Reconceptualizing Delusion: Strategy, Dysfunction, and Epistemic Injustice in Psychiatry

2.1 Predictive coding and adaptation

John Matthewson and Paul Griffiths (2017) identify four ways in which a functional trait might go wrong, one of which is particularly instructive here: a trait does what it is supposed to do, but its Normal conditions for doing so are ones where something else has gone wrong for the organism (2017, 454-5). This is likely the kind of picture Garson has in mind when he speculates that delusions might have “an adaptive or functional role in compensating for perceptual abnormalities, or for yielding an appearance of meaning in a seemingly absurd or cruel world” (Garson 2022, 12). Perhaps forming a belief, as against not forming a belief, is adaptive, lest one live in the paralysis of uncertainty. This idea seems to be suggested in Garson’s discussion of Snyder when he says “[d]elusions are the only thing that enable me to continue to navigate my environment; the delusions rescue me from madness” (Garson 2022, 226).

Some predictive coding approaches are a natural home for thinking in these terms. Indeed, the idea that delusions are biologically, rather than merely psychologically adaptive has only been defended in this context.3 As we have seen, some versions of the predictive coding approach focus on the importance of resuming learning and engagement with the environment. On these approaches, delusions are adaptive insofar as they maintain behavioural interactions in the face of abnormal prediction-error signaling (Fineberg and Corlett 2016; Mishara and Corlett 2009). Or, as Garson puts it, delusions could be seen as “an attempt to buffer the mind from events or perceptual experiences that would otherwise totally disrupt our ability to get around the world” (Garson 2022, 252). We see then a route from strategy to adaptation.

We make two points here. First, as noted earlier, it is more common for predictive coding accounts of delusion to cast them as by-products of a single dysfunction (Lancellotta 2021, 56). Accounts which cast them as adaptive responses which resume learning represent deviation from these “standard” predictive coding accounts. That’s no objection of course, but we might ask which of the dysfunction or adaptation versions of the predictive coding approach to delusions is to be preferred. Lancellotta (2021) has argued that the dysfunction approach is superior to the adaptation account for several reasons (including that the former is simpler and more compatible with the available empirical evidence).

In addition, some philosophers who think delusions arise from doxastic dysfunction have identified prediction errors as the site of such dysfunction (Miyazono 2019, 65). Recall that someone who understands delusions as adaptive strategies needn’t deny that they are responses to dysfunction elsewhere (e.g. in perceptual mechanisms). However, predictive coding accounts are built upon a denial of any sharp distinction between perceptual and doxastic mechanisms. Fineberg and Corlett for example propose a single impairment in prediction error, occurring in three stages: (1) delusional mood, in which “attention is drawn to irrelevant stimuli”, (2) delusion formation, in which “explanatory insight occurs and flexible processing is disabled”, and (3) explaining things with the delusion, in which the delusion becomes habitual and “enables patients to stay engaged with the environment and exploit its regularities” (Fineberg and Corlett 2016, 4). It looks, then, like the error can characterize either perception or belief. Indeed, Fineberg and Corlett themselves note that on their model “top-down and bottom-up processes sculpt one another” (2016, 5), and Corlett and Paul Fletcher (2014) suggest that prediction error dysfunction could result in deficits in both experience and belief. So although delusions could be an adaptive strategy in the face of dysfunction elsewhere (i.e. in perceptual mechanisms), given that predictive coding approaches deny a sharp distinction between perception and belief, it is difficult to isolate the delusional response from the dysfunction which characterises the single impairment in prediction error.

All told, predictive coding accounts which take delusions to be adaptive responses to anomalous experiences break from more standard predictive coding approaches, and face challenges arising from that break regarding theoretical complexity and consistency with the empirical evidence. It might also be difficult to characterise the delusion as formed by mechanisms of belief “performing their evolved functions perfectly well” (Garson 2022, 251-2), given that the approach does not sharply distinguish perceptual and doxastic mechanisms.

