Madness Explained
Page 47
Once the various psychotic complaints have been explained in this way, no ghostly illness remains that also requires an explanation. The complaints (particular classes of behaviours and experiences that have been singled out because they sometimes cause distress) are all there is.
And yet, I would not be surprised if the account I have so far given seems incomplete to many readers. Although it begins to explain what happens when someone has these kinds of unusual experiences, it does not seem to explain how the complaints come into being in the first place. In short, it seems to say very little about aetiology. This criticism lies at the heart of Mojtabai and Rieder’s objections to the symptom-orientated approach,2 which we briefly considered at the end of Chapter 6.
In the following three chapters I will try and counter these arguments by showing how the symptom-orientated approach can be expanded to provide a richer account of aetiology than has been possible within the Kraepelinian paradigm. In doing so, I will attempt to do justice to the complexity of psychiatric problems. As we have seen, symptoms sometimes congregate together, although not in a simple way that is amenable to a categorial system of classification. Symptoms also wax and wane with time, often unpredictably. These types of complexity are difficult to explain within the Kraepelinian paradigm but I will show that they can be readily understood within the context of the symptom-orientated approach.
I would like to begin with a few words of warning. The argument I will outline in the first part of this chapter will probably seem quite abstract. I ask the reader to stick with it, because it will help us to make sense of some of the scientific findings that will be presented later on. Also, because I will sometimes have to extrapolate from limited data, some of the conclusions I will reach in this chapter and the two that follow will be fairly speculative. In these circumstances I ask the reader to focus on the general principles that I am attempting to expound (about which I feel fairly confident) rather than the details, which may have to be amended as more evidence becomes available.
Thinking Carefully about Causal Processes
Many people assume that the main aim of psychiatric research is to establish the cause of each psychiatric disorder. Often, psychologists and psychiatrists encourage this assumption by prejudging some kinds of causal factors as pre-eminent. (For example, Mojtabai and Rieder seem to regard genetic causes as particularly important, whereas the British historian of psychiatry German Berrios3 seems to discount explanations unless they point to abnormalities in particular brain regions.) However, a moment’s reflection will reveal that causal factors come in many different shapes and sizes. Even a simple physical event is the product of a complex network of causal effects and circumstances. Consider an example offered by the philosopher Ted Honderich.4 In order for a match to ignite it must rub against a rough surface with a certain force. However, the match also has to be of a certain chemical composition and it must have been kept in a place that is dry for a period prior to ignition. Which of these factors should be singled out? To pursue this question further, we could choose to explain the ignition of the match in terms of its design and manufacture, or the actions of the person who strikes it. Rather than selecting one factor as pre-eminent, we need a way of classifying these different kinds of causal events and thinking about the relationships between them.
For our purposes it will help to distinguish between three general kinds of causes. (These distinctions are fairly arbitrary but will aid our thinking.) Aetiological factors lie somewhere in the individual’s past, and will be the focus of the next two chapters. In contrast, the factors that will concern us in the present chapter play a more immediate role in determining whether complaints occur at particular points in time. These in turn can be divided into events and circumstances that appear to elicit or trigger symptoms, and those that help to keep symptoms going once they have occurred.
The former can be described as proximate determinants because they occur at approximately the same time as symptoms. Proximate determinants provoke or exacerbate the cognitive abnormalities that coincide with complaints. For example, unpatterned environmental noise can be thought of as a proximate determinant of auditory hallucinations, as the presence of this kind of stimulation immediately increases the likelihood that someone will misattribute their inner speech to an external source.
Events and circumstances that keep complaints going once they have become established are sometimes described as maintenance factors. As the reaction-maintenance principle reminds us, some maintenance factors are psychological. For example, in Chapter 10, I described research by the American psychologist Susan Nolen-Hoeksema which has shown that some ways of trying to cope with depression, particularly ruminating, tend to prolong depressive episodes. Similarly, in Chapter 12, I discussed Tony Morrison’s proposal that paranoid beliefs are sometimes maintained by safety behaviours. Other maintenance factors are social or environmental. Poverty, stigmatization and unemployment all serve to keep psychotic people locked out of society and into the role of the psychiatric patient.5 Of course, this kind of social marginalization can also be a consequence of patients’ complaints. Friends, neighbours and potential employers may shun people who begin to have psychotic experiences.
Proximate determinants and maintenance factors affect the course of complaints – whether they wane so that the individual recovers, or whether they worsen so that the individual becomes more disabled. It will not be possible to provide an exhaustive account of these factors in a single chapter. Some (for example, the impact of environmental noise on hallucinations) have been described in previous chapters, so I will avoid repeating discussion of them here.
