Delusional beliefs have puzzled psychologists and psychiatrists for well over 100 years (Janet, 1898; Jung, 1907/1960; Jaspers, 1913/1963). Delusions are usually defined as fixed and false beliefs that continue to be held even when the person is presented with conflicting evidence (American Psychiatric Association, 2013; Kendler et al., 1983). The most common form of delusion are persecutory delusions, when individuals state beliefs that they are going to be harmed or harassed, or that they are being watched (American Psychiatric Association, 2013). These behaviors can lead to many problems in a person’s life and much suffering.

There have been many criticisms of such a definition, and especially whether there is any difference between “clinical” delusions and strongly held religious or other beliefs, and how “false beliefs” can even be determined and by whom (cf. Georgaca, 2000, 2004; Murphy et al., 2018). The criteria of falsity or implausibility have been particularly questioned by critics who highlight that the distinction between plausibility and implausibility is almost impossible to resolve (Bentall, 2018; Freeman & Garety, 2014; Freeman et al., 2002; Garety & Freeman, 1999; Harper, 2004; Murphy et al., 2018), and that because of the usual power imbalance between a “client” and an “expert,” it is the beliefs of the “client” that get pathologized, and labeled as false and delusional (Palmer, 2000), whereas they can also be seen as just different. A practical solution for some of these problems has been to consider delusions as those beliefs that are unusual or false within a particular society, community, or culture, although this hides, rather than solves, the issue (Feyaerts et al., 2021; The Lancet Psychiatry, 2021).

This article will first argue instead for a contextual reconceptualization of “delusional beliefs” in which the above issues do not even arise, and that further means that “therapy” does not need to consist of persuading the person out of their reputably false and fixed beliefs. We will first review the properties of delusions, followed by a brief overview of the current explanations for delusional beliefs. We will then present a new account of delusions based on discourse analysis (Edwards & Potter, 1993) and social contextual analysis (Guerin, 2016, 2020a). The main difference with this alternative perspective is to focus on delusions, like all behaviors, as being shaped by the life contexts of the person and especially the social, cultural, and societal life contexts. Rather than focusing on the truth or falsity of the beliefs, our analysis focuses on delusions as normal social storytelling gone wrong. The questions then become: What has shaped the person into making such belief statements within their life world? What does stating these beliefs do for the person?

What are the Properties of Delusions?

Karl Jaspers first set out the idea of delusions by giving several properties to define them as beliefs (Jaspers, 1913/1963). He followed the traditional psychological and common-sense idea of “beliefs,” that people had beliefs “inside” them as a reflection of their life experience, and that these internally stored “beliefs” were occasionally reported in speech. But for delusions, unlike the “normal” beliefs of most people, such beliefs were strongly held, resistant to change, sudden or without a prior history in their life, and idiosyncratic or out of context. They were also mostly unusual or at least not mundane beliefs such as “cars have wheels” (Bentall, 2018).

The first column of Table 1 lists the commonly suggested properties of delusions, with similar terms used by other writers in brackets. The problem is that all these properties can also be found in the “normal” beliefs of other groups and with similar intensities (such as religious, political and sporting groups, or even families; Billig, 1991). None of these properties, therefore, works well as a definition of delusions, because these properties perfectly encapsulate the beliefs of others who are considered “normal” in their own groups. For example, someone who has recently joined a Pentecostal Church will espouse strongly held beliefs that start suddenly and are resistant to change, and to an atheist these new beliefs are idiosyncratic and unusual (an omnipresent god who talks directly to them?). Within the Pentecostal Church, however, these beliefs are neither idiosyncratic nor unusual, and, moreover, they should be held strongly and be resistant to change as signs of faith.

Table 1 The main properties of persecutory delusions for both a traditional and a discursive view of belief

So, when we assume the common way of conceptualizing human beliefs that has been followed in psychology, we are led to these sorts of conundrums, even though we sort of know what Jaspers and everyone since has meant. Beliefs in a Pentecostal Church seem to make sense within that milieu whereas being watched by the CIA or being hunted by some unknown assassins do not seem to make sense in any groups or milieus (unless the person is part of a group of conspiracy theory believers). Put in more contextual or behavioral terms, the Pentecostal Church has made those beliefs functional within their community whereas the person said to have persecutory delusions is not usually part of a relevant group.

