The 'rhetorical concession': a linguistic analysis of debates and arguments in mental health.

BACKGROUND
Debates about psychiatric diagnosis can be contentious and problematic.


AIMS
To examine the rhetorical devices used by mental health professionals when debating fundamental issues of psychiatric diagnoses online.


METHODS
Linguistic analysis of online articles and blogs.


RESULTS
Asymmetrical themes of; assertion of authority, rhetorical concession, appeal to moral clarity, and positioning the opponent were found, at the expense of evidence-based dialogue.


CONCLUSIONS
These argumentative linguistic themes adversely impact important discussions within mental health, which have the potential to be more effective through avoiding polemical rhetoric themes.


Introduction
The field of mental health care is notorious for fundamental conceptual disagreements, which frequently lead to emotionally charged and disrespectful comments (Bajwa et al., 2018). Mental health is a disputed field, with colleagues disagreeing about the fundamental nature of mental health problems, causal mechanisms, and what constitutes appropriate care (Kinderman, 2013). The language used in these debates reflects this (Bajwa et al., 2018). One area of current debate concerns the use of diagnosiswhich may either benefit the individual using mental health services or potentially increase the risk of discrimination and cause harm (Schlier & Lincoln, 2019). Excessive use of a medicalised approach (that is, relying on biomedical explanations, diagnosis and medication) may increase coercion (Verbeke et al., 2019) and has been criticised from a number of perspectives, perhaps most notably the United Nations Special Rapporteur on right to health (P uras, 2017). On the other hand, quasi-medical diagnosis is seen as a fundamentaland necessarypart of conventional psychiatric practice (Frances, 2013).
Language use, in this context, is important. Respectful dialogue between colleagues requires appropriate discourse (Bajwa et al., 2018). And, probably because of the contentious nature of the subject-matter, discourse in mental health can become strained, especially on social media. On platforms such as Twitter (perhaps designed to be confrontational), the ubiquitous tag 'pill-shaming' directed at anyone who raises concerns about the benefits or safety of psychiatric drugs (Wessely, 2018) and professionals describe both colleagues and people who use mental health services as "toxic" (Huda, 2020a), liken them to scientologists, and imply that they have links with 'Nazis', 'Klansmen' (Huda, 2018), 'colonialists' (Huda, 2020b) and fascists of various kinds (Bolstridge, 2017, Huda, 2020c. This is, as Bajwa et al. (2018) point out, unprofessional and unhelpful. But this language of confrontation also speaks to the frameworks of understanding that different people bring to the debate. The principle of linguistic relativity, or the Sapir-Whorf hypothesis, suggests that, because language choice reflects our thought processes, the language used can give an insight into how people think (Mehl & Pennebaker, 2003). Our language not only reflects and reveals the concepts guiding our thinking, but is also part of the mechanism of change. How we speak and how we think reflect underlying cognitive schemas, so speaking differently may lead to thinking differently (O'Neill, 2015).
Linguistic analysis is the scientific study of language, grammar, syntax, morphology, vocabulary and tone. It focuses on individual words used within context, to establish emergent themes through a careful syntax. How we use words in language is important, language conveys information about experience, thoughts and feelings. Within linguistics, word valence can indicate emotional processes; such as, inward focusing pronouns (the use of "I", "My") for firstperson singular expressions having a more emotional effect in writing (Lyons et al., 2018).
Linguistic analysis methods have been used in studies within mental health to build computational models for detecting stigma. Li et al. (2018) found that incorporating linguistic analysis methods into online detection of stigma can be beneficial to improving the performance of stigma reduction programs. Media representations of mental health problems influence readers' understandings and attitudes towards those with a psychiatric diagnosis (Bowen et al., 2019). Greater understanding of the intricacies of language use allows us to explore the effects of linguistic characteristics, such as how we internalize and process language.
Such linguistic analysis may also significantly complement more traditional modes of research, extending and adding value to qualitative approaches (Billington, 2019). Kaszynska (2015) argued that techniques of linguistic analysis could help explore what she termed the "visceral nature of cultural experiences", and identify "moments of subtle mental change and personal breakthrough, cognitive revaluation, interactive mind" (Davis et al., 2020).

