Unfinished Business
In this chapter I have tried to tell a story about how hallucinations have become amenable to psychological analysis. I have focused on auditory hallucinations, and have shown that they occur when people mistake their inner speech for the speech of others. This mistake seems to result from a failure of the psychological mechanisms that normally allow us to tell the difference between the things that we are thinking and the things that we are hearing. However, as a model of hallucinations, the story is obviously incomplete.
Some researchers, notably Ivan Leudar and Phil Thomas, have questioned whether it is accurate to characterize hallucinating people as poor at source monitoring, because people who hear voices usually make very clear distinctions between their hallucinations and their thoughts.75 The implication of this observation is that the hallucinating person experiences only some mental events as alien, and correctly identifies others as self-generated. A related objection to the account that I have offered is that it has treated source monitoring as an ‘all-or-nothing’ phenomenon. The experiences of patients who have pseudohallucinations or who suffer from thought insertion suggest that source monitoring can fail to varying degrees or in different ways. As philosophers G. Lynn Stephens and George Graham have pointed out, thought insertion seems to be particularly difficult to explain, because patients having this symptom describe their experiences as thoughts but deny that the thoughts are their own.76
It seems likely that the selective failure of source monitoring can be partly explained by taking into account the environmental factors that are known to influence source-monitoring judgements. However, other factors almost certainly play a role. Earlier, when I discussed the work of Marcia Johnson, I pointed out that source monitoring in ordinary people is less efficient for automatic thoughts than for thoughts that require considerable cognitive effort. (When you struggle hard to think of something, you know that it is you who is doing the thinking – a modern spin on Descartes’ famous dictum ‘cogito ergo sum’.) It therefore follows that highly automatic thoughts are particularly likely to be misattributed to an alien source. Of course, as we discussed in the context of depression, negative thoughts about the self are usually highly automatic, so it is not surprising that psychiatric patients often mistake these kinds of thoughts for voices. Interestingly, when Tony Morrison and Caroline Baker questioned a group of hallucinating patients about their experiences, they found that they reported, in addition to their voices, many more intrusive thoughts than ordinary people. These thoughts were usually interpreted as distressing and uncontrollable.77
A second possibility is that people’s reactions to their hallucinations help to determine which particular thoughts are misattributed in the future. Tony Morrison draws a parallel between hallucinations and intrusive thoughts that is helpful in this context.78 People who are troubled by obsessions often struggle to suppress them, a strategy that can be counterproductive. Just as the injunction not to think of a white bear immediately evokes an image of a snow-coloured furry animal, so too, in ordinary people at least, the effort not to think of ideas that are disturbing appears to guarantee that these ideas recur.79 As Morrison has pointed out, in hallucinating patients, this effect (another example of the reaction-maintenance principle) might well lead to further experiences of the very thoughts that the individual believes are alien.
A third possibility, suggested by G. Lynn Stephens and George Graham, is that thoughts are misattributed elsewhere when they are experienced as ego-dystonic (that is, not consistent with the individual’s current beliefs about the self). As they point out, this might happen for a variety of reasons. For example, taking up a suggestion by the psychologist and philosopher Louis Sass,80 they suggest that the hallucinating person may be unable to construct an integrated representation of his thoughts, emotions and actions, in order to make sense of his hallucinated thoughts. Alternatively, perhaps the hallucinating patient denies that she is the agent of an experience because she cannot explain it in terms of her conception of her own intentional psychology (‘Someone like myself would never have a thought like this’). (The availability of an alternative explanation for an experience – for example, that it is a ghost – presumably makes this more likely.) Interesting though these speculations are, they have yet to be properly tested by researchers.
Even if one or more of these three proposals should turn out to be supported by future research, they probably fail to provide a completely satisfactory account of the hallucinating person’s source-monitoring difficulties. Here are a few questions that remain unanswered: Why do some patients experience their intrusive thoughts as voices outside themselves, some experience them as voices inside their heads, and still others experience them as thoughts that have been inserted into their minds by other people? (Possibly because, once source-monitoring failures occur, different patients form different theories in order to explain their experiences.) Why do hallucinations typically begin during periods of stress or trauma? (Possibly because these periods tend to provoke a flood of intrusive, automatic thoughts about the self.) More importantly, perhaps, why is it that some people seem to be more vulnerable to source-monitoring failures than the rest of us? Could it be that they never learn to monitor efficiently in the first place, but manage to function adequately until their source-monitoring ability is further compromised by some kind of trauma? (I will discuss this idea further in Chapter 18.) Or could it be that people who later hallucinate have some kind of subtle dysfunction of the neural circuits that are responsible for source monitoring? (Henry Szechtman’s findings suggest that these circuits are located in the right anterior cingulate, but do not tell us whether they are damaged in hallucinating people.)