The practical problem we will address in this article is that although the beliefs of someone said to have persecutory delusions do not appear to be functional as social behaviors, unlike the Pentecostal Church example, this will, of course, depend upon how thoroughly anyone has bothered to look. So, the final listed property in Table 1 relies on a judgment as to whether the stated beliefs appear to be socially functional with an observable group, how well we have observed the beliefs being stated within their contexts, and how well we have observed the functional outcomes of stating such beliefs. Within usual restrictions of clinical practice, it is nearly impossible to make such observations but still leaves it open that other contextual methodologies might make this criterion more useful (Guerin et al., 2018).

Review of Explanations for Delusions

There have been four main approaches in mainstream psychology and psychiatry for understanding and trying to explain delusions: by looking at personal and demographic patterns; by treating delusions as behaviors interpretable through various theoretical notions; by treating delusions as faulty cognitive processing; and by treating delusions as normal attributions but ones which have gone wrong in some way. Table 2 lists most of the variations of these.

Table 2 Four main types of explanations for delusions and their variants

Personal/Demographic Patterns

The first approach to “explaining” delusions is to document correlations with features of the person’s life. Those people diagnosed with delusions are said to be in a higher risk category, for example, or have more discrimination against them or lower education levels (Fearon et al., 2006). There are also correlations with various social indicators, such as having insecure attachments in life (especially early life), a strong need to belong, complex or poor social relationships in life, and more emotional distress (Freeman, 2016; Lyon et al., 1994; Lyon et al., 1999; Randjbar et al., 2011; Stevens et al., 2009; van der Ven & Selten, 2018).

Although these correlations can be helpful in pointing us in new directions, they do not explain anything by themselves. However, they can possibly lead to more explanatory models and some have been taken up in cognitive and attribution models. Some will also figure indirectly in social contextual explanations for delusions.

Faulty Behavior Patterns

More observable units of behavior have also been linked with delusions. Overreliance on immediate stimuli in the environments and focus of attention are two examples, although these once again do not explain because we need to know how they occur in the first place, and how they more directly link to delusions (Harris et al., 2014; Merrin et al., 2007; Salzinger & Serper, 2004). Some cognitive and attribution models use these more concretely, however. Other more behavioral links treat delusions as “misguided” searching after meaning or ways of reducing stress (Freeman et al., 2002; Kinderman & Bentall, 1997). Another “explanation” has been that those with delusions have developed a “pessimistic explanatory style,” so they emphasize exaggerated and negative explanations in their life (Alloy et al., 1984). Once again, we still need to know with all these how they arise and how they lead to delusional behaviors in particular, rather than any other behaviors.

Faulty Cognitive Processing or Verbal Learning

Most of the recent attempts to explain delusional behavior focus on cognitive explanations, and these also currently include the attributional models which follow. The basic proposition is that something (not usually specified) has led to a fault in the person’s cognitive processing system, and this leads them to make delusional beliefs (Lincoln et al., 2010).

In this vein, it is said that people who state delusions are led to: more readily jump to conclusions in their reasoning; reason about the world in faulty or biased ways; not search properly for information that might change their delusions; process their environment in faulty ways (including the attentional problems in the previous section); learn to only work with one single explanation for events in their world; finish processing information about events too quickly (need for closure); by making hasty decisions because ambiguity is not tolerated and a single firm decision (belief) is adhered to; be inflexible in their processing; not be able to judge other people well; and have built self-schemata that are faulty in various ways (Aakre et al., 2008; Buck et al., 2020; Moritz et al., 2018; Ward & Garety, 2019). These are abstract models as to why someone would say delusional beliefs to others, and the evidence is therefore indirect (Lee et al., 2004; Lincoln et al., 2010; Morrison, 2001; Randall et al., 2003). Once again, these are useful as correlates to explore, but they are not explanatory without details of how these difficulties developed, what they do (functions), and why they persist.

