Face-Processing

Impairments

and the Capgras Delusion

ANDREWW. YOUNG, IAN REID,SIMON WRIGHT and DEBORAHJ. HELLAWELL
Investigations of two cases of the Capgras delusion

might prove as useful in understanding delusional
misidentification as it has proved in neuropsychology.

found that both patients showed face-processing impair

ments encompassing identification of familiar faces,

To this end, we investigated the face-processing

recognItion of emotional facial expressions, and matchk@g abilities of two patients with the Capgras delusion,

of unfamiliar faces. In neither case was there any
Impairment
findings

of recognition

are consistent

memory

with

the

for words.

idea

that

the

using procedures developed in cognitive neuro

These
basis

psychological

of

investigations.

the Capgras delusion has k@damage to neuro-anatomical
_ways

respansI@Ie for appropriate

emotional reactions

Case reports

to famliar visual stimuli. The delusion would then
represent the patient's attempt to make sense of the

The investigation

fact that

experienced the Capgras delusion, MC and ML. Both were

these

visual

stimuli

no longer

have

appropriate

involved two patients

who had

affective significance.
tested before discharge from hospital, when they were no
British Journal of Psychiatry (1993), 162, 695—698 longer in the florid state of being convinced of the veracity
of the delusion.

The Capgras delusion involves the belief that certain
people, usually close relatives, have been replaced
by ‘¿dummies'
or impostors. Although there has been
a great deal of interest in psychodynamic accounts
of the delusion, this has been tempered by the
observation that in many cases there is brain injury

Case 1

A 66-year-oldmarried woman, MC, was admitted to a
psychiatric ward following a two-week history of increasing
agitation and confusion. Her husband was particularly
distressed by the fact that she believed he had been

(Förstlet a!, 1991). This is supported by findings that
patients with the Capgras delusion perform poorly on
standard neuropsychological tests of face processing
(Ellis & Young, 1990). These tests often require
subjects to match photographs of faces of unfamiliar
people by picking out those which are pictures of the
same person, such as the Benton Test of Facial
Recognition (Benton eta!, 1983). However, it is now
clear that neuropsychological impairments can affect

‘¿replaced'by an ill-intentioned
aggressive.

matching,

and

recognition

important

At interview MC was confused, disorientated and

to test a range of

face-processing abilities in order to determine which
abilities are impaired and which intact, rather than

carried out. No new specific abnormality

relying on a single test.
Some

authors

have

suggested

a link

between

matching

of unfamiliar

faces

(Young,

the

1992).

To date, though, there have been no reports of
systematic testing of familiar-face recognition in
Capgras cases.
Ellis & Young (1990) have suggested that detailed
testing of different types of face-processing ability

to account

for

her acute confusion was found and, in particular, her
diabetic state was well controlled with no evidence of

Capgras delusion and the neurological condition
prosopagnosia. However, the central feature of
prosopagnosia is that familiar faces are not recognised,
and some prosopagnosic patients can show fairly
accurate

She suffered from diabetes mellitus, ischaemic

agitated. She appeared neither depressed nor elated, and
apart from her firmly held delusional belief about her
husband she had no other psychotic symptoms. An initial
diagnosis of acute confusional state supervening on chronic
mild dementia was made, and a series of investigations

of

emotional expressions have been described (Young,
1992). It is, therefore,

who was trying

heart disease, and had recently been diagnosed as suffering
from a mild dementing illness, probably vascular in nature.
In the past she was diagnosed as suffering from a manic
depressive illness, but had had no symptoms or treatment
on this account in recent years. Despite her physical
problems and mild dementia, she was living independently
at home before the onset of the symptoms leading to her
admission.

face processing in a number of dissociable ways: dis
sociable impairments of familiar-face recognition,
unfamiliar-face

impostor

to ‘¿cheat'
her. In addition, she was disorientated and

infection. Her confusional

state fluctuated from day to day,

and then gradually resolved without specific treatment, as
did her conviction that her husband had been duplicated.
On recovery, she clearly recalled this belief, but dismissed
her experience as ‘¿imaginary'.
As far as could be
determined, MC had never held such beliefs before and was
eventually

discharged

home to the care of her husband.

