Delusions have fascinated clinicians, researchers and philosophers for centuries. While the DSM-IV-TR required that delusional beliefs be “false,” “based on incorrect inference” and “firmly sustained,” the updated DSM-5 defines delusions on the basis of their fixedness alone, providing no commentary on their veracity or the reasoning used to reach them.
1,2 This shift in definitional emphasis, from the content of beliefs to the tenacity with which their proponents cling to them, parallels a historic shift in the general approach to delusions and the mechanisms driving them. While psychoanalytic “motivational” models of delusions once dominated, framing delusions as ego defenses protecting against distressing unconscious conflict with their unique content essential to understanding the psychodynamic processes at play, more recent cognitive neuroscience approaches suggest instead that a single unifying mechanism may be responsible for the broad range of delusional beliefs.
3 Within these latter models, it is the way in which the belief is developed and maintained, rather than content of the belief itself, which is most of interest.
William James observed that delusions, in many cases, were “certainly theories which patients invent to account for their abnormal bodily sensations.”
4 In their original 1923 paper reporting the “illusion of doubles,” Capgras and Reboul-Lachaux proposed an “agnosia of identification” triggered by disconnect between cognitive and emotional recognition of faces:
Some faces that [the patient] sees with their normal features, the memory of which is not altered in any way, are nevertheless no longer accompanied by this feeling of exclusive familiarity which determines…immediate recognition… The patient, whilst picking up on a very narrow resemblance between two images, ceases to identify them because of the different emotions they elicit. (Capgras & Carrette 1924)
5
While Capgras went on to propose a psychodynamic formulation of the syndrome based on oedipal conflict,
5 modern studies in patients with what came to be known as the Capgras delusion have borne out his original hypothesis. Using skin conductance response (SCR) as a marker of autonomic activity, Ellis and Young
6 found a SCR deficit in Capgras patients presented with familiar faces, a finding that others have replicated.
7,8 This model of the Capgras delusion has at its core a fundamental abnormality of “covert,” or affective, facial recognition in the presence of preserved “overt,” or visual, recognition. In some patients with this abnormality—although notably not in all of them
9—this discrepancy is explained by way of delusion about a lookalike impostor.
Capgras, a so-called monothematic delusion,
10,11 is the best studied of all the specific delusions in large part due to its relative simplicity and the frequency with which it is encountered. Monothematic delusions in general have lent themselves more easily to scientific study due to their highly circumscribed nature. In addition to Capgras, other commonly encountered types of monothematic delusions include Frégoli, in which strangers are believed to be close friends or relatives in disguise
12; Cotard, in which the individual feels that he or she is dead
13,14; mirror agnosia, in which the individual’s own reflection is believed to be someone else
15; reduplicative paramnesia, the belief that a person or place has been duplicated
16,17; anosognosia, delusional denial of illness (often left hemiplegia after right hemisphere stroke)
18,19 and the related phenomenon of asomatagnosia, in which a body part is denied as one’s own
20; delusional jealousy (Othello Syndrome)
21; delusional fidelity (Reverse Othello Syndrome)
22; and erotomania (de Clérambault syndrome).
23Polythematic delusions, in contrast, involve multiple delusional beliefs, sometimes interrelated, covering a wide range of topics. The complex, multiform delusions professed by famed mathematician John Nash during the height of his psychosis exemplify this phenomenon: Nash believed, at various times, that he was at the center of a secret effort to build a world government and that he would be the Emperor of Antarctica within this government; that his picture was on the cover of
Life magazine disguised as Pope John XXIII; that he was the left foot of God and that God was walking the earth through him; and that he was a Go board upon which a “first-order” game was being played by his two sons while a “second-order” game pitted him, as he put it later, “in an ideological conflict between me, personally, and the Jews collectively.”
