Phenomenological and neurocognitive perspectives on delusions: A critical overview

Abstract

There is considerable overlap between phenomenological and neurocognitive perspectives on delusions. In this paper, we first review major phenomenological accounts of delusions, beginning with Jaspers’ ideas regarding incomprehensibility, delusional mood, and disturbed “cogito” (basic, minimal, or core self-experience) in what he termed “delusion proper” in schizophrenia. Then we discuss later studies of decontextualization and delusional mood by Matussek, changes in self and world in delusion formation according to Conrad's notions of “apophany” and “anastrophe”, and the implications of ontological transformations in the felt sense of reality in some delusions. Next we consider consistencies between: a) phenomenological models stressing minimal-self (ipseity) disturbance and hyperreflexivity in schizophrenia, and b) recent neurocognitive models of delusions emphasizing salience dysregulation and prediction error. We voice reservations about homogenizing tendencies in neurocognitive explanations of delusions (the “paranoia paradigm”), given experiential variations in states of delusion. In particular we consider shortcomings of assuming that delusions necessarily or always involve “mistaken beliefs” concerning objective facts about the world. Finally, we offer some suggestions regarding possible neurocognitive factors. Current models that stress hypersalience (banal stimuli experienced as strange) might benefit from considering the potential role of hyposalience in delusion formation. Hyposalience – associated with experiencing the strange as if it were banal, and perhaps with activation of the default mode network – may underlie a kind of delusional derealization and an “anything goes” attitude. Such an attitude would be conducive to delusion formation, yet differs significantly from the hypersalience emphasized in current neurocognitive theories.

THE PHENOMENOLOGICAL APPROACH TO DELUSIONS

The phenomenological approach focuses on delusion as a phenomenon, on its subjective or lived dimension: what it is like to have a delusion. A crucial feature of phenomenological psychopathology is its emphasis on the mode, manner, or form of the experience in question (1,2). The content of an experience, and the supposedly erroneous nature of beliefs presumably asserted or assumed by the patient, are less important than how the delusional world seems to be experienced and what sort of reality or existence the patient might ascribe to it.

Heidegger (3) referred to the latter dimensions of existence as “ontological”, and distinguished them from object-oriented (what he called “ontic”) modes of understanding experience. Delusions often involve a mutation in the ontological framework of experience that can alter the overall sense of existence and the world, including changes in felt reality-status, time, space, and self-experience (4–7).

Phenomenology is acutely sensitive to the potential variety of orientations a “deluded” patient might have. Also, it stresses that delusions may defy ready comprehension or empathy by normal individuals, since they sometimes involve radical changes in grounding structures and assumptions of human experience. Schizophrenia patients are often aware of the difficulty of conveying their experiences and the likelihood of being misunderstood (8). D.P. Schreber, the schizophrenia patient (of a grandiose paranoid sort) who is perhaps the most famous delusional patient of all time, spoke of matters that “lack all analogies in human experience and which I appreciated directly only in part with my mind's eye” (9, p. 123). “To make myself at least somewhat comprehensible I shall have to speak much in images and similes, which may at times perhaps be only approximately correct” (9), p. 2).

Recognizing this diversity and this difficulty in communication and comprehension suggests the need for in-depth, qualitative exploration of the patient's experiences. This requires replacing typical structured interviews with semi-structured formats (10) capable of illuminating how behavioral signs and symptoms are reflective of “alterations of the structures of experiencing” (11, p. 546).

Jaspers and “delusion proper”

The formulation of delusions offered by K. Jaspers, the father of phenomenological psychopathology, set the agenda for much subsequent psychiatric theory and investigation. Unlike contemporary uses of the term, in which “delusion” applies to aberrant beliefs in schizophrenic, paranoid, manic, psychotically depressive, and organic conditions, Jaspers considered “delusion proper” or “primary delusion” (1), pp. 95,98) (also known as “true” delusion) to be characteristic of schizophrenia.

“True” delusions could, thought Jaspers, be distinguished from delusion-like ideas, because the latter could be understood within the context of underlying personality or as exaggerations of normal emotions and affects. True delusions – which involve conviction, certainty, imperviousness, and impossibility as (merely) associated features – require learning about “the primary experience traceable to the illness” (1), p. 96), of which the apparently incorrect judgment is a secondary product.