Method and ethical considerations
The present study aimed to investigate the linguistic and rhetorical strategies used by professionals taking contrasting stances in the key debates within mental health. It used publicly available material and therefore individual participant consent was deemed unnecessary, since the use of blogs in the public domain implies consent (Eastham, 2011).
Nevertheless, the current research was covered by a generic ethical approval from the University of Liverpool Health and Life Sciences Research Ethics Committee (Human participants, tissues and databases; number 4873). Blogs were also further anonymised by the researchers for privacy protection. The present study was not supported by external funding.

Materials
A selection of six of 'blog' publications were examined, including perspectives critical of psychiatric diagnoses, articles supportive of psychiatric diagnoses and blogs which contain both approaches. A purposive, as opposed to representative, strategy was used to select blogs for analysis, as this was most suited to the research aim of contrasting specifically divergent perspectives (Campbell et al., 2020). All the blogs examined were in English; five by authors from the United Kingdom and one from the United States of America.

Source 1: (psychiatrist 'A')
Psychiatrist 'A' falls strongly on the side of psychiatric diagnostic language, stating firmly that mental illnesses are just thatillnessesand that a diagnosis is necessary for patients to optimize treatment.
'A' relies heavily on the force of the personal pronoun "I". This illustrates the first identified linguistic theme, the use of asserted authority. But authority is combined with a second linguistic themerhetorical concession.
'A' uses the adjective 'sad' to describe an emotional reaction to those who believe that mental health does not exist; "It's quite strange, quite sad I think, that after decades of anti-stigma campaigns, work by charities like Mind, so many confessions of mental illness and progressive government policies, that some people still doubt mental illness even exists". However, the use of the word 'sad' is not a disclosure of personal emotion by the author, but is rather a rhetorical attempt or tactical concession to claim the status of humane regret in service of an argument. Similarly, the adjective 'proud' is used to establish their own credentials and authority: "But I am very proud to defend it tonight". 'A' similarly states that mental illnesses are "simply" illnesses which involve the brain. Here, the word 'simply' does not add to the logic of the statement, but instead serves as a rhetoric of apparent objectivity; a non-arguable statement of authority.
This rhetorical concession is delivered through the use of reductive adverbs and adjectives inserted as small tonal persuaders. 'A' offers an apparent concession; that "some""illnesses" are caused by events in the outside world. This represents 'A' offering an apparently reasonable rhetorical concession to the larger audience and holding back the remaining argument before leading with "but" to assert the original stance in favour of pathological explanations. The position of humane reasonableness is maintained in the following paragraph where they argue that the mind 'deserves' its own specialty; one that also 'respects' social (not just biological) factors. Again, this may look as though they are heading towards the 'middle ground' of pragmatic viewpoint, but they are still keeping their argument on the importance of the label of diagnosis and strict psychiatric criteria. This occupation of the 'middle ground' seems to serve two functions; as 'a place for thinking', evaluating the strength of differing perspective, but also as a strategic or rhetoric device whereby polemical extremes are contrasted with the writer's own implied balanced position.
'A' goes on to speak in the language of finding a "cure"even if only, and apparently modestly "sometimes" (adverb). However, 'A' remains clearly in the camp of believing that there is a 'disease'. Mental illnesses are "plainly" (like the adverb 'simply' earlier) real. There are further examples of asserted authority in 'A's fifth paragraph, with an expert academic description of symptoms, using the term 'syndrome' to describe mental illnesses. "Even find a cure" highlights the fundamental pathological centrepiece of their argument: 'cure' suggests that there is a 'disease' and allows 'A' to revert into the language of traditional psychiatry. To which 'A' adds that mental illnesses are 'plainly real'-suggesting expert professionals can perceive this easily and find medicalized solutions.
'A' takes an apparently modest and defensive posture; stating that they lack years of experience, but then counters with the years of time spent with patients and states that they were a patient themselvesfurther to reinforce their personal credibility: 'I don't have the years of research experience that the other speakers have; all I have are years spent with patients and my own time as a patient'. This may look conciliatory, but the key phrase 'all I have' is assertiveness cloaked in English modesty. 'A' then introduces an apparently objective image of a brain scan of a 'depressed' against a 'non-depressed' patient, but without comment on the research. This again illustrates the process of rhetorical concessionapparently conceding a point in the debate, only to use it as an argumentative device.
'A's third paragraph introduces a less polemical tone towards mental health being pathologically caused and 'cured'. It states that the mind 'deserves' its own specialty; one that 'respects' social factors (not just biological). These key words offer an implicitly moral language within an apparently pragmatic approach for nevertheless maintaining the importance of the label of diagnosis and strict psychiatric criteria. An ostensible humane discourse contains within it an assertive or even aggressive argument, and appeal to moral clarity. Indeed, 'A' occasionally uses language that reflects more overtly the polemical stance, for instance in the common sense tone of: "Let's be clear on another thingpsychiatry is not "growing out of control", you only need to read the headlines about the shameful ways our services are being cut to see that", "Some mental illness or disorders are quite plainly real", "The choice is your own, but you have to understand what voting against mental illness would mean", and finally, "One short step away from get over it. I thought we'd moved past all that". This final comment also suggests that 'A' is subtly positioning the opponentcreating a perspective or point of view for anyone who might disagree which renders them (for example) morally bankrupt or scientifically illiterate.

