When Inferring to a Conspiracy Might Be the Best Explanation1

M R. X. Dentith

1. INTRODUCTION

While philosophers have been late in coming to the analysis of these things we call "conspiracy theories," it seems that — as a discipline — many of us analyze them with much more sympathy than our peers in the social sciences. In a raft of papers and books, starting with Charles Pigden's "Popper Revisited, or What Is Wrong with Conspiracy Theories?" (1995), philosophers like Brian L. Keeley (1999), Juha Räikkä (2009a), Joel Buenting and Jason Taylor (2010), Lee Basham (2011), David Coady (2012), and myself (2014) have argued that as conspiracies occur — and that theories about conspiracies sometimes turn out to be warranted — conspiracy theories cannot automatically be dismissed just because they are called "conspiracy theories."

This does not mean that philosophers consider belief in conspiracy theories to be the exemplar of rational thinking; the current findings in the Philosophy of Conspiracy Theories (to coin a new discipline) simply show that belief in conspiracy theories is not prima facie irrational. Rather, the kind of issues often held up as being a problem for belief in conspiracy theories tend to also be examples of issues common to a great many other beliefs that we do not typically think of as prima facie suspicious.

However, there still remains the view among some philosophers — and a great many thinkers in other disciplines — that even if explanations of events citing conspiracies can be warranted, conspiracy theories themselves are still unlikely. For sure, many of the complaints such conspiracy theory theorists (to coin another name) have about belief in conspiracy theories get phrased in terms of conspiracy theorists seeing conspiracies where none exist, or being prone to making bad inferences. Yet at the heart of these complaints — as we will see — are claims either about the unlikeliness of conspiracies or conspiracy theories. These claims of unlikeliness are then meant to explain why most ordinary people (and a large number of academics) are justified in treating conspiracy theories as a kind of suspicious belief. Yet, as we shall see, it is not obvious that conspiracy theories are unlikely. Indeed, when we understand what this supposed unlikeliness means, it throws an interesting light on how we sometimes avoid talking about just how much conspiratorial activity might be going on around us. Not just that: as we will also see, our judgments about the likeliness of conspiracies and conspiracy theories, in turn, affect our judgments as to when some theory about a conspiracy might just qualify as an inference to the best explanation.

1.1 Philosophers and the Unlikeliness of Conspiracy Theories

Let us start with the philosophers. Karl Popper, in The Open Society and Its Enemies, considers conspiracy theories to be unlikely. Why? Because such theories take it that history is largely the result of a succession of successful conspiracies. However, Popper believes that as most of us accept conspiracies are both rare and seldom successful, conspiracy theories are just unlikely (1969). While Popper accepts that conspiracies occur, belief in what he calls the "conspiracy theory of society" is irrational because conspiracy theories are rarely warranted. Popper's most modern echo is Quassim Cassam, who argues that conspiracy theorists suffer from the epistemic vice of gullibility. As such, while he — like Popper before him — admits conspiracies occur, conspiracy theories — being the kind of thing gullible conspiracy theorists believe — simply turn out to be so unlikely as to be untrue (2015).

Neil Levy argues that conspiracy theories which conflict with official theories — theories that have been endorsed by some authority — are prima facie unwarranted. As such, they turn out to be unlikely compared to their non-conspiratorial rivals (2007).

Pete Mandik takes a slightly different tack, and argues that when there is conflict between an official theory of the type "shit happens" and a conspiracy theory, then we have no good reason to prefer the conspiracy theory (2007). Conspiracy theories are, for Mandik, no more likely than their non-conspiratorial rivals. Indeed, because conspiracy theories often portray a world of complex causation — which might be better understood as the result of the law of unintended consequences — we are justified in thinking conspiracy theories are prima facie unlikely. This, in turn, justifies our preference for the hypothesis that "shit just happens."

Mandik's view echoes an argument put forward by Steve Clarke. Clarke argues that conspiracy theories are examples of dispositional explanations (explanations which rely on some central claim of someone intending for an event to happen). He argues that we should prefer situational explanations (explanations which rely on some central claim about the context or situation under which a series of events occurred), because situational explanations are better than claiming some event was the result of intentional activity. Given that most of the rival explanations to conspiracy theories — so Clarke claims — are situational in character, we should think conspiracy theories — as dispositional explanations — are unlikely, at least compared to their situational rivals (2002).

Peter Lipton — in what is admittedly only a passing reference in his book Inference to the Best Explanation — thinks that conspiracy theories may very well be unlikely, using such theories to illustrate how to disambiguate what it is we mean by "best" when parsing talk of inference to the best explanation.

