Conspiracy theories and conspiracy theorizing

Conspiracy Theories and Conspiracy Theorising Steve Clarke Centre for Applied Philosophy and Public Ethics, Charles Sturt University LPO Box 260, ANU, Canberra, ACT 2601, Stephen.Clarke@anu.edu.au/stclarke@csu.edu.au June 2002 Abstract The dismissive attitude of intellectuals towards conspiracy theorists is considered and given some justification. It is argued that intellectuals are entitled to an attitude of prima facie skepticism towards the theories propounded by conspiracy theorists, because conspiracy theorists have an irrational tendency to continue to believe in conspiracy theories, even when these take on the appearance of forming the core of degenerating research programmes. It is further argued that the pervasive effect of the ‘fundamental attribution error’ can explain the behaviour of such conspiracy theorists. A rival approach due to Brian Keeley, which involves the criticism of a sub-class of conspiracy theories on epistemic grounds, is considered and found to be inadequate. 1. Introduction Conspiracy theorists, like creation scientists and astrologers, are advocates of ideas that are generally quite popular. There are large numbers of people who believe that the United States’ military has conspired to keep the public uninformed about visits by alien life forms (Shermer 1997, pp. 91-3.). Others believe that Elvis Presley conspired to fake his own death (Brewer-Giorgio 1988). Some even believe that Ludwig Wittgenstein led a secret double life as a Soviet spymaster (Cornish 1997). Like the ideas promulgated by creation scientists and astrologers, the ideas promulgated by conspiracy theorists are very unpopular amongst intellectuals. “That’s just a conspiracy theory,” say the intellectuals, apparently feeling entitled to dismiss such theories simply on the grounds that they involve conspiracies. No matter how hard conspiracy theorists have tried to make their cases, they are not accorded the same hearing in intellectual circles that a proponent of a non-conspiratorial explanation would be accorded. This situation is more than slightly mystifying. Unlike creation scientists and astrologers, conspiracy theorists have epistemic ‘runs on the board’ A measure of success in the game of cricket., and lots of them. Conspiracies have consistently taken place throughout history. Elvis Presley may or may not have faked his own death, but Richard Nixon really did conspire to cover up his involvement in the Watergate burglary, and Cecil Rhodes really did conspire to provoke conflict between the British Empire and the Boer Republics. No doubt history plays its part in explaining the hostility of intellectuals towards conspiracy theories. Conspiracy theorising has long been favoured by Populists, who are almost invariably anti-elitist, and therefore generally anti-intellectual as well. Some intellectuals may dismiss conspiracy theories simply on the basis of guilt by association with anti-intellectual Populism. Opposition to perceived elites and to elite cultures, such as the culture of intellectuals, is endemic in Populist thought. Lazer offers a ‘working definition’ of Populism as the belief that “the majority opinion of the people is checked by an elitist minority” (Lazer, 1976, p. 259). Although Populist movements are almost invariably anti-elitist, they are not invariably anti-intellectual, for the simple reason that intellectuals are not always perceived as being part of an elite. The Russian Populist Narodnik movement of the 1870s was a political grouping predominantly consisting of anti-elitist intellectuals. American Populists of the late nineteenth century were particularly noted for their propensity to indulge in conspiratorial reasoning, a feature of their thinking that the historian Richard Hofstadter highlights, informing us that: There was something about the Populist imagination that loved the secret plot and the conspiratorial meeting. (Hofstadter 1955, p. 70) and: At the so-called grass roots of American politics there is a wide and pervasive tendency to believe  I hasten to add that the majority of Americans do not habitually succumb to this tendency  that there is some great but essentially very simple struggle going on, at the heart of which there lies some single conspiratorial force ... (Hofstadter 1955, p. 16) Although guilt by association with anti-intellectual Populism can go some way to explaining the contempt that intellectuals have shown towards conspiracy theories, there are at least two reasons to think that it cannot explain it all. First, guilt by association is a fallacious form of reasoning and intellectuals, generally being intelligent people, can reasonably be expected to be somewhat resistant to the lure of fallacious forms of reasoning. Second, many American intellectuals have shown a marked sympathy for the anti-elitist tendencies of Populism that lead Populists to be anti-intellectuals. For such intellectuals, association with anti-elitist movements such as Populism, would be a cause of virtue by association rather than guilt by association. Liberal intellectuals in America may be part of an elite culture, however they are often self-consciously and somewhat reluctantly part of such a culture. Hofstadter paints an unflattering portrait of nineteenth century American Populism, stressing the continuities between it and 1950s McCarthyism (1955, 1971). However, in the main, American historians have been quite sympathetic to nineteenth century American Populism, emphasising the extent to which Populist grievances against the urban East Coast elites of America were legitimate, and portraying Populism as a ‘progressive’ rather than a ‘reactionary’ movement. Canovan describes the development of debate about the status of nineteenth century American Populism amongst historians (1981, pp. 46-51). However much we think that an historical association with anti-intellectual movements can explain the antipathy of intellectuals to conspiracy theories, it is hard to see how it could justify such an attitude. Given that we know that conspiracies have occurred, could intellectuals really be justified in dismissing conspiracy theories merely by pointing to the fact that these are just conspiracy theories? If so, then surely we ought to be able to say why. This paper is an attempt to do so. Although I will not argue that intellectuals are entitled to arrogantly dismiss all conspiracy theories, I will argue that there is an entitlement to an