24,25The deficit in autonomic reactivity observed in Capgras provides support for a general model of delusions as arising out of a core perceptual anomaly, not unlike the model proposed by James over a century ago. Maher
26 first suggested that a sufficiently abnormal perception alone might be enough to produce a delusion. Noting that this model would not explain individuals with abnormal perceptions who did not develop delusions, Davies and colleagues proposed a two-factor theory
10 which accepted this initial proposition and posited a second abnormality, this time of reasoning: faced with an unusual perceptual experience (“abnormal data”
27), the patient in this model develops an explanation that should be rejected but, due to some second failure, erroneously is not.
While the model’s empiric support comes largely from studies of Capgras patients and has been applied primarily to misidentification delusions, its theoretical framework is consistent with and can incorporate other models of delusion. In the “comparator model” of delusions of control, the primary anomalous data are proposed to be a deficit in using “efference copy” generated by corollary discharges from a motor command to predict its sensory consequences, yielding a subsequent mismatch of predicted and experienced behavior that makes possible a belief that the movement was externally controlled.
28,29 A similar hypothesis regarding a mismatch between predicted and actual experience of inner speech has been advanced to explain auditory hallucinations and related delusions of alien thought insertion.
30,31 In the “aberrant salience” model of schizophreniform psychosis, built on an understanding of dopamine as central to mediating the experience of stimulus significance, a hyperdopaminergic state leads to an inappropriately heightened degree of salience assigned to internal and external events. Delusions arise as a cognitive attempt to explain this powerful, fundamental feeling about an event’s importance.
32 Anosognosia, pathologic unawareness of a neurologic deficit, is a delusional denial in which a failure of normal sensory feedback (e.g., from a paralyzed limb, or from the visual system in the case of Anton syndrome) allows for the development of a belief that should be, but is not, rejected on the basis of overwhelming evidence that it is incorrect.
33 In all of these cases, as in Capgras, the perceptual anomaly can exist without the delusion and thus must not be sufficient to produce it. A second process interfering with the rejection of implausible beliefs must be invoked.
Coltheart, Davies and colleagues propose that while a delusion can be triggered by any of a variety of abnormal perceptual experiences, the second factor common to all delusions is likely a defective “belief evaluation system” housed in the right frontal lobe.
34 Devinsky similarly suggests that delusions arise when an unfettered left hemisphere “creative narrator” is allowed to confabulate explana-tions for experiences without the ongoing “monitoring of self, memory, and familiarity” normally offered by the right frontal lobe. Here, we examine the case for a right hemisphere contribution to delusion production by way of four interrelated lines of evidence relating to its roles in nonverbal communication; perceptual integration; attentional surveillance and anomaly/novelty detection; and belief updating.
The Right Hemisphere and Delusions: A Brief History
When structural or functional imaging abnormality can be demonstrated with delusions, the right hemisphere is frequently implicated.
35–39 Anosognosia has long been observed to occur disproportionately after right hemisphere stroke as compared with left, as have asomatognosia, in which the paralyzed limb is disowned, and somatoparaphrenia in which there is a delusional belief about the true identity or source of the limb.
20 Delusional supernumerary limbs have also been reported with right hemisphere lesions.
40 Delusional misidentification syndromes in particular show a right hemisphere association,
38,41,42 with specific case reports and series associating right hemisphere pathology with reduplicative paramnesia
16,43,44; the Cotard delusion
45,46; the Capgras delusion
35; mirror agnosia
47–49; Fregoli syndrome
50,51; and Othello syndrome.
52,53 In a recent review of 61 case reports of delusional misidentification syndromes associated with specific lesions, Darby and Prasad found right hemisphere lesions in 92% of cases, with right frontal lobe lesions present in 63%.
54 Preexisting bilateral hemispheric pathology likely accentuates this effect: Levine et al, in a study of 25 right hemisphere injured patients, found the existence of preexisting brain atrophy to be significant in predicting a delusional syndrome, with no clear significance attributed to lesion size or location within the right hemisphere.
55 Several functional imaging studies examining Alzheimer’s disease (AD) patients with and without delusions found an association between the presence of delusion and relative right hemisphere (usually right frontal and/or temporal) hypometabolism or hypoperfusion.
