Stuart Kelly
Department of Language and Linguistics
University of Essex
October 1984
Revised April 2004






The medical model

The anti-psychiatric model

The psychedelic model



Over-inclusive thinking

Distinguishing thought from language



Its significance to psychiatry

Schizophrenic words

Schizophrenic discourse

Cohesion in schizophrenic language

Language perception skills in schizophrenia

Attention and distractibility

Perception of schizophrenic language

Poetry and schizophrenic language

Is there a schizophrenic language?

Information processing

Making sense of schizophrenic utterances






References and works consulted







Since the very earliest attempts to classify the various forms of insanity, the novel and often bizarre way in which some schizophrenics use language has been singled out as one of the most striking features of schizophrenia, differentiating the phenomenon from other psychotic states and from non-psychotic persons.

It is all too easy for a normal listener, heeding the clinical descriptions of psychiatric textbooks, to disregard such utterances as merely bizarre and incomprehensible verbal manifestations of the incoherence and loose associations supposedly characteristic of ‘schizophrenic thought’. Nevertheless, it seems to me that, although ‘schizophrenic language’ may be difficult to follow, and on first hearing appear to be devoid of meaning, if a schizophrenic takes the trouble to communicate, surely courtesy alone dictates that, the listener should make some attempt to understand.

Leech (1974) asserts that,

‘…the human mind abhors a vacuum of sense.’ (p.8)

When confronted by a schizophrenic, one has two alternatives. Either to disregard his/her utterance as nonsense or try and fill the vacuum.

I would suggest that, in some cases at least, It may be not only the transmitter, but also the receiver that is faulty. How many times do we admit to being ‘on a different wavelength’ from another person whom we find difficult to understand, never considering for one moment that, what we hear is nonsense?

Why, then, can we not afford the same benefit of the doubt to the utterances of a schizophrenic?

‘Schizophrenic language’ has a significant role to play in the diagnosis of thought disorder in schizophrenia. Its importance as a diagnostic tool is enhanced in the absence of any other major clinical signs, and very often, the linguistic performance of a patient, during an interview with a psychiatrist is enough to tip the diagnostic balance in favour of schizophrenia.

Yet, there are many examples of widely varying accounts of the nature of schizophrenia. Such diversity of opinion abounds throughout the many nations and cultures of the world that, even individual psychiatrists are often unable to agree and may make different diagnoses in the same patient. So difficult and unreliable is the diagnosis of schizophrenia that, even non-psychotics have been thus labelled and may continue to be so.

Yet, confident in their diagnostic role, psychiatrists continue to identify schizophrenia with alarming regularity. How can we be sure that, such misjudgement does not apply to the equally subjective assessment of incoherence in the language of schizophrenics?

What, if anything, is different about ‘schizophrenic language?

How does it differ from normal language?

How is it to be identified as deviant?

Above all, is there a schizophrenic language?

These and other questions have been addressed by several psychiatrists, psychologists and linguists over many decades. Studies of ‘schizophrenic language’ have taken many forms and produced a large amount of experimental and observational data. Very often, the results and conclusions reached from them have been inconsistent and even contradictory. After more than half a century of research, the quest for the essential definitive characteristics of ‘schizophrenic language’ continues.

(back to top)


The medical model

The first comprehensive and detailed description of schizophrenia (dementia praecox) was made by Kraepelin (1919) who considered it to be a form of insanity (dementia) with a premature (praecox) onset. The term schizophrenia was coined by Bleuler (1950), who conceived of the disease as a group of disorders, which he called the schizophrenias.

Although the schizophrenias comprised several signs and symptoms, those which were common to all and, in Bleuler’s view, characteristic clinical signs of schizophrenia, included changes in emotional state, a tendency to withdraw from the real world into a state of fantasy and a disorganisation of logical thought. An important characteristic feature was said to be a disorder in the association of ideas.

Although both Kraepelin’s and Bleuler’s accounts of schizophrenia represented a significant step forward in the identification and classification of mental disorders, there was a considerable amount of vagueness, which allowed a variety of interpretations to be made, and a strong possibility of misdiagnosis.

In an attempt to improve the diagnostic criteria for schizophrenia, Schneider (1959) proposed a number of ‘first rank symptoms’. Any one of which would be sufficient for a diagnosis to be made, provided that no organic disease could be shown to be present.

These first rank symptoms were:-

Auditory Hallucinations

The patient may hear voices which are not his/her own and which talk about him/her in the third person.

Primary Delusions

False beliefs or conclusions, which are maintained despite overwhelming evidence to the contrary and, in contrast to secondary delusions, do not arise from any hallucinatory source.

Experience of external control over one’s thoughts

For example, the patient may believe that his/her thoughts are not his/her own, and that they have been inserted by some other person or force (thought insertion). Alternatively, the belief that thoughts have been removed (thought withdrawal). Thirdly, the patient may believe that others aware of what he/she is thinking as though they are able to read his/her thoughts (thought broadcasting).

Although Schneider was successful in narrowing the descriptive criteria for schizophrenia, he only proposed that these symptoms were a sufficient condition for a diagnosis to be made or confirmed. He did not imply that all, or any of them, should necessarily be present for schizophrenia to be diagnosed. Even without the presentation of any of these symptoms, a patient may still be considered to be schizophrenic. The psychiatrist is at liberty to select his own criteria and make a diagnosis accordingly. Since psychiatrists, themselves, are not always in agreement, about what does or does not constitute schizophrenia, there must necessarily be some disagreement about who is or is not schizophrenic.

Katz, Cole and Lowry (1969) demonstrated that, even with a common rating scale, a group of psychiatrists were not always able to agree about the extent to which certain symptoms were presented by a particular patient. In addition, when the assessments of British and American psychiatrists were compared, it was observed that, the British perceived fewer features as abnormal than did their American colleagues. (note 1). Passsamanick, Dinitz and Lefton (1959) also showed that, even within the same hospital, a, verdict may vary from one psychiatrist to another.

Further revealing yet disturbing results were obtained in an experiment by Rosenhahn (1973). Comprising several normal volunteers, Rosenhahn’s group had themselves admitted into mental hospitals, complaining only that they had been hearing noises. Although they had been instructed to carry on their daily lives as normally as possible within the constraints imposed by a mental institution, all but one were diagnosed as schizophrenic. The periods of hospitalisation lasted from seven to fifty-two days and, on discharge, all were said to be ‘in remission’.

The experiences of his volunteer group were enough to convince Rosenhahn that,

‘… it is clear, then, that we cannot distinguish the sane from the insane.’ (p.257).

It would seem, then, that the label, ‘schizophrenic’ can be attached to anybody!

Bickford (1973), a psychiatrist with many years experience in the diagnosis and treatment of schizophrenia, acknowledges the enormous difficulties to be encountered when attempting to define the phenomenon.

‘…Delineating schizophrenia is as difficult as depicting the shadow of a fluttering butterfly. It cannot be touched or

examined microscopically and no radiological or chemical test confirms its presence. (p. 794).

Another psychiatrist, Laing (1979), admits that,

‘… I have difficulty in actually discovering the signs and symptoms of psychosis in persons I am myself interviewing.’ (p.28).

Of the problems of diagnosis in schizophrenia, Bickford has this to say,

‘... few complaints are diagnosed with more frequency and less accuracy and it tends to be an expression of the psychiatrist’s attitude rather than an assessment of signs and symptoms.’ (p.794).

Quite clearly, even for experienced psychiatrists themselves, the phenomenon of schizophrenia is not well defined and its identification is, at best, hazardous. However, one thing is, in Bickford’s opinion, quite certain.

‘… if a schizophrenic patient recovers, then the diagnosis was wrong.’ (p.795)

The medical model, for all its shortcomings is, perhaps, the most widely accepted account of the nature of schizophrenia. However, a number of others have been proposed. Often the proponents of these alternative models are, themselves, trained within the traditional psychiatric framework and continue to be practising psychiatrists. Of the many models of mental illness, (Siegler and Osmond, 1974), two seem to me to be worthy of mention. Both views have become renowned throughout the medical fraternity and have met with disapproval and even hostility (note 2) from those working from the established traditions of the medical model.

