Journal of Mental Health (1993)2, 223-238

Deconstructing the concept of ‘schizophrenia’

RICHARD P BENTALL

Department of Clinical Psychology, Whelan Building, Liverpool University PO Box 147,

Liverpool L.69 3BX

Abstract
In this articleI challengetwo assumptions that inform much of traditional research into psychosis: the
assumption that psychotic disorders fall into discrete types such as schizophreniaand the assumption
that psychotic experiences and behaviour have no meaning. Illustrating my argument with recent
research on hallucinationsand delusions carried out at Liverpool University I show how scientific
progress can be made by abandoning the concept of schizophrenia and focusing on particular
symptoms. In addition to the scientific advantages of this approach, it has the further advantage of
linking the experiencesof mentally ill people to the experiences of ordinary people.

Introduction In this article, in which I summarise some

‘Schizophrenia’ has been described as
amongst the most serious of illnesses afflict-
ing modem society. Between one half and
one per cent of people in the developed world
can expect to be diagnosedas ‘schizophrenic’
at some time in their lives (Torrey, 1987).
Yet our understandingof ‘schizophrenia’has
not kept pace with our understandingof other
serious diseases like cancer or heart disease.
Many different theories have been proposed
to account for the disorder, implicating vari-
ables as diverse as childhood trauma, early
family environment,life stresses, neuropsych-
ological, neurophysiological and neuroana-
tomical abnormalities, diet, genetic endow-
ment and even exposure to viruses (Neale &
Oltmans, 1980). Indeed, it is no exaggeration
to say that every variable known to influence
human behaviour has, at one time or another,
been identifiedas a potential cause of ‘schizo-
phrenia’.
arguments I have developed at greater length
in a series of previous publications, I will
suggest that it is time to abandon the concept
of ‘schizophrenia’, mainly because it has
outlived its usefulness. This suggestion is
usually attributed to sociological, anti-
psychiatric or even antiscientific attitudes.
However, I will attempt tojustify my position
on scientificgrounds, rather than on the basis
of ethical or philosophical arguments. This is
not because ethical or philosophical argu-
ments are irrelevant to the question of how
we construe mental illness but because I want
to show that it is possible to challenge the
traditional approach to psychiatric disorder
using the kinds of evidence which should be
highly valued by psychiatric researchers.
Adopting this approach, I will argue that the
concept of ‘schizophrenia’ has impeded the
development of adequate scientific accounts
of psychotic behaviour, that there are ways of
thinking about psychotic behaviour which do

223
224 Richard P Bentall
not depend on the kinds of diagnostic classi-
fications in wide use today, and that these
ways of thinking are more consistent with
recent developments in the psychological
and neurobiologicalsciencesthan what might
be called ‘the standard approach’ in psycho-
pathology. By way of illustrating this latter
point, I will describe some of the research
carried out by my colleagues and myself at
Liverpool University, in order to show how it
is possible to do meaningful scientific re-
search into psychotic behaviour without re-
course to the concept of ‘schizophrenia’.
The Standard Approach to
‘Schizophrenia’
It will help to remind ourselves of the
origins of the concept of ‘schizophrenia’.
Most current research into severe psycho-
pathology takes as its starting point an
approachto psychiatricclassificationwhich
has evolved from the pioneering work of
Emil Kraepelii in the second half of the last
century. Although he lacked the statistical
methods available to modem researchers,
Kraepelin was one of the first psychopa-
thologists to systematically collect data on
the course and outcome of psychiatric dis-
orders, believing that such data were cru-
cial for the development of an adequate
diagnosticsystem(Berrios&Hauser, 1988).
When thinking about how these data might
be interpreted he made a number of as-
sumptions which seem over-simple when
judged by the standards of today, but which
were reasonable for his time:
‘Judgingfrom our experience in internal
medicine it is a fair assumption that similar
disease processes will produce identical
symptom pictures, identical pathological
anatomy andan identical aetiology. If; there-
fore, we possessed a comprehensive knowl-
edge of any of these three fields - pathologi-
cal anatomy, symptomatology, or aetiology -
we would at once have a uniform and stand-
ard classificationof mental diseases. A simi-
lar comprehensive knowledge of either of the
other two fields would give us not just as
uniform and standard classifications, but all
of these classifications would exactly coin-
cide. ’
(Kraepelin (1907; quoted in Reider, 1974)
Applying this kind of reasoning to the data
he had collected from his patients, Kraepelin
believed that he had identified two major
types of psychiatric disorder. On the one
hand there was dementia praecox (under
which categoryhe grouped together a number
of syndromespreviously describedby Morel,
Heckerand Kahlbaum),characterisedby early
onset followed by a very poor outcome, and
which was later renamed ‘schizophrenia’by
Bleuler(l911). Ontheotherhand,therewere
the manic depressive psychoses character-
ised by a later onset and better outcome. This
distinction has informed all subsequent sys-
tems of psychiatric classification (see Boyle,
1990,for a historical account beginning with
Kraepelin and endingwith the third edition of
the American Psychiatric Association’s Di-
agnosticand StatisticalManual) and has been
described by Kendell and Gourlay (1970) as,
‘One of the cornerstones of modem psychia-
by’. It is a distinction which has become
increasingly problematic over the years be-
cause many patients exhibit signs and symp
toms belonging to both hypothetical syn-
dromes. This latter observation has led some
authors to suggest there may be a third cat-
egory of ‘schizo-affective’ psychoses
(Kasanin, 1933).
It is hard to overemphasise the impact of
Kraepelin’s model on the research method-
ologies which have been employed when
investigating psychotic disorders. One im-
plication of Kraepelin’scategorical approach
to psychiatric classification is that patients