For Jaspers, primary or true delusion involves a direct and unmediated experience that is un-understandable in light of previous experiences or beliefs. It is rooted in some indescribable alteration of personality (in the sense of a structural basis for subjectivity) or mode of consciousness of the patient: “a transformation in [one's] total awareness of reality” (1), p. 95) that remains “largely incomprehensible… and beyond our understanding” (1), p. 98). This primarily distinguishes it from “normal belief”, “overvalued idea”, and “delusion-like idea” (the latter found in other “psychotic” conditions as well as in schizophrenia).

Most subsequent phenomenological investigations have followed Jaspers’ lead by focusing on delusion in schizophrenia, especially on three types of global change that contribute to the distinctive “incomprehensibility”: delusional mood; disturbance of basic, minimal, or core self; and certain alterations in the apparent reality-status of delusional beliefs.

Jaspers suggested that delusions and delusional perceptions often arise out of feelings of unidentified significance (12). Occurring most notably in schizophrenia, this psychological state involves feelings of strangeness and tension and of suggestive yet ineffable meaning: “perception is unaltered in itself but there is some change which envelops everything with a subtle, pervasive and strangely uncertain light” (1), p. 98). The second feature involves mutations of basic self-experience: “The individual, though he exists, is no longer able to feel he exists. Descartes’ cogito ergo sum (I think therefore I am) may still be superficially cogitated but it is no longer a valid experience” (1, pp. 122,578; see also (13,14). The third feature of “delusion proper” (1), p. 95) or of schizophrenic “incomprehensibility” (less explicitly stated by Jaspers: see 1, pp. 95,99,105,106) concerns the experienced reality-status of the delusional beliefs or quasi-beliefs (i.e., the degree or kind of reality that the delusional world is experienced as having).

Delusions may help to make a kind of rational sense out of what would otherwise be felt as disturbing, and otherwise inexplicable, changes in the very foundations of the experiential world – i.e., out of the torturous sense of things seeming “just so” (disconcertingly specific or precise) or bristling with strange, cosmic, intentional significance (15), p. 61), or of the disconcerting experiences of losing one's sense of self or of living in a shared and objective world.

Among the many interesting phenomenologically-oriented theorists in Jaspers’ wake who wrote on delusions (these include Minkowski, Binswanger, Ey, Rümke, among others (6,14,16)) are P. Matussek and K. Conrad, psychiatrists who brought phenomenology together with Gestalt psychology.

Matussek: decontextualization

Matussek described a “loosening of the natural perceptual context” (17), p. 90) that allows “individual perceptual components” to float free of standard anchoring in commonsense unities or scenes; this, in turn, opens the way to attributions of exaggerated or peculiar significance and a bringing-together of isolated elements into relationships of delusional meaning that may be kaleidoscopic but can become rigidified. The delusional significance is typically “experienced directly as inherent in the object” (17), p. 98), as if rooted in sensory perception.

Matussek describes a distinctive combination of passivity with activity. The patient experiences a rigid gaze, feeling “held captive” (17), p. 94) by the object or objects. Yet he also has an exaggerated “ability” to “focus his attention on such isolated details” (17), p. 93) and may indeed take “pleasure” in this “fixing [of] attention” (17), p. 94), thereby further transforming the perceptual field via a “prolonged gazing” difficult for normal individuals to sustain. One patient stared at a swinging cord of a light-switch so intently that, after a time, he came to feel that not the cord but the wall and background were moving back and forth; this led, in turn, to the thought that the world was coming to an end (17), p. 93).

Matussek stresses decontextualization and framing of isolated features for “apperception of delusional significance” (17), p. 94) and generation of delusional beliefs. The very salience or framing of particular items may suggest that something of particular importance or “weighting” is being indicated, conveying an aura of intensified “symbolic content” (17), p. 95). Dis-embedding and a certain derealization can also allow these emerging significances to link up more easily with other, isolated features or meanings floating outside any standard or practical context, and perhaps also separated from normal sorts of reality-claims. “One is much clearer about the relatedness of things, because one can overlook the factuality of things. They don't exist so one has nothing to do with them” (17), p. 96), said one patient, who reported special, revelatory access to meanings ignored by others and stated: “my ability to see connections had been multiplied many times over”.

Conrad: apophany and anastrophe

K. Conrad (18) offers a compelling account of global alterations in delusional mood. He refers to the first stage of delusion formation as the trema – theatrical jargon referring to stagefright before the play begins. The patient senses that something is in the offing. Most commonly he feels a sense of threat and associated anxiety, depression, inhibition, or indecisiveness (18), pp. 93,105); there may also be anticipatory excitement, exaltation, manic-like euphoria. Despite some difficulties with concentration, there is no clouding of consciousness, but hyper-vigilance, feelings of being hyper-awake (18, pp. 230,249).

