Self-injury as self-cutting: the exclusion of overdoses in the present

The new DSM-5 category of non-suicidal self-injury (NSSI) excludes self-poisoning, which is described as ‘intentional self-inflicted damage to the surface of his or her body’. With the specification of surface, self-poisoning is ruled out.¹² However, general hospitals still include both cutting and poisoning under ‘self-harm’ in their statistics. As seen in the Introduction, there is in the literature a strong differentiation of motives between cutting and overdosing – a differentiation that deals exclusively with self-cutting. These studies tend to be smaller scale, qualitative, and interview-based.

This differentiation between self-cutting and self-poisoning is varied and complex. It is largely achieved through four interlinked strategies, which can be labelled as: general assertion, motivational ambiguity, visibility and clinical management. General assertions are often rather sweeping statements, such as Tantam’s and Huband’s claim in 2009 regarding the ‘very different cultural and psychological roots of self-injury and self-poisoning’. For them, self-injury means solely ‘cutting, burning or otherwise damaging the skin and its underlying tissue’.¹³

In 2006, clinician Leonard Fagin begins with a general assertion, but then develops this into a comment on the motivations behind the behaviour:

I see self-injury as different from self-poisoning, where substances (usually drugs) are ingested, usually in order to die, cry for help or obtain temporary respite from unhappiness or unbearable distress, and I believe that people who poison themselves have different characteristics from those who injure themselves.¹⁴

As far back as 1988, Barent Walsh’s and Paul Rosen’s book, Self-Mutilation, contains the following passage based on a criterion of visibility, and then develops into an argument about ambiguity of motivation. This passage is quoted by Armando Favazza in 2011 as ‘the best explanation’ for maintaining the difference between self-cutting and self-poisoning:

In the case of ingesting pills or poison, the harm caused is uncertain, ambiguous, unpredictable, and basically invisible. In the case of self-laceration the degree of self-harm is clear, unambiguous, predictable as to course and highly visible. In addition, self-laceration often results in sustained or permanent visible disfigurements to the body, which is not the case with overdose. In various ways, therefore, these two forms of self-harm are quite different; the danger in combining them in a single category is that these important differences (and their clinical implications) are overlooked.¹⁵

In 2007 Jan Sutton differentiates between the behaviours along precisely these lines, arguing that self-poisoning is invisible and self-cutting visible, and therefore motivations for self-poisoning are ambiguous whereas for self-cutting, the intent is clear. She claims that ‘self-injury is now well recognised as a coping mechanism and survival strategy, whereas the intent behind self-poisoning is less clear ... It could be a botched suicide attempt, it could be an accident, it could be a cry for help, or it could be a means of temporarily escaping from emotional turmoil’.¹⁶

As well as visible versus invisible harm, and ambiguous versus clear motivation, the behaviours are further separated by clinical management strategies. In 2008 Pengelly et al. contribute to a debate about ‘harm minimisation’, building upon National Institute for Clinical Excellence (NICE) guidelines from 2004. Their guidelines include: ‘If you feel you must cut, only use clean, sharp instruments to reduce the risk of infection and complications. Keep tetanus protection up-to-date ... Avoid alcohol and drug use as you may inflict worse wounds than intended ... Gradually reduce the severity of your injuries. Leave more time between injuries’. This practical minimisation attitude disappears when it comes to poisoning, as they state: ‘Do not take tablets. There are no safe overdoses – even “small” overdoses can kill’.¹⁷ Whilst self-cutting can be managed and minimised, self-poisoning must be prohibited. This feeds off and feeds into the stronger association with death that self-poisoning acquires between the late 1970s and the present. It is important to stress that I am not contesting any of this advice, merely pointing out that in terms of visibility, motivation and management, self-poisoning and self-cutting are strongly differentiated. All this effort confounds Favazza’s assertion in 2011 that ‘the British literature still does not make this distinction’ between self-injury and overdosing.¹⁸

However, Favazza is partially correct – there is a British literature that persists in combining self-poisoning and self-cutting – primarily general hospital–based psychiatric epidemiology. These professionals largely conduct studies from accident and emergency departments as well as attempt to record the prevalence of self-harm that does not present to hospital but is established by retrospective questionnaire. A 2010 report by the Royal College of Psychiatrists states that ‘[f]or the purpose of this report we define self-harm as an intentional act of self-poisoning or self-injury irrespective of the type of motivation or degree of suicidal intent. Thus it includes suicide attempts as well as acts where little or no suicidal intent is involved (e.g., where people harm themselves to reduce internal tension, distract themselves from intolerable situations, as a form of interpersonal communication of distress or other difficult feelings, or to punish themselves).¹⁹ To take another recent example, Hawton, Saunders and O’Connor define their object of study in a 2012 Lancet paper thus: ‘Self-harm refers to intentional self-poisoning or self-injury, irrespective of type of motive or the extent of suicidal intent’. Self-poisoning and self-cutting are thereby combined. But it is not as simple as that, as they differentiate the behaviours in terms of incidence, claiming ‘Self-cutting is the most common method of self-harm in adolescents in the community’ whereas for ‘adolescents presenting to hospital after self-harm ... self-poisoning is by far the most common [method]’. They also differentiate by motive: ‘[I]ndividuals who self-harm by cutting differ somewhat from those who take overdoses, with suicidal intention more often indicated for self-poisoning, and self-punishment and tension relief for self-cutting’.²⁰ This is the same motivational differentiation shown above: suicidal intention against tension relief, which maps reliably onto self-poisoning against self-cutting.