should be studied according to their diag-
noses. Not surprisingly, therefore, the most
common paradigm for studying the psycho-
ses involves comparing patients with a diag-
nosis of ‘schizophrenia’ with individuals di-
agnosed as ‘normal’ (or perhaps as suffering
from some other kind of disorder), on the
assumptionthat those sufferingfrom ‘schizo-
phrenia’ will have something in common
(hopefullyof aetiologicalsignificance)which
will be absent in the case of those not suffer-
ing from the disorder. This approach is
almost as dominant in clinical psychology as
it is in psychiatry. Sarbin and Mancuso
(1980) surveyed The Journal of Abnormal
and Social Psychology (later renamed The
Jouml of Abnormal Psychology) between
the years 1959 and 1978, finding that 374
papers totalling 15.3 percent of the journal
space used the presence or absence of a
diagnosis of ‘schizophrenia’ as an independ-
ent variable in this way and there is no sign
that psychologists’ enthusiasm for this kind
of research has diminished in the years fol-
lowing their report. Of course, if ‘schizo-
phrenics’ do not, in fact, have something
uniquely in common which is of aetiological
significance then this strategy is doomed to
yield weak and inconsistant findings.
When looking for variables on which to
compare ‘schizophrenic’ and ‘nonschizo-
phrenic’ individuals, most psychological re-
searchers have assumed that ‘schizophrenia’
is a disorder which reflects dysfunctions in
the biological systems that sustain core cog-
nitive processes. This assumption reflects
Bleuler’s (191 1) belief that ‘schizophrenic’
symptoms are products of an underlying dis-
order of association and also Jaspers’ (1963)
later assertion that true psychosis is
‘ununderstandable’ and hence inexplicable
in terms of a patient’s personality or life
experiences. The popularity of this approach
in recent years to some extent reflects a
Deconstructing schizophrenia 225
prevailing scientific Zeitgeist in which it is
widely assumed that, ‘There is no such thing
as apsychiatry which is too biological’(Guze,
1989) and in which many clinicians believe
that the model form of psychiatric treatment
is by means of drugs prescribed following a
psychiatric interview lasting approximately
fifteen minutes (Andreasen, 1984).
An important but not often articulated dis-
tinction which merits attention in this context
is that between cognitive deficits and cogni-
tive biases. Deficits concern what the patient
cannot do and are typically studied using
emotionally neutral tests, such as traditional
memory and attention tasks. Biases, on the
other hand, concern the way in which the
processing of information is affected by its
content, and are typically studied by requir-
ing the individual to memorise or think about
emotionally significant material. It is impor-
tant to study cognitive deficits because they
provide evidence about the role of neuro-
psychological impairment in psychiatric dis-
orders. Indeed, it is natural to look for dys-
functions in the biological mechanisms sus-
taining cognition when gross deficits in the
ability to attend, memorise and reason are
observed. On the other hand, it is also impor-
tant to study cognitive biases because human
mental life has the property described by
philosophers as ‘intentionality’,which is to
say that it is a feature of all mental states
(thoughts, desires, beliefs, as well as what in
ordinary language are known as intentions)
that they are always about something (Tallis,
1991). This human ability to mentally repre-
sent the world and behave in response to that
representation depends in turn on the ability
to process information according to its con-
tent. When individuals show evidence of
abnormal cognitive biases in the absence of
cognitive deficits it is natural to assume that
these biases reflect past learning and unusual
life experiences.
226 Richard P Bentall
There is good evidence that cognitive defi-
cits can be found in patients diagnosed as
‘schizophrenic’ and some creative accounts
of these deficits have been proposed in recent
years (e.g. Frith & Done, 1987; Gray et al.,
1991; Hemsley, 1992). However, the possi-
bility that cognitivebiases may also be impli-
cated in the behaviours and experiences of
psychotic patients, and that these behaviours
and experiencesmay be in some sense ‘mean-
ingful’, has been almost entirely ignored by
contemporary researchers (Bentall, 1992a).
despite the fact that such biases have been
fruitfully studiedin the context of other kinds
of psychiatric disorder, notably anxiety and
depression.
There is no convincing evidence of a
schizophrenia syndrome
I have outlined two assumptions which I
believe inform most contemporary efforts to
understand psychosis: the assumption that
psychosis falls into discrete categories such
as ‘schizophrenia’ and the assumption that
the symptoms experienced by psychotic pa-
tients lack meaning and invariably reflect
underlying cognitive deficits. I now want to
outline some evidence that gives cause for
questioning the first of these assumptions.
Because I have reviewed this evidence else-
where (Bentall, Jackson & Pilgrim, 1988;
Bentall, 1992b,c)the account given here will
be relatively brief.
To begin with, it will be helpful to remem-
ber that the concept of ‘schizophrenia’ has
been used to denotea strikingly wide range of
behavioursand experiences,including hallu-
cinations, delusions, passivity experiences,
cognitive disorganization, disordered and
impoverished speech, anhedonia, flat affect
and social disabilities. These behaviours and
experiences are rarely a l l present in the same
individual. As a consequence ‘schizophre-
nia’ is a disjunctive concept, such that two
people may receive the diagnosis while hav-
ing little if anything in common (Bannister,
1968). Over the years different ‘schizophre-
nia’ researchers have highlighted different
symptoms as important when determining
whether the diagnosis should be applied to
particular individuals (Koehler, 1979).
Two criteria are important when we con-
sider the scientificusefulness of any diagnos-
tic system, namely the reliability and validity
of the diagnostic classifications it yields.
Reliability refers to the extent to which dif-
ferent observers can agree about the applica-
tion of a particular diagnosis. Validity is a
more complex criterion and refers to the
extent to which a diagnosis is meaningful in
terms of its internal consistency and
covariation with a range of scientifically in-
teresting variables. It is worth noting that a
diagnosiscan be reliable without being valid,
as in the case of the fictitious ‘Bentall’s
Disease’ which has the following ‘first rank‘
symptoms:long fingernails (greater than on9
inch on average), red hair (natural, not dyed)
and a preference for the music of Pink Floyd
(ownership of at least four albums). Al-
though this syndrome could be diagnosed
with a high degree of reliability (it would be
a relatively simple matter to measure the
length of fingernails, determine hair colour
and count Pink Floyd albums) applications
for research grants to study Bentall’s Disease
are not likely to be well received.
Because early studies of the reliability of
psychiatric diagnoses yielded disappointing
results, a number of investigatorshave devel-
oped structured psychiatric interviews, for
example, The Present State Examination
(Wing, Cooper & Sartorius, 1974), for reli-
ably eliciting symptom data from patients,
and have proposed operational criteria for
awarding diagnoses such as ‘schizophrenia’.
Perhaps the most widely used operational