There is one behavior analytic interpretation of delusions that focuses on rule insensitivity, which is a starting point for the present account (Monestès et al., 2014). It proposed that delusions are like rules and people who show clinical delusions showed some tendency on a laboratory task to be insensitive to rules. However, it is not clear why these people develop such insensitivity, although the account proposed in the current article would suggest a history of poor language interactions in their life situations, as detailed below. So, this model can be seen as a good starting point for the following account, but only where the delusions are in the form of rules. As will be suggested below, not all delusions are like this and most consist of much more elaborated stories and claims of multiple beliefs.

Attributional Approaches

The final grouping of explanations, and the most recent and popular, moves partly away from cognitive explanations and into the social functioning of language. This is where the present article will move even further away from “cognitive” approaches for attributions to even more discursive approaches (Edwards & Potter, 1993; Guerin, 2016, 2020b).

The basic idea of attribution phenomena is that those people deemed to have delusions are trying to explain the happenings in their worlds but for various reasons get this wrong (e.g., Livet et al., 2020; Maher, 1999). This basic proposition can be couched in cognitive theory or not, and some couch this as “personality styles of attribution” that have developed over time (including the pessimistic explanatory style and poor self-schemata mentioned earlier).

One general version, following Kelley (1973) and social psychology, is that things are going wrong in the person’s life and so they learn to blame this on (attribute this to) other people to remove any responsibility from themselves. This is an exaggerated form of what is called the “self-serving bias” in social psychology and elsewhere (De Michele et al., 1998; Shepperd et al., 2008; Stewart, 2005). Any negative events are the fault of someone else. In this sort of way, it is said that people will develop delusional beliefs about the world and their place in the world, because they have been trying to find ways to protect their “sense of self” (Beese & Stratton, 2004; Diez-Alegria et al., 2006; Kinderman & Bentall, 1997; Lyon et al., 1994; Young & Bentall, 1997).

The evidence for such attributional accounts is more directly observable than for the other cognitive explanations, because people’s talk can be recorded in a way that “cognitive processes” can only be inferred (Bentall & Kaney, 2005; Garety & Freeman, 1999; Kinderman & Bentall, 1997, 2000; Lyon et al., 1994). However, most attributional models then go on to interpret (see Table 2) what is found in terms of abstract cognitive models (Anderson & Slusher, 1986), including what is this “self” and why it even needs protecting. We will see later that this is not a necessary step.

Discourse Analysis and Models of “Mental Health” Behaviors

After this brief review of the properties and explanations for delusional behaviors, we will suggest another way to view the phenomena (rather than the explanations that have been given), and then explore more the practical and clinical implications for this.

The main point of this is to show that “beliefs” are not cognitive or knowledge structures, but behaviors we use for managing our social relationships. There are many earlier formulations of language use as a social behavior spanning discourse analysis, social anthropology, traditional behavior analysis, interbehavioral psychology, and sociology (Bakhtin, 1984; Farr & Moscovici, 1984; Godlier, 2011; Kantor, 1977, 1981; Kantor & Smith, 1975; Mead, 1934; Skinner, 1957; Vološinov, 1973; Vygotsky, 1997; Wertsch, 1985). Although these have provided some good starting points, they do not have observations and details and remain abstract (cf. Guerin, 1997, with Skinner, 1957, ch. 7, for example). There are also several contextual/behavioral approaches which could be used here (Drossel et al., 2007; Hughes & Barnes-Holmes, 2016a, 2016b; Törneke, 2021) but these also are either abstract and measured indirectly or hide the “social” functions in abstract notions such as “relational frames,” “emotional variables,” or “motivational variables.”

We are using Social Contextual Analysis here because it has stronger details outlined between uses of language and a person’s social and societal life contexts, details drawn from social science research, especially in regard to analyzing “beliefs” as active verbal behaviors with audiences (Guerin, 1994, 2016, 2020a). This is not inconsistent with most of the those listed above, but the details provide more opportunities for analysis, observations, clinical applications, and research.

The fundamental point for a discursive contextual model is that we need to change how we think and talk about “beliefs” in the first place. The model needs to change from the standard account: that through experience, reasoning and logic, people build up “inside them” a store of “facts” or “propositions” about how they see the world, and these are their beliefs, which can be changed through arguments, persuasion, and reasoning.

