Abstract

Delusions are beliefs of interest to researchers in cognitive science, psychiatry, psychology, and philosophy. They are formed on the basis of evidence that do not properly support their content and that are resistant to counterevidence. Often these beliefs have contents that are bizarre, and they can impair a person’s functioning by prompting anxiety or distress. For example, suppose that Fred believes his wife has been replaced by an imposter (Capgras delusion), or Claire believes that she has ceased existing (Cotard delusion), or Mohammed believes that his arm is being controlled by alien forces (delusion of control). This characterization is not intended to give necessary and sufficient conditions or to capture all and only delusions. Indeed, there are numerous debates about the nature of delusions: whether delusions are beliefs, what the content of delusions might be, how delusions are formed and maintained, how these processes interact with social factors, and the epistemic standing of people with delusions. The oddity of delusions makes them a useful test case for a wide range of topics currently studied by cognitive scientists.

History

Part of the theoretical interest of, and controversy surrounding, the topic of delusion is how to capture what delusions are. If researchers in cognitive science and cognate disciplines are to study delusions, it is crucial that the target is a stable one, lest researchers risk talking past one another. Of course, the phenomenon of delusion itself reaches far back into human history, even if not always under that label (for a historical overview, see Shepherd, 2022). Historically, being deluded and being mad amounted to the same thing (Jaspers, 1913; Ball & Ritti, 1881 cited in Berrios, 1981). This can be seen in Paul Hoff’s tracing of the term delusion as entering medical vocabulary at the start of the 19th century (as the German Wahn, madness; Hoff, 2006). In Great Britain at least, the term was applied to perceptual disorders in 1830 and then to wrong beliefs from 1850 (Berrios, 1981).

At the start of the 20th century, three pragmatic criteria (i.e., criteria not present in every case but which could nevertheless serve as guidelines in clinical and research settings) for delusion were identified: (1) subjective feeling of certainty, (2) incorrigibility (i.e., difficult to correct), and (3) wrong or impossible contents (Jaspers, 1913). Much of the spirit of this understanding survives today. On reflecting on the question of what a delusion is, Karl Jaspers stated that it would be “superficial and wrong” to capture delusions just as incorrigibly held false beliefs (Jaspers, 1913: 78). Nevertheless, the use of the term “delusion” understood as “wrong beliefs” has a significant history, beginning in the 19th century and very much still operating today with, of course, some further specification (Berrios, 1991).

Core concepts

One common starting point for thinking about the nature of delusion—at least to a level sufficient for further discussion—is to look to diagnostic criteria. In the two most recent editions of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV and DSM-5), the definition of delusion is given as the following:

A false belief based on incorrect inference about external reality that is firmly sustained despite what almost everyone else believes and despite what constitutes incontrovertible and obvious proof or evidence to the contrary. The belief is not one ordinarily accepted by other members of the person’s culture or subculture (e.g., it is not an article of religious faith). When a false belief involves a value judgment, it is regarded as a delusion only when the judgment is so extreme as to defy credibility (American Psychiatric Association, 2000: 765; American Psychiatric Association, 2013: 819).

The DSM-5 also characterizes delusions when they appear in a list of key features of psychotic disorders. They are described as follows:

[…] fixed beliefs that are not amenable to change in light of conflicting evidence. Their content may include a variety of themes (e.g., persecutory, referential, somatic, religious, grandiose). […] Delusions are deemed bizarre if they are clearly implausible and not understandable to same-culture peers and do not derive from ordinary life experiences. […] The distinction between a delusion and a strongly held idea is sometimes difficult to make and depends in part on the degree of conviction with which the belief is held despite clear or reasonable contradictory evidence regarding its veracity (American Psychiatric Association, 2013: 87).