attitude of prima facie skepticism towards the theories propounded by conspiracy theorists and I will identify a reason which grounds this entitlement. Conspiracy theories are not a common topic of philosophical discussion; however I am not the only philosopher who has discussed them. In a recent article Brian Keeley has attempted to explain their apparent popularity and identify their underlying deficiencies. Others are Pigden (1995) and, somewhat elliptically, Popper (1966). Keeley has undertaken an important project but has gone about completing it in the wrong way. Keeley attempts to identify a sub-class of conspiracy theories which he describes as Unwarranted Conspiracy Theories (UCTs). These have crucial epistemic deficiencies that go unrecognised by conspiracy theorists according to Keeley. It will be shown that Keeley’s case against UCTs is exaggerated and confused. He intends to attack a class of conspiracy theories, but through a slide in his reasoning, ends up attacking the reasoning patterns of conspiracy theorists, arguing that they have a tendency to fail to recognise the epistemic weaknesses of the conspiracy theories that they favour and that they have a tendency to respond inappropriately to evidence which conflicts with these theories. Although Keeley has not identified an underlying failing common to conspiracy theorists, it is perhaps no accident that he ends up attacking the reasoning patterns of conspiracy theorists, because this is where the real problem lies. I will argue that conspiracy theorists are typically victims of a form of cognitive failure. They are unusually prone to committing what social psychologists refer to as the ‘fundamental attribution error’. The fundamental attribution error has recently become a ‘hot topic’ in moral philosophy. See for example Campbell (1999), Doris (1998), Doris (forthcoming), and Flanagan (1991). I will describe this error, and explain how it leads conspiracy theorists to hold on to theories which they would otherwise abandon. I then conclude with a few words in favour of the conspiracy theorist. Although conspiracy theorists do commit a cognitive error that leads them to prefer theories which are otherwise less plausible over theories that are otherwise more plausible, the activities of the conspiracy theorists are not to be condemned outright. The prevalence of conspiracy theorising is beneficial to us in several ways. Before I go on, here are a couple of definitions. Charles Pigden defines a conspiracy as “ ... a secret plan on the part of a group to influence events partly by covert action.” (Pigden 1995, p. 5) Keeley defines a conspiracy theory as: ... a proposed explanation of some historical event (or events) in terms of the significant causal agency of a relatively small group of persons  the conspirators  acting in secret.” (Keeley 1999, p. 116) Although one author offers a definition of a conspiracy and the other a definition of a conspiracy theory, it is apparent that the two authors are largely in agreement. Keeley’s theories are effectively theories about Pigden’s conspiracies. The only substantial difference between them is that Keeley is more restrictive about the number of conspirators acceptable in a genuine conspiracy. The reader may accept either definition in the argument that follows. 2. Epistemological Problems Sometimes it is thought that conspiracy theorists are simply people who are gripped by a desire to believe the truth of their ‘pet’ theory, and because of this desire they are not motivated to go and find evidence for the pet theory to the extent that they would that normally be motivated to find evidence for a given theory, before accepting that theory. It is plausible to believe that people are initially attracted to belief in conspiracies because of the emotions that these stir in them. Hume made a similar claim on behalf of miracles when he wrote that “The passion of surprise and wonder, arising from miracles, being an agreeable emotion, gives a sensible tendency towards the belief of those events, from which it is derived.”(Hume 1988, p.150). Belief in the ability of conspirers to carry out their plans, while duping others into believing in a ‘cover’ story, may inspire similar feelings. But while it is plausible to hold that there are people who are emotionally attracted to belief in conspiracy theories, anyone who concluded that because of these feelings conspiracy theorists would go on believing in conspiracies theories on the basis of less evidence than they would otherwise require to substantiate belief, cannot have had much contact with conspiracy theorists. Because conspiracy theorists almost always wish to see conspiracies exposed, they are typically quite dedicated in their search for evidence relevant to their favourite conspiracy theory and are usually able to overwhelm you with a deluge of evidence in favour of that theory. Gail Brewer-Giorgio, the author of Is Elvis Alive? (1988), is one such typical conspiracy theorist. Her argument for the conclusion that Elvis Presley faked his own death brings together a vast array of evidence. The conventional explanation of the death of Elvis Presley, in 1977 at age 42, has a meagre evidential base, referring to Elvis’ heart condition which is explained by appealing to facts about his lifestyle. Proponents of the received non-conspiratorial view about Elvis’ death could presumably acquire more evidence for their view. The conspiratorial rival theory that Brewer-Giorgio mounts is rich in detail, explaining, inter alia, why Elvis’ middle name is misspelt on his tombstone (Elvis was superstitious and wouldn’t want his name correctly spelt on a tombstone when he was in fact alive), and why Elvis’ casket was unusually heavy (it contained a wax dummy and an air conditioning unit to stop the wax dummy from melting). Brewer-Giorgio can also explain away the apparent plausibility of the conventional explanation of Elvis’ death. Given that Elvis wished to fake his own death it would have suited him to appear to be in poor health, so as to add plausibility to his cover story. Conspiracy theories invariably seem to be based on more evidence than their immediate rival, the non-conspiratorial ‘received view’. This is because they explain all that the non-conspiratorial received view explains  the apparent plausibility of the non-conspiratorial received view is a consequence of the success of the cover story or ‘cover up’, according to conspiracy theorists  and then go on to