The faulty science, doomism, and flawed conclusions of Deep Adaptation

As members of Extinction Rebellion and other climate movements, we have been overjoyed at the success of our movement in ringing the alarm about climate and ecological breakdown, and in applying pressure to the UK government, as well as other governments worldwide. As members of the science community, we have also found comfort in a movement dedicated to telling a truth that has for decades been obscured by corporate public relations campaigns and misinformation.

Many scientists support Extinction Rebellion or are active members, lending some immediate authority to our message of climate and ecological emergency. The need for peaceful civil disobedience has been explicitly supported by over a thousand scientists. Arrested Extinction Rebellion activists received support during their trials from high-profile scientists acting as expert witnesses. As scientists ourselves, we support our movement’s goal of halting greenhouse gas emissions and biodiversity loss rapidly and equitably, but we also know that doing so successfully requires clarity about what science can and cannot tell us. Such clarity is especially important now. In the past few years we have seen a troubling trend: a few figures in the climate movement using science — or what looks like science — to justify increasingly dire and prophetic, but ultimately unsupported, claims about the future.

The most influential example of such climate doomism is undoubtedly Professor Jem Bendell’s ‘Deep Adaptation’, a self-published 2018 paper which holds that accelerating climate change has guaranteed social collapse within the next few decades. Hundreds of thousands of people have downloaded ‘Deep Adaptation’ and the paper has significantly impacted the ideology and strategy of climate movement organizations like Extinction Rebellion. People have changed their life plans based in large part on this paper’s predictions. It is therefore past time to show that Deep Adaptation is wrong — not least because Bendell’s brand of doomism relies heavily on misinterpreted climate science that undermines the credibility of his claims. In fact, Deep Adaptation consistently cherry-picks data, cites false experts, puts forward logical fallacies, and disregards robust scientific consensus. Bendell defends himself by offering unsupported reasons for activists and the public to distrust mainstream climate science. In all of these regards, Deep Adaptation mimics the practices that deniers of global warming have wielded for decades.

Why is it important to deconstruct Deep Adaptation now, in the midst of a global pandemic? In short, the fatal verdict handed down by Deep Adaptation brings with it a bundle of personal and strategic implications with the potential to cripple us as a movement. The flawed science of Deep Adaptation supports flawed socio-political conclusions. The pandemic makes the divergence between these flawed conclusions and the ones we ought to draw all the more apparent. Where Deep Adaptation implies that scientific understanding can no longer save us from catastrophe, COVID-19 has shown the critical importance of science-based policy. Where Deep Adaptation backs away from questions of equity and distribution in the face of disaster, COVID has shown that (in)justice only becomes more important under such circumstances. The people who have suffered most under the coronavirus will also suffer disproportionately from climate change. Conversely, the same people who oppose climate justice and malign climate science also bear central responsibility for disastrous COVID-19 responses in countries like the US, England and Brazil. The coronavirus pandemic may open a window for policy shifts to begin an equitable transition away from our carbon based economy — in which case we cannot allow a faulty quasi-ideology like Deep Adaptation to mislead us.

To be totally clear, we argue that all of the following are simultaneously true:

1. There is an unprecedented global climate and ecological emergency. If governments do not undertake enormous measures to mitigate climate change, then some form of “societal collapse” is plausible — albeit in varying forms and undoubtedly far worse for the poorest people.

2. Policymakers and society at large are not treating this grave threat with anything approaching sufficient urgency.

3. The climate crisis is dire enough in any case to justify urgent action, including mass sustained nonviolent disruption, to pressure governments to address it swiftly.

4. However, neither social science nor the best available climate science support Deep Adaptation’s core premise: that near-term societal collapse due to climate change is inevitable.

5. This false belief undermines the environmental movement and could lead to harmful political decisions, overwhelming grief, and fading resolve for decisive action.

6. Respecting the distinction between the coming hardships and unstoppable collapse clarifies our agency to minimise future harm by mitigating and adapting to climate change, whilst freeing us from moral and political blinkers.

Deep Adaptation: unfounded doomism

Deep Adaptation is just one prominent case of a stubborn class of doomist narratives. Doomism has always occupied an influential place within the western environmental movement. It was present during the first Earth Day, fifty years ago, in concern over the coming ‘population bomb’. When one instance of doomism becomes discredited or disproven, another appears, generally following a re-examination of the state of environmental degradation. The resulting dire findings are then used to justify a fatalist ideology or response.

This most recent iteration of environmental doomism occurs as a widespread and growing scientific consensus confirms that we are in a planetary crisis. The climate movement has fought hard just to bring this science back into the public narrative. But a movement based on science that is opaque to anyone without relevant expertise will always depend on writers, journalists and academics in interpretive roles. Because of this dependence, the climate movement is also structurally vulnerable to doomist intellectuals who claim that science supports their ideas.

Deep Adaptation exploits this vulnerability by appearing not only rigorous and scientific, but revelatory. The paper itself is framed as an act of academic rebellion: it was first published as a blog post after it was rejected for publication by the 'Sustainability Accounting, Management and Policy Journal', and the author claims it is one of the first academic articles to “conclude that climate-induced societal collapse is now inevitable in the near-term''. His framing paid off. Deep Adaptation has since been downloaded over 450,000 times and prominently featured in Extinction Rebellion media. In fact, Bendell contributed a chapter to the Extinction Rebellion handbook, and has spoken for the movement many other times. His paper received a boost from VICE’s favourable coverage, and from the Financial Times.

Why is it so popular?

Deep Adaptation has clearly resonated with many people, including professional scientists. Why is it so popular? We should separate this into two distinct questions. First, what immediate appeal does Deep Adaptation have to the first-time reader? Second, where does it get its significant and lasting support base?

One immediate appeal is the paper’s blunt language strikes a chord for people anguished at society’s obliviousness to the huge climate and ecological threat. Governments continue to massively subsidise fossil fuels without accompanying pollution controls, and the non-binding national commitments made under the Paris agreement are grossly inadequate even to limit global temperature rise to 2˚C. Once one appreciates the scale of this negligence it is difficult to see roads filled with petrol-burning cars, or new airport expansions, and not feel totally powerless. Deep Adaptation openly discusses the emotional anxiety induced by thinking about such colossal problems: the author writes that they “still make my jaw drop, eyes moisten, and air escape my lungs.” These are familiar feelings for us and many others. Relatedly, Deep Adaptation talks about why many deny climate change’s implications and the necessity of action — a sensitive discussion too-often framed in solely political terms. Part of the paper’s value is its willingness to discuss the current, affective, and emotional impacts of the crisis.