2.2 Explanationism and adaptation

Let us return now to explanationism, which could conceive of delusions as strategies to explain anomalous experiences. From here, we won’t find an easy route to the idea that these strategies are adaptive ones. The orthodox position in the empiricist framework is the two-factor theory, which posits (1) an experiential abnormality and (2) an abnormality in belief. A delusion may well be formed as a strategy to make sense of anomalous experience, but it is a bad strategy, and the choice of a delusional content arises from abnormalities in mechanisms of belief formation or evaluation.

Part of what motivates two-factor approaches is that delusional explanations of anomalous experiences are bad explanations. Indeed, Max Coltheart and colleagues suggest that any theory of delusion formation should answer two questions:

Whilst all explanationists agree that the first question is answered by appeal to anomalous experience, the second question is taken by two-factor theorists to be answerable only by appeal to a second factor (an abnormality of belief) (Davies 2009, 72). Some go further, suggesting that delusional explanations for anomalous experiences are not only poor, they’re “unintelligible” (Nie 2023; cf. Sullivan-Bissett and Noordhof 2024, 3-5), or “nonstarters”, and “the explanations of the delusional patients are nothing like explanations as we understand them” (Fine et al. 2005, 160). Delusions may well be strategies, but they are bad ones.

We are not two-factor theorists (one of us has defended the one-factor approach elsewhere, Sullivan-Bissett 2020, 2024c; Noordhof and Sullivan-Bissett 2021, 2023). In our view, the selection of the delusional hypothesis can be explained by appeal to a range of normal (which is not to say good, or rational) ways of forming beliefs that we find across human psychology, and without appeal to doxastic abnormalities. However, something being within a range of statistically normal responses does not make it adaptive, and here we can usefully distinguish between everyday dysfunction and abnormal dysfunction. This distinction allows for a defence of delusions as strategies, without abnormalities in belief, but stops short of the idea that these strategies are adaptive.

Doxastic dysfunction of a minor kind is likely ubiquitous. It is not exactly rare for us to fail to update background beliefs in light of new evidence, miscount, be temporarily forgetful, misuse rules of inference, and so on. We think of these as cases of everyday dysfunctions of belief formation. Might delusions involve dysfunctions of this kind? We think so (it would be strange if delusions were immune to the kind of everyday dysfunction that causes error in belief). Delusions involving dysfunctions of this kind would place them alongside other irrational beliefs. It is also a familiar fact that there can be motivational influences on belief; such influences are most obviously at work in cases of self-deception, but they might also be of use in explaining the formation and maintenance of delusions with welcome contents (something Garson considers on page 169).

Delusions might be conceived of as strategies, but, we say, they are strategies mounted on a host of everyday doxastic dysfunctions and, in some cases, the epistemically inappropriate influence of motivation. A better strategy when faced with the uncertainties of an anomalous experience would be to form a non-delusional belief about the origins of the experience (e.g. I am ill, or I am hallucinating). The anomalous experiences with which people with delusions must deal might represent abNormal conditions for the proper performance of belief formation. However, even on this view, things aren’t going well doxastically for the subject with a delusion, and delusion formation is helped along by various errors in reasoning that might also go into explanations of other kinds of irrational beliefs. It is thus difficult to see how delusions might be conceived of as stemming “from mechanisms that are performing their evolved functions perfectly well” (Garson 2022, 251-2).4

None of this is to deny that the delusional hypothesis may hold gifts not held by the non-delusional hypothesis. And the benefits that may accrue as part of delusion’s sense-making have been recently discussed in a few contexts. For example, Rosa Ritunnano and colleagues have used their case study of Harry to show that some 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). This is no doubt important, both for a more comprehensive understanding of the nature of delusion and its impact on people’s lives, as well as in the clinical setting. But there’s no route here to a claim of adaptation.5

There is also the term epistemic innocence, which Garson suggests is used by Lisa Bortolotti and Ema Sullivan-Bissett “to describe the apparent reasonableness of some delusions, given the experiences that fostered them” (Garson 2022, 260). These authors do argue that the framework of epistemic innocence allows for a more nuanced assessment of the epistemic and pragmatic benefits of delusions, but they also deny that the delusion is overall epistemically good (Bortolotti 2015, 490; Sullivan-Bissett 2018, 924), and have it that delusions “compromise good functioning to a considerable extent” (Bortolotti 2015, 496).