Before discharge, but after recovery, a single-photon
emission computerised tomography (SPECT) scan was
carried out. This showed asymmetrical

tracer uptake with

patchy perfusion deficits consistent with a multi-infarct

695

696

YOUNG ET AL

process, with a marked perfusion deficit in the left
frontal

area.

Case 2
A 60-year-oldwidow,ML, wholivesalone, wasbroughtto
the accident and emergency department by her son, with

a one-weekhistoryof confusionand persecutorydelusions.
She believed people wished to enter her house and throw
stones through her window. She said that they were going
to kill her son and that she herself was going to die, being
convinced

that she had suffered

a heart attack

that day.

She claimed to have psychic powers and to be able to predict
the future. She also complained of a voice telling her to
leave her house.
She was a well nourished, fit-looking woman who was
anxious and agitated, fatuous in her affect, and laughed

inappropriately.Physicalexaminationrevealednothing of
note except an equivocal right plantar response. Routine
laboratory investigations revealed an elevated erythrocyte
sedimentation

rate (ESR)

at 35 mm/h,

which

returned

to

normal.
During the next 48 hours ML's mental state fluctuated
between

periods

of complete

rationality

and episodes

when she was convinced that the nurses were trying
to poison her. Five days after admission she left the
ward without permission and returned home. Over the
next three weeks she barricaded herself in her
house, saying that her female general practitioner was
having sex with her, and that her son was an impostor.
She repeatedly threw buckets of water over people who
tried to visit her and required readmission under the
Mental

Investigation of face-processingabilities
Tests were used to assess MC and ML's abilities to determine

facial expression, to identify familiar faces, and to match
unfamiliar faces. A control group comprised men aged over
60 years tested at the Radcliffe Infirmary, Oxford. Although

the controlsweremenand both patientswerewomen,there
is no reason to expect substantial sex differences in the test
results. In addition,

a standard

test involving

recognition

memoryfor facesand words(Warrington, 1984)wasgiven
to both MC and ML, whoseperformanceswerecompared
to the published norms (which include both men and
women).

The results are summarised

in Table 1.

The ability to recognise emotional facial expressions was
tested with 24 photographs of facial expressions, presented

oneat a timewitha listof sixpossibleemotionlabelsprinted
below each face (Young et a!, 1990).The task was to choose
the correct label to describe the facial expression. Both
Capgras patients performed poorly (MC, 16/24 correct;
ML, 10/24 correct) in comparison to 34 men from the
Radcliffe Infirmary control group (mean 21.10/24 correct,
s.d. 1.70), showing defective ability to interpret facial
expressions of emotion.
To test ability to recognise

familiar

people,

20 very

familiar faces, 20 moderately familiar, and 20 unfamiliar
faces, were presented one at a time in random order.
For each face the patient was asked whether or not it
was a familiar person and, if so, his/her occupation and
name. Control data were taken from 28 men from the
Radcliffe Infirmary control group. The task has previously
been used in the investigation of delusional misidentification
by Young

et a! (1990).

The data of interest

concern

the

Health Act.

On this admission she claimed that her son had been
replaced by two impostors, which over the next week
Table 1
became four. This delusion also included her daughter-in
Capgras patients' (MCand ML)performances on face
law and her granddaughter. She said of her son, “¿There's The
processing tasks, and means and standard deviations for
been someone like my son's double which isn't my son...
control subjects of comparable age
I can tell my son because my son is different. . . but you
have got to be quick to notice it.―She renamed one of her
MCMLContr
son's alleged impostors ‘¿John
Yates' and said this one
Meanols
(s.d.)Facial
wished to rape her.
Physical examination revealed slightly decreased power
expressions
in the right leg, and brisk knee reflexes. An electro
Labelling
16/24―10/24―21.10/24(1.70)Identification
cardiogram (ECG) showed the S1Q3T3configuration of
pulmonary embolism; blood gas estimation showed a
of familiar faces
lowered p02 (9.94 kPa) and a slightly lowered pCO2
Very familiar faces
occupation
8/20―
(4.2 kPa). An electroencephalogram (BEG) revealed a mild
name
5/20―
7/20― 16.25/20(1.15)
(2.81)Unfamiliar-face
background excess of slow activity with scattered sharp
Unfamiliar faces
elements and no focalisation. Computerised tomography
correct rejections
20/2013/20―20/2018.86/20
of the head showed dilatation of extracerebral spaces filled
with cerebrospinal fluid, although not advanced for her age.
matching
From these investigations, it was presumed that ML was Bentontest
39/54'
suffering from an episode of delirium precipitated by (2.16)Recognition
Disguisetest
18/24'35/54―44.22/54
22.30/24(3.07)
hypoxia