What typically ensues in this delusional or pre-delusional mood is the apophany. “Apophany” comes from a Greek word meaning “to become visible or apparent”; it refers to an abnormal, sometimes excruciating sense of meaningfulness – tantalizing but typically unidentifiable – that can affect “internal” (body, stream of consciousness) as well as “external space”. The patient attributes these changes to the external world and searches for clues to render the new unpredictable changes more comprehensible.

Often the patient experiences the world as somehow false, inauthentic, or insinuating, and as referring somehow to himself: “I have the feeling that everything turns around me” (18), p. 161). Conrad uses the term anastrophe (literally: turning-back or turning-inward) to capture this self-referential, introversive, or self-observing quality – what could be termed a form of “hyper-reflexivity” (15).

Apophany and anastrophe are two sides of a coin. Changes in perceived environment (e.g., sense of things being oddly significant, false, or planned) elicit reflection and inhibit spontaneous engaged activity; yet these changes themselves can only occur in the presence of a veritable “spasm of reflexion” (18), pp. 167,199).

Conrad describes progression from subtle to blatant alterations of inner space (“inner apophany”) that occur during anastrophe. The inward focusing actually has an externalizing effect, transforming proprioceptive sensations or internal mental threads into something felt as distanced and alien (15,19). An intriguing example is described in a classic article by Tausk and later discussed by Sass as exemplifying hyperreflexivity (15): the patient Natalija's illness began with mild experiences of estrangement from herself and culminated in a full-blown delusion regarding a distant “influencing machine” that determined her every thought, sensation, and movement; it crystallized the apophany of her inner space under conditions of “reflexive spasm”.

These mutations have been formulated recently as an alteration of minimal or core self or ipseity (the basic sense of existing as a unified and vital subject of experience), with two complementary facets: hyperreflexivity and diminished self-affection – also associated with disturbance in cognitive/perceptual “grip” or “hold” on the external world (20–22). “Hyperreflexivity” describes how experiences normally tacit emerge into focal awareness, where they are experienced as objects separate from self-as-subject, whereas diminished self-affection (also termed diminished self-presence) describes decreased sense of existing as an experiencing consciousness or lived body.

Self-disturbance “destabilize[s] the natural ontological attitude and may throw the patient into a new ontological-existential perspective, an often solipsistic framework, no longer ruled by the ‘natural’ certitudes concerning space, time, causality, and noncontradiction” (11), p. 544), thereby facilitating experiences of the world as staged or mind-dependent, and a grandiose sense of gaining access to deeper layers of reality. Although self-disturbances involve atmospheric and categorically subjective qualities, there is growing evidence that they can be meaningfully measured, thereby bridging phenomenology and scientific demands for reliable measurement (23,24).

Reality-testing and quasi-solipsism

With its sensitivity to kinds of reality subjectively felt or ascribed within delusion, phenomenology is cautious about assuming that delusions are “false beliefs” about external reality. Phenomenologists question the general applicability of the standard “poor-reality-testing” formula assumed in DSM-IV, DSM-5, ICD-10, and much of psychiatry.

Jaspers remarked on “the specific schizophrenic incorrigibility” of (true) delusions, and their peculiar tendency to be associated with irrelevance for action (“inconsequentiality”): “Reality for [the patient] does not always carry the same meaning as that of normal reality… Hence the attitude of the patient to the content of his delusion is peculiarly inconsequent at times… Belief in reality can range through all degrees, from a mere play with possibilities via a double reality – the empirical and the delusional – to unequivocal attitudes in which the delusional content reigns as the sole and absolute reality” (1), pp. 105-106).

But even when delusions reign as “sole and absolute”, it is not clear that they are necessarily experienced within the “natural attitude” – the commonsense orientation that takes objects and persons as objectively real and intersubjectively available (2,25). It is noteworthy that patients who claim absolute confidence in their delusion, may nevertheless not act on its basis. Schreber, for example, often does not make claims about the external or interpersonally shared world, claims that could be supported or refuted by evidence independent of the experience itself. His delusional beliefs are frequently described in a way that gives them a coefficient of subjectivity – as when he speaks of appreciating his psychotic experiences “only in part with my mind's eye” or “inner eye” (9), pp. 110,123,136,157,234,235,312). At times he even makes the solipsistic or quasi-divine claim that “seeing” – awareness itself – is “confined to my person and immediate surroundings” (9), p. 322).

