The Discursive Construction of Mental Illness in UK Newspapers (1984-2014): A Critical Corpus Linguistic Analysis

This thesis explores the representation of mental illness in the UK press. Specifically, it addresses the following central research questions: ‘What do the terms ‘mental health’ and ‘mental illness’ refer to, ‘How are people with mental illness named and referred to in reports on mental illness’, ‘What are the salient transitivity processes in news reports on mental illness’ and ‘Do press reports on mental illness accurately portray the symptoms of specific mental illnesses?’ In order to investigate these questions, I designed and constructed the MI 1984-2014 corpus, which comprises 50,972,932 words of UK local and national news articles from 1984-2014. The stretch of time covered by the corpus is an important period for legislation related to mental health, including the 1983 Mental Health Act and the amendments to this act in 2007. I use frameworks drawn from corpus linguistics (e.g. keyness and collocation analysis) and Critical Discourse Analysis (e.g. naming and transitivity analysis) to analyse the MI 1984-2014 corpus. The main findings from my study are as follows: (i) lexical items in the semantic domain of mental health and illness are undergoing semantic change (e.g. the term ‘mental health’ is being used more frequently to refer to states of mental illness via a process of socially motivated and euphemistic language change); (ii) with regard to naming practices and, in particular, naming practices that anti-stigma initiatives have identified as problematic and stigmatising, the press use identity-first forms (identified as stigmatising by mental health advocates) to refer to people with mental illness (e.g. ‘a schizophrenic’) more often than person-first forms (such as ‘a person with schizophrenia’); despite this, early evidence suggests the press are increasingly adopting person-first language, which is the linguistic structure promoted by mental health advocates; (iii) with reference to transitivity, whilst the press overall represent the process of having mental illness as ‘suffering’, first-person accounts from people with mental illness are proportionally 4-times more likely to refer to having mental illness as ‘experiencing’ it (e.g. “I was experiencing psychosis”). I found that, overall, reports that include symptoms of mental illness are inaccurate, or are reported in contexts that are too specific to serve the purpose of properly informing the public about mental illness. On the basis of these findings, I argue that it would be beneficial for journalists and mental health charities to make a number of changes to the way they write about mental health. One basic but important change for mental health charities would be to take account of linguistic evidence prior to creating guidelines stipulating prescribed linguistic forms for discussing mental illness in the press. A further important change for journalists would be to more accurately depict the symptoms of mental illness in news articles and ensure that symptoms are contextualised appropriately (e.g. not used in reference to violent attacks). This thesis is offered as a contribution to the developing field of medical humanities. It provides findings and methods for examining further the issue of the press representation of mental illness and the related impact on society (and on individuals in society) that this can have