Bridging the great divide: Conspiracy theory research for the 21st century

Abstract

This article starts from the observation that research on conspiracy theories is currently thriving, but that it is also fragmented. In particular there is an increasing divide between disciplines with culturalist and qualitative approaches, such as history, cultural studies and ethnology, and disciplines with quantitative and empirical approaches, such as psychology and political science. The article argues that this ‘great divide’ has to be bridged for research to arrive at a genuine understanding of conspiracy theories. As a first step in such a bridge-building process, the article engages, from the vantage point of (American) cultural studies, with research done on the other side of the divide, namely in psychology and political science. It summarizes the work done in these fields and evaluates it critically, concentrating on methodology; the assumptions about the dysfunctionality of conspiracy beliefs; the circularity of some arguments; the focus on individual rather than social and collective aspects; the lack of engagement with the definition of the phenomenon; and the neglect of cultural and historical difference. The article ends with recommendations for future collaborative research projects.

Introduction

Research in conspiracy theories is currently thriving. Scholars in fields as diverse as history and analytical philosophy, anthropology and media studies, literary studies and political science, or psychology and cultural studies have produced an impressive body of work and greatly enhanced our understanding of the phenomenon. This research now spans half a century, but in the last two decades the pace and volume of publication have increased considerably. In their research on conspiracy theories, scholars often arrive at vastly different results and conclusions. While this is not problematic as such, it is a cause for concern because scholars seldom pay attention to work done in disciplines that proceed from different assumptions and employ different methods. In fact, there is currently in conspiracy theory research what we would like to call a ‘great divide’. This divide separates disciplines with culturalist and qualitative approaches, such as history, cultural studies, ethnology or religious studies, from disciplines with quantitative and empirical approaches, such as psychology and political science. While some scholars stay completely within the confines of their disciplines, some take into account what neighbouring disciplines have to say about the issue. However, to date no study engages seriously with work conducted on the other side of the great divide. At most, studies from the other side are mentioned in passing in the introductions to books or articles, and then quickly dismissed or simply by-passed. As a result, research in the field is increasingly fragmented.

In this article we undertake a first step to bridge the great divide and outline routes that future research programmes need to follow. Maybe paradoxically, at first sight, this initial step is a sustained critique of work on the other side. This critique, though, is not meant to burn the few feeble bridges over the divide that may still exist, but is intended to achieve exactly the opposite. After all, transdisciplinary dialogue can only truly begin when everybody involved is clear about the strengths but also the weaknesses of all disciplines involved, as only this enables scholars to meaningfully disagree, to find common ground, and to develop joint research projects.

In what follows we therefore critique work in psychology and political science from the specific perspective of (American) cultural studies. We begin by summarising the research done in these fields and then proceed to discuss it. Much of the research undertaken in these two fields has brought a real energy to the study of conspiracy theories, but it is also in danger of ignoring some of the basic insights of work carried out in textual studies. We highlight six potential areas of concern: the methodologies used in some of these studies; their assumptions about the dysfunctionality of conspiracy beliefs; the circularity of their argument; the focus on individual rather than social/collective aspects; the lack of engagement with the definition of the phenomenon; and the implications of cultural and historical difference. We then end by recommending future steps in overcoming the great divide.

Research in psychology

Despite the foundational work of Hofstadter's (1964) psychohistorical approach to conspiracy theories, the discipline of psychology was initially slow to investigate the topic, partly because it regarded them merely as a fringe curiosity. However, as researchers have come to recognise that the phenomenon is widespread and has potentially serious consequences, there has been a significant flourishing of empirical studies in the last decade (for a more detailed discussion of the field, see Swami and Coles, 2010; Bost, 2015).