However, this nuance in the epidemiological studies is not often reported by the literature focusing upon self-cutting alone, even though – despite some differences – they present extremely similar clinical pictures. Thus, Sutton is widely understood when complaining that hospital statistics under the term ‘self-inflicted injuries’ contain 90% overdoses: ‘What sort of image does that [term] conjure up? Overdosing? I doubt it. Cutting? Highly probable ... mention the word “self-harm”, and it immediately conjures up images of people cutting themselves’.²¹ Recent books on self-harm have titles like The Tender Cut (2011) and Blades Blood and Bandages (2012), and recent novels about self-injury are entitled Cut (2009) and Scars (2011), leaving little doubt about the methods of self-harm employed.²² Whilst Sutton is right that there is a mismatch between the stereotypes that the term ‘self-injury’ conjures up (self-cutting), and the majority of people figuring in hospital statistics – 90% self-poisoning – this has not always been the case, as this book has shown in detail.

The scope of this chapter is not broad enough to focus upon all aspects of self-cutting, and instead focuses upon just one: the way in which self-cutting becomes conceptualised as a behaviour motivated by internal emotional states, rather than as a communication. This approach is in order to show how self-cutting becomes different from self-poisoning, which as we have seen, is intimately connected to communication and the social setting. As Shelly James points out in a recent dissertation, the reasons most often put forward for deliberate self-harm centre upon the relief of distress, a way of regulating or combatting emotional numbness. She also notes that social aspects remain under-explored.²³ James’s dissertation may well be part of a swing back towards more socially embedded explanations but, if so, such a shift has yet to gather much pace or influence. The focus of this chapter, drawing upon the contrasts with self-poisoning, means that some important parts of the self-cutting stereotype are not addressed. The part of this focussing-down that has caused me the most disquiet is the lack of attention to some of the gendered aspects of self-cutting. The idea that it is an extension of grooming behaviour (said to be more prominent in the female psyche) or a practice rooted in vicarious menstruation, cannot be fully explored here. The analytical heart of this book is the epidemic of self-poisoning, not the practice of self-cutting, and difficult choices have to be made. This makes the following a rather partial and fragmented account of self-cutting, but hopefully a full and coherent account of the ways in which self-cutting and self-poisoning interact. As we saw in the introduction’s analysis of successive editions of Myre Sim’s textbook, some awareness of self-cutting, wrist-cutting or wrist-scratching, linked to affect regulation, emerges between the end of the 1960s and the mid-1970s, with a significant nod to North American clinicians. Cutting becomes archetypal in the 1980s, and as will be discussed in the conclusion, resonates with neurochemical explanations of human behaviour.

British clinicians and self-injury: inpatients and American influence

How does self-cutting or self-injury emerge in Britain? In what ways and through which channels does awareness crystallise and stabilise? Sarah Chaney has written of the various self-mutilating practices in Victorian literature and psychiatry, but at issue here is the specific phenomenon of self-cutting that emerges in the 1960s – something that Chaney acknowledges as rather different: ‘self-cutting, often regarded a prevalent method of self-harm in the mid- to late-twentieth century, is not emphasised in nineteenth-century writings’.²⁴

In British psychiatry, the story of self-cutting begins in Chicago. The principal study consistently referenced throughout the early British and American work on self-cutting is by Daniel Offer and Peter Barglow, psychiatrists at the Institute for Psychiatric Research and Training, which is commonly referred to by the acronym PPI. PPI is part of the private, Michael Reese (General) Hospital in Chicago, and in 1964 it is a ‘psychiatric establishment [which] has a national reputation, especially for its research and teaching functions’.²⁵ Offer’s and Barglow’s study concerns an ‘outbreak’ of self-mutilation amongst adolescent and young adult inpatients over the nine months between November 1958 and August 1959, comprising ‘approximately 90 incidents of self-mutilation’. Although they relate that ‘[i]solated incidents of self-mutilation had occurred periodically during the eight-year history of the institution’,²⁶ the scale of this outbreak is unprecedented. PPI has a largely psychoanalytic or ‘dynamic’ approach, but Offer’s and Barglow’s conceptual approach

follows the social-field multilevel approach illustrated by the hospital studies of Stanton and Schwartz. A field method was used because it became apparent early that self-mutilation was a complex product of many interacting and interdependent factors. Its ramifications extended throughout most of the hospital structure, and etiological factors could not be meaningfully evaluated in isolation.²⁷

Offer and Barglow still use the language of suicide to a significant extent, claiming that ‘the self-mutilation incidents were “suicidal gestures” rather than “suicidal attempts”’, where the latter signifies a genuine attempt to kill oneself. They argue that in all bar one incident,

‘secondary gain’ was involved, and some conscious effort to gain gratification from the environment was seen. Increased prestige in peer group, desire for more attention from staff, competition with group members, expression of anger toward family or hospital personnel, were frequently encountered motives.

