Conspiracy Mentality Versus Belief in Conspiracy Theories

Clarity

Nera (2024, this issue) insightfully diagnoses a lack of clarity in the language used to describe the relationship between conspiracy mentality and belief in conspiracy theories. Before commenting on these points, we try to be clear on our own terminology. By conspiracy mentality, we mean the abstract, “generalized political attitude” in which forces are perceived to conspire to manipulate world events, intended to be measured by the Conspiracy Mentality Questionnaire (CMQ; see Bruder et al., 2013; Imhoff & Bruder, 2014; see also Stojanov & Halberstadt, 2019). By belief in conspiracy theories, we mean the general tendency to endorse specific conspiracy theories across domains, measured by scales such as the Belief in Conspiracy Theories Inventory (BCTI; Swami et al., 2010).¹

One of the ambiguities that Nera (2024, this issue) identifies surrounds the nature of conspiracy mentality – what is it, exactly? What, for example, is a mentality, and what makes the conspiracy mentality more than a set of beliefs (see Williams et al., 2022)? What motivations, structures, and cognitive processes form part of this mentality, over and above its belief content (see also Sutton & Douglas, 2020)? As Nera writes, describing conspiracy mentality as a disposition also raises important questions. Terms such as “mentality” carry surplus meaning over the tendency to believe in conspiracy theories. This meaning needs to be specified theoretically and validated empirically (Sutton & Douglas, 2020).

The surplus meaning in this construct risks conflating the description of belief in conspiracy theories with its explanation. Indeed, as Nera (2024, this issue) notes, it affords two radically different understandings of the relationship between the conspiracy mentality and belief in conspiracy theories. The first views the relation as isomorphic. That is, the conspiracy mentality may be a latent construct in scales such as the CMQ and scales such as the BCTI. In other words, it may be the generalized tendency to believe in conspiracy theories (Frenken & Imhoff, 2021). This view appears to be straightforwardly asserted by Imhoff, Zimmer et al. (2022, emphasis added):

The second understanding views the relation as causal. The conspiracy mentality may cause belief in conspiracy theories. This alternative understanding of the relation appears to be postulated by Imhoff and Bruder (2014, emphasis added):

This ambiguity suggests that the conspiracy mentality does not have a stable conceptualization: Is it supposed to be the general tendency to endorse conspiracy theories, or is it supposed to cause this tendency? It seems incoherent to view a variable as both a cause of another variable and the same thing as that variable. Thus, in the following sections, we ask whether and how the construct might help researchers understand the psychological appeal and consequences of conspiracy theories when viewed in isomorphic or causal terms.

Isomorphism

If we view the two constructs as isomorphic (the conspiracy mentality is the general tendency to endorse conspiracy theories), then the conspiracy mentality cannot explain what causes this general tendency. It rather offers a new way to characterize (talk about) this tendency (adapted from Moscovici, 1987), and, through associated scales, new ways to measure it. Taking the isomorphic view, therefore, entails that the contribution of the conspiracy mentality to the literature depends on advantages of the scales that have been designed to measure it (e.g., the CMQ), relative to scales such as the BCTI.

Imhoff, Bertlich et al. (2022) offer three reasons to prefer the CMQ. One is that the CMQ is less skewed than the BCTI. This matters, since skewness is typically an undesirable psychometric characteristic. However, nature is not always normally distributed, and as we argue elsewhere in this issue (Trella et al., 2023), we might expect stigmatized and unconventional beliefs to be skewed positively. Thus, the skewness of scales such as the BCTI might be an authentic reflection of the phenomenon we are trying to understand. A second claimed benefit is that the CMQ may be more stable over time. This is generally a desirable characteristic of a scale designed to capture a stable disposition. However, this greater stability is also open to a less favorable interpretation: the CMQ is less sensitive to important and informative fluctuations. The third claimed benefit, that the CMQ is free from contamination by the specific content of conspiracy theories, is moot. Some evidence (e.g., Swami et al., 2011) suggests that endorsements of novel and familiar conspiracy theories are so correlated that familiarity may not be very important. Further, it overlooks potential sources of content contamination in the CMQ itself. For example, participants living in authoritarian regimes may feel uncomfortable endorsing a CMQ item such as “Politicians do not usually tell us the true motives for their decisions.” For them, compared to participants in democracies, such items carry a specific and rather menacing surplus meaning.

