Conspiracy theories and their epistemology

You may believe that on 20 July 1969, astronauts Neil Armstrong and Buzz Aldrin landed on the surface of the Moon aboard the Lunar Module Eagle. Or you might think that the moon landing was a fake, made in a film studio under the direction of Stanley Kubrick. You might believe that American politics is characterized by the opposition of two parties, the Republicans and the Democrats, which alternate in power according to the results of regular democratic elections. Or you might think that beneath this reassuring surface there is a war going on between good and evil forces––with Donald Trump leading the former in an attempt to fight a satanic and pedophile ring secretly ruling the country. You may believe that the Earth is approximately spherical. Or you may think that this is what “they” want us to believe, when, in fact, the Earth is flat.

Each pair consists of mutually inconsistent views. If you choose the first element, you believe what is usually called the “official version” of a fact or event, while if you choose the second, you buy into what is called a “conspiracy theory” about it.

It is widely assumed in popular conversation (though not universally accepted, especially in the philosophical debate) that the label “conspiracy theory” denotes (or should denote) a problematic genre, grouping together theories that are epistemically flawed for reasons having to do with the fact that they posit a conspiracy. Yet, positing a conspiracy for explanatory purposes is not, in itself, sufficient to render a theory defective in the relevant way. The stories we read in history books about the assassination of Julius Caesar, the Watergate scandal and the collapse of the Twin Towers on 9/11 give conspiracies a prominent explanatory role. Yet, no one would call them “conspiracy theories”, as they are backed up by a solid historical record. This suggests that the unfortunate epistemic feature common to what we call “conspiracy theories” is not simply (or not necessarily) that they attribute to a conspiracy a causal explanatory role. Instead, I want to suggest that their most problematic feature is the role these theories assign to the hypothesis of a conspiracy in accommodating the available evidence. Let me explain.

It is a reasonable suggestion, explored at length in contemporary epistemology, that justificatory relations are grounded in explanatory relations. This means, roughly, that when the evidence E justifies believing some proposition P this is because (and to the extent that) the truth of P is part of the best available explanation of why E exists. To illustrate, suppose that Amanda’s fingerprints are found on the handle of the knife with which Tom was stabbed. This fact seems to justify the belief that Amanda must have held the knife in her hand (possibly during the crime). For this hypothesis explains better than anything else why Amanda’s fingerprints have been found there.

With this account of justification in play, let’s return to conspiracy theories. As we have seen, attributing to a conspiracy a causal explanatory role is not necessarily illegitimate. However, explaining an event is not the only thing that the hypothesis of a conspiracy might do for a theory. A different and arguably more problematic role is this. Suppose a theory, whether or not it adduces a conspiratorial explanation of an event, is faced with evidence it cannot explain. Suppose, further, that it contradicts an existing official explanation which is widely accepted and promoted by the relevant epistemic authorities. This evidence seems to refute the theory in question and to support the official version instead. However, the theory can be expanded so as to explain this recalcitrant evidence. One way to do so is by incorporating the additional claim that there is a conspiracy aimed at hiding the truth and to say that the adverse evidence with which it is faced has been doctored by the conspirators. Say in this case the theory attributes to a conspiracy a meta-evidential role.

The distinction between a conspiracy’s causal explanatory role and its meta-evidential role is unheard in the contemporary philosophical conversation on conspiracy theories. And while most theorists agree (though they don’t call it in this way) that a conspiracy theory must attribute to a conspiracy a causal explanatory role, my suggestion here is that what is really distinctive of conspiracy theories, as pejoratively understood, is that they credit a conspiracy with the meta-evidential role.

The difference between the two roles is easily obscured by the fact that many (perhaps most) theories that ascribe a meta-evidential role to a conspiracy also ascribe a causal explanatory role to it. Take, for example, the 9/11 “inside job” conspiracy theory. The theory explains the collapse of the Twin Towers as the result of a controlled demolition secretly organized by the US government, and consequently explains contrary expert evidence––such as the National Institute of Standards and Technology’s report that no evidence of a controlled demolition has been found––as coming from a federal agency complicit in the same conspiracy.

However, a theory does not have to ascribe a causal explanatory role to a conspiracy in order to ascribe a meta-evidential role to it, and thus to be a conspiracy theory in the sense under consideration here. The flat earth theory illustrates this point well. The theory’s main claim is not about a conspiracy, it is about the shape of the Earth. So, the theory doesn’t attribute to a conspiracy any explanatory causal role. Crucially, however, the theory’s main claim is unable to explain the available evidence. This is constituted, for instance, by photographic evidence depicting a globe from space. The hypothesis that the Earth is a globe explains the existence of this evidence better than the hypothesis that the Earth is flat. So, this evidence justifies the former theory and disconfirms the latter. Things possibly change when a conspiracy is attributed the meta-evidential role, though. Paired with the extra-hypothesis that a group of conspirators wants to hide the true shape of the Earth from the world, the claim that the Earth is flat can accommodate the existence of evidence showing otherwise. The conspirators, after all, have power. So, that they should exercise it and try to pollute the evidential environment so as to place the true shape of the earth beyond anyone’s epistemic reach is exactly what should be expected.

Two epistemological consequences deserve emphasis if one accepts the claim that the defining feature of a conspiracy theory is that it fulfils the meta-evidential condition.

The first thing to notice is that, on this characterization, a theory can alter its status along the way, start as a conspiracy theory and grow into something different as new and better evidence is amassed which the theory’s main claim is able to explain without attributing to a conspiracy the meta-evidential role. When this happens, the theory ceases to be a conspiracy theory and its main claim may become justifiably believable on the strength of the new evidence. This could happen, for instance, if new scientific data were uncovered showing that the Coronavirus was engineered in a Chinese military lab. A claim that used to be part of a conspiracy theory­­––when it was made in spite of the scientific consensus that the Coronavirus originated in wildlife––would become a respectable scientific hypothesis.

A second thing to notice is that a conspiracy theory is not necessarily irrational from an epistemic point of view even when it retains its status. A conspiracy theory is epistemically justified only if it supplies an explanation of the evidence which is good enough to outperform the explanation provided by the official theory. This explanation always involves the suggestion that part of the evidence has been doctored by the conspirators, which can make appear it epistemically suspect. However, nothing forecloses the possibility that, under certain specific circumstances, such an explanation may be better than alternative (including official) explanations. Whether such circumstances exist and how likely they are is thus a question that, based on the account I have described, can be considered open-mindedly.

Kristina Flour on Unsplash