There is much British literature that focuses on the relationship between inpatient institutions and psychopathology – for example, Russell Barton’s Institutional Neurosis (1959), and John Wing’s and George Brown’s Institutionalism and Schizophrenia (1970). Illustratively, the opening chapter of the latter is entitled ‘Disease and the Social Environment’. This period also sees the dawn of so-called anti-psychiatry, in which the sociological anthropology of Erving Goffman is so influential. Indeed, as one historian expresses it, this is a time when ‘the diagnosis was social’.³⁰ As we have seen in previous chapters (especially Chapter 2), this socially focused outlook has roots in the Second World War. Tom Main, heavily involved in the second Northfield Experiment, addresses the British Psychological Society in 1957 in what becomes one of his best-known publications. Simply entitled ‘The Ailment’, Main draws attention to the ways in which certain psychiatric inpatients absorb disproportionate energy and attention from staff, creating problems, cliques and divisions within and between clinical staff.³¹ Despite this established seam of sociological influence, the inpatient literature on self-cutting moves away from social explanations and – slowly and unevenly – emphasises internal psychopathology.

From imitation and epidemics to internal psychopathology and internal tension

In 1963, polymath and psychiatrist Colin McEvedy writes a dissertation on self-inflicted injuries for his diploma in psychological medicine (DPM) at the Institute of Psychiatry in South London. He joins the Maudsley in 1960 upon leaving his national service with the air force, and impresses the institute’s director, Aubrey Lewis. McEvedy is best known for a controversial reinterpretation of ‘Royal Free Disease’, a 1955 epidemic among nurses at the Royal Free Hospital characterised by fatigue and ambiguous neurological signs. He and his co-author Bill Beard argue that this is a form of conversion hysteria, a conclusion that angers many.³² McEvedy also publishes on hysterical epidemics in secondary schools: one of ‘overbreathing’ amongst schoolgirls in Blackburn, and one of vomiting, abdominal pain and ‘faintness’ in Portsmouth.³³ He is also well-known for his historical atlases. His analysis of self-inflicted injuries centres upon an outbreak of self-cutting in Bethlem and St Francis Hospitals, and a group of 13 patients in particular. I have been unable to obtain permission from McEvedy’s next of kin to quote from this unpublished work, so I shall paraphrase throughout.

The only paper he finds that deals specifically with self-mutilation is Offer’s and Barglow’s (1960).³⁴ His work is partially concerned with the ways in which the behaviour might be learned or transmitted, but he also speculates upon the internal psychological reasons for behaviour that he considers to be bizarre and outside of recognised syndromes and symptom patterns. Crucially, of his opening case, Kay R., he relates that when discussing her behaviour with others he is questioned about the nature of her suicidal intent – using a continuum from a hysterical gesture designed to procure sympathy, to a ‘genuine’ attempt at self-killing. What strikes McEvedy is that he does not feel able to place Kay R. on this continuum. She does not seem to fit.³⁵

Thus, McEvedy sees existing explanations as inadequate – this is neither an attempt at killing oneself, nor an attempt to elicit sympathy, nor a reaction to stress. The spectrum of possible action utilised by Stengel and Cook in the 1950s – between a social-stress reaction and a determined attempt to kill oneself – cannot accurately capture the actions of Kay R., the archetypal self-cutter in McEvedy’s estimation.³⁶ He attends to the symbolism of these acts in a precise and sophisticated manner, arguing that even though popular opinion might hold that slashed wrists are lethal – and thus somebody might genuinely attempt to die by performing that action – he reasons that Kay R. must have realised quite swiftly that her wrist-cutting was more or less nonlethal, given that she kept surviving.³⁷ McEvedy’s work shows how a certain form of self-cutting comes to light in Britain through its separation from socially embedded understandings of psychopathology – and this despite his more famous work on hysterical epidemics.

There are a number of intellectual assumptions that make this Bethlem and St Francis group a coherent set of 13 patients. The purported similarities between the cases exist alongside an awareness of the varied nature of their pathological behaviours. They are selected because they repeatedly self-lacerate, but also share many other characteristics. They are all young and female, and their psychological problems are seen to take many antisocial forms, including screaming, rudeness and obscenity, smashing windows and crockery, or swallowing unusual objects and taking overdoses.³⁸ Additionally, ten of the thirteen cases are thought to simulate illnesses or fits, to exhibit conversion symptoms, to have fits that are not considered totally genuine as well as hallucinations thought hysterical rather than psychotic.³⁹ These patients selected for their cutting might manifest disturbance in very different ways, but these varied outcomes are thought to be rooted in (the same) impulsive paroxysm.