account for evidence that the received view is unable to explain. Once a conspiracy theorist has become committed to a conspiracy theory she is able to account for almost any relevant evidence that is presented. It is either evidence of the cover-up, which the conspirers are attempting, or it is evidence of discrepancies in the received explanation. Strictly, none of this can be described as ad hoc. A theory that involves an attempt by some people to deceive other people is a theory which involves reasons both to expect a cover-up and flaws in the cover-up. Keeley makes this point quite forcefully (1999, p.121). It is not hard to understand why conspiracy theorists are able to accommodate new evidence in their theories. What is hard to understand is why conspiracy theorists continue to be motivated to do so. Conspiracy theories often have the appearance of forming the core of what Lakatos referred to as ‘degenerating research programmes’ (Lakatos 1970). Research programmes are very roughly what Kuhn refers to as ‘paradigms’, research traditions built around a core theory in which participants are dedicated to advancing the case for that core theory, and dedicated to protecting it from the effect of apparently contradictory evidence by making modifications to the ‘protective belt’ of auxiliary hypotheses and initial conditions. Lakatos urged us to make judgements regarding the success of research programmes, distinguishing between progressive and degenerating research programmes according to the way in which these relate to new evidence. A progressive research programme is one in which novel predictions and retrodictions are made that are generally successful. In a degenerating research programme successful novel predictions and retrodictions are not made. Instead auxiliary hypotheses and initial conditions are successively modified in light of new evidence, to protect the original theory from apparent disconfirmation. Social scientists and social commentators are not usually expected to make exacting predictions and retrodictions. Nevertheless it is often possible to discern the expected consequences of a social theory. For example, if classical Marxist social theory is true then we can reasonably expect heightened economic instability in capitalist countries at some stage in the not-too-distant future. Because conspiracy theories typically involve an ongoing conspiracy it is often particularly clear what consequences can be expected to result from a conspiracy theory being true. When Bernstein and Woodward formulated the core of the ‘Watergate Conspiracy’, they were led to reasonably expect the complicity of a number of individuals in the conspiracy and to make rough and ready predictions, as well as retrodictions, about the behaviour of these conspirators. These predictions and retrodictions turned out, for the most part, to be true and the Watergate conspiracy has gained acceptance as a successful conspiracy theory. The story of the uncovering of the Watergate conspiracy is recounted in Bernstein and Woodward (1974). In contrast, the theory that asserts that Elvis Presley faked his own death is at the core of what appears to be a clear example of a degenerating research programme. If Elvis Presley really was alive now (over 20 years after his supposed death) then we would reasonably expect to have some firm evidence of his activities at some time in the past two decades. The behaviour of his relatives, who Brewer-Giorgio alleges are co-conspirators and probably in continuing contact with Elvis, would also be expected to be highly unusual in a number of ways. But no reliable evidence of the Presley relatives attempting to contact Elvis after 1977, or inadvertently revealing their complicity in a conspiracy, is forthcoming. Lakatos did not say exactly when it becomes irrational to cling to a degenerating research programme, but there clearly are cases where a research programme has degenerated beyond the point where it is reasonable to hold on to it. The problem with conspiracy theorists is that they usually seem loath to give up conspiracy theories when this occurs. It might perhaps be thought that we do not need an explanation for the tendency of people to continue to believe in conspiracy theories even when these exhibit signs of having become degenerating research programmes. After all, significant numbers of scientists continue to remain committed to research programmes when these show signs of clear degeneration and are no longer dominant in their field. However, there appears to be a relevant difference between these ‘elderly holdouts’ in science, and conspiracy theorists. In the case of scientists there are compelling sociological explanations for their continued commitment to unsuccessful ideas. A scientist who has invested a career in a particular research programme stands to lose much of her credibility by renouncing past intellectual commitments. Furthermore, the scientists may work as part of a research team that is committed to the degenerating research programme. She may not be able to join another research team which is participating in a different research programme. Even if she has doubts about the viability of the degenerating research programme she may find it impolitic to make these public. Renouncing her chosen research programme could involve ending her research career. There is a wealth of recent literature in science studies which provides sociological explanations for the intellectual commitments of scientists. Biagioli (1999) contains the best of most of it. In stark contrast the conspiracy theorist typically stands to make substantial social gains by abandoning commitment to a particular conspiracy theory. Abandoning conspiracy theories means ending the ridicule of intellectuals and this can allow one the realistic prospect of a return to intellectual ‘polite society’. It should be conceded that the thinking of some conspiracy theorists could be importantly affected by cultural factors which help to explain their continued allegiance to degenerating research programmes. In some cases conspiracy theorists may belong to a sub-culture of fellow theorists which acts to significantly affect their judgment. It is plausible to believe that some subcultures, such as the subcultures of religious cults, can affect the reasoning of their members in such a way as to make them believe things that they would otherwise not believe. Religious cults expend much effort to indoctrinate or ‘brainwash’ their members. However, such cultural explanations will not go very far to explain the continuing popularity of degenerating research programmes with conspiracy theories at their core. Conspiracy theorising is most popular amongst members of the general public who are perhaps affected by, but not actively indoctrinated by, subcultures of conspiracy theorists. Conspiracy theorising may be culturally transmitted, but in most cases it is not strongly culturally maintained. 