Spilt Ink
On Conspiracies, Rationality and Politics
A recent online exchange on how best to respond to conspiracy theories raised some interesting questions about the nature of rationality. The debate was prompted by a series of tweets from Chris Kavanagh, an anthropologist and co-host of the podcast Decoding the Gurus. He expressed his frustration at Scott Alexander, a psychologist and leading online “rationalist”, who had just published his second lengthy assessment of the claim that the drug ivermectin was effective in curing Covid. This claim had been made by a handful of doctors, against the mainstream medical advice, and promoted by online personalities and anti-vaxxers. Despite the fact that Alexander ended up agreeing with the mainstream, Kavanagh complained that Alexander’s intervention had lent credibility to a controversy which had been completely manufactured by conspiracy theorists:
This 10,000-word article by Scott Alexander following up on his previous 15,000 one on Ivermectin is a good illustration for me of the limitations of the rationalists. So much ink spilt to arrive at a conclusion relevant experts reached long ago & still seemingly little appreciation that researchers are not capable of replicating what conspiracy theorists & anti-vaxxers generate because their success relies on misrepresentation & zealotry…. It’s indulgent & potentially misleading as it leans into the framing of this being a legitimate area of scientific controversy when it is much more akin to debating with 9/11 truthers. If studies had supported Ivermectin as an effective treatment it would have been adopted by medical and public health authorities. That willingness to follow the evidence does not exist on the pro-Ivermectin side. It should not take 25,000 words and a personal meta-analysis to recognise this.
Alexander in his reply couldn’t see a problem with rationally discussing the conspiracy theorists’ claims. Indeed he thinks the real danger lies in ignoring or ridiculing them, as that approach simply reinforces their commitment to the conspiracy theory. Kavanagh then responded with his own essay.
On my reading, Kavanagh’s main criticism of Alexander’s approach is that it is not possible to have a rational discussion with conspiracy theorists because they lack “a willingness to follow the evidence.” Alexander addresses this point directly, arguing that conspiracy theories are rational. Prefacing his argument with a biographical note, Alexander claims that he “learned about rationality” from a conspiracy theory, thanks to a youthful interest in Graham Hancock’s Fingerprints of the Gods (1995). In that book, Hancock argues that Atlantis was an advanced civilization which existed thousands of years before the Bronze Age. Although Alexander doesn’t explain why he thinks this is a conspiracy theory, one can identify certain traits associated with the colloquial sense of the term: Hancock advances outlandish claims about prehistory which contradicts the expert consensus; he employs convoluted arguments and ambiguous evidence; and he maintains the truth of his claims to this day, against all the criticisms and evidence marshalled against it.
Alexander recounts that by taking Hancock’s theory seriously, and undertaking his own critical analysis (he didn’t think Hancock’s critics at the time were of much help), he learned how to evaluate evidence; especially with respect to coincidences, anomalies and the narratives used to explain them. He even taught himself geology and explored underwater sites thought to contain sunken pyramids. Eventually he concluded that Hancock was wrong, but this wasn’t because his arguments were irrational. After some investigation, Alexander decided the same could be said for conspiracy theories in general:
[M]y conclusion is that conspiracy ecosystems fall prey to the exact same biases that all of us have, including experts and correct people. But experts and correct people have slightly less of them, have better self-correction mechanisms, and manage to converge on truth, whereas conspiracy theorists have slightly more of them and shoot off into falsehood…. Conspiracy ecosystems’ don’t have some exotic reasoning style we can’t possibly understand. They have normal reasoning, and are just slightly worse at applying CONSTANT VIGILANCE than everyone else. [The link is to an essay where Alexander argues that a fundamental skill of rationality is being able to notice when you are confused about something.]
