface describes shock at finding Wegener’s ideas had long gained some acceptance:
“And although I had only just learned of these ideas two years before, my English flatmate could pull out the dog-eared copy of Arthur Holmes’s 1945 textbook she had read in elementary school. … Working as a professional geologist in Australia – I learned – often from mildly indignant colleagues – not only that many Australian geologists knew about and believed in the idea of continental drift in the 1940s and 1950s but also that in several instances they were ridiculed at international meetings or on visits to the United States by rude and ignorant Americans.”
There just was not a consilience of data for decades, so there were a bunch of competing hypotheses, split between 2 main camps, and with different rules (I think).
When the post-WW II data became available, with a clear mechanism, a fight that had gone for decades essentially evaporated, and plate tectonics was widely accepted as a strong consensus.
I remind the sceptics that the consensus was that there was no AGW. Arrhenius went up against the mighty Angstrom in the early 1900s. The evidence won but it took decades.
The pseudo-sceptics are right. Just a century late. And on the wrong side.
That may be a fair point. My personal view would be that the first way in which you describe it is defensible, while the second is too strong. However, I got the impression that Cox was not suggesting that we should make statements that are stronger than we can defend, simply that we should avoid making statements that appear less certain than we are. One could argue that saying “we don’t know, but it’s probably between 1.5 and 4.5 could be perceived as weaker than “our best evidence indicates that it is probably between 1.5 and 4.5.
I’m not entirely sure I’m following you. If we want to communicate with the public and with policy makers we have to attempt to define/present our best understanding. That – as I understand it – is what the IPPC tries to do with its reports. It may not be perfect, but – in this context – perfection may be unachievable. So, yes, there may be some formally defined consensus view that we can never achieve, but my simple point was that – given the IPCC reports – it is possible to get a good sense of our current (or almost current) best understanding.
Steve Bloom, because there will be a rapid absorption of CO2 over the first two centuries after emissions fall to effectively zero, ESS of peak CO2 concentration can be a poor predictor of the final response to emissions. Indeed, because of a rough coincidence of timescales for equilibriation of the radiative imbalance and the draw down of CO2, the TCR of the peak concentration is the best predictor of the temperature response over the next few thousand years. Of course, that assumes that net emissions fall to zero. If emissions remain above a few percent of peak emissions (and certainly if they remain above 10%) CO2 levels will not fall and ESS becomes the best predictor of the future response (ie, in several thousand years). In any event, you cannot talk sensibly about the future unless you are aware of both possibilities, and frame the discussion accordingly.
I note that recent calculations of net volcanic emissions have increased substantially, which means calculations of the rate of draw down of CO2 may also need to rise, favouring the TCR side of the equation.
I don’t quite understand the relevance of this. How does this imply that the rate of draw down of CO2 may rise.
Oh dear. Brendan O’Niell is a professional contrarian from the Spiked stable of intellectual anti-environmentalists pro-business sorts. So that means anything he writes is most likely to be wrong, not always, but often enough not to be worth bothering read it. He has a history of not understanding climate change, or rather, like spiked authors, seeing it through their own non-reality based political lense. http://www.theguardian.com/commentisfree/2007/jan/11/flyinginthefaceofthescie
Yes, I’ve just discovered this.
Anders, CO2 levels stayed essentially constant through the holocene. Therefore net draw down for preindustrial conditions approximately matches volcanic emissions. Until recently the best estimate of volcanic emissions was less than the best independent estimate of the rate of draw down suggesting that either volcanic emissions were higher, or the rate of draw down lower than then estimates. More recent data suggests volcanic emissions higher than the best estimate of the rate of draw down, suggesting either that that estimate is too high, or that the rate of draw down is too low (or both). The carbon cycle model used by David Archer (for example) uses a preindustrial rate of draw down matching earlier estimates of volcanic emissions, and so is too low in any event. The difference is not enough to mean we get rid of the effects of a spike in CO2 emissions in less than several thousands of years, but means you can have a higher net emission rate with stable CO2 (at rates around 5-10% of current emission rates (rough figures).
Okay, that makes sense. Thanks.
guthire: thanks, local knowledge is always useful.
Yes, science has a weird way of formulating things that invites being misunderstood by the public. When talking to the public a term like confidence interval is much closer to what we mean as the word uncertainty. Even if in the scientific literature the latter is used most and they mean the same thing.
Some time ago you had a list with such words, explaining what the scientists means and what the public hears. It is good to be aware of that.
ATTP: “Essentially, the best evidence we have today is the consensus scientific view which represents our best understanding. It could, and will, change as we gather more data and improve our models and theories. It could even up being very wrong, but it’s still the best we have today. All of those who like to point out that consensus views have been wrong in the past, should recognise two very obvious truths : there are also very many consensus views that have turned out to be correct, and – even though some consensus views did turn out to be wrong – those views were still be the best evidence of their day.”