There has been significant criticism of the DSM’s characterization of delusion. In particular, there are four main points of contention. First, in the glossary definitions, delusions are described as false beliefs, but the bare possibility of true delusions should not be ruled out. Suppose there was someone with a belief exactly like a delusion (in the way it is formed, sustained, and so on), which happened to be true. Arguably such a case should nevertheless be counted as a delusion (Davies et al., 2001). Second, it is often not clear what would count as “incontrovertible proof or evidence” against the delusion because some delusions are, at least in practice, unfalsifiable (Coltheart et al., 2011: 274). Conversely, some authors have argued that people with delusions have the capacity to appropriately respond to evidence concerning their delusion after all (Flores, 2021). Third, the culture clause has been criticized as atheoretical—it is not clear why a belief should be excluded from the category of delusion just on the basis that is shared, and indeed, this exemption excludes the possibility of group delusion (Ross & McKay, 2017). Fourth, delusions are characterized as beliefs; this characterization has been subject to substantial debate.

Perhaps the most important concept for understanding delusion is belief. There is no uncontroversial characterization of belief, but there are a couple of features around which most researchers agree. The first is that beliefs are formed and revised on the basis of evidence, and the second is that they are motivationally efficacious with respect to behavior, that is, they make people act when combined with the right desires, across contexts. This might well contrast with other mental states like imagination, which do not have these features.

Another important piece of conceptual apparatus is the distinction between monothematic and polythematic delusion. Monothematic delusions concern a single topic and can arise in otherwise healthy individuals (Coltheart et al., 2007), whilst polythematic delusions are often more elaborated, involve more than one topic, and occur in the context of mental disorder. There is, then, a very strong (although not exceptionless) relationship between a delusion being monothematic and circumscribed on the one hand and a delusion being polythematic and elaborated on the other (Davies et al., 2001).

A final core concept especially important for research on monothematic delusion in particular is aberrant or anomalous experience, which many researchers have taken to play a key role in the formation of these beliefs. For example, in the case of a delusion of alien control, in which a subject might believe that some of her bodily movements are not authored by her but something external to her, such experiences might include one’s intentions being incongruent with one’s actions. In other cases, the anomalous experience might take the form of systematic visual and auditory hallucinations. Yet, other anomalous experiences might best be understood as experiences of absence. For example, in Capgras delusion (where one believes that a familiar person has been replaced by an imposter), subjects experience a lack of affective response to somebody with whom they are close. In Cotard delusion (where one believes that one is dead or has ceased existing), this lack of affective response may be generalized to one’s entire environment (Young et al., 1992), or there is an experience of depersonalization (see Gerrans, 2024 for a discussion of the experience associated with Cotard delusion). Some experiences involved in delusion have been hypothesized to arise from neuropsychological damage (see Coltheart, 2010).

Questions, controversies, and new developments

Are delusions beliefs?

Belief is often taken to be a core concept in delusion research, and scholars have noted that beliefs are the kinds of things that are responsive to evidence and involved in the production of action. Whether delusions are properly described as beliefs has been subject to substantial and ongoing debate. Although the idea that delusions are beliefs is present in the diagnostic criteria, and indeed in much of the literature, some characterizations of belief make the idea of delusions understood this way seem rather implausible.

Take the content of Fred’s delusion: my wife has been replaced by an imposter. The question of whether Fred is properly described as believing that content matters. For if Fred believes that his wife has been replaced by an imposter, he might be expected to behave in particular ways, have evidence for his claims, respond to counterevidence against his claims, and so on. However, delusions are typically thought to not respond to evidence, and they can fail to guide action in the manner of belief. At first—albeit superficial—blush, then, delusions do not look like beliefs.

In this debate, two options are available. The first is doxasticism about delusion, which is the view that delusions are beliefs. Doxasticists loosen strict ties between belief and evidence or belief and action and treat delusions as one of many kinds of imperfect belief (Bortolotti, 2009). Alternatively, non-doxasticism about delusion has it that the apparently bad fit of delusions into the category of belief should be taken seriously and that delusions are better captured as a different kind of mental state. Non-doxasticist positions have been developed according to which delusions are in fact imaginings (misidentified by their subjects as beliefs; Currie, 2000), cognitive feelings (Dub, 2017), or as somewhere in between beliefs and imaginings (Egan, 2008).