By showing that many apparently unrelated events flow from a single source and many apparent coincidences are really related, such a [conspiracy] theory may have considerable explanatory power. If only it were true, it would provide a very good explanation. That is, it is lovely. At the same time, such an explanation may be very unlikely, accepted only by those whose ability to weigh evidence has been compromised by paranoia. (2004, 59–60)

Lipton distinguishes two notions of "best": the most likely explanation, and the one that provides the most understanding (with respect to some account of the explanatory virtues), which he calls the "loveliest explanation." Lipton considers conspiracy theories to be problematic because, while they have some lovely features (if they were true they really would provide a very good explanation as to why some event occurred) they are unlikely compared to their rivals. This is because Lipton assumes that conspiracy theories only seem likely because conspiracy theorists suffer from paranoia. For Lipton, it is evidence (or the lack thereof) that makes conspiracy theories unlikely because conspiracy theorists are paranoid, and thus see evidence of conspiracies where none exist.

1.2 Non-philosophers and the Unlikeliness of Conspiracy Theories

These philosophers who think that conspiracy theories are unlikely are in good company. For example, Cass Sunstein and Adrian Vermeule (foreshadowing Cassam) claim conspiracy theorists suffer from a "crippled epistemology." Conspiracy theories are unlikely because the:

[A]cceptance of such [conspiracy] theories may not be irrational or unjustified from the standpoint of those who adhere to them within epistemologically isolated groups or networks, although they are unjustified relative to the information available in the wider society[.] (2009, 204)

Which is to say that conspiracy theories look likely to conspiracy theorists, but only because they typically consort with, and gain information from, other like-minded individuals.

Similarly, Michael J. Wood and Karen M. Douglas argue that conspiracy theorists are, typically, more interested in disputing rival and official theories than they are promoting their own conspiracy theories. They characterize conspiracy theories as a kind of "negative belief," one which calls into question another explanation, and is indicative of a worldview in which most of what we are told is a lie. They consign belief in such theories to something akin to paranoia (Wood and Douglas 2013).

Jovan Byford differentiates conspiracy theories from "legitimate analyses of secrecy," arguing — like Popper before him — that as conspiracy theorists see the world as the product of successive and successful conspiracies, conspiracy theories are unlikely. This is because they do not reflect the way in which the world actually works (2011).

Meanwhile, Robert Brotherton and Christopher C. French build into their definition of what counts as a "conspiracy theory" that such theories are "an unverified and relatively implausible allegation of conspiracy, claiming that significant events are the result of a secret plot carried out by a preternaturally sinister and powerful group of people." As such, they take it as a given that conspiracy theories are going to be unlikely (2014).

Much of this kind of work accepts that conspiracies occur, but holds that conspiracy theorizing — the generation of, or coming up with, conspiracy theories — is a suspicious kind of activity to engage in. As Sander van der Linden writes:

Clearly, people and governments have conspired against each other, throughout human history. Healthy skepticism lies at the very heart of the scientific endeavor. Yet there is something fundamentally dangerous and unscientific about the nature of conspiracy theorizing. (2015)

There is something chilling about this kind of sentiment. "Yes," the proponent of this view agrees, "conspiracies occur. Just don't go around suspecting people of conspiring; that's unhealthy!" While often this suspicion of conspiracy theorizing is couched in terms of conspiracy theorists suffering from some ominous and psychological pathology, a failure by said theorists to think critically about their conspiracy theories, or the inability for adherents of conspiracy theories to contemplate non-conspiratorial alternatives, at root this suspicion stems from some variety of the claim "Look, conspiracies are unlikely, or even if they do occur, conspiracy theories are unlikely, right?" Yet this latter claim — no matter what we believe about the psychology of conspiracy theorists — is not something we should accept without further examination. This is particularly important because while many of us might reasonably think claims about conspiracies should be evaluated according to the evidence, many theorists — as we have seen — argue that we can dismiss such claims out of hand merely because they are conspiracy theories.

2. UNLIKELY COMPARED TO WHAT?

If we are told something is unlikely, we need to ask "In relation to what?" Likeliness is a relative thing. One argument for the relative unlikeliness of conspiracy theories is to claim that conspiracies are unlikely, say, because conspiratorial activity is taken to be rare or, if not rare, seldom successful. Popper — echoed by Byford — argues that conspiracy theorists see the world as the product of successive and successful conspiracies, a position many conspiracy theory theorists take to be obviously false; conspiracies are either unlikely, or, when they do occur, seldom successful. Brotherton and French — in a similar vein — posit that conspiracy theories are unlikely because they are based upon unverified and relatively implausible claims of conspiracy. For theorists of this ilk, conspiracy theories turn out to be unlikely, because of the unlikeliness of conspiracies.