Even if the consensus view turns out be wrong, much of our understanding we gained within that view will still be right. When relativity or quantum mechanics showed classical mechanics wrong, bridges did not suddenly collapse. And when QM is shown wrong, our computers will keep on buzzing.
Thus I feel that the term, “best estimates given current evidence”, is much too weak. That is again a term that suggests much more uncertainty as is actually there.
Pekka Pirilä says: “Formation of consensus is an integral and essential part of the scientific process. Nothing in the above has been formalized, but all is widely accepted in clear cases. What remains are questions like:
– How do we know that there’s a consensus and how strong that is?
– When we observe the presence of clear consensus, we may still lack knowledge on it’s precise content.”
Given that consensus is only used as an heuristic rule, there is not that much use in quantifying it perfects. A ballpark figure is more than sufficient.
Given that the consensus is different for every different question you can think of, you will generally have to guess the number yourself (as a scientist).
I agree and I did wonder about expanding on that. In a sense it’s typically more evolution than revolution. Rarely does something turn out to be “wrong” in some complete sense. Our understanding may change, but that’s mainly because we understand it in more detail than because our earlier understanding has no value.
Yes, there probably is better way to frame this, but I can’t think of one immediately 🙂
Conaensus seems controversial. A suppose a layman would be even more confused if we said we shared a paradigm.
General arguments of philosophical nature are fully valid in their own space, but relating them to real world cases cannot done without introduction of some arbitrariness and subjectivity. What Brian Cox wrote was correct on that general level. It’s relevant also for the real world cases, but real world cases offer basis also for counterarguments.
Concerning IPCC it’s totally reasonable to ask, how much bias the procedures of IPCC introduce. In case of physical sciences such bias is usually small, but surely not totally absent. In the case of the issues of WG2 and WG3 it’s fully possible that the biases are substantial.
On potential biases of WG1 we have seen both views like that presented by Trenberth:
and a great variety of claims in the opposite direction. On WG2 and WG3 some may be suspicious on the chapter, where Richard Tol was a coordinating lead author, while others have their problems elsewhere.
IPCC produces worthwhile reports, but are these reports an unbiased presentation of the real scientific consensus?
To be fair to laypeople, there appear to be a number of people who are supposedly experts in a related field, who also seem rather confused.
But isn’t that all that’s relevant here. He’s referring to – as I understand it – bloggers, journalists, and policy makers. He’s referring to real world situations, not to an idealised, academic scenario.
Possibly not in some idealised sense, but I don’t think anyone credible would suggest that this bias significantly influences what one would conclude based on these reports. Additionally, the reports are written by a selection of experts and so is presumably influenced by their biases, but one could still argue that the report represents the average consensus position of a representative sample of experts. This may differ slightly if the sample of experts were different, or would differ from what one would get if you could objectively analyse the literature, but since that’s probably not possible, it’s not obvious that that is all that relevant.
I guess that there is no way to truly determine an absolutely perfect and entirely objective consensus view. That doesn’t mean – in my view at least – that one can’t determine a view that is a reasonable representation of the consensus position and that would be suitable for use by policy makers (which is really what this is about).
In modern science, consensus on some heavily-studied question = best approximation to reality, with expectation that approximations will improve over time.
1) As noted earlier, during Wegener’s time, there was nothing like a widespread consensus.
2) People often cite H.pylori as a consensus-junking discovery, but I don’t think that’s quite right: there is a class of problems where some early assumptions settled in and weren’t tested very strongly. Another of this sort is the assumption that human influence on climate really started with the Industrial Revolutionh, and that is getting challenged by Ruddiman and co, and I think starting to change.
3) The Earth is flat is actually a good approximation on a small scale. paper maps work OK, although can provide surprises for those on bicycles. Topographic maps are better.
The Earth is a sphere is much better.
The Earth is an oblate spheroid is better yet.
The Earth is an oblate spheroid with regional gravity variations that matter.
4) John Tukey:
“Far better an approximate answer to the right question, which is often vague, than an exact answer to the wrong question, which can always be made precise.”
-The future of data analysis. Annals of Mathematical Statistics 33 (1), (1962), page 13.
[Mod : Sorry, not interested in unverified conspiracy ideation.]
2) People often cite H.pylori as a consensus-junking discovery, but I don’t think that’s quite right: there is a class of problems where some early assumptions settled in and weren’t tested very strongly
Also, that kind of scenario appears common in medicine and nutrition – consider the back-and-forth on cholesterol, trans-fats, eggs, etc. That’s also what stands out in the examples used in articles about the “replicability problem” – they are almost always some combination of medicine and social sciences.