secondary

to pulmonary

embolism.

ML was treated with chlorpromazine (75 mg three times
daily) and her delusional beliefs settled (e.g. “¿there
are no
longer two Johns, but there used to be―).The neuro
psychological tests reported here were carried out during
this period, before her discharge.

memory

WarnngtonRMT:faces 41/50
WamngtonRMT:words41/5027/50―
49/50
Asterisked
scores
‘¿z>1.65.P<0.05;

*

are significantly below the control mean:
“¿z>2.33, P<0.O1; “¿z>3.10. P<O.OO1.

697

CAPGRAS DELUSION AND FACE PROCESSING
ability to give correct occupations and names to the 20 very
familiar faces, and the rate at which the 20 unfamiliar

faces

were misidentified. Data for the moderately familiar faces
are not included

in Table 1, because the control subjects

showed high variance in their responses to these.
Results showed severely impaired ability to identify
familiar faces for both MC and ML. They scored more
poorly than controls for both occupations and names of
highly familiar people. However, neither patient showed

any tendencyto misidentifyany of the unfamiliar people
(20/20 correct rejections), which indicates that they did not
confabulate responses.

The Benton Test of Facial Recognition(Benton et a!,
1983)was also given to both patients. In this test, subjects
have to choose which of six photographs ofunfamiliar faces
are pictures

of the same person

as a simultaneously

presented target face photograph. The test includesitems
involving

choice

of identical

photographs,

and trans

used in most previous studies. Both MC and ML
showed severe impairments which encompassed
recognition of familiar faces and recognition of facial
expressions as well as unfamiliar-face matching. In
neither case, however, was there any disturbance of
recognition

memory

for words, which shows that

MC and ML's problems in face-processing tasks
cannot be attributed to lack of attention to the tasks,
or any more general failure to understand what they
were being asked to do. Similarly, the fact that both
patients were able to reject 20 out of 20 of the un
familiar faces included in the familiar-face recognition
task shows that they were not confabulating or
confused about what they were doing.
Neuropsychological

impairments of face processing

are typically found after brain injury affecting the

formations of orientation or lighting, which are pooled into

right cerebral hemisphere (Young, 1992). Hence our

an overall composite score.
Both MC and ML produced overall scoreswhich were

casesareconsistent

impaired

both

in terms

of the test's

norms

and

in

comparison to 32 men from the Radcliffe Infirmary control

group (see Table 1).
A secondunfamiliar-facematchingtest examinedability
to match disguisedfaces. Becauseof time constraints, this
wasonlygivento MC. Two separatetest sheetswereused,

findings of face-processing impairments in Capgras
with reports which have emphasised

the role of right hemisphere disease in the production
of the Capgras delusion from both neurological and
neuropsychological

lines of evidence (Cutting, 1991).

We do not claim that the two patients suffered
exclusively right hemisphere

damage.