(back to top)

The anti-psychiatric model

Probably one of the best-known opponents of the medical model is the American (anti-) psychiatrist, T.S.Szasz. Although he is reputed to deny the existence of mental illness, Szasz does acknowledge that such phenomena exist, though he would repudiate the concept of madness as a disease to be treated by medical means. According to Szasz’s definition,

What health is can be stated in anatomical and physiological terms’ (Szasz, 1960, p.114)

In the medical model of schizophrenia, this is obviously not the case since anatomical and physiological characteristics are among the few that seem not to be considered appropriate,

The anti-psychiatric model portrays the schizophrenic (and other mental patients) as the victim(s) of persecution by the medical fraternity on behalf of the rest of society. The schizophrenic is considered abnormal because s/he does not conform to the social, cultural or political norms of the society to which s/he belongs, and having been given the label mad, deviant, dissident or schizophrenic, s/he will act out that role.

Szasz (1973) asserts that,

‘ … the concept of mental illness is analogous to that of witchcraft.’ (p19).

In such an analogy, it is not difficult to see how the inquisition, torture and burning at the stake have merely been replaced by medicine, electro-convulsive ‘therapy’ and psychosurgery.

Although the anti-psychiatric model may be open to criticism from its opponents as alarmist, reactionary or even heretical, (see note 2), it can claim some support from recent and current practices in contemporary psychiatry.

The incarceration of dissidents in the former Soviet Union, so often deplored by the western media was one particularly good example.

As research evidence has clearly shown, (Katz et al 1969, Passamanick et al 1959), the diagnostic criteria for schizophrenia are open to wide interpretation and in spite of the fact that, to western observers, such people were clearly not psychotic, if soviet psychiatrists determined that, a particular deviation from the cultural or political norm is indicative of psychosis, then even the most conscientious among them must make a diagnosis accordingly and treat the ‘disease’ in the appropriate medical manner.

In England and Wales also, the Mental Health acts of 1959 and 1983 both provided for compulsory detention in hospital for an indefinite period of time for certain persons on the instruction of a court of law!

The Mental Health Acts, while outwardly ensuring detention on grounds of ill health, nevertheless stipulate that the authority to discharge a person detained under these circumstances, lies not with the consultant in charge of treatment, but with a medically unqualified politician – the secretary of state!

It is not long since the disease of ‘moral insanity’ could be diagnosed in young women whose ‘symptoms’ included having contravened the cultural norm by becoming pregnant whilst unmarried.

The claim of the anti-psychiatric model that, psychiatry is the instrument by which society regulates the behaviour of non-conforming individuals is not, it seems to me, without some foundation.

(back to top)

The Psychedelic model

A second alternative point of view of the nature of schizophrenia is taken by a British psychiatrist, R.D.Laing. Like Szasz, he too recognises that schizophrenia exists.

‘…it is clear, at least, that some people come to behave and experience themselves and others in ways that are strange and incomprehensible to most people including themselves.’

(Laing, 1979, p.86).

However, he too, refuses to recognise it as a disease,

‘In using the term schizophrenia, I am not referring to any condition that I suppose to be mental rather than physical, or to an illness like pneumonia, but to a label that some people pin on others under certain social conditions.’ (p.86).

Laing (1977) views the schizophrenic experience as a journey through an inner world. The experience of the schizophrenic is not only different from the experiences of others, but such an experience is inaccessible to the non-schizophrenic and has more in common with the transcendental experience of the oriental mystics in that it,

‘… goes beyond the horizons of our common, that is our communal sense.’ (p.109).

In Laing’s view, the schizophrenic is as an explorer embarked upon an enlightening and mind-expanding trip. Laing even goes so far as to predict that, if the human race survives

 ‘… future men will, I suspect, look back on our enlightened epoch as a veritable age of darkness … they will see that what we call "schizophrenia" was one of the ways in which, often through quite ordinary people, the light began to break through the cracks in our all-too-closed minds (op.cit. p.107)

There would seem to be a connection here between the two models advocated by both Laing and Szasz. Will future men look back in horror and disbelief as we now look back on the witch hunts, the trials and the persecution of the mediaeval religious establishment?

(back to top)


Over-inclusive thinking

according to the medical model, disorders of thought in schizophrenia may manifest themselves in forms other than delusions and hallucinations. Schizophrenic thought has been said to be under-inclusive and the patient may be unable to incorporate related ideas into a single concept. More often, however, the thinking of schizophrenics has been described as over-inclusive (Payne 1962; Epstein 1953) when the patient has exhibited a tendency to associate several unrelated ideas and incorporate them all into one single concept.

Epstein (1953) compared a group of schizophrenics with a control group of normal subjects. He devised a test where both groups were required to underline those words in a given list, which they considered to be related to a stimulus word. For example;

MAN____ arms, shoes, hat, toes, head, none

One over-inclusion was scored for each inappropriate word selected and for every omitted appropriate word, one under-inclusion was scored. Results showed that, both groups scored equally on under-inclusions but the schizophrenics scored almost twice as many over-inclusions as under-inclusions. The normal group, however, scored a near equal amount of over- and under-inclusions. Epstein found, however, that no clear division existed between the performance of some schizophrenics and most normals. Instead there was a level at which the over-inclusion scores of normals and schizophrenics overlapped.

Epstein’s experiment may attract some criticism, though. Although he claimed to be testing for peculiarities in thinking, it is far from certain to what extent he can expect to make inferences about the thought processes of his subjects by interpreting the results of a test of verbal skills.

Payne (1962), in his object classification test, dispensed with the word association task. Instead, he required his subjects to group together a number of shapes varying in size, colour and material. Chronic schizophrenics and acute schizophrenics were grouped separately and compared with groups of neurotics, endogenous depressives and normals.

Only schizophrenics produced significantly fewer appropriate responses than normals (neurotics actually scored higher than normals) and only acute schizophrenics scored significantly more inappropriate responses. Nevertheless, only about half of these were found to be more over-inclusive than normal subjects. As Epstein discovered, no clear division could be demonstrated between some schizophrenics and most normals.

Andreason and Powers (1974) were unable to replicate Payne’s findings for over-inclusiveness in schizophrenics. Although they followed Payne’s methods quite closely, they found that, the performance of their schizophrenics was similar to that of the normals in the Payne experiment. Manics, however, were significantly more over-inclusive and performed much more like Payne’s schizophrenics.

According to Andreason and Powers, a possible explanation for this discrepancy could be that, those schizophrenics selected for the Payne experiment were probably diagnosed according to different criteria and may also have included subjects who, besides being schizophrenic, may also have been suffering from affective disorders such as depression or even mania.

It is also possible, given the wide variation in diagnostic criteria, that either study may have included a number of subjects in whom schizophrenia had been misdiagnosed.

(back to top)

Distinguishing thought from language

One great difficulty in assessing thought, whether disordered or otherwise, is that it is an internal event (Chapman and Chapman 1973) and can only be inferred from overt responses. As Chapman and Chapman observe, many researchers have confused thought with language. Research into schizophrenic thought has all-too-often concentrated on the way in which schizophrenics use language (e.g. Epstein, 1953). Indeed Cooper (1980) actually defines thought disorder as a technical term for,

‘… a flow of speech with nonsensical associations’ (p31)

Although it is possibly true that, impaired thought/intelligence, as manifested in dementia or severe learning difficulties, may imply aberrant language, to infer from this that disordered language implies disordered thought is not only logically unsound, but also contrary to the evidence from studies of aphasic patients.

Head (1963) for example, describes several aphasic patients whose use of language is clearly problematic. Nevertheless, a number of these patients, though experiencing great difficulties with speech, are still able to play and win games such as chess and draughts which require them to have retained at least some of their previous capacity for logical thought. Of course, it may be argued that, the kind of language difficulties presented by aphasic patients do not resemble the linguistic peculiarities of schizophrenia and that a direct comparison may be invalid.

(back to top)


Its significance to psychiatry

For more than a century, the utterances of schizophrenics have aroused a great interest among psychiatrists and others concerned with the care and treatment of psychotic patients. Kraepelin (1919) and Bleuler (1950) both remarked on the apparent incoherence and incomprehensibility of schizophrenic discourse. Indeed the use of language by schizophrenics is often described as bizarre and can pose great problems for the listener when attempting to discover exactly what the patient is trying to say. Szasz (1972) considers that,

‘…communication by means of conventional language… constitutes one of the central areas of interest for psychiatry.’(p24)

In the absence of other diagnostic signs, the language of schizophrenics is often of great significance to the psychiatrist in the evaluation of thought disorder, which is considered to be a primary feature of schizophrenia. Yet, as Forrest (1973) observes, although any psychiatrist might profess to be able to identify an utterance as schizophrenic, when asked to indicate how s/he is able to make this judgement, s/he may be less confident in his/her reply admitting that it is a ‘feeling one gets’(p286).