Possibly somewhat off topic, so moderate away if needed, but I’m struck by Judith Curry’s recent series of Twitter comments (both her own and retweets). Just in the last two days she has retweeted or linked to Andrew Montford, Matt Ridley, Brendan O’Neill and even Lawrence Solomon, while transparently mischaracterizing points by ATTP and Dana Nuccitelli (“You are saying that there are no natural sources and sinks [of CO2]?”)
Is she even *pretending* to act as a scientist in the public forum anymore?
She promised a post on Salby. Should be a good test…
@- Science has no formal ways for deciding, what should be considered as well established truth.
But there is a long established informal method. People who hold to ideas contradicted by known science are laughed at. Something like the Heliocentric solar system becomes an established truth(!) when there is no possible evidence that could refute the evidence already surpporting the theory.
@- When relativity or quantum mechanics showed classical mechanics wrong, bridges did not suddenly collapse. And when QM is shown wrong, our computers will keep on buzzing.
I would put my money on relativity being ‘wrong’. Note the problems with defining big G.
afreman: there are always problems with medical / social sciences ..
a) Humans are not electrons, so some kinds of studies are impossible, and replications that work in chemistry or physics … don’t. In climate, there aren’t handy extra Earths as a control group.
b) Some experiments are technically possible ,but will never ever happen.
For example, a definitive study would be to take, say 100,000 12-year-olds, split into 2 matched groups, try to get half of them to start smoking/vaping, and make sure the other half doesn’t, 12-22. Then follow the two groups through the rest of their lives, to see cessation rates for the first group, how many of the second group start using nicotine, comparing disease issues, etc, and with DNA sequencing for all (as there are genetic influences.)
That would be definitive and we’d learn something … but not going to happen.
Seems you have already been proved correct. See attps latest. It’s astonishing!
H Pylori is a really bad example of a paradigm shift from a mistaken scientific consensus.
There was not a definitive medical consensus about the cause of gastric and duodenal ulcers, beyond some hand-waving about stress and diet and then an immune disorder when that became the fashionable explanation of the day. Treatment was for the symptoms and the enable healing, antacids to surgery.
The supposed shift was finding that an extremophile bacteria could live in the stomach and was always present in ulcer cases, which could then be cured by eliminating the bacteria can make it look as if the previous explanation (such as it was) and form of treatment had been overturned.
But the novel detection of H Pylori was as much the result of improved methods of growing and detecting unsuspected bacteria as the novelty of the idea that one could be detected in stomach acid. Once the methodology to detect it became established it is found as a common commensal passenger, probably present in over half of any population over fifty.
Most do not get ulcers.
While the large majority of people with ulcers have H Pylori present it is clear that it is a factor that is necessary but not sufficient as a cause. There may be a genetic susceptibility linked to immune responses, which are in turn linked to stress and diet….
What treating H Pylori does is cure people of their peptic or duodenal ulcers, not because the bacteria is a direct cause but because the disease does not progress to that stage of damage in people without the common commensal intestinal resident.
Causation in biology makes climate science look so simple…..
Izen: thanks for the extra info.
Although for different reasons, H.pylori and Wegener are the two cases I’ve seen msot often by people saying “See consensus can be wrong! Hero X got it right and wasn’t believed, but then science finally changed.”
John, I’m not familiar with all the backstories for the names on this list, but at least a few look like candidates for scientists that fought against the scientific consensus. Zweig (quark theory), Zwicky (Dark Matter), Robert Bakker (fast, warm-blooded dinosaurs) and Chandrasekhar (black holes) are stories that I’m at least somewhat familiar with.
Of course, the big difference between the examples above and climate skeptics is the way the cases were argued.
Einstein did not prove General Relativity by exposing a conspiracy of Newtonian Physicists who were hogging all the grant money. Continental Drift was not accepted because someone Audited Geosyncline theory. H.pylori was not identified by someone declaring that the Stress-Stomach-Acid model was ‘just a model’.
The interesting thing when you actually look at their ‘laundry list’ of examples is that it’s very short (given the scope is ‘every scientific discovery or engineering advance ever’), many of the examples are ancient (i.e. pre-1900), many are basically engineering, and at least two (Deep abiotic oil, and nanobacteria) appear to be false.
Those nanobacteria are fun indeed. The newest research all suggests that it is abiotic material, fitting with the original consensus that bacteria can’t be that small.
The abiotic oil hypothesis has some natural attraction for the WUWT crowd, as similar argumentation is used: the mainstream view is supposedly insufficiently proven, and hence the alternative is at least as good an explanation.
Abiotic methane is another matter. Small amounts of CH4 can be generated within the crust and very shallow mantle by reaction of CO2 and H2O with the abundant mineral olivine as it is transformed to serpentine minerals, iron oxide and other secondary minerals. Longer chain hydrocarbons and economic deposits of petroleum? Unlikely to the point that the scientific community, from those carrying out basic research to those working in the oil & gas industry, completely disregards the hypothesis.