The basic approach of most discourse analyses (Edwards, 1997; Edwards & Potter, 1993; Eggins & Slade, 1997; Gee & Green, 1998; Kitzinger, 2000; Potter, 1996) and social contextual analyses (Guerin, 2003, 2016, 2020a) is that language is a strategic behavior or response that people use to do things to other people (never to the world directly). The analyses that flow from this are not about whether what is said is true or not, but about what that language does in its social context: who is involved, what is at stake, what strategies are used, and what are the outcomes that affect future discourses (cf. Hymes, 1974). In this way, beliefs can be thought of more as tools for people to carry out their social behaviors than propositions or “rules” stored in the head.

People use language to navigate their lives, attain resources, develop and maintain social relationships, and avoid social relationships, and these can be done using common sense statements, propositions, cajoling, humor, compliments, casual banter, arguments, attributions, delusional beliefs, or word salad. The point is that discourses have effects on people who speak that same language whether or not they are true, and whether or not they even make sense. (cf. the “My head is a lettuce” word salad example in Guerin, 2020b; and Galletly & Crichton, 2011). As we will see later, language that does not make sense or is patently untrue can still be effective in affecting other people’s behaviors, but the strategies for using such discourses are very different.

A first important point from discursive contextual analysis is that the language we use comes from our communities, society, media, groups, family and other social persons, and we do not invent it as individuals through reasoning (e. g., Farr & Moscovici, 1984). A second important point is that we end up speaking with what has been shaped through our interactions with people and groups. This does not mean that we just go along with everything our groups say, because we can be shaped into using self-language to present our “selves” (Mead, 1934) by taking the opposite line to our groups, or by presenting our “self” in a negative way (both common strategies in life). The key thing to analyze, therefore, is how the person has been shaped by their discursive communities into these ways of talking and what are the concrete and usually observable social outcomes.

A third point to follow is that to change our discourses we therefore need to change the social shaping by the relevant groups if this is possible. The person needs new groups to talk to, with new concrete outcomes from their talk, new opportunities of talk, or new support (shaping) for saying new things from new people (including a therapist). They do not need to be persuaded or “talked out of” what they are currently saying, which often just leads to a competition between persuasion and resistance.

A final point is that the “power” for any talking or writing to actually affect someone or change what they do does not arise from the words themselves. Any “force” of language comes from the social relationships between speaker and listeners, writers and readers, and the histories of those material social relationship exchanges. There are now large literatures in psychology, sociology, and elsewhere on the social properties of different language forms and the contexts in which they might be used (summaries in Guerin, 1997, 2003, 2004, 2016).

The Contextual World of Those with “Mental Health” Behaviors

To do discourse analysis we therefore need to know about the person’s world or life contexts, especially their social and cultural relationships, but some general points are already known for those with “mental health” diagnoses (summaries in Guerin, 2017, 2020c, 2022; Johnstone & Boyle, 2018). Research has shown that the contexts of individuals with “mental health” diagnoses are usually bad life situations involving multiple negative events. Many people have proposed that “mental health” behaviors are shaped when people are trying to deal with extremely bad life situations, such as living with traumatic events, abuse, poverty, threats of all sorts, oppression or severe behavior restrictions, violence, etc. (Guerin, 2020c; Boyle & Johnstone, 2020; Johnstone & Boyle, 2018; Read & Sanders, 2010), and not because of any brain disease, chemical imbalance, or “cognitive dysfunction.” People are adapting to their bad worlds to get along and survive but this is not working out well and their opportunities and their responses are very limited in the bad situations. In such circumstances, few of their behaviors have any real effect to change anything, including their uses of language which is a primary way that humans normally deal with life situations.

