Figure 13.3 Diagram indicating the influence of self-discrepancies on attributions, and the subsequent impact of attributions on later self-discrepancies, as found in two experiments by Kinderman and Bentall (2000).
good excuse-making, as in ‘I’m sorry I’m late but the traffic was dreadful’) led to few changes in discrepancies between beliefs about the self and ideals (‘I’m still the sort of person I’d like to be’) and very few discrepancies between beliefs about the self and beliefs about the opinions of other people (‘and other people still think I’m wonderful’). However, external-personal attributions (‘I’m sorry I’m late but the police deliberately set all the traffic lights to red to stop me from getting here on time’), although also resulting in few discrepancies between beliefs about the self and ideals (‘I’m the sort of person I’d like to be’), led to increased discrepancies between beliefs about the self and the perceived beliefs of others (‘but other people hate me’). Of course, as we have seen, it is precisely these kinds of attributions that are made in excess by paranoid patients.19
Paranoid patients can therefore be accommodated within the model of the attribution–self-representation cycle if we assume that, like depressed patients, their stored knowledge about the self is very negative but, unlike depressed patients, they have somehow learnt to avoid making internal attributions for negative events, and instead tend to make external-personal attributions. Because of the cyclic relationship between attributions and self-representations, these efforts to attribute negative experiences to external causes may not be sufficient to ensure that their current beliefs about themselves are always positive.
This theory neatly ties together the evidence on attributions and the paranoid self, but leads us to conclude that it may be impossible to make simple predictions about the way that paranoid patients will answer self-esteem questionnaires. Indeed, it is easy to see why some patients appear to have what Peter Trower and Paul Chadwick describe as ‘poor me’ paranoid delusions (they express relatively positive opinions about themselves and believe themselves to be the innocent victims of imaginary conspiracies) while others appear to have ‘bad me’ delusions (they seem to have low self-esteem and apparently believe that others have good reasons for wanting to harm them).20 Depending on the precise nature of their stored self-knowledge, the circumstances in which they find themselves, and the extent to which they are able to attribute their most troubling experiences to external-personal causes, either of these types of delusional systems is a likely outcome of many iterations of the attribution–self-representation cycle.
Of course, sceptics might argue that a theory of this sort should not be trusted because it cannot be properly tested. (This is precisely the reason that many modern psychologists have rejected the theories of the early psychoanalysts.) Against this argument it can be said that some phenomena just are unpredictable (the weather is a familiar example), and establishing when this is the case is an important kind of scientific achievement. (This is one of the main goals of the mathematical science of non-linear systems theory, more popularly known as chaos theory.)21 More importantly, as we saw in Chapter 10, it is possible to make other kinds of predictions that can be subjected to experimental investigation.
For example, the theory suggests that under some circumstances the attributions of paranoid patients will not always be excessively self-serving. If something happens to activate their underlying negative beliefs about themselves, their subsequent attributions should be temporarily more pessimistic. We have already seen evidence that this actually happens. Figure 10.5 (p. 259) shows the data from an experiment in which depressed patients, paranoid patients and ordinary people were asked to make attributions for hypothetical negative events before and after completing an anagram task that included impossible questions. Before the task, the paranoid patients made excessively external attributions for negative events, an observation that is consistent with the results of previous studies. However, afterwards, their attributions had shifted to become much more internal (almost like those of depressed patients), presumably because their failure on the anagrams task had brought some of their negative beliefs about themselves to the surface. It is difficult to see why these changes would happen if the attribution–self-representation cycle was not a real phenomenon.
Why External-Personal Attributions?
There is one remaining puzzle about the role of attributions in paranoia. Given that paranoid patients appear to avoid making internal attributions for negative events, why don’t they just do what most of us do, and make external-situational attributions? After all, as we have seen, external-situational attributions are benign. They keep us from thinking bad things about ourselves, and from attributing malevolent intentions to other people.
There are two main explanations for the paranoid patient’s failure to generate situational explanations. First, hyper-vigilance to threat-related information might lead the paranoid person to notice person-relevant information more than ordinary people. Second, paranoid patients may suffer from some kind of cognitive disability that prevents them from noticing situational information. As there is quite good evidence for both of these mechanisms, they are clearly not mutually exclusive.
Attention and attributions
In the last chapter, we saw that paranoid patients excessively attend to22 and recall23 threat-related information. More intriguingly, they seem to be more proficient than ordinary people at identifying negative emotional expressions on the faces of others.24 The availability of this kind of information at the forefront of their minds is likely to make them attribute negative social interactions to something about the other people involved, rather than to themselves or to circumstances. For example, we can imagine a paranoid patient having a minor argument with an acquaintance and, detecting a subtle expression of suppressed anger, concluding that the acquaintance maliciously set out to provoke a disagreement.