If the delusion is believed in the context of something like a (subjectively experienced) natural attitude, one ought to act in relation to that belief. But if the delusion is felt to be true only for me, in my mind's eye and for me alone (or, at least, only for me and my delusional others), the contradiction is resolved: one need hardly seek evidence for an experience (akin, in some respects, to an imaginary realm) that makes no claim with regard to normal intersubjective reality; one will hardly take action in actuality with regard to what one senses as existing in a purely or quasi-virtual realm.

This provides a phenomenological way of accounting for at least some instances of the famous “double bookkeeping” of which Bleuler (26) spoke. The very unreality of the delusional world may, in fact, be the feature that most recommends it, perhaps providing, in some instances, the deepest motivation for dwelling in the delusional realm – as Sartre suggests of the “morbid dreamer” who thereby escapes the anxieties and demands of real-world experience (15,27,28).

Summary

The phenomenological approach has focused almost exclusively on delusions in schizophrenia, which (contrary to current usage) are sometimes understood to be the only “true delusions”. Phenomenological accounts of delusional mood/atmosphere discuss loosening of perceptual context, promiscuous linkages between de-contextualized elements, a sense of being at the center of things, and alienating hyperreflexivity.

Such a perspective helps to clarify how paranoid, metaphysical, influence-related, bodily, and solipsistic delusions – together with altered experience of self, intersubjectivity, and felt reality – might develop on the basis of overall structural or ontological changes associated with delusional mood and its aftermath. One may wonder whether these constitute heterogeneous features, and to what extent they derive from or reflect some central disruption. The notion of ipseity disturbance (20,22,29) is one hypothesis regarding such a trouble genérateur (see (21,23).

NEUROCOGNITIVE APPROACHES

We turn now to a consideration, from a phenomenologist's standpoint, of contemporary neurocognitive theories of delusion. We focus on prominent recent representatives of two domains: a) the phenomenological hypotheses of minimal-self or ipseity disturbance (20) and hyperreflexivity/alienation (15), introduced above; b) neurocognitive models (30) postulating aberrant salience (31) and prediction error offered by Fletcher and Frith (32) and especially Corlett et al (33), with later consideration of efforts to relate the concept of the default mode to schizophrenia and delusions (34–36).

We consider three, somewhat overlapping, issues: a) the consistency – within certain limits – between phenomenological perspectives and the prediction error and salience dysregulation models; b) the problematic tendency of contemporary neurocognitive (as well as cognitive-behavioral) accounts to adopt an overly homogenizing and insufficiently “ontological” perspective on “delusion”; c) some ways phenomenology might broaden the contemporary neurocognitive vision, expanding the range of testable hypotheses. Given phenomenology's focus on delusions in schizophrenia, we mostly deal with that particular syndrome or disorder.

Compatibility of phenomenology with prediction error and salience dysregulation accounts

The prediction error model assumes that principles of Bayesian inference (method whereby probability estimates are continuously updated as additional evidence is acquired) play a crucial role in human perception and learning. Applied to delusion, the model introduces three postulates.

The first is that the patient experiences an exaggerated sense of strangeness, novelty or surprise – due to “disrupted prediction-error signaling” (37), pp. 2387,2388) – of the significance of what, objectively speaking, are minor discrepancies between perceptual expectancies and perceptual input, thereby “engender[ing] prediction error where there ought to be none” (33), p. 357) (“noise in predictive learning mechanisms”).

The second is that this disconcerting sense of anomaly, whereby objectively “redundant or irrelevant environmental cues” (33), p. 347) nevertheless seem “strange and sinister”, readily gives rise to delusions that serve to account for the sense of anomaly.

The third is that this over-emphasis or over-“weighting” (32) also skews future perceptual and cognitive expectancies (“empirical prior beliefs” (33), p. 347)) in unrealistic, dysfunctional, often delusional directions, because of the distorting influence of input that should have been filtered out as irrelevant or inconsequential (as mere noise) and the attempt to understand the anomalous experience. Because “the errors are false” (32), p. 55), no amount of associated updating of beliefs can ever adequately model the world; hence “prediction errors will be propagated even further up the system to ever-higher levels of abstraction”. Dysfunctions in the mesocorticolimbic dopamine system, right prefrontal cortex, and glutamate transmission may underlie these dynamic processes.