criteria are those in DSM-ILI-R (American
Psychiatric Association, 1987), but many
others have been proposed. Unfortunately,
different operational criteria do not always
diagnose the samepatients as ‘schizophrenic’
(Brockington, Kendell & Leff, 1978;
McGuffin, Farmer & Harvey, 1992). As a
consequence, disagreements between clini-
cians about who meritsthe diagnosisof schizo-
phrenia have been replaced by, “A babble of
precisebut differing formulationsof the same
concept” (Brackington et a]., 1978).
A number of tests can be evoked to address
the validity of the ‘schizophrenia’ diagnosis.
For example, the concept of ‘schizophrenia’
should map on to a cluster of symptoms
which tend to occur together in real life.
Research in which multivariate statistical
methods have been applied to symptomsdata
collected from large groups of patients have
generally failed toreveal such acluster(S1ade
& Cooper, 1979; Blashfield, 1984). Indeed,
some authors have argued that the available
evidencefrom these kinds of studies points to
two (Crow, 1980) or even three (Liddle,
1987) independent ‘schizophrenia’ syn-
dromes.
A quite different way of addressing the
validity of a psychiatric diagnosis involves
determining whether it predicts course, out-
come or response to treatment. Ordinary
people, when consulting a physician, gener-
ally hope that the doctor’s diagnosis will
allow predictions about what is likely to
happen in the future and about which kinds of
treatment are likely to be useful. In the case
of psychotic disorders, however, the course
of illness is extremely variable and is better
predicted by social variables than by symp-
toms (Ciompi, 1984; Sartorius et al., 1987).
Moreover, although neuroleptic drugs are
usually regarded as the treatment of choice
for patients diagnosed as ‘schizophrenic’,
many ‘schizophrenic’ patients consistently
Deconstructing schizophrenia 22 7
fail to respond toneuroleptics (Warner, 1985;
Brown & Hertz, 1989) and some respond to
other drugs such as lithium (Delva &
Letemendia, 1982) or benzodiazapines
(Lingjaerde, 1982). Interestingly, in those
few trials in which drugs have been assigned
to patients irrespectiveof diagnosis,response
to neuroleptics seems to have been better
predicted by particular symptomsrather than
by broad diagnostic classifications (Kendell,
1989).
Taking all this evidence together, it is dif-
ficult to see why modem researchers con-
tinue to take the concept of ‘schizophrenia’
seriously. Indeed, we are inevitably drawn to
an important conclusion: ‘schizophrenia’
appears to be a disease which has no particu-
lar symptoms,which has no particular course
and which responds to no particular treat-
ment. It is therefore not surprising that
aetiological research has revealed that it has
no particular cause (Bentall et al., 1988).
One obvious implication of this account is
that, if we wish to make progress in the
understanding of serious psychiatric disor-
der, the concept of ‘schizophrenia’ will have
to be abandoned. In this regard the concept of
‘schizophrenia’ is similar to a number of
other concepts; for example ‘phlogiston’and
‘the luminiferous ether’, which were widely
employed by scientists for a time but which
turned out to be scientifically misleading in
the long term.
Perhaps one reason why the concept of
‘schizophrenia’has continuedto beemployed,
despite its poor validity, is because it is widely
believed that to abandon it would amount to
rejecting science and falling down some kind
of Langian abyss. It will, therefore, be useful
to anticipate a common objection to the argu-
ments I have just outlined, which is neatly
encapsulated in Kety’s (1974) famous re-
mark that, ‘If schizophrenia is a myth it is a
myth with a strong genetic component’. This
228 Richard P Bentall
objection has recently resurfaced in reviews
of Mary Boyle’s (1990) historical
deconstruction of the ‘schizophrenia’ con-
cept which, in many ways, parallels the argu-
ments sketched out here. Farmer (1991), for
example, takes Boyle to task for failing to
acknowledge recent developments in genet-
ics and the neurosciences which, she main-
tains, point to the existence of a wide range of
biological abnormalities in ‘schizophrenic’
patients.
In response to this kind of objection it
might be argued that the achievements of
biological psychiatry have been overstated
(see,for example,Marshall’s 1990review of
research on the genetics of ‘schizophrenia’),
a point which is conceded at times even by
biological scientists (Charlton, 1990). How-
ever, this response does little to answer the
widely held conviction that the identification
of biological abnormalities - any biological
abnormalities - in patients diagnosed as
‘schizophrenic’givessubstancetothe ‘schizo-
phrenia’ concept. In fact, the real problem
faced by those who would wish to defend the
concept of ‘schizophrenia’ on such grounds
is an embarrassment of riches. Indeed, the
genetic (Read, Potter & Gurling, 1992),
neuropsychological and neurochemical
(Seidman, 1984; Jackson, 1990) abnormali-
ties found in association with the diagnosisof
‘schizophrenia’ are so diverse and yet also so
nonspecific that they provide further evi-
dence against the existence of a single, dis-
crete ‘schizophrenia’ disease entity.
Recent biological findings, then, give us
few grounds for retaining the concept of
‘schizophrenia’ but, on the contrary, should
encourage us to abandon it. On this view
there is little point in carrying out further
research in which patients diagnosed as
‘schizophrenic’ are compared to individuals
diagnosed as normal or as suffering from
some other kind of disorder. Clearly, if we
are to abandonthis kind of research and at the
same time continue to inquire into the causes
of psychotic disorders new research strate-
gies are required. One obvious possibility
would be to try and develop a new system of
psychiatric classification, perhaps by divid-
ing the disorders now subsumed by the term
‘schizophrenia’ into subtypes such as those
described by Liddle (1987). In my view, any
such new typology would have to be dimen-
sionalrather than categorical,given that sub-
stantialresearchnowshowsthat ‘schizotypal’
characteristics (and even sometimes full-
blown ‘symptoms’ such as hallucinations or
paranoid ideas) are present in a substantial
minority of individuals who have not re-
ceived psychiatric treatment and who are, in
all other respects,quite normal (see Claridge,
1987,1990for detailed reviews of this litera-
ture).
A more radical alternative strategy would
be to abandon the attempt to classify psy-
chotic experiencesandbehavioursaltogether.
As Bannister (1968) suggested some time
ago, instead of trying to study ‘schizophre-
nia’ we could make what we observe in the
psychiatricclinicthe subjectof ourinvestiga-
tions. After all, patients rarely complain of
‘schizophrenia’; they complain about spe-
cific ‘symptoms’ such as hearing voices or
feeling persecuted. The fact that each symp-
tom of ‘schizophrenia’ can occur in the ab-
sence of any others suggest that each may be
associated with a particular profile of cogni-
tive abnormalities. It should therefore be
possible to identify those abnormalities im-
plicated in each symptom in the hope that this
will eventually allow us to construct symp-
tom-specific aetiological models.
As promised at the outset, in what remains
of this article I will illustrate the value of this
approach by describingsomerecent research
which my colleagues and I have carried out
intohallucinationsand persecutory delusions.