Despite this variation, it is self-injury that McEvedy investigates, and the patients all show injuries – some caused during aggressive outbursts (for example, window-smashing) and some cuts deliberately self-inflicted. He also mentions that a separate record is kept of overdoses and of any ‘bizarre’ swallowed objects. What is interesting, not to mention odd, to the sensibilities of the twenty-first century, is that injuries inflicted by window-smashing are included with self-cutting. This is perhaps because – as mentioned by clinicians below – window-smashing does not necessarily involve injury to oneself. However, in the controlled inpatient environment, it may appear as an obvious way to procure the sharp edges needed for self-cutting). Less jarring, but no less important, is the fact that overdoses are kept separate. This is one of the first examples that I have found of self-inflicted cutting (even though it includes window-smashing) being kept explicitly separate from overdosing. The claim is also made that self-laceration follows a remarkably consistent pattern, with the left wrist being cut most commonly (just less than half of all self-cutting incidents). Various behaviours are downplayed in order to cohere the group, as McEvedy refers to these patients – supposedly for reasons of brevity – as ‘cutters’.⁴⁰ In this study the behaviour of ‘cutters’ is significantly differentiated from socially embedded hysterical, cry-for-help communications as well as earnest suicidal attempts. There is a sense in which this behaviour is new and unsettling (repeatedly labelled ‘bizarre’) and that it repeatedly (if subtly) confounds existing categories. This clinical object is not self-evident, but is the result of human analysis and intervention.

McEvedy’s separation of this action from the social setting is based upon a differentiation between ‘spontaneous’ and ‘susceptible’ cutters – between those who perform the action on their own initiative, and those who try it in response to another patient’s actions. There is a subtle relationship between these groups, but there is little doubt that the copying is – ultimately – secondary to the unity of the syndrome of cutting. McEvedy notes, with regret, that he would like to re-categorise the patients in order to separate out those who perform the act without any imitation of others. However, isolating these non-imitators requires too much of a reworking of the material. He has no doubt that some of his Bethlem group are only classed as repeated self-lacerators (those with five or more cutting incidents) because they happen to be present during the self-cutting epidemic.⁴¹ This makes it clear that the key to the syndrome is in the internal impulse, not the social imitation.

McEvedy argues that the most notable aspect of behaviour is the apparently unprovoked mood swings – bringing emotional states to the fore. These are distanced from the social environment in the case of Kay R., who is said to have cut herself over and over, regardless of environment or levels of stress. McEvedy finally distances cutting from a socially motivated phenomenon because the impulse (presumed to underlie all the behaviours – from cutting forearms to swallowing dominoes) seems so unorthodox that it cannot be explained by mere social pressure or stress. He reasons that there must be something preventing the (supposedly suicidal) impulse from being conventionally expressed.⁴² That it might also be connected with hysterical, susceptible imitators does not change the fact that the behaviour begins in a pathological, emotional, internal impulse rather than a disordered social setting.

McEvedy is not entirely sure what might replace the powerful aetiological force of the social environment or imitation, but he speculates that the so-called ‘spontaneous’ cutter’s personality has not only a high level of hysterical traits, but also something that he labels ‘hostile tension’.⁴³ This is the first mention of ‘tension’ as a key motive force for ‘cutters’ in Britain – something that is well established in the contemporary literature, and it emerges in McEvedy’s work as characteristic of the cutter who is not simply copying.⁴⁴

In sum, McEvedy makes a particular effort to distance the patients from socially embedded motivations. The research is not written up into any research articles and thus does not garner much attention or influence, although it is available at the library of the most influential psychiatric training institution in Britain, the IoP. Unlike his work on the ‘Royal Free Epidemic’, which is explained as conversion hysteria, this study remains largely obscure.⁴⁵ It is referenced by a study from a Plymouth adolescent unit in 1968, a more influential unpublished study in 1972, and a published article in the British Journal of Psychiatry in 1975 (all of which are considered below). Its significance lies in the fact that it isolates a group of female psychiatric inpatients, focuses upon one particular symptom, and presents in such a way that the impulse underlying that particular symptom is important, rather than its status as an epidemic behaviour or shedding light upon the social organisation of a hospital. Although bizarre and unorthodox to McEvedy, self-cutting in response to ‘hostile tension’ seems very familiar to us.⁴⁶