56–59One meta-analysis of functional neuroimaging studies examining time perception in healthy controls and schizophrenic patients showed significantly decreased activation of most right hemisphere regions during timing tasks as compared with controls, suggesting a role for right hemisphere dysfunction in the time perception abnormalities of schizophrenia.
60 In a series of small studies examining the P300 component of auditory event-related potentials (thought to reflect conscious attention to a stimulus), patients with psychotic depression and delusional misidentification disorders showed reduced right frontal P300 amplitude, and the delusional misidentification patients showed an additional P300 amplitude reduction in the right parietal region as well as increased P300 latency in the central midline brain region.
61Some authors have reasoned that if right hemisphere underactivity allows delusions to occur, then stimulating the right hemisphere might suppress them. Studies of cold water caloric vestibular stimulation (CVS) suggest that this may be true, at least to some extent. CVS produces widespread, largely contralateral hemispheric activation. Left-sided CVS can resolve, transiently, left hemispatial neglect,
62,63 somatoparaphrenia,
64 and anosognosia
65 after right-sided stroke. Levine and colleagues reported improvement in delusions and anosgnosia in schizoaffective disorder and schizophrenia
66 after left ear as opposed to right ear CVS, and another group reported a case of improvement in conversion disorder following left CVS.
67But even with multiple lines of inquiry providing highly suggestive circumstantial evidence for a right hemisphere role in delusion production, direct evidence and specific neurophysiologic models of the relationship are lacking. Right hemisphere dysfunction remains a slippery suspect: present at the scene of delusions too often to be chalked up to chance, but not often enough to be implicated directly, and at times occurring without any delusion at all. Here, we explore four of the right hemisphere’s purported functions in depth and suggest that these functions, taken together, subserve a right hemisphere-dominated grip on reality that becomes increasingly tenuous the more impaired these functions become. While right hemisphere lesions likely do not “create” delusions per se, and while there is no clear single anatomic location or network to blame when delusions arise, we suggest that these four right hemisphere functions, when intact, work in tandem to provide at least a partial barrier against delusional belief. When one or a combination of these functions fails, delusions may arise.
Pragmatic Communication
While the left hemisphere enjoyed early celebrity status in the mid- to late 19th century thanks to the work of Dax
68 and Broca
69 localizing speech and language there, the right hemisphere has since proven itself to be a major mediator of human experience, at times by way of more abstract, less tangible modulatory effects on cognition, emotion, and verbal and behavioral output. Following an initial awareness of its role in visuospatial orientation beginning in the 1940s,
70,71 what was previously thought of as the “minor” hemisphere has subsequently become known to play a major role in spatial attention,
72 mental manipulation of objects in space
73,74; and body image.
18,75 Previously dismissed as lacking language, the right hemisphere is crucial for social communication, mediating comprehension of emotional content through interpretation of prosody, facial expressions and gestures.
76 It is thought to control spontaneous facial expression of emotion, with the left face shown to express emotion more intensely in healthy people
77 and right hemisphere-damaged individuals demonstrating relative reductions in spontaneous facial expressivity.
78 Individuals with right hemisphere injuries have difficulty understanding verbal humor,
79 idioms,
80 and metaphors.
81,82 They perform worse than left-hemisphere aphasic patients on tests of connotations of words, even while their understanding of word denotation is generally preserved.
83,84 Some functional imaging studies support a role for the right hemisphere in the interpretation of metaphor,
85,86 although others have disputed this.
87,88 Beyond interpreting content at the sentence level, the right hemisphere is thought to play a key role in organizing complex narrative material: right hemisphere-injured individuals have difficulty making correct inferences
89,90; distilling central themes (the “gist” of a narrative) from complex linguistic material
91–93; integrating elements of a story into a coherent narrative
94,95; selecting appropriate endings to jokes
79; and assessing plausibility of individual story elements.
94 Right hemisphere injured individuals have difficulty distinguishing lies from jokes and have demonstrated deficits in theory of mind.