Here then, is the theoretical terrain as we now see it (Figure 3).

Fig. 3 Delusion as dysfunction versus three strategy approaches

Overall, when it comes to understanding the formation and maintenance of delusions, our view is that the distinction between (mere) dysfunction and strategy (the top row of Figure 3) is not where the theoretical goods lie. Rather, we should explore the kind of strategy in play, and, in particular, whether the doxastic dysfunctions prompting or facilitating its execution are adaptive, or involve everyday or abnormal dysfunction (the bottom row of Figure 3). We think this question is more illuminating with respect to the nature and role of delusion, and it is one on which major accounts of delusion formation might be distinguished.

Of course, even if delusions arise from doxastic dysfunction of some kind, they can nevertheless be considered worthwhile epistemic contributions to interpersonal exchange, and we suspect that this is something on which we and Garson agree, but also something which is not kept in firm view in psychiatry. We turn now then to epistemic injustice and delusion. Here we think there is theoretical fruit in the distinctions of both rows of Figure 3. That is, sometimes the distinction between (mere) dysfunction and strategy might make a difference to ameliorating epistemic injustice in psychiatric encounters, whilst other times it is the distinction between kinds of strategy that makes the difference.

3.1 Credibility in clinical interactions

In its most general formulation, testimonial injustice occurs when a communicator sustains an “identity-prejudicial credibility deficit” (Fricker 2007, 28), whereby they are perceived as having deflated credibility due to the operation of some identity prejudice. Consequently, their testimony is dismissed or distrusted. A key way in which identity prejudice deflates credibility is via stereotypes (Fricker 2007, 17).6 Abdi Sanati and Michalis Kyratsous argue that people with delusions are frequently stereotyped as being “bizarre, incomprehensible, and irrational” (Sanati and Kyratsous 2015, 484), and are consequently vulnerable to testimonial injustices in clinical contexts. Here, we focus on irrationality, since this dimension of the stereotype relates most directly to credibility, although we touch on bizarreness and incomprehensibility in our discussion of intelligibility later (sect. 3.3).

According to Sanati and Kyratsous, the irrationality ascribed to people with delusions is not restricted to the irrational delusional belief(s), but is generalised to other beliefs the person holds (2015, 483). Thus, communicators with delusions are stereotyped as globally epistemically irrational. The irrationality that is stereotypically ascribed is not the unremarkable, everyday irrationality that even communicators without delusions exhibit. Instead, the stereotype should be understood as attributing an important difference in the degree or kind of irrationality characteristic of communicators with delusions (Palafox-Harris 2024, 261).

This stereotype deflates perceptions of testimonial credibility, as someone whose belief-system—indeed “their general psychic life” (Sanati and Kyratsous 2015, 484, our emphasis)—is characterised by global irrationality may not be perceived as an epistemically competent or credible communicator. If that’s right, the testimony of people with delusions so stereotyped will not be given uptake in clinical interactions, and delusional communicators sustain testimonial injustices.

Let’s have this basic picture in the background, as we turn to accounts of delusion and how they might impact on testimonial injustices related to assessments of credibility. If we think that what matters is the accuracy of the stereotype deflating credibility, then it is not (mere) dysfunction versus strategy which might make a difference here (because both kinds of approach, taken broadly, are consistent with the stereotype). What matters is whether the strategy involves everyday or abnormal dysfunction.