Delusional behaviors in particular, for example, have been shown to correlate with unemployment, migrant-status, low-income, low-education-achievement, living-alone, use of illicit drugs and alcohol, trauma, and abuse (Degenhardt & Hall, 2001; Janssen et al., 2004, 2006; Scott et al., 2007). Stigmatization and marginalization of individuals with these life-contexts further disenfranchises them from opportunities and limits their ability to exchange with others. Thus, one can view these language events (including attributions, talk, delusions, and thoughts) as sophisticated and/or desperate strategies in bad life situations, just like those used by all others, to gain social relationships and resources (cf. Galletly & Crichton, 2011). Victor Frankl (2006/1949) famously wrote this of mental health, that “An abnormal reaction to an abnormal situation is normal behavior.”

The practical upshot from a discursive contextual analysis is that when we hear any language being used, we must not ask whether what is being said is true or false, but ask instead what life contexts have shaped those pieces of language use—How do they fit into the person trying to engage in their life paths or merely trying to survive their situations?; What do they do to listeners?; What have been the outcomes? For those who have been labeled with “mental health” labels, their behaviors are likely to have been shaped by very bad life situations and opportunities, and, in particular, their uses of language have likely become exaggerated because they have been unable to get effects in their social relationships, which are typically difficult, complex, or both.

A Discursive Contextual Model of Delusional Behaviors

We will suggest two specific ways to analyze delusions as strategic discourses shaped when living in a bad or difficult social world. The first way expands on the mainstream clinical literature and though it also treats delusions as beliefs, the way we conceptualize beliefs is radically altered from the mainstream clinical literature utilizing discourse analysis and contextual/behavioral analyses. The second treats delusions more originally and specifically as forms of storytelling gone wrong.

Applying Discourse Analysis to Delusions as Beliefs

The first discursive analysis we can make is to treat delusions as beliefs, but with the understanding that beliefs are just verbal behaviors that have been socially shaped to use as tools in managing our social relationships and lives. The questions for delusions then become: What makes such a belief useful?; How can such beliefs be strategically used; how are the properties of such beliefs (as opposed to statements labelled as “attitudes” or “opinions,” for example) be made to work?; and How have the specific contexts shaped these specific beliefs? Here is a typical paranoid delusion treated as an internal “belief”:

Erica decided to get a coil fitted. Except this was no ordinary coil. She only realised as she walked away from the sexual health clinic that they had fitted a camera in her womb. This made perfect sense. It was MI5’s way of tracking her. (Filer, 2019, p. 26)

It would be claimed by mainstream psychology that Erica “has an internalized belief” that MI5 is following and tracking her, and that she (roughly) attributes many otherwise innocuous events in her life as stemming from this belief, and that she reasons in a faulty way to explain such events in her life. Traditional approaches could also add that she does this to protect her “self-esteem” by attributing responsibility for difficult events in her life to this external agent (MI5), or that she has jumped to a single (monocausal) conclusion (about the MI5) and has not considered or attended to other perhaps contradictory explanations (see Table 2).

Once conceptualized in the traditional way as an internal delusional belief, the obvious therapeutic approach is to try and talk Erica out of this belief, usually by presenting alternative scenarios or reasons for the MI5 not being involved or giving her observations to prove it wrong (e. g., CBT). But we will assume though that she “holds this belief” with certainty, is resistant to change, she has no history or life context of the MI5, and this belief is unusual for her social groups (Table 1). With these properties, such a persuasive approach is unlikely to succeed.

If, on the other hand, we apply the present discursive contextual model to the quote above, we must treat this “belief” as a way of doing things to people in Erica’s social relationships. From the brief information given, we are not sure why she would have been shaped into using such beliefs. We would need to look to her bad life contexts and probably poor social relationships for more information, with her probably having difficulty getting heard or acknowledged in her main life social contexts, which would be a situation that would shape her telling more and more extreme forms of beliefs to get at least some effects (consequences) from her listeners, even if not ideal effects.

What we can look at more definitively, however, are the social properties of her beliefs and how these are likely to play out in her social interactions. Table 1 in the right-hand column lists how the standard properties of “delusional beliefs” (left-hand column) might equate to discursive contextual properties (Guerin, 2003, 2004, 2016).