According to Alexander, there are two explanations for why people persist in believing false theories despite the evidence. One is that such people are are incapable of thinking rationally about some topic because of intractable cognitive biases, such as “trapped priors” (which is when one’s past experience overrides the capacity to learn from new experience). He suspects that extreme instances of such cognitive issues may be genetic. The other is that conspiracy theories are often presumed to be irrationally motivated by critics, e.g. when Hancock’s critics dismissed him as racist and so not worth taking seriously. When those who entertain a conspiracy theory are ridiculed or ignored, and a reasonable explanation for why they are wrong is not forthcoming, then they tend to double down on their commitment to the theory and form a poor opinion of the so-called experts. It is in order to avoid this scenario that Alexander thinks it is important to start with the assumption that everyone is rational, and address their fringe beliefs in a way consistent with this assumption. Anything else amounts to treating the mainstream view as beyond criticism, an attitude he believes is akin to fideism (the idea that the sincerity of one’s religious faith is undermined by subjecting it to rational scrutiny).
Kavanagh argues in his response that Alexander’s assumption of rationality is undermined by both the cognitive biases which make people susceptible to conspiracy theories (in particular, he thinks it stems from overestimating one’s own competence when it comes to evaluating evidence), and the pseudo-rationality which characterizes what he calls “conspiracy ecosystems.” This pseudo-rationality includes the use of emotive anecdotes, exaggeration and misrepresentation. Kavanagh argues that rather than trying to evaluate every controversial opinion on your own, it is better for most people to become aware of their limitations; take note of the relevant context, such as the possible influence of conspiracy ecosystems; and learn to identify reliable expert sources and what constitutes a consensus in a field.
Now we could just leave it at that, noting that both sides make some good points, and either agree with Kavanagh that one ought not to help spread conspiracy theories by engaging with them, or accept Alexander’s point that only by engaging with them can we hope to loosen their grip. But there is a feature of conspiracy theories which they both pass over, namely the conspiracy. Conspiracy theories get their name because they fundamentally distrust the evidence which is used to justify the consensus opinion they oppose. It is not just that they think such evidence is wrong, and they have a better theory which explains the facts. Rather they reject the facts, and think that the mainstream view is based on lies, designed to mislead and cover up the truth.
This seems to tilt the argument in Kavanagh’s favor, because it confirms that it is pointless trying to argue with conspiracy theorists, since they will not trust the evidence which contradicts their position. But does that make them irrational? I think one has to be careful here. Governments, corporations and other institutions lie all the time, about all sorts of things. Sometimes they get scientists and other experts to lie on their behalf. For example, the justifications provided by the U.S. and the U.K. for the 2003 invasion of Iraq. A number of people questioned the evidence at the time. Were these people conspiracy theorists? In a sense they were, since they believed that the consensus view of the experts (at least as publicly presented) was not only wrong, but intended to deceive. But these critics weren’t being irrational. They were proven right. It was a conspiracy.
I’m not suggesting by any stretch that most, or indeed many, conspiracy theories will be proven right. However, I don’t think that being skeptical of official sources or experts is inherently irrational. Not even Kavanagh is suggesting that. He notes in his reply that mainstream research can suffer from “very real problems,” and that there exist cases where the medical consensus has been proven wrong. So it is certainly possible that rational people may reject the facts that are widely accepted as the basis for justifying the mainstream position. Is consensus even possible in this case?
Alexander’s assumption of rationality depends on the possibility of reaching a consensus, or “to converge on truth” as he puts it. That’s why he thought it was important to engage conspiracy theorists in rational discussion, so as to move them closer to the consensus view. Only when the other side proves incapable of consensus did it make sense to withdraw the assumption of rationality. But conspiracy theories seem to challenge this link between rationality and consensus. If we are prepared to assume rationality on the part of those whom we cannot reach an agreement with, what then?
The Fools and the Wise
Someone who attempts to answer this question is the French philosopher Jacques Rancière. In a short essay titled “The Fools and the Wise”, written shortly after the January 6 riots at the U.S. Capitol in 2021, Rancière begins with the premise that those who believe in the conspiracy theory that motivated the riot, namely that the presidential election was rigged, were not being irrational. Indeed, the fact that the rioters did not believe the official version of events, or even the most obvious or simplest explanation, is a sign of their intelligence. This is not to say that Rancière believes that the election was rigged. Instead, like Alexander, he wants to draw attention to the underlying rationality of conspiracy theories:
[I]f people reject the obvious, it is not because they are stupid, it is to show they are intelligent. And intelligence, as is well known, consists in being wary of facts and questioning the purpose of the enormous mass of information unleashed on us every day. To which the answer is, quite naturally, that it is to deceive people, because what is displayed in plain sight is generally there to cover up the truth, which we need to be able to discover — hidden under the false appearance of the facts presented.