Corlett et al's (33) prediction error formulation offers a rich and stimulating neurobiological account. They aim to provide a comprehensive and unifying framework, postulating “a singular dysfunction” (33), p. 361), or “single factor, prediction error dysfunction for delusion formation and maintenance”, underlying all delusions, across diagnostic groups – a view consistent with a “complaint-oriented approach” (38), p. 221) and the U.S. National Institute of Mental Health's initiatives to focus on symptoms or “problem behaviors” (39), p. 635) and “aberrant systems that implement psychopathology” (39), p. 633) rather than syndromes (40). Our own preference is for transcending this potentially misleading polarity between symptoms and syndromes, by emphasizing how “syndromes” (such as schizophrenia) can embody distinctive global modes of psychological life that may render symptoms (such as delusion) more heterogeneous than they otherwise appear.

The compatibility of phenomenological and prediction error (and related) accounts has been discussed in two recent articles on salience dysregulation and source monitoring as possible neurocognitive correlates of the disturbance of minimal-self/ipseity emphasized by phenomenologists (41,42; see also 43 for an alternative discussion of phenomenology/neurobiology/delusion). Though Corlett et al's (33) model particularly emphasizes “prediction error”, it incorporates both the aberrant salience theory of Kapur (31) and Frith's model of source monitoring (44–46).

Already in 1992, Sass (15) discussed neurocognitive accounts consistent with his emphasis on the “hyperreflexive” and alienated aspects of schizophrenic subjectivity; he linked the latter directly to phenomenological accounts of delusion formation by Conrad and Matussek, and to both Hemsley and Gray's model of disturbed expectancies (“weakening of the influence of regularities” that normally orient and constrain ongoing perceptual processes) (47, p. 18; 15, p. 69); and Frith's (45) diminished efferent “feedback from willed intentions” (neurocognitive feedback indicating to oneself that one's own bodily movement is an intentional action) (15), p. 435) undermining normal self-experiences of possession and control.

Important details of the neurocognitive model have obviously changed over two decades. The psychological processes now postulated are not very different, however, since the key element of both Hemsley's expectancy hypothesis (prominently cited by Corlett et al, (33) and the current prediction error model is “mismatch between expectation and experience” (33), p. 345) and the consequent prominence of (inappropriate) salience and surprise. Corlett et al (48) present their Bayesian prediction error model as an extension, in terms of “underlying neurochemistry”, of the perspectives offered by Hemsley and Kapur. Sass (15,49) notes that this disruption of expectancies would affect experience not only of the external world (Hemsley's original emphasis; Conrad's external apophany), but also of core-self and lived body (Conrad's internal apophany). Although these self-aspects are not stressed in Kapur's (31) account of salience dysregulation, the issue of agency is extensively discussed in prediction-error theory, with prediction errors in self-generated action explaining both delusions of motor control and excessive sense of agency (33).

It is noteworthy, however, that phenomenologists have emphasized (e.g., (15) the potential contribution of disturbed self-experience – especially of lived-body or body-schema (implicitly sensed body-awareness that serves as background to our experience of the world) – to constituting the overall ontological transformations, including those pertaining to the external world, inherent in delusional mood or delusion formation. These ontological transformations, expressed and grounded in distortions of the lived-body (as locus of basic selfhood), are, in any case, necessary considerations for accounting for the “bizarreness” (50) of many schizophrenic delusions and experience more generally, which seem less a matter of straightforward threats from the mundane environment than more foundational alterations of experience of self, lived-body, or world. Indeed, many schizophrenic delusions are closely interwoven with altered bodily ways of being, which they in a sense express (51).

Consider Natalija's classic influencing-machine delusion and Schreber's solipsistic delusional world of “nerves” and “rays”. Each manifests, in a particularly blatant way, the experience of being alienated from one's own corporeal and mental feelings and movements (while also taking these feelings and movements – even one's own subjectivity – as prime objects of attention), together with concomitant derealization of the external world (15,28).

Such delusions might be better captured by perspectives that favor “enactive” or “radical embodied” approaches to cognition (52,53), inspired by phenomenology's – especially Merleau-Ponty's (54) – stress on how world-experience is imbued with the perceiver's own implicit sense of bodily capacities and dispositions. Here there is at least the appearance of conflict with the more intellectualistic-sounding prediction-error formulations, e.g., regarding “false inference” (33), p. 346, emphasis added) and “maladaptive beliefs that misrepresent the world”, with “belief” defined (within the Bayesian framework) as “the subjective probability that some proposition about the world is true” (32), p. 50, emphasis added). The “beliefs” and “propositions” at issue pertain to spontaneous bodily action/perception, yet are described on the model of intellectual conjecture or formulation, reminiscent of the explicit “conjectures and refutations” of scientific theorizing (55; see also 56 and Merleau-Ponty's critique of intellectualism (54)); they seem to pertain to facts within the world rather than to alterations in global foundations or self-world structures. Closer to the spirit of phenomenology (or, at least, its vocabulary) is an attempt to link disturbances of predictive precision with diminished “self-presence” (ipseity) and, in turn, with associated delusions (57).