If this idea is correct, it should be possible to influence ordinary people to make paranoid attributions for negative experiences by making them focus their attention on the actions of others. This effect has been demonstrated by social psychologists Ehud Bodner and Mario Mikulincer in a series of experiments conducted at Bar-Ilan University in Israel. In these experiments, ordinary people were asked to explain their performance after they had failed to solve a series of problems that were in fact unsolvable. When their attention was focused on themselves (for example, they could see a video camera pointing towards them, and their own faces on a nearby television screen), they made depressive attributions. However, when their attention was focused on the experimenter (because the video camera was pointing at the experimenter and the television screen was showing the experimenter’s face) the participants in the experiments were especially likely to make paranoid attributions. Interestingly, these effects only occurred when the participants believed that they had uniquely failed to solve the problems. If they knew that everyone had failed to solve them, they made neither depressive nor paranoid attributions.25
The difficulty of understanding situations
The second possible explanation for the paranoid patient’s tendency to make external-personal attributions is an inability to make situational attributions. To see why these kinds of attributions might present special difficulties, we need to look again at what happens when we try to explain things.
In Chapter 10, I argued that this process probably involves some kind of mental search that is terminated when a suitable explanatory construct is found. We rack our brains until we arrive at a satisfactory explanation and then stop. I also suggested that this search almost always begins with current beliefs about the self. In the case of ordinary people, if we can immediately think of aspects of the self that match the event we are trying to explain, we will attribute the event to an internal cause. On the other hand, if we are unable to find a characteristic of ourselves that matches the event, we will find ourselves searching for some other kind of cause, most likely something ex
ternal to ourselves. (Paranoid patients going through the same process presumably reject most candidate internal attributions for negative events, and move smartly onwards in order to search for external causes.)
Studies carried out by social psychologist Daniel Gilbert at the University of Texas suggest that this further search for external causes does not happen in a single step. Gilbert required ordinary people to make attributions about the behaviour of other people in relaxed circumstances and also when their minds were being kept busy with a competing task. Without the competing task, the participants in these experiments tended to make external attributions that were situational whereas, when their minds were otherwise occupied, they tended to make attributions that implicated fixed traits in the other people. Gilbert concluded that, when accounting for the behaviour of other people, we first of all make external-personal attributions and then later discount these in favour of more situational explanations in the light of whatever information is available about the circumstances. This second step requires mental effort, so it tends not to occur when our minds are occupied with other tasks.26
There might be many reasons why paranoid patients lack the mental resources necessary to generate external-situational attributions. It is possible that, during a psychotic episode, some of the attentional and memory deficits we considered in Chapter 7 limit the extent to which they can think about situational factors. It is also possible that the tendency to jump to conclusions and a strong motivation to avoid ambiguity (characteristics of deluded patients that we discussed in the last chapter) result in a rush to attributional judgement, so that the second step described in Dan Gilbert’s model is skipped. Another intriguing possibility is that the failure to understand other people’s mental states contributes to this deficit.
When I earlier considered Chris Frith’s suggestion that paranoid patients suffer from an impaired theory of mind (ToM),27 we saw that, overall, the available evidence suggests that psychotic patients in general seem to suffer from difficulties of this sort (although only when they are ill). That is, these difficulties do not appear to be specifically connected to paranoia. Nonetheless, I would now like to suggest that ToM deficits, even though they may not be specific in this way, might indirectly influence paranoid thinking by limiting the ability of patients to make situational attributions.
To see why an inability to think about the mental states of others might have this effect, imagine that you are walking down a road and a friend passing in the opposite direction ignores you. What do you think? Most people considering this type of event generate some kind of excuse for the friend, such as ‘She’s having a bad day’ or ‘She’s worried about something.’ Many of these kinds of explanations have a situational flavour – they explain the friend’s behaviour in terms of circumstances that are affecting her. Moreover, these kinds of explanations also require you to take your friend’s point of view, in other words, to use your ToM skills. If you are unable to do this, and especially if you are anxious to avoid blaming yourself, you may well attribute your friend’s behaviour to some kind of simple disposition or trait (‘He’s a bastard’ or ‘He’s selfish and unreliable’).*
In an attempt to test this prediction, Peter Kinderman, Robin Dunbar and I administered a ToM measure and also a measure of attributional style to a large group of students.28 As we had predicted, students who performed relatively poorly on our ToM task made more paranoid-style external-personal attributions than students who performed well, a finding that has since been replicated.29 Obviously, this study was limited by our use of students, rather than of people who were actually suffering from psychiatric symptoms. The necessary studies of the relationship between ToM skills and attributions in patients are currently being conducted by some of my postgraduate students.
