davies two-factor account

8/3/2019 Davies Two-Factor Account

1/27

Monothematic Delusions: Towards a Two-Factor Account

Davies, Martin, 1950-Coltheart, Max.Langdon, Robyn.

Philosophy, Psychiatry, & Psychology, Volume 8, Number 2/3,

June/September 2001, pp. 133-158 (Article)

Published by The Johns Hopkins University Press

DOI: 10.1353/ppp.2001.0007

For additional information about this article

Access Provided by King's College London at 09/22/10 9:09AM GMT

http://muse.jhu.edu/journals/ppp/summary/v008/8.2davies.html
http://muse.jhu.edu/journals/ppp/summary/v008/8.2davies.htmlhttp://muse.jhu.edu/journals/ppp/summary/v008/8.2davies.html


2/27

DAVIES, COLTHEART, LANGDON, AND BREEN / Monothematic Delusions I 133

2002 by The Johns Hopkins University Press

Monothematic

Delusions: Towards aTwo-Factor Account

ABSTRACT: We provide a battery of examples of delu-sions against which theoretical accounts can be tested.Then we identify neuropsychological anomalies thatcould produce the unusual experiences that may lead,in turn, to the delusions in our battery. However, weargue against Mahers view that delusions are falsebeliefs that arise as normal responses to anomalousexperiences. We propose, instead, that a second factoris required to account for the transition from unusualexperience to delusional belief. The second factor inthe etiology of delusions can be described superficial-ly as a loss of the ability to reject a candidate for belief

on the grounds of its implausibility and its inconsis-tency with everything else that the patient knows, butwe point out some problems that confront any attemptto say more about the nature of this second factor.

KEYWORDS: belief, experience, cognitive neuropsychia-try, misidentification delusions, schizophrenia, affec-tive nonresponse, attributional biases, mirrored self

Introduction:Definition and Examples

AMONGSTSEVERALENTRIES for the word de-lusion, the Oxford English Dictionary(1989) lists this:

3.a. Anything that deceives the mind with a falseimpression; a deception; a fixed false opinion or beliefwith regard to objective things, esp. as a form ofmental derangement.

Martin Davies, MaxColtheart, RobynLangdon, and Nora Breen

There is more than one idea here, and thedefinition offered by the American PsychiatricAssociations Diagnostic and Statistical Manualof Mental Disorders (DSM) seems to be based onsomething similar to the second part of the OEDentry:

Delusion: A false belief based on incorrect inferenceabout external reality that is firmly sustained despitewhat almost everyone else believes and despite whatconstitutes incontrovertible and obvious proof or evi-dence to the contrary (American Psychiatric Associa-tion 1994, 765).

There is much in these putative definitions that isintuitively helpful for the psychological study ofdelusions, but equally, they raise many ques-tions. The DSM definition prompts us to ask at

least the following:

Is a delusion always a belief? Might some delu-sions have some but not all of the characteristicproperties of a belief? For example, Currie (2000)suggests that some delusions may actually be imag-inings that the delusional subject misidentifies asbeliefs.

Does the belief have to be false? It may indeed bepart of the ordinary meaning of the term that adelusion must be false. But if a true belief is sus-tained in just the same way as a delusion is sus-tained, despite its implausibility and in the face ofall the available evidence, then it seems that, forthe purposes of psychological theory, it should begrouped together with delusions.


3/27

134 I PPP / VOL. 8, NO. 23 / JUNESEPTEMBER 2001

Does the belief have to be based on inference? Asubject might arrive at a delusional belief simplyby taking a misleading perceptual experience asveridical. Forming a belief in this way does notobviously involve a step of inference.

Does the belief have to be about external reality?Some delusional beliefs concern the subjects ownbody or even the subjects own thoughts.

Does the belief have to be different from whatalmost everyone else believes? If a bizarrely im-plausible belief is formed and sustained in waysthat are characteristic of delusions, then it seemsthat, for the purposes of psychological theory, itshould be grouped together with delusions even ifmany other subjects believe the same thing.

There is research aimed at improving on the

DSM definition. In Figure 1, we display part of aquestionnaire that is being used by Stanley Catts

and a research team at the University of New SouthWales to elicit responses to two draft definitionsof delusions.1 However, it is a highly non-trivialtask to provide a definition that meets the needs

of both psychological theory and clinical practice.Here, we shall not attempt to refine the definitionof delusion. We shall simply assume that delu-sions are beliefs and proceed to some examples.

The definition of delusions

The clinical literature suggests that conventional definitions of delusions have weaknesses. For instance, DSM-IVdefines delusions as false beliefs. However, many authors have challenged the falsity criterion.

We have drafted two definitions, one using the persons subculture as the reference point for deciding on thepresence of a delusion (Definition One) and the other attempting to use aspects of the form and content of thebelief to make this decision (Definition Two).

Please write your comments about each of these definitions in the spaces provided and, after this, draft thedefinition you use in clinical practice.

Definition One: A delusion is a belief not in keeping with the beliefs held by persons of the same cultural andsocioeconomic background. The evidence supporting it is frequently inadequate or bizarre. The belief isresistant to contradictory argument or evidence that would be accepted by a person of equivalent cultural andsocioeconomic background. Irrespective of the plausibility or bizarreness of the delusion, it is held with astrong sense of conviction at least some of the time.

Definition Two: A delusion is a problematic belief that is highly personalised or idiosyncratic. Though the

content of the belief may have some understandable link to the persons psychological background, it is out ofkeeping with the holders knowledge or experience. Usually the evidence offered in support of the belief isinadequate, overly personal, or bizarre. The belief has the quality of being self-evidential and unfalsifiable(irrefutable) either because of the nature of the belief or because of the refusal of the holder of the belief toadmit anything that would falsify the belief. Irrespective of the plausibility or the bizarreness of the belief, it isheld with a strong sense of conviction and reality most of the time.

FIGURE 1: THE PORS QUESTIONNAIREONTHEDEFINITIONOFDELUSIONS

Ancient Examples:Two cases of Misidentification

Ancient literature provides many stories ofmisidentification, including the horrific case of

Agave who is tricked by the god, Dionysus, intomurdering her own son, Pentheus, under theimpression that she is killing a lion (see Way1988: Bacchae, lines 127778, 128183).

Cadmus: Whose headwhose?art thou bearing inthine arms?Agave: A lionsso said they which hunted it . . . .Cadmus: Gaze, gaze on it, and be thou certified . . . .Seems it to thee now like a lions head?Agave: No!Wretched!Wretched!Pentheushead I hold!

Rather similarly, Hercules is duped by the

demon, Lyssa, into killing his own sons, thinkingthat they are the sons of Eurystheus (see Way1988: The Madness of Hercules).

In both of these examples, the misidentifica-tion arises from misperception; the protagonist isdeceived by the gods. This constitutes a delusionaccording to another of the OED entries: The

fact or condition of being cheated and led to


4/27


believe what is false. It is not so clear, however,that these are misidentification delusions in themodern sense of the term, for a false belief thatarises from misleading experience is not neces-

sarily a delusion. Perhaps our intuitive judgmentson these two cases are conditioned by the fact

that there is nothing available to Agave or toHercules that constitutes incontrovertible andobvious proof or evidence to the contrary. Infact, Herculess wife and child do try to persuadehim that they are his wife and children, but heseems oblivious. To this extent, his case seems

closer to a delusional belief.

Varieties of Delusion: Monothematicor Polythematic; Circumscribed orElaborated

A monothematic delusion is simply one that isspecific to a particular topic. As Stone and Young(1997) note, monothematic delusions that resultfrom brain injury are often also circumscribedorunelaborated. Monothematic and circumscribeddelusions can be contrasted with thepolythemat-ic and elaborateddelusions or delusional systemsthat are characteristic of some schizophrenic pa-tients. The following case description refers to apatient who was interviewed by Robyn Lang-don. The patient had polythematic and elaborat-

ed delusions, some of which were interwovenand some of which stood apart. The delusionalthemes included somatic concerns, paranoia, ideasof reference, grandiosity, and loss of boundaryexperiences.