Although there is now an increasing recognition that conspiracy theories are a mainstream phenomenon, the focus of psychological work has often started from the assumption that conspiracy theories are illogical, unfalsifiable and riddled with contradictions (e.g. Wood, Douglas and Sutton 2012). Early researchers in the field tended to take for granted that conspiracy theories are held by distinctive kinds of people with identifiable and flawed psychological characteristics: conspiracy theorists. Instead of investigating the structural, historical and cultural features of conspiracy theories, much work in psychology has sought to profile believers, and enumerate the personality and cognitive factors involved in what is usually termed – in a phrase that evokes an unwarranted level of diagnostic precision – ‘conspiracy ideation’. Following Hofstadter (albeit not always directly), some researchers have investigated the supposed link between conspiracy thinking and forms of psychopathology: if not full-blown paranoia as such, then schizotypy (cf. Darwin, Neave and Homes, 2011; Barron et al., 2015), or related traits of a ‘damaged’ psyche such as low levels of trust, suspiciousness, obsession with hidden motives, heightened sensitivity to threat, feelings of alienation, cynicism, uncertainty, powerlessness, anxiety and loss of control (e.g. Abalakina-Paap et al., 1999; Goertzel 1994; Swami, Chamorro-Pemuzic and Furnham, 2010; Grzesiak-Feldman, 2013; Jolley, 2013). Although some researchers (e.g. Swami, Chamorro-Pemuzic and Furnham, 2010) have found some correlations between conspiracy thinking and elements of the so-called Big Five personality differences (e.g. a negative relation to Agreeableness, connected with asuspicion of others) and have suggested that individual differences in ‘conspiracy ideation’ are stable over time (Imhoff and Bruder, 2014), others have found that the conspiracy theorist does not have a distinctive personality (Brotherton, French, and Pickering, 2013) and that circumstantial factors are needed to trigger the personality traits.

Instead of focusing on conspiracist personality as such, some psychologists have investigated the heuristics, cognitive biases and other forms of supposedly faulty reasoning involved in ‘conspiracy ideation’, such as mistaken causal attribution and an overreliance on intentionality (‘fundamental attribution error’), a faulty estimation of probability (‘conjunction error’), and a ‘stickiness’ to beliefs in the face of contrary evidence (‘confirmation bias’) (see Brotherton and French, 2014, 2015). Conspiracy theories have also been compared to other forms of ‘flawed’ ways of seeing the world, such as paranormal, New Age and fundamentalist outlooks (e.g. Whitson, Galinsky and Kay, 2015). Indeed, Goertzel (1994), one of the first psychological studies and still influential, argued that conspiracy beliefs are ‘monological’, i.e. they serve as a complete worldview, such that people who believe in one conspiracy theory, tend to believe in them all (see also Swami, Chamorro-Pemuzic and Furnham, 2010). However, this view has recently been challenged by other researchers (e.g. Sutton and Douglas, 2015), who find that sometimes conspiracy beliefs are topic-specific.

In terms of methodology, most psychology studies employ questionnaires that rank the respondent on a set scale of conspiracy belief, and then test out psychological and sociological variables that might be associated with high or low rates. There has been a proliferation of different scales, such as the Belief in Conspiracy Theories Inventory (Swami, Chamorro-Pemuzic and Furnham, 2010), the Generic Conspiracist Beliefs Scale (Brotherton, French and Pickering, 2013) and the Conspiracy Mentality Scale (Imhoff and Bruder, 2014), but as yet no agreement on a single measure. Although most surveys measure belief in well-known conspiracy theories, some researchers include imaginary scenarios (e.g. Leman and Cinnirella, 2007). Other researchers in both political science and psychology have also begun to move beyond surveys to experimental manipulation of attitudes in their quest to identify the variables and mechanisms involved in ‘conspiracy ideation’, e.g. Whitson, Galinsky and Kay (2015) find that people who have been induced into experiencing a sense of emotional uncertainty or a loss of control are more likely to draw on conspiratorial interpretations of events. Many researchers are increasingly concerned with the harmful social and political effects of conspiracism, with findings that exposure to conspiracy theories makes it less likely, for example, for people to try to reduce their carbon footprint or have their children vaccinated (Jolley and Douglas, 2014a, 2014b). Others have conducted experiments to show that belief in potentially harmful conspiracy theories can be reduced with a task that increases analytic thinking (Swami et al., 2014).