Deep Adaptation also correctly identifies many of the ideological barriers that have stymied environmental protection so far: “The West’s response to environmental issues has been restricted by the dominance of neoliberal economics since the 1970s. That led to hyper-individualist, market fundamentalist, incremental and atomistic approaches.” We and many others agree with the thrust of this argument; permanently solving the ecological crisis requires much more fundamental societal shifts than merely trusting deregulated market forces, corporate social responsibility initiatives, or personal carbon footprinting.

A crucial strength of the Deep Adaptation paper is the general idea that we need to brace for serious impacts from climate change. Already, many countries in the Global South have been suffering these impacts for years, and people in frontline communities in (for example) Louisiana, Puerto Rico and across the western U.S. face greater risks of climate-exacerbated disasters each year. The oceans’ heat uptake lags behind that of the atmosphere, meaning seas are expected to continue to rise past 2100 even in a low emissions scenario. Cutting emissions in line with the 1.5°C target would still lead to around a metre of global sea level rise over the next few hundred years, and cannot rule out the possibility of the West Antarctic Ice Sheet contributing an additional 0.2-1.7m by 2100, exceeding the hundreds of millions of people who will already be at risk from flooding. Therefore, even the most rapid emissions reductions programme must be accompanied by major adaptation efforts, as the effects of the climate crisis will only get worse.

The second question, why Deep Adaptation maintains strong support from a small and committed community, is harder to answer. In private conversations some have speculated that Deep Adaptation provides a kernel of shared grief about which shared identity coalesces. Bendell has encouraged this development and goes as far as to frame Deep Adaptation as something akin to a spiritual movement, while also sheltering online groups like the “Positive Deep Adaptation” Facebook page from dissenting views. (The page discourages any discussion of climate change mitigation.) We trust the good intentions of setting up a compassionate space for people to share feelings of grief and loss, as many of us share similar feelings. Nonetheless, the central premises shared by these Deep Adaptation groups are based fundamentally on faulty science.

Exaggerated tipping points

Supporters of Deep Adaptation often claim that understanding social collapse requires a holistic understanding of physical science, economics, culture, food systems, and so forth. Bendell himself has argued that opinions about the effects of climate change on human society “are not climate science,” and we agree.

Yet at its base, Deep Adaptation rests entirely on a climate-science-based argument contained in the sections titled ‘Our Non-Linear World’ and ‘Looking Ahead’. Bendell writes in a letter to the editor attached to Deep Adaptation that “the summary of science is the core of the paper as everything then flows from the conclusion of that analysis”. The science Bendell purports to summarise revolves around two concepts: so-called “climate tipping points” and “nonlinearity”. These concepts are crucial to understanding both the paper and the sub-movement it has inspired.

What are tipping points? Any complex physical system like the climate can feature self-reinforcing feedback effects. Arctic ice melt is one example: as the earth warms, the highly reflective ice melts, and the darker surface below absorbs more heat, warms faster, and melts more ice. Climate feedback effects normally balance each other so that the climate stays relatively stable. In theory, a particular feedback can grow until the system crosses a key “tipping point,” after which the feedback becomes the dominant factor determining the rate of global warming, overpowering any human intervention. Passing tipping points is a worrying thought and a real potential problem, but it is one that must be considered in context.

The core of Deep Adaptation’s argument depends on two particular feedback loops: Arctic ice melt and methane release from permafrost. Although the discussion of these processes occupies only a small section of the paper, Bendell’s argument that societal collapse is now inevitable, and thus the basis for the Deep Adaptation philosophy, stands or falls depending on whether they are correct. They are not.

Arctic ice claims are overblown

As the atmosphere warms, more Arctic sea ice melts and less refreezes each year. The receding ice reveals more of the ocean and, as darker water reflects less sunlight than white ice, the surface absorbs more incoming energy from the sun. This “ice albedo effect” is a well-established part of climate modelling and, like any other positive feedback, a real cause for concern. But it is not, as Bendell claims, a near-term existential threat. A summary of the relevant research explained by Dr. David Armstrong McKay, a postdoctoral researcher on climate tipping points, shows that the overall warming expected as the result of ice-free summers is about 0.15°C globally, which would be primarily concentrated in the Arctic — a fraction of the goal set by the Paris Agreement of limiting global warming to 2°C.

Compare this summary of multiple studies to Deep Adaptation’s treatment of the same topic:

“One of the most eminent climate scientists in the world, Peter Wadhams, believes an ice-free Arctic will occur one summer in the next few years and that it will likely increase by 50% the warming caused by the CO2 produced by human activity (Wadhams, 2016). In itself, that renders the calculations of the IPCC redundant, along with the targets and proposals of the UNFCCC.”

This passage shows the author’s tendency to dismiss the work of hundreds of scientists based on one person’s estimates. It is, needless to say, a bold move. In this case it is also a bad one, because even a cursory search reveals that the magnitude of Wadhams’ prediction is off, and so is his timeline.

Half of the warming caused by anthropogenic CO2 so far equates to around 0.4˚C, because one-third of the warming experienced so far is due to emissions of other greenhouse gases such as Nitrous Oxide. However, this estimate of the warming from an ice-free Arctic summer is likely two times too large. The summary of the science suggests that this amount of excess warming would only occur if the Arctic became ice-free all year round, and not just in the summer.

In 2016, Wadhams made claims predicting ice-free summers by 2018 that were condemned by multiple scientists as “low credibility”. These other scientists were right: Arctic sea ice continues to deteriorate but not nearly as fast as Wadhams predicted, and it still persists through the summer. This means that Deep Adaptation’s claims on ice are wrong on two counts: 1) an ice-free summer won’t happen as soon as it claims, and 2) that ice-free summer won’t cause nearly as much warming as it claims.

If Wadhams’s position were widely shared by experts, one might excuse Bendell’s reliance on this single source. But other climate scientists have repeatedly cautioned against trusting Wadhams’ predictions. This is a pattern we’ll see again: Deep Adaptation invokes an extreme prediction by an outlier scientist (or even non-scientist), and then seemingly implies that we should trust that prediction because it goes against the consensus. Such unfounded trust is inconsistent with the demands of Extinction Rebellion and the broader environmental movement to “Listen to the science,” and “Tell the Truth.” Indeed, it aligns Deep Adaptation with fringe conspiracy theorists, who seek out single extreme views, rather than reflecting on all available evidence.

