The illusion created by the media

Judging by the contents of the media, we are confronted with a wave or explosion of conspiracy theories and believers in them. Is there a single monthly magazine that has not published an article on these famous conspiracies and their network of supporters? Is there a daily newspaper that has not taken the pretext of a recent poll to discuss conspiracy theorists? We should note that, according to a CSA^{Footnote 1} poll for Atlantico (in January 2015), belief in the existence of a conspiracy following, for example, the Charlie Hebdo and Porte de Vincennes attacks, seems to be greater among young people, working-class people and Front National sympathisers (see also the poll by Opinion Way for the UEJF,^{Footnote 2} conducted in February 2015 with similar results).

A priori, we are confronted with a wave of irrationality. But what seems irrational and illogical in a framework of examination strongly marked by the Cartesian rationalism and mathematical proof taught in schools can perhaps acquire ‘rationality’ and its own logic in a different framework (Reference DelouvéeDelouvée, 2011; Reference Kalampalikis and HaasKalampalikis and Haas, 2008). In fact, if individuals have ‘good reasons to believe’ (according to the definition used by Reference BoudonBoudon, 2002; see also Renard in this issue), we can question the ‘reasons for the belief’ in collective beliefs in general and in conspiracy theories in particular by using the theoretical framework of social thinking (Reference GuimelliGuimelli, 1999; Reference Rouquette and MoscoviciRouquette, 1973, Reference Rouquette2009). We will use such a framework in discussing the questions in this article.

However, if this increase in the recourse to conspiracy theories remains debatable – no study to date has shown an effect of time on recourse to such theories – we must note that the number of scientific studies has increased recently (see in particular Reference Brotherton and FrenchBrotherton and French, 2015; Reference Dieguez, Wagner-Egger and GauvritDieguez et al., 2015; Reference Douglas and SuttonDouglas and Sutton, 2015; etc.).

One of the classic recurring questions addressed by these studies is that of the reasons why people believe in conspiracy theories and, furthermore, the question of how to counter them. Studies that examine the age, gender and level of education of believers in them all conclude that there is no link between these factors and adherence to conspiracy theories. If any link exists between belief in the paranormal or a sense of personal powerlessness and belief in conspiracy theories, it is very limited. The study by Reference Raposo, Pantazi and KleinRaposo et al. (2015: 155) is a good example of this type of investigation into adherence:

Adherence to a belief or belief in adherence?

A scientific colloquium on collective beliefs or a sociocultural event discussing the question of rumours or conspiracy theories would be a good setting for the propagation of particular rumours or conspiracy theories. For, what do we do when we discuss and analyse the human functions at work in the spreading of rumours or conspiracy theories? We must perforce reproduce the theories, describe them and explain them and in so doing we transmit them and spread them more widely.

Belief in conspiracy theories must be viewed in the form of a continuum. It is not a case of conspiracy theorists being found on one side and those who are not conspiracy theorists on the other. The more a person believes in one particular theory, the more that person will tend to believe in other conspiracy theories. The similarities between theories are very strong. Reference Moscovici, Graumann and MoscoviciMoscovici (1987) talks of a ‘conspiracy mentality’ to define this way of thinking. Reference Swami, Chamorro-Premuzic and FurnhamSwami et al. (2010) term it ‘conspiracist ideation’.

Since then, different scales for this have been developed. Reference GoertzelGoertzel (1994), Reference Abalakina-Paap, Stephan and CraigAbalakina-Paap et al. (1999) and Reference Swami, Chamorro-Premuzic and FurnhamSwami et al. (2010) each suggest a scale based on responses to a list of specific conspiracy theories (assassination of J F Kennedy, deliberate spread of AIDS, the moon landing, the death of Princess Diana and so on) by asking participating individuals to position themselves on the scale and indicate the degree to which they agree with each of the theories on the list. These scales, termed specific, are problematic because of the historical and local character of some of the theories listed. To avoid this problem, Reference Brotherton, French and PickeringBrotherton et al. (2013) and Reference Bruder, Haffke and NeaveBruder et al. (2013) propose scales that they term ‘generic’. Participants were asked to indicate their position on, respectively, 15 and 5 more general statements such as ‘I think that general public is not informed about a lot of very important things that happen in the world’ and ‘I think that events which don’t have any apparent link are often the result of secret activities’. More recently, Reference Lantian, Muller and NurraLantian et al. (2016) have proposed a scale based on a single statement: ‘I think that the official version of events given by the authorities often conceals the truth’. The instructions presented with these scales seem to play a very significant role in the studies.