A Thirty-One-Year-Old Woman with Chronic

Schizophrenia. The patient had been ill for sevenyears. At the time of the interview, she reportedolfactory and somatic hallucinations but no au-ditory or visual hallucinations. She noticed anoccasional unexplained powdery smell about herbodysomething like the smell of baby pow-

derand was distressed by the experience ofphysical blows raining down on her head on adaily basis. Marked paranoia was present; sheavoided all contact with her family, believingthat they would harm her if they knew of herlocation. She also avoided public places, being

generally distrustful of other people. She believedthat people followed her with their eyes and

gossiped about her whenever she went out inpublic. The only places that she was prepared tofrequent on a relatively regular basis were thegym (at times when other people were unlikely

to be present) and a local church that she hadstarted attending. She had a history of somatic

concerns. At the time of the interview, she wasslightly concerned that her genitals had beenchanging in shape and size but was not particu-larly distressed, because the changes had appearedto stop with everything going back to normalsince she had resumed going to church. She was

also keeping an eye on one of her teeth; there wassomething special about this tooth. At the timeof the interview, she did not quite know whatthis something special was but was sure that if

she kept watching the tooth, its importance wouldbecome apparent; it was going to be a special

sign for her. She also described a special freckleon her body; this was a freckle that God hadmade especially for her to mark her as his cho-sen. She believed that she had a special relation-ship with God; God had stepped in to save herfrom her enemies, and this was because she had a

mission. That was why she had returned to thechurch: to find out about her mission for God.Finally, the patient reported thought broadcastdelusions; she believed that her thoughts seepedout of her head like signals that could be under-stood by other people. This meant that she had

to be very careful about what she thought; sheshould only think good thoughts.

Just as polythematicity and elaboration co-occur in the delusional systems of some schizo-phrenic patients, so also monothematicity and

circumscription tend to go together, but it isworth noting that monothematicity and circum-scription are, strictly speaking, independent of eachother. In particular, a neurological patient mayelaborate a monothematic delusion to some ex-

tent. Furthermore, although psychiatric patientsoften have polythematic and elaborated delusionalsystems, they sometimes have delusions that aremonothematic and at most only somewhat elab-orated. Here are two more descriptions of pa-tients who were interviewed by Robyn Langdon.The first case is an example of an individual with a

more or less circumscribed monothematic delusion:


5/27


A Twenty-One-Year-Old Man with PsychoticSymptoms. The patient had started presentingwith psychotic symptoms two years earlier dur-

ing his second year at university. At the time ofthe interview, he was receiving, and responding

well to, medication. His only concern was theoccasional recurring thought that he was physi-cally unwell. In the past these concerns had de-veloped into a fervently held hypochondrial de-lusion that he was diseased and would die fromthis disease; this despite doctor after doctor fail-

ing to find anything at all wrong with his body.Now, he said, he was trying to stop these thoughtsfrom getting a firm hold by using reality testing.No other symptoms were present.

The second case is an example of an individu-

al with a monothematic but somewhat elaborat-ed delusional system:

A Thirty-Three-Year-Old Man with ChronicSchizophrenia. The patient had been ill for four-teen years. At the time of the interview, he was

preoccupied and distressed by the firm belief thathe had no internal organs. Although his doctorshad told him that this was a physiological impos-sibility, and despite some acknowledgement onthe part of the patient that he could not quiteunderstand how such a thing was possible, the

patient said that he could not rid himself of thebelief. The patient also expressed the belief thatspirit doctors had come to his room one night toperform a magical operation in order to removehis internal organs. This happened, he believed,because he was being punished by God for some

evil or sin that he had committed, although hewas uncertain about the nature of that sin. Themost distressing aspect of the delusion for thispatient was the pervasive worry that, when hedied, he would be rejected from heaven because

he was no longer a proper human being.

Modern Examples: Eight Monothe-matic Delusions

We now present a battery of eight monothe-matic delusions against which a theoretical ac-count of delusions might be tested.

(1) The first is the Capgras delusion: One of myclosest relatives has been replaced by an impos-tor. Although it is often remarked that the Capgrasdelusion usually concerns someone close to thepatient, Ellis and de Pauw (1994, 320) note, Clear-

ly the [Capgras delusion] does involve the beliefthat some individuals have been replaced by dou-bles or robots. But these are not necessarily peoplein close relationship with the patient. Often neigh-bours, nursing staff, doctors, and other patientsare the objects of what some have termed hy-poidentification. This is one of several consider-ations that tell against a psychodynamic accountof the delusion. One patient investigated by Younget al. (1993) said, Theres been someone like mysons double which isnt my son . . . . I can tell myson because my son is different . . . but you haveto be quick to notice it. The first major report ofthe Capgras delusion was in 1923, and the delu-

sion is not especially uncommon. There are hun-dreds of published cases (Capgras and Reboul-Lachaux 1923; Berson 1983; Signer 1987; Frstl,et al. 1991).

(2) The second delusion in our battery is the Cotarddelusion: I am dead (Cotard, 1882; Berrios andLuque 1995). It is worth noting that, in fact, onlyone of Cotards original eight pure cases involvedthe delusional belief that the patient herself wasdead; more common were self-deprecatory delu-sions and negation of self (see Young and Leaf-head [1996] for discussion). Young et al. (1992)describe the case of a patient, WI, who, following ahead injury, became convinced that he had died. WIsdelusion was elaborated to the extent that, basedon the heat in South Africa where he was traveling,he also believed that he had been taken to hell.

(3) The third delusion is the Frgoli delusion: I ambeing followed around by people who are knownto me but who are unrecognizable because theyare in disguise (Courbon and Fail 1927; de Pauwet al. 1987; Ellis et al. 1994).

(4) The fourth is mirrored-self misidentification: Theperson I see in the mirror is not really me (Breenet al. 2000b).

(5) The fifth delusion in our battery is reduplicativeparamnesia (Breen et al. 2000b). A patient, DB,

who had suffered a right parietal stroke, affirmedthat her husband had died and was cremated fouryears earlier (which was true) and that her hus-band was a patient on the ward in the same hospi-tal that she was in (which was not true). DBseemed to have two separate, and inconsistent,records or dossiers for her husband.

(6) The sixth delusion is one that is sometimes foundin patients with unilateral neglect. Unilateral ne-


6/27


glect occurs in people who have had major dam-age to one hemisphere of the brain (almost invari-

ably the right hemisphere). Since the right hemi-sphere controls the left half of the body, thesepatients normally have paralyzed left arms andleft legs. They show a marked deviation of eyes,head, and trunk away from the left side of spaceas if they are captivated by the right side of theirworld. They are liable to collide with objects ontheir left side, they leave the food on the left halfof their plates, they do not dress or clean their leftlimbs, and they do not respond to conversation ifthe speaker is to their left. Some patients withunilateral neglect actually deny ownership of theirleft arm or left hand, even when it is placed so thatthey have no difficulty attending to it. In one

striking case reported by Edoardo Bisiach, thepatient (PR) insisted that his left arm was theexaminers and accepted the consequence that theexaminer had three arms (Bisiach 1988, 469. Forfurther discussion of PR, see Bisiach and Gemini-ani [1991, 34]. For further examples of denial ofownership of a contralesional limb, see Critchley[1953]. For reviews of unilateral neglect, see Bisi-ach and Berti [1995] and Vallar [1998]). Bisiachand Geminiani (1991, 3233) also provide a vividdescription of a patient, LA-O, who denied own-ership of her left hand:

On request, she admitted without hesitation

that her left shoulder was part of her bodyand inferentially came to the same conclu-sion as regards her left arm and elbow, given,as she remarked, the evident continuity ofthose members. She was elusive about theforearm but insisted on denying ownershipof the left hand, even when it had been pas-sively placed on the right side of her trunk.

She could not explain why her rings hap-pened to be worn on the fingers of the alienhand.

The final two delusions in our battery are

characteristic of schizophrenia (Frith 1992):

(7) The seventh is the delusion of alien control:Some-one else is able to control my actions. I am apuppet and someone else is pulling the strings.

(8) The eighth is the delusion of thought insertion:Someone elses thoughts are being inserted intomy mind.

Since we are offering a battery of monothematicdelusions, we should note that loss of boundarydelusions, including alien control and thoughtinsertion, often co-occur with other delusions inschizophrenia. However, they can also be found in

isolation. In this paper, we are mainly concernedwith monothematic delusions that result from braininjury, but we hope that our account can beextended to cover delusions, particularly mono-thematic delusions, arising in psychiatric patients.

We end this introductory section by noting theincidence rate of delusional beliefs. Approximate-ly one person per hundred will be diagnosed ashaving schizophrenia. Since delusions are first-rank markers of schizophrenia, virtually all thesepatients will have delusional beliefs. Other psy-

chiatric patients who experience delusions might

be diagnosed as having schizoaffective disorder,affective disorders with psychotic episodes, orschizophreniform disorder. The incidence of de-lusions as a result of brain injury is lower thanthe incidence of schizophrenia, but overall, we

can take it that, at a minimum, between one andtwo people per hundred will have delusionalbeliefs at some time in their lives.

1. Delusions and Experience

Brendan Maher (1974, 1988, 1992, 1999)famously defends the view that delusions are

false beliefs that arise as normal responses tounusual experiences. (Maher has been primarilyconcerned with the role of anomalous experienc-es in the etiology of delusions in schizophrenia.)A recent article provides an outline of his ac-count (Maher 1999, 55051):

Capgras delusion

Cotard delusion

Frgoli delusion

Mirrored-self misidentification

Reduplicative paramnesia

Unilateral neglect

Alien control

Thought insertion

Table 1. Eight monothematic delusions


7/27


Delusional beliefs, like normal beliefs, arise froman attempt to explain experience.