Research in political science

Compared to the enormous body of psychological studies that have come out in the past two decades, there is still relatively little research in political science where scholars were much slower to engage with the issue. Undoubtedly a factor in this lack of interest was the discipline's primary focus on partisanship, ideology and issue positions during the 1950s and 1960s, and later the dominance of the rational choice paradigm, which held that opinions were rational as opposed to purely social-psychological. Conspiracy theories, cast by Hofstadter and most of his predecessors as irrational and unscientific, did not really fit either of these paradigms. However, the widespread conspiracy theories concerning Barack Obama appear to have finally motivated research in this field. In recent years, there has been a significant flourishing of empirical studies in political science, which, like recent studies in social psychology, also mainly rely on questionnaires to detect the factors that fuel belief in conspiracy theories, and to measure how widespread they are. All of these studies come to the conclusion that conspiracy theorising is not a fringe phenomenon but something done by a considerable number of Americans both in the past and the present. Most Americans, they find, endorse at least one conspiracy theory, while individual conspiracy theories are usually believed in by about one quarter or one third of the population (e.g. Oliver and Wood, 2014; Sunstein and Vermeule, 2009; Uscinski and Parent, 2014). Most studies find that standard demographic factors (race, class, gender, age, education etc.) are not particularly significant predictors of conspiracy belief.

The studies differ, however, with regard to what causes they identify for the belief in conspiracy theories. Sunstein and Vermeule hold that conspiracy theories are the result of ‘crippled epistemology’ (2009: 211), i.e. they arise when people lack information to understand what really happened. Uscinski and Parent conclude that ‘conspiracy theories are for losers’ (2014: 130), suggesting that they arise among groups who feel threatened, powerless and insecure. By contrast, Oliver and Wood come closest to a psychological explanation by arguing that conspiracy theories are caused by the predisposition to attribute events to the machinations of invisible forces, and to perceive the world as a Manichean struggle between good and evil (2014: 953). Finally, Sunstein and Vermeule argue that conspiracy theories are far more a matter of the political right than of the left, whereas Uscinski and Parent, and Oliver and Wood reject this claim. They both contend, though, that ideological assumptions and political convictions determine which conspiracy theories individuals believe in.

Since these studies all hold that conspiracy theories are a widespread phenomenon in American culture, they all more or less explicitly reject the correlation between conspiracy theories and personality disorders so prominent in psychology. Because they identify a cause, misinformation, that in theory could be remedied, Sunstein and Vermeule openly reflect on possible cures for conspiracism. They suggest, however, that once people have begun to believe in a conspiracy theory, it is almost impossible to convince them otherwise. This finding has been recently corroborated in further studies (Nyhan, Reifler and Ubel, 2013; Nyhan and Reifler, 2015). Uscinski and Parent as well as Oliver and Wood largely refrain from offering remedies for conspiracism, because their findings suggest that conspiracy theories are such an integral part of American culture that they will not go away.

Convergence and contradiction in conspiracy theory theory

Some commentators (e.g. Bost, 2015) see this ferment of research activity in psychology and political science as laying the groundwork for a full understanding of the psychological traits, habits of mind, circumstantial triggers and demographic variables that underpin conspiracy theories. According to this optimistic account, there is an increasing convergence of empirical results, and instead of additional studies of particular factors, now the focus should be on putting all the piecemeal insights together into a broad synthesis. However, it can also seem that the explosion of interest in the topic from psychologists and political scientists is not leading toward a broad agreement on the nature, causal origins, psychological effects and social consequences of conspiracy belief, but instead is increasingly fragmented and fractious, and at times fundamentally misguided. There is no consensus, for example, on whether belief in conspiracy theories is monological or issue-specific; whether conspiracy beliefs are abnormal and irrational or mainstream and (at times) justifiable; whether conspiracy theories are on the rise or are constant; and whether conspiracy theories inevitably lead to social harm or whether they might have some individual or social utility. One explanation for these disagreements would be that earlier research is now being challenged by newer investigations that are more accurate (because of their sample size, greater precision in isolating variables etc.). Another explanation would be the more general anxiety that is currently besetting social science (and social psychology in particular), that the findings of many studies are exaggerated at best, and impossible to replicate at worst (Open Science Collaboration, 2015; Carey, 2015). While the problem of replicability is indeed a troubling one, we want to argue more precisely that the lack of convergence into an established consensus is the result of some deeper and more troubling contradictions that are built into much of the research in psychology and political science into conspiracy theories. We consider five areas of potential disagreement.