In discussing this work I will attempt to
challenge the second assumption underlying
much contemporary research into psychosis
which I identifiedearlier: the assumptionthat
psychotic disorders are meaningless and thus
reflect the exclusive influence of cognitive
deficits rather than biases. Again, because of
limitations of space the following treatments
will be relatively brief.
Hallucinations
It will be helpful to begin our discussion of
hallucinations with a few observations. Al-
though usually regarded as pathological, hal-
lucinatory experiences are highly valued in
some societies (Bourguignon, 1970) and are
reported by a substantial minority of other-
wise normal individuals in the developed
world (Slade & Bentall, 1988; R o m e &
Escher, 1989). There is considerable cross-
cultural variation in the extent to which hal-
lucinations are reported by ordinary people,
and in the kinds of hallucinations reported to
psychiatrists, with visual hallucinations be-
ing much more commonly recorded in unde-
veloped countries than in the West (Al-Issa,
1978; Sartorius et al., 1986).
Researchhas shownthathallucinationstend
to occur under conditions of stress (Slade,
1973) and are accompaniedby stress-related
changes in the autonomic nervous system
(Cooklin, Sturgeon & Leff, 1983). Auditory
hallucinations are also most likely to occur
under particular environmental circum-
stances, particularly when the individual is
exposed to restricted or unpatterned stimula-
tion (Margo, Hemsley & Slade, 1980;
Gallagher, Dinin & Baker, in press). There is
some indicationthat these kinds of hallucina-
tions are accompanied by ‘subvocalisation’
or small movements of the speech muscles
(McGuigan, 1966; Inouye & Shimizu, 1970)
although some of the relevant data is equivo-
Deconstructing schizophrenia 229
cal (Green & Kinsbourne, 1990). Consistent
with the idea that auditory hallucinations are
associated with speech activity, remote
telemetered EEG data indicates that the left
frontal areas of the brain are activated when
patients hear their ‘voices’ (Stevens &
Livermore, 1982). Moreover, voices tend to
be suppressed when patients engage in verbal
tasks such as humming or naming objects
(Margo et al., 1980; James, 1983).
The simplest way of accounting for these
findings is to suppose that hallucinations
occur when internal, mental processes are
misattributed to an external source. Various
versions of this theory have been proposed in
recent times. For example, Hoffman ( I 986)
has suggested that auditory hallucinations
reflect a speech-planning disorder, so that
voices are heard when the individual experi-
ences unintended speech acts. Frith and Done
(1987), on the other hand, have argued that
the misattribution of speech acts to an exter-
nal source may reflect a failure in a
neuropsychological mechanism responsible
for monitoring actions. In my own version of
this theory (Bentall, 1990), I have suggested
that the hallucinator’s misattribution of an
internal event to an external source may re-
flect both a failure to accurately discriminate
between internal and external events and
‘top-down’ influences: beliefs and expecta-
tions of the sort which normally guide our
interpretation of perceptions.
Investigating the way in which hallucinat-
ing and nonhallucinating people discrimi-
nate between self-generated and external
events has proved quite difficult, and at Liv-
erpool we have tried to address this problem
using avariety of strategies. In ourfmt study,
using signal detection theory, we compared
hallucinating and nonhallucinating patients
on the one hand, and students who scored
high or low on a hallucination questionnaire
on the other (Bentall & Slade, 1985). Sub-
230 Richard P Bentall
jects were asked to listen to a series of pas-
sages of white noise, half of which contained
a voice quietly speaking a single word. By
observing subjects’ judgements about when
they thought they had heard the voice it was
possible to calculate two values: perceptual
sensitivity (roughly, how well the individu-
al’s hearing is functioning) and perceptual
bias (roughly, the individual’s willingness to
believe that a stimulus has been presented in
conditions of uncertainty). Hallucinatingpa-
tients and hallucinating students showed a
greater bias towards detecting stimuli than
their respective controls but no differences
were observed on the measure of perceptual
sensitivity.
In a more recent study (Bentall, Baker &
Havers, 1991) we have employed a ‘reality
monitoring’paradigminwhich subjectswere
asked to discriminate between memories of
their own thoughts and memoriesof informa-
tion they had heard. The subjects- hallucinat-
ing psychiatric patients, nonhallucinating
psychiatricpatientsand normal controls- had
to answer simple clues which varied in diffi-
culty (e.g. ‘Thinkof a vehicle beginningwith
C’; ‘Think of a vegetable beginning with 0’)
and also listened to a list of easy and difficult
paired associates (eg. ‘Footwear-shoe’;
‘Country-Norway’).One weeklaterthey were
given a list of words which included the
answers they had given to the clues, words
from the list of paired associates and new
words. Hallucinators did not perform worse
thanthe psychiatriccontrolsin termsof over-
all accuracy. However, they were modestly
but significantlymore likely than the psychi-
ahic controls to mistaketheir own answersto
the difficult clues for words that they had
heard.
Clearly, much more work needs to be car-
ried out to determine why hallucinators mis-
take their thoughts for voices. However, the
failure to identify group differences in per-
ceptual sensitivity on the signal detection
task, together with the failure to identify
group differences in overall accuracy on the
reality monitoring task, might be construed
as evidence that core cognitive deficits are
not solely responsible for the hallucinator’s
difficulties. and that ‘topdown’ or concep
tual processes must therefore be implicated.
This hypothesis is consistent with the ob-
served cultural differences in the experience
of hallucinations - if one is brought up to
expect to see dead ancestors then imagined
dead ancestors are likely to be classified as
‘real’- and with the differences in perceptual
bias observed between hallucinators and
nonhallucinatorson the signal detection task.
Of course, it would be helpful to have direct
evidence that beliefs and expectations influ-
ence the hallucinator’s judgement about
whether a perceived event is self-generated
or originating from an external source. In
fact, at least two studies (Mintz & Alpert,
1972; Young et al.. 1987) have shown that
hallucinatorsrespond more readily than con-
trols to suggestions to see or hear particular
kinds of events.
This observation raises the question of
whether or not it might be possible to educate
hallucinating patients to reamibute their ap-
parently alien thoughts to themselves. In our
latest work, we have attempted to develop
just such a therapeutic strategy, which a p
pears to hold somepromise for some patients
(Bentall, Haddock & Slade, in press; Had-
dock, Bentall & Slade, in press) although it is
too soon to say whether this approach will be
generally effective.
Delusions
In contrast to research on hallucinations,
the investigation of cognitive processes in-
volved in delusions has proved to be more
straightforward. In Liverpool, we have fo-