Overall, as McKay summarizes: “The first ice-free summer in the Arctic will happen sooner than originally thought and likely sometime in the next few decades, but is hard to predict exactly when because of large natural variability on top of the human-driven warming trend. And while losing the summer sea ice will drive significant regional warming and may increase mid-latitude weather extremes, the global warming boost will be modest (estimated at +~0.15-0.2˚C, of which around half has already happened) and won’t happen in one sudden jump.” The first summer without sea ice is now predicted to occur before 2050 in all emissions scenarios, but this, and its effect on warming, still falls short of Wadham's predictions. This is a far cry from “runaway climate change.”

Methane claims are misleading

The second tipping point often used to justify claims of near-term catastrophe is the release of methane from methane clathrate compounds buried in seabeds and seabed permafrost. The “clathrate gun hypothesis”, originally proposed in a 2003 paper, suggests that warming in the Arctic could release massive quantities of methane within a few years: up to 50 giga-tons, or a twelve-fold increase in atmospheric methane concentration, by one estimate. Methane is about 84 times more potent a greenhouse gas than carbon dioxide over the first twenty years after emission, which means that this release would cause a powerful feedback loop in which rising temperatures cause more methane to escape, and so on.

But multiple more recent reviews of the scientific literature have all come to an opposing conclusion: “it seems unlikely that catastrophic, widespread dissociation of marine clathrates will be triggered by continued climate warming at contemporary rates (0.2˚C per decade) during the twenty-first century.” Much of the methane released from these marine sources never reaches the atmosphere in the first place, in part because microbes in the ocean’s water and sediment digest it before it bubbles out of the ocean. A comprehensive review found little evidence that these sources are contributing to an increase in atmospheric methane. In fact, a scientist at NASA debunked similar claims back in 2013 — but Deep Adaptation has almost single-handedly resurrected them.

Deep Adaptation revisits the methane question through suggestive allusions to real-time methane measurements. It references the “[methane] data published by scientists from the Arctic News website,” claiming that they are “consistent with this added methane coming from our oceans, which could in turn be from methane hydrates”. Arctic News is a blog which the science editor of the fact-checking website Climate Feedback highlighted for containing all sorts of unscientific claims as early as 2014. For example, the blog previously predicted a global temperature rise of a full 20 degrees Celsius by 2040 using a physically unjustified polynomial trendline one could fit in Excel.

Aside from confusing global with local arctic temperature increase (the polar regions experience much greater increases than the globe as a whole), this particular prediction was made by fitting an arbitrary curve to some arctic temperature data. These kinds of predictions are unscientific in the same way that the Trump administration’s recent “cubic curve fit” to COVID-19 cases was — they use no scientific principles to ground the prediction, so they aren’t accurate. (The COVID-19 fit predicted that mid-May would bring a negative number of virus deaths.)

Multiple sources refute that recent increases in methane come from the Arctic, and deny that the world is near a methane tipping point. As the Tipping Points Project explains: although there has been an increase in methane since 2007, and at a faster rate since 2015, this is “mostly from either tropical or sub-tropical sources (e.g. from farming or wetlands), fossil fuels (e.g. natural gas leaks), and/or a slowdown in how quickly methane breaks down in the atmosphere. This modelling also indicates that after 2007 methane has mostly come from outside polar regions, which would rule out Arctic permafrost or methane hydrates as the driver of the recent increase.”

The most recent exhaustive review of this topic was the IPCC Oceans and Cryosphere report, published in September 2019. The IPCC found that humans still control methane-induced warming: “As with total carbon emissions, there is high confidence that mitigation of anthropogenic methane sources could help to dampen the impact of increased methane emissions from the Arctic and boreal regions.” In fact, one of the main papers the Cryosphere report cites concludes that, “Despite large uncertainties associated with future projections of Arctic natural methane emissions, our current best estimates of potential increases in natural emissions remain lower than anthropogenic emissions. In other words, claims of an apocalypse associated solely with Arctic natural methane emission feedbacks are misleading, since they guide attention away from the fact that the direction of atmospheric methane concentrations, and their effect on climate, largely remain the responsibility of anthropogenic GHG emissions.”

Large methane contributions from permafrost are not impossible, but it is a gross misrepresentation of the science to say that it is likely, let alone inevitable, that non-anthropogenic methane emissions have or will become the dominant warming factor regardless of our actions going forwards.

Tipping cascades to the rescue?

Deep Adaptation exaggerates tipping points and promotes false science on arctic sea ice melt and methane emissions to fuel its doomist narrative. We and others have shown that these claims are unfounded, and yet supporters double down on them, searching for evidence to confirm a position which had already been chosen.

Since Deep Adaptation was published, a paper in PNAS introduced the concept of “tipping cascades.” Supporters of Deep Adaptation now regularly use it to support the idea of inevitable runaway heating. “Tipping cascades” refer to multiple tipping points, each one sequentially triggering the next, which could hypothetically take global warming out of human control. We should not ignore the possibility of positive feedback cycles such as this. But other climate scientists have pointed out several crucial caveats to this paper. Firstly, this was what is called a “perspectives” paper – a speculative think piece which made a suggestion, not a detailed numerical study backed up by new modelling or experimental evidence. It might well be that more detailed studies later show that no such cascade is possible. Secondly, the lowest temperature at which this cascade was hypothesised to begin was 2°C higher than pre-industrial times, a mark which we definitely have not yet reached. Further, this was a lower bound, meaning that uncertainty would likely place that limit higher, not lower. It is crucial to note that the timescale over which these changes were hypothesised to occur is centuries to millennia, not the few decades that would be required to support claims of near-term societal collapse.

It is important to clarify a few further aspects of tipping points. First, we must not confuse tipping points that cause further planetary warming with ones which cause isolated damage to a local ecosystem; the latter, while contributing to a crisis of their own, do not contribute significantly to the chance of a climate-driven global collapse. Second, a tipping point is not the same thing as a feedback loop. For one, feedbacks exist which don’t constitute tipping points – because they are too weak, for example, or can only provide a finite amount of additional warming before all the ice is melted. Additionally, some tipping points exist in systems which do not contribute to a global feedback loop. Recognizing the complexity within the broad concept of tipping points helps us to understand the scale and scope of these various threats.