The search for answers to these questions will probably keep psychologists busy for a few years yet.
15
The Language of Madness
Language is the light of the mind.
John Stuart Mill1
In this review of what is known about psychotic complaints I have held back until the end discussion of the symptom that has been subjected to the most extensive psychological analysis. I have done so because much of the relevant research has been misdirected by false assumptions made by the founders of modern psychiatry. We will see that even the term commonly used to describe this symptom – thought disorder– is fraught with difficulty.
Misunderstanding the Problem?
I have presented examples of thought disorder in earlier chapters, when discussing the work of Eugen Bleuler in Chapter 2 and when debating the boundaries of madness in Chapter 5. On the surface, at least, the speech of the thought-disordered patient seems jumbled and incoherent – in extreme cases it is sometimes described as a ‘word salad’. When encountering thought-disordered patients for the first time, the neophyte clinician, like the layperson, might be forgiven for assuming that here, at last, is a symptom that unequivocally meets Karl Jaspers’ criterion of ununderstandability. This was certainly how Kraepelin regarded the speech of mad people, as revealed in the following case presentation to medical students:
The patient I will show you today has almost to be carried into the room, as he walks in a straddling fashion on the outside of his feet. On coming in, he throws off his slippers, sings a hymn loudly, and then cries twice (in English), ‘My father, my real father!’ He is eighteen years old, and a pupil of the Oberrealschule [higher-grade school], tall, and rather strongly built, but with a pale complexion, on which there is very often a transient flush. The patient sits with his eyes shut, and pays no attention to his surroundings. He does not look up even when he is spoken to, but he answers beginning in a low voice, and gradually screaming louder and louder. When asked where he is, he says, ‘You want to know that too? I tell you who is being measured and is measured and shall be measured. I know all that, and could tell you, but I do not want to.’ When asked his name, he screams, ‘What is your name? What does he shut? He shuts his eyes. What does he hear? He does not understand; he
understands not. How? Who? Where? When? What does he mean? When I tell him to look he does not look properly. You there, just look! What is it? What is the matter? Attend: he attends not. I say, what is it, then? Why do you give me no answer? Are you getting impudent again? How can you be so impudent? I’m coming! I’ll show you! You don’t whore for me. You mustn’t be smart either; you’re an impudent, lousy fellow, such an impudent lousy fellow I’ve never met with. Is he beginning again? You understand nothing at all, nothing at all; nothing at all does he understand. If you follow now, he won’t follow, will not follow. Are you getting still more impudent? Are you getting impudent still more? How they attend, they do attend,’ and so on. In the end, he scolds in quite inarticulate sounds…2
Quoting this encounter more than half a century after it was published, the British anti-psychiatrist R. D. Laing drew attention to Kraepelin’s own understanding of the exchange. According to Kraepelin:
Although he understood all the questions, he has not given us a single useful piece of useful information. His talk was… only a series of disconnected sentences having no relation whatever to the general situation.3
Research into thought disorder following Kraepelin has been dominated by a set of assumptions that was made explicit by Eugen Bleuler. Recall that, according to Bleuler, loosening of the associations is one of the most important features of the disorder. Elaborating on this hypothesis, he argued that the speech of psychotic patients is difficult to understand because ‘Fragments of ideas are connected in an illogical way to create a new idea’, and ‘New ideas crop up which neither the patient nor the observer can bring into any connection with the previous stream of thought.’ Bleuler described the results of simple psychological experiments in support of this analysis. For example:
In experimental investigations of associations, we find a notable frequency of ‘mediate associations’. I suspect that only the lack of sufficient observation has been responsible for our inability to demonstrate them more frequently in the thought-processes of our patients. The above mentioned example [a patient had associated the death of a relative to the word ‘wood’], the association ‘wood (wood-coffin) – dead cousin’ may be considered as a mediate association.4
Although a careful reading of Bleuler reveals that he did not believe speech to be merely an overt manifestation of thought, his account suggested to many later investigators that thought disorder is literally a disorder of thinking. Because thinking was seen as a proper object of psychological inquiry by even the most hard-nosed biological psychiatrist, the thinking of psychotic patients therefore began to receive special attention from psychologists. The consequence was a plethora of studies and theories published from the 1930s onwards. Some researchers suggested, for example, that schizophrenia patients fail to follow the rules of logic.5 Others, such as the Russian developmental psychologist Lev Vygotsky6 and the American neurologist Kurt Goldstein,7 argued that schizophrenia patients were incapable of dealing with abstract ideas.