Methodology

As we have seen, many of the political science and psychology studies use survey questionnaires to produce a statistically representative picture of conspiracy belief, with a troubling proliferation of different scales to measure conspiracy disposition. One problem with the reliance on surveys is that it operates with a very crude notion of what it means to believe in a conspiracy theory, which is usually modelled as a continuum. For one thing, the category of ‘don’t know’ is ambiguous: it might denote active scepticism towards an ‘official’ version of events or an active rejection of a conspiracy theory, but it might mean nothing more than a non-committal sense of agnosticism in the face of the information overload surrounding contentious historical events. People are more likely to entertain as possibly true beliefs with which they are familiar and which they know are fairly widespread (Crocker et al. 1999). As Sunstein and Vermeule (2009) suggest in a discussion of what they call ‘conspiracy cascades’, in the era of opinion polls belief in conspiracy theory can become a self-fulfilling prophecy, because respondents are less likely to reject completely ideas that they know considerable numbers of other people take seriously. It is also perfectly possible that many respondents in these questionnaires who give positive replies are not die-hard believers in particular conspiracy scenarios, but are willing to assent to a proposition ‘as if’ it were true, partly because of a world-weary despair of not knowing what to think about complex phenomena such as global warming, and partly for the kitsch entertainment value of conspiracism as a hip, alternative stance (see Knight, 2001).

Dysfunctionality

As we have seen, much research in psychology and, to a lesser extent, political science has tended to start from the assumption that belief in conspiracy theories is a result of cognitive bias or a ‘crippled epistemology’ (Hardin, 2002; see also Groh, 1987; Robins and Post, 1997). One problem with this line of inquiry is that it is possible that loosely agreeing with a conspiracy theory (not least in response to a questionnaire) involves very little actual cognition: often people turn to conspiracy theories as a shared resource of ready-made stories and explanations precisely to avoid having to think in detail about troubling events. Entertaining belief in conspiracy theories is thus not so much a sign of individual cognitive dysfunction, as participation in a collective and iterative process of sharing stories, much like rumours and urban legends (for more on this point and a wider consideration of the problems of psychological approaches to conspiracy theories, see Byford, 2011). Conspiracy theories might not be the result of delusional thinking or an overly suspicious personality, but instead might fulfil the individual and collective need to make sense of troubling events (van Prooijen and Jostmann, 2013), often by blaming scapegoats, and in the process making the down-trodden believer feel part of an elite minority who have managed to pierce through the veil of obfuscation.

More problematic still is the assumption that belief in conspiracy theories is socially abnormal, if not positively bizarre, the reasons for which are to be found in the idiosyncrasies of individual psychopathology. However, as the more recent findings from opinion poll data indicate, belief in conspiracy theories is demographically mainstream and, as some social psychologists have begun to recognise, in tune with more ‘normal’ modes of cognition such as ‘magical thinking’ (cf. Oliver and Wood, 2014). It does not make much sense to analyse conspiracy theories as aberrational modes of thinking when, say, three quarters of the American public believe that there was a conspiracy behind the Kennedy assassination. Nor is it particularly fruitful to document the correlations between ‘conspiracy ideation’ and fundamental attribution bias or the propensity to find big causes to explain big effects, as the latter are shared by both believers and sceptics alike (cf. Leman and Cinnirella, 2007). Likewise, the focus on the individual cognitive differences of conspiracy theorists as a discrete class of people potentially comes unstuck in cases of conspiracy rumours that turn out to be true. If believers in a Watergate conspiracy or an official cover-up in the Hillsborough football stadium disaster had been included in some of the psychological studies of conspiracy ideation before those stories were confirmed as proven conspiracies, would they have manifested the same traits as those who believe in stories that have not been proven?