Furthermore, we suggest that a disadvantage of the CMQ arises from the different approach it takes to measurement. Since the CMQ is comprised of relatively abstract statements and does not include any specific conspiracy theories, it is logically possible for a participant to endorse it (e.g., score above mid-point), but not to endorse even one conspiracy theory (the strong correlation between CMQ scores and belief in conspiracy theories notwithstanding). This risk does not confront scales such as the BCTI, which measures mean endorsement of various conspiracy theories. Thus, the BCTI treats this as an observable tendency: One can literally observe people doing it. Observable variables are distinct from latent variables, whose existence and nature can only be inferred by mathematical inference from variables that can be observed (Kmenta, 1986). Using the CMQ to measure the general tendency to endorse conspiracy theories means regarding this tendency as a latent variable whose existence can only be inferred. Whatever its other advantages, this takes the research one step further away from studying relationships between observable variables, which is surely a fundamental objective of an applicable behavioral science.

Such differences in the items of the CMQ and scales such as the BCTI suggest that they are really not measuring the same thing. Elsewhere in this issue (Trella et al., 2023), we report that people view endorsement of BCTI items versus CMQ items as more stigmatizing and likely to give rise to dispute. These effects were associated with the greater abstraction of conspiracy mentality statements and were not explained away by the relative implausibility of specific conspiracy statements. These preliminary results suggest that endorsing the two constructs has different social-psychological consequences.

In sum, the isomorphic view of the conspiracy mentality entails that it can offer no causal explanation of belief in conspiracy theories. Its contribution to the literature instead depends in large part on the vocabulary and scales it offers. The methodological and empirical considerations we have raised cast doubt on the superiority of the CMQ (see also Swami et al., 2017). It seems more likely to us that it is better suited to measure a different construct. Ultimately, we propose that the key problem with the isomorphic view lies deeper in its conceptualization. Relative to the general tendency to endorse conspiracy theories, the conspiracy mentality construct entails additional content by postulating a “mentality.” Whatever this mentality is, it must be surplus to the tendency to endorse conspiracy theories. Further, the conspiracy mentality does not include key content of belief in conspiracy theories, such as commitment to epistemically risky narratives that oppose widely accepted understandings (Douglas & Sutton, 2023: see Sutton & Douglas, 2022; Trella et al., 2023). We therefore agree with Nera (2024, this issue) that a causal understanding of the conspiracy mentality is more coherent.

Causality

Although it is coherent to think of the relationship between conspiracy mentality and belief in conspiracy theories in causal terms, Nera (2024, this issue) rightly argues that the nature of this relationship is unclear. As Nera suggests, there is a danger that audiences might think of the causal relation in circular terms (i.e., people believe in conspiracy theories because they are disposed to believe in conspiracy theories). As Nera also suggests, however, Imhoff and Bruder (2014) provide a definition of the conspiracy mentality (e.g., as a generalized political attitude) and so did not necessarily put forward a circular argument. Nonetheless, this definition is so minimal as to leave its causal relevance a mystery (a.k.a., “black box”). In this section therefore, we add to Nera’s critique some observations about the strength, nature, and mechanisms of any possible causal relationships.

First, it is surely not realistic to regard conspiracy mentality as the cause of belief in conspiracy theories. Human beliefs and behaviors have multiple causes (e.g., Kim, 1989): as a rule, no one variable is entirely caused by any other variable. Therefore, the conspiracy mentality should be regarded as a hypothetical cause. We also agree with Nera (2024, this issue) that any causal relationship may be bidirectional: the conspiracy mentality may be an effect as well as a cause of belief in conspiracy theories (for longitudinal confirmation of this, see Granados Samayoa et al., 2022; for experimental confirmation, see Trella et al., 2024, this issue). This speaks against the conspiracy mentality as the sole or dominant driver of belief in conspiracy theories and suggests instead more complex, nuanced relationships.