This debate has taken place in what has been called the Cartesian tradition of belief formation, according to which one can entertain (that is, hold in conscious thought) a proposition and then endorse that content in belief—or not. An alternative approach to belief acquisition is the Spinozan model, according to which merely representing a proposition leads immediately to believing it (see Gilbert et al., 1993). This latter approach to belief acquisition radically loosens the constraints and makes belief much easier to come by, thus potentially making delusions as beliefs so understood more plausible (an idea recently explored by Bongiorno, 2022).

What kind of belief?

Even for doxasticists, delusions are at least somewhat different from beliefs about, for example, bus timetables and the weather, so is there a further way they can be characterized—as a specific kind of belief—which might help explain why they are different?

Delusions are usually characterized negatively: they are considered disruptive of broader functioning and have a bad relationship to evidence. However, there are a few promising avenues for identifying the benefits of delusions. None of these avenues would suggest that delusions are overall good, they rather recommend taking a more nuanced view as to the overall value of a delusion. One way of capturing this is through the lens of epistemic innocence, in which delusions are understood as epistemically imperfect cognitions that nevertheless deliver significant epistemic benefits not otherwise available to the subject (Bortolotti, 2020). Psychologists and philosophers have also explored the relationship between delusion and meaning. Recently, it has been argued that delusions can enhance a person’s sense of meaningfulness, understood as “the extent to which one’s life is subjectively experienced as making sense and as being motivated and directed by valued goals” (Ritunnano et al., 2022: 110).

It is also relevant to consider the proper functioning of belief in the context of evolution and then ask whether delusions are adaptations or malfunctions of belief. That delusions are malfunctioning beliefs is prima facie quite plausible (see Miyazono, 2018 for the most thoroughgoing defense of this idea, (cf. Sullivan-Bissett forthcoming). However, some researchers have suggested that delusions (or a subset of them) might actually be biological adaptations in which the mechanisms responsible for their formation were selected for their role in deceiving others into social alliances (Hagen, 2008). Another kind of adaptive approach can be located in the prediction error framework. According to such an approach, delusions are adaptive insofar as they maintain behavioral interactions in the face of abnormal prediction-error signaling (Fineberg & Corlett, 2016).

How are delusions formed?

Related to the question of what kind of belief delusions are is the question concerning how they are formed. Empirical work in cognitive psychology has focused primarily on explaining how monothematic delusions (those concerning a single topic) are formed. There are two broad approaches to delusion formation: empiricism and rationalism. According to empiricism, the anomalous experiences often associated with monothematic delusions play a causal role in the formation and maintenance of those beliefs. So, for example, the discrepancy between Mohammed’s intentions regarding the movement of his arm and the actual movements of his arm is a key component of the explanation as to why Mohammed believes that his arm is being controlled by alien forces. To take another example, when Fred looks at his wife and does not feel what he is used to feeling, that absence of an expected response is part of the explanation for why he believes that the woman in front of him is not his wife but an imposter. On the other hand, according to rationalism, anomalous experiences are downstream of delusional beliefs and do not play a role in their genesis (Campbell, 2001).

Both empiricism and rationalism recognize a distinction between perception and belief, and the main point of disagreement might be construed as which comes first in the causal account of delusion formation. A different way of considering matters is given through prediction-error approaches, which are built upon the denial of a sharp distinction between perceptual and doxastic mechanisms. Rather, such views hold that perceptual processing involves generating predictions about sensory input based on hypotheses about the world. Some prediction-error theories hypothesize that aberrant prediction error signals cause problems in the allocation of attention in people with delusions (Fletcher & Frith, 2009). This might result in people with delusions paying undue attention to particular things or events (those which defy expectations), and a delusion is formed to explain those things or events. Such views might be distinguished from empiricism and rationalism on the grounds that “top-down and bottom-up processes sculpt one another” (Fineberg & Corlett, 2016: 5).

How do delusions get their contents?

There is a debate within empiricism concerning the relationship between the anomalous experience and the content of the delusion. According to endorsement models, the content of the anomalous experience is identical to the content of the delusion (the delusional belief is thus an endorsement of what is presented in experience). Explanationist models, on the other hand, state that the content of the anomalous experience falls short of the content of the delusion but the delusion explains the anomalous experience. Take a case of perceptual delusional bicephaly, in which a subject believes that she has a second head, and suppose that the subject hallucinates (visually and auditorily) a second head. On an endorsement approach, the subject might experience that she has a second head, and that experiential content is simply endorsed in belief. For an explanationist, the hypothesis that one has a second head is devised to explain the various experiential data, but the delusional belief goes further than what is given in experience.