Source 2: (psychiatrist B)
'B' is also an assertive advocate for psychiatry. In the first paragraph of the selected text, 'B' notes that "maybe" people may get a wrong diagnosis or "maybe" their diagnosis has changed over time, indicating a less driven argument. "However", 'B' insists that labels aren't particularly stigmatizing as much as they are functional. "I've been practicing for a long time", "I understand that people are miserable": as with 'A', this personal emphasis is more to do with the polemics of argument than scientific reasoning. This is another example of the themes of rhetorical concession and of the assertion of authority.
This therefore begins to shape and pre-determine the rest of 'B's argument. "I think that most people who worry their diagnosis will cause others to judge them negatively simply don't tell people that a doctor says they have disease X"the word disease almost slipped in under cover of its not being admitted. 'B' states that "many" will say a diagnostic label is unfair, and "many" will worry about the judgements of others, but "personally", which is an emphasised tone, they do not find it so: "sometimes, people behave in distressed or dysfunctional ways and the fact that they have a psychiatric disorder is obvious, this is not because of the title of the label, it's because of the symptoms of the illness and the persons behaviour". The tell-tale word is "obvious": the illness is identified as the problem, not the psychiatrist. In the last paragraph, 'B' invites any who have suffered distress from the diagnostic label "alone" to "come forward": it is more a challenge than an invitation.

Source 3: (psychiatrist C)
'C' is a psychiatrist with a more middle-grounded stance towards diagnosispossibly reflecting his own moderate opinions, but perhaps also taken as a rhetorical or strategic position. In the first paragraph of selected text, 'C' writes that there has been a "crisis" taking place within psychiatric diagnoses. This already has a less polemically aggressive stance: 'C" is not defending psychiatric models, but accepts that there have been fundamental flaws. 'C' then asserts an authoritative, detached, tone by stating figures and objective informationrather than using 'I' in abundance to describe experience and assert dominance such as 'A' did. This is still, however an illustration of the theme for the role of authority. 'C' states that "25% of us qualify for a mental disorder and unless checked, the DSM may turn diagnostic inflation into future hyperinflation".
Within their '12 Tips of Diagnosis', 'C' states that there is "no bright line demarcating the boundary between mental disorder and normality", and that therefore everyday experiences should not be medicalised mechanically. Stating that there is no "bright line", no "clear" indication for categorisation, 'C' constitutes an example of someone in the middle grounded stance. They also state that milder problems may resolve "without treatment", further arguing for their pragmatic viewpoint. This is an example of rhetorical concessions offered, or conciliatory stances proposed, as a mechanism for levering support for their overall stance.
In their third paragraph, 'C' states that "environmental stresses" may cause difficulty with diagnoses. This seems to refer to issues affecting neural functioning; such as trauma, stresses, life hassles and experiences. 'C' also states that adolescents are 'difficult to diagnose', stating that initial diagnoses are 'unstable' and 'inappropriate'. However, after offering these concessions, 'C' reverts back to the more psychiatric stance by saying that models are still valued and reliable: "If you hear hoof-beats on Broadway, think horses, not zebras! When in doubt, go with the odds. Exotic diagnoses may be fun to think about-but you almost never see them. Stick with the bread and butter". Once again, this illustrates the themes of assertion of authority and rhetorical concession.