Let’s start with strategy accounts which appeal to abnormal dysfunction, e.g. an account which explains the belief that a loved one has been replaced by an imposter by appeal to a cognitive deficit in belief evaluation. A conception of delusion that posits abnormality in belief lends theoretical justification to the credibility deficits communicators with delusions sustain in clinical interactions. Firstly, it lends theoretical support to the idea that people with delusions are characterised by irrationality that is different in kind or degree from the everyday irrationality of others. This is because positing an abnormal dysfunction in doxastic mechanisms provides a way to demarcate delusional irrational beliefs (those that are the product of abnormal doxastic dysfunction) from other unremarkable irrational beliefs (those that are not the result of abnormal doxastic dysfunction).7 Thus, a conception of delusions as resulting from abnormality in belief reinforces the idea that people with delusions are irrational in a different way or to a greater extent than those without delusions, precisely because the irrationality associated with delusion can be traced back to an abnormal dysfunction in belief which people without delusions do not have.

Secondly, positing an abnormal doxastic dysfunction in those with delusions might also legitimise the generalisation of irrationality from one site (a particular delusional belief) to the presumed broader epistemic irrationality communicators with delusions are stereotypically taken to possess. It might be thought that an abnormal dysfunction in belief would lead to many of a person’s beliefs being affected.8 Conceptualising delusions as the products of abnormal doxastic dysfunction thereby sanctions the credibility deficits communicators with delusions can sustain in psychiatric contexts, because, on such a conception, people with delusions can be characterised as irrational (and thus, not credible) in a way which is global and remarkable (that is, different in an important respect from everyday irrationality). Therefore, we suggest that accounts of delusion which appeal to an abnormality in belief can contribute to testimonial injustices in psychiatry.

Let us turn to strategy accounts which deny abnormal dysfunction, e.g. an account on which the belief that a loved one has been replaced by an imposter can be explained as an unremarkable (which does not mean rational) response to weird perceptual experience, without appeal to cognitive deficit (or other abnormalities in belief). Such accounts do not support the stereotype that communicators with delusions are globally epistemically irrational, and therefore do not provide theoretical justification for downgrading the testimonial credibility of people with delusions. On an account that takes delusions as strategies for explaining anomalous experience without positing abnormal doxastic dysfunction, the irrationality of delusional beliefs is not different in kind nor in degree from other irrational beliefs. Instead, delusional beliefs can be explained by appeal to a range of normal responses to anomalous experience, utilizing cognitive processes which are “in no important respect different from those by which normal beliefs are formed” (Maher 1992, 262). A strategy approach of this kind does not reinforce the stereotype that people with delusions are globally epistemically irrational. If people with delusions are not irrational in a remarkable way, we avoid pre-emptively discrediting the testimony of delusional communicators on considerations of credibility.

Of course, testimonial injustices involving people with delusions might still occur in clinical interactions even if we take delusions to be strategies facilitated by everyday doxastic dysfunction. Delusional beliefs are, after all, irrational (even if they are irrational in an unremarkable way), and can contain content that is bizarre and difficult to comprehend, such as the belief that I am dead (a variation of Cotard delusion). Delusional testimony is therefore vulnerable to being distrusted or discounted even if we do not take delusions to result from abnormal doxastic dysfunction. Nevertheless, we suggest that if we’re in the business of reducing vulnerability to testimonial injustice, it is not (mere) dysfunction versus strategy but everyday versus abnormal dysfunction within strategy approaches that could make the difference. That’s because divergence on theoretical justification for the credibility deficits sustained by communicators with delusions is found between strategy accounts, not between (mere) dysfunction versus strategy.

However, in clinical interactions, a communicator with a delusion may have her claims dismissed not only on the basis of irrationality (and therefore, of lacking credibility), but because her claims are deemed to be irrelevant or unintelligible. With respect to injustices arising from this, we think that the distinction between (mere) dysfunction and strategy is key. Let us turn to that now.