Overall the PNAS paper constitutes a thought-provoking argument: what we do now may matter indefinitely, so we should do all we can to avoid that long-term risk. It does not say that we might already be too late, rather that “decisions occurring over the next decade or two could significantly influence the trajectory of the Earth System for tens to hundreds of thousands of years.” We agree with this argument, but we’ve also shown that it lends no weight to claims of inevitable near-term societal collapse.

Non-linear but not unstoppable

To support the idea that we have already crossed some tipping points, Deep Adaptation leans heavily on the concept of nonlinearity. While “nonlinear” has a strict mathematical definition, Deep Adaptation appears to use the term in the colloquial sense, which describes something whose change is not directly proportional to the input. We are currently receiving a brutal lesson on one kind of nonlinear relationship — exponential growth — from COVID-19.

Fortunately, “nonlinear” does not mean “unstoppable,” even by this casual definition. Coronavirus spread exponentially at first, but in most countries social distancing has now begun to reduce the rate of infection so that, with persistence and luck, the number of new infections will begin to reduce exponentially. The spread and containment of coronavirus has been a nonlinear process throughout, but, as countries like South Korea and New Zealand have shown, it can still be controlled. (Using the strict mathematical definition of nonlinear – differential equations whose coefficients are functions of the variables to be solved for – does not change our fundamental point: non-linear does not immediately imply uncontrollable runaway behaviour.)

The Deep Adaptation paper does not make the distinction between nonlinear and unstoppable. It regularly implies that no nonlinear process can be controlled, and appears to confuse nonlinear change with unstoppable exponential growth. For example, Deep Adaptation claims that observed temperature and sea level rise “are consistent with non-linear changes in our environment that then trigger uncontrollable impacts on human habitat and agriculture”, and that measurements of atmospheric methane “are consistent with a non-linear increase - potentially exponential”.

But exponential growth is only one possible type of nonlinear change. COVID-19’s spread remains nonlinear even while we constrain it below exponential levels. Similarly, the height of a person over time changes nonlinearly, because 40-year olds are not normally twice as tall as 20-year olds. Human growth is nonlinear, but it certainly isn’t exponential. The height example illustrates another key point: extrapolation from recent data without any understanding of the driving process will predict the wrong outcome. Someone with no understanding of human biology might watch the rapid growth of a newborn and conclude that an 80-year old person should be 100 feet tall, but we know better.

Climate scientists like NASA’s Gavin Schmidt have already criticized Bendell for misunderstanding the nature of nonlinearity. Bendell responded by “emphasis[ing] the importance of real time observations” of quantities like methane release from permafrost. He also writes in the original paper that the IPCC incorrectly assumes linear as opposed to nonlinear increases in quantities like sea level rise.

This response is still wrong. Recent history is littered with faulty predictions made based on short-term observations and no mechanistic understanding. These predictions range from claims that global warming had stopped during the “global warming hiatus,” to others arguing that it had entered a runaway cycle, as we encountered in the Arctic sea ice section above. Climate science avoids these pitfalls because it models the underlying (and nonlinear) processes driving climate change by incorporating as much basic physics, chemistry and biology as possible, whilst also matching historical observations. This process can include many feedback effects between temperature changes and natural carbon stocks, so global climate models are fundamentally nonlinear. But they avoid the trap of lending too much importance to short fluctuations in the climate. As a result, a recent review of 15 different historical models used since 1970 found “no evidence that the climate models evaluated in this paper have systematically overestimated or underestimated warming over their projection period.”

Tipping points are still less important than human emissions

The most important points we should take away from this discussion are that 1) the tipping points that Deep Adaptation refers to threaten us far less than it claims; 2) mainstream climate science has predicted warming fairly accurately; 3) Bendell’s criticisms generally misunderstand how mainstream climate science models work; and, most importantly, 4) human activity, rather than climate feedback processes, still comprises the dominant influence on climate change.

For example, as noted above Deep Adaptation mentions the danger posed by releases of methane from Arctic permafrost. But the magnitude of methane emissions from the Arctic is dwarfed by those from human activity: about 2-10 million tons per year from permafrost and hydrates versus around 100 million tons from fossil fuel use. We also have other evidence that recent increases likely originate from increased fossil fuel use, not coming from the arctic. Therefore, the key takeaway is that we are still in control of the majority of warming, both present and future. We have primary agency in the outcome, it is not at all “too late.”

Deep Adaptation contains many more scientific claims which we could question: for instance, it mentions dire predictions about biodiversity loss but cites only one paper, on the extreme end of those predictions, which has been strongly criticised, and we discuss this pattern of referencing below. But Deep Adaptation brings up so many spurious claims that it would be a massive undertaking to thoroughly refute them all. Instead, for now we should focus on the fact that almost all of the climate science claims underlying Deep Adaptation’s predictions of societal collapse are wrong. Such strong predictions require strong evidence, but that evidence just isn’t there. Bendell states that his conclusion about societal collapse “flows from the conclusion of [the scientific] analysis,” but this crucial scientific analysis is wrong. The rest of the paper does not make up for this failing.

Societal collapse is not guaranteed

Despite the above quote from Bendell, the author has rightly stated elsewhere that climate science alone cannot confirm or deny predictions about societal collapse. After all, societies have collapsed in the past for any number of reasons to do with both social practices and the environment. Besides being true, this claim is a good hedge: it leaves Deep Adaptation potentially immune to criticism that focuses only on climate science, because to refute the paper’s claims one also has to refute the proposed scenarios of collapse. Fortunately for all of us, those scenarios are also implausible.

Here is one collapse scenario that Deep Adaptation seems to take seriously: “inevitable methane release from the seafloor leading to a rapid collapse of societies will trigger multiple meltdowns of some of the world’s 400 nuclear power-stations, leading to the extinction of the human race”. A grim picture. How does it hold up against accepted science? We have already seen that large-scale methane release is not inevitable, and does not promise the levels of warming Deep Adaptation seems to expect — a fact which by itself disproves the claim. Yet Bendell also fails to explain why societal collapse might cause any of these plants to spontaneously meltdown. Nuclear plants do plan for “Station Blackout” events in which contact with the outside world is cut off. For example the AP1000 design being built at the Vogtle site in the US includes planning, equipment, and supplies to “provide safety functions for an indefinite time” in the event of station blackout. Newer designs are intended to be “walk away safe”.