Lorenz (1961) also confesses to what she considers to be a paradox,

‘… while we recognise schizophrenic language when we see it, we cannot define it.’ (p603)

If psychiatrists are unable to define schizophrenic language, who can? Is the linguist or psycholinguist in a better position to unravel the mysteries of schizophrenic language?

(back to top)

Schizophrenic words

Although carried out supposedly as a test of the disordered association of ideas or of over-inclusive thinking, a variety of word association tasks were among the earliest attempts to unravel the mystery of schizophrenia by discovering the distinguishing features of schizophrenic language.

Though the study by Kent and Rosanoff (1910) revealed that, schizophrenics and other psychotics had a tendency to give more unusual responses than normals ( clang associations, neologisms and associations to previous stimuli), there were those who showed no significant deviation from the normal associations.

A later study by Sommer, DeWar and Osmond (1960) attempted to demonstrate that schizophrenics associate to associations, such that, if a stimulus S would normally be expected to elicit the response R1, and R1 would normally stimulate the response R2, then it was hypothesised, that the schizophrenics would give R2 as a response to the original stimulus S.

Unfortunately, the results did not show this to be the case, but the finding of Kent and Rosanoff, that schizophrenics give more unusual responses than normals, was upheld. It was also shown that, not only do schizophrenic associations differ from normal associations, but that schizophrenic associations vary between schizophrenic subjects. Schizophrenics, it seems, are also less likely to give the same response on different occasions.

Williams (1966) conducted several experiments to test the verbal performance of Schizophrenics. Using a cloze procedure, she found that, when asked to guess the missing word in a sentence, schizophrenics performed poorly when compared with a control group of normal subjects. She suggested that, schizophrenics associated with a single word rather than to context. She also noted that, when asked to complete a list of words, the schizophrenics made the following three kinds of incorrect responses:-

Associations to an early word in a list.

good: bad cruel: kind slow: INDIFFERENT

Associations to the last word.

spade: dig pen: PENCIL

Associations to the list as a whole.

Gallop: canter trot: HORSES

Williams also discovered that, when asked to complete verbal passages of varying length and contextual constraints, normal subjects, under time pressure, responded similarly to schizophrenics. Both groups tended to associate to the first words in the sequences rather than to contextual constraints.

Silverstein and Harrow (1982) carried out a continuous word association test, requiring subjects to produce up to ten responses to each of twenty familiar and frequently occurring nouns. Significant differences were observed between the schizophrenics and a control group. The degree to which words were related was greater in normals than in schizophrenics. So, too was the commonality of responses. The schizophrenic group also produced more idiosyncratic responses than did the normal control group.

Clark (1970) however, observes that the effects of word associations depend on ‘ the rules the player has followed’. (p272). Even normal players, when allowed to take their time, may react with ‘rich images, memories or exotic verbal associations’ which, according to Clark, give way to ‘idiosyncratic, often revealing, one word responses.

Since these are, allegedly the type of responses typically produced by schizophrenic speakers, perhaps now is the time to examine the rules by which each player plays the game.

The interpretation of words by schizophrenics has also received some attention. It was found (Chapman, Chapman and Miller 1964) that schizophrenics opt for the stronger (unmarked) meaning of a word regardless of its context in a sentence. Subjects were asked which of three alternatives would be the appropriate interpretation of a given sentence.

For example,

When the farmer bought a herd of cattle, he needed a new pen.

A) He needed a new writing implement.

B) He needed a new fenced enclosure.

C) He needed a new pickup truck.

The results of this experiment suggested that, schizophrenics were more likely to choose the inappropriate, though stronger meaning response (A) than was the case with normal subjects.

The reliability of these findings has been placed in doubt, however. It has been pointed out (Schwartz 1982) that, a strong possibility exists for subjects to give an appropriate response by chance. By eliminating the irrelevant alternative (C), subjects have a 50% chance of guessing the appropriate answer. Schwartz cites the experiment of Boland and Chapman (1971) where this seems to be a plausible explanation for the fact that 42% of appropriate and 52% of inappropriate responses were made by schizophrenics.

Studies of prisoners (Rattan and Chapman 1973) and other non-schizophrenics (Naficy and Willerman 1980) have both demonstrated similar responses to those of schizophrenics. It was even found (Neuringer, Kaplan and Goldstein 1974) that schizophrenics also respond with weaker meanings than non-schizophrenics do.

It is likely, as Schwartz suggests, that a lowered intellectual functioning is responsible for the response bias shown in these studies, rather than a diagnosis of schizophrenia.

The proposal that schizophrenics have a tendency to confuse literal and figurative meanings of words (Chapman 1960) failed to be supported by the results of a comparative study of schizophrenics and a group of general medical patients (Eliseo 1963).

Having found no significant difference between the two groups, Eliseo concluded that, the misinterpretation of words was a characteristic of chronic illness in general rather than being specific to schizophrenia.

Given the results of Rattan and Chapman’s study of prisoners, Eliseo’s results may justify a further conclusion that, the misinterpretation of words may be due not only to chronic illness, but also to prolonged institutionalisation. In either case, it is unlikely that such misinterpretations could be attributable to schizophrenia alone.

(back to top)

Schizophrenic discourse

It has been suggested (Miller 1965) that, a study of individual words and associations does not tell us much about language, whether disordered or otherwise. Later work has tended to concentrate on larger units of schizophrenic language.

Although Brown (1972) concluded that, an unusual mode of thought might be present in schizophrenia, he could find no reason to believe that a schizophrenic language exists. In contrast to this, Chaika (1974) analysed the utterances of one schizophrenic whose dialect she describes as uneducated, and showed several features, which she considered to be indicative of schizophrenic language.

These features were:-

1) sporadic disruption of the ability to match semantic features with sound strings comprising actual lexical items in the language.

2) Preoccupation with too many of the semantic features of a word.

3) Inappropriate noting of phonological features.

4) Production of sentences according to phonological and semantic features of previously uttered words.

5) Disruption in the ability to apply rules of syntax and discourse.

6) Failure to self-monitor.

Although Chaika may answer the obvious criticism, that inferences about the language of schizophrenics are unreliable when they arise from a study of one single schizophrenic, (she concedes that, these six features may not be presented by all schizophrenic speakers), she does not seem to consider that some, at least, of these features may be attributable to the ‘southern uneducated dialect’ of her subject (p261).

In opposing Chaika’s view, Fromkin (1975) presented evidence to suggest that, those features cited by Chaika also occur in normal speech. She collected over 6,000 utterances from speakers who were not only normal, but included academics such as university professors. The utterances collected by Fromkin were said to have been produced spontaneously during what she describes as scholarly meetings.

From this enormous sample of data, Fromkin quotes several examples of the confusion of antonyms, which Chaika presented as evidence of feature (1). Neologisms abound in ‘slips of the tongue’, and several errors in normal speech reveal an apparent disruption in the application of linguistic rules. Fromkin also points out that, the failure to self-monitor also occurs in normal discourse. When people make ‘slips of the tongue’, they do not always attempt to correct themselves. She concludes that, if any of Chaika’s features are unique. ‘then they are unique to the class of HUMAN speakers’.

There would seem to be one point on which Chaika and Fromkin are in agreement. Chaika observed that, apart from what she calls gibberish, the phonology of schizophrenic language does not appear to be deviant. Fromkin suggests that, even gibberish, though not resembling any of the words of the language, nevertheless, conforms to the phonological rules of the language, just as the spoonerisms and other errors in the language of normal speakers. (Fromkin 1971).

Although Fromkin’s observations would suggest that normal speech errors are identical to those of schizophrenics, her data do not indicate whether they occur with the same frequency. It could well be the case that, schizophrenics make the same errors as normal speakers, only more often. Other experimental results may also point to features, unobserved by Chaika or Fromkin, which may be characteristic of schizophrenics.

Rochester and Martin (1979) report on their studies of the discourse of schizophrenic speakers. They tested three groups of ten subjects, who were assessed independently by two psychiatrists, and classified as either normal, thought-disordered schizophrenic or non-thought-disordered schizophrenic.

Three-minute samples were taken from those parts of an interview where thought - disorder was said to have occurred. Two other speech situations were a cartoon description and a brief narrative, where subjects were asked to re-tell in their own words what had been read to them by one of the researchers.