3. Unwarranted Conspiracy Theories Keeley, as was already mentioned, attempts to identify a class of conspiracy theories that are unwarranted because of their epistemic deficiencies. In addition to satisfying his definition of conspiracy theories, Keeley’s unwarranted conspiracy theories (UCTs), have the following characteristics: (1) A UCT is an explanation that runs counter to some received, official, or “obvious” account. (2) The true intentions behind the conspiracy are invariably nefarious. ‘Nefarious’ seems overly strong. If Elvis Presley really did conspire to fake his death is it hard to see his intentions as actually being nefarious. This is certainly not how Brewer-Giorgio (1988) portrays him. (3) UCTs typically seek to tie together seemingly unrelated events. (4) ... the truths behind events explained by conspiracy theories are typically well-guarded secrets, even if the ultimate perpetrators are sometimes well-known public figures. (Keeley 1999, pp. 116-7). A fifth characteristic is formulated around the activity of conspiracy theorists: (5) The chief tool of the conspiracy theorists is what I shall call errant data (Keeley 1999, p. 117). As used by Keeley “errant data” is a relative term. Relative to a received theory data is errant if it is unaccounted for, or contradictory to, that received theory. Before we consider Keeley’s arguments for the epistemic weaknesses of UCTs, we need to say a bit more about his list of characteristics for UCTs. Although the theories that have these characteristics are meant to have the epistemic deficiencies which Keeley argues for, the list of characteristics is looser than a set of necessary and sufficient conditions for a conspiracy theory being unwarranted. Some unwarranted conspiracies do not have all of the listed characteristics, and some conspiracies that do have all of them are conspiracies which we are warranted in believing in. Keeley mentions Watergate and the Iran-Contra affair, as conspiracies that have all the characteristics of a UCT, but which we are nevertheless warranted in believing in (Keeley 1999, p. 118). Perhaps the epistemic deficiencies common to UCTs are somehow overridden in these cases? Keeley does not address this question. It is not clear whether Keeley believes that necessary and sufficient conditions for an important class of UCTs could be articulated or not. However, he clearly believes that he has articulated a characterisation of UCTs in sufficient detail to allow for these to be usefully criticised as a class. I am concerned to attack these criticisms and I will set aside questions about the worthiness of his characterisation of UCTs. Keeley considers the explanatory virtues of UCTs which he claims are ‘subtly flawed’. The explanatory virtues of UCTs are the very sources of their epistemic deficiencies, and this is what makes it difficult for us to properly assess them; a situation which goes some way to explaining their enduring popularity, or so he contends. The epistemic virtues of conspiracy theories are that they provide unifying explanations and that they explain errant data. An epistemic over-reliance on the ability to explain errant data is the first flaw of UCTs, according to Keeley. Keeley suggests that it is appropriate to place great stress on explaining errant data in the natural sciences, but inappropriate in social explanation, because we ought to be modest about our ability to gather reliable data about the human world. Much of what we accept as the data to be explained in social theorising will turn out to be false. Apparently, because of this consideration, we respond appropriately to the conspiracy theorist who challenges us to explain errant data by simply shrugging our shoulders. The other flaw of UCTs, according to Keeley, is that these have an inbuilt undermining feature. If we accept them then we will have reason to abandon our confidence in the trustworthiness of the people and the institutions that are involved in the conspiracy. Keeley sees this problem as very serious as the following quote suggests: These theories throw into doubt the various institutions that have been set up to generate reliable data and evidence. In doing so, they reveal just how large a role trust  in both institutions and individuals  plays in the justification of our beliefs. (Keeley 1999, p. 121) Acceptance of UCTs threatens to put us in a position where our confidence in authorities is so eroded that we are no longer warranted in holding any beliefs which are socially produced, according to Keeley. Thus, we are left in a position analogous to the lamentable position of no longer being able to accept the legitimacy of the mechanisms that warrant belief production in the sciences; no longer being justified in believing that “the platypus is a mammal and that gold is an atomic element” (Keeley 1999, p. 121). Such epistemic endpoints appear to embody a degree of skepticism that is too high to be acceptable to anyone. Keeley’s first charge against conspiracy theories is not at all telling. Keeley is right that social data is generally less reliable than the data that natural scientists locate, but he fails to inform us as to why this consideration should tell particularly against the errant data that conspiracy theories explain. It appears to be a consideration which tells equally against all social data. Furthermore, it is very hard to see how this could be a consideration which tells against errant data more than against non-errant data. Errant data is only errant in relation to an accepted theory and to discount errant data on grounds which apply to both errant and non-errant data, would be to prejudice oneself in favour of data simply because it happens to be explained by the received theory. The second charge is slightly more telling, but only slightly more. Keeley is surely right to argue that trust in institutions and individuals plays a crucial role in the justification of our beliefs, as a number of philosophers have recently emphasised (Coady 1992, Lipton 