96 Their speech, described by the aphasiologist Myers in 1977 as “copious and inappropriate… confabulatory, irrelevant, literal, and occasionally bizarre,” is adequate at the sentence level of language but fails at its pragmatic function.
90,97While schizophrenia has been associated at various times with both left-
98 and right-sided
99 dysfunction, numerous studies have identified deficits in the pragmatic aspects of communication and understanding nonliteral speech
100–102 and facial expressions,
103 and it has been suggested that this deficit in discourse-level communication may in fact be a core feature of the illness.
104Perceptual Integration
The right hemisphere is thought to play a dominant role in our ability to integrate disparate perceptions into an overall “gist” or “gestalt” comprehension.
93,105 Studies of visuospatial processing using hierarchical visual stimuli—e.g., a large letter made up of smaller letters—have long suggested a model of lateralized function in which the left hemisphere preferentially attends to an item’s component parts (“local” processing) while the right hemisphere attends to the item’s overall contour and gestalt impression (“global” processing).
106–108 A similar phenomenon is demonstrated in music, with the right hemisphere proposed to play a role in global appreciation of melodic contour and meter while the left deals with the more local features of pitch intervals and rhythm.
109,110 Some authors, noting parallels in right hemisphere patients’ visuospatial and verbal deficits, have speculated that these may be two faces of a single central failure of perceptual and ideational integration—i.e., global processing—based in the right hemisphere. Wapner and colleagues suggested that their right hemisphere patients’ difficulties organizing and comprehending narratives might reflect a broader deficit in handling complex ideational materials.
94 Benowitz and colleagues, finding a strong correlation between deficits in verbal story recall and visuospatial organization in right hemisphere injured patients, considered the same.
91 Myers showed a correlation between visuospatial integration and interpretive language ability, hypothesizing that the tangential, overinclusive speech seen in right hemisphere injury might reflect a higher level difficulty with conceptualizing situations and using contextual cues to distinguish relevant from irrelevant details.
97Devinsky, in a discussion of the spectrum of disorders of body image and ego boundaries found in right hemisphere injury, argued for a dominant right hemisphere role in the most fundamental synthesizing task of all: construction of the corporeal and psychological self.
75 Bogousslavsky and Regli described a “response-to-next-patient-stimulation” phenomenon in 11 right hemisphere stroke patients in which these patients were observed to follow commands directed to other patients as though they were directed to them; interpreted by the authors as a variant of perseveration, this behavior might also suggest the presence of impaired ego boundaries in which self and other are not clearly demarcated.
111 Here, too, we find parallels in schizophrenia. Patients with schizophrenia have difficulties with complex visuospatial processing
112,113 similar to those seen right hemisphere patients. Authors have long suggested a primary role for heteromodal perceptual integration deficits in driving what Borda and Sass have called a disorder of “basic-self experience” or “ipseity disturbance”: here, failure to adequately integrate the multimodal sensory experience of existing as a “self” in reality disrupts a patient’s “grip” or “hold” on that reality.
114 Where the hold on reality has been disrupted, delusions can seep in. Postmes and colleagues suggest that such perceptual incoherence creates a “sensory vacuum” into which the brain pours imagined or remembered experiences in an effort to reinstate coherence: “thus, sensory coherence will be restored at the expense of reality monitoring,” and delusions and hallucinations “can be regarded as a ‘solution’ for incomprehensible, incoherent multisensory experiences.”
115Attentional Surveillance, Self-Monitoring, and Novelty/Anomaly Detection
The right hemisphere provides ongoing attentional surveillance of both hemifields in the visuospatial realm
72 and is largely responsible for vigilance and detecting novel or incongruent stimuli across all perceptual modalities.
116–119 It is believed that it serves the same function at a heteromodal, conceptual level as well, providing ongoing monitoring of the self and its relationship to the environment and functioning as what Ramachandran has called an “anomaly detector.”
120 The left hemisphere, focused on processing at the local level, seeks to establish order and consistency between individual features; it is, as Gazzaniga writes, “constantly looking for order and reason, even when there is none—which leads it to make mistakes.”