Even more troubling are cases where a belief in a conspiracy seems a perfectly justifiable – if not strictly rational or true – response to a known history of previous conspiratorial abuse. For example, Harriet Washington (2006) has documented the long history of the mainstream American medical community's dubious science, therapeutic neglect and outright abusive treatment of African Americans, that long predates and continues beyond the now well-known case of the Tuskegee syphilis trials. For Washington (as for other commentators, e.g. Turner, 1993; Knight, 2001), lumping all conspiracy theories together under the generic explanation of ‘black paranoia’, and ignoring the social and historical factors such as social marginalisation that inform such beliefs, can lead to psychopathologising what should more accurately be understood as collective, warranted fears. In short, a focus on the lack of trust in individuals diverts attention away from the much needed investigation of the untrustworthiness of actual medical institutions and practices.

Circularity

Although recent research in psychology and political science has made progress in specifying the personality traits, cognitive mechanisms and social characteristics involved in conspiracy thinking, in most cases these are correlations. Various cognitive biases might often be associated with conspiracy thinking, but in themselves do not cause it. There is still surprisingly little hard information about the causal mechanisms involved in conspiracism, or why certain people and not others turn to conspiracy theories in particular historical moments. Indeed, at times there is a troubling circularity in the findings. Are conspiracy theories the result of a ‘crippled epistemology’, or do they cause it? As far back as Hofstadter's pioneering analysis, there has been a tendency for explanations of conspiracy theories to beg as many questions as they answer. Hofstadter (1964) insisted that the recurrent outpouring of conspiracy rhetoric is a result of an upsurge in paranoia (albeit not in a strictly clinical sense), and many other researchers have followed his underlying assumption that conspiracy theories are a sign of delusional thinking, often with quasi-Freudian assumptions about the supposed psychosexual origins of the condition. But what explanatory force does the notion of the ‘paranoid style’ actually provide, if paranoia is a mental condition marked out by beliefs that are usually highly idiosyncratic, unlike conspiracy theories that draw on communal narratives, tropes and images? Paranoia for Hofstadter and those following in his footsteps (e.g. Pipes, 1997; Robins and Post, 1997) is offered as the explanation for the turn to seeing conspiracies everywhere, yet ‘paranoia’ in these discussions seems to mean little more than the tendency to see the world through the lens of conspiracy theories. Likewise, the fact that people who score highly on a particular scale measuring conspiracy disposition might tell us little more than the fact that ‘conspiracy theorists’ are people who tend to view the world in terms of conspiracy theories. What makes conspiracy theories a distinctive way of explaining the world is not to be found solely in the psychology of individual believers, but in the shared structural elements of the conspiracy theories themselves. Researchers therefore need to investigate the cultural work conspiracy theories perform in different places and times, and the social relations that conspiracism both enables and curtails.

Definitions

The confusion in the empirical research literature on the nature and causal effect of conspiracy ideation is in part caused by the lack of agreement on a definition of the very term ‘conspiracy theory’. It is often treated as if it is a timeless and stable concept, whereas it only became popularised in the 1960s and 1970s, and carries considerable ideological baggage. Some commentators lament that the lack of consensus and precision in defining the phenomenon hampers progress in empirical research, a problem that supposedly could be avoided if researchers could agree on its basic features. While there is a broad convergence on the idea that conspiracy theories see a small group of nefarious conspirators working in secret to bring about a change in the course of history, this working definition leaves out the important component that at least in the contemporary period conspiracy theories often – perhaps inevitably – involve ‘stigmatized knowledge’ (Barkun, 2013; see also his contribution to the present volume). However, this is not merely an additional criterion that helps to make the definition incrementally more precise. Instead it implies that what makes a conspiracy theory distinguishable from other interpretations of current events is not solely an effect of the individual psychology of the believer or the structure of the belief, but is a product of the particular content and social function of the story that is told in opposition to received wisdom.