The processes by which deluded persons reasonfrom experience to belief are not significantly dif-ferent from the processes by which non-deluded

persons do. Defective reasoning about actual personal normal

experience is not the primary contributor to theformation of delusional beliefs.

The origins of anomalous experience may lie in abroad band of neuropsychological anomalies.

1.1 Mahers Account of PrimarySchizophrenic Delusions

Maher suggests that the delusions that Jaspers(1963) called primary, as well as those that hecalled secondary, can be brought within the scopeof this general account. (See Maher [1999, 552]:

I hope to show that Jasperss distinction be-tween primary and secondary delusions can besatisfactorily encompassed with the single modelalready described.) Jasperss account of prima-ry delusions is problematic in a number of re-spects (see Eilan 2000), but the basic idea is that

a primary delusional belief does not arise fromearlier psychological events in a way that is intel-ligible from the subjects point of view; it ispsychologically irreducible and incomprehensi-ble. A secondary delusion, in contrast, emergesin a comprehensible way from other psychologi-

cal events or, as Jaspers puts it, out of a givenpersonality and situation. He includes amongsecondary delusions, or delusion-like ideas,the transient deceptions due to false perception,etc. (Jaspers 1963, 107) Consequently, it wouldbe natural to classify as secondary a delusion

that arises from an attempt to explain an unusu-al perceptual experience.

Given this contrast between primary and sec-ondary delusions, it is initially surprising thatMaher should seek to apply his account to pri-mary delusions, but Jaspers also speaks of a

general delusional atmosphere (Jaspers 1963,98) and says:

Delusions proper [primary delusions] are the vaguecrystallizations of blurred delusional experience anddiffuse, perplexing self-references which cannot besufficiently understood in terms of the personality orthe situation; they are more the symptoms of a diseaseprocess (Jaspers 1963, 107).

In line with these and other remarks, Maherproposes that the subject of a primary schizo-phrenic delusion experiences feelings of signifi-cance that are endogenously generated but may

become attached to aspects of the subjects men-tal life, such as thoughts and images, and to

perceived features of the external environment.A vivid description of this kind of experience

is provided by a schizophrenic patient, NormaMacDonald:

I became interested in a wide assortment of people,events, places, and ideas which normally would makeno impression on me . . . .I . . . felt that there was someoverwhelming significance in all this, produced eitherby God or Satan, and I felt that I was duty-bound toponder on each of these new interests, and the more Ipondered the worse it became. The walk of a stranger

on the street could be a sign to me which I mustinterpret (MacDonald 1960, 218).

Roderick Anscombe, who has compiled a richcollection of patients descriptions of schizophren-

ic experience, comments: The significance isspurious, but the patients feel impelled to reactby making sense of it in some way (Anscombe1987, 249).

On Mahers account, a subject may offer adelusional explanation for a generalized and per-

sistent feeling of significance (Maher 1999, 560):[I]f no specific concrete change can be found,

and the feeling of significance persists, every-thing must have changed in some fundamentalway. He continues, If everything in the envi-ronment is significant, the significance is likely to

be widespread and profound. Themes of worlddissolution, that people have all become robots,or are the walking dead, that the individual hashimself died and is looking at the world frombeyond life, are some of the common possibili-ties. Alternatively, it may be that when a subject

experiences these feelings of significance, an ap-parently significant train of thought and its asso-

ciated imagery come to provide content for adelusional belief. Thus Anscombe, who suggeststhat feelings of significance are the result of adisorder of attention, says that certain of the

persons own thoughts may be imbued with asignificance that is out of proportion to their realimportance, simply because they happen to cap-


8/27


ture the attentional focus (Anscombe 1987, 252).In either case, the delusional belief then givesmeaning to the sense of significance that attachesto perceived features of the environment.

Maher illustrates how this account of the eti-ology of primary delusions works by considering

the classic example ofa patient who looked atthe marble tables in a caf and suddenly becameconvinced that the end of the world was com-ing (Maher 1999, 559; the example is fromArthur 1964, 106). For this patient, the experi-ence of seeing the marble tables was full of sig-

nificance, shot through with meaning about theending of the world, but, Maher suggests, therewas nothing anomalous about the visual experi-ence itself; there was, for example, no abnormal-

ity in visual information processing. Thus it isnot that the visual experience had some unusual

property that figured as a causal factor in thegeneration of the delusion. Rather, the delusionarose in some way from feelings of significance.Then, as a causal result of the presence of thedelusional belief with its specific content, a par-ticular significance came to be attached to the

perceptual experience of seeing the marble ta-bles. John Campbell (this issue) describes thiscase as involving a top-down loading of theperception by the delusional content . . . .

If we focus on the visual perceptual experi-ence of the marble tables, then the fact that the

significance attached to that experience was caus-ally posterior to the delusional belief makes itinitially surprising that Maher should includethis kind of example within the scope of hisaccount. However, what Maher stresses in thiscase is not the causal role of the visualperceptualexperience but rather the causal role of the feel-ings of heightened significance that coincidedwith viewing the marble tables:

The delusional meaning comes as given by the com-bination of experience [of feelings of significance] andother coincidental elements of consciousness, but it isthe experience [of feelings of significance] that createsthe combination; it is not the delusional explanationthat creates the experience (Maher 1999, 561).

In general, the anomalous experience that figuresin the etiology of primary schizophrenic delu-sions, on Mahers account, is not a perceptual

experience; it is a compelling, generalized, andpersistent feeling of significance.

Lesser feelings of significanceexperienced asthe sense that something has changed, that some-

one is familiar, that some event was surprising,or that something wonderful is about to hap-

penoccur in everyday life (Maher 1999, 552-58). A vague sense that something has changed,for example, normally prompts a search for aspecific difference, and it is a common enoughoccurrence that such a search may end in a quitefalse view about what it is that has changed.

When the feeling of significance is so much morecompelling, generalized, and persistent, is en-dogenously generated, and co-occurs coinciden-tally with various thoughts, images, and percep-

tual experiences, bizarrely false beliefs may ariseand may then condition other aspects of the

subjects experience.This account of primary delusions does not,

of course, rule out the possibility that in othercases, feelings of significance, of difference, or offamiliarity might attach directly to particularperceptions of objects in the environment. The

resulting experiences could then figure in theetiology of delusions that would be classified assecondary. The specific contents of those delu-sional beliefs would be intelligible in the light of thecontent and nature of the antecedent experiences.

Mahers account of primary schizophrenic de-

lusions is also entirely compatible with the ideathat delusions may arise in an intelligible wayfrom unusual experiences that result from braininjury. Indeed, he says

[D]elusional interpretations of circumscribed anoma-lies of experience arising from psychopathology arenot confined to schizophrenia . . . .[T]he model ofdelusion formation . . . posits that the basic origin liesin the anomalous experience, regardless of how thatanomaly arose (Maher 1999, 566).

1.2 Anomalous Experiences in theEtiology of Monothematic Delusions

For the purposes of the argument in this pa-per, we agree with Maher that unusual experi-

ences are one factor in the etiology of monothe-matic delusions and particularly of delusions thatresult from brain injury. Our main claim is that


9/27


there must be a second factor as well. The argu-ment for this claim comes in Section 2 (below).In the remainder of this section, we focus on thepoint of agreement. Is it possible to identify ex-

periences that could figure in a psychologicallyintelligible way in the etiology of the eight delu-

sions in our battery?2

The Capgras Delusion

Hadyn Ellis, Andy Young, and colleagues pro-pose that the Capgras delusion (One of myclosest relatives has been replaced by an impos-tor) arises from a deficit in face processing thatis a kind of mirror image of prosopagnosia (see,for example, Ellis and Young 1990). The facerecognition system is intact, but there is a loss of

affective responses to familiar faces.3 This is aneuropsychological explanation of the delusion,not a psychiatric explanation. It has to do withthe way our bodies respond emotionally to fa-miliar faces compared with the faces of strang-

ers. According to Ellis and Young, the Capgraspatient has an experience of seeing a face thatlooks just like a close relative (usually the spouse),but without the affective response that wouldnormally be an integral part of that experience.This hypothesis about the patients experience is

supported by the finding that Capgras patients

have a reduced galvanic skin response to facesand in particular do not respond more to famil-iar faces than to unfamiliar faces (Ellis et al.1997; Hirstein and Ramachandran 1997). Braindamage has disrupted the connections in the pa-

tients brain between the face recognition systemand the autonomic nervous system. The delusioncan then be explained in terms of the patientstrying to make sense of this peculiar experience:

When patients find themselves in such a conflict (thatis, receiving some information which indicates thatthe face in front of them belongs to X, but not receiv-

ing confirmation of this), they may adopt some sort ofrationalizing strategy in which the individual beforethem is deemed to be an impostor, a dummy, a robot,or whatever extant technology may suggest (Ellis andYoung 1990, 244).