It was observed that, schizophrenics tended to use shorter clauses than normals, both in the interview and in task based experiments. Although no difference was noted within clauses, schizophrenics were found to pause longer at clause boundaries (Rochester, Martin and Rupp 1978), and an inverse relationship was seen to exist between the length of the pause and the degree to which lay judges assessed the coherence of the discourse. After a long pause (greater that 5 seconds), the following clause was two to three times more likely to be judged incoherent than after a short pause (less than 5 seconds). Maher (1972) would seem to support this finding when he reports language disturbances at ‘… terminal points in an utterance such as full stops and period points’ (p13).

Andreason (1974a, 1974b) defined 18 characteristics of schizophrenic language, which were considered to be of excellent to acceptable reliability between raters. Using 12 of these variables, 95,2% of schizophrenics and 68,8% of manics could be identified from three groups of manics, depressives and schizophrenics.

It has also been pointed out however (Morice and Ingram 1982), that, although such features as poverty of content may be consistent in extreme cases, they become rather more subjective when assessing speech, which more closely resembles that of normal speakers. They suggest that such variables may be as unreliable as looseness of association which, itself, may be only loosely determined.

Andreason, herself, also admits that, since one of the assessors in her experiment had also helped to collect the speech samples, and was not ‘blind’ when evaluating them, the assessment of some subjects may have been influenced by this prior familiarity.

Using their own criteria, a total of 98 ‘syntactic variables of complexity, integrity and fluency’, Morice and Ingram developed language profiles for a group of schizophrenics. They found that 17 of these 98 variables gave 95% accuracy in differentiating schizophrenics from both manic and normal subjects. Although the linguistic differences appeared to be predominantly syntactic, other factors could have been involved. Drugs and/ or anxiety appeared to have had no significant effect, but matching of subjects was acknowledged to be imperfect, owing to their limited availability. In particular, the social and educational backgrounds were ill matched and may need to be taken into account when considering the results. In addition, though the context was identical for all subjects, it is possible that the perceived context may have varied between subjects.

Having considered all these factors, Morice and Ingram conclude that, observed differences may be attributed to specific effects of schizophrenia, but they do cautiously suggest that, rather than being due to specifically linguistic factors, these differences may bear more relation to a general cognitive or information processing deficit.

(back to top)

Cohesion in schizophrenic language

According to Halliday and Hasan (1976), cohesion is,

‘… the means whereby elements that are structurally unrelated to one another are linked together.’ (p27).

Rochester, Martin and Thurston (1977) studied the cohesive ties in schizophrenic discourse and found significant differences in the way in which two groups of schizophrenics tied their sentences together.

Conjunction, a tie that expresses logical relations between sentences, was used less frequently by thought-disordered schizophrenics than by others. Schizophrenics also made relatively more use of lexical cohesion such as hyponyms and synonyms.

Because conjunction tends to make a stronger bonding between sentences than does lexical cohesion, Rochester and her associates predicted that the discourse of thought-disordered schizophrenics, because of its looser connections, would be more difficult to follow than the discourse of no-thought-disordered schizophrenics. When lay judges were asked to indicate which of the samples were most difficult to follow, their responses supported the prediction.

It was also noted that schizophrenics made more use of phonetic ties such as rhymes, puns and homonyms than did normal subjects. This would appear to be consistent with the clang associations reported in word association tests (Kent and Rosanoff 1910), and to the fourth of Chaika’s characteristic features. (see above).

While hypothesising that the schizophrenic is an adequate user of linguistic rules and uses a lexicon familiar to native speakers of English, Rochester (1978) asserts that the schizophrenic fails in his/her verbal communication by his/her failure to ‘account for the listener’s needs’. (p228).

Besides using fewer cohesive links between clauses, the schizophrenic does not provide the listener with enough information about which parts are new and which are already given. Those schizophrenics classified as thought-disordered (Rochester and Martin 1977, 1979) were found to be less proficient in the ‘art of referring’ (Martin and Rochester 1978) or ‘phoricity’ (Halliday and Hasan 1976) which serves as an instruction to the listener to,

‘Retrieve from elsewhere the information necessary for interpreting the passage in question’ . (p41).

Although no significant difference was noted in the interview situation, ‘situational references’ were more frequently used by thought-disordered schizophrenics than by either normal subjects or non-thought-disordered schizophrenics.

The following are given as examples of (1) a thought disordered schizophrenic speaker and (2) a normal speaker: -

(a) she’s kni

well he’s got yarn in his hand and on his feet

and she’s winding

and imagine winding a ball of wool off of a man who’s in the stocks.

(2) (a) here we have a chap in the pillory or in the stocks with both

his hands and his feet being held by the device.

and we have a woman who is seated by his side doing her


and she has raveled the yarn all about his hands and feet.

And is now raveling up the yarn into a ball getting ready for knitting.

Without additional reference to the situation being described (in this case a cartoon), the account by the thought-disordered schizophrenic (speaker 1) is clearly more difficult to comprehend than that of the normal speaker whose description includes the relevant information required by the listener. The thought-disordered schizophrenic seems to rely on the listener’s prior knowledge of the contextual prerequisites, which other researchers (Bransford and Johnson 1972) have shown to be necessary both for the comprehension and recall of language. Hardly surprising therefore, is the fact that lay judges, presented only with speech samples, found thought-disordered schizophrenics more difficult to comprehend than either of the two other subject groups.

Rochester and her colleagues first hypothesised that problems of communication in schizophrenia might be due to an ‘interpersonal process’. However, after reassessing the results of these and other tests, particularly work on clause boundary pauses, their findings led them to conclude that’

‘… information processing problems rather than interpersonal failure lay at the root of the schizophrenic’s communication difficulties.’ (Rochester 1978, p279).

(back to top)

Language perception skills in schizophrenia

Using material devised by Miller and Isard (1963), Gerver (1967) compared the perception of sentences by three groups of subjects; schizophrenics, other current psychiatric in-patients and normal subjects. Against a background of white noise, each subject heard recorded sets of grammatical meaningful sentences, ungrammatical meaningful sentences and ungrammatical meaningless wordstrings. Schizophrenics were found to be less proficient than other groups in the number of words that hey were able to recall correctly, and were generally less able to distinguish auditory signals from background noise. However, no difference was noted in the extent to which all three groups used syntactic and semantic cues as a guide to the correct perception and recall of sentences. Gerver concluded that, the use rather than the knowledge of linguistic rules might be at fault in schizophrenia.

Rochester, Harris and Seeman (1973) investigated whether schizophrenics are able to break down sentences in the way that normal speakers do. Using ‘clickology’ tests similar to those pioneered by Bever and his associates, (Bever Lackner and Kirk 1969, Bever Lackner and Stolz 1969), Rochester compared the performances of schizophrenics and a control group.

As in the experiments by Bever, subjects heard a sentence in one ear while, in the other ear, an audible click was presented. Subjects were instructed to indicate as far as possible, where in the sentence, the click had occurred. It was discovered that, schizophrenic listeners responded differentially to sentences that were ‘acoustically identical but syntactically distinct.’

This suggests that schizophrenics are aware of the syntactic features of sentences, particularly at clause boundaries. However, when compared with a normal group, schizophrenics were not as accurate at locating the clicks, and when requested to reproduce the sentences they had heard, the schizophrenic group had more difficulty than the normal subjects did.

Clearly, the results of these and other experiments, both by Gerver and by Rochester and her colleagues, would indicate that, whatever the linguistic problems of schizophrenics might be, in perceiving language, at least, syntactic capabilities cannot be shown to be at fault.

(back to top)

Attention and distractibility

Both Rochester and Gerver have suggested that a possible cause of the difficulties encountered by schizophrenics in the perception of language, may be an attentional deficit. Maher (1972) has also proposed that language disturbances in schizophrenia may be a consequence of ‘... an inability to maintain attentional focussing. (p12)

Schneider (1976), in a dichotic listening experiment, instructed his groups of delusional schizophrenics, non-delusional schizophrenics and normal subjects to perform a shadowing task. The passage to be shadowed was heard through one ear while, in the other ear, a distractor message was heard. The topic of the distractor was either physical (from a high school textbook), material about the hospital, at which all psychotic subjects were currently receiving treatment as in-patients, or a topic relating to the delusions of the delusional schizophrenics.