  • First, rather than her belief being strong “inside her,” her presentation of these beliefs in a strong manner has clear social/discursive properties. Many properties might be idiosyncratic for Erica’s life and history, but some general ones are that listeners are less likely to challenge what is said if it is presented strongly, and this can be used to make bluffing easier. If her life is full of difficult or complex social relationships, this is a useful property when presenting her beliefs in conversation and she will also at least get some responses from any strangers she tells (including an ever-patient therapist) if her normal audiences still ignore or challenge her.

  • Second, presenting her beliefs in a way that listeners recognize she is resistant to change them is also a great way to bluff and be heard, with listeners also less likely to challenge if it is clear the beliefs will not be shifted (except for persistent therapists).

  • Third, having sudden or new forms of beliefs can itself have useful discursive effects on the person’s regular audiences, showing that the speaker has gone beyond their circle of acquaintances and is learning new topics they do not even know about; the listeners are also less likely to have standard or rehearsed rebuttals or rejoinders to such new and surprising belief presentations, so their attention can be engaged more by the speaker and challenges be less successful. This will be an especially potent strategy when the person (“with” delusions) has spent a lot of time thinking about how to answer all the common sorts of responses that might be made and has prepared retorts to these (as most of those with chronic delusions have done).

  • Fourth, having unusual or out of context beliefs shares some of the properties of the third point, but in addition this also means that presenting these beliefs can work on groups of strangers just as easily as acquaintances and family; listeners are less likely to understand and so might ask for further information and hence engage with the speaker for longer and prolong the conversations; such beliefs can also be used easily and quickly with any new audiences.

  • Fifth, using unusual beliefs shares most of the properties of the fourth point above, by putting listeners at a disadvantage in gaining control of the conversation, including the common social strategy of just exiting from difficult conversations.

  • Sixth, for the delusions (especially paranoid delusions) that directly involve the person (“MI5 is tracking me” rather than, “MI5 is tracking us all”), this form also has some distinctive discursive properties which can be strategic in social interaction. Firsthand accounts are more persuasive because listeners have less knowledge than the speaker to contradict or challenge what is said (Beattie & Doherty, 1995; Guerin, 2016). Although firsthand accounts might not engage the listener as much as one that includes them (“MI5 is tracking you as well!”), it does mean that the speaker can appear more certain and knowledgeable. This is also part of how talking about “self” (Table 2) can get messed up and exaggerated.

So, although the details require personal interaction and a contextual history with Erica, the point is that these properties of delusional beliefs (left-hand column) can be easily seen instead as properties of discourse and conversation that can help Erica engage with people (right-hand column). In a usually negative world of people not listening, not responding, always challenging, and dismissing whatever she says, engaging in beliefs with these “delusional” properties opens the way for longer conversations with any listener at a disadvantage (especially strangers), and gaining more attention for whatever this is useful for within their world.

A final point about the properties of delusional beliefs is that beliefs that involve the person themself are usually the ones reported as delusions, but a discursive analysis would predict that all types should be found, but that they will have different strategic uses and outcomes (Table 4 will have more on this). The category of paranoid beliefs, in fact, signals that the belief is something about the person with the belief (they are being chased, watched, etc.). But other beliefs more like conspiracy beliefs (we are all affected) are also likely to be prevalent amongst those living in bad life situations but they might not get discussed as delusions because they are more difficult to tell from truth when the person themself is not directly involved.

Applying Discourse Analysis to Delusions as Storytelling

Most delusions are complex and well thought out, and do not appear as a single reported “belief,” although they can certainly be presented and interpreted that way. Rather than being presented in conversation as simple stated beliefs, they are more likely to be part of telling a longer story involving several “beliefs.” This means that we must also examine the conversational properties of using delusions as part of telling stories, rather than only as stand-alone beliefs.

Telling stories as part of any discourse has a long history and, in terms of discursive properties, is a double-edged sword (Bakhtin, 1984; Baumann, 1977, 1986; Baumann & Sherzer, 1989; Doerfler et al., 2013; Guerin, 2004, 2016; King, 2003; Nash, 1990; White & Epstein, 1990). If you can sustain a long story, rather than just stating some isolated beliefs, then you have sway over the conversation and can engage your listeners for longer and in other ways, and this has many useful social properties. However, in order to sustain telling a whole story, other story properties are necessary in your telling, such as having subplots during your story or ways of entertaining your listeners along the way.