As we have

each showing a 3 x 4 matrix of faces in which each of the
four faces in the top row appeared three times in disguised
or undisguised forms elsewhere on the sheet (see Young et
a!, 1990, for an illustration). MC was impaired at matching

noted, both cases involved an organic psycho
syndrome, and other cases of the Capgras delusion
in the context of such conditions have been reported

thesefacesin comparisonto the 10controlsof comparable
age (range 63—75
years) from Young et al's (1990)study.
The WarringtonRecognitionMemoryTest(Warrington,
1984)was givento both patients. In this test, recognition
memoryistestedseparatelyfor facesand words.In the faces

showed evidence of reduced blood flow in the left

part of the test, 50 faces are shown at the rate of one every

three seconds for a ‘¿pleasant
or unpleasant' decision, and
then recognition memory is tested by presenting each of

the facespaired with a distractor, with the subject having

(Förstl et a!, 1991). Moreover,

MC's SPECT scan

frontal area, as well as other less marked perfusion

deficits.
Face-processing impairments in Capgras patients
have been taken to imply a connection between the
Capgras

delusion

and the neurological

condition

prosopagnosia, in which people with occipito
temporal brain lesions lose the ability to recognise

to choose which has been seen before; a similar procedure

familiar faces. This claim is made more plausible by

is usedwithwords.ML's scoreof 27/50correctfor the faces

the observation that there are some similarities
between the brain lesions which can produce the
Capgras delusion and prosopagnosia (Lewis, 1987).

part of the test was not significantly different from chance.

It was also severely impaired in terms of the test's norms

(Warrington, 1984), and markedly different from her
excellent score of 49/50 correct on recognition memory for
words. MC performed both parts of the test satisfactorily
(faces 41/50 correct; words 41/50 correct).
Discussion

For both MC and ML, the Capgras delusion arose
in the context of an organic psychosyndrome. How
ever, at the time when neuropsychological testing was
carried out, neither patient was markedly confused

However,

it is necessary to treat this possible link

carefully (Ellis & Young, 1990). It seems unlikely that
there is any direct and sufficient causal connection;
prosopagnosic

patients do not usually experience the

Capgras delusion and, conversely, Capgras patients
still can recognise a number of familiar faces. This

is evident both in the observation that Capgras
patients still recognise that the ‘¿double'
looks like
their husband, wife or whoever, and in our finding
that although MC and ML's performance in recog

or disorientated, and both were able to co-operate
fully.

nising highly familiar faces was severely impaired,

Our findings show that face-processing impair
ments in patients with the Capgras delusion are not
restricted to the unfamiliar-face
matching tasks

our findings

it was by no means completely abolished. Moreover,
of face-processing

impairments

for

MC and ML were made when they were no longer
in the florid state of being convinced of the

698

YOUNG ET AL

veracity of the delusion. These points imply that
some additional factor is involved.
Lewis (1987) and Ellis & Young (1990) have
suggested that the basis of the Capgras delusion lies
in damage to neuro-anatomical pathways responsible
for appropriate emotional reactions to familiar visual
stimuli. The delusion would then represent the
patient's attempt to make sense of this puzzling
change. Of particular interest in this respect is Kaney

promise for uncovering the basis of this bizarre
phenomenon. Things need not be as strange as they
seem.
Acknowledgements
We gratefullyacknowledgethe support providedby ESRCgrant
R000231922. The SPECT scan was kindly conducted by Dr K.
Ebmeier and colleagues from the MRC Brain Metabolism Unit at

& Bentall's(1989)fmding that peoplewith persecutory the Royal Edinburgh Hospital. We thank Dr Freda Newcombe
delusions tend to attribute hypothetical negative
events to external rather than internal causes.
Persecutory delusions and suspiciousness are often
noted in Capgras cases, and were prominent clinical
features for both MC and ML at the time when they

were experiencing the Capgras delusion (MC thought
the duplicate husband was ill-intentioned, and ML
voiced a number of persecutory delusions). They

may, therefore, contribute to the fundamental mis
interpretation
in which the patients mistake a
change in themselves for a change in others (the
delusion that because altered affective reactions
make people seem strange, they must have been
‘¿replaced').
Our hypothesis is therefore that the Capgras
delusion reflects an interaction of impairments at two
levels: one set of contributory
factors involve
anomalous perceptual experience; the other factors
lead to an incorrect interpretation
of this. On this
account, the Capgras delusion typically involves close
relatives because these would normally produce the
strongest reactions, and hence suffer the greatest
discrepancy.
Since substantial parts of the pathways which
imbue visual stimuli with affective significance are
close to those involved in visual recognition, one