Later in the 1960s, D.W. McKerracher, a clinical psychologist at Rampton secure hospital in Nottinghamshire, publishes two articles of note with a number of different colleagues, all working at the hospital. There is an established literature on prison self-mutilation that emphasises self-harm as a response to the confined space, or to perceived injustices.⁴⁷ A secure hospital environment can feed off that frame of reference, but its status as a secure psychiatric hospital means that staff are likely to give close consideration to internal and psychopathological factors. In 1966 a comparison of the behavioural problems of male and female prisoners in the hospital is published. In 1968 there emerges a specific study of self-mutilation in ‘female psychopaths’.⁴⁸ These articles are important because they show how self-mutilation becomes more strongly established through understandings of internal psychopathology. The behaviour is seen as less outward-looking, social and communicative in its meaning, and more internal and emotional. It is seen as a ritualistic behaviour predominantly performed by females, and it clearly troubles the clinicians. However, it is also grouped together (in the second study) with the practice of ‘window smashing’. Again, the familiarity of some of the observations jars with this detail. This marks it out, as with McEvedy’s study, as an inpatient phenomenon, and we have glimpsed it in McEvedy’s study, too. The significance of window-smashing is difficult to ascertain: it is often seen as merely an expression of vandalism and also as connected to experiences of confinement. A psychologist at a Durham remand centre in 1981 conducts a study on the smashing of cell windows. He observes that ‘window smashing is predominantly “expressive behaviour”, stemming from boredom and frustration’.⁴⁹ This obviously does not exhaust the significance of this action but it is relevant that this problem (for McEvedy as well as McKerracher and colleagues) is predominantly being studied in people who are confined.

The first study, by McKerracher, Street and Segal, is a general comparison of behavioural problems, with focus on aggression in particular. The appearance of general, non-specific aggressive behaviour in women is seen as less understandable than is that same behaviour in men. They comment that ‘men seldom seemed to indulge in aggression merely for the sense of release obtained from it’: by implication, women did. The aggressive outbursts are tentatively characterised as ‘displacement activity which helps patients to avoid experiencing feelings of anxiety and subjective stress’. They argue explicitly that the ‘aggression of the females, however, seems more emotional than instrumental, and erupts spontaneously whenever they feel angry, tense, anxious or even depressed’. The most striking formulation for contemporary accounts of self-cutting, however, is the following: ‘They seem to experience feelings of internal stress which build up to such a state of tension that violent activity becomes essential’.⁵⁰ However, it must be born in mind that they are talking about all female aggression – to property, to themselves, to staff, verbal threats, threatening suicide or even refusing food. A whole host of behaviours can be reduced to these emotional outbursts. The aggression is theorised in terms largely independent of the confined surroundings, with explanations focused upon ‘a stronger primary drive level of anger’ and ‘lower frustration thresholds’.⁵¹

With different colleagues, McKerracher publishes specifically upon ‘self-mutilation’, focusing upon a group of ‘female psychopaths’.⁵² The authors refer briefly to the results of the previous study with the claim that ‘female patients were significantly more prone than males to mutilate their own bodies and smash hospital property’. They note that these incidents are normally regarded as ‘hysterical’, and they compare two groups: one that ‘indulged in self-mutilation and smashing of windows’ and another, slightly smaller group, that does neither. The authors expand upon this relationship between self-mutilation and window-smashing, observing that ‘[m]any of them had smashed windows for the purpose of self-mutilation though this should not be taken to imply that all window-smashers are necessarily self-mutilators’. They quote a personal communication from a colleague who claims that

in some patients window-smashing is indeed a means of self-mutilation, and in fact these patients often adopt other means to the same end; but there are other patients who regularly smash windows in a way that causes them no injury at all, and their actions can only be regarded as aggressive towards others and not towards themselves.⁵³

The Rampton clinicians read Offer’s and Barglow’s ‘important’ study as claiming that ‘attention-seeking, prestige-gaining and tension reduction were the main goals of self-mutilation’, and that ‘the major dynamic was aggression turned against the self’. As we have seen above, the study from PPI in Chicago is significantly more focused upon the former: the social, institutional and epidemic aspects of the behaviour.⁵⁴ What is striking about the Rampton study of incarcerated patients, ‘who could loosely be termed feeble-minded psychopaths’, is that there are a number of links with current literature, specifically on role of the cutting as ritualistic behaviour and in reducing internal psychic tension.

It is hypothesised that ‘the acts of self-mutilation and window-smashing may have a ceremonial or ritual quality’ that is made habitual by the positive reinforcement of ‘tension reduction’.⁵⁵ It is important to clarify the difference between the anthropologically influenced ideas of ritual cutting practices (such as penile subcision) which are often excluded from contemporary ideas of self-cutting, and a more general description of ritualistic practices, which suggest the establishment of an informal but highly habituated set of actions that a person might perform before and after carrying out the act. It is notable that ritual and tension reduction (prominent in the current literature on self-cutting) are here assumed to play a role that underlies both self-mutilation and window-smashing.