This is not to say that the patient must be awarethat there is a reduced affective response.4 Thedeficit in unconscious affective processing might

just have the result that the patient has a generalfeeling that something is different or thatsomething is different, and the difference is re-lated to this person in front of me. The delu-

sional belief would then arise as a specific hy-pothesis about what has changed. (It might even

be that the deficit in unconscious affective pro-cessing has the result that it seems to the patientthat something is different, and the difference isthat this person in front of me is not my spouse;see section 4.2 below.)

The Cotard Delusion

In the case of the Capgras delusion, it is plau-sible that when the patient sees a familiar face,there is a reduced affective response, and as a

result, the patient has an unusual experience. Wecan now ask whether experiences caused by lossof affective responsiveness can help explain oth-er delusions, such as the Cotard delusion (I amdead). One suggestion made by Young et al.

(1992) is that the Cotard patient has the samekind of experience as the Capgras patient butexplains the experience in a different way (seesection 3.1 below). Another possibility (whichmay be more plausible) is that the neuropsycho-logical anomaly in the Cotard patient is a more

general flattening of affective responses to stimu-

li. This might result in a loss of strong emotionalexperiences and a feeling of emptiness (Ram-achandran and Blakeslee 1998; Gerrans 1999,2000; Young 2000). Alternatively, the patientmight have a general feeling that everything is

different.The Capgras delusion and the Cotard delu-

sion may arise from brain injury, but they alsooccur in patients suffering from psychosis. Thepatient who in the early stages of psychotic break-down described his experience in the following

way, seems to be heading in the direction of anihilistic delusion like the Cotard delusion(McGhie and Chapman 1961, 109):

Things just happen to me now and I have no controlover them . . . .At times I can t even control what Iwant to think about. I am starting to feel pretty numbabout everything because I am becoming an objectand objects dont have feelings.

Anscombe comments on this passage:


10/27


The thoughts and actions that he performs do nothave his stamp upon them, because in some basic,experiential sense, it does [not] feel as ifhe has initiat-ed them . . . . It becomes harder for the patient to bean agent on his own behalf . . . .Objects don t have

feelings; only people who connect with themselves do(Anscombe 1987, 25455).

The Frgoli Delusion

If experiences caused by flattened affectiveresponses can lead to delusions, might some oth-er delusions be explained in terms of experiences

that are caused by anomalously heightenedaf-fective responsiveness? Following a suggestionby Ramachandran and Blakeslee (1998), we in-vite you to suppose that you were to suffer a

form of brain damage with the result that youhad strong affective responses to most faces, not

just the faces of people that you know. Whatexperience would that give rise to? It might seemto you that most of the people that you see arepeople who you know (even people who arereally complete strangers). If this were indeed thenature of your experience, you might naturally

wonder why, if these are people that you know,you cannot recognize them. A possible explana-tion might be that, although these are people thatyou know, they are in disguise. The suggestion is,then, that experiences caused by anomalouslyheightened affective responses to faces might lead

to the Frgoli delusion (I am being followedaround by people who are known to me but whoare unrecognizable because they are in disguise).(In fact, Ramachandran and Blakeslee [1998, 171]suggest that a heightened affective responsivenessto faces might cause a subject to have repeated

experiences of seeing the same familiar face.)

Reduplicative Paramnesia

Nora Breen et al. (2000b) suggest that a simi-lar explanation in terms of heightened affectiveresponses might be offered for DBs reduplicativeparamnesia. Recall that DB affirmed that herhusband had died and was cremated four yearsearlier and that her husband was a patient on the

ward in the same hospital that she was in.It would not be quite right simply to take over

the idea of a heightened affective response even

to unfamiliar faces and to propose that it seemedto DB that some stranger on the ward was famil-iar to her, with the result that the stranger wasmisidentified as her husband, for DB did not

identify anyone who was actually present as herhusband. Instead, she said that her husband was

on the ward in the hospital and she mentioned,for example, that he was talking to other peopleabout having a stroke. What Breen et al. (2000b)suggest is that DB had a heightened sense ofpersonal significance attached to certain remem-bered events. DB remembered an occasion when

a patient on the ward had some visitors; perhapsshe also remembered that the patient talked tohis visitors about having a stroke. She then ex-plained the sense of personal significance at-

tached to this memory by the delusional hypoth-esis that the patient was her husband.5

Alien Control

In the case of the delusion of alien control

(Someone else is able to control my actions),Christopher Frith and colleagues propose thatthe delusion arises when internal monitoring ofself-initiated action is lost (Frith 1987, 1992;Frith and Done 1989; Mlakar et al. 1994). (Foran authoritative review of the role of efference

copies in the internal monitoring of actions, see

Jeannerod 1997, chap. 6.) When visual and ki-nesthetic feedback of movement is intact, but thesense of self-initiation is lost, the resultant expe-rience is of being a puppet with someone elsepulling the strings. (For vivid descriptions of the

alien control experiences of schizophrenic pa-tients, see Spence et al. 1997.) Evidence for thehypothesis that the alien control experience isthe result of a failure of monitoring is providedby the finding that patients with alien controlexperiences have difficulty correcting errors in

their performance on a video game when theymust monitor their movements in the absence ofvisual feedback (Frith and Done 1989).

Thought Insertion

Frith (1992) extends the idea of a failure tomonitor the initiation of action to include thefailure to monitor the initiation of thought orinner speech. Just as failure to monitor the initia-


11/27


tion of action leads to experiences of alien con-trol, so also (Frith 1992, 81): If we found our-selves thinking without any awareness of thesense of effort that reflects central monitoring,

we might well experience these thoughts as alienand, thus, being inserted.

Frith also appeals to a failure to recognize thatinner speech is self-initiated in his account ofauditory hallucinations (Frith 1992, 73), but itmay be that to give an account of both auditoryhallucinations and thought insertion, we need todistinguish two components in our conception of

first-personal ownership of thoughts. Mythoughts are mine in the sense that they occur inmy mind, in inner space rather than in extra-personal space; they are also mine in the sense

that I initiate or generate them. The experienceof thought insertion results from a failure to

recognize self-initiation, but the inserted thoughtis still experienced as occurring in inner space.6

The experience of auditory hallucinations seemsto be different from the experience of thoughtinsertion (even when the inserted thoughts areverbalized). It may be suggested that auditory

hallucinations result from a combination of defi-cits: a failure to monitor the initiation of innerspeech and a loss of the sense of subjective priva-cy that is normally associated with hearing onesthoughts in the minds ear. (For a detailedaccount of auditory hallucinations, see David

1994.) If the sense of subjective privacy were tobe lost while recognition of self-initiation re-mained intact, then a patient might have an ex-perience that could figure in the etiology of thedelusion of thought broadcasting (for furtherdiscussion, see Langdon and Coltheart 2000).

Denial of Ownership of a Limb in UnilateralNeglect

We asked whether it is possible to identifyexperiences that could lead to the eight delusionsin our battery. So far we have suggested that sixof the eight delusions might have a basis in expe-riences resulting from anomalous flattening orheightening of affective responses or from fail-

ures of monitoring. In the case of the seventhdelusion in the batterythe denial of the owner-ship of a limb in patients with unilateral ne-

glectit is plausible that the delusion is to beexplained, at least in part, in terms of the unusu-al experience that results from paralysis and theloss of kinesthetic and proprioceptive feedback

from the arm. Bisiach and Geminiani (1991, 20)speak of the feeling of alienness of the limbs

contralateral to the brain lesion.This leaves us with just the eighth of our

delusions to consider. Does mirrored-self misi-dentification arise from an unusual experience ofoneself as seen in a mirror?

Mirrored-Self Misidentification

It is important here to note that different typesof brain disruption and, correspondingly, differ-ent types of cognitive deficit might give rise to

unusual experiences of mirrored-self. One of thepatients described by Breen et al. (2000b), FE,had a disorder of face processing and, for exam-ple, performed poorly on a face-matching task.In his case it is plausible that the visual appear-

ance of his own face seen in a mirror was unusu-al to the extent that the way that the person inthe mirror looked now was different from theway he remembered himself looking in earlieryears. In contrast to the case of FE, anotherpatient with a delusion of mirrored-self misiden-

tification, TH, showed no significant face-pro-

cessing disorder.7

However, when TH was asked to take in hishand various objects that were held up behindhis shoulder, and so visible only in a mirror, herepeatedly attempted to reach into the mirror,

scratching on its surface, or attempted to reachbehind it. What TH had lost, it seems, was theability to interact fluently with mirrors, eventhough he had not lost his encyclopedic knowl-edge about mirrors, since he was able to provideappropriate answers to such questions as, What

do people use mirrors for?At the beginning of this subsection, we said

that we agreed with Maher that unusual experi-ences are one factor in the etiology of monothe-matic delusions. In what way, then, did TH havean abnormal visual perceptual experience of mir-

rored space or of himself as seen in a mirror?Based on his reaching into the mirror when askedto take reflected objects in his hand, it is certain-


12/27


ly natural to say that he saw those objects asbeing in a space behind the glass, but there ismore than one account that might be given con-cerning the nature of our normal visual percep-

tion of mirrored space, and so it is not immedi-ately clear whether THs visual experience isabnormal.