Although schizophrenics were found not to be generally inferior to normals in shadowing with distraction, and the loudness of the distractor had no greater effect on the performance of schizophrenics than on normal subjects, it was noted that the performance of the group of delusional schizophrenics was considerably impaired by the distractors pertaining to their delusions.

Of course, as Schneider admits, this does not indicate that the selective attention of delusional schizophrenics is in any way inferior to that of the other groups. Indeed, it has been shown (Moray 1959) that a personalised distraction is more likely than any other to similarly affect the performance of normal subjects. In Schneider’s opinion, the manner in which schizophrenics direct their attention is more in doubt than their ability to attend.

In a different shadowing task (Pogue-Guile and Oltmanns 1980), subjects were instructed to shadow a passage in one ear while, in the other ear, an irrelevant passage was to be ignored. At the end of the task, subjects were then asked to answer five questions about the content of the relevant passage.

As in Schneider’s experiment, no groups differed significantly in their shadowing accuracy. Distraction, while slightly impairing the performance of schizophrenics, did not significantly affect the abilities of any of the four groups in the experiment.

However, of all the errors committed in the shadowing task, those which were described as semantically irrelevant, were more frequently made by schizophrenics than by any other group. This was found to be the case regardless of the presence or absence of the distractor passage. Schizophrenics were also less accurate than other groups in their ability to correctly recall the content of the shadowed passage and, with the introduction of the distractor, this deficit was even more marked.

Since the distractor element of this experiment seemed only to have any discernible effect on content recall, rather than on shadowing accuracy, Pogue-Guile and Oltmanns considered that difficulties experienced by schizophrenics, in the perception of language, may be due not solely to problems of selective attention, but may be related to a more general deficit in information processing.

(back to top)

Perception of schizophrenic language

Several writers have made reference to the often-cited incomprehensibility of schizophrenic language (Gerver 1967, Payne 1968). Believing that the key to this incomprehensibility might be found at a level higher than the sentence, Rutter (1979) tested this notion by a technique of reconstructing passages of schizophrenic and normal speech, sentence by sentence. Rutter’s subjects for this experiment were all students and presumed to be normal. Each was presented with sentences on separate cards, all of which, with the exception of the first sentence, were in random order. The task simply required the subjects to try and reconstruct the original passages by fitting the sentences together in the correct order.

Results showed that those students attempting to reconstruct schizophrenic passages were less successful, in stringing together three or more sentences, than those students whose passages were from normal speakers. With strings of two sentences, however, Rutter found that no group differed significantly from the other.

It is difficult to decide what conclusions to reach from this. Although partly supporting the observations that schizophrenics make less use of cohesive ties than normal speakers (Rochester and Martin 1979), and that they are less proficient in the art of referring (Martin and Rochester 1978), Rutter was only able to demonstrate such a lack in strings of three or more sentences.

When 42 psychiatrists were requested to identify schizophrenic thought-disorder in passages from two manics, two schizophrenics and two authors (Andreason, Tsuang and Canter 1974), none of the schizophrenic passages were judged to show signs of thought-disorder, but those from the works of writers frequently were.

Again what is one to conclude from this amazing result?

42 incompetent psychiatrists? Two thought-disordered authors?

More likely, I would suggest, is the fact that in the absence of non-linguistic, clinical observations, even the psychiatrist is unable to identify the schizophrenic speaker.

Honigfeld (1963) attempted to test the notion that ‘it takes one to know one’. It was hypothesised that the language of schizophrenics would be more easily understood by other schizophrenics than by normals. Honigfeld took speech samples from a schizophrenic, a volunteer under the influence of the hallucinatory drug, psilocybin, and an excerpt from a newspaper article. Using the cloze procedure, the comparative abilities of college students and schizophrenics to understand each sample were assessed, but no evidence was found to support the hypothesis.

Undaunted, Honigfeld did not reject the hypothesis, but asserted that the lack of supportive evidence may have been due to several factors. In particular, the speech samples used in his experiment contained no neologisms and showed little deviation from normal usage. The schizophrenics, he claimed, may also have been atypical, since they were all patients in a veterans hospital and tended, perhaps, to be ‘more chronic’ than patients in other institutions.

However, Honigfeld’s explaining away of his results must be regarded with some suspicion. Can he really justify his failure by accusing his subjects (and sources of data) of breaking the rules?

It would appear that, despite its reputed unintelligibility, the language of schizophrenics might not always be recognised as such when assessed in isolation from other clinical signs. Furthermore, any features, which may be alleged to differentiate schizophrenic language from any other, do not seem to be recognisable even by schizophrenics, themselves.

(back to top)

Poetry and schizophrenic language

Andreason, Tsuang and canter (1974) have shown that, even to a group of psychiatrists, the task of distinguishing the psychotic from the creative writer may not be accomplished on the basis of language samples alone. In a comparison of schizophrenic language and the language of poets, Forrest (1976) points out that one poet, at least, has made,

‘… comprehensive use of every aberration of language described by Bleuler… in the language of schizophrenics.’ (p294-5)

In likening the language of schizophrenics to that of Shakespeare, Forrest asserts that the schizophrenic may be,

‘… intending more of what he is doing than we give him credit for.’ (p297)

He also remarks on the tendency of some schizophrenics to give new meanings to existing words, and also to create new words (neologisms). Although considered to be abnormal when it occurs in the discourse of a schizophrenic, it is often put to good use in poetry and other creative writings. This is what Forrest considers to be a normal process of,

‘stretching language to serve new uses.’ (p294)

As examples of this stretching of language, Forrest cites Lewis Carroll’s word ‘slithy’, which he created by blending the two words slimy and lithe. Shakespeare too, found new uses for words, as in the following examples quoted by Forrest: -

Cudgelling one’s brain.

Beggaring all description.

Reed (1970) also examined similarities between the language of schizophrenics and the poetic works of Shakespeare, Dylan Thomas, and E.E. Cummings. He found that, not only do poets employ similarly unusual associations and condensations, but deviations in typography and syntactic structure were also not unlike those to be found in the writings of some schizophrenics. He also points out that some poets have remarked on their lack of control over the poetic ideas, which seem to spring spontaneously to mind and have ‘no connection with the consciousness or the will.’ (p422)

This seems to bear an uncanny resemblance to the experience of thought-insertion, which, as one of Schneider’s first rank symptoms, would be sufficient to diagnose schizophrenia.

Keefe and Magaro (1980) compared the performance of schizophrenics, other psychotics, and normal subjects using the Graded Alternate Uses Test, which is alleged to be a test of creativity. They found that, in one measure, at least, schizophrenics scored higher than either of the other two groups. In a different test, however, the higher score of the schizophrenics was thought to be related to age rather than to clinical diagnosis.

Although caution dictated that the test could not be cited as evidence to support a hypothesis that creative people are psychotic, from the results of their experiment, Keefe and Magaro concluded that,

‘… the possibility exists that one can study a form of thought that may apply to creatives and schizophrenics.’ (p397)

So echoing an earlier conclusion reached by Hasenfus and Magaro (1976) who, after reviewing the evidence, asserted that,

‘… the schizophrenic ‘deficit’ does in some instances equal creativity.’ (p348)

In a survey involving 56 professional writers, who were tested using the Minnesota Multiphasic Personality Inventory (MMPI), Barron (1972) found that, of the 78 items on the schizophrenia scale, only 18 helped to distinguish the writers from a group of patients. Intensive interviews with the creative writers revealed that, about an equal number of them and of schizophrenics reported similar personal experiences including a preference for solitude (withdrawal?) and a rejection of many common values of society (deviance?).

Further similarities have been found in the attentional strategies both of schizophrenics and of creative people (Dykes and McGhee 1976). However, it was concluded that, whereas the widening of attention in schizophrenics is involuntary and impairs their performance, the creative person is able to,

‘… cope with this above average influx of stimuli … without risk of cognitive overload.’ (p54)

Yet, even this would appear to be at variance with the feelings of poets (Reed 1970) who, though possibly regarding the experience as pleasant, nevertheless feel that their lack of control can be,

‘… agitating, exhausting, and at times embarrassing.’ (p422)

Although one would hesitate to suggest that, similarities of language and personal characteristics point to evidence of psychosis in writers and poets, there may be reason to suppose that, the schizophrenic turns to poiesis in order to relate those experiences, which are inaccessible to most people and cannot be expressed in a more conventional form of language.

(back to top)

Is there a schizophrenic language?