cused our research on delusions of persecu-
tion. Although diagnostic manuals such as
DSM-III-R describe such beliefs as sympto-
matic of psychosis and qualitativelydifferent
from nondelusional beliefs, they clearly lie
on a continuum with ordinary beliefs and
attitudes(Strauss, 1969;Harrow,Rattenbury,
& Stoll, 1988). Indeed, Fenigsten & Vanable
(1992) have recently published a question-
naire which seemsto validly detect subclinical
paranoid traits in otherwise ordinary indi-
viduals.
It seems likely that the abnormal beliefs of
psychiatric patients reflect a variety of fac-
tors. For example, it is entirely possible that
beliefs of persecutionsometimesreflect genu-
inepersecutory experiences(Kaffman, 1983;
Mirowski & Ross, 1983). Maher (1974) has
argued that perceptual processes are also
involved and that many abnormal beliefs are
best seen as rational attempts to make sense
of anomalous experiences. Although Maher
was been unable to provide specific evidence
in favour of his hypothesis, and although it is
clear that abnormal beliefs often emerge in
the absence of anomalous perceptions
(Chapman & Chapman, 1988), it is equally
clear that some delusions are indeed percep-
tion-driven. This seems to be especially true
of delusional misidentifications in which a
loved one is believed to have been replaced
by a robot or impostor. Patients with these
particular delusions seem to suffer from
neuropsychological impairments of facial
recognition which rob them of the feeling of
familiarity usually experienced when en-
countering a well-known person (Ellis &
Young, 1990).
In our own research we have focussed
primarily on those cognitivebiases which are
usually regarded as reflecting motivational
factors. In one of our fmt studies, we em-
ployed Peterson et al.’s (1983) Attributional
Style Questionnaire (ASQ to compare the
Deconstructing schizophrenia 23 I
ways in which deluded and depressed pa-
tients reason about success and failure expe-
riences. We were able to show that, whereas
depressed patients, relative to normal con-
trols, tended to attributefailure to themselves
and success to external causes (consistent
with the theory of cognitive processes in
depression suggestedby Abramson,Seligman
and Teasdale, 1978), our deluded subjects
showed a relative tendency to blame others
for failure experiences and themselves for
success (Kaney & Bentall, 1988). This find-
ing, which was substantially replicated by
Candido and Romney (1990), can be ex-
plained by hypothesising that persecutory
delusions have a protective function, allow-
ing the individual to maintain a fragile sense
of self-esteem(Bentall, Kinderman & Kaney,
in press). In fact, research with normal sub-
jects has consistently indicated that ordinary
people have a ‘self-serving bias’ towards
blaming others for failure and themselves for
success, especially in situations where there
is a personal threat (Taylor, 1989). The ab-
normal cognitive style of paranoid patients
can therefore be thought of as an exaggera-
tion of a normal defensive process.
In arecent study (Kaney & Bentall. 1992),
we have studied this bias using another kind
of task. Subjects were asked to play two
computergames, one preprogrammedto yield
a successful outcome and one preprogrammed
to yield a failure outcome. After each game,
our subjects were asked to estimate the extent
to which they believed that they had control-
led the outcome. Consistent with previous
research by Alloy and Abramson (1979) the
depressed subjects in this study were ‘sadder
but wiser’, and returned low estimates of
control after each game. The normal subjects
showed the expected self-serving bias to-
wards claiming more control in the ‘win’
condition than the ‘lose’ condition. As pre-
dicted by our account of paranoid thinking,
232 Richard P Bentall
this bias was present to a significantly greater
degreein patients suffering from persecutory
delusions.
We have explored other kinds of reasoning
biases in deluded patients, for example in
their attributions for the behaviour of others
(Bentall, Kaney & Dewey, 1991). However,
in our most recent studies we have tried to
find directevidence of the defensivefunction
of delusional ideas. One way in which we
have tackled this problem is by looking again
at what happens when an individualattributes
an event to a particular cause. Petersonet al.’ s
ASQ requires subjects to think of a likely
cause for events such as ‘winning a prize’ or
‘goingon adate which turns out badly’. Once
they have thought of a likely cause and then
written it down the subjectsthen have to self-
rate the cause on a seven-point internality
scale (caused by self versus caused by other
people or circumstances). By comparing our
subjects’ self-ratingswith internality ratings
of theircausesprovidedby independentjudges
we were able to show that it was the ratings
more than the actual causes generated which
discriminatedbetweenthe groups(Kinderman
et al., 1993). Indeed, the deluded subjects
often generated causes for negative events
which they selfrated as unequivocally exter-
nal (to do with otherpeople or circumstances)
but which the judges rated as unequivocally
internal.
What kind of attributions would be made,
then, by deluded patients on an attributional
style measure which did not require an
effortful consideration of blame of the sort
required when completing self-ratings on the
ASQ? In order to answer this question we
employed a ‘pragmatic inference task’ de-
signed by Winters and Neale (1985). In this
test subjects listen to stonesdescribingthem-
selves involved in endeavours which have
either positive or negative outcomes. After
each story, the subjects are required to com-
plete multiple-choice questions, some of
which concern factual elements of the story.
For each story, however, one question re-
quires subjects to choose between two possi-
ble causes of the described outcome, one
internal and one external, which are equally
implied in the text (for example, in one story
about setting up a successful business, sub-
jects are asked to say whether success was
due to hard work or lack of competition). On
this task, our depressed subjects chose inter-
nal causes more often for negative than for
positive outcomes, consistent with their per-
formance on more conventionalattributional
style measures. However, the deluded sub-
jects, in direct contrast to their performance
on the ASQ,also attributed failure to them-
selves more often than they attributed suc-
cess to themselves (Lyon, Kaney & Bentall,
in press).
This finding is strong evidence that
persecutory delusions reflect cognitive bi-
ases which have a motivational function.
Moreover, they indicate that, underlying
persecutory delusions, there lies a cognitive
organization,particularly with respect to the
self, which is in many ways similar to that of
depressed patients. While the aetiology of
this cognitive organization must remain, at
present, a matter of speculation,it is difficult
to see how a purely biological account of
delusionscan be offered, although biological
factors may be involved to some degree
(Bentall, Kinderman & Kaney, in press).
The account of persecutory delusions that
emerges as a consequence of looking at mo-
tivational factors may well have implications
for therapy. In particular, it seems that di-
rectly challenginga patient’s delusional sys-
tem is no more liiely to be fruitful than
challenging the core religious or political
beliefs of ordinary people, which also reflect
strong motivational factors. Interestingly,
there are reports of successful cognitive-be-