As Nera (2024, this issue) argues, understanding the causal relevance of the conspiracy mentality requires theoretical elaboration of the construct and associated causal mechanisms. To this end, we offer some suggestions. The conspiracy mentality comprises an abstract representation of some of the core political or quasi-political sentiments of conspiracy theories.² This minimal understanding of the construct is sufficient to derive some causal hypotheses, since it identifies semantic affinities between the conspiracy mentality and belief in conspiracy theories. Semantic affinities do not guarantee and therefore do not necessarily trivialize causal relationships. Indeed, it is well established that people can hold multiple logically inconsistent beliefs (Arnulf et al., 2018; Billig, 1996). There must therefore be psychological processes by which semantically related constructs are causally related.

Fortunately, three known processes yield causal relations of this sort and allow us to hypothesize about what lies inside the black box. First, patterns of activation in the associative networks representing people’s world knowledge favor belief coherence (Goertzel, 1994; Williams et al., 2022; Wood et al., 2012). In these networks, associations grow over time between mutually coherent beliefs – including conspiracy theories (Goertzel, 1994; Williams et al., 2022), and potentially, more abstract conspiracy mentality propositions (Sutton & Douglas, 2014; Wood et al., 2012). Second, people may follow normative reasoning processes, including induction and deduction, to infer one type of conspiracy representation from the other (Trella et al., 2024, this issue). Third, the semantic similarity between conspiracy mentality propositions (e.g., CMQ items) and conspiracy theories (e.g., BCTI items) means that similar cognitive operations are required to process them. Such semantic similarity can be estimated with natural language comprehension algorithms and may help account for the observed correlation between the conspiracy mentality and belief in conspiracy theories (as it might for many other variables in psychology; see Arnulf et al., 2018, 2022).

Conclusions and Recommendations for Researchers

We endorse and have tried to elaborate upon Nera’s (2024) critique by trying to understand how the conspiracy mentality and the associated scale (CMQ) may contribute to research on the psychology of conspiracy theories. We conclude that the conspiracy mentality is not the general tendency to endorse conspiracy theories. We conclude that the scale does not seem to be superior or even straightforwardly comparable to scales that attempt to assess this tendency more directly (e.g., the BCTI). However, we have identified hypothetical mechanisms that might causally connect conspiracy mentality, as both an antecedent and a consequence, to belief in conspiracy theories. This has been possible because the construct and scale capture an abstract representation of conspiracist ideas without the more socially problematic and epistemically risky content of conspiracy theories themselves. This may be valuable to the research literature. For example, using scales such as the CMQ and BCTI simultaneously might tease apart the social, political, and psychological ramifications of belief in conspiracy theories from those of the abstract, conspiracist distrust embodied by the conspiracy mentality. This would allow empirical exploration of the hypothesis that the latter distrust might be a heuristic aid to democratic processes by encouraging people to be suspicious of what powerful interests are doing and what they are being told (e.g., Briggs, 2004; Enders & Smallpage, 2018; Huntington, 1983; Sobo, 2021). Some promising divergences between the CMQ and measures such as the BCTI are indeed already being observed in meta-analysis (Stasielowicz, 2022), and in the small number of studies that run them both at the same time using covariance analysis (e.g., Pan et al., 2023; for review, see Trella et al., 2024, this issue). Toward these aims, informed by Nera’s (2024, this issue) critique and our elaboration, we recommend that researchers:

1. 1.
   Elaborate the conspiracy mentality construct theoretically and validate it empirically.
2. 2.
   Do not use the use the conspiracy mentality terminology or scale when studying belief in conspiracy theories. Instead, use scales that measure belief in specific conspiracy theories.
3. 3.
   Run both conspiracy mentality and belief in conspiracy theories scales simultaneously when resources permit. This will inform efforts to understand (a) what is measured by conspiracy mentality scales and (b) the potentially important differences between abstract and specific conspiracy representations.
4. 4.
   Be wary of postulating higher-order concepts such as conspiracy thinking that formally subsume both conspiracy mentality and belief in conspiracy theories. Without careful theoretical and empirical work, this might invite a replication, and worsening, of the conceptual confusion that already exists.