As the large amount of experimental and observational data will testify, the search for linguistic phenomena to characterise the language of schizophrenics has provided little, if any, conclusive evidence which would point to a form of language that is peculiar to schizophrenics. Certainly, many interesting results have been obtained to show that particular features are prominent in the discourse of schizophrenic speakers. Yet, with a few notable exceptions (Rochester et al 1977, 1979; Morice and Ingram 1982), many of the phenomena to be encountered in the utterances of schizophrenics are not uncommon in the every-day language of normal speakers (Fromkin 1971,1975).

Tests of perception (Gerver 1967, Rochester, Harris and Seeman 1973) have indicated that schizophrenics are no less able to use syntactic rules than are non-schizophrenics, and even the most bizarre innovations in language can be found with some frequency in poetry (Forrest 1976, Reed 1970).

The perception of schizophrenic speech by non-schizophrenics has been shown to be less problematic than our intuitions would lead us to believe (Andreason, Tsuang and Canter 1974), and from the earliest studies of language in schizophrenia, word association tests have failed to demonstrate any commonality of response such that, a mutually intelligible language is shared by schizophrenics (Sommer, DeWar and Osmond 1960, Silverstein and Harrow 1982). The cloze analysis of Honigfeld (1963) also failed to support the hypothesis that, schizophrenics have the capacity to understand schizophrenic utterances with any greater ease than the non-schizophrenic does.

From the evidence available, it seems doubtful that a form of language exists, which is exclusively schizophrenic. Nevertheless, it is undoubtedly the case that, on some occasions, some schizophrenics may produce utterances which, to the listener, whether normal or schizophrenic, may be extremely difficult to understand. Yet, the many studies have failed to demonstrate that any communication difficulty, which may be experienced, is due solely to a linguistic deficit on the part of the schizophrenic. More likely, it seems, is the possibility that, a more general cognitive disorder is responsible, and current research tends to favour the hypothesis that, an information processing problem may lie at the root of the schizophrenic’s communication difficulties (Rochester 1978, Schwartz 1982).

Information processing

(back to top)

Several authors have come to the conclusion that, the schizophrenic’s language problem may not be purely linguistic. It has been proposed instead, that the unusual features, reported to be prevalent in the discourse of schizophrenics, may be attributable to a more general cognitive deficit in information processing (Rochester 1978, Pogue-Guile and Oltmanns 1980, Morice and Ingram 1982).

In reporting the results of their ‘clickology’ tests, Rochester and her associates suggest that any of three alternatives might account for the apparent lack of success by schizophrenics in reproducing sentences that they have heard.

Firstly, they speculate that, the schizophrenic carries out some additional processing, which normal subjects do not. For example, the studies by Cohen (Cohen and Cambi 1967, Lisman and Cohen 1969) may indicate a difficulty in editing out associations, which could, conceivably, present the schizophrenic with an extra processing load.

Secondly, the possibility exists that, the processing mechanism, itself, may be impaired. Perhaps owing to a deficit in attentional focussing (Maher 1972, Schneider 1976), the schizophrenic may experience some difficulty in processing the same information as normal language users.

Either of these alternatives, a faulty mechanism or a processing overload, would be consistent with the observation that, schizophrenics pause longer at clause boundaries (Rochester and Martin 1979) and the report (Maher 1972) of language disturbances at terminal points in a sentence.

A third alternative was also considered. Observations of the recall ability of schizophrenics (Gerver 1967, Rochester et al 1973) would seem to support the view that, the schizophrenic’s ability to perceive, store and retrieve syntactic information may be at fault. It is also likely that a similar problem exists with semantic information (Pogue-Guile and Oltmanns 1980). Unlike normal listeners, (Sachs 1967), schizophrenics, it seems, retain fewer semantic than syntactic aspects of information.

Reed (1970) proposed that a filter mechanism, which enables the listener to select only that information, which is relevant to the task in hand, may be impaired in the schizophrenic. Accompanied by a slower rate of processing (Williams 1966), the extra information is likely to overload the short-term memory and result in some items becoming lost. Interestingly, this would appear to compound two of the alternatives suggested by Rochester, who speculated that, either a faulty mechanism or a processing overload would be likely causes of the schizophrenic deficit.

Schwartz (1978, 1982) favours the view that, the faulty mechanism of schizophrenia is to be found in the process of pigeon-holing. Besides the selection of only relevant information (filtering), the normal listener is able to select from a number of items, those, which constitute a response set. For example, from a list of words, those that can be subcategorised as colours may be allocated to one set, while shapes may constitute another.

The ability of schizophrenics to accurately shadow a passage in dichotic listening tasks (Schneider 1976, Pogue-Guile and Oltmanns 1980), provide evidence that schizophrenics are able to filter out irrelevant material, but when two messages are heard simultaneously, and through both ears, the results are rather different (Helmsley and Richardson 1980). Schizophrenics, in this case, are less proficient than normals in their shadowing performance.

This difference, Schwartz claims, supports the pigeon-holing hypothesis in preference to filtering as a probable defective mechanism in schizophrenia. He argues that, in the dichotic listening tasks, by allowing attention to be directed to the ear of entry as a physical cue, the mechanism being used adequately by schizophrenics is that of filtering. The study by Helmsley and Richardson, however required that attention be focussed on the meaning of the relevant passage, and as the Pogue-Guile and Oltmanns experiment also indicated, this is one of the areas, which poses far greater perceptual difficulties for schizophrenics than for non-schizophrenics.

Schwartz asserts that attention to meaning involves the pigeon-holing mechanism and concludes that this is impaired in schizophrenia.

Schwartz has not escaped criticism, though. It has been pointed out (Knight 1982) that the two types of task cited by Schwartz may have been ill-matched, and comparisons between the two may be less revealing than they appear. A second criticism (Knight and Sims-knight 1982) questions the reliability of citing as evidence, experiments that were designed to test other hypotheses.

Schwartz, however is aware of these problems and in an earlier work (Schwartz 1978), admitted that further work was required, which clearly separates filtering from pigeon-holing.

Of course there remains a great deal of scope for further research into this particular aspect of schizophrenic language

(back to top)

Making sense of schizophrenic utterances

Although the evidence suggests that schizophrenics are adequate users of linguistic rules, the utterances of some schizophrenics can pose grave problems for the listener, whether normal or otherwise (Rutter 1979, Honigfeld 1963).

In shadowing tasks (Pogue-Guile and Oltmanns 1980), schizophrenics were found to make far more semantic errors than normals, and in the recall of the content of the shadowed passage, schizophrenics were also shown to be less proficient than normal subjects.

Word association tests have shown that more idiosyncratic responses are made by schizophrenics than normals (Kent and Rosanoff 1910, Sommer et al 1960,Silverstein and Harrow 1982) and the over-inclusion reported by some researchers (Epstein 1953, Payne 1962) may suggest an over-extension of semantic fields to include those items, which are only vaguely related. As a result, something resembling a mental thesaurus would enable the schizophrenic to make innovative use of existing words, so stretching the language to serve new uses (Forrest 1976).

It has been hypothesised that, the often experienced communication difficulties of the schizophrenic may be attributed to an information processing deficit, rather than to any specifically linguistic disability. Indeed the pigeon-holing hypothesis may well provide some explanation for the allegedly over-inclusive thinking of the schizophrenic. It is not difficult to see how over-inclusive semantic categories would be consistent with an impaired ability to allocate various items of information to the appropriate response sets.

The failure of the schizophrenic to make enough use of context (Williams 1966, Chapman et al 1964) and to provide contextual information to the listener (Rochester and Martin 1977,1979, Martin and Rochester 1978) would also create difficulties for the listener, attempting to discover the meaning of an utterance.

While being conscious of the risk of making sense out of nonsense, it is possible, nevertheless, by taking account of the context of an utterance, and stretching some lexical items beyond the bounds of their normal usage, to derive something meaningful from a schizophrenic utterance.

Consider the following: -

Poison needle blasting gun.

We’re on the moon now, Stuart.

The system is affecting me here (head) and giving me a pain here (abdomen).

The first of these, though appearing at first, to be a random string or word salad, is easy to interpret. What poet could better describe a syringe?

Statement 2 is clearly false since, both the utterer and the addressee are firmly situated on Earth. The man could be deluded or suffering from hallucinations (he is, after all, schizophrenic). However when asked for further information, his assertion that, the price of cigarettes in the hospital shop has increased yet again, may shed some light on the matter. Though he could quite easily have said that prices have gone up, that they have rocketed, are sky high, or that they are astronomic, instead he created a metaphor of his own. It is interesting to note that, although he appears to have used the poetic ploy of stretching language, his own metaphor is not unrelated to the many, well-worn, existing, idiomatic expressions.