1998). One of the reasons against believing in the Watergate conspiracy, when it was initially propounded, was that this would involve the undermining of our warrant for the beliefs that we had accepted on the basis of the testimony of Nixon, Mitchell, Ehrlichman et. al. We might similarly hesitate to believe allegations that the Bureau of Alcohol, Tobacco and Firearms (BATF) conspired in the Oklahoma City bombing because this would involve undermining our warrant for beliefs that were accepted because of the testimony of representatives of the BATF. These sorts of considerations ought to be taken into account when deciding whether or not to accept social theories, but it is difficult to believe in the case of most UCTs, that they are the very serious considerations which Keeley takes them to be. Potentially, conspiratorial thinking could lead us to doubt beliefs as entrenched in our current thinking as the belief that the platypus is a mammal. But this is a far cry from the usual degree of undermining that accepting a UCT involves. The Watergate conspiracy (which has all of the characteristics of a UCT) involved the participation of government officials at the highest level and its acceptance ought to undermine our confidence in the truth claims of those particular officials. However, it is unclear why it would erode our confidence in other government officials and agencies, much less induce pervasive and debilitating skepticism. Consider how the Watergate conspiracy might be adapted to embrace skepticism as far reaching as skepticism about the taxonomic status of the platypus: in his increasingly paranoid state of mind Nixon became convinced that the platypuses’ taxonomic status disposed voters in favour of the Democratic Party, and so he ordered Gordon Liddy to interfere with the documentation of basic science so as to have the platypus falsely taxonomised. I take it that it is clear that no actual version of the Watergate conspiracy involves such allegations, or allegations remotely like these. Indeed it is hard to see that any of the conspiracy theories Keeley contemplates have the effect of undermining very entrenched mundane beliefs. Keeley does make an attempt to show that acceptance of one particular conspiracy theory, the theory that the Holocaust was faked, leads to an unacceptably general level of skepticism. He informs us that belief that the Holocaust was faked is akin to belief that World War II did not happen as well as to various other unlikely beliefs (Keeley 1999, p. 123). The case is made too swiftly to be convincing. Keeley makes much more of the importance of the second flaw of UCTs than he is entitled to because of a slide in his reasoning. He observes that some contemporary conspiracy theorists have a tendency to react to criticism of their preferred theories by adapting these theories and increasing the number of conspirators involved in the alleged conspiracy. Apparently there are some conspiracy theorists who began by alleging that the BATF was conspiratorially involved in the Oklahoma City bombing and now allege a conspiracy involving the collusion of the FBI and parts of the press. Keeley contends that such a pattern of development in reasoning sets these conspiracy theorists down the slippery slope towards ‘almost nihilistic’ degrees of skepticism. This pattern may often be followed, but the fact that it is often followed is an observation about the fallacious reasoning patterns of some contemporary conspiracy theorists and this is simply not relevant to the epistemic evaluation of UCTs as a class. In any case, a sophisticated conspiracy theorist could advance a conspiracy theory with the reach which Keeley contemplates, and resist Keeley’s slippery slope to pervasive skepticism, by introducing the following consideration. Due to the nature of conspiracy, we can generally expect conspiring agents to be more reliable than they would otherwise be about matters which are not directly relevant to the conspiracy. Suppose that Nixon really did have Liddy interfere with all scientific records so as to have the platypus falsely taxonomised. If Nixon’s concern is to deceive us about platypuses and not about other animals, then this need not undermine our confidence that the wombat is a mammal and that the kookaburra is a bird. It is important to conspirators that their conspiracies go undetected. Part of one’s cover as a conspirator is to ensure that one has a good reputation, and this will typically involve maintaining high standards of honesty and epistemic responsibility on all issues other than those directly relevant to the conspiracy. It might be thought that I am making too much of the distinction between conspiracy theories and conspiracy theorists. Aren’t UCTs just those theories which are propounded by the contemporary conspiracy theorists whom Keeley attacks? No. UCTs are a set of theories that occupy a region of ‘logical space’ that Keeley has roughly located through his characterisation of them. Contemporary conspiracy theorists may have persistently advocated the less warrantable of these, but this tells us nothing about the warrant that more acceptable but unfancied conspiracy theories, which Keeley counts as UCTs, might actually deserve. Keeley has undertaken to say something important about the warrantability of UCTs as a class, but he has failed to do this. It might also be thought that I am being too strict about what Keeley means by the term ‘theory’. I have accused conspiracy theorists of remaining committed to research programmes even when these exhibit clear signs of degeneration. If Keeley was to be interpreted as using the word ‘theory’ to mean developmental series of views changing over time, then a theory in his sense would be something akin to Lakatos’ conception of a research programme, and Keeley could be understood as joining me in claiming that conspiracy theories are unwarranted when they form the core of degenerating research programmes. However, Keeley appears to rule out this charitable interpretation when he tells us that “... a conspiracy theory deserves the appellation ‘theory’, because it proffers an explanation of the event in question.” (Keeley 1999, p. 116). Theories are understood by Keeley as single explanations and not as developmental sequences of explanations. Keeley greatly overstates the strength of his case against UCTs, and as I have shown, he conflates it with a case he develops against the explanatory stratagems favoured by some