(RadcliffeInfirmary,Oxford)for allowingus to use control data
collectedfor her own studies.
References
BENTON, A. L., HAMSHER, K. DE S., VARNEY, N., et al (1983)

Contributions to Neuropsychological Assessment:A Clinical
Manual. Oxford: Oxford UniversityPress.
CU-I-rING,J. (1991) Delusional misidentification

and the role of the

right hemisphere in the appreciation of identity. British Journal
ofPsychiatry, 159, 70—75.
ELLIS, H. D. & YOUNG, A. W. (1990) Accounting

for delusional

misidentifications. British Journal of Psychiatry, 157, 239—248.
FORSTL,H., ALMEIDA,0. P., Owar@i,A. M., et al (1991) Psychiamc,

neurological and medical aspects of misidentification syndromes:

a reviewof 260cases. PsychologicalMedicine,21, 905—910.
BURNS, A., JAc0BY, R., et al (1991) Neuroanatomical
correlates of clinical misidentification and misperception in senile

dementia of the Alzheimer type. Journal of Clinical Psychiatry,
52, 268—271.
KANEY, S. & BENTALL, R. P. (1989) Persecutory

delusions

and

attributionalstyle.BritishJournal of MedicalPsychology,62,
191—198.

LEwIs,S. W. (1987) Brain imaging in a case of Capgras' syndrome.
British Journal of Psychiatry, 150, 117—121.
WARRINGT0N,E. K. (1984) Recognition

Memory Test. Windsor:

NFER-Nelson.
Youi@o,
A. W. (1992)Facerecognitionimpairments.Philosophical
Transactions of the Royal Society, B335, 47-54.
ELLIS, H. D., SZULECKA, T. K.,
processing
impairments
and delusional

et al (1990) Face
misidentification.

Behavioural Neurology, 3, 153—168.

would expect that few brain lesions will compromise
emotional

reactions

to visual stimuli without

also

affecting other visual functions involved in recognition
to some extent, and this would be consistent with
our observations of defective face-processing abilities
for MC and ML. Most Capgras patients will thus

*A. W. Young, BSc, PhD, DSc, CPsychol, FBPsS,
Professor of Psychology, Department of Psychology,
University of Durham, Science Laboratories, South
Road, Durham DHJ 3LE; I. Reid, MB, ChB,

show defective face-processing abilities because these
are, for neuro-anatomical
reasons, likely to occur
along with the fundamental problem in affective

Department of Psychiatry, University of Edinburgh,
Royal Edinburgh Hospital, Edinburgh; S. Wright,

reactions.
The Capgras delusion is one of the most striking

Psychiatry, Northern General Hospital, Sheffield;

neuropsychiatric

problems,

but the use of tests to

determine whether or not measurable cognitive
deficits play any role in the creation and maintenance
of the delusion is still at an early stage. However,
the present results indicate that the approach has

BMedBiol, MRCPsych, Lecturer

MB, ChB, MRCPsych,

D.

J.

Hellawell,

Senior

in Psychiatry,

Registrar,

University

Department

BSc, MPhil, Senior

of

Research

Assistant, Department of Psychology, University of
Durham, Science Laboratories, Durham
Correspondence

Face-processing impairments and the Capgras delusion.
A W Young, I Reid, S Wright and D J Hellawell
BJP 1993, 162:695-698.
Access the most recent version at DOI: 10.1192/bjp.162.5.695

References

This article cites 0 articles, 0 of which you can access for free at:
http://bjp.rcpsych.org/content/162/5/695#BIBL

Reprints/
permissions

To obtain reprints or permission to reproduce material from this paper, please write
to permissions@rcpsych.ac.uk

You can respond
to this article at
Downloaded
from

http://bjp.rcpsych.org/letters/submit/bjprcpsych;162/5/695
http://bjp.rcpsych.org/ on February 8, 2015
Published by The Royal College of Psychiatrists

To subscribe to The British Journal of Psychiatry go to:
http://bjp.rcpsych.org/site/subscriptions/