Another argument for the relative unlikeliness of conspiracy theories is that they are unlikely because even if we accept that conspiracies occur, given a choice between a conspiracy and a non-conspiracy theory, the non-conspiratorial explanation will just be more likely, all things considered. This kind of view can be found in the works of Cassam, who takes it conspiracy theorists are gullible (and thus treat conspiracy theories as more likely than they really are); Sunstein and Vermeule (whose argument is a more sophisticated take on Cassam's), who argue conspiracy theorists only see conspiracy theories as likely because they suffer from a "crippled epistemology," born out of existing in isolated epistemic communities; and Levy (as well as Mandik) who claims we have no good reason to think conspiracy theories will ever be more likely than their rivals.

Yet it is not clear that conspiracies or conspiracy theories are relatively unlikely. For example, Kathryn S. Olmsted's book Real Enemies: Conspiracy Theories and American Democracy, World War I to 9/11 — in which Olmsted covers topics such as the secrecy behind the United States's entry into the First World War, the ills of the McCarthy Era, and the way in which the official theory of 9/11 was sometimes misrepresented by the authorities for political point-scoring — reads as a litany of US government–run conspiracies. Her calm and dispassionate historical analysis of a century of US political practice presents conspiracies not so much as deviation from the norm, but, rather, as standard operating practice (2009). We can add to this numerous other examples; the Gulf of Tonkin Incident in 1964; the Ford Pinto Scandal of 1977; the Snowden revelations of 2013, concerning the National Security Administration (NSA) in the United States covering up the existence of a mass surveillance program; and the Volkswagen Emissions Scandal of 2015. In each of these cases the idea that the perpetrators were up to no good was labeled as a "conspiracy theory." These examples are but the tip of an iceberg.

So, if, there really is anything to the claim conspiracy theories are relatively unlikely, then the debate about said likeliness depends on:

(a) how you define what counts as conspiratorial, and

(b) whether you accept the claim that any official theory which cites a conspiracy as a salient cause is no longer considered to be a conspiracy theory.

Understanding how our definitions of these key terms rules in or out certain kinds of activities or theories as conspiratorial ends up being important not just for our understanding of these things called conspiracy theories, but also for working out whether some claim about a conspiracy can ever qualify as being part of the best explanation for some event.

2.1 What Counts as a Conspiracy?

Conspiracy theories posit the existence of a conspiracy, where the conspiracy is the salient cause of some event. The most minimal conception of what counts as a conspiracy, then, must satisfy the following three conditions:

The Conspirators Condition: There exists (or existed) some set of agents with a plan.

The Secrecy Condition: Steps have been taken by the agents to minimize public awareness of what they are up to.

The Goal Condition: Some end is, or was desired, by the agents.

These conditions are individually necessary and jointly sufficient for some activity to be classified as conspiratorial, and it is fair to say that some beliefs about the likeliness or unlikeliness of conspiracy theories hinge on finessing or questioning such a minimal definition of what counts as a conspiracy.

For example, Mandik argues that theories about known conspiracies — citing such examples as the official theories of 9/11, the Watergate Affair, and the Iran-Contra deal — fail to be proper conspiracy theories. Why? Because the aforementioned conspiracies were not kept secret; they were not conspiracies in the sense that we typically talk about when discussing conspiracy theories (2007). Yet the view that a conspiracy is not a proper conspiracy unless the conspirators keep their activities properly secret is incredibly idiosyncratic. If that were the case, how could we have any belief in the existence of conspiracies? Indeed, if we accept Mandik's view, then it is not even clear that conspirators could believe in the existence of their own conspiracies. After all, the fact someone knows about the conspiracy means it is not being kept properly secret. Not just that, but if someone blows the whistle on a conspiracy, does that mean — under Mandik's view — that the conspiracy is no longer a conspiracy because it is no longer secret?

Mandik challenges the idea of known conspiracies being the kind of thing we mean when people talk about conspiracy theories. However, when it comes to the secrecy condition of conspiratorial activity, all we need say is that some conspirators will be more successful at keeping their existence and activities secret than others. If we restrict talk of what counts as conspiratorial to cases of proper secrecy, then not only are we using a restrictive definition (with some strange corollaries), but it just follows from said definition that conspiracy theories will be unlikely. After all, if a conspiracy must be kept properly secret, then the associated conspiracy theory turns out be irrational to believe. This is because it will not be based upon any good evidence of a conspiracy.