contemporary conspiracy theorists. Neither of Keeley’s arguments against UCTs establish that UCTs are significantly less epistemically reputable than other social theories. He does establish that, sometimes, the acceptance of UCTs involves a degree of undermining of beliefs which had been warranted by those who are alleged to be participating in the conspiracy. But it is not even clear that this problem is something that ought to count particularly against conspiracy theories. Consider the situation I am in when I decide whether or not to accept a non-conspiratorial Marxist social theory which involves the view that the beliefs of most, or all, of those whose testimony I have hitherto relied on have been ideologically distorted. If I am to accept such a theory then I will undermine the warrant for many of my current beliefs. The problem of beliefs being undermined when theories are accepted is a problem that is shared by some conspiratorial and some non-conspiratorial social theories. 4. Conspiracy Theorists Instead of attempting to home in on the epistemic flaws of a significant class of conspiracy theories, as Keeley has, I will focus my attention on the cognitive failures of a significant class of conspiracy theorists  those conspiracy theorists who continue to hold on to conspiracy theories even when these take on the appearance of forming the core of degenerating research programmes. If we can identify a consistent form of cognitive failure amongst such conspiracy theorists, then we can go much of the way to justifying the attitude of intellectuals who dismiss conspiracy theories out of hand. The intellectuals can be shown to be entitled to assume (perhaps implicitly) that, like most conspiracy theorists, the conspiracy theorist being ignored is likely to be the proponent of a degenerating research programme, and the continued advocacy of such a research programme is likely to be the result of cognitive failure on the part of the conspiracy theorist. There may be many reasons why individual conspiracy theorists remain committed to their favoured conspiracy theories even when these exhibit clear signs of degeneration. However, if we are to justify the attitude of those who dismiss conspiracy theories on the grounds that these are propounded by conspiracy theorists we need to identify a factor which is present in the overwhelming majority of cases of such conspiracy theorising. The factor which I have identified as being common to the thinking of conspiracy theorists who hold on to degenerating research programmes, is that they commit what social psychologists call the ‘fundamental attribution error’. This is a form of cognitive error which is endemic to human thinking and which leads to a variety of unfortunate consequences (Nisbett and Ross 1980, 1991; Ross and Anderson 1982). Social psychologists studying our interpretation of the behaviour of others make a primary distinction between situational and dispositional explanations of behaviour. When I ask you to explain what caused Manfred’s motoring accident, you could provide a dispositional explanation by citing what you take to be features of Manfred’s personality. For example, you could tell me that Manfred is (disposed to being) careless. Alternatively you could appeal to relevant features of the situation that Manfred was in to explain the occurrence of the accident. You would be doing this if you told me that the accident was caused by the difficult driving conditions presented by the wet road that Manfred’s car was on. Of course there is nothing stopping you from appealing to both sorts of explanation when seeking to account for the events in question. Typically though, people presenting causal explanations single out one explanatory factor as ‘the cause’. According to many social psychologists, humans systematically make the error of severely overestimating the importance of dispositional factors, as well as the concomitant error of severely underestimating the importance of situational factors, when seeking to understand and explain the behaviour of others. This error is both very widespread and of a significant magnitude. Social psychologists have marshalled compelling evidence in favour of its existence. In one experiment which provides strong evidence in favour of the existence of the fundamental attribution error research subjects were given a set of essays and recorded speeches on controversial topics and invited to make inferences about the beliefs of the authors of the pieces. The subjects persistently inferred from the fact that a given author had written an essay or presented a speech in favour of (for example) the legalisation of marijuana that she was in fact in favour of the legalisation of marijuana, even when they were specifically informed that the authors and speakers were acting under instructions to argue for a particular side in a debate. A very strong situational factor was underapreciated by research subjects who exhibited a high degree of confidence in their own ability to explain the behaviour of the authors and speakers in dispositional terms (Jones and Harris 1967). In another experiment, the Darley-Batson experiment, a large group of research subjects who were students at the Princeton Theological Seminary were asked to prepare a presentation at a particular time and location, which was to be recorded. The seminarians were given directions to the location, which involved walking past a person slumped in a doorway coughing and groaning  a person who was obviously in need of assistance. The purpose of the experiment was to find out who amongst the research subjects would offer assistance. As it turned out, 63% of the research subjects who were unhurried offered some form of assistance. These research subjects were examined on a number of dispositional criteria, none of which were found to significantly correlate with their behaviour. However, one apparently trivial situational factor that was introduced was found to make an extremely significant difference. When seminarians who walked past the person in need of assistance were told beforehand that they were late and needed to hurry, the assistance rate dropped from 63% to 10%. This level of difference between hurried and unhurried subjects is a far cry from the expectations of psychologists and members of the public who were asked to predict the outcome of the experiment (Darley and Batson 1973). Not only did people not expect that this situational factor would be