A normal subject who is given the informa-tion that a sheet of glass is a mirror can make useof a complex set of visuo-motor transformations

and reach fluently into the space in front of theglass to grasp reflected objects. One way of de-scribing this situation is to say that a normalsubjects visual perception of mirrored spacechanges once she learns that a sheet of glass is amirror. Before the change, she sees objects as

being in a space behind the glass, as if the glasswere a window; after the change, she sees reflect-ed objects is being in the space in front of theglass, and this is where she reaches to grasp thoseobjects. On this way of describing the situation,a normal subject who knows that a sheet of glassis a mirror has a visual perceptual experiencethat is quite different from THs experience when

he looks in a mirror. A normal subject sees re-flected objects as being in the space in front ofthe glass, but TH still sees reflected objects asbeing in a space behind the glass, as if the glass

were a window. If this is the correct way todescribe the situation, then THs visual percep-tion of mirrored space is certainly different fromthat of a normal subject.

In contrast, the situation might be describedin a rather different way. It might be said that a

normal subjects visual perception of mirroredspace is the same both before and after she learnsthat the sheet of glass is a mirror. Once she learnsthat it is a mirror, it is her dispositions to reachand grasp that change as she draws on a set oftransformations that recode spatial information

derived from the visual perception into mirrored-space coordinates. On this second way of de-scribing the situation, what normal subjects haveand TH lacked is not a particular kind of visualperception of mirrored space but just the reach-ing and grasping abilities that are grounded inthe accessibility of those visuo-motor transfor-mations.

Capgras delusionUnusual experience of faces or a sense that something is different as a result of flattened affective responses

Cotard delusionLoss of strong emotional experiences and a feeling of emptiness or a sense that everything is different as aresult of global affective flattening

Frgoli delusionUnusual experience of people as a result of heightened affective responses

Reduplicative paramnesiaUnusual experience as a result of heightened affective responses or a heightened sense of personal significanceattached to remembered events

Alien control, thought insertionLoss of experience of self-initiation of action or thought

Unilateral neglectLoss of kinesthetic and proprioceptive experience of the arm and a feeling of the arm as being alien

Mirrored-self misidentificationUnusual experience of ones own face seen in the mirror or experience of reflected objects as if they were on theother side of the glass with loss of the ability to interact fluently with mirrors

Table 2. Experiences that could lead to delusions


13/27


However, we suggest that this second descrip-tion is, as yet, incomplete. Before a normal sub-ject realizes that a sheet of glass is a mirror, shemay see her mirrored-self as someone just like

her (but left-right reversed) in a space behind theglass, but once she learns that it is a mirror, shesees her reflected self as herself. It seems quiteplausible that this difference at the level of con-scious awareness is a result of the accessibility ofthe visuo-motor transformations, even if the sub-

ject is not actually required to do any reachingand grasping. Since the visuo-motor transforma-tions were inaccessible for TH, it seems plausiblethat TH saw his reflected self as someone justlike him, but not as himself. This difference from

normal visual perception of mirrored-self wouldbe a result of THs neuropsychological anomaly.

In this subsection of our paper, we have iden-tified experiences that could lead to the eightdelusions in our battery (Table 2) and have alsoindicated the kinds of neuropsychological anom-

alies that could produce those experiences (Table3). We are agreeing, then, with Maher when hesays (Maher 1999, 551): The origins of anoma-lous experience may lie in a broad band of neu-ropsychological anomalies. However, accord-

ing to Maher, the delusion itself can be seen as anormal response to the unusual experience thatresults from a neuropsychological anomaly. Thiswe wish to question. On our proposal, a secondfactor is required to account for the transitionfrom the experience to a delusional belief. Wepropose that an unusual experience is perhaps

necessary, but it is not sufficient. We now turn tothe argument for that claim.

2. The Argument for a SecondFactor

Suppose it is agreed that Table 2 offers aplausible account of the kind of experience thatis associated with each of the eight monothemat-ic delusions in our battery. On Mahers view,

simply suffering from any one of these experi-ences would be sufficient to produce a delusion,because a delusion is the normal response tosuch unusual experiences. It follows that anyonewho has suffered neuropsychological damage thatreduces the affective response to faces should

exhibit the Capgras delusion; anyone with a righthemisphere lesion that paralyzes the left limbs and

leaves the subject with a sense that the limbs arealien should deny ownership of the limbs; any-one with a loss of the ability to interact fluentlywith mirrors should exhibit mirrored-self misi-

dentification, and so on. However, these predic-tions from Mahers theory are clearly falsified byexamples from the neuropsychological literature.

Reduced Affective Response to Faceswithout the Capgras Delusion

Tranel et al. (1995) describe a number of pa-tients with damage to regions of both frontal

lobes of the brain. Like the Capgras patients ofEllis et al. (1997), these patients did not show thetypical affective response to familiar faces; theydid not discriminate autonomically between fa-miliar and unfamiliar faces. However, these pa-tients, who we assume have the same kind ofexperience of faces as Capgras patients, were not

delusional. (It is possible to question the assump-

Capgras delusion, Cotard delusion, Frgoli delusion, reduplicative paramnesiaReduced or abnormal affective processing

Alien control, thought insertionFailure of internal monitoring

Unilateral neglectLoss of kinesthetic and proprioceptive feedback

Mirrored-self misidentificationDisorder of face processing or inaccessibility of visuo-motor transformations for mirrored space

Table 3. Neuropsychological anomalies that could lead to delusions


14/27


tion that the patients studied by Tranel et al.[1995] have the same kind of anomalous experi-ence as Capgras patients, since they had bilateralventromedial frontal lesions, whereas at least

some Capgras patients have different lesion sites.For further discussion, see Langdon and Colth-

eart [2000, 18990].)

Global Affective Flattening withoutthe Cotard Delusion

People with severe depression experience glo-bal affective flattening, as do those schizophrenicpatients who exhibit the symptom known asflattening of affect. In both cases, the delusionthat one is dead can be absent.

Many patients who are classified as suffering

from the Cotard delusion do not explicitly claimto be dead but have other self-nihilistic delu-sions. We assume that these versions of the Co-tard delusion are also associated with unusualexperiences that result from affective flattening.

However, it is possible to have self-nihilistic ex-periences without becoming delusional. GalenStrawson describes the following case:

A friend who recently experienced depersonalisationfound that the thought I dont exist kept occurringto him. It seemed to him that this exactly expressedhis experience of himself, although he . . . knew, ofcourse, that there had to be a locus of consciousnesswhere the thought I dont exist occurred (Strawson1997, 418).

Loss of Experience of Self-Initiationof Action without the Delusion ofAlien Control

Patients with the alien control delusion saythat an alien is controlling their actions, butpatients with depersonalization disorder may say(among other things) that it is as ifan alien werecontrolling their actions (Davison and Neale

1998). According to the DSM, an individual suf-fering from depersonalization disorder may feellike an automaton and may have a sensationof lacking control of [his or her] actions. How-ever, The individual with Depersonalization Dis-order maintains intact reality testing (e.g., aware-ness that it is only a feeling and that he or she isnot really an automaton) (American Psychiatric

Association 1994, 50002). Therefore, the aliencontrol experience is not sufficient for the aliencontrol delusion.

Loss of Kinesthetic and Propriocep-tive Experience of a Limb withoutDenial of Ownership

The occurrence in some unilateral neglect pa-tients of the delusional belief involving denial ofownership of a limb is not well understood, but

there are many patients who exhibit the patternthat is crucial for the argument of this section.These are patients in whom damage to the senso-ry and motor areas of the right hemisphere hascaused paralysis and loss of kinesthetic and prop-rioceptive feedback from their left limbs, who

(presumably) share the unusual experience of thedelusional patients and who neglect their leftlimbs, but who are not deluded concerning theownership of those limbs. Bisiach and Geminiani(1991, 20) attribute the feeling of alienness ofthe limbs contralateral to the brain lesion to arange of patients not all of whom deny owner-ship of a contralesional limb.