121 These mistakes create inconsistencies within the explanatory model and between the model and reality which Ramachandran argues are explained away by the left hemisphere until some anom-aly threshold is reached, at which point the right hemisphere “forces a Kuhnian paradigm shift”
120,122 in order to develop an alternate, more workable hypothesis.
There is evidence that the right hemisphere plays a role in novelty detection, and problems with novelty detection have been linked to delusions.
123 Novelty detection is a function attributed to the hippocampus, specifically dopamine-related gating at CA1; now, emerging evidence shows that dysfunction in this novelty detection mechanism is related to psychosis and delusion.
124,125 Notably, delusions correlated positively with the difference of the functional connectivity of the right hippocampus with the frontal lobe, suggesting that alterations of fronto-limbic novelty processing may contribute to the pathophysiology of delusions in patients with acute psychosis.
123Without appropriate salience given to novel and anomalous stimuli, right hemisphere injured patients are inappropriately blasé about bizarre occurrences and confabulate explanations for how these might fit into a previously established framework. In a story retelling task, while controls and left hemisphere injured patients looked puzzled on hearing nonsensical story elements and left them out on retelling, right hemisphere patients not only readily accepted these odd elements but added justifications for them.
94 Rather than being totally insensitive to incongruities in the narrative, right hemisphere patients seemed “at least tangentially aware that something does not fit and yet, are either unwilling or unable to frankly label the anomalous element as such.”
94Anosodiaphoria, the inappropriate lack of concern about one’s illness that can occur with (and often outlasts) anosognosia after right hemisphere stroke, may similarly be understood as a failure to be adequately impressed by the very salient fact of one’s own new neurologic deficit. Studies of insight in Alzheimer’s Disease have shown a correlation between impaired insight and decreased right temporo-occipital perfusion on SPECT imaging
126 and right lateral and dorsolateral frontal cortical perfusion on FDG-PET.
127 Furthermore, AD patients with lower right insula volume have worse awareness of memory (metamemory).
128 In frontotemporal dementia (FTD), patients with right frontal dominant disease present with the behavioral variant of FTD which is characterized with reduced symptom awareness as compared with the nonfluent, agrammatic aphasia FTD patients who have left frontal predominant disease and in whom symptom awareness is more frequently intact.
129Belief Updating
The related tasks of recognizing that an explanatory model has become outdated and shifting allegiance to a new, more workable one comprise the function of belief updating. The frontal lobes facilitate changing cognitive set, with the right frontal lobe in particular dominant for updating beliefs and avoiding repetitive responses.
130,131 The right dorsolateral prefrontal cortex (DLPFC) is thought to play a major role in problem-solving in complex, “ill-structured” situations.
132 Drake and colleagues, in a series of studies in healthy individuals, showed that counter-attitudinal messages were more persuasive, and disagreeing statements more readily recalled, when heard from the left
133,134; they hypothesized this represented increased openness to cognitive set adjustment with relatively increased right hemisphere activation. In studies using mixed-handedness as a marker for relatively stronger nondominant hemisphere influence,
135 mixed-handers were found to be more gullible and easily persuaded
136; more apt to experience sensory illusions
137; more prone to magical thinking
138; and more likely to internalize false personality trait characterizations
139 than strong left- or right-handers. Strong-handers, meanwhile, are suggested to be less sensation-seeking
140; more likely to prefer authoritarianism and conservative politics
140,141; and more likely to retain beliefs in creationism from childhood despite extensive scientific evidence for evolution.
140 Sharot and colleagues showed an increased ability to incorporate new negative information into preexisting belief frameworks after transient disruption of the left—but not right—inferior frontal gyrus with repetitive transcranial magnetic stimulation (rTMS).
142 Cacioppa, Petty and Quintanar, using electroencephalography (EEG) to monitor cortical activity during exposure to pro- and counter-attitudinal beliefs, showed a relative shift in activity from left to right as subjects considered issues for longer periods of time.