This focus upon individual reinforcement due to tension reduction is a less socially focused way of explaining the behaviour than Offer and Barglow, for example, but this is not the whole story. The Rampton clinicians argue that the supposedly ‘horrifying form’ of the ‘compulsive “acting-out”’ is linked to the ‘restrictions of a security environment’ and the ‘limited range of activity available’. This feeds the ‘[s]uppressed interpersonal aggression occurring in a personality that has low thresholds of boredom and feelings of frustration’. The social environment is considered very important here, but in a way that bears much more explicitly upon individual, psychological needs – unlike McEvedy’s analysis, where the social setting functions more to explain the transmission and imitation of the behaviour. McKerracher and colleagues class the mutilators and window-smashers as more ‘obsessive-compulsive, phobic, and pre-occupied with bodily complaints’. There is no significant engagement with the sociological, epidemiological side of the preceding literature. As noted above, this is part of a more general shift towards personality types, frustration thresholds and psychic tension, and a step away from learning, contagion and imitation. This may be linked to their finding that ‘[s]urprisingly, hysteria was not a discriminating characteristic’,⁵⁶ but whatever the cause, it seems highly significant in retrospect.

The same year, a study on epidemic self-injury by P.C. Matthews is published from an adolescent unit in Plymouth. The focus here is upon the spread and control of the behaviour, and includes a ‘sociogram’ that plots so-called ‘ratings of social power’ between the adolescents to try and make sense of the spread of the behaviour. The inner feelings of these patients are mentioned, but not analysed in any significant sense, as they do not seem to cohere in any logical manner.⁵⁷ The article does not give much space to the inward-looking, inner tension, psychopathology stance – it remains much more focused upon the epidemic, contagious nature of the symptoms, using self-mutilation to shed light on other potentially transferrable or imitative behaviours. The focus is upon management strategies to stop behaviour spreading rather than on investigation into the significance of the mutilation itself. So we can see that the shift described in this chapter is not a straightforward chronological progression from social fields to internal tension, but is a partial and uneven shift. However, as we go from the 1960s into the 1970s, this shift becomes increasingly apparent.

Presuming the social and confounding expectations

In 1970, an article is published from the Maudsley Hospital by J.P. Watson, one which gives us some insight into a doctor’s expectations about self-mutilation motives and how these might be confounded or modified by clinical experience. Crucially for the shift being described here, the expectations concern the social setting, and the clinical results privilege internal emotional states. Watson describes one patient (as opposed to a group) and focuses upon the relationship between the patient and himself (her doctor). The article’s central concern is why the patient might cut herself, and what various interpersonal relationships might have to do with it. In order to measure these relationships, Watson uses a ‘repertory grid technique’, a formalised way of processing interpersonal data regarding social roles and relationships. It is based upon American psychologist George Kelly’s Personal Construct Theory. Various factors or relationships are rated a number of times by the patient according to their strength or significance in the individual’s world-view (or personal construct).⁵⁸

The patient is admitted to an unnamed psychiatric unit because of reported anxiety and depression that leads to her cutting herself ‘with glass or razor blades on her arms and face’. This very visible behaviour is initially ascribed to ‘difficulties with a boy-friend, G., of whom her parents disapproved’. Watson expects this social frame of reference to be an adequate explanation: ‘When I began psychotherapy I thought that disturbed relationships with both parents and the unhappy experience with G. were probably the most important determinants of the patient’s self-mutilant behaviour’.⁵⁹ However, according to the repertory grid,

the elements ‘having the same thoughts in my head for a long time’, and ‘wanting to talk to someone and being unable to’, not the elements concerned with persons, were the situations ranked as most likely to make her cut herself, feel angry and depressed, and think people were unfriendly.⁶⁰

Although G. does feature rather more significantly than Watson expects, Watson’s own presence, and that of her parents are not reported to be significant. Watson initially suggests that this may have to do with denial – given that the grid is a self-report technique – then concludes that this is ‘a complex matter, but I think it likely that the “person” elements seemed to her less likely to upset her and make her cut herself than the “talking” and “thoughts” elements’.⁶¹ Even though the article is based upon an individual, there is still much scope for relational, interpersonal aetiology (along the lines of self-poisoning explanations). However, this is rejected, seemingly on the basis of the patient’s own reported statements. This shift in explanation becomes more and more established as the 1970s progresses.

In the same year, a dissertation for an MSc in clinical psychology at the IoP, entitled ‘Self-Mutilation’, is completed by psychologist Anthea Keller. Self-cutting is seen to have two possible causes, which possibly interact – internal and the external. Keller recognises that the understanding of self-cutting as internally focused tension relief is prominent in the literature, and she references three articles from North America published in the late 1960s. All these references feature relatively regularly in the current literature on self-cutting, at least until the early years of the twenty-first century.⁶² This might also be seen as the beginning of the explicit (referenced) influence of the North American psychoanalytic, internalist studies of the late 1960s on studies carried out in Britain – with previous influences limited to the sociologically minded Offer and Barglow.