havioural treatments for delusions (Watts,
Powell & Austin, 1973; Chadwick & Lowe,
1990)and, consistent with the account I have
just outlined, these treatments avoid chal-
lenging patients’ beliefs head-on.
Qualifications
In this article I have, first, suggested that
there are good scientific grounds for aban-
doning the concept of ‘schizophrenia’ and,
second, tried to show how the strategy of
focusing our studies on the specific difficul-
ties and experiences of psychoticpatientscan
yield important findings. Aside from the ob-
vious scientific benefits of such an approach,
I hope that readers will agree that it has the
effect of making psychotic behaviour more
comprehensible and thus more obviously re-
lated to ordinary behaviours and experiences
than might have otherwise been supposed. I
would like to finish by considering some
further implications of this approach.
First, it is worth restating that a symptom-
orientated approach to psychopathology is
not antibiological as is sometimes supposed.
There is absolutely no reason why biological
investigators should not study particular
classes of psychotic experience and behav-
iour in the way I have advocated in this
article. Indeed, some authors (e.g. Bomstein
et al., 1989) have argued that neuropsych-
ological data on psychosis can be best under-
stood in relation to symptoms rather than
broad classifications. An important caveat to
this recommendation is that biological re-
search should always go hand-in-hand with
social and psychological research. One im-
plication of the intentionality of psychotic
symptoms is that explanations in terms of
gross neurobiologicaldysfunction will never
provide a complete account of the causal
pathways which lead to particular psychotic
experiences. Indeed, data on the association
Deconstrucring schizophrenia 233
between biological abnormalities and psy-
chopathologyis intrinsicallyincomplete with-
out an understanding of the cognitive func-
tions implemented by the relevant biological
systems.
Second, the generalizabilityof the strategy
I have outlined to other types of psychopa-
thology is worth considering. A number of
authors have argued for the abandonment of
categorical classificationsof the neuroses on
the basis of very similar arguments to those
given here. For example, depressionandanxi-
ety, although usually regarded as separate
kinds of disorders, covary considerably in
practice and the effects of antidepressants
and minor tranquillisers do not seem to be
diagnosis-specific (Tyrer, 1990;Goldberg &
Huxley, 1992).If we look in more detail at the
concept of ‘depression’ we see that it appar-
ently denotes a variety of behaviours and
experiences - for example low self-esteem,
guilt, dysphoria, apathy, fatigue and loss of
appetite- which have defied subclassification
into types (Craighead, 1980). The apparent
contradictions which exist between different
theories of ‘depression’ (for example, theo-
ries implicating self-deprecating cognitive
biases, cf. Abramson, Alloy & Metalsky
(1988) and those implicatingcircadian disor-
ders, cf. Healy & Williams (1988)) begin to
dissolve if we assume that different abnor-
malities are associated with different symp-
toms (for example,that circadian dysrhythmia
is implicated in fatigue whereas self-depre-
cating cognitive biases are associated with
low self-esteem).
Third, although I have argued that no con-
sistent pattern of association is observed be-
tween different kinds of psychotic symp-
toms, it must be acknowledged that many
patients experience more than one symptom.
Any adequatetheory of psychopathologywill,
therefore, have to explain why several differ-
ent symptoms are often present in the same
234 Richard P Bentall
person. The solution to this problem is not to
suppose, as psychiatric writers have tradi-
tionally done, that covariation of symptoms
in one person inevitably indicates a single
underlying cause, but to recognise the rich
and varied ways in which symptoms can be
functionally interconnected. Taking an ex-
ample from the depression literature, low
self-esteem,if it leads to inactivity in the face
of threatening life events, may precipitate
circadian dysrhythmiaand hence fatigue. On
the otherhand, in otherpatientsfatiguecaused
by circadian dysrhythmiamay be interpreted
as evidence of personal inadequacy,with low
self-esteem following as a consequence
(Healy & Williams, 1988). In the psychotic
domain it is clear that the delusions of some
patients are driven by anomalous perceptions
(Maher, 1974) but it is equally clear that, in
other cases, hallucinatory experiences are
driven by abnormal beliefs (Bentall, 1990).
Finally, I wish to return to the ethical as-
pects of psychopathology which I have pur-
posefully avoided until this point. Although
arguments against the concept of ‘schizo-
phrenia’ areoftendismissedasantipsychiatric
or even antiscientific, the contribution of
values to psychiatric decision-making can-
not be avoided. As Szasz (1961)has so force-
fully demonstrated, the attribution of ‘ill-
ness’ inevitably involves a valuejudgement,
although he was wrong to suppose that this
was the case only in the field of psychiatric
medicine (Bentall & Pilgrim, 1993). Recog-
nition of the fact that values influencepsychi-
atric decision-making need not delay scien-
tific research into the causes of psychosis;
precisely the same strategy to that outlined
here could be used to investigatethe causesof
positively valuedandhence ‘nonpathological’
experiences such as ‘happiness’ (Bentall,
1992d).In our pursuit of the causes of insan-
ity, however, we need to be sensitive to the
values and experiences of those individuals
who are the subject of our inquiries. Indeed,
our years of studying ‘schizophrenic’experi-
ences at Liverpool has taught us that thiskind
of work is best conducted in active partner-
ship with those who have first-hand knowl-
edge of psychotic states.
Itisinthis way thatwhatmightbecalledthe
‘gung-ho’ biological approach which typi-
fies much current psychiatric thinking so
conspicuously fails users of psychiatric serv-
ices: not because biological factors are unim-
portant but because the biological approach,
when taken to extremes, encourages an us-
and-them distinction between the ‘mad‘ and
the ‘sane’. In reality, of course, madness and
sanity are but points on a continuum along
which we are all inclined to move during
difficult moments in our lives.
Acknowledgement
Much of the research summarised in this
paper has been supported by grants from the
Medical Research Council(to RichardBentall
and Peter Slade for the study of cognitive-
behaviouralinterventionsfor patients suffer-
ing from auditory hallucinations) and the
Wellcome Trust (to Richard Bentall for the
study of cognitive processes involved in de-
lusions and hallucinations).
References
Abrarnson, L,Y.,Seligman.M. E. P. & Teasdale.
J. D. (1978). Learned helplessness in humans:
Critique aud reformulation. Journal of Ab-
normal Psychology,78.40-74.
Al-Issa, I. (1978). Social and cultural aspects of
hallucinations. Psychological Bulletin. 84.
570-587.
Alloy, L. B. & Abramson. L.Y. (1979). Judge-
ment of contingency in depressed and
nondepressed students: Sadder but wiser?
J o u m l of Experimental Psychology: Gen-
eral, 108,441485.