Furthermore, as Bratich (2008) and other cultural studies scholars have argued, the very label ‘conspiracy theory’ is not simply a neutral, objective description of a particular mode of oppositional belief (whose psychological and demographic factors could in theory be identified precisely), but a pejorative dismissal of other people's worldview. No one willingly admits to being a conspiracy theorist, because the term itself is in effect an insult. This means, however, that some of the results presented in Uscinski and Parent (2014) are flawed because they draw on the frequency of Internet posts containing the term ‘conspiracy theory’ to measure how widespread such theories are. But since conspiracy theorists rarely ever use the term to refer to their ideas, they fly under the radar of this study. Accordingly, the aim of producing empirical, value-neutral research on the phenomenon of ‘conspiracy theory’ is misguided, because the term itself is not value-neutral.

Quantitative research into the personality traits, cognitive habits and demographic factors that lead to conspiracy thinking is skating on thin ice without a historical and sociological understanding of how the very object of study came into being. That history has yet to be fully told, but we already know enough (Bratich, 2008; Fenster, 2008) to be cautious about producing a definition of the phenomenon that all researchers can take as their starting point. If the term ‘conspiracy theory’ has its origins in demarcating a particular set of beliefs as illegitimate and undesirable, then we need to be more attuned to the politics of delegitimisation in different historical moments.

Cultural and historical contingency

Researchers in political science and psychology are usually quick to emphasise that, despite the comparatively small sample size involved in their particular study, they are able to produce statistically meaningful generalisations about conspiracy theories and the people who believe in them. This research (especially in psychology) is, however, prone to the wider problem of what has come to be known as the ‘WEIRD’ phenomenon, i.e. experimental studies and (to a lesser extent) opinion polls are often based on a very selective sample drawn from a population that is Western, Educated, Industrialised, Rich and Democratic – and often taking part in surveys and experiments for college credit.

This is particularly problematic for studies that seek to explore the correlation between ‘conspiracy ideation’ and seemingly fundamental human cognitive traits. Even after applying statistical techniques that try to extract representative data from the selective sample, the suspicion is that the findings are indeed skewed towards a Western – and, more often than not, specifically American – way of thinking. Very little quantitative research to date has been done on conspiracy theories in other parts of the world (one notable exception is Swami, 2012; and there are some beginnings of research by anthropologists, e.g. West and Sanders, 2003). Generalising the results of partial studies conducted in the United States is problematic because conspiracy theories appear to have a very different social function and political status in other parts of the world and political regimes. In many present-day Middle Eastern countries, for example, talk of conspiracy theories is not abnormal and aberrational but very much part of how both elites and ordinary people think about current affairs (Gray, 2010; Rabo, 2014). Although the recent move to exploring the psychological needs that conspiracy theories fulfil for individual believers is welcome, much of this research fails to acknowledge that the needs might vary in different cultural and political situations.

We might also be suspicious that even statistically robust studies of conspiracy theories are in fact quite specific to the contemporary United States, and their conclusions do not necessarily apply to other Western nations, let alone other non-Western cultures and political regimes. The political scientists and social psychologists are often quite aware of the limitations of their studies and are very happy to encourage future research that explores other cultures, but their papers often rarely make even the most perfunctory nod to cultural specificity. In some cases they are expressly trying to determine cognitive mechanisms and political factors that transcend individual differences, but in general there is a disturbing cultural imperialism at work that scales up the partial findings into universal claims. Of course, it might turn out that the findings based on these ‘WEIRD’ studies are indeed applicable to other cultures and political regimes, but that cannot be assumed in advance. There is also a tendency to treat conspiracy theories as interchangeable, focusing on ‘typical’ examples drawn from a small roster of current favourites, which has the effect of bracketing off any engagement with the specific content of particular theories and the particular political, historical and cultural context in which they arise and gain meaning.