3.2 Relevance in clinical encounters

Kidd and Carel argue that:

Clinicians, scaffolded by medical institutions and the theoretical framework that informs clinical practice and policy, are the arbiters of what is relevant in clinical encounters. It is the clinician’s job to glean which information is relevant to a patient’s condition from sources such as the patient’s testimony, testimony from others (e.g. family members), and their medical history. Naturally, what exactly is considered relevant to clinical practice will turn on the theoretical framework for health and ill-health that is at work. If we are the papists from the Middle Ages, then the patient’s moral and spiritual character—their vices, their sins—are relevant to diagnosing and treating their malady (Garson 2022, 28-29). If, however, we are Kantians, then we need only investigate which of the mind’s faculties are erring (Garson 2022, 79-94).

If delusions are mere dysfunctions, the clinician’s role would be to diagnose and treat the dysfunction from which the delusion arises. On such an approach, there is little room for the patient’s personal narrative or interpretation. If a delusion is simply the result of a broken doxastic mechanism inside the person with the delusion, then the clinician need not look further than the dysfunction to determine how best to treat them; the patient’s subjective experiences and personal interpretations sit outside the bounds of clinical relevance, and need not be given clinical uptake.

This narrow conception of relevance affects the epistemic agency of a communicator with a delusion in at least two related ways. Firstly, it constrains the scope or degree of her participation in a given clinical encounter, as the clinician alone has “special authority to delineate and treat” mental disorders (Garson 2022, 240). If the rich subjective life of the person with delusions is not clinically relevant in any substantive sense, whilst the clinical interpretation of the subject’s experiences is considered authoritative, the patient’s perspective need not be solicited, appealed to, nor given uptake. Indeed, the extent of the patient’s participation might be exhausted by “confirming biographical details or reporting symptoms” (Kidd and Carel 2017, 181). Thus, people with delusions might be very restricted in what areas of clinical investigation they can participate in—they can report symptoms, but not interpret them—and in the degree to which their participation is sought or taken seriously by clinicians.

Secondly, this narrow sense of clinical relevance delimits the kind of participation a patient can perform. According to Christopher Hookway (2010), participating in collaborative epistemic projects “is not simply a matter of exchanging information” (2010, 156). Instead, participation involves asking questions, putting forward ideas, evaluating explanations, considering possible alternative explanations, and so on. By constraining clinical relevance in a way that excludes patient contributions beyond providing information about symptoms and history, people with delusions are precluded from performing the richer participative role Hookway describes. In so doing, delusion-as-(mere)-dysfunction takes delusional communicators as sources of information for clinical inquiry, instead of as collaborators in the epistemic pursuits of diagnosis and treatment.

Kidd and Carel (2017) argue that “ill persons (…) are typically regarded as the objects of the epistemic practices of medicine rather than as participants in them” (2017, 181). Therefore, a (mere) dysfunction-centric view of clinical relevance treats the person with delusions as an epistemic object (Fricker 2007; McGlynn 2021), or perhaps as a truncated subject (Pohlhaus 2014), but not properly as a participant in clinical encounters. In restricting patient participation in these two related ways, delusional communicators sustain participatory injustices (Hookway 2010). Both of these effects undermine the epistemic agency of the person with delusions, whilst bolstering the epistemic authority of the clinician, thereby consolidating the epistemic privilege of healthcare professionals (Carel and Kidd 2014) and unequal power dynamics in clinical interaction. Thus, (mere) dysfunction accounts of delusion undermine epistemic agency by tightly circumscribing what is relevant to the clinical encounter, to the exclusion of the patient’s rich subjective experiences.