Even if these plants did somehow collectively melt down, global nuclear meltdown would not come close to causing human extinction. There are 450 nuclear reactors currently operating. If each simultaneously released a Chernobyl-level amount of radioactive material, the average radiation dose per square mile over the earth would be far less than the dose per square mile which the whole of Ukraine suffered as a result of the actual Chernobyl accident. Clearly all human life was not wiped out in soviet-era Ukraine (the resulting mortality across the country was roughly 4000 people out of 51 million), so whilst this hypothetical scenario would be a profound tragedy, it would not come anywhere near causing human extinction. (In contrast to global nuclear war, which remains an existential threat. Warming will likely exacerbate the risk of nuclear war by inflaming geopolitical tensions – a strong argument against a passive, non-interventionist stance on climate mitigation. However, if the argument is really “climate change makes nuclear war more likely”, then that should be made clear, as disarmament would also be an effective preventative measure.)

In a paper using multiple unreliable sources, this claim about nuclear meltdowns still stands out. The apocalyptic prediction of “chain nuclear meltdowns” comes from Guy McPherson, a retired ecologist who spreads misinformation about climate science in order to package and sell a “near-term human extinction” narrative. In 2008, McPherson predicted the end of civilisation by 2018, and in 2012 he predicted that global warming would kill much of humanity by 2020.

McPherson made one of his most eyebrow-raising claims in 2012: “atmospheric oxygen levels are dropping to levels considered dangerous for humans, especially in cities.” This patently false statement appears to stem from a misrepresentation of both the effects of carbon particulate air pollution, and the famous Keeling Curves of atmospheric CO2 and oxygen concentrations. Oxygen levels are indeed dropping as we combust fossil fuels — but they are dropping at an average rate of about 20 parts per million, or 0.002%, per year, so we would have to wait several hundred years to see any noticeable difference. Air pollution in cities is a serious health issue, but due to high levels of carbon particulates, not low levels of oxygen. McPherson’s atmospheric oxygen claim was made on the same webpage that Deep Adaptation cited as the source for the implausible nuclear meltdown scenario. His claims are so ludicrous that this citation casts immediate doubt on the integrity of the rest of Deep Adaptation’s references. Not unrelatedly, these references also include far fewer peer-reviewed papers than one typically finds in an academic article.

More generally, Deep Adaptation gives strikingly little explanation about how collapse might happen, beyond scattered discredited sources like McPherson, and instead relies on emotionally compelling descriptions of what collapse means for us. Bendell writes that “the evidence before us suggests that we are set for disruptive and uncontrollable levels of climate change, bringing starvation, destruction, migration, disease and war.” Any serious prediction should back up such a statement with a line of reasoning. But this is not a serious prediction — rather, it’s a highly effective emotional appeal to fear, masquerading as science. We see as much when Bendell writes that “when I say starvation, destruction, migration, disease and war, I mean in your own life. With the power down, soon you wouldn’t have water coming out of your tap. You will depend on your neighbours for food and some warmth. You will become malnourished. You won’t know whether to stay or go. You will fear being violently killed before starving to death.” This description of collapse is big on pathos, but low on explanations of how or why such a scenario is inevitable, or even plausible. A close reading of this section of Deep Adaptation reveals no such explanation.

Intentionally or not, Deep Adaptation strikes a skilful balance between attempting to discredit mainstream scientific sources, postulating frightening tipping points, and appealing to our fear of the future and our own justified distrust of the institutions meant to protect us, all to conceal the lack of serious evidence for its own predictions.

Patterns of misinformation

We wrote earlier that the climate movement bears structural vulnerabilities to people who misinterpret climate science for a public audience. Since the vast majority of people in the climate movement don’t have time to review the scientific literature for ourselves, we can at times be duped by narratives like the ones employed in Deep Adaptation. In this way, Deep Adaptation bears striking similarities to writing by climate change deniers. A report by Geoffery Supran and Naomi Oreskes on the #ExxonKnew scandal breaks down some of the different techniques used by fossil-fuel-funded contrarians to make junk science seem convincing. In Exxon’s case the mistakes are deliberate, intended to delay regulations which would hurt their profits, but they needn’t be. The report describes five techniques of science denial, each of which appears in some form in Deep Adaptation, albeit in support of the opposite message:

1) Fake experts: Promoting dissenting non-experts as highly qualified though they have not published any actual climate research and/or received any relevant education. Guy McPherson, who has no expertise in climate science yet periodically claims prophetic predictive powers, nevertheless receives a serious treatment in Deep Adaptation.

2) Logical fallacies: Logically flawed arguments that lead to false conclusions. Deep Adaptation argues that nonlinearities in the climate system lead to accelerating impacts, and that, therefore, society will collapse within the next couple decades. But the therefore in the argument is unsubstantiated — Deep Adaptation simply provides no logical argument to back it up. (‘Societal collapse’ is not even defined anywhere in the text.) Arguing from the premise that you “feel that [collapse] is inevitable,” to the conclusion that it actually is inevitable, is a straightforward logical fallacy.

3) Impossible standards: Demanding unrealistic standards of certainty before acting on the science. This technique can manifest as demanding that the IPCC’s predictions all be near-perfect. Deep Adaptation claims that “The observed phenomena, of actual temperatures and sea levels, are greater than what the climate models over the past decades were predicting for our current time.” While the observed warming has been slightly higher than the mean of the results of the model runs used by the IPCC, it has fallen well within their expected range.

4) Cherry picking: Selectively choosing data that supports a desired conclusion that differs from the conclusion arising from all the available data. As discussed above, Deep Adaptation dismisses the work of the entire IPCC related to methane by citing a single paper about methane clathrates, with no mention of the multiple papers which disagree with its conclusions.

5) Conspiracy theories: Proposing a secret plan among a number of people, generally to implement a nefarious scheme such as conspiring to hide a truth or perpetuate misinformation. Deep Adaptation invites its readers “to consider the value of leaving mainstream views behind,” as if those views are inherently flawed by virtue of being mainstream. The original blog post is titled “The study on collapse they thought you should not read – yet.”