These views might seem equally plausible, or at least equally deserving of consideration, when considering cases like perceptual delusional bicephaly, in which the content of the anomalous experience might be taken to be sufficiently rich to be endorsed in belief. But in other cases, specifically Capgras, in which the relevant experience is the absence of something expected, it has been suggested that endorsement models will fail to provide a convincing characterization of the experience that is endorsed (for discussion, see Bongiorno, 2020).

How many factors?

Another debate within empiricism relates to how many factors are needed to explain the formation or maintenance of a delusion and, in particular, whether there are one or two. It is important to note that a factor is not a mere cause. Delusions have many causes. Rather, a factor is a cause with the feature of being abnormal and is explanatorily relevant to delusional belief formation in particular. The abnormality of factors is widely recognized. For example, one-factor theorists have understood a factor as “an abnormality that explains the formation of abnormal beliefs” (Noordhof & Sullivan-Bissett, 2021: 10279), whilst two-factor theorists have taken the second factor to be “a departure from what is normally the case” (Davies et al., 2005: 228). Furthermore, when discussing how to defend the one-factor approach, one suggestion has been to show that the proposed second factor in fact describes “a rationalization process which is within the normal range” (Gerrans, 2002: 48). This would only be a defense of one factor over two if the second factor is taken to pick out an abnormality. One-factor theories claim that the only factor in the formation or maintenance of a delusion is anomalous experience (a view developed by Maher, 1999, elsewhere, and more recently by Noordhof & Sullivan-Bissett, 2021).

Two-factor theories state two factors are needed, and such theories can be distinguished by their characterization of the second factor. Proposals for the second factor can be roughly grouped as characterizing a performance error, reasoning bias, or belief evaluation deficit. The claim that the second factor is a performance error (as opposed to a failure of competence) is the claim that subjects with delusions have the capacity to, for example, form or evaluate beliefs appropriately, but they fail to put that capacity into practice (Gerrans, 2001).

Other two-factor theorists have sought to identify a bias in reasoning as the second factor. One suggestion for this is a data-gathering bias understood in terms of jumping to conclusions, thought to be supported by empirical data gathered from the beads task (e.g., Garety et al., 1991). In this task, subjects are presented with two opaque jars of beads containing two colors in opposing ratios (e.g., 80:20 and 20:80). Subjects are asked to say when they are confident that they know which jar the beads are being drawn from. Those with delusions are found to request fewer beads before deciding on the jar, hence the charge of jumping to conclusions. However, recently it has been argued that when participants coded as “careless” are removed from the data, the relationship between holding delusion-like beliefs and jumping to conclusions is “severely attenuated” or “disappeared entirely” (Sulik et al., 2023: 757).

Another suggestion is a bias towards observational adequacy, in which subjects privilege observational data (that given in experience) over minimizing adjustments to one’s beliefs (Stone & Young, 1997). One issue with this is that if subjects with delusions had a bias of this kind, they should also be more easily persuaded by visual illusions, but there is no evidence that this is the case (Davies & Coltheart, 2000).

Finally, some have argued that attributional biases might play the envisaged role of the second factor, particularly for certain kinds of delusion. There is some evidence, for example, that people with persecutory delusions are more likely to attribute the cause of positive events internally (i.e., to themselves) and attribute the cause of negative events externally (i.e., to other people or circumstances), and the opposing pattern has been found in subjects with depressive delusions (Kaney & Bentall, 1989). There remains a question of whether attributional biases in fact characterize delusions of various kinds (Langdon & Coltheart, 2000: 193–197). Furthermore, it has been argued that it is a familiar fact that one’s mood can influence the beliefs one forms, and so this would not be something that could explain delusion in particular (Langdon & Coltheart, 2000).