Source 4: (psychologist D)
The clinical psychologist 'D' is another 'middle grounded' stance member, and states frankly that they are a "long term critic" of psychiatric diagnoses, adopting a position seemingly more moderate than Psychiatrists 'A' and 'B'. Like other contributors, 'D' uses both appeals to authority and appeals for moral clarity: "the ones most affected by this argument are not us as professionals".
They begin from a phenomenological camp, however: "I've been increasingly frustrated by the myths and over-generalisations that get repeated and recycled in the diagnosis debate". The use of "I've been frustrated by" is not unlike the assertiveness of 'Psychiatrist A'. 'D' discusses the use of biological tests in relation to sleep disorders, anorexia and neurocognitive disorders, and states that 'psychiatric diagnoses can be reliableif not sometimes excellently reliable'. 'D' also makes comparisons with conventional medical diagnoses, also made involving human judgement, and the role of a committee decide on a diagnosis: although they "aren't perfect, they aren't invalid". However, 'D' also states that; 'the ones most affected by this argument are not us as professionals'. Here, the language of "we" is used manipulatively to argue that; "we need a system to support everyone"; "I think we'd be better off if we treated diagnoses more like tools and less like ideologies" and "we should strive to ensure a range of options is available; both diagnostic and non-diagnostic".
This is designed to suggest that although 'D' is fundamentally a phenomenologist, they are willing to accept and even suggest other approaches to improve care. However, 'D's argument does not speak to the 'paradigm shift' that, for example, Psychologist 'F' proposes when, for example, they use the term 'disability' instead of 'illness'.
Generally, 'D's comments avoid polarizing rhetorical tropes, however, their conclusion tends towards positioning the opponent: "Finally, I think we'd be better off if we treated diagnoses more like tools, and less like ideologies", "And if you are really against classifying experience and behavior in any way, I recommend you stop using language, because it relies on exactly this". "Saying the DSM is about 'labelling distress' when many DSM diagnoses do not will get nothing more than an eye roll from me". The point here is that there is no immediate reason to suppose that anyone of a contrary point of view would necessarily 'treat diagnoses like ideologes' or would be opposed to classifying experience 'in any way'.

Source 5: (psychiatrist E)
'E', a psychiatrist with pronounced 'critical' views (at least in the blog analysed), states that; "Sadness is a normal emotion, and low mood is a common experience found in which we all have to deal with". 'E' states; "As we all know what it is to feel sad, we can all have some understanding of why to suffer from clinical depression is so horrible; just think of feeling like that all the timeand an episode of clinical depression may last for months'. 'E' also states that "Though an episode of depression may be associated with a difficult life event such as loss of job, what is not so commonly understood is that clinical depression can also be triggered by experiences that are not necessarily associated with recognisably saddening events". The phrase "not so commonly understood" asserts professional authority; they understand people's perceptions of mental health; however 'E' has their own viewpoint which is deemed more correct, and takes a more holistic and less aggressive defensive stance for the role of psychiatric models. 'E' goes on to state that 'psychiatrists are aware of mental health disorders occurring along with depression', so 'Psychiatrists are medical doctors. We make medical diagnoses, taking all physical and mental factors as a whole'. This is another example for the theme of authority. 'E' uses this position of authority to question, as well as promote, his profession. "Picking up the physical symptoms and having the breadth of medical knowledge that comes with being trained in medicine, is sometimes about recognising what doesn't fit in with a familiar pattern and exploring it".
One key phrase 'E' uses is that; "Having depression doesn't always mean you will need to take antidepressants". This seems to differentiate 'E' from 'A' and 'B', as 'E' accepts alternative viewpoints and does not emphasise the need for medicalisation. This may be calculated to enhance 'E's perceieved authority.