143Patients with right hemisphere injuries, predictably, have difficulty updating their beliefs. In 16 unilateral anterior temporal lobectomy patients given a problem solving task, Rausch found that while all patients had difficulty solving problems as compared with controls, patients with left temporal lobectomies were more likely to shift from a hypothesis even when it was correct, while right temporal lobectomy patients tended to maintain a hypothesis even when told it was not.
144 Right hemisphere patients perseverate more than left hemisphere patients on measures of design fluency
145 and number fluency
146 and they have more difficulty suppressing previously learned cognitive sets when switching tasks.
147 On the Wisconsin Card Sorting Test (WCST), an executive function task known to produce strong activation in the DLPFC, particularly on the right,
148 patients with right frontal lobe tumors and patients with schizophrenia made perseverative errors at a rate that was equal to each other and significantly greater than that of normal controls and patients with left frontal and nonfrontal tumors.
149 Another functional imaging study examining WCST performance after head trauma showed an inverse relationship between perseverative responses and metabolism in the right, but not left, dorsolateral prefrontal cortex and caudate nucleus.
150The Right Hemisphere at the Interface of Self, Environment, and Reality
Woven together, these threads reveal a picture of a right hemisphere that is essential for our ability to create and maintain accurate appraisals of mental objects holistically and in context—be they simple visuospatial figures, complex narratives, or the self. Returning to the two-factor theory of delusions, it follows that a primary somatic/perceptual abnormality creates an inconsistency in a previously functional explanatory model, which the relatively preserved left hemisphere does its best to explain while keeping the model intact. Overly drawn to the left hemisphere task of connecting individual dots, the right hemisphere injured patient is unable to appreciate that the picture thus created is bizarre and incoherent. If anomalies are noted, they lack the cognitive and emotional valence usually accorded to strange or surprising occurrences and do not capture the attention the way they should.
Sass and Byrom have described this phenomenon in schizophrenia as an “anything-goes orientation” in which patients “quickly identify, accept and take in stride phenomena that most people would find anomalous.”
151 In the recent past, the default mode network (DMN) has gained attention as a possible mediator of function and dysfunction of ‘real-time’ thought and belief monitoring and constraints. The DMN is associated with daydreaming, imaginative planning, and stimulus-independent reflection, perhaps acting as a threshold between consciousness and behavior. Studies of patients with schizophrenia consistently demonstrate DMN hyperactivation (i.e., impaired suppression) on a variety of cognitive tasks
152 as well as decreased connectivity to task-positive right frontal networks.
153,154 Sass and Byrom proposed that an overactive default mode network (DMN) might be partly responsible for the “hyposalience” attributed to experiences by schizophrenic patients - experiences that should trigger alarm bells for strangeness but nevertheless do not.
151 Gerrans, drawing on studies demonstrating an anticorrelation between DMN activity and activity in task-focused networks, has proposed a model in which the task-negative DMN is inhibited (“supervised”) by right prefrontal task-positive networks; when these fail, a person with a disinhibited DMN is free to generate a range of beliefs across a broad spectrum of likelihood, without the constraints usually imposed by the reality-based, environment-surveying right prefrontal cortex.
155 Unimpressed by major inconsistencies and unable to revise beliefs, patients cling to old explanatory models while simultaneously acknowledging, either explicitly or implicitly, the existence of contradictory information. This may explain, in part, the widely observed phenomenon in delusional patients referred to as “double bookkeeping,” in which the delusion is upheld even while other statements or behaviors suggest that the patient, on some level, knows it is not true.
156Discussion
It is probably not the case that right hemisphere lesions directly cause false beliefs; more likely, without the complex cognitive skill set normally offered by the right frontal lobe, there may be fewer barriers preventing their occurrence. Our experience of reality is mediated in part by the stories we tell ourselves to explain it. With deficits in comprehension of metaphor, difficulty interpreting nonverbal conversational cues, and impaired attention to unexpected events, right hemisphere patients are at a disadvantage when it comes to collecting the evidence they need to build a workable explanatory hypothesis for any unusual or novel experience; and having constructed a hypothesis, they are unable to evaluate its validity in context of their preexisting knowledge. Their tangential, over-inclusive speech and difficulty with organizing complex narrative materials likely reflects a core deficit in filtering out irrelevant data – even when those data are their own memories. Because of their difficulties with nonliteral communication and affective regulation, their friends and loved ones really do behave differently around them, further widening the gap between the patient’s expectation of the world and what is actually experienced and making it even more difficult for the patient to keep up with reality.