important, it appears that they had extreme difficulty in adjusting their beliefs in light of such results. Pietromonaco and Nisbett (1982) asked research subjects who had read about the Darley-Batson experiment to predict the rate at which people would help others in a similar situation. Despite the strong evidence of the Darley-Batson experiment in favour of the situational factor of being in a hurry being very explanatorily significant, the subjects made only insignificant adjustments to their predictions about the importance of being in a hurry and no adjustments at all to their assessment of the importance of dispositional factors. Of course this result is just what we should expect if the fundamental attribution error really is as generic as social psychologists suggest. As explanations, conspiracy theories are highly dispositional. When conspiracies occur it is because conspirators intend them to occur and act on their intentions. The conspiratorial dispositions of the conspirators play the role of the cause in a typical explanation which involves a conspiracy. In most cases the received view, the conventionally accepted non-conspiratorial alternative to a particular conspiracy theory, is a situational explanation. If you accept the received view that Elvis Presley’s funeral occurred because he died, as a result of a heart condition, you would explain his death by appealing to the situational factor of the state of his health. However if you believe that Elvis Presley faked his own death, you would account for the occurrence of the funeral by appealing to a dispositional explanation, because you presumably hold that he and his fellow conspirators went out of their way to deceive the public as to the fact that he remains alive. If you believe that the US military leadership are reluctant to discuss the ‘Roswell Incident’ because there is no such incident to discuss, you are basing your belief on a situational factor. By contrast, if you believe that the US military leadership are conspiring to keep the public unaware of contact with alien species, which occurred at Roswell, New Mexico, you would presumably explain the US military leadership’s persistent denials of knowledge of the incident by appealing to their disposition towards conspiratorial paternalistic behaviour. To be able to explain why conspiracy theorists remain committed to degenerating research programmes with conspiracy theories at their core, in the face of the degeneration of these research programmes, we have to appreciate what is typically involved in the giving up of a conspiracy theory. To give up a conspiracy theory in favour of a non-conspiratorial alternative is typically to abandon a dispositional explanation in favour of a situational explanation. But this involves overcoming the fundamental attribution error, which is to go against our cognitive instincts. This can be done, but it is psychologically difficult for us to do. Those who continue to believe in conspiracy theories, when it is intuitively clear to the majority that the time to abandon such theories is well overdue, may simply be people who are more in the grip of the fundamental attribution error than most. Of course the proponents of a conspiracy theory will not simply feel that a dispositional conspiracy theory is better than its non-conspiratorial situational alternative despite its degeneration; they will make efforts to rationalise their preference. One way they can do this is by appealing to the unifying power of conspiracy theories. Dispositional explanations, such as conspiratorial explanations, can appear to exhibit more unifying power than situational explanations, because dispositional explanations can relate the occurrence of events within the context of an intended plan. Because conspiracy theories typically involve highly elaborate plans they will usually exhibit great unificatory power. Situations, by contrast, can appear to be ‘one off’ events and explanations appealing to them can appear to lack unificatory power. But this contrast is fallacious. If you examine the circumstances of Elvis Presley’s natural death closely enough you will be able to relate it to other natural events, and with sufficient persistence you will be able to relate all of these within the scope of physics, thereby furnishing yourself with an explanation with more unificatory power than any dispositional explanation can provide. Some will object to the idea that there are cognitive errors at all, appealing to conceptual arguments that aim to rule out the possibility that we could be systematically irrational. Stein (1997) contains a discussion of such conceptual arguments, focussing particularly on one form of these, arguments for the conclusion that claims of human irrationality are self-undermining. Others will attempt to base objections on evolutionary theory and ask how it is that we could possibly have evolved to make systematic cognitive errors. Unfortunately there is good evidence to show that we are disposed to make a variety of such systematic cognitive errors (Kahneman et. al. 1982.). For example, it is well established that we frequently make systematic errors about the extent to which small samples are representative of larger populations (Tversky and Kahneman 1982). It may seem hard to believe that humans have evolved to make a systematic error about something as important to us as the interpretation of the behaviour of other people. Nevertheless this is possible. The fundamental attribution error primarily occurs when we make judgements about the behaviour of people outside contexts of familiarity (Nisbett and Ross 1991). For the most part, human evolution probably occurred when early humans were members of close-knit tribal groups, in which members would have been highly familiar with other members of the group and in which they were not often exposed to their behaviour outside of familiar contexts. It may also be that a heightened awareness of dispositional factors in the understanding of the behaviour of others was to our evolutionary advantage, even if this came at a cost to our understanding of the importance of situational factors. If another person, with whom I am in close contact, is disposed to conspire against me then it is very important that I am aware of this. If I commit the error of mistaking their hostile dispositions for a situational factor then I potentially expose myself to much danger by continuing to interact with them. In general it may be better to ‘err on the side of caution’, and mistake situational factors for dispositional factors, rather than take the chance of misreading people’s possibly hostile dispositions. 