Table 3 presents a few of the social properties of storytelling as a discursive form (Bakhtin, 1984; Baumann, 1977, 1986; Baumann & Sherzer, 1989; Carranza, 1999; Charlick-Paley & Sylvan, 2000; Cheshire, 2000; Doerfler et al., 2013; Guerin, 2004, 2016; Humphreys, 2000; Jackson, 2002; King, 2003; Kwansah-Aidoo, 2001; Langellier & Peterson, 1993; Mumby, 1987; Nash, 1990; Robinson, 2000; Schiffrin, 1990; Shuman, 2005; Silverstein & Cywink, 2000; Sunwolf & Frey, 2001; White & Epstein, 1990).

Table 3 Some social properties of storytelling as a discursive form

Of importance for delusions is that stories must be entertaining or attention-getting to sustain a longer part of any conversation and have real effects, and this shapes the way stories are told over time and the content even within one telling. In particular, using attention-getting and unusual features in a story will be shaped by audiences unless the teller of the story has the social status or power within the group of listeners to hold their attention when their stories are boring. Baumann (1986) presented research, for example, which tracked the same storytellers telling their same stories over decades, and showed that as their audiences changed over that period, their stories were also subtlety shaped in length and content. For Baumann’s storytellers, an important reshaping occurred when they stopped telling stories “around the fire” and started telling the “same” stories on stage to larger audiences in community halls (with mostly strangers present).

The important point to learn from this literature about those telling “delusions” is that with a life full of difficult circumstances and recalcitrant audiences, stories will be shaped into more and more extreme versions with more extreme details added to sustain the story-telling. This can be seen in “normal” storytelling (Baumann, 1986) and is likely a part of how extreme delusions (beliefs or stories) are gradually shaped over time.

As a final way of looking at the social and discursive properties of telling delusions as stories, Guerin and Miyazaki (2006) compared four forms of stories for their social/discursive properties—rumors, urban legends, gossip, and “serious stories.” Table 4 reproduces three of these with delusional stories added for comparison.

Table 4 Social properties of stories compared across different forms of stories

The point here is that similar story forms such as rumor and urban legends get shaped and repeated because of their particular mixes of discursive social properties, not because of something inherent in the category use or cognitive processing properties. For example, if one takes a rumor and makes it specifically about someone known to both the speaker and listener rather than known in general, makes it less ambiguous, possibly more humorous, and more scandalous or unusual, then you now have what is generally categorized as gossip. If you were to take that same rumor and make it directly about the speaker themself, make it more unusual, less credible, and perhaps make it shorter, then this will now be the type of story labelled delusional. The differences are not categorical or essential but dependent on how we name the particular mix of social or discursive properties which are included by the speaker and have been shaped by audiences. When not focused directly on the storyteller themselves, then delusions become close to what are normally called rumors, urban legends or conspiracy theories (Guerin, 2004).

Advantages of Contextualizing a Person’s Discursive World for Clinical Practice

We have presented an alternative way to think about, explore, and deal with delusional behaviors, by treating them as discourses that have been shaped in fairly extreme and negative social contexts (Johnstone & Boyle, 2018). The properties that have historically been used to categorize and diagnose delusions (Table 1) are not wrong per se but turn out to have been shaped as strategic properties for managing social conversations for difficult life contexts. There are clear advantages to this in that such contextual analyses become potentially more observable, because we can observe conversations and the responses and audience outcomes of conversations, even though other research methods might be required (Guerin et al., 2018).

Further, this clears up the long-standing philosophical discussions (Jaspers, 1913/1963) around whether delusions can be said to be true or false, because this issue is no longer of relevance (cf. Bentall, 2018; Georgaca, 2000, 2004; Murphy et al., 2018). The criteria of falsity or implausibility have been particularly questioned by critics who highlight that the distinction between plausibility and implausibility is almost impossible to resolve.

