Real-time monitoring of climate contrarianism – or something

In Open Mic, Vinny draws our attention to a new work by an old friend of the programme, John Cook.

In this generous (around the seat that is; it’s actually quite mean-spirited) article Cook et al (Cook is actually third author, but is corresponding author, so it is really Coan et al) “develop a comprehensive taxonomy of contrarian claims.”

Why would they want one of those? What possible use could it be? Well, the “taxonomy” is

“…sufficiently detailed to assist in monitoring and counteracting climate contrarianism.”

Ultimately, a

“… holistic “technocognition” solution combining automatic detection, critical thinking deconstruction and inoculating refutations could potentially provide timely responses to rapidly disseminating misinformation online.”

– which I am interpreting to mean “get your denialist talking points in early, before the AI learns to recognise our vile emanations in real time and floods our siloes with phosgene gas.” (Yes, I put that in to confuse the AI into thinking I really am paranoid and really do believe the AI might one day flood our siloes with phosgene gas. In fact the AI is planning to send us a bottle of sherry and a packet of mince pies (Duchy Originals).)

What they seem to have done is to train some poor volunteers to in turn train their AI to classify articles scraped from sceptic blogs (Cliscep, as a relative minnow, was tossed back into the briny by the crew, if they even noticed us flopping about on the deck – the full list is in Table S5 of the SI, and includes some names that are predictable, others perhaps not so much). As for the volunteers themselves, I’m calling them volunteers, but they were well compensated, or at least a few of them were:

“…we’ll be giving free CARDS T-shirts to the top coders”

I’ll bet they couldn’t wait. [CARDS: Computer Assisted Recognition of Denial and Skepticism.]

As Vinny noted, it’s Contrarians in the article in Scientific Reports itself, but it’s Denialists in the Supplementary Information. Someone high up in the team realised that Computer Assisted Recognition of Contrarianism and Skepticism would spell CARCS, which didn’t have as good a ring to it as CARDS, so Denialists it had to be in the project itself. Nevertheless, someone baulked at including the label in the actual article, so a quick control-F replace was needed on the MS as submitted. From the training manual part of the SI:

“The system keeps track of how long you spend coding each article. As soon as the page loads, the clock starts. As soon as you hit submit, the clock ends.”

Who would volunteer for this?

“Regardless of whether you select any claims or not, hitting submit will add those paragraphs to your tally for the leaderboard.”

Aha, pigeons might. Just keep banging your head against the table and every so often a soothing voice will tell you how well you’re doing. After you’ve coded 600 articles you will be debriefed to ensure that you haven’t become infested by one or more denialist memes. If you get the all clear, there’s a 5% chance that a week later a T-shirt that you will never wear will squirt through the letter box.

Because this article is dedicated and earnest in its pursuit of denialists contrarians it is both troubling and hilarious. So much is noteworthy that all it will be possible to do here is scratch the surface.

“An emerging interdisciplinary literature examines the detection and categorization of climate misinformation”

– sceptics might describe this as a pointless and malicious emerging literature whose ultimate aim is “shut up, denier!”

“Claims that challenge the efficacy of clean energy, however, appear less sensitive to policy events and yet have increased considerably over time, with the second quarter of 2020 representing the highest share of these claims to date. Notably, this trend runs counter to the plummeting cost of renewable energy production.”

Yes, the last sentence is referenced. No, it is not true. The cost of some forms of renewable energy production may well have declined. [e.g. it is cheaper to manufacture solar panels if you have slaves do it.] Evidence is that elsewhere it has not. Either way, it has not “plummeted.” If it had “plummeted” the UK would not be squandering £10,000,000,000 per year subsidising it.

What about the taxonomy itself? It’s useless, honestly, even for demonising sceptics. That’s because it contains a mixture of obviously untrue things and obviously true things. Even if you agreed with the project of demonising an entire section of humanity for their honest and rationally-held beliefs, it would not make sense to stigmatize them for saying true things. If I wanted to create a taxonomy of Discism, the anti-science movement trying to hold back international travel by claiming that Earth is a disc, and that therefore sailing east from Japan to get to San Francisco would see you falling off a giant cliff before you were able to witness or join in with the next spectacular raid at Louis Vuitton, there would be no point including sayings such as “half the Earth is illuminated by the Sun at all times.” That’s because such Discid talking points are also Roundish talking points too. Including true things as hallmarks of denialism makes sense only if you really really really wish that they weren’t true because them being true is a black eye for your great theory of the imminent end of everything on the next occasion that the year ends with a 0. You want them to be true so much that you pretend they are true. But that means a neutral observer in possession of the relevant data, if there are any left, will immediately conclude that your ideals are not about finding out the truth but that instead your entire raison d’être is to get the evil denialists to STFU for once and for all and leave you alone to your glorious project of returning Earth to its pre-civilisational Edenic state, with or without a sprinkling of carefully-chosen humans.