We see, then, that there is substantial compatibility between contemporary neurocognitive and phenomenological accounts. Phenomenology, however, with its commitment to subjective dimensions, stresses the constitutive role of basic self-experience and ontological aspects of delusional mood as crucial to characterizing delusions.

The paranoia paradigm

Our second point criticizes the tendency to view “delusion” (using this term in the broad, contemporary sense) as a distinct, even modular phenomenon, and to assume the possibility of a unifying account that cuts across diagnostic entities and phases of illness, offering “a unifying explanation for delusions with disparate contents” (33), p. 346) by postulating “a singular dysfunction” (33), p. 361) or “single factor”.

We acknowledge the remarkable theoretical ingenuity and subtlety especially of Corlett et al's (33) hypotheses, based on an overarching prediction error model that views mind/brain as an inference machine. To us, however, the Bayesian prediction error model lacks face validity as an account of (at least) such delusions as those involving profound guilt, death, or bodily disintegration, or wealth and power, that are common in severe depression and mania, or of solipsistic grandeur and metaphysical revelation found in some chronic/withdrawn patients with schizophrenia (e.g., (58,59). Such delusions seem likely to be more decisively linked to different psychological states, not the “powerful and uncomfortable experience” (33), p. 353) of uncertainty and surprise, but rather self-hatred, despair, devitalization, elation, or the possibilities for grandiosity or awe inherent in solipsistic withdrawal. The emphasis on surprise, novelty, or strangeness may apply best to paranoid delusions, representing what we term the “paranoia paradigm”: the tendency to view literal beliefs about external threat or attack as constituting the prototypical instance of delusion.

Paranoid/persecutory delusions frequently occur in schizophrenia and delusional disorders (60–62), which, historically speaking, are the delusional conditions par excellence; they are common in affective psychoses and certain neurological conditions (63). This is one reason why such delusions may play a disproportionate role in inspiring neurobiological and cognitive theories. Another reason concerns the understandability and at least rational possibility (non-bizarre quality) of many paranoid delusions (especially outside schizophrenia): this means that they more readily conform with both commonsense and standard psychiatric assumptions about the nature of beliefs. This understandability, however, may align this kind of delusion with what Jaspers termed a “delusional idea” – something that can be “comprehend[ed] vividly enough as an exaggeration or diminution of known phenomena” (1, p. 577) (namely, anxious scanning and sense of threat).

Some would stress that psychotic delusions are not, in fact, categorically different from what Jaspers termed “delusion-like” ideas, overvalued ideas, and normal beliefs (64). Such a view has been influential in cognitive models of delusions (65), cognitive-behavior therapy (CBT) for delusions (66), psychometric measurement of delusions (67), and neurocognitive attempts to model delusions as “maladaptive beliefs that misrepresent the world” (33, p. 346). The operative assumption is that all such phenomena lie on a continuum, and that all involve claims about the world that are believed by the patient more or less from within the (subjectively experienced) “natural attitude”. There is merit in questioning absolute and categorical claims. But, as indicated, consideration of lived dimensions of key delusional experiences suggests the importance of appreciating certain qualitative differences regarding implicit reality-claims or experiences of reality.

Some delusions, especially persecutory, may well involve “mistaken beliefs” that prompt action. But, as Jaspers’ observations suggest, not all delusions are of this nature. Many schizophrenia patients, especially perhaps withdrawn individuals, describe forms of double bookkeeping or quasi-solipsistic subjectivization (27,28). Patients with grandiose and guilt-based delusions are less likely to act on delusions, despite being more certain and more impervious to counter-evidence (68,69; see also (70,71). Such delusions may pertain to what the patient experiences (albeit ambiguously) as occurring only in another realm where the usual sources of refutation – or motivations for action – are rendered phenomenologically and logically irrelevant, or at least dubious. This begs for a distinction between empirical or ontic delusions and those of a more autistic, solipsistic, or ontological nature (4,15,28,72). Both should, however, be recognized as ideal types, involving transitional and overlapping possibilities rather than dichotomous forms. Cognitive scientists tend to base their models of delusion formation on formal logic and notions about scientific discovery (e.g., (73). But once the limitations of the belief model or “paranoia paradigm” are recognized, the notions of aesthetic or imaginative experience may seem equally or more relevant (27,58).