Science endeavours to guard against these dangers through several methods: peer review, logical positivism, consensus, and expertise. The process of peer review helps filter out mistakes, fallacies, and ignorance of existing work. Practicing logical positivism puts the burden of proof on the author making the claim. Consensus asks for agreement across multiple analyses, because the evidence presented in a single paper may be flawed or incomplete. Academic bodies require proof of a certain level of relevant subject expertise. Because of these methods, the IPCC synthesises the work of hundreds of authors, drawing on thousands of papers in order to make its conclusions. It is possible that a piece of scientific work that hasn’t used these safeguards is still correct, but it’s far less likely, especially if it implies that the peer-reviewed science is completely wrong.

Deep Adaptation sidesteps the IPCC and in doing so ends up using each of the techniques above. The paper gives at least two justifications for conducting a separate analysis of the recent scientific literature. First, it claims that “one needs real-time data on the current situation and the trends that it may infer” in order to understand the real implications of recent warming — which requires going beyond the IPCC, since it generally synthesizes science published a few years previously. We have already seen how this claim gets it wrong. In large, chaotic systems like the climate system, erratic behaviour can lead to short-lived trends that merely seem catastrophic when extrapolated further into the future. Real-time data is important in that it provides more information about the climate system and allows us to constantly refine our predictions. It does not demonstrate that recent non-linear changes will continue into the future.

Second, Deep Adaptation argues that the IPCC “has done useful work but has a track record of significantly underestimating the pace of change,” concluding that it is better to turn to individual “eminent climate scientists” whose predictions have been more accurate. We’ve already shown how many of those “eminent climate scientists” have made far worse predictions than the IPCC. But even if individual experts occasionally make better predictions than the IPCC, that does not mean that their methods are better. Because climate science studies a massive, complex, and chaotic system, there is always the chance that a lucky individual will make a prediction that turns out to be closer than those made by models. Most of the lone experts cited in Deep Adaptation turned out not to be so lucky, but even if they had made spot-on predictions at one time or another, that would not give reason to trust them above the entire IPCC.

Deep Adaptation relies on three further claims to evidence the bias towards conservative predictions in “mainstream” climate science. First, it claims that “the information available to environmental professionals about the state of the climate is not as frightening as it could be” due to the tendency towards cautious understatement inherent in the scientific process. Deep Adaptation cites Brysse et al. (2012), who point to a tendency towards “moderation” within the scientific community which may have contributed to the IPCC’s underestimation of sea level rise, ice sheet deterioration, and the threat of methane leakage from thawing permafrost. We agree that there are structural forces for moderation in the science community, but this does not merit the logical leaps that Bendell makes. In other words, just because the IPCC has sometimes given (slightly!) conservative predictions on particular questions, that does not justify disregarding its overwhelming body of evidence and concluding that societal collapse is inevitable. There is also a legitimate critique to be made of the way that the IPCC’s maze of qualifiers and caveats undermines its severe message. However, this is not the same thing as downplaying or underestimating the actual science.

Deep Adaptation also argues that climate scientists, like all people, “avoid voicing certain thoughts when they go against the social norm around them and/or their social identity,” and that this social mechanism prevents sustainability experts from speaking the full truth on climate change. According to Deep Adaptation, they might be even more prone to such psychological blocks than the average person because their investment in achieving status within existing social structures makes them “more naturally inclined to imagine reform of those systems than their upending.” As scientists, we take some small satisfaction in seeing this faulty argument, so often used to portray scientists as alarmists, now used for the opposite end. Yet multiple climate scientists made grim public statements around the release of the IPCC’s Special Report on 1.5°C of Warming (some of which can be found on Extinction Rebellion’s own website), and many participated in mass civil disobedience with Extinction Rebellion. These forceful public actions in support of radical, science-based change show that many climate scientists are more outspoken than Bendell gives credit for.

Lastly, Deep Adaptation claims that none of the major institutions involved in climate research or activism have “an obvious institutional self-interest in articulating the probability or inevitability of social collapse.” This is a compelling argument in that institutions such as the World Wildlife Foundation (WWF), which must demonstrate their impact to donors, might be structurally disincentivized from admitting the full scope of the crisis for fear of seeming ineffective — although we are not aware of conclusive evidence either way, and WWF has made many dire warnings about the precipitous decline of wildlife in the past. While many people no doubt appreciate the call for bolder engagement by academics, it is unclear how institutional incentives prevent academics from making dire predictions. Some of Bendell’s own sources, like Wadhams, gained prominence in exactly this way: by making bold but weakly supported claims about, for example, Arctic ice sheet collapse.

The precautionary principle, that lack of full scientific certainty is not required before forestalling a threat, is often invoked around discussions around societal collapse. But whilst caution and urgent action on climate are certainly justified, responses should still be based on scientifically-plausible scenarios. This is the point at which the expertise of scientists is vital; they are the ones who can tell you whether or not a sudden catastrophic release of methane or human extinction due to a chain of nuclear meltdowns is actually plausible.

Damaging the movement

Deep Adaptation is undeniably an emotionally compelling work, especially when one does not dig too deep into its details and references. Maybe this explains the sense felt throughout some parts of Extinction Rebellion that the paper has become an integral part of our motivation; that without the Deep Adaptation agenda, we would have less reason to take radical actions and make radical demands of our governments.

In fact the opposite is true: besides being based on flawed science, Deep Adaptation actively harms us, our movement, and the people we are fighting for.

Most obviously, if societal collapse were truly inevitable, it would make no sense to practice mass civil disobedience against governments that would shortly fall apart. Instead, we in the climate movement should view societal collapse as a distinct possibility among a number of long-term outcomes — and work as hard as possible to prevent it by transforming our societies. The possibility of averting this disaster in fact justifies a greater extent of nonviolent rebellion than we have accomplished thus far: our governments are failing to protect us even though they can. If they couldn’t, rebellion would be both pointless and unjustified.

A defeatist outlook also removes the agency from future acts of harm. A narrative that destruction is inevitable justifies continued destruction, but ignores the human choices which cause it. For example, if we assume that the Amazon rainforest is doomed to die, we might ignore the horrible injustices that brought it to this point, or be less inclined to support the indigenous and front-line communities still fighting to defend their land. If Native ways of life were truly doomed by natural forces then it would make more sense to support those people in relocating than in fighting for lost land. But that’s not the case. Instead, extractivism remains the larger threat to Indigenous and marginalised people, and catastrophic levels of further extractivism are not a physical inevitability.