More Understanding of the Ununderstandable
In this chapter I have argued that the paranoid world view arises from the tendency to make extreme self-protective attributions, together with the failure (for whatever reason) to take into account situational causes of events. Of course, this account does not mean that biological processes are not involved – attributions, self-representations and associated cognitive processes are presumably generated by circuits in the brain. Clues about the specific neural circuits that may be involved have emerged from studies of patterns of regional cerebral blood flow in deluded patients,30 and in patients with positive symptoms in general,31 which have implicated the left lateral prefrontal cortex among other areas. More recently, Nigel Blackwood and Rob Howard at the Institute of Psychiatry in London have begun a programme of research (in which I am a collaborator) focusing specifically on paranoia, in which fMRI is being used to study relevant cognitive processes in both patients and ordinary people. Initial findings indicate that, in ordinary people, attention to threatening statements relevant to the self activates a neural network involving the left lateral inferior frontal cortex, the ventral striatum and the anterior cingulate.32 It remains to be seen whether these areas are more highly activated in paranoid patients.
The reader will have noticed that many of the psychological processes implicated in the model I have described in this chapter are the same as those involved in abnormal mood. Indeed, the model is represented schematically in Figure 13.4 as a variation of the models of depression and mania previously presented in Figures 10.7 and 11.5. These parallels suggest a close relationship between depression, mania and paranoia, which bridges the Kraepelinian divide between dementia praecox and the affective psychoses.
In fact, I am not the first to suggest a close relationship between paranoid symptoms and depression. Previous researchers have noted that psychotic patients with paranoid symptoms are more likely to have a history of depression than those without.33 Some studies have also shown that paranoid personality characteristics are frequently observed in the close relatives of patients diagnosed as suffering from an affective disorder,34 but less commonly observed than might be expected in the relatives of patients diagnosed as suffering from schizo-phrenia.35 Some years ago, American psychologists Edward Zigler and Marion Glick pointed out that preoccupation with the self is a central feature of depression, paranoia and mania, but not of other symptoms of psychosis.36 Their own research also led them to conclude that social and occupational functioning is usually good before the onset of a paranoid or depressive illness, but typically poor in people who later develop other psychotic symptoms.
Zigler and Glick proposed that both paranoid schizophrenia and delusional disorder are therefore forms of camouflaged depression.
Figure 13.4 The model of paranoid thinking (unique features shaded).
This theory takes Kraepelin’s distinction between schizophrenia and the affective psychoses for granted, but draws the line between the two types of disorder in a novel way. It will be evident from earlier chapters that I do not think that this is the right solution to the problem of classifying the psychoses.
14
The Illusion of Reality
Trintano: Your Excellency, haven’t we seen each other somewhere before?
Rufus T. Firefly (Groucho Marx): I don’t think so. I’m not sure I’m seeing you now. It must be something I ate.
Duck Soup (1933)
Hallucinations may seem unpromising candidates for psychological analysis. The hallucinatory experiences described by patients often seem so frightening and senseless that it is easy assume that they are exclusively the product of some kind of brain malfunction.
When I first attempted the psychological treatment of patients who heard voices, I was not convinced that I had anything useful to offer. I was just one year out of the Liverpool training course and going through my divorce, and my mind was more focused on personal problems than on clinical innovation. The Mersey Regional Forensic Psychiatry Service, where I was working, was responsible for the care of psychiatric patients who had committed criminal offences or who were considered to be dangerous to others. Most patients resident in the service’s inpatient f
acility had been diagnosed as suffering from schizophrenia. However, before my arrival, the psychologists in the service had focused their efforts mainly on non-psychotic patients attending the outpatient clinic (mostly sex offenders), and had left work with the inpatients to nurses and psychiatrists. It was only because I had expressed a tentative interest in psychosis that one of the consultant psychiatrists encouraged me to see a few patients who were proving difficult to treat with medication.
Many years later, I can remember very little about the first patient I saw, a young woman who told anyone who cared to listen about her desire to kill old people. As she was at times quite paranoid, and had a history of violence, it seemed prudent to take this strange and frightening fantasy seriously. Unfortunately, although she freely admitted to an indiscriminate hatred of anyone above the age of 65, she was in other ways uncooperative. Unable to make any sense of her strange obsessions, or the voices she occasionally disclosed, I eventually admitted defeat and left decisions about her management to my wiser and more experienced colleagues.
Madness Explained Page 40