Functional Relationships
I will begin by introducing an abstract concept from mathematics that will facilitate our discussion. (Don’t worry, no knowledge of maths will be required to follow my argument!) In mathematics, the association between two variables is described by a functional relationship. For example, y = f(x) indicates that y is a function of x; in other words, as x changes, y changes also in some orderly fashion. The way in which y changes will, of course, reflect the precise mathematical equation that is used to express the relationship. For example, if y = x2, y will increase ever more dramatically for small increases of x; the graph of this particular relationship is a parabola.
This concept of a functional relationship should not be confused with the idea that certain types of processes have a function for the individual. There is nothing wrong with this idea, which is commonplace in biology (for example, we known that the function of the heart is to deliver blood to various parts of the body) and, indeed, I have resorted to it at various points in the preceding chapters (when I asserted that the human brain has evolved to sustain complex relationships, or that the function of the self-serving bias is to maintain positive beliefs about the self). However, this is not what I mean by a functional relationship in the present context.
Functional relationships provide a language for describing the interrelations between events in nature. To take a simple example from the physical world, the laws of classical mechanics describe how the acceleration and velocity of a moving body change when a force acts on it. As this example shows, mathematical specifications of functional relationships often lead to descriptions that are more precise than those that can be given in ordinary language. Unfortunately, there have so far been few attempts to construct mathematical descriptions of the processes involved in psychotic phenomena, although there are no reasons for supposing that such descriptions will not be possible in the future.
Even in the absence of mathematical models, the idea of a functional relationship remains useful in psychopathology, and the term functional analysis* is sometimes used to indicate the effort to describe, in general terms, those relationships that may be important in determining psychological complaints.6 This approach has at least three strengths. First, it is relatively devoid of philosophical baggage. By describing a network of relationships between different groups of eve
nts that ultimately lead to complaints, it is not necessary to single out any particular event or class of events as the ultimate cause. Second, this kind of description can alert us to the possibility that many of these relationships are non-linear. (A linear relationship is one in which a change in one variable leads to a linear increase or decrease in another; even the simplest relationships in nature are usually more complex than this.) Third, a functional analysis can highlight reciprocal interactions in which two or more processes influence each other, either directly or indirectly. As we have already seen, these kinds of interactions can be very important.
At this point, the reader may be forgiven for wondering precisely where this argument is heading. It may help to consider whether any functional relationships have been described in the preceding chapters. In fact, the accounts I have given of the psychological processes involved in symptoms have been packed with them. A give-away is the presence of the following symbol:
Figure 16.1 The symbol for a functional relationship.
So, Figure 10.7 (p. 269) shows some functional relationships that appear to be important in depression, Figure 13.4 (p. 345) shows some that are involved in paranoid delusions, and Figure 14.3 (p. 368) shows some that seem to play a role in auditory hallucinations. These examples establish that there is nothing mysterious or magical about functional analysis. However, as I will now show, it is an approach that enables us to cut through much of the confusion that exists about the complexity of psychiatric phenomena.
At last, the problem of psychiatric classification solved (?)
First, let us return to the problem of psychiatric classification. When considering the evidence reviewed in Chapter 4 some readers may have been struck by an apparent paradox. I concluded that chapter by arguing that the Kraepelinian paradigm seems to propose at the same time too many categories of disorder and also too few. The similarities between some definitions of schizophrenia and bipolar disorder seemed to push us towards the einheitspsychose or unitary psychosis concept. On the other hand, the factor-analytic evidence seemed to suggest that each of these diagnostic categories could be split into a number of separate components or symptom clusters (for example, according to Peter Liddle’s model, positive symptoms, negative symptoms, and symptoms of cognitive disorganization). At the same time, it is not immediately obvious why any symptoms should go with any other (for example, why some one who is suffering from delusions should have a high but less than certain probability of experiencing hallucinations). This question lies at the heart of the problem of psychiatric classification. Indeed, it is the problem of psychiatric classification, as a successful attempt to resolve it will make the problem go away.
Kraepelin’s attempt to solve the problem can itself be expressed as a set of functional relationships. In fact, I have already done so in Figure 1.2, a part of which is reproduced below as Figure 16.2. According to this model, symptoms go together because they are functionally related to the same underlying biological abnormalities. Most people who have worked in the field of psychiatry over the last century or so have implicitly endorsed this model, even when they have disputed Kraepelin’s version of it. Indeed, this model has been the driving force behind the quest to find the causes of schizophrenia and manic depression.
Figure 16.2 Kraepelinian paradigm expressed as a set of functional relationships.
Of course, the Kraepelinian paradigm is not the only possible model of the relationship between aetiological variables and symptoms. Symptoms could be functionally connected to several causal variables. In fact, these kinds of relationships are portrayed in many of the models outlined in previous chapters. For example, Figure 14.3 shows how hearing voices can be influenced by beliefs, environmental noise and emotional arousal.