Finally, once the truth or falsity of delusions is dispensed with as irrelevant, practical clinical treatment can move away from trying to persuade or convince those speaking “delusions” out of those beliefs, and into more concrete ways of changing those audiences who have shaped such discourses and changing their bad life conditions. What is clear from this article, and a discursive contextual approach to delusions, is that rather than trying to change the delusion itself, through stopping, replacement, contrary evidence, or logical persuasion (CBT), the treatment needs to change the person’s discursive practices and audiences along with their bad life contexts: if we can change their relevant social contexts so they can begin to get some sort of effects or consequences from their language use, then the delusions will fade or “drop out” rather than stop or be “cured.”

Table 5 presents the main points that can be used by clinical psychologists and therapists to contextualize delusions as beliefs and stories, and the sorts of questions that can elicit something of a person’s current and historical discursive practices. These can provide new ways to explore working with people who are currently diagnosed with delusional behaviors, using contextual methods whether for clinical practice or research (Guerin, 2022; Guerin et al., 2018).

Table 5 Contextualizing delusions as beliefs and stories, and the sorts of questions that can elicit something of a person’s current and historical discursive practices

The basic approach of any contextual therapies should be to reduce the person’s stress if they are in crisis, and after contextual analyses, work to change their bad life situations (Guerin, 2022). The latter is often difficult, especially when chronic or societal bad life conditions exist, and reliance is therefore usually made on talking to a therapist. But this counts as a new audience for the person and though only an intermittent and small audience, the therapeutic talking does have discursive properties of distinctiveness and professional consensus that might help the client change their talk to their usual audiences (Guerin, 2022).

For those with highly verbal issues, such as anxieties, delusions, and “eating disorders,” changing audiences can be expanded beyond just the therapist, and work towards establishing new audiences with more useful responses to the client’s discourses (Guerin, 2022). Some steps in doing this might be as follows, but many other ways can be tried (cf. Schwartz & Goldiamond, 1975).

As a first step, after working to change the person’s nondiscursive world, the person’s discursive life context should be found by letting them tell their story, prompting for how conversations and making requests has worked (or not) for them with the different people in their lives. How have they been able to talk during their lives and with what outcomes?

A second step is to provide them with new ways to use their language and get real effects from their words. This might be by therapists simply doing something they ask, or by directly telling them how something they said made them feel (Guerin, 2022).

A third step, if their usual social relationships are unlikely to change, is to support them making social connections with new people (probably an inadvertent function of peer-support groups) but ensuring that they can talk, tell stories, and ask requests of people that are reasonable and that get some effect. For example, sociolinguistics has shown that “adjacency pairs” are common effects for normal conversational requests (Clark & Schaefer, 1989), such as questions will be followed by answers, compliments by thanks, etc. These could be embedded into therapy and rehearsed, and sociolinguistics has numerous other ways that could also be pursued in clinical practice (see summary in Guerin, 1997).

A fourth step would be to practice conversational strategies for when things go wrong in real contexts, with either new audiences or with previous social relationships. The person needs a way of responding (as we all do) when a listener is oppositional or recalcitrant, other than by exaggerating what is being said even further. Some good examples (inadvertently) are given by (Dee, 2012).

Overall, people stating delusions need to find new opportunities to engage with new audiences and new contexts where they can talk in more varied ways but, most important, that their new conversations have a discernible useful effect on their listeners in some way. Some of this is probably done inadvertently inside and outside of therapy currently, with natural changes in the person’s life contexts, and also through the therapist being a new audience who responds to all their talk unconditionally and does not bully, neglect, or avoid the conversation (Guerin, 2022). Once again, the goal is not to stop the delusions but to give the person new social contexts for conversation and storytelling in which telling delusions is irrelevant and will slowly extinguish.

In these ways, the responding to clients is the treatment (Guerin, 2022). The main changes for therapy are that the importance of therapy is not in any direct “treatments” given but from the way the client is treated within normal and fair social conversations. Their talk gets a more normal effect from such exchanges with a therapist, and they do not need to pursue getting conversational effects by exaggerating, becoming louder or more extreme, or by shocking the listener. From this, the social and discursive contexts will be changed, and the delusions will not be “cured” but they will no longer be necessary in the person’s life and seem to magically disappear.