The role of social or outward-looking factors is seen to be more difficult to isolate or pin down with any precision, according to Keller, due in part to difficulties patients have in talking about it. In a similar way, Keller divides incidents into two classes: ‘group’ and ‘individual’ cuttings. The former concerns the sociologically influenced literature considered above, such as the studies at PPI by Offer and Barglow and Strauss et al. Keller claims that, apart from a higher proportion of men in some studies, there are no substantial differences between the two groups. She then argues that any ‘group’ cuttings only occur when individuals already have a predisposition to the behaviour.⁶³ In this way, under the veneer of parity, the individual cuttings are in fact made more significant, being the root cause of any group cuttings that may occur. This echoes McEvedy’s rooting of the behaviour in ‘spontaneous’ cutters who then influence ‘susceptible’ ones. For reasons both practical and theoretical, the group is secondary here to the individual inclination.

Despite this, some recognisably ‘social’ or relational factors are broached. Keller mentions that visits of parents and setbacks during therapy have been seen as significant in the aetiology of cutting. However, it is also claimed that virtually every published investigator of self-mutilation emphasises the role of building tension (which may not have any obvious reason behind it) that then overwhelms patients and causes them to try to reduce the tension by self-cutting or smashing windows.⁶⁴ Patients might cut when alone or feeling lonely, but also, confusingly, when in the presence of an important person. It is unclear to Keller why cutting happens in the latter scenario, given that solitary, affective relief is the dominant explanatory frame here. Window-smashing has not entirely retreated from consideration, but it is clear that the sociological, group-epidemic focus is fading, being replaced by a model of internal affective regulation – something that corresponds quite closely to today’s understandings of self-injury.

Brian Ballinger publishes a study on self-mutilation in 1971, comparing two populations from Dundee: one group from Strathmartine Hospital for the ‘mentally subnormal’, and the Royal Dundee Liff Hospital, a psychiatric inpatient institution. Right away, Ballinger makes clear that he is not talking about window-breaking, restricting the study to acts that are ‘painful or destructive ... committed by the patient against his own body’, and excluding‘[a]ccidents, tearing clothes, window-breaking, swallowing dirt and refusal of food’.⁶⁵ Frustrated outbursts are here only included if they involve damage or pain (in the assessment of the staff) to the patients’ bodies. There is a sizeable literature, which has been mentioned, around repetitive self-damage performed by those with severe learning difficulties; Ballinger’s study is an explicit attempt to compare two recognised categories of self-injury: that of the ‘mentally subnormal’ and that of the ‘mentally ill’. These categories remain very separate today, with little attention on the former.

The methods of injury are seen to differ, but with significant overlap. Patients in the ‘subnormality hospital’ are reported to self-injure by ‘picking, striking, scratching, banging, biting, pulling hair out and rubbing’. The psychiatric patients, on the other hand, injure themselves by ‘scratching, picking, striking, rubbing, cutting and tying string round fingers’. There is no sense here that self-cutting is an archetypal form of injury. Self-injury is seen as more prevalent in subnormality hospitals than psychiatric hospitals, with 15% of patients in the former institution engaging in self-injury, compared with only 3% of the latter. It is seen as related to the social setting: ‘environmental restriction, boredom and frustration played a part in worsening self-injury in many patients’.⁶⁶ Here again it is not a smooth (teleological) progression from social explanations to internal ones, but this article is useful in showing how the concept of self-injury encompasses a number of distinct, but overlapping, inpatient populations, and becomes ever more visible – and differentiated – throughout the 1970s. Psychiatric self-injury is a clear, definable object here.

North American influence and the triumph of internal tension

One of the MPhil dissertations submitted at Institute of Psychiatry in 1972 is entitled ‘Wrist-cutting: a Psychiatric Enquiry’. Little is known of the author, Samuel Stuart Anthony Waldenberg and, as far as I can make out, the research does not form the basis for research articles in psychiatric or medical journals. The study sits on the ‘thesis’ shelves in a newly refurbished section of the IoP, amongst the dissertations of some of the luminaries of British psychiatry, including Sidney Crown, Michael Rutter and Murray Parkes, as well as Neil Kessel and Norman Kreitman (and Colin McEvedy). Despite Waldenberg’s relative obscurity, the study is referenced by a number of subsequent published texts on self-cutting.⁶⁷ The dissertation contains a sheet showing when it is signed out of the IoP, and by whom. Names on the sheet include Alec Roy, who publishes on self-mutilation later in the 1970s (see below), and then on depression, suicide and schizophrenia, and Dinesh Bhugra, currently professor of mental health and diversity at the IoP, and author of – among many other books – the Maudsley Monograph, Culture and Self Harm: Attempted Suicide in South Asians in London (1994).⁶⁸ As with McEvedy’s work, I have been unable to obtain the necessary permissions from Waldenberg to quote from this unpublished thesis, so I shall paraphrase throughout.