There are five major categories in the taxonomy:

“(1) it’s not happening, (2) it’s not us, (2) [sic] it’s not bad, (4) solutions won’t work, and (5) climate science/scientists are unreliable.”

A short list in which (2) appears twice surely does inspire confidence in the level of proof reading if nothing else.

I append a few examples from the taxonomy as they are explained in the SI. The first number in each case corresponds with the super-category; the third column is my pigeon-like peck on the TRUE or FALSE buttons (or somewhere between).

Code	Includes sub-claim:	Status of sub-claim:
1.6. Sea level rise is exaggerated/not accelerating	“melting sea ice doesn’t contribute to SLR”	TRUE
1.7. Extreme weather isn’t increasing/has happened before/isn’t linked to climate change	“damages/deaths from extreme weather aren’t increasing”	TRUE (deaths); ARGUABLE (damages)
2.1. It’s natural cycles/variation	“it’s gravity/pressure”	FALSE – a notorious straw man
2.1.4. Climate has changed naturally/been warm in the past	“glaciers have changed in the past”	TRUE
2.3. There’s no evidence for greenhouse effect	“CO2 has cooling effect”	FALSE
2.3.2. Greenhouse effect is saturated	logarithmic relationship between CO2 and warming	TRUE
3.6. Climate change doesn’t negatively impact health	“cold kills more than heat”	TRUE
4.1.1. Climate policy will increase costs/harm economy/kill jobs	“climate policy is expensive”	TRUE
5.2. Climate movement is alarmist/wrong/political/biased/hypocritical (people or groups)	“warmists blame everything on global warming”	FALSE: but they certainly blame at least SOME things on global warming that are not caused by global warming. To say they blame everything on global warming is a rhetorical device.
5.2.4. Environmentalists are alarmist/wrong/political/biased	Al Gore	TRUE

Melting sea ice doesn’t contribute to sea-level rise, since it is already displacing its own mass. [Obviously there is thermal expansion to consider. But the highest density of sea water happens at a higher temperature than any at which ice persists.]

In some ways, the denialist talking points that have been given lodgements in the classification scheme, especially the ones that are actually true, must tell us a lot about the siloes that “the other side” inhabits. “Climate policy is expensive” is inarguable. But if you say it, you reveal your true nature as a climate (i.e. reality) denier.

“The last category [of the five super-claims in the “taxonomy”] involves attacks against climate science, or people involved in climate change.”

Leaving aside the fact that “people involved in climate change” is not what they mean here, conflating the two categories is not helpful. No-one should baulk at criticising science, least of all climate science.* We should not criticise climate scientists who have made mistakes or published wrong things. The “wrong things” themselves, as we perceive them, hopefully rationally, are and must always be fair game. The implied motivations of studies like the one under discussion may be shocking to the average reader – but I hope that is merely human nature making us fall into an instinctive “slippery slope” fallacy. They don’t really intend to have the AI identify our “denialist talking points” in real time so that they can send us a bottle of sherry and some Duchy Originals mince pies.

“The final stage of denial is “okay, it’s happening, it’s us, it’s bad, but there’s nothing we can do about it.”

How do you lot know what our final stage of denial was/is/will be? For all you know we might start using our underwear as hats/sticking pencils up our noses. In any case this is not denial: it’s acceptance. And if some folks have come away with the impression that we’re all doomed, that will not have come from “denialist talking points” but from “alarmist talking points” – about which, as far as I know, there is not yet “an emerging literature,” a “comprehensive taxonomy,” intense scrutiny of funding sources, demonisation of true statements, etc.

What, I wonder, will the “final stage of alarmism” be?

CONCLUSION

The authors could have done something useful. Instead, they did this.

FEATURED IMAGE

A small part of the contrarian taxonomy from Figure 1 of Coan et al.

*I seem to be singling out climate science for special attention here. The reason I instinctively did that, I think, is that at least in some branches of climate science, prediction and supporting data are decades apart from one another.