The aim of this article is not to discuss conspiracy theorists or conspirators as such or the question of the real extent of belief in conspiracy theories. Recognising that those who believe in this type of theory will be more inclined than those who do not believe in them to invest time, money and energy in spreading them, justifying them and proselytising for them in a certain way (Reference BronnerBronner, 2003, Reference Bronner2006), it is a question of shifting the focus and examining the respective roles of knowledge of, adherence to and transmission of collective beliefs. The focus therefore should be on those who spread and transmit these much discussed conspiracy theories. There is a difference between those who create these theories and those who propagate them. We intend to examine the latter group.

Why do we believe in these ‘false’ theories?

Social science and humanities researchers have established some answers to this question. The most trivial answer, and that most frequently voiced, is that some of these theories prove to be true (Reference KeeleyKeeley, 1999). Think of the Watergate scandal or of a more recent example, the Wikileaks scandal, and we can see proof that some conspiracies really do exist. Does this justify a routine blaming of conspiracies as the cause of historical events? The difficulty in countering some of the arguments put forward by conspiracy theorists is therefore seen by some as a justification for these theories. This is essentially a question of rhetoric (see Reference Danblon and NicolasDanblon and Nicolas, 2010, Reference Nicolas2014) but what is termed ‘confirmation bias’ also often assumes a primary role.

While the scientific method is classically based on hypothesis and proof, starting from a hypothesis that is then tested against empirical data to reach a conclusion, a conspiracy theorist starts from a conclusion and interprets all facts in terms of this conclusion. It is, therefore, impossible to contradict such reasoning from a stance of the scientific method: any argument against the theory will be re-interpreted. These two opposing strands of argument can never meet on common ground. Conspiracy theories, just like all naive theories about the world around us (see the social representation theory), also give an impression of simplified control over that world (Reference Newheiser, Farias and TauschNewheiser et al., 2011), an impression of control (Reference Imhoff and BruderImhoff and Bruder, 2014) which is however a fallacy. On this point, we could also refer to attribution theory and of course this list is not exhaustive. Lastly, before asking why we believe in conspiracy theories, perhaps it would be better to examine how far we believe in them. Or whether we believe in them at all.

Reference DurkheimDurkheim (1912) and Reference VeyneVeyne (1988) each analyse a specific case of belief, which indicates that it is possible to spread knowledge of and belief in something without necessarily or actually believing it.

From a very young age, children talk to and play with their favourite toy or security blanket as if it were a living being. Do they believe that, like Pinocchio, an inanimate object can come to life and respond? Do they hope for a response? If you watch them playing and talking to their toy, there is no doubt that they do. Reference DurkheimDurkheim (1912) explains that children do not really believe that the toy is alive. If the toy were to kick them, they would be the most surprised. They know perfectly well that the toy isn’t alive but, when they play with it, they temporarily think that it comes alive. However, they do not really believe it:

In his essay on constitutive imagination ‘Did the Greeks believe in their myths?’, Reference VeyneVeyne (1988) examines the status of belief and truth. In the manner of the ancient Greeks, one can hold contradictory beliefs as soon as one recognises that there can be multiple frameworks of truth. Therefore, myth must be envisaged as discourse. Belief in a myth exists outside of the alternatives of the true and the false. The question of veracity does not arise. The Greek myths were stories like those we tell children and of which we accept, temporarily, the chain of events and the characters. A captivating story plays an important role in explaining events.

Do we always believe our beliefs?

In Witchcraft, Oracles and Magic among the Azande (Reference Evans-Pritchard1937), Evans-Pritchard reveals that the Azande's beliefs in magic are eminently coherent and logically interlinked with each other. The apparent irrationality in this can be explained by the fact that the pragmatic explanation (the how) and the mystical explanation (the why) are not contradictory and do not necessitate any conflict in interpretation on the part of the Azande: an Azande^{Footnote 3} ‘does not see a witch charge a man, but an elephant. He does not see a witch push over sorcerer a granary, but termites gnawing away its support. He does not see a psychical flame igniting thatch, but an ordinary lighted bundle of straw. His of how events occur is as clear as our own’ (Reference Evans-PritchardEvans-Pritchard, 1937: 72).