It is important to mention here, as part of a complete discussion of false beliefs arising after brain injury, the phenomenon of confabulation. Notably described by Korsakoff as “pseudoreminiscences” occurring in patients with chronic alcohol use, seemingly out of proportion to the degree of cognitive impairment otherwise present, confabulation was historically thought of in terms of memory dysfunction with false or distorted memories arising to fill amnestic patients’ gaps in recall.
157 Berlyne, in 1972, defined confabulation as “a falsification of memory occurring in clear consciousness in association with an organically derived amnesia.”
158Not all amnestic patients confabulate, however, and subsequent treatments of the topic have introduced the importance of ongoing reality-monitoring for effective use of memories relevant to the individual’s current situation. Kopelman first distinguished between “spontaneous” and “provoked” confabulation, arguing that the former likely required some frontal dysfunction in addition to a memory deficit, whereas the latter might be a normal response to the experience of impaired memory.
159 In a subsequent study of 11 brain injured patients, Bajo, Kopelman and colleagues found an association between severity of confabulation and severity of memory impairment and executive dysfunction.
160 Schnider notes that spontaneous confabulation “constitutes a syndrome of profound derangement of thought,” rather than of memory per se, “in which the concept of ongoing reality in thinking and planning is dominated by a patient’s past experiences and habits rather than true ongoing reality; the confabulations are simply the verbal manifestation of the thought disorder.”
161 Noting that cases of spontaneous confabulation reported in the literature invariably involve lesions of the anterior limbic structures, and specifically the posteromedial orbitofrontal cortex (OFC) and its connections, he suggests a role for this set of structures in monitoring ongoing reality and “constantly suppressing activated, but currently irrelevant, memories.”
161 Confabulation, then, becomes a frontally-mediated disorder of distinguishing “now” from “not-now,”
162 in which memories and associations are allowed to bubble to the surface of conscious experience in a disorganized, “incoherent and context-free” way.
159 Gilboa has proposed an overarching failure of “strategic retrieval” of memories in confabulation, of which temporal confusion is one symptom.
163 He describes two interacting memory evaluation systems: an intuitive “feeling of rightness” attached to retrieved memories based on how well they fit with an overall cognitive schema; and a conscious monitoring process that checks these memories for internal and current contextual coherence. The first judgment, “rapid, automatic, and relatively impenetrable to reasoning,” is thought to be housed in the ventromedial prefrontal cortex (VMPFC); the latter system, in the dorsolateral prefrontal cortex (DLPFC). When this system breaks down, confabulation may occur.
163 Cabeza and colleagues, showing increased left prefrontal cortex (PFC) activity on PET imaging of recall tasks and increased right PFC activity on recall tasks, similarly hypothesized a “production-monitoring” framework in which the left PFC generates semantically guided information “whereas the right PFC is more involved in monitoring operations, including the evaluation and verification of recovered information.”
164Whether or not delusions and confabulation in neurologic patients are in fact two distinct processes, or if they are merely variations along a single spectrum, has yet to be agreed upon in the literature. At minimum, it is likely that their mechanisms overlap, and they may interact with each other. Coltheart takes this approach in a discussion of provoked confabulation—instances in which the confabulated content is offered not spontaneously, but in response to some question or task—noting that delusional patients, amnestic patients, and healthy controls all can be shown to confabulate when asked to provide explanations for their own behavior in instances where a good explanation is lacking.