Source 6: (psychologist F)
Psychologist 'F' is a clinical psychologist: their blog highlights the more protective tone of the argument against over-diagnoses. True to the rhetoric of polite and reasonable concession to the other side, 'F' states that they have the "utmost respect" for medical colleagues, "but the medicalisation of childhood has gone too far". This highlights the deployment of rhetorical concession through this use of ironic English.
'F' uses the word "understanding" as a term for describing symptoms as everyday life experiences rather than illnesses. 'F' states that the 'consequences [of inappropriate diagnosis] could be catastrophic, especially for vulnerable populations such as children and older people'. The word "catastrophic" is therefore played off against 'vulnerable", positioning any opponent who wishes to avoid catastrophe and (in an appeal for moral clarity) claims to care about vulnerable people. For 'F', patients' personal qualities and emotions are valued in human terms: 'We need to step back from the 'diagnosis & treat' mentality and understand people a little more.' But, here, both the word "we" and the phrase "a little more" are again rhetoricalrecruiting the reader as an ally (one of the 'we') and reasonable and modest and humane in the face of the 'catastrophic'. This is showing deliberate contrast, positioning the opponent -'you are catastrophic, I respect humanity, I am therefore reasonable'. 'F' states that; "a child's behavior is highly dependent on context", and that "people change, agreement with specific criteria does not confer validity": this is an argument against rigid and static diagnoses that offer strict categorical models for what may shifting contextual behaviours. In a paragraph discussing validity, 'F' states: 'The poor validity of psychiatric diagnoses-their inability to map onto any entity discernable in the real world-is demonstrated by their failure to predict course or indicate which treatment options are beneficial, and by the fact that they do not map neatly onto biological findings, which are often nonspecific and cross diagnostic boundaries'. This is humane understanding now adopting the language of science: prediction, refutation, mapping, non-specific.
'F' also states that; "diagnoses are used as pseudo-explanations for troubling behaviours", driven by social anxieties and political concerns. 'F' states that DSM-5 is a "potential threat" to the public, as though it were a disease or a weapon itself. 'F' seeks a more nuanced language, closer to human life; "psychological distress is a normal part of human life" alongside "individual formations being necessary when therapeutic decisions are made". This highlights the concept of situational, individual factors that psychologists support when defending the argument of phenomenology. This is still however labelling and positioning the opponent. 'F' assertively positions opponents by defining their arguments for them, and in language they may not have chosen themselves.
Finally, 'F' relies heavily on appeal to moral clarity, for example, the use of "should" with "we": "A valid and humane system for identifying, describing and responding to distress should reflect these principles. We need a wholesale revision of the way we think about psychological distress". "We need to step back from the 'diagnosis & treat' mentality and understand people a little more. There is no reason to assume that these phenomena reflect underlying illnesses".