Thus, at least some delusions, especially in schizophrenia, are not straightforward instances of mistaken belief, in the usual sense of the term, or of “false inference” (33), p. 346) that “misrepresent” objective reality. To recognize this should force us to raise questions regarding underlying pathology (e.g., the role of hypothesized “deficits” or biases in cognitive processes, including supposed “reasoning abnormalities”, “top-down reasoning impairments” or lesions in “belief evaluation” regions of the brain), and also regarding potential psychological treatments: for instance, can proffering counter-evidence really be the true therapeutic element in successful CBT for psychosis (74)?

Expanding the range of neurocognitive hypotheses

The tendency to conceive of delusions in terms of hypersalience and surprise, typically involving a sense of ontic threat from external reality, may be related to a certain asymmetry within the prediction error account.

Both Fletcher and Frith (32) and Corlett et al (33) do acknowledge that prediction error can be awry through overweighting but also through under-weighting or under-emphasizing of prediction error or perceptual anomaly. Thus, Fletcher and Frith (32) speak of “relatively augmented response to stimuli that should be neutral” (32, p. 53), but also of “relatively suppressed response to stimuli that [since unexpected] should be relevant and important”, noting evidence for both in schizophrenia (see also (37). Despite this recognition, their accounts of delusion formation clearly emphasize exaggerated prediction-error signaling: how “excessive bottom-up signaling” (48), p. 517) or “aberrant novelty… signals drive attention” (33), p. 347) or “grab attention” (32), p. 55), demanding “an explanation or an updating of belief… [which in turn] forms the germ of a delusional belief”.

A striking feature of schizophrenia's characteristically paradoxical nature is that schizophrenic persons seem unusually capable not only of finding the banal to be strange (which would accord with hypersalience), but also of finding the strange banal (hyposalience).

Sass (15) cites the neglected work of Polyakov (75) on disturbances of “probability prognosis”. Using tachistoscopic studies, Polyakov demonstrated exaggerated openness in schizophrenia: enhanced ability to quickly identify, accept, and take in stride phenomena that most people would find anomalous, strange to the point of being difficult to recognize. Related tendencies are mentioned by Fletcher and Frith and by Corlett et al, who speak, respectively, of dream-states in which experiences “are often bizarre yet accepted without question” (32), p. 52) and of enhanced capacity, in schizophrenia, to recognize a concave mask rather than “correcting” to the standard convexity of a face (33), p. 352). They do not, however, clearly relate this propensity (for taking anomalies in stride) to delusions.

We suggest that this latter propensity may often foster an attitude – call it an “anything-goes” orientation – that is intimately related to the characteristically schizophrenic loss of commonsense or natural self-evidence (76,77), and thereby to delusion formation as well. An “anything-goes” orientation obviously undermines the overall grounding in habitual, commonsense reality that ordinarily bolsters our ability to recognize and reject that which is eminently implausible; hence this orientation may facilitate the genesis, acceptance, and persistence of objectively implausible meanings and connections that can be central in delusion formation, especially in some “bizarre” and solipsistic delusions characteristic of schizophrenia. The lack of constraints may foster a kind of “pathological freedom” (15), p. 127) that is characteristic of schizophrenia. Recall Matussek's patient: “One is much clearer about the relatedness of things, because one can overlook the factuality of things” (17), p. 96). In this sense it could underlie some – though not all – instances of the so-called “jumping-to-conclusions” style studied by CBT theorists.

This mode of experience – “bizarre-as-banal” – may be associated with underweighting rather than overweighting of the significance of prediction errors, involving distinct patterns of activity in neural pathways and in neurotransmitters such as dopamine and glutamate. Psychologically, it may be associated with passive withdrawal and hyperawareness, perhaps characterized less by overemphasis on than by relative indifference to anomaly. Ipseity or minimal-self-experience seems likely to be altered in this default, anything-goes mode, probably in the direction of disengagement and diminished presence or vitality. Further, this orientation may have some relationship with what has been termed the “default-mode network” (DMN) activity, already found to be related to the generation of positive symptoms, hallucinations, and possibly delusions as well (36,78–80).

The DMN – first identified by Raichle (81) in 2001 – is activated when there is withdrawal from practical, world-oriented activity in favor of self-referential processing, autobiographical recall and mind-wandering (82,83). The network includes medial prefrontal cortex, posterior cingulate/retrosplenial cortex, and left and right inferior parietal lobules (36). Various experts have mentioned a “give and take” between DMN and “task-positive” systems (“reciprocal patterns of activation and deactivation”) (82), p. 1276), since DMN normally suppresses the attention and salience systems or central-executive network systems (which involve dorsolateral prefrontal cortex and parietal regions) activated in situations requiring response selection and working memory that presumably increase weighting of anomalous perceptions (84).