165 Using Capgras as an example, he suggests that the patient, deeply attached to a belief in an imposter but with no plausible conscious explanation for this belief, might confabulate evidence to explain it. Linking this to Gopnik’s discussion of the human “drive for causal knowledge” (the successful culmination of which she likens to orgasm),
166 Coltheart comments that the satisfaction derived from reaching an explanation for a previously unexplained behavior may override any explanatory implausibility: “it is better to have an explanation for a piece of one’s behaviour—any explanation, no matter how bizarre—than to have none.”
165One theme common to discussions of both delusions and confabulation is that of a two-step process in which there is an initial automatic, preconscious, autonomic or affective experience; and a secondary conscious mechanism in which this experience is evaluated with reference to a larger context. Linking these two steps, in most instances, is a thought; but whether the affective experience generates the thought (as in Ellis and Young
6; Kapur
32; and Davies et al.
10), or the thought is experienced as entering consciousness already somatically or affectively tagged (as in Gilboa
167 and Damasio
168) remains yet to be determined.
The ventromedial prefrontal cortex (VMPFC) looms large in all of these conversations. The idea that ‘covert’, affective, autonomic, unconscious bodily processes might surreptitiously influence conscious decision-making has been advanced most notably by Damasio, whose “somatic marker hypothesis” proposes that emotion, as registered in the brain by its association with transient autonomic and visceral changes, alters conscious cognition at an unconscious level.
168 Damasio argues that the VMPFC is essential for linking unconscious, affective awareness with conscious cognition, likely due to its strong reciprocal connections with the hippocampus and amygdala.
168,169 Experimentally, patients with VMPFC injury do not generate appropriate skin conductance responses (SCRs) when shown emotionally charged stimuli.
170 On a gambling task, they cannot adjust their behavior to account for biases that are too subtle to identify overtly but that nevertheless drive normal controls and patients without frontal injury to alter their behavior.
171 Outside of the laboratory, these patients cannot link preconscious emotional awareness and intuition with conscious decision-making, leading to poor choices particularly in the social and interpersonal realm as well as in risk assessment and outcome prediction; in the lab and in life, they continue to play from “bad decks” long after everyone else has perceived a bias and changed course. The VMPFC almost certainly plays a central role in drawing upon the contextual cognitive and somatic “knowledge” held in the brain and body, respectively, in order to interact with the current environment in a way that makes sense. The CA1 region of the hippocampus is monosynaptically connected to the VMPFC
172; with the CA1 region being crucial to comparator function/ novelty detection and evaluation,
173,174 the VMPFC is part of this critical circuit for reality evaluation. It is clear, however, that injury to the VMPFC alone is not enough to produce delusions. The DLPFC is typically implicated in measures of executive function and the right DLPFC in particular has been shown to play a role in suppressing perseverative behaviors, navigating complex situations, and exerting “inhibitory cognitive control on affective impulses, being therefore particularly critical to limiting the influences of impulses in decision making behavior.”
175 It is possible that the DLPFC provides the “top-down” experiential monitoring counterpart to the VMPFC’s “bottom-up” ‘intuitive’ sense that is not accessible to consciousness, with delusions being allowed to arise when there are lesions to both of these cortices or to essential connections between the two. Papageorgiou’s findings of reduced right frontal P300 amplitudes in both patients with delusional misidentification syndrome (DMS) and psychotic depression, but increased midline P300 latency only in DMS, would seem to support this multifocal model.
176 When this injury or disconnection occurs on the right, the right hemisphere-specific deficits detailed above may make it difficult for patients to compensate when interpreting the world around them.
Delusions then, are neither a necessary outcome of right hemisphere injury nor solely dependent on the right hemisphere for their production; but it is significantly easier for them to emerge when failures of pragmatic communication, perceptual integration, attentional surveillance, and belief updating are superimposed on impairments of executive function and preconscious autonomic processing. It is not clear whether any one deficit is more important for the production of delusions than the others, and it may be the case that different combinations of deficits produce different delusional presentations. Future experimental work in this area will continue to clarify specific right hemisphere networks and their contribution to confabulated, delusional, and nonpathologic belief, and further extend our understanding of the fragile nature of our relationship to reality and the remarkable resourcefulness of the injured brain.