Deep Adaptation hurts Extinction Rebellion and the broader climate movement in a number of other ways:

1. It demotivates us

The belief that near-term societal collapse is inevitable takes a serious toll on the mental health of many people. That’s bad in itself, and it’s also the wrong way to bring people into our movement. In fact, a recent study surveying 50,000 people found that “individuals who believe that climate change is unstoppable were less likely to engage in behaviours or support policies to address climate change.” Telling the full truth about the climate emergency is at this point a radical and extremely powerful act; telling the fatalist tale of Deep Adaptation is not. Deep Adaptation encourages a kind of paralysis when it claims that “there is no ‘effective’ response” to the crisis. Again, that’s just wrong: we know that there are many effective responses, and in order to make them happen (and happen fairly) we need democratic political power which can overcome the people and corporations opposed to action on climate change. Building that power is the main goal of the climate movement.

It is worth noting that for some people Deep Adaptation has the opposite effect. Rather than paralyzing, it resolves a painful cognitive dissonance that comes from seeing the yawning gap between our actual response to climate change and the actions we should take. In the face of this gap, Deep Adaptation says “you’re right, everyone else is catastrophically wrong, and you’d best accept the consequences.” This narrative can be a relief, despite the grief it entails, because it brings a simplifying clarity to our decisions. But, again, it’s simply wrong. Regardless of whether or not we subconsciously want to believe Deep Adaptation, actually following Bendell to his logical conclusions entails a split from reality. If it feels right, it feels right in much the same way a cult feels right to its members: by depriving them of information.

2. It delegitimises us

Needless to say, doomist scientific malpractice is especially damaging for any movement whose first demand is to “Tell the Truth.” Using sloppy science hands ammunition to the denialist opposition, allowing them to claim that things aren't nearly as bad as we say and dismiss our demands. By framing arguments like those presented in Deep Adaptation as the sole alleged justification for Extinction Rebellion’s protests, deniers and delayists can dismiss the entire movement as “alarmist.” Of course, most of the opposition will continue to argue in bad faith no matter what — but there is a second reason not to go beyond the bounds of reasonable justification: we lose support. Clumsily attacking the scientific consensus and quoting junk science makes it harder for major figures in science to lend their explicit support, denying Extinction Rebellion a valuable source of legitimacy (and even legal support) for a fight we know is justified. This reasoning applies just as well for the broader climate movement, whose opposition has spent years sowing mistrust in science.

3. It obscures our long-term vision and planning

Believing that the end is nigh undermines the kind of long-term planning that will be crucial to Extinction Rebellion’s continued success. We’ve already seen this effect in practice: one of the small group of Extinction Rebellion protestors who, against the wishes of the wider movement, made the poor strategic decision to disrupt the London underground in Canning Town reportedly referenced the debunked, doom-laden predictions of Guy McPherson as justification. Others have thoroughly critiqued how actions blind to structures of oppression severely damage the movement by setting different struggles against each other, rather than stressing how issues of climate intersect with issues of race, class and gender.

A belief in near-term collapse or extinction engenders a depth of desperation which makes the kind of long-term planning we need in order to live through the climate crisis redundant. Even if the climate movement succeeds in instigating a rapid post-carbon transition, we and many other movements will have much more work to do over the next century to ensure that this transition actually happens, and happens justly. That work will require planning on multi-year or decadal scales, and ensuring that institutions like governments, banks, and universities — not to mention entire industries — contribute to the structural transformation that needs to occur. Simply passing legislation does not guarantee success; neither will mass protests, on their own, bring about the kind of wide-reaching changes to infrastructure and daily practices that we need. And that work will remain important whether or not we manage to stay below 2°C of warming. An unscientific belief in near-term societal collapse undermines this vital, far-sighted dedication.

4. It ignores real aspects of a potential collapse

Further, Deep Adaptation’s vague framing of “collapse” ignores important aspects of severe societal disruption. It’s not a binary; there can be varying degrees of collapse, and different kinds of collapse. At every step, it always makes sense to keep fighting, both to prevent further damages and to recover from past ones. There is no better demonstration of a persistent fight than the ongoing resistance of indigenous peoples to colonialism and environmental destruction. Deep Adaptation takes a different kind of inspiration from indigenous struggles: Bendell references the abilities of Native leaders to come up with new forms of hope as they were confined to reservations — what Deep Adaptation calls “their new lifestyle”. There are myriad problems in this framing of the attempted destruction of Indigenous societies: most prominently, it ignores the genocidal intent of the US government, the wilful and violent land grabs of settlers, and the resulting rampant despair in Indigenous communities. We would do far better to look for guidance on the climate crisis from indigenous resistance to colonial rule and extraction than in strategies for passive resistance. Indigenous communities have been fighting the climate and ecological fight for generations; it's past time that settler communities step up and join that fight.

Secondly, Deep Adaptation ignores the fact that if some form of climate-spurred collapse occurs, it will be imposed by a set of people and institutions who will likely dodge its effects, at least for a time. These are the oil majors, networks of libertarian organizations, and other industries all bent on obstructing both public understanding of climate breakdown and any real social response. Their tactics include sowing doubt about climate science, influencing public opinion against climate policy and tying the reality of climate change to contested partisan ideas. Trade groups like the American Petroleum Institute pour millions of dollars into making fossil fuels seem essential for American greatness. Together with think tanks like the Heritage Foundation and the Global Warming Policy Foundation, these organizations mobilized to defeat and weaken international climate agreements like the Kyoto Protocol and Paris Agreement. Academics and journalists have documented years of activity by this “climate change countermovement,” and it continues to fight to prolong the use of fossil fuels today. Discussing the failures of the climate movement without a serious focus on this countermovement is like blaming an army’s soldiers for losing a battle in which it was outnumbered ten-to-one.

5. It is incompatible with environmental and social justice

Gross inequality and injustice are fundamental to climate change. The vast majority of emissions have bolstered the livelihoods of a small number of middle-class and, most of all, extremely wealthy Westerners. Meanwhile the majority of impacts will hit places like African nations or India, which emits at most a seventh as much CO2 per person as the US. Considerations of racial, social, and environmental justice therefore need to be a central part of any response, because not discussing them is still taking a stance.

Extinction Rebellion has been criticised many times for a weak position on environmental justice. No such consideration is found in Deep Adaptation. (None of the words “justice”, “equality”, “racial”, or “colonial” appear once in 13,181 words of text.) The audience for Deep Adaptation is clearly Western, with collapse described as “a situation where the publishers of this journal would no longer exist”. This means there are only two ways to interpret the unwritten equity implications of Deep Adaptation’s framing of collapse, both of which directly conflict with a justice-centric response.