Here Rancière highlights the role that skepticism plays, not just in conspiracy theories but in the exercise of intelligence or rationality. And this skepticism is tied to another feature of rationality, the ability to draw connections between facts and events, and by doing so establish an overall structure or logic which reveals the truth behind appearances. This can work in reverse as well, by denying facts because they don’t integrate with the whole. In this way, Rancière establishes the rationality of even the most absurd conspiracy theories, such as the argument made by some anti-Semitic Marxists that the Holocaust couldn’t have happened because it doesn’t fit with their historical model of capitalism. (I don’t know who Rancière is referring to here. Roger Garaudy perhaps?) He also mentions those who think that Covid is a government conspiracy to control the population: presumably the logic here involves positing a global regime of such totalitarian control that nothing like a pandemic could happen without its permission.
Of course, being rational doesn’t mean that one falls for every ridiculous conspiracy theory. Rather the point is that conspiracy theories, as Alexander puts it in his reply to Kavanagh, consists of “normal reasoning.” And when Rancière points out that such reasoning involves integrating certain facts into some overall conceptual structure or model, while excluding other facts, he is effectively talking about Alexander’s anomalies, contradictions and the narratives used to explain them.
Assuming that Trump supporters are rational then, why do they believe Trump and his outrageous lies? Rancière proceeds to reject some of the standard explanations. He doesn’t think it is because they are brainwashed by the right-wing media or because they are ignorant of the facts: “we live in a world where means of information, means of verifying information, and commentaries that ‘decipher’ all information, are abundantly available to all.” More significantly, he rejects the idea that all of Trump’s supporters believe Trump in the sense that they take what he says to be true. After all, seventy-five million people voted for Trump. It is more likely that many of them believe Trump simply because they want to, because they derive a certain pleasure from seeing the world as depicted in his rhetoric. It is not truth but sentiment (in the sense of both feeling and opinion) which unite his supporters. This is what Rancière calls “la communauté sensible” (loosely translated here as “community of sentiment”).
Evidence for this thesis was published just the other day. Internal communications at Fox News, made public as part of a lawsuit, revealed that Fox personnel did not believe in the veracity of Trump’s claim of a rigged election. Nevertheless, they understood that for their audience, the need to affirm Trump’s sentiments regarding the election was more important than establishing the truth. In order to keep their audience happy and loyal to the network, they had to talk about these false claims as if they were true.
Both Alexander and Kavanagh would probably describe the influence of sentiment in this respect as “motivated reasoning”. For them it represents a failure of rationality, or at least a lapse in the “constant vigilance” which Alexander thinks rational people must exercise in order to “converge on truth”. But Rancière is making the point that rationality does not always have to be in the service of truth; it can also be used to construct fictions. And “fiction” here does not necessarily mean the opposite of truth, something false, but rather a way of constructing a sense of reality. A fiction is an imagined configuration of events, people, things and feelings. It is as much a part of the social sciences as it is novels; employed by politicians, journalists, and economists to name a few.
If sentiment is not opposed to rationality, but instead what motivates the use of rationality to construct a sense of reality, what can we say about the sentiments which motivate Trump’s supporters? Again Rancière rejects the conventional wisdom. It is not that Trump gives a ‘populist’ expression to a segment of the population’s anger or resentment. Rather, Trump is seen to embody their collective will in a way which is consistent with the interpretation of democracy as representative. This sentiment is reinforced by a personalized forms of address between the leader and his followers made possible by social media. In addition, Trump appeals to his supporter’s sense of superiority; playing “not on a feeling of inequality to be repaired but on a feeling of privilege to be maintained against all those who would want to attack it.” This sentiment is one around which powerful emotions of hate are generated. It is a “passion for inequality,” one which is not limited to the rich enjoying their superiority over the poor, but is something anyone can partake in because it is always possible to find someone else whom you consider your inferior.