In addition to focusing primarily on Western varieties of conspiracy thinking, quantitative researchers also tend to concentrate solely on the present day (with a few notable exceptions, e.g. Uscinski and Parent, 2014). This lack of historical comparison is an inevitable consequence of the virtual impossibility of conducting experiments and opinion surveys retrospectively, in the absence of meaningful time series data. Yet, unlike the tendency to geographical universalisation that could be corrected by expanding the scope of research to non-Western populations, the default focus on the present is harder to rectify. Even within its own terms, much of the psychological research assumes that an individual's belief in conspiracy theories does not have a significant aetiological history, it being merely a result of socialisation in childhood. Admittedly some social psychology research has begun to consider in broad brush terms some of the circumstantial triggers and life history episodes that might contribute to conspiracy thinking, but so far there has been little attention in empirical and experimental research to the role, for example, that moments of conversion play in individual conspiracy belief.

There has been even less attention in the psychology and political science literature to the institutional, ideological and cultural factors that make conspiracy theories more attractive in particular historical moments. Do conspiracy theories flourish in times of social and economic stress, or are they constant? Have conspiracy theories always been stigmatised, or is this a comparatively recent phenomenon? Do conspiracy theories only emerge with the Renaissance or the Enlightenment, or can they be traced back to Ancient Greece and Rome? What role do conspiracy theories play in the wider history of ideas? Are we currently living in a golden age of paranoia, and, if so, why? Has the internet increased the popularity of conspiracy theories, or has it merely made them more visible? Although there is not yet a consensus on any of these issues, a considerable body of research already exists that engages with the changing historical character and function of conspiracy theories, albeit primarily in the US and Western Europe. Scholars such as Bailyn (1967), Wood (1982), and Butter (2014) have argued that throughout much of American history conspiracy theorising was not stigmatised but, on the contrary, considered a perfectly rational way of making sense of the world, one that was employed not only by most members of society but especially so by its elites. Olmsted (2008) argues that conspiracy theories have grown in popularity and significance since the beginning of the twentieth century, largely as an intelligible counter-reaction to the increasing size, secrecy and suspicious behaviour of the American state itself. Other historians (e.g. Zwierlein and Graaf 2013) have focused not on the rise of populist conspiracy fears about the government in the twentieth century, but on the way that European states in the early modern period developed repressive security measures in reaction to fears – often unfounded – about perceived enemies among their populations, which in turn led to increased popular suspicion of the authorities. In a different vein, intellectual and cultural historians (e.g. Melley, 2012; Horn, 2013; Boltanski, 2014) have studied the connections between conspiratorial narratives and ideological structures of feeling¹ that have emerged in different historical moments and cultures. What this large body of diverse work (the full extent of which we can only gesture at here) has in common is the challenge it presents to the generalising conclusions in much of the research in political science and psychology into conspiracy theories.

Recommendations, or: bridging the great divide

So far in this article we have been highlighting some of the methodological and conceptual problems involved in the recent research in political science and psychology on conspiracy theories. This should not be taken to imply that work in history, literature and cultural studies is without its own shortcomings, the most serious of which is the lack of systematic investigation or generalizable findings.² Although much of the recent quantitative research is imaginative in its design and impressive in its statistical rigour, it suffers from a lack of engagement with or knowledge of the large body of work done in other fields. For example, neither Oliver and Wood (2014), one of the most influential and comprehensive recent studies of conspiracy theories in political science, nor the edited collection The Psychology of Conspiracy (Bilewicz, Cichocka and Soral, 2015) includes any references to the work of Fenster (2008), Barkun (2003) or Olmsted (2008), which each provide lengthy and well researched examinations of the phenomenon.