Failure to Recognize Ones Face in theMirror because of a Disorder ofFace-Processing without Mirrored-Self Misidentification

We stressed that different kinds of cognitivedeficit could play a role in mirrored-self misiden-tification. In the case of patient FE, it is plausiblethat he has a disorder of face-processing and thatthe visual experience of his own face as seen in amirror is anomalous (see Breen et al. 2000b, 87,10102). Patients with severe prosopagnosia areunable to recognize familiar faces and even unableto recognize themselves in a mirror, but theprosopagnosic patient does not have the delusion-

al belief that the person in the mirror is not him.

Loss of Ability to Interact Fluentlywith Mirrors without Mirrored-SelfMisidentification

We suggested that in some cases of mirrored-self misidentification, one factor may be a loss ofthe ability to interact fluently with mirrors (seeBreen et al. 2000b, 9192, 101). Binkofski et al.


15/27


(1999) have investigated thirteen patients wholack this ability to varying degrees. Patients clas-sified as mirror agnosic reach towards the vir-tual object in the mirror and are not capable of

changing their behavior; TH would belong inthis category. Patients classified as mirror ataxic

do learn to guide their arms towards the realobject but not in a fluent way; they continue tomake many errors. For the purposes of our argu-ment, the crucial question is whether these pa-tients suffer from mirrored-self misidentification.This question is not answered in the Binkofski et

al. (1999) report, but Binkofski has informed us(pers. comm.): As far as our parietal patientsare concerned, none of them had difficulties inmirror self-recognition.

Now Maher (1999, 550) maintains, The pro-

cesses by which deluded persons reason fromexperience to belief are not significantly differentfrom the processes by which non-deluded per-sons do. However, he also allows (Maher 1999,566), Many normal people have anomalousexperiences but do not develop delusions. From

this it is argued that something additional isnecessary. This is roughly how we have argued,although we have not restricted ourselves to theanomalous experiences of normal subjects.8 Whatis Mahers own response to examples of anoma-lous experience without delusion?

One response to this is that the kinds of anomalousexperience that deluded patients have appeared to bemuch more intense and prolonged than those thatoccur in the population in general. Their intensity andduration is determined by the prolonged duration orfrequent repetition of the pathological state creatingthe experience (Maher 1999, 566).

This response can be adapted to our examples ofnon-delusional patients who have unusual expe-

riences as a result of neuropsychological anoma-lies, but in our view, the response thus adapted is

not really adequate.It is certainly not generally true that delusions

arise only after a prolonged period of anomalousexperience. A delusion may be present very shortly

after the onset ofthe pathological state creatingthe experience. Thus consider again patient LA-O, described by Bisiach and Geminiani (1991,3233):

Shortly before admission [to hospital] she had sud-denly developed left hemiplegia [i.e., paralysis of theleft side of the body] without loss of consciousness.Alert and cooperative, she claimed . . . that the lefthand did not belong to her but had been forgotten in

the ambulance by another patient . . . .Two days laterthe symptoms had partially faded . . . .She volunteeredthe information that only on that very morning shehad begun to recognize her left hand as her own whilebeing perfectly reminiscent of her past denial.

In this example, and surely in others, the re-sponse adapted from Maher will have to place

weight on the notion of the intensity rather thanthe duration of the experience. However, althoughthe duration of an experience can, in principle,be readily measured, it is not so clear how theintensity of an experience is to be quantified.

(For this purpose, measurement of the intensityof an anomalous experience must not, of course,appeal to whether the subject is delusional.)

In any case, an experiment by Cahill et al.(1996) casts doubt on any proposal that theintensity of an experience is the crucial factor inexplaining why some subjects who have that

kind of experience become delusional while oth-ers do not. Subjects in this experiment heardtheir own voices through headphones, but theexperience was anomalous because the pitch ofthe voice was distorted to varying degrees. A

pilot study showed that normal volunteers wereable correctly to attribute the heard voice tothemselves, despite the pitch distortion. Howev-er, schizophrenic patients often attributed theheard voice to another agent, with one subjectgoing so far as to say, I think its an evil spiritspeaking when I speak (Cahill et al. 1996, 209).The frequency of attributing the heard voice toanother agent was correlated with the currentseverity of the patients delusions and also withthe degree of pitch distortion. Thus, delusionalpatients tended to make a false attribution of theheard voice to another agent, while normal sub-

jects presented with the same distorted stimulihad no difficulty in correctly attributing the heardvoice to themselves. Cahill and her colleaguesnote that these results do not support Mahersproposal that erroneous beliefs reflected theoperation of normal reasoning processes in the

context of the experience of abnormal percep-tions (Cahill et al. 1996, 209). Instead,


16/27


We conclude that the hallucination-like reports elicitedby our paradigm [the attribution of the heard voice toanother agent] resulted from an interaction between anunusual perceptual experience (distorted auditory feed-back) and an abnormal mechanism for belief forma-

tion present in deluded patients (Cahill et al. 1996, 201).

It is not easy to resolve all the issues here, butwe know of no reason to suppose that in all the

examples described above, the absence of a delu-sional belief can be explained in terms of thelower intensity or shorter duration of the unusu-al experience by comparison with delusional pa-tients. Consequently we conclude that, while anunusual experience produced by a neuropsycho-

logical anomaly such as a deficit in perceptual oraffective processing may be one factor in the

etiology of a delusion, there must also be at leastone other factor present if a delusion is to occur.Our proposal is that for a monothematic delu-

sion to occur, two factors must be present. Onefactor is a neuropsychological anomaly with somemanifestation in the experience of the subject,but this is not a sufficient condition for the oc-currence of delusions. The nature of the firstfactor varies from delusion to delusion and frompatient to patient, and as Tables 2 and 3 indicate,

there are different kinds of neuropsychologicalanomalies and different kinds of unusual experi-ences associated with different kinds of delusion.

In contrast, on the boldest version of our propos-al, the nature of the second factor is the same forall deluded patients. It is the second factor that

explains the difference between a mirrored-selfmisidentification patient such as TH and a pa-tient with mirror agnosia or between a patientwith the alien control delusion and a patientwith alien control experiences as part of deper-sonalization disorder.

The argument for a second factor is, of course,strengthened if we can find other examples ofnon-delusional subjects who have neuropsycho-

logical anomalies and consequent unusual expe-riences. Indeed, the ideal situation would be tohave non-delusional cases corresponding to each

of the eight delusions in our battery. There ismore work to be done here. We do not know ofany such non-delusional cases that correspond tothe Frgoli delusion, to reduplicative paramne-sia, or to thought insertion, but in our view,

based on the examples of non-delusional pa-tients that we have mentioned in this section, theargument for the two-factor proposal is alreadystrong.

3. Biases in the Formation ofBeliefs

In the last section, we argued that cases of aneuropsychological anomaly and consequent un-

usual experience without delusional belief showthat Mahers account of delusions is not ade-quatea second factor is required. The problemfor Mahers claim that delusional beliefs are nor-mal responses to unusual experiences can also beseen if we consider two possible routes from

experience to belief. A subject might arrive at abelief by constructing and adopting an explana-tion of the occurrence of an experience, just as ascientist explains the occurrence of empirical data,or a subject might arrive at a belief by simplytaking perception to be veridical.9 Consider nowthe two possible routes from unusualexperienceto delusionalbelief, routes that Maher must re-gard as normal.

Suppose, first, that delusional patients areaware that there is something unusual or anoma-lous about their experience and that they con-

struct an explanation for the occurrence of this

unusual feature. This seems to be the route thatMaher has in mind when he says:

[T]he explanations (i.e., the delusions) of the patientare derived by cognitive activity that is essentiallyindistinguishable from that employed by non-patients,by scientists, and by people generally . . . .[A] delusionis a hypothesis designed to explain unusual perceptualphenomena (Maher 1974, 103).

The suggestion that delusions arise from the nor-mal construction and adoption of an explana-tion for an unusual feature of experience faces

the problem that delusional patients construct

explanations that are not plausible and adoptthem even when better explanations are avail-able. This is a striking departure from the morenormal thinking of non-delusional subjects whohave similar unusual experiences. Consider, for

example, the following two possible dialoguesbetween a clinician and a right-hemisphere-dam-aged patient:


17/27


C: Why cant you move this arm?P: Because it isnt my arm; its yours.

versus

C: Why cant you move this arm?

P: Because I had a stroke in my right hemisphere;would you like to see my CT scan?

(Anscombe [1987, 250] makes the same pointabout Mahers account as it applies to schizo-

phrenic delusions: [S]chizophrenic patients comeup with explanations of their perceptual aberra-tions, particularly heightened significance, thatare not sane. The sane explanation is that theyhave schizophrenia, or something like it.)