This kind of model edges us towards an understanding of why symptoms sometimes but not invariably occur together. Perhaps they have some functional relationships in common. For example, in Chapters 12 and 13, we considered evidence that theory-of-mind deficits (the inability to understand the beliefs, thoughts and intentions of other people) may contribute to paranoid ideas. In Chapter 15 we saw that these kinds of deficits also appear to play a role in disordered communication. These relationships are portrayed together in Figure 16.3. Note that, according to this amalgamated model, we should expect paranoid delusions and disordered speech occasionally to occur together (because they are both influenced by theory-of-mind deficits). However, we should not expect this to happen very often, because theory-of-mind deficits have only an indirect influence on paranoid delusions (see Chapter 12), and because both symptoms are influenced by processes that they do not share in common.
Even this kind of model must be an oversimplification. One complication is that complaints sometimes cause each other. In Chapter 12, we considered Brendan Maher’s anomalous perception model, which
Figure 16.3 Overlapping functional relationships implicated in paranoia and incoherent speech.
suggested that delusions are sometimes rational interpretations of anomalous experiences.7 The most obvious way that this can happen is when a patient makes a delusional interpretation of a hallucinated voice. (Recall that, in Sondra Escher and Marius Romme’s prospective study of children who heard voices, some 9 per cent developed delusional interpretations of their experiences.)8 However, in Chapter 14 we also saw that hallucinations can be facilitated by certain kinds of beliefs (for example, as demonstrated in experiments in which patients have been given various kinds of suggestions).9 Which of these accounts of the relationship between hallucinations and delusions is correct? Given the evidence, both must be, as illustrated in Figure 16.4. If
Figure 16.4 Hallucinations can influence delusions, and delusions can influence hallucinations.
delusions can provoke hallucinations and hallucinations can provoke delusions, it is no wonder that they very often occur together. However, the fact that they sometimes occur in isolation shows that this association is not inevitable. (For this reason, in order to understand the mechanisms that are responsible for these complaints it has been necessary to consider them in isolation.)
Some depressive symptoms experienced by psychotic patients may be explained in this way. As we have already seen, abnormal mood often precedes the development of hallucinations and delusions. Although, in many patients, the high levels of depression experienced during an episode reduce as they recover from their psychotic episode, about a third go on to develop a severe post-psychotic depression.10 Max Birchwood and his colleagues at the University of Birmingham have shown that this type of depression is often a consequence of patients feeling humiliated and entrapped by their illness, and their expectation that they will be doomed to a low social status as a consequence.11 In this case, patients’ appraisals of one set of complaints (delusions and hallucinations) lead to new complaints (depression).
A final complication concerns less direct ways in which one complaint can increase the probability that another will occur. For example, there is evidence that social isolation and the absence of social support can maintain psychotic symptoms–the more that people who experience psychosis live in impoverished and unsupportive social networks, passing each day with very little contact with other people, the more likely it is that their complaints will persist or even worsen.12 It is easy to see how psychotic people can become isolated, either because their delusions leave them in fear of others, or because they are shunned by others who are frightened of them. But it is also easy to see how isolation can further provoke difficulties. Without the company of others, the paranoid person is not able to put his fears to the test. In the absence of external stimulation in the form of conversation, the source-monitoring problems of the person who hears voices will be exacerbated.
The lessons we have just learnt can be summarized as a general principle of functional interconnectedness, which might be stated formally in the following way:
Psychiatric complaints are connected by a rich network of functional relationships, which may be
causal, or reflect shared causal pathways. Co-occurrence of complaints depends on the strength and number of these functional relationships. The more functional relationships connecting two complaints, and the stronger they are, the more likely it is that they will occur together.
This principle gives a much richer account of the way in which symptoms cluster together than the Kraepelinian paradigm. It explains why some symptoms tend to occur together (for example, hallucinations and delusions) but also why attempts to break down psychosis into a small number of categories (schizophrenia, bipolar disorder and so on) have failed.
The Ups and Downs of Mental Health
We must now turn to the second level of complexity that I promised to address at the outset of this discussion. In ordinary language, people who suffer from psychiatric problems are sometimes said to be ‘unstable’. This description reflects our common experience that psychological well-being can vary from day to day, or even from hour to hour.
In the case of those of us who have been lucky enough to avoid the psychiatric clinic, these fluctuations may be quite mild. Occasionally we ‘get out on the wrong side of the bed’. We have good days and we have bad days. In contrast, day-to-day variations in the frequency and intensity of psychotic complaints may be much more marked. Sometimes superimposed on these fluctuations are changes that take place over longer intervals of weeks or months. We therefore speak of people having ‘relapses’ or ‘episodes’ – periods of time when their difficulties are quite severe – interspersed by periods of ‘remission’ when symptoms are attenuated, or have disappeared altogether. Unfortunately, much of the research that I reviewed in Part 3 has consisted of ‘snapshots’ of the relevant psychological processes. In the absence of further elaboration, these snapshots provide a rather misleading picture of what must be happening in the minds of both ordinary people and psychiatric patients. We are more like movies than still photographs.