American PsychiatricAssociation.(1987). Diag-
nostic and Statistical Manunl of Mental Dis-
orders. 3rd edn, Revised . Washington:
APA.
Andreasen, N.C. (1 984). The Broken Brain: The
Biological Revolution in Psychiatry. New
York, Harper & Row.
Bannister, D. (1968). The logical requirementsof
research into schizophrenia. British Journal
ofpsychiutry, 114, 181-188.
Bentall, R.P. (1990). The illusion of reality: A
review and integration of psychological re-
search on hallucinations. Psychological Bul-
letin, 107, 82-95.
Bentall, R.P. (1992a). Psychological deficits and
biases in psychiatric disorders. Current Opin-
ion in Psychiatry, 5.825-830.
Bentall, R.P. (1992b).Theclassificationof schizo-
phrenia. In :Kavanagh,D.J. (Ed.) Schizophre-
nia: An overview and practical handbook.
London, Chapman & Hall.
Bentall, R.P. (1992~).Reconstructing psychopa-
thology. The Psychologist, 5,61-65.
Bentall, R.P. (1992d). A proposal to classify
happiness as a psychiatric disorder. The Jour-
nal of Medical Ethics, 18,9498.
Bentall, R.P., Baker, G. & Havers, S. (1991).
Reality monitoring and psychotic hallucina-
tions. British Journal of Clinical Psychology,
30,213-222.
Bentall, R.P., Haddock, G. & Slade, P.D. (in
press). Cognitive-behaviourtherapy for audi-
tory hallucinations: From theory to therapy.
Behavior Therapy.
Bentall. R. P.. Jackson, H. F.. & Pilgrim, D.
(1988). Abandoning the concept of “schizo-
phrenia”: Some implicationsof validity argu-
ments for psychological research into psy-
chotic phenomena. British Journal of Clini-
cal Psychology, 27, 156-69.
Bental1,R. P., Kaney. S. & Dewey,M. E. (1991).
Paranoia and social reasoning: An attribution
theory analysis. British Journal of Clinical
Psychology, 30. 13-23.
Deconstructing schizophrenia 235
Bentall, R.P., Kindennan, K, and Kaney, S. (in
press). Self, attributional processes and ab-
normal beliefs: Towards a model of
persecutory delusions. Behaviour Research
and Therapy.
Bental1,R.P. &Pilgrim,D. (1993).Thomas Szasz,
crazy talk and the myth of mental illness.
British Journal of Medical Psychology, 66,
69-76.
Bentall, R.P. &Slade, P.D. (1985). Reality testing
and auditory hallucinations: A signal-detec-
tion analysis. British Journal of Clinical
Psychology, 24,159169.
Bemos, G. & Hauser, R.(1 988). The early devel-
opment of Kraepelin’s ideas on classification:
A conceptual history. Psychological Medi-
cine, 18, 813-821.
Blashfield, K. (1984). The Classification ofPsy-
chopathology: NeoKraepelinian and Quanti-
tative Approaches. New York, Plenum.
Bleuler E. (1950). Dementia Praecox or the
Group of Schizophrenias. (originally pub-
lished 1911, English translation; Zinkin E.).
New York. International Universities Press.
Bomstein, R.A., Nasrallah. H.A., Olson, S.C.,
Coffman, J.A., Torello. M. & Schwarzkopf,
S.B. (1989). Neuropsychological deficit in
schizophrenic subtypes: Paranoid,
nonparanoid and schizoaffective subtypes.
Psychiatry Research, 31, 15-24.
Bourguignon, E. ( 1970). Hallucinations and
trance: An anthropologist’s perspective. In :
W. Keup (Ed.) Origins and Mechanisms of
Hallucinations. New York, Plenum.
BoyIe. M. (1990). Schizophrenia: A Scientific
Delusion. London, Routledge.
Brockington. LF., Kendell, R.E. & Leff, J.P.
(1978).Definitionsof schizophrenia:concord-
ance and prediction of outcome. Psychologi-
cal Medicine, 8, 399-412.
Brown, W.A. & Hertz, L.R. (1989). Response to
neuroleptic drugs as a device for classifying
schizophrenia. Schizophrenia Bulletin, 15,
123-129.
236 Richard P Bentall
Candido, C. & Romney, D.M. (1990).
Amibutional style in paranoid vs depressed
patients. British Journal of Medical Psychol-
ogy, 63,355-363.
Chadwick,P. &Lowe,C.F. (1990).The measure-
ment and modification of delusional beliefs.
Journal of Consulting and Clinical Psychol-
ogy, 58,225232.
Chapman, L. J. & Chapman, J.P. (1988). The
genesis of delusions. In: T. F. Oltmanns & B.
A. Maher, (Eds) Delusional Beliefs. New
York, Wiey.
Charlton, B.G. (1990). A critique of biological
psychiatry. Psychological Medicine, 20.3-6.
Ciompi, L. (1984). Is there really a schizophre-
nia?: The longterm course of psychotic phe-
nomena. British Journal of Psychiatry. 145,
636-640.
Claridge, G.S. (1987). ‘The schizophrenias as
nervous types’ revisited. British Journal of
Psychiatry, 151,735-743.
Claridge. G.S. (1990). Can a disease model of
schizophrenia survive? In :Bentall, R.P.(Ed)
Reconstructing Schizophrenia. London,
Routledge.
Cooklin,R,Sturgeon,D.&Leff,J.P.(1983).The
relationship between auditory hallucinations
and spontaneousfluctuation of skin conduct-
ance in schizophrenia British Journal of
Psychiatry, 142,47-52.
Craighead, W.E.(1980). Away from a unitary
model of depression. Behavior Therapy, 11,
122-128.
Crow, T.J. (1980). The molecular pathology of
schizophrenia: More than one disease proc-
ess? British Medical Journal, 280.66-68.
Delva, N.J. & Letemendia, F.J.(1982). Lithium
treatmentin schizophreniaandschizoaffective
disorders. British Joumal of Psychiatry, 141,
387-400.
Ellis. H.D. & Young,A.W. (1990). Accounting
fordelusionalmisidentifications.BritishJour-
nal of Psychiatry, 157, 239-248.
Falmer, A. (1991). A review of ‘Schizophrenia:A
scientific delusion’ by Mary Boyle. British
Journal of Clinical Psychology, 30,280.
Fenigsten, A. & Vanable, P.A. (1992). Paranoia
and self-consciousness. Journal of Personal-
ity & Social Psychology, 62, 129-138.
Frith, C. D. & Done, D. J. (1987). Towards a
neuropsychology of schizophrenia. British
Journal of Psychiatry, 153,437-443.
Gallagher. A.G. Dinin, T.G. & Baker, LJ.V. (in
press). The effects of varying auditory input
on schizophrenic hallucinations: A replica-
tion. British Journal of Medical Psychology.
Goldberg. D. & Hwdey. P. (1992). Common
Mental Disorders. London, Routledge.
Gray, J.A.. Feldon, J., Rawlins, J.N.P., Hemsley.
D.R. & Smith, A.D. (1991). The
neuropsychology of schizophrenia. Behav-
ioural und Brain Sciences, 14.1-84.
Green, M.F.& Kinsbourne, M. (1990). Subvocai
activityand auditoryhallucinations: Cluesfor
behavioral treatments? Schizophrenia Bulle-
tin, 4,617-625.
Guze. S . (1989). Biological psychiatry: 1s.there
any other kind? Psychological Medicine, 19,
3 15-323.
Haddock,G.,Bentall,R.P. & Slade, D. (in press).
Psychological treatment of chronic auditory
hallucinations:Two case studies.Behavioural
Psychotherapy
Harrow, M., Rattenbury, F. & Stoll, F.(1988).
Schizophrenicdelusions: An analysisof their
persistence, of related premorbid ideas and
threemajor dimensions. In: T. F.Oltmanns&
B. A. Maher, (Eds) Delusional Beliefs. New
York, Wiley.
Healy, D. & Williams, J.M.G. (1988).
Dysrhythrma, dysphoria and depression: The
interaction of learned helplessness and
circadian dysrhythmia in the pathogenesis of
depression. Psychological Bulletin, 103,163-
178.
Hemsley, D.R. (1992). Disorders of perception
and cognition in schizophrenia. Revue
Europeene dePsychlogieApplique, 42.104-
114.
Hoffman, R. E. (1986). Verbal hallucinationsand
languageproduction processes in schizophre-
nia Behavioural and Brain Sciences, 9,503-
548.