Alternatively, if delusions are conceived of as strategies, the clinician’s role is to “discern the secret purpose that madness is trying to fulfil” (Garson 2022, 1) and “target the situation or event or arrangement” (Garson 2022, 12) to which delusions are a strategic response. This conception of madness carries a much broader notion of relevance to clinical practice, and we think this broadening comes with thinking in terms of mere strategy, regardless of what kind of dysfunction the strategy involves. The personal interpretations and experiences of the communicator with a delusion are not automatically discredited as irrelevant. Rather than being relegated to the periphery of clinical investigation (if not excluded altogether), on a delusion-as-strategy framework the patient’s rich subjective experiences are the very substance of clinical interest: in order to diagnose what the delusion is a response to, the clinician needs to look at what is happening in the patient’s life, how they are experiencing their world, and how they make sense of their experiences.

This broader notion of relevance can scaffold the patient’s epistemic status and agency in clinical encounters. Firstly, it means that more of the subject’s testimony is clinically relevant compared to a (mere) dysfunction account; their personal experiences, interpretations, and meaning-making matters to a clinical inquiry that takes delusional beliefs to be doing something, that is, serving some purpose for the person who holds them, even if that purpose is helped along by everyday or abnormal dysfunction. In this way, the scope and degree of the person with delusion’s participation is not so tightly circumscribed as on a conception of delusion as (mere) dysfunction. Secondly, conceptualising delusions as strategies allows for the patient to perform the richer participative role Hookway (2010) describes. If a patient’s perspective is clinically relevant in a substantive way, then the patient is empowered to put forward their own interpretations, to participate in evaluating alternative interpretations, and to ask relevant questions. In this sense, delusional communicators are collaborators or co-investigators in the shared epistemic project of understanding their delusional beliefs, rather than the objects (or truncated subjects) of that epistemic project. Therefore, taking delusions as strategies for explaining anomalous experience can mitigate the participatory injustices people with delusions are vulnerable to in clinical encounters.

Moreover, we take the broader conception of clinical relevance contained in delusion-as-strategy to be more closely aligned than the narrow sense of relevance on delusion-as-dysfunction with frameworks like phenomenology, co-production, and expertise-by-experience, all of which take seriously the lived experiences of people with psychiatric conditions and promote epistemic agency. A strategy approach to delusion appears a natural ally to these more collaborative approaches to healthcare, which are often actively engaged in the project of ameliorating epistemic injustice in psychiatry (e.g. Ritunnano 2022; Carel and Kidd 2014).

3.3 Intelligibility in clinical contexts

As well as defining standards of clinical relevance, Kidd and Carel (2018, 230) emphasise that a conception of health and ill-health also defines standards of clinical intelligibility, that is, it sets the authoritative language for discussing medical conditions by producing the concepts, terms, and diagnostic categories used to interpret and talk about illness. In so doing, a conception of health and ill-health can hermeneutically marginalise (Fricker 2007, 153) patients. Hermeneutical marginalisation occurs when a group is not afforded equal participation in the creation of shared hermeneutical resources. Patients are hermeneutically marginalised when the interpretative resources available to make sense of and talk about an experience of ill-health are produced by healthcare professionals, rather than by people with similar experiences.

Kidd and Carel argue that aspects of patient experience “cannot gain purchase within an exclusively naturalistic conception of health”, which is equipped to interpret “physiological dimensions of the process of illness” rather than subjective experiential dimensions (Kidd and Carel 2018, 228-229). A similar thing might be said for (mere) dysfunction accounts of delusion. For example, claims of meaning in delusion are not easily understood against a theoretical backdrop according to which delusions simply are, or arise from, dysfunction. When the theoretical conception of delusion is ill-equipped to interpret subjective experience, such experiences might either be dismissed as unintelligible, or shoehorned into the language and style the theoretical conception is equipped to interpret. Kristen Steslow argues that adopting the medicalised language used by clinicians in order to render oneself intelligible involves “forsaking the uniqueness of [one’s] own perspective, understanding, and expression” (2010, 30). In this event, the subject’s interpretation is given clinical uptake, but something is lost in translation. In either case, whether the patient’s interpretations are dismissed as unintelligible, or made intelligible only by translation into ill-fitting medicalised concepts, the person with delusions sustains a hermeneutical injustice (Fricker 2007, 155): they are unfairly hindered in their own interpretative efforts by a theoretical framework that cannot accommodate such efforts.