Suppose, alternatively, that a delusional belief

arises because a patient takes an unusual percep-tion to be veridical. If delusional patients simplybelieve what they perceive, implicitly assumingthat perception is veridical, then they do whatwe all normally do as we use perception to findout about the world. Nevertheless, if the sugges-

tion is that the route from unusual experience todelusional belief goes via the unreflective accep-tance of veridicality, then it faces a problem. Fordelusional patients seem to proceed from experi-ence to belief even when there are overwhelmingreasons not to trust experience. This is once

again in contrast with the more normal thinkingof non-delusional subjects who have unusual ex-

periences as a result of brain injury.In our view, these problems reinforce the doubt

about Mahers account. We agree with Youngwhen he says:

We think that the Capgras delusion represents justone among a number of ways in which people mighttry to explain similar anomalous perceptual experi-ences to themselves, and that to properly understandthis delusion we need to understand not just the per-ceptual anomaly but also the factors which create andsustain the relatively bizarre impostor explanation (Young2000, 63; see also Young 1998, 4041; 1999, 572).

Young mentions attributional biases and rea-soning biases, but recent research suggests that

we should consider data-gathering biases insteadof reasoning biases. For example, Garety andFreeman (1999, 131) say: People with delu-sions do not, it seems, have a probabilistic rea-soning bias, . . . but have a data-gathering bias.

In our view, neither attributional biases nor data-gathering biases can play the role of the secondfactor in the etiology of monothematic delusions.In the remainder of this section, we consider

them in turn.

3.1 Attributional Biases

Consider again the first possible route from

experience to belief, i.e., delusional patients con-struct an explanation for the occurrence of anunusual experience. This account faces the prob-lem that delusional patients construct explana-tions that are not plausible and adopt them evenwhen better explanations are available. Why do

delusional patients prefer the less good explana-tion?

Richard Bentall and others propose that per-secutory delusions result from biases in the kindsof explanations that subjects give of their ownbehavior and the behavior of other people (Ben-

tall 1994, 1995; Kinderman 1994; Bentall andKinderman 1998). In particular, patients withpersecutory delusions tend to blame other peoplewhen something goes wrong. The appeal to dif-ferences in attributional style may go some waytowards explaining the difference between delu-

sional and non-delusional subjects who have thesame kind of unusual experience. It also opens

the possibility of explaining different delusionsas differently biased attempts to explain the sameunusual feature of experience.

Thus, for example, concerning the Capgras

and Cotard delusions, Young (2000, 65) suggests:Although these delusions are phenomenally dis-tinct, they may represent the patients attemptsto make sense of fundamentally similar experi-ences (see also Young et al. 1992). The basicidea here is that, starting from the same unusual

feature of experience, the Capgras delusion aris-es from an externalizing attributional style thatis associated with feelings of persecution while theCotard delusion arises from an internalizing at-tributional style that is associated with depression.

This is a fascinating, and in many ways plausi-ble, suggestion that gathers support from a casein which a single subject experienced the Cotardand Capgras delusions in sequence (Wright et al.1993). However, the account of the Capgras and


18/27


Cotard delusions that appeals to differences in

attributional style cannot be fully general sincethere are cases ofconcurrent Capgras and Co-tard delusions (Joseph 1986; Wolfe and McKen-zie 1994; Butler 2000).

We accept that attributional biases may, insome cases, explain why a particular delusionalhypothesis is prioritized as a candidate explana-

tion of an unusual experience, but the appeal toattributional biases does not adequately addressthe argument for a second factor in the etiologyof delusions, because it does not explain why thehypothesis is actually adoptedand maintainedas

a belief.

3.2 Data-Gathering Biases

Consider now the second possible route fromexperience to belief, i.e., delusional patients sim-ply believe what they perceive. This account fac-

es the problem that delusional patients seem to

proceed from experience to belief even whenthere are overwhelming reasons not to trust ex-perience. Why do delusional patients trust expe-rience when there are good reasons not to do so?

Philippa Garety and others propose that one

factor in the etiology of delusions is a tendencyto jump to conclusions based on insufficient evi-dence (Huq et al. 1988; Garety et al. 1991; Gare-ty and Hemsley 1994; Garety and Freeman 1999;see also Dudley et al. 1997a,b). Results in proba-bilistic reasoning experiments with delusionalpatients do not reveal a total inability to reason

probabilistically, but subjects do show a biastowards early acceptance of hypotheses. Doesthis bias in data gathering offer an explanationof how delusions arise based on unusual experi-ences and why they are maintained?

We accept that a tendency to jump to conclu-sions may help to explain why the delusional

hypothesis is initially adopted as a belief, but itdoes not explain why the delusional belief ismaintained so tenaciously. Indeed, in Garetysexperiments, delusional subjects jump to conclu-

sions more rapidly than normal subjects but alsotend to abandon existing hypotheses based on

very little evidence. The appeal to a data-gather-ing bias, like the appeal to attributional biases,falls short as a response to the argument for asecond factor in the etiology of delusions.

4. The Second Factor

The strength of the two-factor account ofmonothematic delusions is the evidence that a

first factor (often a neuropsychological anomalyin perceptual or affective processing) is demon-

strably present in many patients with delusions,yet is not itself sufficient to cause the delusion.The weakness of the account as it currently standsis the inadequate characterization of the nature

of the second factor. The second factor might bedescribed as a loss of the ability to reject a candi-date for belief on the grounds of its implausibili-ty and its inconsistency with everything else thatthe patient knows. However, attempts to say inmore detail what this loss of ability amounts to

face many problems. In this final section, wedescribe some of these.

4.1 Two Problems: Monothematicityand Appreciation of Implausibility

In this paper we are concerned with monothe-matic delusions,10 but the very fact ofmonothe-maticity already presents a challenge. The pa-tients that we have described typically have just

one delusional belief or a small set of relateddelusional beliefs. Outside the specific domain ofthe delusion, their belief systems are not abnor-mal. If these patients suffer from some kind ofdeficit or abnormality in their belief evaluation

system, what is it that protects them from adopt-ing bizarrely false beliefs in other domains?

At least some delusional patients show con-siderable appreciation of the implausibility oftheir delusional beliefs. This is evident in thefollowing extract from an interview with a manwho thought that his house and family had been

replaced by duplicates (Alexander et al. 1979, 335):

Capgras delusion = loss of affective responsiveness +a personal bias to externalize blame

Cotard delusion = loss of affective responsiveness + apersonal bias to internalize blame

Table 4. Summary of the attributional-bias account


19/27


E: Isnt that [two families] unusual?S: It was unbelievable!E: How do you account for it?S: I dont know. I try to understand it myself, and itwas virtually impossible.

E: What if I told you I dont believe it?S: Thats perfectly understandable. In fact, when I tellthe story, I feel that Im concocting a story . . . . It s notquite right. Something is wrong.E: If someone told you the story, what would youthink?S: I would find it extremely hard to believe. I shouldbe defending myself.

Delusional patients with this kind of apprecia-tion of the implausibility of their delusional be-

lief are not surprised that the people aroundthem refuse to accept the belief as true. (Young

[1998, 37] says, Capgras delusion patients canbe otherwise rational and lucid, able to appreci-ate that they are making an extraordinary claim.If you ask what would you think if I told you mywife had been replaced by an impostor?, youwill often get answers to the effect that it wouldbe unbelievable, absurd, an indication that youhad gone mad.) If these patients suffer from aloss of the ability to evaluate beliefs for plausibil-ity and consistency with other things that they

know, how are they able to judge that otherpeople will find their belief implausible?

We can see how these two problems arise for

putative accounts of the second factor in theetiology of delusions if we briefly consider theidea that delusional patients are unable to make

appropriate and effective use of stored knowl-edge. There is evidence, for example, that schizo-phrenic patients may be prone to over-inclu-sive thinking, classifying an airplane in thecategory birdbecause it has wings and flies (Chenet al. 1994). A categorization problem of this

kind is consistent with the idea that schizophren-ic patients ignore stored information about ex-ceptions. In line with that idea, recent research

by Sellen et al. (2000), using a conditional infer-ence task, suggests that high-schizotypy normalsubjects make less use of information about ex-

ceptions than low schizotypy subjects.11 It mightbe proposed, then, that monothematic delusionsare the result of failure to use information inencyclopedic or semantic memory to assess

the plausibility of a candidate for belief. Anysuch proposal faces both the monothematicity

problem and the appreciation of implausibilityproblem. If delusional patients cannot use infor-

mation in semantic memory to reject the delu-sional belief, then why do they not end up with a

host of other bizarrely implausible beliefs? Ifdelusional patients cannot use information insemantic memory to assess the plausibility oftheir delusional belief, then how do they appreci-ate so clearly that other people will indeed find itimplausible?

4.2 The Two Routes from Experienceto Delusional Belief

In Section 3, we distinguished two possible

routes from unusual experience to delusional be-lief. On the first route, delusional patients areaware that there is something unusual or anoma-lous about their experience, and they constructan explanation for the occurrence of this unusual

feature. On the second route, delusional patientssimply believe what they perceive; they implicitlyassume that perception is veridical. Consider nowhow the Capgras delusion might arise by one orother of these routes.