5. Conspiracy Theorising I have suggested that the fundamental attribution error may have survived in the human population because in most cases it was not particularly harmful and because the opposite error of overestimating situational factors to the exclusion of dispositional factors was potentially very harmful indeed. Perhaps this asymmetry should be taken into account when we consider our attitude towards the activity of contemporary conspiracy theorists. Most conspiracy theorists who manage to make the headlines these days produce theories that are harebrained and lacking in warrant. But few are actually harmful. Furthermore, there are several things that can be said in favour of conspiracy theorising. First, the conspiracy theorist challenges us to improve our social explanations. If a non-conspiratorial social explanation is better articulated as a result of the challenge of a conspiracy theory then that is all to the good. Second, the conspiracy theorist occasionally identifies a genuine conspiracy. Giving a thousand conspiracy theories some consideration is a small price for us to pay to have one actual nefarious conspiracy, such as the Watergate conspiracy, uncovered sooner rather than later. The prevalence of conspiracy theories confers a third benefit upon us, which is that it helps to maintain openness in society. Government agencies have a tendency to be less than forthcoming with information which might prove embarrassing to them, but which the public would prefer to have made available. The information gathering activities of conspiracy theorists can help to prevent such secretiveness. The conspiracy theorist may be a victim of cognitive error, but it is perhaps to our advantage that they remain in error. Although we would not wish to fall victim to the fundamental attribution error, it can sometimes be to our advantage that others do. Perhaps we should thank the conspiracy theorist for remaining vigilant on our behalf. They may still be out there. Thanks to an anonymous referee, an audience at Charles Sturt University, John Bigelow, John Campbell, Dean Cocking, Neil Thomason, Robert Young and particularly Jakob Hohwy, for helpful comments on earlier versions of this paper. Steve Clarke is a research fellow in the Centre for Applied Philosophy and Public Ethics and a lecturer in the School of Humanities and Social Sciences at Charles Sturt University. He is the author of a book, Metaphysics and the Disunity of Scientific Knowledge (Aldershot: Ashgate, 1998), as well as a number of articles on topics in applied ethics, metaphysics and the philosophy of science. These have appeared in such journals as the American Philosophical Quarterly and the Australasian Journal of Philosophy. References Bernstein, C. and Woodward, B. 1974. All the President’s Men. New York: Simon and Schuster. Biagioli, M. 1999. The Science Studies Reader. New York: Routledge. Brewer-Giorgio, G. 1988. Is Elvis Alive? New York: Tudor. Campbell, J. 1999. “Can Philosophical Accounts of Altruism Accommodate Experimental Data on Helping Behaviour?”. Australasian Journal of Philosophy 77: 26-45. Canovan, M. 1981. Populism. New York: Harcourt Brace Jovanovich. Coady, C.A.J. 1992. Testimony. Oxford: Oxford University Press. Cornish, K. 1997. The Jew of Linz —Wittgenstein, Hitler and their Secret Battle for the Mind. London: Random House. Darley, J. M. and Batson, C.D. 1973. “From Jerusalem to Jericho: A Study of Situational and Dispositional Variables in Helping Behaviour”. Journal of Personality and Social Psychology 27: 100-19. Doris, J. M. 1998. “Persons, Situations and Virtue Ethics”. Nous 32: 504-30. Doris, J. M. 2002. Lack of Character, Personality and Moral Behaviour. Cambridge: Cambridge University Press. Flanagan, O. 1991. Varieties of Moral Personality. Cambridge MA: Harvard University Press. Hofstadter, R. 1955. The Age of Reform. New York: Alfred A. Knopf. Hofstadter, R. 1971. The Paranoid Style in American Politics, and Other Essays. London: Jonathan Cape. Hume, D. 1988. An Enquiry Concerning Human Understanding, edited by A. Flew. La Salle: Open Court. Jones, E.E. and Harris, V.A. 1967. “The Attribution of Attitudes”. Journal of Experimental Social Psychology 3: 1-24. Kahneman, D., Slovic, P., and Tversky, A. 1982. Judgement Under Uncertainty: Heuristics and Biases. Cambridge: Cambridge University Press. Keeley, B. 1999. “Of Conspiracy Theories”. The Journal of Philosophy 96: 109-126. Lakatos, I. 1970. “Falsification and the Methodology of Scientific Research Programmes”. In Criticism and the Growth of Knowledge, edited by I. Lakatos and A. Musgrave, 91-195. Cambridge: Cambridge University Press. Lazer, H. 1976. “British Populism: The Labour Party and the Common Market Parliamentary Debate”. Political Science Quarterly 91: 259-277. Lipton, P. 1998. “The Epistemology of Testimony”. Studies in the History and Philosophy of Science 29: 1-31. Nisbett, R. and Ross, L. 1980. Human Inference: Strategies and Shortcomings of Social Judgement. Englewood Cliffs: Prentice Hall. Nisbett, R. and Ross, L. 1991. The Person and the Situation. New York: McGraw-Hill. Pietromonaco, P. and Nisbett, R. 1982. “Swimming Upstream Against the Fundamental Attribution Error: Subjects’ Weak Generalizations from the Darley and Batson Study”. Social Behaviour and Personality 10: 1-4. Pigden, C. 1995. “Popper Revisited, or What is Wrong With Conspiracy Theories?”. Philosophy of the Social Sciences 25: 3-34. Popper, K. R. 1966. The Open Society and Its Enemies, Vol. 2: The High Tide of Prophecy: Hegel Marx and the Aftermath. 5th Edition. London: Routledge. Ross, L. and Anderson, C. A. 1982. “Shortcomings in the Attribution Process: On the Origins and Maintenance of Erroneous Social Assessments”. In Judgement Under Uncertainty: Heuristics and Biases, edited by D. Kahneman, P. Slovic and A. Tversky, 129-52. Cambridge: Cambridge University Press. Shermer, M. 1997. Why People Believe Weird Things. New York: W.H. Freeman. Stein, E. 1997. “Can We Be Justified in Believing That Humans Are Irrational?”. Philosophy and Phenomenological Research 57: 545-65. Tversky, A. and Kahneman, D. 1982. “Belief in the Law of Small Numbers”. In Judgement Under Uncertainty: Heuristics and Biases, edited by D. Kahneman, P. Slovic and A. Tversky, 23-31. Cambridge: Cambridge University Press. Endnotes 32