At most, this body of recent research tends to refer to Hofstadter (1964), still the default starting point for both academic studies and journalism. The lack of engagement operates both ways, though: it is equally true that very few works on conspiracy theories by historians and cultural scholars make extensive reference to the recent research in psychology and political science. Yet it is important to note that many of the issues and hypotheses currently being explored in the empirical literature have already received discussion – sometimes quite extensive – in the textual studies of conspiracy theories. Indeed, the wave of sociology, cultural studies and literary studies books that began to emerge in the late 1990s expressly developed a critique of Hofstadter's pathologising approach, and provided substantial evidence that conspiracy theories have been and continue to be mainstream and are thus not necessarily a sign of delusional thinking – even if they are now, in the United States at least, more likely to be stigmatised that in previous historical eras (Butter, 2014). In short, we need to read across the ‘great divide’ to ensure that we are not endlessly reinventing the wheel. We also need to become more familiar with studies conducted in languages other than English, and this bilingual special issue is a welcome addition to this task.

Where should the study of conspiracy theories go from here? There are many exciting and important research projects that are being conducted within particular disciplinary traditions, methods and debates, and these additional contributions to their respective fields are still much needed. However, the study of conspiracy theories has now reached a moment, we would argue, that requires a comparative and transdisciplinary approach to fully comprehend the phenomenon in all its complex historical, regional and cultural variation. Some recent intriguing work in social psychology (e.g. Franks et al., 2015) has begun to introduce qualitative methods (semi-structured interviews) into the mix, with results that suggest that some of the conclusions drawn from the more familiar quantitative work in the field will need revising. While this development is welcome, we would encourage an even bolder use of mixed methods. Franks acknowledges that conducting interviews with conspiracy researchers as research subjects has the inherent difficulty that many are wary of experts and authorities in general, and are especially suspicious of psychologists who wish to categorise them as ‘conspiracy theorists’. However, historical archives provide us with a large body of rich qualitative data in the form of factual and fictional writings that can be mined without the need for negotiating the complexities of institutional ethical clearance. Conversely, it would be useful to bring some of the quantitative techniques of textual data mining to the analysis of conspiracy texts, albeit with more nuance than was practical in studies such as Uscinski and Parent (2014). Likewise disciplines such as social anthropology – which have been oddly slow to consider conspiracy theories as a topic – could bring much needed rich ethnographic detail to the study of conspiracy theories ‘in the wild’ rather than in the lab. Innovative studies of the vectors of transmission of conspiracy theories (proposed by a team of medical anthropologists at Durham University), for example, need to be combined with the kind of forensic historical analysis made by Selvage (2015) of the ways that the AIDS conspiracy theory developed – a story that includes the conspiratorial revelation that the KGB were involved in spreading the rumour as part of a disinformation campaign.

What we are recommending is not a contest between disciplines to have the last word on conspiracy theories as an object of inquiry, but collaborative projects that are (a) aware of work carried out in other fields; (b) informed by different methodologies; and (c) able to make meaningful comparisons between different regions, political regimes and historical moments – or, at the very least, are aware of the limitations of any conclusions they might reach. The starting point would need to be the recognition that no matter what psychological traits are involved, conspiracy theories are essentially social constructs. At a basic level bridging the divide between quantitative and cultural approaches might lead to comparative surveys of conspiracy belief in Western and non-Western countries, or ethnographic fieldwork being carried out among different conspiracy communities. Future research might also involve, for instance, designing opinion polls and experiments to test out the insights generated by analytical textual studies. Conversely, conclusions drawn from empirical research (e.g. that conspiracy belief is associated with a sense of powerlessness) could be tested out against the large body of historical materials to be found in different cultural traditions. A full-blown collaborative research project would seek to examine the history, culture, psychology, lived experience and politics of a particular conspiracy theory.

The institutional framework for these kinds of transnational and transdisciplinary projects will be enabled by COMPACT: Comparative Analysis of Conspiracy Theories, a four-year EU-funded COST Action running from 2016–2020, led by the present authors. The aim of this project is (1) to make it easier for scholars to bridge the ‘great divide’; (2) to lay the groundwork for future collaborative research projects into conspiracy theories that brings together textual and empirical approaches, spanning different cultures and historical moments; and (3) to team up with EU political administrators, NGOs, journalists and others to consider the social and political consequences of conspiracy thinking.