Here, we are confronted with an example of social logic particular to certain beliefs. In order to best understand this social logic, we must call on the concept of social thinking. Social thinking is the term for so-called natural thinking about a social phenomenon and, at the same time, thinking that occurs through social factors (Reference Rouquette and MoscoviciRouquette, 1973, Reference Rouquette2009). This social thinking is part of daily life, the thinking that one hears expressed in cafés, reads about on social media or in the press, or in conversations that appear to be harmless or trivial. This is not scientific or rational thinking but it is thinking that follows its own logic, a system of logic based on the social interaction of the individuals who express it.

Whether we talk about and spread a particular rumour depends on who we are and to which groups we belong; we draw on particular stereotypes or we call on specific events in the collective memory of those groups. The same is true for how we talk about and pass on to others certain conspiracy theories and not others. In these cases, that which seems to us to be irrational in normal life now seems logical, according to a social logic, and that which normally seems wrong or biased no longer seems so. It makes sense.

In social psychology, beliefs can be defined by the interaction of three elements (see in particular Reference Apostolidis, Duveen and KalampalikisApostolidis et al., 2002):

• • doxa: my opinions, the things I know, the things I am familiar with;

• • faith: the forms of adherence, ‘I believe’ = I adhere to this theory, I agree with it;

• • ritual: the rules for action and communication.

Social beliefs combine these three elements: awareness (states of opinion, the climate of opinion regarding the doxa), adherence (the form of adherence to the belief) and propagation (rules for action and communication linked to ritual). What is the link between these three elements? It appears that it is possible to have awareness and propagation without adherence, but also awareness and adherence without propagation. A link between the three elements is therefore not essential. In the case of rumours and conspiracy theories, is it possible to have awareness and propagation without adherence?

Empirical examples

Rumour is the preferred expression of social thinking since it is the result of a collective process. Rumours involve both real participants and real situations from daily life and they also express, in a symbolic manner, the fears and hopes that characterise social groups and that cannot be formulated in intellectual, abstract language. Starting in 1945, the Allport brothers (Gordon Willard and Floyd Henry) were the first to undertake experimental research into the process of rumour (Reference Allport and PostmanAllport and Postman, 1945). With Lepkin, Floyd Henry Allport showed that the more a rumour (in this case during war) is believed, the more widely it is spread and accepted (Reference Allport and LepkinAllport and Lepkin, 1943). However, other authors have also shown that rumours can be spread by someone who does not necessarily believe the rumour but recounts it simply because the rumour is new, because the person thinks it could possibly be true, because recounting it can reduce tension or allow the person to be the centre of attention and so on. It is not because one believes the rumour that one recounts it and the reverse may also be true.

The factors influencing adherence affect the level of involvement (the more a rumour affects them directly, the more individuals will tend to believe it (Reference Allport and LepkinAllport and Lepkin, 1943; Reference RouquetteRouquette, 2009)) just as much as the plausibility and credibility (of the rumour; Reference ClémentClément, 2006) and its compatibility with the information expected by the individuals and their attitudes. This is the case for rumours about conspiracy theories (Reference GalamGalam, 2002). However, these authors largely assume that adherence is initially present when investigating the factors which affect the degree of that adherence.

To undertake our first study, we selected 12 rumours (e.g. ‘Michael Jackson isn’t dead’, ‘France is going to default on its debts’, ‘the moon landings were faked’, etc.) using several selection criteria: their subject (conspiracy theory, danger, private life), duration, scope (global, local, individual), period (recent, historical, timeless), credibility (highly plausible to highly implausible) and so forth. These assertions were represented as being ‘rumours’ or as ‘statements’ (variable labelling).

The study was conducted using an online questionnaire which consisted of six stages: knowledge of the rumours, belief in/adherence to these rumours, distraction task, recall task (the participants had to recall and describe as many rumours as possible), rating the rumour for discussion (e.g. Is it a good subject for conversation?) and provision of personal data (age, sex, knowledge of websites such as Hoaxbuster, etc.).