Speaker 3 would appear to be exhibiting one of the first rank symptoms of schizophrenia – external control over his thoughts. But what is the system that affects him and what connection is there between the system’s affects and the man’s abdominal pain?

Roget’s Thesaurus lists system as a synonym of order. A category which also includes progression, series and arrangement. As a synonym of plan, this word is associated with, among others, continuance and invention. This information, by itself, may of course, be less than helpful without some additional clues.

The speaker, in this instance, often spoke of his fear that, the human race might be endangering its very existence by allowing technology (the system?) to progress at a faster pace than humanity itself. Little wonder, then, that the man experienced severe pain – from an ulcer!

Of course, every one of the above interpretations is no more than mere speculation. Perhaps many more, equally plausible explanations exist. Nor can one discount the possibility that, any uncovered meaning is the result of filling the vacuum of sense so much abhorred by the human mind (Leech 1974).

Nevertheless, it may well be the case that, the schizophrenic speaker has an important meaningful message to transmit to any person equipped (or disposed) to receive it.

(back to top)


I have endeavoured to demonstrate that, owing to the diversity of professional opinion, both within and without the medical model, as to precisely what constitutes schizophrenia, the probability of a uniform diagnosis between nations, institutions and even individual psychiatrists, is remote.

For this reason, it seems to me that, serious doubts must be raised concerning the psychiatric status of the subjects of many experiments conducted over several decades and in a variety of locations. It is unlikely that, all schizophrenic patients have been selected according to the same criteria, and many may have had other disorders to complicate the diagnosis and distort the experimental results.

To their credit, Andreason and Powers (1974) have acknowledged that, this may have some bearing on the discrepancies observed between their results and those in a similar experiment by Payne (1962).

There is also the possibility that, some subjects may have been diagnosed as schizophrenic for the purpose of one experiment when in another, by a different researcher, the diagnosis would have been considered inappropriate.

A further factor, which seems to have been overlooked in most instances, has been pointed out by Schwartz (1982). Even if an accurate diagnosis has been made and no affective disorder is shown to be present, the heterogeneous nature of the group of schizophrenias described by Bleuler may well have affected the composition of subject groups in the various experiments.

It is, no doubt, assumed that subjects in experimental situations have been representative of the schizophrenias as a whole, yet few attempts have been made to determine whether or not any particular clinical type of schizophrenia dominates any of the groups. No experiments, to my knowledge, have differentiated the simple schizophrenic from catatonic, hebephrenic etc. Although paranoid and non-paranoid were tested separately (Keefe and Magaro 1980), it has not been satisfactorily demonstrated whether a language disorder, if it exists at all, is peculiar to one type of schizophrenia, more than one, or is a general characteristic of the group as a whole.

The failure to distinguish between the two concepts of thought disorder and language disorder may also have introduced a further complicating factor. Though some experimenters have been aware of the imprudence of equating language with thought (Chapman and Chapman 1973, Rochester and Martin 1979), it is possible that, a large number of researchers have fallen foul of some rather dubious assumptions.

If a diagnosis of thought disorder is made by a psychiatrist on the basis of what he perceives as disordered language, then it should come as no surprise to the linguist to find that, his thought-disordered subjects, as assessed by the psychiatrist, display unusual linguistic characteristics. Clearly, the psychiatrist and the linguist together have successfully demonstrated that, disordered language is observable in persons whose language is disordered. Needless to say, such a conclusion reveals nothing of value and cannot possibly bring us any closer to an understanding of the utterances of schizophrenics.

In view of the contradictory opinions of psychiatrists, and the inconclusiveness of many of the observational and experimental data, it seems to me that, the subjective and often inconsistent evaluations of language disorder in psychiatric patients are at best, an extremely unreliable guide to the diagnosis of schizophrenia.

However, studies of the discourse of schizophrenics may yet have a role to play in psychiatry. The ability to recognise and act upon the anxieties and preoccupations of all psychiatric patients is an essential skill of the nurse and other therapists concerned with the care and intervention in psychotic crises.

Although it would be naïve to disregard the possibility that, meaning might be attributed where none exists, the clues to these preoccupations may well be found underlying the utterances of schizophrenics. An understanding of these utterances, in spite of its limited and doubtful application as a diagnostic tool, may yet prove to be an invaluable adjunct to the therapeutic process.

(back to top)


1) The suggestion that a possible cure for American schizophrenics would be to cross the Atlantic (Clare 1970), although a cynical viewpoint on behalf of British psychiatrists, is nevertheless, not without some support from this observation.

2) Szasz (1972) briefly mentions the consequences of publishing The Myth of Mental Illness, and relates that, within one year of publication, the New York Commissioner for Mental Hygiene demanded that, he (Szasz) be dismissed from his university position because he did not believe in mental illness.

(back to top)

References and works consulted

Andreason, N.C (1979a) Thought, language and communication disorders I Clinical assessment, definition of terms and evaluation of their reliability.Arch. Gen. Psychiatry (36) 1315 – 1320

Andreason, N.C. (1979b)Thought language and communication disorders II Diagnostic significance.Arch. Gen. Psychiatry (36) 1325 - 1330

Andreason, N.J.C. and Powers, P.S. (1974)  Overinclusive thinking in mania and schizophrenia. Brit. Journ. Psychiatry (125) 452 - 6

Andreason, N.J.C., Tsuang, M.T. and Canter, A. (1974) The significance of thought disorder in diagnostic evaluations. Comp. Psychiatry (15, 1) 27 – 34

Barron, F. (1972)The creative personality: akin to madness Psychology Today July 1972 43 – 44, 84 – 85

Bever, T.G., Lackner, J.R. and Kirk, R. (1969) The underlying structures of sentences are the primary units of immediate speech processing.Perception and Psychophysics (5) 225 – 234

Bever, T.G., Lackner, J.R. and Stoltz, W. (1969) Transitional probability is not a general mechanism for the segmentation of speech. Journ. Exp. Psychology (79) 387 – 394

Bickford, J.A.R. (1973) Schizophrenia The Practitioner, Jan. 1973  Reprinted Nursing Times June 21st 1973

Bleuler, E. (1950) Dementia Praecox or the group of schizophrenias (Transl. J. Zifkin)  N.Y. International Universities Press (repr. 1966)

Boland, T.B. and Chapman, L.J. (1971)  Conflicting opinions from Broen’s and Chapman’s theories of schizophrenic thought disorder. Journ. Abn. Psychology (78) 52 – 58

Bransford, J.D. and Johnson, M.K. (1972) Contextual prerequisites for understanding. Some investigations of comprehension and recall. JVLVB, (11) 717 – 726

Brown, R. (1972) Schizophrenic language and reality. Amer. Psychologist 3 -17

Chaika, E. (1974) A linguist looks at schizophrenic language Brain and Lang. (1) 257 – 276

Chapman, L.J. (1960) Confusion of figurative and literal usages of words by schizophrenic and brain damaged patients Journ. Ab. Psychology (60) 402 – 406

Chapman, L.J. and Chapman, J.P. (1973) Disordered thought in Schizophrenia Englewood Cliffs, N.J., Prentice Hall

Chapman, L.J., Chapman, J.P. and Miller, G.A. (1964) A theory of verbal behaviour in Schizophrenia In Maher, B.A., (Ed). Progress in experimental personality research London, Academic Press

Clare, A. (1980) Psychiatry in Dissent (2nd Edition) London, Tavistock

Clark, H.H. (1970) Word association in linguistic theory In Lyons, J. (Ed) New Horizons in Linguistics Harmondsworth, Penguin (repr. 1977)

Cohen, B.D. and Camhi, J. (1967) Schizophrenic performance in a word association task Journ. Abn. Psychology (72) 240 – 246

Committee on the Judiciary, U.S. Senate (1972) Abuse of psychiatry for political repression in the Soviet Union Washington, U.S. Govt.