Interestingly, in schizophrenia, there seems to be poor deactivation of the DMN even when such persons are attending to external stimuli (36,84,85), suggesting “reduced engagement with the external world” (82), p. 1276). This finding suggests that a dreamlike, introversive, or subjectivistic orientation can prevail even during ostensible engagement in practical action.

The importance of this sort of introverted orientation is discussed by Sass (15) in relation to the “hypofrontality” (deactivation of dorsolateral prefrontal cortex) often found in schizophrenia. Sass argued that this hypofrontality should be understood not as decline of higher or more abstract functions (a standard view at the time), but rather as withdrawal from “goal-directed action” (15), p. 389), “orientation to the external world”, or pragmatic cognition in preference for “more introverted and detached modes of cognition” (15), p. 556) – modes that could underlie characteristic “transformations of self and world” (15), p. 390) that occur in schizophrenia, including perceptual disorganization, abnormal salience, and introversion conducive to delusions and hallucinations.

Gerrans (34) has offered an original but compatible discussion of the default mode's role in generating delusions. Whereas he emphasizes suppression of “reality-testing”, we speak of an alternate ontological modality in which intersubjective reality is suspended.

Thus, there may be at least two abnormal modalities (“overweighting” and “underweighting” of prediction errors), apparently opposed, both common in schizophrenia and both related to development of delusions – albeit not necessarily delusions of the same type. But what of the relationship between these two modes?

A possibility consistent with Kapur (31), Fletcher and Frith (32), and Corlett et al (33) is that the overweighting would have pathogenetic priority, coming first, and then giving rise to an underweighting that results from fatigue or withdrawal. Neurobiologically, hyperactivation of salience network would be followed by hyperactivation of DMN and concomitant suppression of salience and attention systems. Psychologically, anxious hyperawareness of anomaly would be followed by psychological withdrawal or non-reactivity, perhaps because constant strangeness accustoms one to the strangeness of the strange. But in some schizophrenic patients – perhaps those of a more “disorganized” type and with insidious onset – the “anything-goes” orientation might be more crucial to the initial formation of delusions. Delusions are less prominent in such patients, but they certainly occur, and clearly demand explanation in any comprehensive account.

Another possibility, however, is that a kind of “anything-goes” orientation could have a more intrinsic relationship to the banal-as-bizarre experience. Many patients might clearly recognize or perhaps sense – at some level of implicit but encompassing (ontological) awareness – that a world in which things loom up as salient for no apparent reason (banal-as-strange, due to “overweighting”) is, in fact, a distinctly different world involving different “feelings of being” (86,87) and in which standard assumptions or reality-criteria are suspended. The very fact of these changes would undermine the “natural attitude”, leading to feelings that the world is somehow unreal and therefore no longer subject to standard constraints of logic and realism.

Perhaps, then, even without an underweighting of Bayesian prediction errors, there could be a shift to an “anything-goes” attitude, albeit one that may also differ from the more purely permissive, less anxiety-ridden “anything-goes” of the prototypical disorganized schizophrenia patient. This would accord with some findings indicating reduced functional network connectivity in schizophrenia, especially reduced mutual suppression between salience and default-mode networks (84), such that “systems normally inhibited by the DMN (e.g. the salience system) may slip from its control” (82), p. 1276), allowing DMN and salience system to be activated simultaneously.

Obviously we are in the realm of speculation. These hypotheses derive, in large measure, from patient self-reports and theoretical conjectures grounded in phenomenological philosophy and psychopathology. They are, however, also grounded in the neurobiological findings and hypotheses of contemporary neuroscience.

Summary

The phenomenological perspective is congruent with (and has anticipated) many aspects of contemporary neuroscientific approaches to delusions, including models emphasizing salience dysregulation, prediction error, and hyperactivation of DMN.

A phenomenological approach supports the relevance, for many delusions, of the now-prominent overweighting-of-prediction-error model, while also suggesting other aspects of, and pathways to, delusion that are not addressed by this model. These latter include forms of distorted ipseity or minimal-self and also alterations of ontological dimensions neglected outside phenomenology – such as overall derealization, solipsism, and an “anything-goes” orientation.

In our opinion, the diversity of these factors raises doubts about the wisdom of viewing delusion as a unitary phenomenon – whether in experiential, psychological, or neurobiological terms.