Discussion
Linguistic analysis of the rhetorical devices used by mental health professionals when debating fundamental issues of psychiatric diagnoses online revealed themes of; assertion of authority, rhetorical concession, appeal to moral clarity, and positioning the opponent. In general, polemical language was seen to have been used repeatedly within the blogs examined, creating a potential barrier for communicative purposes between the authors.
In each blog, conceptual shifts or concessions were visible. These blogs seem designed to create a reaction, or make a point, rather than address the real issues. Clearly all clinicians use rhetorical and argumentative devices; they all engage in the debate that they want to win.
There was, however, evidence of a more genuine, less rhetorically strategic, middle ground, a holding ground for multiple thoughts and the complex relations between them, as shown particularly with clinicians "C" and "D". Nevertheless, these accounts still included "rhetorical concessions". This can be seen as a form of 'convenience convergence', which inhibits the genuine propositions that the middle-grounded clinicians make. The term "convenience convergence" was created on the basis of diplomatic positions focusing selectively on certain points of an argument in order to win, rather than the underlying argument or other points of an argument being addressed.
In these accounts, polemical tones tended to override the fundamental arguments. In this analysis, polemics between authors were preventing the emergence of solutions. This splitting into warring oppositions has occurred for years, and the risk is that this will continue. The differing conceptual models produce combative rhetoric. Debates online are still polemical.
An apparent asymmetrical debate was observed in the toand-fro argument between phenomenological and pathological linguistics. Arguments and counter-arguments are generally presented as either statements or as emotive human cases. Factual claims are pitted against emotional accounts, such as 'antidepressants have a certain percentage of recovery out of a given population' (this being a factual claim or statement) compared to the more personal accounts (such as the negative effects of stigma within mental health). It is falsifying rhetoric that is at fault here in the service of polemical opinion and vested interests: rhetoric and tone that get in the way of thought. There is a good rhetoric (the use of compositional techniques such as the capacity of sentences to include different clauses) that enables responsive and responsible thought, responding visibly (by syntax and vocabulary) and aurally (by tone and attention) to the pressures of a complex case/situation.
There is a temptation to see the 'middle ground' of an emotive debate as almost automatically balanced and wise. The rhetorical device (much-used by journalists in political debate) that anyone positioned between two different perspectives automatically holds the most wise and judicious position is common … but remains a rhetorical position, and equally lacking in evidence and logic. All those debating the important issues addressed here, including those occupying the 'middle ground', need to improve their abilities to adhere to logic and relatively inclusive rather than predetermined evidence, and reduce reliance on polemic and rhetoric in their writing. Even the middle-ground, which shows evidence of 'convenience convergence' (only picking certain points of an argument to 'win'), employs rhetorical devices even when appearing diplomatic. The 'middle ground' too often may have an appearance of 'having the better argument' simply because of residing in the middle groundwhich is a rhetorical position in itself.
This analysis demonstrates the potential of nuanced thinking and where it gets sabotaged. These pieces of writing should be creating a space for thought, not driven by pre-determined polemical extremes, but equally not merely splitting the difference. That creation of a richer speculative space for investigation allowed for thought to challenge opinion and to make of blog posts not necessarily a polemic but more ideally a place for wider dissemination of thought and a more productive public debate. If the thoughts (and hence the symptomatically more complex syntax they demand) were sufficiently deep and intervolved, rather than split into simple extremes, it would help ensure a more authentic discourse. A genuine area for thinking needs to be created, and that area need not be 'in the middle': it can move between positions because it allows, at best, more mobility of thought in relation to a greater density of evidence.
The middle ground is counterintuitive when it is just another strategy, however it is helpful when it allows in more thoughts otherwise made mutually extreme, or when it thus becomes a holding-ground for specific thinking, tried out in its language. Blogs in this crucial human area should not succumb to the lowest common denominators for public debate in this genre.
The debates within mental health are important. Debates about diagnosis, cause, the use of medication are all important. Facts need to be engaged, rather than asserting polemical justice. Precisely because these are important debates, arguments should be won on logic and evidence, rather than rhetorical devices and argument. It is not to do with 'logic' as such (and 'facts' too are contested, not merely neutral), but it is to with genuinely troubled or careful thinking inflected by evidence and by a variety of competing considerations that affect and even pressurise syntax.

Disclosure statement
Bethany Garner: None Peter Kinderman: Peter Kinderman is employed by the University of Liverpool. Over the course of his career, he has received research grants from the National Institute of Health Research, the Medical Research Council, the Economic and Social Research Council, the Wellcome Trust, the Youth Justice Board for England, various NHS Trusts, the Department of Health, the European Commission, the British Psychological Society, and the Reader, and personal fees from the Department for Constitutional Affairs, legal counsel, the BBC, Smoking Gun Media, GLG Group, True North Productions, and Compass Pathways Ltd. He has also received royalties on published work in the field of mental health from Palgrave Macmillan, and Little, Brown Book Company, and travel and hospitality from a variety of organisations in relation to speaking and other professional activity. He has a remunerative secondment contract through his employers to act as a Clinical Advisor for Public Health England. He is a Member of the Council for Evidence Based Psychiatry, a director of Kyrie Therapeutic Farm Ltd, and a former President of the British Psychological Society. He owns shares in Alphabet (the parent company of Google), and Twitter, and is a Director and owner of Kinderman Consulting Ltd.
Philip Davis: None

Funding
The author(s) reported there is no funding associated with the work featured in this article.

Data statement
The data that support the findings of this study are in the public domain.