If it is understood that collapse in Western nations would only come alongside much worse outcomes in developing ones, then the passive response advocated by Deep Adaptation is tantamount to lifeboat ethics. Work on adaptation and resilience are extremely important, but they are most important in the places which will be hit hardest. Bendell’s narrative of inevitable collapse conveniently ignores the debt owed by wealthy nations to exploited ones, regardless.

If instead we are meant to understand that no-one is safe from the spectre of societal collapse, and that we are therefore all doomed together and equally, then this is the polar opposite of a justice-oriented approach. Aside from being completely inconsistent with climate science, this framing is harmful in the same way that “All Lives Matter” as a response to “Black Lives Matter” is harmful. It wilfully ignores both the fact that the crisis will disproportionately impact marginalised groups and the continuing role of racially-charged colonial-era power structures in a system which has failed to address that crisis.

6. It distracts from what is most important

The narrative that collapse is now inevitable also distracts readers from the most important responses to the climate crisis. We need to remove socio-political power from the Exxons, Kochs, and Trumps who are actively exacerbating the climate crisis and work towards defining and enacting systemic changes throughout the rest of society. If we psychologically abandon all the complex institutions and structures we live in and rely on (as Deep Adaptation advocates) then we will not see reasons to repurpose them. But that’s exactly what we need to do.

Moving past post-scientism

So how can we in the climate movement do better?

We should publicly disavow the message that near-term collapse is inevitable, or that climate-induced total human extinction is plausible. There is uncertainty, but not so much that one can claim anything will happen. Stepping back from previous unsupported statements would help show that although the movement is broad, they are still fundamentally motivated by the truth.

We should be much stricter and more careful with science messaging. Major spokespeople should be citing science appropriately, and defer to expertise on the physical science and direct impacts. When slip-ups are inevitably made, statements should be corrected based on scientific feedback instead of doubling down. Representatives should understand the difference between junk science and the valid reasons to be worried about spiralling long-term societal impacts.

Extinction Rebellion, 350.org, and other groups now have an international network of sympathetic scientists and other experts to call upon. The movement is in a unique position to confront political and cultural incumbents across the UK with the emergency message, delivered by those with the necessary understanding and authority. We must embrace this network, and use it.

In the longer term Extinction Rebellion could organise public debates on the science of the emergency. Inviting discussion by authority figures in science would show that the movement is open to criticism and adaptation, while still shifting the window of public debate towards the necessary urgency. It would also help highlight the mainstream media’s failure so far to communicate the scale of the crisis.

We are right to sound the alarm, and to maximise our impact we need to ensure that our message is grounded in sound science. Greta Thunberg exhorts us to “Unite behind the science”, because the truth is bad enough.

Our choice is not new

There is a Greek myth about Persephone, the vegetation goddess abducted to the Underworld by Hades, god of Death. Fearing that he would lose his stolen bride, the god planned to bind Persephone to the Underworld by forcing her to eat its food.

Persephone fasted until her hunger compelled her to eat. When Persephone finally begged her captor for food, he presented her with three choices. On the right he placed the ambrosia of the underworld, a sweet food laced with melancholy that leads those who eat it to resent the living and embrace death. On the left he placed a vessel filled with nectar from his own garden, which dulls painful memories of the world above. In the middle he placed a bitter pomegranate, whose seeds awaken one to the horror of death and provoke a restless longing for life.

Hades knew that Persephone would find the ambrosia and nectar all but irresistible. Once she ate either her fate would be sealed. But Persephone chose to eat the pomegranate, accepting the accompanying anguish in exchange for the chance to return home.

The environmental movement faces a similar choice. On the one hand we feel justifiably alienated from a society which appears damned by its own excess. Some of us abandon it as irredeemable. On the other hand, seeing the enormous scale of the obstacles we face, our world constricts. Forgetting the possibility of revolutionary change, we content ourselves with small concessions and political manoeuvring.

Hades’ nectar is the post-political environmentalism of ineffectual carbon markets and greenwashing: a seductive barbiturate. Ending the fossil fuel industry’s dominance could by some measures require the largest expropriation of private wealth in history; facing odds like that it can seem foolish, if not disastrous, to gamble on the chance of revolutionary change. When we drink the nectar our lives become simpler, our ends narrower.

Deep Adaptation is ambrosia, and like all ambrosia the kernel of truth within Deep Adaptation only makes it more enticing. Things are far worse than newspapers, cable news, politicians, and so forth commonly let on. We face a more awful reality with every round of failed COP negotiations. There are real reasons for despair.

Yet despite all the talk in Deep Adaptation of accepting the harsh reality of social collapse, the idea has a sickly-sweet attraction: believing that social collapse is inevitable provides a level of certitude and clarity which makes many choices simpler. But we must remember that the choice to believe in an inevitable collapse is itself a luxury, a form of escapism only available to those with the time and resources to plan for its consequences.

Choosing the pomegranate means accepting the truth with all the pain it brings. The truth is that most of the world has been systematically lied to, betrayed, and exploited to prop up a failing system of fossil-fuelled growth which will not go away easily. We know who is to blame. We know that we need to reform the government from the city to the international level to meet the demands of global and intergenerational justice.

This knowledge is far from comfortable because it requires real work: not just nonviolent disobedience, but educating and training others, and gaining political power. As an argument and a philosophy Deep Adaptation ultimately requires none of these things. By rejecting Deep Adaptation we recommit ourselves to understanding how we can live well within planetary limits, imparting that knowledge to others, and excising the rot and paralysis from our politics.

The authors would like to thank Professor Julia Steinberger, Megan Ruttan Walker, Dr Chris Wymant and Dr Alison Green for input, and to Dr David McKay, Professor Richard Betts, Professor Rich Pancost, Dr Scott Archer-Nicholls, Dr James Dyke and many other scientists for checking the claims in this article.

For more detail on what climate tipping points do and do not imply for the planet, the science outreach project climatetippingpoints.info is an excellent resource. It is run by two postdoctoral researchers in climate science based at Stockholm University, and was seed-funded by a project of the Engineering Physical Science Research Council - the body which funds most physical science research in the UK. It gives detailed descriptions of the science of methane clathrates and arctic albedo, including fact-checking of incorrect claims made in Deep Adaptation and elsewhere.