Inouye, T. & Shimizu, A. (1970). The
electromyographic study of verbal hallucina-
tion. Journal oflvervous andMental Disease,
151.415422.
Jackson, H.F. (1990). Are there biological mark-
ers of schizophrenia? In :Bentall, R.P. (Ed)
Reconstructing Schizophrenia. London,
Routledge.
James, D.A.E. (1983). The experimental treat-
ment of two cases of auditory hallucinations.
British Journal of Psychiatry, 143.5 15-516.
Jaspers, K. (t963). General Psychopathology
(trans. J. Hoenig & M.W. Hamilton). Man-
Chester, Manchester University Press.
Kaffman (1983). Paranoid disorders: Family
sources of the delusional system. Journal of
Family Therapy, 5 , 107-116.
Kaney, S. & Bentall. R. P. (1988). Persecutory
delusionsand attributionalstyle. BririshJour-
nal of Medical Psychology, 62, 191-8.
Kaney, S. & Bentall, R. P. (1992). Persecutory
delusions and the self-serving bias: Evidence
from a contingencyjudgement task. Journal
of Nervous andMental Disease, 62,19 1- 198.
Kasanin, J. (1933). The acute schizoaffective
psychoses. American Journal of Psychiatry,
13.97-126.
Kendell, R.E. (1989). Clinical Validity. In: Rob-
ins, L.N. & Barrett, J.E. (Eds) The VaZidityof
Psychiatric Diagnosis. New York, Raven
Press.
Kendell, R.E. & Gourlay, J.A. (1970). The clini-
cal distinction between the affective psycho-
ses and schizophrenia British Journal of
Psychiatry. 117.261266.
Kety, S.S.(1974). Fromrationalization to reason.
American Journal of Psychiatry. 131, 957-
963.
Kindennan P., Kaney S., Morley S. & Bentall
R.P. (1992). Paranoia and the defensive
attributional style: Deluded and depressed pa-
tients’ attributions about their own attribu-
tions. British Journal ofMedical Psychology,
65.371-383.
Deconsrructing schizophrenia 237
Koehler. K. ( 1979).First rank symptomsof schizo-
phrenia: Questions concemingclinicalbounda-
ries. British Journal of Psychiatry, 134,236-
248.
Liddle, P.F. (1987). The symptoms of chronic
schizophrenia: A reexamination of the posi-
tive-negative dichotomy. British Journal of
Psychiatry, 151, 145-151.
Lingjaerde, 0. (1982). Effects of the
benzodiazapine derivative estazdam in pa-
tients with auditory hallucinations: A
multicentre double-blind cross-over study.
Acta Psychiatrica Scandinavica, 63,339-354.
Lyon, H.M., Kaney, S.& Bentall, R.P. (in press).
The defensive function of persecutory delu-
sions: Evidence from attribution tasks. Brit-
ish Journal of Psychiatry
Maher. B. A. (1974). Delusional thinking and
perceptual disorder. Journal of Individual
Psychology, 30.98- 1 13.
Margo, A., Hemsley. D. R. & Slade. P. D. (1980).
The effectsof varying auditory input on schizo-
phrenic hallucinations. British Journal of
Psychiatry. 139, 122-127.
Marshall, R. (1990). The genetics of schizophre-
nia: Axiom or hypothesis? In : Bentall, R.P.
(Ed.) Reconsrmcting Schizophrenia . Lon-
don, Routledge.
McGufin, P., Farmer, A. & Harvey, I. (1991). A
plydiagnostic application of operational cri-
teria in studies of psychotic illness. Archives
of General Psychiatry, 48,764-770.
McGuigan, F.J. (1966). Covert oral behaviour
and auditory hallucinations. Psycho-
physiology, 3,73-80.
Mintz, S . & Alprt, M. (1972). Imagery vivid-
ness, reality testing and schizophrenic hallu-
cinations. Journal of Abnormal Psychology,
19,310-316
Mirowsky, J. & Ross, C.E. (1983). Paranoia and
the structure of powerlessness. American
Sociological Review, 48,228-239.
Neale, J. & Oltmans, T.F. (1980). Schizophrenia.
New York: Wiley.
Peterson, C., Semmel. A., Von Baeyer, C.,
Abrarnson. L. Y.,Metlasky, G. I. & Seligman,
238 Richard P Bentall
M. P. E. (1982).The Attributional Style Ques-
tio~aire.Cognitive Therapy and Research,
6,287-300.
Read, T., Potter, M. & Gurling, H.M.D. (1992).
The genetics of schizophrenia. In: D. J.
Kavanagh, (Ed) Schizophrenia:An Overview
Md Practical Handbook. London. Chapman
& Hall.
Reider, 0.(1974). The Origin of Our confusion
about schizophrenia. Psychiatry, 37,197-208.
Romme, M. & Escher, A. (1989). Hearing voices.
Schizophrenia Bulletin, 15,209-216.
Sarbin. T.R. & Mancuso. J.C. (1980). Schizo-
phrenia: Diagnosis of Moral verdict? OX-
ford, Pergamon.
Sartorius,N., Jab1ensky.A.. Korten. A., Ernberg,
G.,Anker, M., Cooper. J.E.&Day. R. (1986).
Early InanifeStatiOnS and first Contact h i -
dence of schizophrenia in different cultures.
Psychological Medicine, 16,909-928.
Warner, R. (1985). Recoveryfrom Schizophre-
Sartorius,N., Jablensky,A., Emberg. G..hff. 1..
Korten, A. & Gulibant, W. (1987). Course of
schizophrenia in differentcounuies: 'Ine
Warn, F. N., Powell, E. G. & Austin, S. V.
results of a WHO comparative5-year follow-
In : Hafner* H.. W.G*
GattazT
Janzarik, W. (Eds) Search for the Causes of
'(1973).Themodifcationof abnomalbeliefs.
Schizophrenia .Berlin: Springer.
Wing, J.K., Cooper, J.E. & Sartorius, N. (1974).
Slade,P. D. (1973). The psychological hvestiga-
iion and treatment of auditory hallucinations:
a second case report. British Journal of Medi-
~ aPsychology,
l 46,293-296.
Winters, K. C. & Neale. J.M.(1985). Mania and
s1ade3 p*D- Bentall, (1988). senrory
Deception: A scientific analysis of hallucina-
tion. London, Croom Helm..
'lade* P'D. cooper* (1979). some 'Oncep
tualdifficultieswiththe term 'schizophrenia':
An alternativemodel. British Journal of So-
cial and Clinical Psychology, 18,309-317.
Seidman, L.J. (1984). Schizophrenia and brain
dysfunction: An integration of neurodiagnostic
findings. Psychological Bulletin, 94, 1%
238.
Stevens,J.R.&Livermore, A.( 1982).Telemetered
EE€i in schizophrenia: Spectral analysis dur-
ing abnormal behaviour episodes. J o u d of
Neurology, Neurosurgery andPsychiatry, 45,
385-395.
Strauss, J. S. (1969). Hallucinations and delu-
sionsas points on continua function. Archives
of General Psychiaw, 21,581-586.
szasz, T.S. (1961). The Myth of Mental Illness.
New York, Harper and Row.
Tallis, R. (199 1) A critique of neuromythology.
In :Tallis, R. &Robinson, H. (Eds) Pursuit of
Mind. London. Carcanet.
Taylor, S.E. (1989). Positive Illusions. New
YO&. Basic Books.
Tomey, E.F.(1987).&valence studies in schizo-
phrenia. British Journal of Psychiatry, 150,
598-608.
Tyrer, P. (19%). The division of neurosis: A
failed classification. Journal of the Royal
sociery of Medicine, 83,614616.
nia: pvchiatry and Political Economy. Lon-
don, Routledge and Kegm Paul.
British Journal of Medical p s y ~ ~ l o g&,
y,
356363.
TheDescription and of psychj-
atric sy,,,ptomr. London, CambridgeUniver-
sity press.
low selfesteem. Journal o f A b n o d psy-
chology,94, 282-290,
Young,H.F.,Bentall,R.P.,Slade,P.D.&Dewey,
M.E. (1987). The ofbrief&@uctions and
suggestibility in elicitation of hallucina-
tionsin and psychiatric subjects. Jour-
nal of Nervous and Mental Disease, 175,41-
48.
Reprintsfrom Dr Richard B e n d , Departmentof
Clinical Psychology, Whelan Building. Liver-
pool University. PO Box 147, Liverpool L69
3BX.