Cooper, D. (1980) The Language of Madness Penguin, England

Dykes, and McGhie, A. (1976) A comparative study of attentional strategies of schizophrenic and highly creative normal subjects. Brit. Journ. Psychiatry (128) 50 – 56

Eliseo, T.S. (1963) Figurative and literal misinterpretations of words by process and reactive schizophrenics. Psychol. Reports (13) 871 – 877

Epstein, S. (1953) Overinclusive thinking in a schizophrenic and a control group. Journ. Consulting Psychology (17,5) 384 – 388

Forrest, D.V. (1976) Nonsense and sense in schizophrenic language. Schiz. Bulletin (2) 286 – 301

Fromkin, V.A. (1971) The non-anomalous nature of anomalous utterances Reprinted in Fromkin, V.A. (1973) Speech Errors as Linguistic Evidence The Hague, Mouton

Fromkin, V.A. (1975) A linguist looks at ‘A linguist looks at schizophrenic language. Brain and Language (2) 498 – 450

Gerver, D. (1967) Linguistic rules and the perception and recall of speech by schizophrenic patients. Brit. Journ. Soc. And Clin. Psychology 204 – 211

Hasenfus, R.M. and Magaro, P.A. (1976) Creativity and schizophrenia. An equality of empirical constructs. Brit. Journ. Psychiatry (129) 346 – 349

Head, H. (1963) Aphasia and Kindred Disorders of Speech. Vol. 2 N.Y. Hafner

Helmsley, D.R. and Richardson, P.H. (1980) Shadowing by context in schizophrenia. Journ. Nervous and Mental Diseases (168) 141 – 145

HM Government (1960) Acts of Parliament 1959 London, HMSO

Halliday, M.A.K. and Hasan, R. (1976) Cohesion in English London, Longman

Honigfeld, G. (1963) The ability of schizophrenics to understand normal, psychotic and pseudo psychotic speech. Diseases of the Nervous System (24) 692 – 694

Katz, M., Cole, J.O. and Lowry, M.A. (1969) Studies of the diagnostic process: the influence of symptom perception,Past experience and ethnic background on diagnostic decisions. Amer. Journ. Psychiatry (125) 937 – 947

Keefe, J.A. and Magaro, P.A. (1980) Creativity and schizophrenia an equivalence of cognitive processing. Journ. Abn. Psychology (89, 3) 390 – 398

Kent, G.H. and Rosanoff, A.J. (1910) A study of associations in insanity Amer. Journ. Insanity (67) 37 – 96

Knight, P.A. and Sims-Knight, J.E. (1982) Evaluating pigeon-holing as an explanatory construct for schizophrenic cognitive deficiencies. Behavioural and Brain Sciences (5) 601 – 603

Knight, R.G. (1982) Language disorder and hemispheric asymmetries in schizophrenia Behavioural and Brain Sciences (5) 603 - 604

Kraepelin, E (1919) Dementia Praecox and Paraphrenia (Transl. R.M. Barclay) Edinburgh, E and S Livingstone

Laing, R.D. (1977) The Politics of Experience and The Bird of Paradise Harmondsworth,  Penguin

Laing, R.D. (1979) The Divided Self Harmondsworth, Penguin 

Lancashire Area Health Authority (1978) Glossary of terms tests and drugs used in psychiatric practice Lancaster, Lancaster Moor Hospital

Leech, G. (1974) Semantics Harmondsworth, Penguin

Lisman,S.A. and Cohen, B.D. (1969) Self-editing deficits in schizophrenia. A word association analysis. Journ. Abn. Psychology (79) 181 – 188

Lorenz, M. (1961) Problems posed by schizophrenic language. Arch. Gen. Psychiatry (4) 603 – 610

Maher, B.A. (1972) The language of schizophrenia. Brit. Journ. Psychiatry (70) 375 – 380

Martin, J.R. and Rochester, S. (1978) Cohesion and reference in schizophrenic speech. In Makkai, A. and Makkai, V.B. (Eds.) The First LACUS Forum 1974 Columbia, Hornbeam Press

Miller, G.A. (1965) Some preliminaries to psycholinguistics AMER. Psychologist (20) 15 – 20

Miller, G.A. and Isard, S (1963) Some perceptual consequences of linguistic rules. JVLVB (3) 217 – 228

Moray, N. (1959) Attention and dichotic listening. Affective cues and the influence of instructions. Quart. Journ. Exp. Psychology (11) 56 – 60

Morice, R.D. and Ingram,J.C.L. (1982) Language analysis in schizophrenia. Diagnostic implications. Aust. And N.Z. Journ. Psychiatry (16) 11 – 21

Naficy, A. and Willerman, L. (1980) Excessive yielding to normal response biases is not a distinctive sign of schizophrenia. Journ. Abn. Psychology (89) 697 – 703

Neuringer, C., Kaplan, H.A. and Goldstein, G. (1974) Schizophrenic avoidance of strong meaning associations to emotional words. Journ. Genetic Psychology (124) 123 – 129

Passamanick, B. Dinitz, S and Lefton, M. (1959) Psychiatric orientation and its relation to diagnosis and treatment in a mental hospital Amer. Journ. Psychiatry (116) 127 – 132

Pavy, D. (1968) Verbal behaviour in schizophrenia. A review of recent studies. Psychol. Bulletin. (10, 3) 164 – 178

Pogue-Guile, M.F. and Oltmanns, T.F. (1980) Sentence perception and distractibility in schizophrenic, manic and depressed patients. Journ. Abn. Psychology (89, 2) 115 – 124

Rattan, R.B. and Chapman, L.J. (1973) Associative intrusions in schizophrenic verbal behaviour Journ. Abn. Psychology (82) 169 – 173

Reed, J.L. (1970) Schizophrenic thought disorder. A review and hypothesis. Comprehensive Psychiatry (11, 5) 403 – 432

Rochester, S.R. (1978) Are language disorders in acute schizophrenia actually information processing problems? Journ. Psychiatric Res. (14) 275 – 283

Rochester, S.R and Martin, J.R. (1977) The art of referring. The speaker’s use of noun phrases to instruct the listener In Freedle, D.O. (Ed.) Discourse Processes. Advances in Research and Theory Norwood, N.J., ABLEX

Rochester, S.R, Harris, J. and Seeman, M.V. (1973) Sentence processing in schizophrenic listeners. Journ. Abn. Psychology (82, 2) 350 – 356

Rochester, S. and Martin, J.R. (1979)  Crazy Talk. A study of the discourse of schizophrenic speakers. N.Y., Plenum Press

Rochester, S., Martin, J.R. and Thurston, S. (1977) Thought process disorder in schizophrenia. The listener’s task. Brain and Lang. (4) 95 – 114

Rochester, S., Thurston, S. and Rupp, J. (1978) Hesitations as clues in coherence. A study of the thought disordered speaker. In Rosenberg, S. (Ed.) Sentence Production. Developments in Theory and Research. N.J, Elbaum

Rosenhahn, D.I. (1973) On being sane in insane places. Science (179) 250 – 358

Rutter, D.R. (1979) The reconstruction of schizophrenic speech. Brit. Journ. Psychiatry (134) 356 – 359

Sachs, J.S. (1967) Recognition memory for syntactic and semantic aspects of connected discourse. Perception and Psychophysics (2, 9) 437 – 442

Schwartz, S. (1978) Language and cognition in schizophrenia. A review and synthesis In Schwartz, S.(Ed.) Language and Cognition in Schizophrenia  Hillsdale, N.J., Elbaum

Schwartz, S. (1982) Is there a schizophrenic language? Behavioural and Brain Sciences (5) 579 – 626

Schneider, K. (1959) Clinical Psychopathology (Transl. M.W. Hamilton) N.Y., Grune and Stratton

Schneider,S.J. (1976) Selective attention in schizophrenia. Journ. Abn. Psychology (85, 2) 167 – 173

Siegler, M. and Osmond, H. (1974) Models of Madness, Models of Medicine N.Y., Macmillan

Silverstein, M.L. and Harrow, M. (1982) Inter-relationships among three measures of disordered thinking in continuous word association Psychol. Reports (50) 803 – 809

Sommer, R, DeWar, M.S. and Osmond,H. (1960) Is there a schizophrenic language? Arch. Gen. Psychiatry (3) 665 – 673

Szasz, T.S. (1960) The myth of mental illness Amer. Psychologist (15) 113 – 118

Szasz, T.S. (1972) The Myth of Mental Illness St. Albans, Paladin

Szasz, T.S. (1973) The Manufacture of Madness St. Albans, Paladin

Taylor,W.L. (1953) ‘Cloze Procedure’ a new tool for measuring readability Journalism Quarterly (30) 415 – 433

Williams, M. (1966) The effect of context on schizophrenic speech Brit.Journ. Soc. & Clin. Psychol (5) 161 - 171