Suppose first that the patient has a visual

perception of someone who looks just like aclose relative and that this is accompanied byawareness that there is a reduced affective re-sponse or by a sense that it doesnt feel likeher, or just by a general feeling that somethingis different. Suppose that the patient sets out toconstruct an explanation for the occurrence ofthis unusual feature of experience. This mightlead to the consideration, and ultimately the adop-tion, of the putatively explanatory hypothesisthat the person who looks like the close relativeis not really a close relative but an impostor.

Intuitively, the patient goes wrong in adopting,and all the more so in maintaining, an explana-

tory hypothesis that should be rejected.Suppose, on the other hand, that the patients

unusual experience represents the situation asfollows: This is someone who looks just like my

close relative but is not really her/him. If thedelusional hypothesis is already part of the rep-resentational contentof the patients perception,


20/27


then the route to a delusional belief involvesnothing more than accepting the perception asveridical.12 In this case, it seems that the patientgoes wrong in not making a critical assessment

of, and particularly in not rejecting, the veridi-cality of this unusual perceptual experience.

The remainder of this section is organizedaround the distinction between the route to be-lief that involves the construction, adoption, andmaintenance of an explanatory hypothesis andthe route to belief that simply involves taking aperception as veridical. We consider in turn pos-

sible accounts of the second factor that assumeeither the first or the second route.

4.3 Delusional Beliefs and ExplanatoryHypotheses

Suppose, once again, that a Capgras patienthas a visual perception of someone who looksjust like a close relative and that this is accompa-nied by awareness that there is a reduced affec-

tive response, or by a sense that it doesnt feellike her, or just by a general feeling that some-thing is different. We need to account for theconstruction, adoption, and maintenance of theexplanatory hypothesis that the person who lookslike the close relative is not really a close relative

but an impostor.Even if we leave aside the question why it

occurs to the patient to construct this bizarrehypothesis (perhaps the idea of an attributionalbias can help here), we need some cognitive ac-count of why the hypothesis is adopted and main-tained, rather than being rejected on the groundsof its implausibility and its inconsistency witheverything else that the patient knows. In partic-ular, we need to answer three questions: First,why can the delusional hypothesis not be refutedby using first-person experience? Second, why

can it not be refuted by using stored encyclope-dic knowledge? Third, why can it not be refuted

by relying on the testimony of other people?In answer to the first question, it may be

helpful to recall, from Cattss draft Definition 2(Figure 1), the suggestion that a delusional belief

has the quality of being self-evidential and un-falsifiable (irrefutable) either because of the na-ture of the belief or because of the refusal of the

holder of the belief to admit anything that wouldfalsify the belief. There are two alternative pro-posals here. One is that the belief is unfalsifiablebecause of its content; the other is that the sub-

ject elevates the belief to such a status that noth-ing is allowed to count against it. The second of

these is quite close to what we are trying toexplain, namely, the adoption and maintenanceof an implausible hypothesis, but the first pro-posal might contribute to an explanation if weinterpret it as saying that a delusional belief has acontent that makes it especially difficult to refute

by using first-person experience. Indeed, this doesseem somewhat plausible in the case of delusionsof misidentification such as the Capgras delu-sion. The delusional hypothesis cannot be readi-

ly refuted by evidence concerning, for example,the observable characteristics of the supposed

impostor or the way that he or she answersquestions.

Even if it is correct that the content of adelusional hypothesis makes it difficult to refuteby using first-person experience, we still need toanswer the second and third questions. We have

already seen (Section 4.1) that an answer to thesecond question is liable to be problematic if itappeals to a general problem in the use of storedknowledge. However, we might speculate thatthere is a more specific problem that arises be-cause of the particular contents of delusions. For

example, it might be that using stored informa-tion to refute the delusional hypothesis wouldrequire the information to be available in explicitor declarative form, while the information thatwould be needed is usually stored in an implicitor procedural form. As for an answer to the third

question, we might speculate that in delusionalpatients, reliance on the testimony of others isimpaired either by a theory of mind deficit (seeLangdon et al. [2002] for a review of theory ofmind deficits in patients with schizophrenia) or

else by a social psychological problem of shrink-ing the in group of people whose opinions canbe trusted.13

These answers to the three questions add upto a complex, underspecified, and highly specu-lative story about the etiology of the Capgrasdelusion or any other delusional belief. However,


21/27


it might be thought that the answers to the firstand second questions do at least offer some re-sources for addressing the monothematicity prob-lem. The reason is that there might be only a

rather narrow class of hypotheses meeting thetwo conditions of being self-evidential or un-

falsifiable in the required sense, while also, forexample, being immune to refutation by storedinformation unless the information is convertedfrom a procedural to a declarative format. Wesuggest, though, that an account based on thesekinds of answers to the three questions about why

the delusional hypothesis is not rejected still facesa version of the monothematicity problem. Anaccount of this kind seems to yield the problematicprediction that a delusional subject will differ from

normal subjects in being liable to adopt and main-tain false beliefs based on misleading perceptual

experiences presented by certain visual illusions.Let us explain what we see as the problem here.

Suppose that a normal subject knows that hertwin sons, Bill and Ben, are the same height andnow sees them through the viewing window ofan Ames room in which Bill is standing in the tall

corner to the left while Ben is standing in theshort corner to the right (see Gregory 1970, 2628). From the viewing position, Bill looks short-er than Ben, but we take it that, even withoutentering the Ames room to measure her sons, thenormal subject will reject the hypothesis that

now, in the Ames room, Ben is taller than Bill.She will prefer the hypothesis that there is some-thing strangeas yet, she knows not whatabout the way things look in this room.

Now consider a delusional patient who alsoknows Bill and Ben well and sees them in the

Ames room. Bill, on the left, looks shorter thanBen, on the right. Suppose that this patient con-structs the hypothesis that people and thingschange their height when they go into this room:Those who go to the left of the room get shorter;

those who go to the right of the room becometaller. It is not a straightforward matter for thepatient to refute this hypothesis by using first-person experience, whether he remains at the view-ing position or enters the room himself to mea-sure Bill and Ben. For according to the hypothesis,both he and his measuring rod become shorter as

he moves to the left to measure Bill and thengrow taller as he moves to the right to measure Ben.

The normal subject makes use of the storedknowledge that height does not change with mere

change of position, but it is fairly plausible thatthis is just the kind of knowledge about the

physical world that might be stored in procedur-al form. The delusional patient may not be ableto make use of this stored knowledge to refutehis hypothesis. Given our answer to the thirdquestion about why it is difficult for him torefute the delusional hypothesis, he may not be

able to make use of the testimony of other peopleeither. In short, we have the prediction that thedelusional patient who sees Bill and Ben in theAmes room is liable to adopt and maintain a

second bizarrely false hypothesis.It is true that we do not have any data to

report about the ways in which delusional pa-tients respond to visual illusions. It might con-ceivably turn out that even patients whose delu-sions are monothematic would indeed arrive at asecond, and unrelated, bizarrely false belief ifthey were placed in the situation that we have

described. Nevertheless, the prediction is onethat we would prefer to avoid. If we assume thefirst route from experience to belief and try togive an account of why the delusional hypothesisis not rejected, then we face the monothematicityproblem in the form of an unwanted prediction.

We have taken some time to explain how themonothematicity problem arises from our specu-lative answers to three questions about why adelusional explanatory hypothesis might be dif-ficult to refute. However, we do not need to takevery much time over the problem about appreci-

ation of implausibility. We are supposing that, invirtue of its content, a delusional hypothesis isdifficult to refute by using first-person experi-ence or by relying on stored encyclopedic knowl-edge, and that a delusional patients reliance on

the testimony of others is also impaired. Againstthe background of these assumptions, it surely ispuzzling that patients should have such a keenappreciation of the implausibility of their delu-sional beliefs.


22/27


4.4 Delusional Beliefs and theAssumption that Perception is Veridical

The prospects for a solution to the apprecia-tion of implausibility problem seem to be better

if we assume the second route from experience tobelief. It is part of the representational content ofa Capgras patients visual perception that Thisis someone who looks just like my close relative

but is not really her/him, and the patient simplyaccepts this perceptual experience as veridical.

Normal subjects usually assume perception tobe veridical; they believe what they perceive. Wemight describe this transition from experience tobelief as apre-potent doxastic response. Howev-

er, normal subjects are also able to suspend theirunreflective acceptance of veridicality and make

a more detached and critical assessment of thecredentials of their perceptual experiences. Wemight conceptualize what happens in the Capgraspatient or other delusional patients as failure to

inhibit a pre-potent doxastic response. (The sug-gestion here is not that delusional patients have ageneral inability to inhi

davies two-factor account

Documents