Other two-factor theorists have located the second factor not in belief formation but in belief evaluation, suggesting that there is a deficit in the belief evaluation of people with delusions (Langdon & Coltheart, 2000). It has been hypothesized that this arises from right hemisphere damage in the frontal lobe, which interferes with the belief evaluation mechanism (Coltheart et al., 2007). According to this view, what needs explaining is not why the subject forms the delusional belief but rather why she maintains it in the face of overwhelming counterevidence.

A new development in research on delusion is the social turn, which draws on social sources of evidence and cultural influences on delusional content. The social turn resists approaches to delusion that focus too narrowly on the individual [see Social Epistemology]. Although the social dimension of delusions has been taken to be key to their understanding before (e.g., Freeman, 2016 on paranoid delusions), there is nevertheless a more recent research program that, perhaps at its boldest, seeks to identify a role for the social in delusion, both in its formation but also its attribution (the conditions under which researchers ascribe delusion; Wilkinson, 2020).

Some social theorists also claim that delusions involve disturbances in social cognition. This pivot to understanding delusions, at least in part, as a social phenomenon has been argued to provide an explanation for several features, including the fact that their contents are often socially themed, most obviously, by implicating other social actors (e.g., delusions of persecution or jealousy; Gold & Gold, 2014; Bell et al., 2021) and that they exhibit resistance to social influence—for example, testimony from peers (Bell et al., 2021; Miyazono & Salice, 2021) [see Social Learning].

Broader connections

It is widely recognized that delusions are not the only beliefs that stand in a poor relationship to evidence; self-deceptive beliefs and conspiracy beliefs in particular have been taken to be a useful comparison point. Self-deceptive beliefs are formed in the presence of information that would justify a different belief (Van Leeuwen, 2007), whilst conspiracy theorists’ hasty dismissal of counterevidence has been described as “arguably the most important feature of conspiracy theories” (McKenna, 2017: 57). Another similarity between delusions, self-deceptive beliefs, and conspiracy beliefs is that in all three cases, there might be motivational influences on belief formation and maintenance. In standard cases of self-deception, a subject believes what she desires to be the case. In the case of conspiracy beliefs, some researchers have posited the influence of motivational biases, for example, the need for uniqueness (Imhoff & Lamberty, 2017), to explain the adoption of a conspiracy belief. The etiology of delusion has also been claimed to involve motivation, at least in some cases. Such cases may include reverse Othello delusion (the belief that one’s partner is faithful), anosognosia, and perhaps others that contribute to increased self-esteem, for example, delusions of grandeur (Bortolotti & Mameli, 2012). There are, of course, key differences as well (for self-deception, see Bayne & Fernández, 2009, and for conspiracy beliefs, see Pierre, 2021). Notwithstanding these differences, it is useful to look at delusions alongside other evidence-resistant beliefs that involve motivation.

Epistemic injustice is a concept introduced to capture the injustice done to a person when they are harmed in their capacity as a knower (Fricker, 2007). For example, a subject’s testimony might not be given the credibility it is due because of prejudices regarding some feature of the subject’s identity (e.g., sex or race). Epistemic injustice has recently been explored in the context of psychiatry. If epistemic injustice arises in cases in which a person is a member of a negatively stereotyped group, then a natural extension of the concept is to subjects with a psychiatric diagnosis, since this also gives rise to negative stereotypes. Indeed, it has been argued those diagnosed with a mental disorder may even be at greater risk of epistemic injustice than those diagnosed with a physical illness, due to “the high prevalence and great power of negative stereotypes of psychiatric illness” (Crichton, 2017: 65). Research has focused on particular groups in the psychiatric context and their vulnerability as a member of that group. Recently, some authors have focused on epistemic injustice and delusion, suggesting that subjects with delusions are at risk of epistemic injustice because of negative stereotypes associated with having a delusional belief, in particular, stereotypes that cast the subject as being “bizarre, incomprehensible, and irrational” (Sanati & Kyratsous, 2015: 484; for a review of delusion and epistemic injustice, see Palafox-Harris, 2024).

Acknowledgments

I am grateful to the Arts and Humanities Research Council for funding my work (Deluded by Experience, grant no. AH/T013486/1). Many thanks to Evan Kidd, Colin Klein, and Asifa Majid for comments on an earlier version of this entry.

Delusion