A particularly influential theory was developed by the American researcher Norman Cameron in the 1940s. Cameron, who was trained both in psychology and psychiatry, observed patients as they attempted psychological tests, and proposed ways of categorizing their unusual responses. For example, some responses could be classified under the term asyndesis, indicating the incorrect connecting of concepts that lacked genuine causal links (as in a patient attempting to explain why the wind blows, with ‘Quickness, blood, heat of deer, length; driven power, motorized cylinder, strength’). Others, in which a term is replaced by an approximately related term, were classified as metonymic distortion (as in a patient saying that he was alive ‘because you have menu three times a day; that’s the physical’).8 Expanding on these ideas, Cameron eventually grouped all of these deviant responses under the single heading of over-inclusion, which he believed was a consequence of failing to limit attention to relevant stimuli.9
When experiments were conducted to test these early ideas, the results were often inconclusive, or could be interpreted in many different ways. In retrospect, it is now obvious that the research conducted during this period was constrained by two major limitations. First, as University of Wisconsin psychologists Loren and Jean Chapman pointed out in a seminal review published in 1973, most investigators had failed to establish that the patients they were studying were actually thought-disordered.10 Because ‘schizophrenia’ was regarded as a lump, it was often assumed that any abnormality in thinking should be observable in all schizophrenic patients, regardless of their actual symptoms.
Second, as the linguists Sherry Rochester and J. R. Martin pointed out in their book Crazy Talk, published in 1979,11 Blueler’s identification of thought disorder with thinking had led researchers to neglect what psychotic patients were actually saying. Flying in the face of this trend, Rochester, of the Clark Institute in Toronto, and Martin, at the University of Sydney in Australia, argued that, as the only evidence of thought disorder is peculiar speech, speech and not thinking should be the focus of the psychopathologist’s inquiries (see Figure 15.1). Furthermore, they pointed out that it is the incomprehensibility of the psychotic person’s speech that leads to the diagnosis of thought disorder. They therefore suggested that the question, ‘What is abnormal about psychotic thinking?’ should be replaced with the more useful question, ‘Why do ordinary listeners find psychotic speech so difficult to understand?’
Will the real symptoms of schizophrenia please stand up?
Rochester and Martin’s approach to thought disorder amounted to something of a revolution in schizophrenia research. At the same time, other investigators were questioning other assumptions about the speech of psychotic patients. Nancy Andreasen, a prominent neo-Kraepelinian who had obtained a Ph.D. in literature before training in medicine, realized that research was only likely to progress if adequate criteria could be devised for determining when speech is psychotic. Like Rochester and Martin, she noted the anomalous status of the term ‘thought disorder’, suggested that it might be abandoned and, after careful consideration, proposed that it be replaced by the more precise but longer term, disorder of thinking, language and communication.12
Figure 15.1 Cartoon (adapted from S. Rochester and J. R. Martin (1979). Crazy Talk: A Study of the Discourse of Psychotic Speakers, New York: Plenum) illustrating the problem of defining thought disorder.
Andreasen developed her Scale for the Assessment of Thought, Language and Communication (TLC Scale) in the hope that it would enable investigators to measure psychotic speech more precisely.13 (In Chapter 5, p. 102 I described the use of this scale in a study of eccentrics.) On the basis of classic and modern accounts of thought disorder, and from observations of patients, she defined twenty different kinds of disordered speech, which are listed in Table 15.1. The scale provides fairly precise definitions, allowing each type of abnormality to be rated on a five-point scale. High inter-rater reliabilities can be obtained for these scores when raters are appropriately trained.
Factor analysis revealed that psychotic speech abnormalities fell into two main types. Items on the TLC Scale that reflected various kinds of speech incoherence fell into a single factor, which Andreasen labelled positive thought, language and communication disorder. However, some items, particularly poverty of speech (saying very little) and poverty of content of speech (speech with only vague content), fell into a second factor, which she labelled negative thought, language and communication disorder. (This distinction clearly maps on to the more general distinction between positive and negative symptoms of psychosis, which we considered in Chapter 4.)
When Andreasen used her scale to compare patients with different diagnoses, she found that some types of speech described in the classic literature of psychiatry – for example clanging, echolalia and neologisms – were so uncommon that they appeared to have very little practical significance. Others, for example tangentiality, derailment, incoherence, illogicality, loss of goal and perseveration, appeared to be equally co
mmon in schizophrenia and mania.14 To Andreasen’s surprise, the manic patients showed more evidence of positive thought, language and communication disorder than the schizophrenia patients, whereas the reverse was true of negative thought, language and communication disorder.15 Positive thought, language and communication disorder, it seems, is not a specific complaint of schizophrenia patients.
Table 15.1 Main types of thought, language and communication disorder described by N. C. Andreasen (1979) ‘The clinical assessment of thought, language and communication disorders’, Archives of General Psychiatry, 36: 1315–21.
Type of disorder
Definition
Example
Poverty of speech
Restriction in the amount of spontaneous speech. Replies to questions are brief and concrete.
Madness Explained Page 44
