Invited commentary on … Word use in first-person accounts of schizophrenia

Summary

The use of pronouns and causal attributions in personal accounts has been analysed to distinguish between schizophrenia and mood disorders. The implications for both cognitive processing and the underlying pathology of symptoms are explored. Context is identified as a key variable in the analysis and interpretation of text.

Although idiosyncrasies of communication are often evident in those experiencing severe mental illness, a difficulty arises in how to analyse them in an objective manner and how to interpret findings in relation to underlying pathological processes. Fineberg et al ¹ have argued that characteristic patterns of writing enable a distinction to be drawn between schizophrenia and mood disorder, and that these differences provide clues about styles of thinking and the evaluation of evidence. They have identified a bias towards third-person, rather than first-person, pronouns in the personal accounts of individuals diagnosed with schizophrenia. External attributions (references to religion and to other humans) are more common in those diagnosed with schizophrenia, whereas those with mood disorders tend to emphasise words relating to the self (such as body and ingestion). Perceptual and causal words are positively correlated by those with mood disorders but disconnected in schizophrenia. A loosening of the link between sensory inputs and causal thinking in the schizophrenia sample may, in part, explain the role of delusions which are an attempt to make sense of troubling changes that the person experiences.

Importance of context

Although Fineberg et al present a compelling case, some caution is indicated as they do not compare texts from an equivalent source. The writings of the sample with schizophrenia were drawn from the Schizophrenia Bulletin, a peer-reviewed research journal. Its editorial guidelines require that contributions ‘be clearly written and organised, and that a novel or unique aspect of schizophrenia be described’. By contrast, the mood disorder texts were taken from ‘personal stories’ published on the website of the Anxiety and Depression Association of America (ADAA; www.adaa.org), a pressure group founded in 1980 as the Phobia Society of America. As a charity, the ADAA campaigns for greater awareness of anxiety and mood disorders and seeks to offer a voice for patients. Their personal stories have an optimistic tone and may have been filtered to de-stigmatise mental illness and create a positive attitude towards treatment. These differences in context might explain the correlation found between perceptual and causal words found in the mood disorder group and the negative association in the essays of the schizophrenia sample. Any doubt about the role of context in framing language may be dispelled by considering a request for a loan written to a bank manager compared with a love letter to a fiancé, albeit by the same author. Furthermore, the paper reveals the power of a discipline to dictate the presentation of an academic study. No text from patient accounts is quoted in the article. Had the research been submitted to a humanities journal, these would have been considered essential on the grounds that form and content are inextricably linked.

I randomly selected a ‘first-person account’ from the Schizophrenia Bulletin and the two-page text was found to contain 84 instances of the pronoun ‘I’ and only four cases of ‘they’. ² Describing himself as ‘a very internally tormented person’, Tim Woodman, the author, attributes his illness to complex family dynamics and interpersonal difficulties at work. His experience, therefore, may not be representative of the schizophrenia sample. This raises a question about the homogeneity of the schizophrenia population and whether the diagnosis embraces a broad range of mental states. Indeed, Fineberg et al acknowledge that the authors of the schizophrenia essays may represent a particularly functional subgroup.

Educational attainment is an important variable determining the quality of written communication. The authors argue that fewer punctuation marks in the schizophrenia sample may reflect ‘a disorganised quality to the language’. However, the experience of marking student essays suggests a more prosaic explanation: that participants may have received differing levels of teaching. The selection of positive and media-friendly accounts for the ADAA website may have introduced a bias towards well-educated and articulate individuals in the mood disorder population.

Does time heal?

Nancy Andreasen, a professor of renaissance literature before training as a psychiatrist, argued that deficits of language were found in patients diagnosed with schizophrenia and mania. ³ The difference between them was that the abnormalities tended to be fixed in the first group, enduring during periods of remission or recovery, while remitting in the mood disorder population. Subsequently, Andreasen observed that the two populations could be distinguished by the qualitative differences: schizophrenia being characterised by poverty of speech and content, such ‘impoverishment’ not being present in the mania sample. ⁴

If there are objective differences in writing between those experiencing schizophrenia and those with mood disorders, it raises the question as to when the distinctive presentational style of the patient with schizophrenia first appears. Does it arise in a prodromal phase or during delusional mood, and to what extent is it muted during periods of remission? The poetry of Ivor Gurney (1890-1937) provides anecdotal evidence. Gurney had suffered an acute breakdown when studying at the Royal College of Music before the First World War. As a private soldier in the Gloucestershire Regiment, he had fought at the Somme and at Ypres, being wounded and exposed to mustard gas. While convalescing in the UK, Gurney became mentally ill and was discharged from the army in October 1918 with a diagnosis of ‘deferred shell shock’. Kavanagh argues that his best work, written between 1919 and 1922, was inspired by his military service at a time before he developed psychosis. In his 1922 poem There is a Man, composed on admission to Barnwood House Hospital, Gurney sought to understand his deteriorating mental state: ‘but the pain is in thought, which will not freely range’. ⁵ His symptoms intensified and in December he was transferred to the City of London Mental Hospital at Dartford with a diagnosis of ‘systematic delusional insanity’. He had become convinced that an external agency was using electricity and wireless to torture him in mind and body. ⁶ Although Gurney had a brief burst of creative energy in 1926, his physical and mental health deteriorated and the final 11 years of his hospitalisation he produced little work judged of literary merit. When Ivor Gurney entered an asylum for a final and protracted admission, he retained sufficient insight to recognise that his capacity for imaginative thought was impaired. The sadness of his illness was that increasingly intense persecutory delusions progressively eroded his creativity and ability to communicate.

While not actively encouraging their patients to write, Fineberg et al consider text valuable evidence in terms of diagnosis and insights into pathology. This is in contrast to the position taken by 19th-century alienists who believed that writing about one’s experience was harmful, to the extent that they often denied asylum patients access to pen and paper. Self-expression was thought to exacerbate mental illness and the ‘talking disease’ was seen as a mark of a hysterical personality. ⁷ John Perceval, a former army officer and son of the assassinated prime minister, endured 17 months in Dr Edward Long Fox’s lunatic asylum in Brislington. Perceval complained in his Narrative that he had been consistently prevented from communicating with others to the extent that his letters were destroyed or censored. ⁸ These ideas endured and in 1951, when R.D. Laing was deployed to the British Army’s psychiatric unit at Netley, he found that patients with schizophrenia were discouraged from talking to each other and to staff during their treatment. Communication was thought to aggravate the psychotic process, akin to ‘giving a laxative to someone with diarrhoea’. ⁹

Conclusions

While Fineberg et al draw a number of persuasive inferences about the relationship between linguistic patterns and cognitive processes, the biggest challenge remains the nature of the relationship between aberrant forms of communication and underlying pathophysiology. Nevertheless, their findings accord with the dopamine-salience hypothesis of Kapur who argued that people who experience schizophrenia attach excessive salience to external events while using biased probabilistic reasoning (a jump-to-conclusions style) to provide an apparently convincing explanation. ¹⁰