Indeed, the problem of unintelligibility might be particularly pronounced in cases of delusion, as the lifeworld of those with delusions might be radically altered from that of the clinician, and because delusional beliefs often have bizarre content which is difficult to comprehend. As we have seen, communicators with delusions are stereotyped as not only irrational, but also bizarre and incomprehensible (Sanati and Kyratsous 2015, 484). In this way, people with delusions have to contend not only with the “communicative roadblocks” generated by an unjust hermeneutical environment which privileges the interpretative resources of clinicians (Palafox-Harris 2024, 260), but also the incommunicability of delusional experience itself.

However, conceptualising delusions as strategies to explain anomalous experience might alleviate some of the apparent incomprehensibility of delusional content. Philip Gerrans argues that “delusions often seem to be situated in chains of reasoning which, while incorrect, are intelligible” (2014, 113, our emphasis). A clinical approach that aims to “discern the secret purpose” (Garson 2022, 1) of a delusional belief and uncover the strategy in a delusion makes intelligible the reasoning behind the formation or maintenance of a delusion—reasoning which might be left obscure on an approach focused solely on diagnosing and treating dysfunctions.

Consider Capgras delusion, where someone believes a familiar person has been replaced by an imposter. Thinking of the Capgras delusion as a strategy for explaining anomalous perceptual experience makes the delusion more comprehensible. That is, even if the reasoning is epistemically faulty in some way (owing to everyday or abnormal doxastic dysfunction), we can nevertheless start to see why the particular delusional hypothesis was adopted or maintained. Understanding the strategy behind the delusion might help clinicians in understanding aspects of delusional experience and delusional content which a dysfunction-centric approach cannot make sense of. In this way, delusion-as-strategy can help to reduce the apparent incomprehensibility of delusions themselves.

Nevertheless, delusional communicators might still be hermeneutically marginalised in psychiatric contexts even on delusion-as-strategy, as we can imagine that patients would be prevented from participating equally in the creation of interpretative resources. However, if thinking of delusions as strategies reduces participatory injustices for the reasons we have discussed, then hermeneutical marginalisation may also be lessened. That is, a strategy approach may afford delusional communicators more participation in the production of shared interpretative resources than a dysfunction-centric approach. Moreover, an interpretative framework that takes delusions to be strategies for explaining or coping with anomalous experiences will be better equipped to comprehend patient interpretations of those experiences because of the standards of intelligibility such a framework carries. For example, claims that a delusional belief is personally meaningful are intelligible against a theoretical backdrop that assumes that delusions can serve a purpose for the person with delusions. If a particular patient interpretation is clinically intelligible, then it would not be dismissed or forced into other pre-existing concepts on the grounds of unintelligibility.

We suggest, then, that reconceptualising delusions as strategies reduces the susceptibility of delusional communicators to hermeneutical injustice in psychiatry for three reasons: (i) the standards of clinical intelligibility are equipped to comprehend patient attempts at interpretation and meaning-making, (ii) there is greater patient participation in creating hermeneutical resources, and (iii) seeking to uncover and understand the strategy behind a delusional belief reduces the apparent incomprehensibility of delusions themselves.

Of course, adopting a strategy approach to delusion (and to psychiatry generally) does not make anyone immune from acting in epistemically unjust ways. Moreover, contingent features of healthcare contexts, such as time and economic pressures, unequal power dynamics, and broader socio-political factors might still obtain even if madness-as-strategy was adopted as the underlying theoretical conception of mental ill-health. Therefore, we do not suggest that reconceptualising delusions as strategies to explain anomalous experience would automatically and wholly mitigate the various epistemic injustices delusional communicators are vulnerable to. However, strategy approaches appear better equipped to promote the epistemic agency of people with delusions than (mere) dysfunction approaches.