Together these sentiments form the basis of a political fiction in which it is rational to deny that Trump lost in a fair election. First, because Trump’s loss frustrates the democratic right of his followers to be represented, and so the election cannot be a legitimate exercise of democracy. Second, because Trump’s loss signifies an attack on the privileged position of his supporters by those deemed to be inferior, a situation which again cannot be allowed stand. Rancière notes that the appearance of both the original flag of the thirteen states and the confederate flag on January 6 seems to symbolize these two sentiments.
Rancière arrives at this analysis, which goes against the conventional wisdom of treating the January 6 rioters as crazy, ignorant and anti-democratic, by assuming the rationality of Trump’s supporters. He is not the only one to make this assumption. I think journalists like Glenn Greenwald, Matt Taibbi and a few others have aggressively pursued a similar thought, defending what they see is the unfairly maligned intelligence of their now predominately right-wing audience. And part of this involves canvassing right-wing conspiracy theories, even when they don’t personally agree with them. They justify this by asserting that while some of these conspiracy theories may be wrong, they represent a rational, anti-establishment alternative to elite sanctioned narratives. Where Rancière differs is he can see that many of these conspiracy theories derive from the same institutionalized sentiments as those of the mainstream consensus. In the case of January 6, there is the sentiment associated with the institution of representative democracy, as well as the sentiment associated with inequality which is reaffirmed almost everywhere in democratic societies. Representative democracy itself is founded on the idea of a minority governing on behalf of a majority. The riot on January 6 was hardly an example of radical egalitarian defiance, as Greenwald & co. naively seem to think.
Rationalism
The excursus on Rancière’s essay was meant to address the question: what are the consequences of assuming the rationality of those we cannot reach an agreement with? The answer is that we have to abandon the idea that rationality is oriented toward consensus and truth. Instead, the point of rational disagreement in the absence of consensus is political. Not necessarily in an ideological or party-political sense, but in the sense that it involves a clash between different fictional configurations of reality, or “communities of sentiment [sensible]”.
Elsewhere Rancière talks about the “partition of the sensible” (“le partage du sensible”): a way of making sense of the world by dividing it up in certain ways: distinguishing between appearance and reality, rational speech and nonsense, competence and incompetence. The “spilt ink” which Kavanagh objects to in Alexander’s engagement with conspiracy theories could be taken as a reference not just to an excess of words, but also to the mixing up of these divisions: the risk of granting speech to the incompetent and ascribing reality to nonsense. Politics for Rancière is the permanent possibility of contesting established ways of dividing up the world, because the very intelligence which is required to participate in these divisions, to understand and conform to them, is the same intelligence capable of partitioning the world differently, of redrawing distinctions along different lines.
Neither Alexander nor Kavanagh seem to be all that interested in discussing politics. But Kavanagh at least can perceive its relevance, as when he takes issue with Alexander’s injunction to ignore the political controversies which the media often use to frame scientific disagreements. Kavanagh claims on the contrary that the political framing can be quite informative. As in the case of conspiracy theories, he recognizes the role that sentiment or “the partition of the sensible” plays in these debates by shaping people’s perceptions. What Kavanagh would probably dispute is the idea that the fictions constructed on the basis of sentiment are indicative of a rationality common to all.
Alexander for his part wants to assume a common rationality, but to do so he conceives of rationality by downplaying what it has in common with fiction; despite the fact he claims to have learned about rationality from a fiction! For Alexander there is only one rational way to configure the world, which ensures that rational people in dialogue will eventually come to a consensus about the truth. But if that were the case, then every disagreement would amount to a misunderstanding. In cases where this misunderstanding could not be overcome, when someone could not be made to understand and persisted in their disagreement, then one is entitled to conclude that they are not rational and (according to Alexander) may even be genetically defective. Far from assuming equality, then, this argument seems to function as just another way of proving inequality.
I suspect that the last thing Alexander wants to acknowledge is politics in Rancière’s sense. Perhaps Alexander’s most famous essay, “Mistake vs Conflict”, sets out to demonstrate that those who treat social problems as a site of conflict, rather than as mistakes to be corrected, are for the most part not being rational. This essay has been cited by the likes of Claire Lehman and Steven Pinker in order to promote a technocratic vision for the management of society. The assumption, let us say fiction, upon which this argument is based is that the liberal capitalist order represents an historical necessity, one of technological and economic progress, rather than a system of domination and exploitation. Any resistance or deviance from this path is misguided and dangerous (see the fate of communism). Thus one ought to abandon misguided calls for revolution and instead learn to manage capitalism and its seemingly never-ending succession of crises.