The results showed that the participants (n = 287; undergraduate students with a mean age of 20.7 years, standard deviation = 1.08) knew about more rumours than they adhered to. On average, they knew about 7.8/12 rumours but stated that they adhered to only 1.3/12. This shows that there is a significant difference between knowing about and believing in a rumour: an individual can know about a rumour but not necessarily believe it to be true. What is more, the labelling of items had no significant effect on knowledge of and recall of the rumour (whether each rumour was categorised as a ‘rumour’ or as a ‘statement’ did not result in a difference in the number recognised or recalled). The categorisation did, however, have a significant effect on adherence: participants believed more in statements than in rumours. Knowledge of the websites had a significant effect on the number of rumours known: participants who were acquainted with the website knew more about the rumours, which seems logical. Knowledge of the Hoaxbuster website also had a significant effect on adherence: participants who were familiar with the site believed less in rumours.

Finally, the majority of participants recalled more than six items. However, in this study, some of the items were complex (long phrases) and the majority were not well known. Given that the actual number of items recalled by the participants was greater than the expected number (cf. studies on short-term memory), this result raises doubt about the low number of rumours the participants reported being familiar with, since if the subjects did not know about the majority of the rumours, they should not have been able to recall as many as they did. There is a difference between ‘saying that one doesn’t know about’ and ‘not knowing about’.

A second study based on a list of 10 conspiracies/conspiracy theories was conducted in order to confirm the results of the first study. Using the categorisation of conspiracy theories established by Reference Brotherton, French and PickeringBrotherton et al. (2013; criminal conspiracy within governments, the cover-up of extraterrestrial landings, malevolent global conspiracies, personal well-being, control of information) and the scale of Reference Swami, Chamorro-Premuzic and FurnhamSwami et al. (2010), a list of 10 conspiracy theories was presented to a further group of 231 students (undergraduate level, mean age 20.6, standard deviation = 1.43). These participants had to read a group of items which were presented either as a simple statement or as a conspiracy theory: ‘The death of Princess Diana was not by accident but by assassination organised by members of the British royal family who did not like her’. The participants then had to reply to the same questions as in study 1.

Although the results are less definite than for the rumours, they are nevertheless encouraging and support the hypothesis that transmission and adherence do not play the same role. These participants were aware of more of these items than they believed in: they were aware of a mean 8.7/10 but they stated that they only believed in 3.1/10. This result was not at all surprising as most of the conspiracies listed were very well known (the least well known was about the role of the ‘Bilderberg group’). The category (conspiracy theory or statement) had no significant effect on recognition or recall. In other words, our participants did not state that they were aware of more conspiracy theories than statements as such and did not recall more of one or the other. Finally, however, we did note that the category tended to have an effect on adherence: participants believed more in statements than in conspiracy theories. In this case, the category or the label played a greater role.

Conclusion

These two studies illustrate the fact that rumours and conspiracy theories can be known and passed on without belief in their content. While they are not rational or logical, rumours and conspiracy theories as forms of manifestation of social thinking depend on their own internal logic and on the social background of the individual.

Social thinking possesses its own logic, which is different from that of Cartesian rationality. It is impossible to try and convince an adherent of a conspiracy theory or rumour of the false nature of that conspiracy theory or rumour because representatives of the two sides are not thinking on the same level. For Reference Pouillon, Izard and SmithPouillon (1979), ‘it is the unbeliever who believes that the believer believes in the existence of God’. It is because we hold that someone else believes that we ourselves are non-believers.

Even though the question of which factors influence adherence to conspiracy theories is an essential one to answer, we hope the points we have raised show that it is also important to consider that dissemination and propagation of them does not necessarily imply adherence to them. The spreading of rumours and conspiracy theorist beliefs makes possible the creation and reinforcement of social links within a group. The transmission of rumours even serves to transmit the norms and values of a social group (see Reference Baumeister, Zhang and VohsBaumeister et al.'s work (2004) questioning students on their knowledge of rumours but, above all, on the network formed of individuals who told them of a rumour and equally of the individuals to whom they had transmitted it). Therefore, an individual who refuses to transmit certain beliefs, holding them to be conspiracy theories or rumours, could even find themselves stigmatised or excluded from their social group.