Here I would have liked to discuss at some length Rancière’s essay “Time, Narrative, Politics” (from Modern Times). He considers an account of rationality virtually identical to that presented in “Mistake vs Conflict”, and notes that it relies on a hierarchy of temporal orders which is correlated with a hierarchy of knowledge. There is the “rational time” of historical necessity, which today is understood to be the time of globalized capital. And then there is the time of everyday life, one of daily routines and repetition. Rationality about human affairs, at least since Aristotle, has consisted in thinking according to the logic of the higher order of time, which today means according to the logic of capital. The “mistake theory” which epitomizes rationality for Alexander involves the management of society by implementing reforms which will synchronize ordinary time with the time of globalized capital: e.g. by raising the retirement age, lengthening the working day, abolishing holidays, and so on. What becomes apparent about this theoretical model when viewed in this way is just how much mistake theory resembles the conflict theory of Marxism, which also proposed a grand narrative about capitalism and defined rationality as the ability to perceive its contradictory logic.
If we situate Alexander’s discussion of rationality within the grand narrative of an historically necessary capitalism, then we can infer why Alexander extends the assumption of rationality to conspiracy theories. It is because he interprets these theories’ concern for uncovering the truth behind appearances as a concern for rational time: e.g. ivermectin is simply being proposed as an effective means of ending the pandemic, allowing ordinary life to once again “return to normal”, i.e. synchronize with the imperative of economic growth. But if we understand the assumption of rationality in Rancière’s sense, as an equality of intelligence by means of which different senses of reality can be constructed, then the sentiment behind such conspiracy theories cannot be so easily ignored. While the sentiment behind the pro-ivermectin side will need to be reconstructed differently to that of the January 6 crowd, it appears to be similarly organized around a sense of privilege. This point amounts to neither a sociological analysis nor a presumption of irrationality, but rather takes seriously the rhetoric of those who promote anti-vaccine conspiracy theories: ivermectin’s supposed effectiveness is proof of the strength, independence, intelligence, faith, etc. of those who refuse vaccinations, as opposed to the weak, dependent, brainwashed, atheists, etc. who believe the lies about vaccinations, which in reality are a tool of control perpetrated by an authoritarian, corrupt, socialist, satanic, etc. government.
When Kavanagh remarks that Alexander’s engagement with the pro-ivermectin side demonstrates the limitations of rationalism, I think he has in mind the rationalist’s ignorance or refusal to consider the political subtext of conspiracy theories. But since Kavanagh seems to share Alexander’s notion of rationality, he can only explicate this point using the language of cognitive bias. This means that Kavanagh must either condemn such political subtexts as irrational, leaving him open to the charge that he is preemptively excluding conspiracy theories from rational critique and so exacerbating the problem of echo-chambers and the like. Or else he ought to accept Alexander’s point that they are merely mistaken and can be brought round to the correct view through rational engagement. Rancière’s approach allows us to see this political subtext in a different light. Not as irrational but as the very demonstration of a dominant form of rationality, one which discovers inequality everywhere it looks. Not as a misstep on the path to consensus, but a disagreement with respect to the way we divide up and perceive the world. It suggests a different mode of engagement. Not a patient, scholarly explanation of why these people don’t understand the evidence. But rather confronting them with demonstrations of equality.
In what might these demonstrations consist? At the end of “Time, Narrative, Politics”, Rancière makes a tentative connection between certain forms of protest and modern literature. He notes that fiction in modern times makes a break with the rationality of fiction which has dominated reasoning about human affairs since Aristotle. Paradoxically, the rationality of modern literature is less “fictional” in the traditional sense than that of the social sciences. Much of Rancière’s recent work is dedicated to trying to establish the ways in which the rationality of modern literature might illuminate the practice of politics. But that is a story for another time.
Note: I have previously discussed Rancière’s notions of democracy and equality here.