Waldenberg’s method is similar to McEvedy’s, built around an ‘accident book’ at the Joint Royal Bethlem and Maudsey Hospitals, in which a record is kept of all the injuries to inpatients that come to the attention of the staff. He notes that a similar book exists at St Francis Hospital.⁶⁹ From these two sources, a group of self-injuring patients is selected for study, with patients interviewed soon after the incident. Their responses to this semi-structured interview are compared with a control group of non-cutting inpatients.

There is an effort to emphasise the cutting over and above a constellation of symptoms. He lists various behaviours occurring in this sample, including truanting, delinquency, the taking of illicit drugs, and supposed sexual deviance: lesbianism, promiscuity and incest. As for more directly and physically harming behaviours, he notes that these patients take overdoses (mostly with no suicidal intent) and engage in self-cutting, window-smashing, self-burning and self-scalding.⁷⁰ It is notable, given the content of previous chapters, that these patients often take overdoses without suicidal intent, but these are not investigated (a point developed below). Window-smashing is still considered an issue (although admittedly minor), as are other supposedly deviant behaviours. It is important that self-cutting is not self-evidently or obviously the behaviour at the centre of these patients’ pathologies: it is made central by the emphases of professional observers. (The same processes of exclusion and emphasis operate in the North American literature.⁷¹)

The key finding of his study, according to Waldenberg, is that internal, emotional gain experienced as a result of cutting is seen to trump any kind of external, social gratification.⁷² Explicitly then, this dissertation constitutes an argument against the socially embedded self-harm analysed throughout this book. It is influenced by a number of North American studies on self-cutting that promote internal, psychological, emotional needs as the roots of self-cutting, especially tension release – studies that continue to influence current models of self-harm. He refers many times to North American studies by Pao (1969), Crabtree (1967), Graff and Mallin (1967) and Grunebaum and Klerman (1969). Indeed, he also mentions Offer’s and Barglow’s sociologically influenced study of 1960 which, he admits, focuses more upon the role of imitation in epidemics of self-cutting. He concedes that psychoanalytic authors such as Pao, Crabtree and Graff and Mallin do not give the role of imitation much consideration.⁷³ Imitation implies a social field, and the idea that the point (underdeveloped in the American psychoanalytic studies) shows how a division is opening up between the internal and external ideas of causation. This division has the potential to separate any self-cutting that might present at A&E departments from the overwhelming mass of socially embedded and understood self-poisoners with whom they are combined in the 1970s analyses of those such as Hugh Gethin Morgan. However, Waldenberg does not make this split according to method. He calls the group ‘cutters’ but argues that some of these patients can distinguish between the feelings that precede a frankly suicidal overdose and those that precede cutting and/or a less serious overdose.⁷⁴ This equates cutting and trivial overdosing and implies that they are prompted by the same state of mind. Thus, the strong differentiation between cutting and overdosing does not seem to stem from here. However, the emphasis on cutting and on internal motivations – explicitly against sociological or epidemic ones – is highly significant. Self-cutting is cast as internally rather than externally motivated, but this internal motivation is also ascribed to trivial overdoses.

Ping-Nie Pao’s study of ‘delicate cutters’ from Chestnut Lodge, Maryland, is praised for the clarity of its descriptions, especially the patients’ subjective experience of cutting. Waldenberg quotes Pao’s account, which uses the words tense, tension, and tenseness in a single sentence. As well as this internal emotional state, Waldenberg does acknowledge the social setting, mentioning interruptions in interpersonal relationships as possible factors that might precipitate cutting.⁷⁵ His literature review is ambivalent about the internal/external divide. He writes that others have noted the relationship between an episode of cutting and interpersonal disturbances, such as the end of visits; others’ works might start with a view of cutting as a purely internally focused activity, but then come to see it as a communication between patient and therapist.

These social motivations exist in tandem with acknowledgement of patients experiencing a painful sense of unreality, or of having no feelings at all, which prompt the patient to cut to try and relieve them. Similarly, the review mentions the dual desire on the part of the patient to punish the parents, but also to obtain their help and support (thereby to communicate with them), alongside rather frank statements that cutting is performed in order to relieve tension and to alleviate feelings of numbness or deadness.⁷⁶

This initial ambiguity about the internal nature of self-injury is disciplined by copious and repetitive intellectual labour (of which Waldenberg is significantly aware) so that the clinical data conforms to his expectations. It is the clearest sign yet that a battle is being fought to de-couple self-cutting from socially focused, communicative action. This is partially fuelled by the belief of the psychiatrists (against other medical and nursing staff) that the behaviour is meaningful above and beyond simple attention-seeking or ‘acting out’. Waldenberg notes that the staff (both nurses and doctors) often react negatively to these patients and label them as manipulative or attention-seeking. He downplays this angle, reasoning that there are numerous ways of seeking attention, yet these patients choose a method that – to him – is extremely unusual, even bizarre.⁷⁷ This particular point is not an aetiological argument formed from psychoanalytical inclinations (like much of the American literature): it has a much more mundane, everyday conflict at its heart – a conflict between those who see some psychiatric patients as manipulative (and therefore communicative) timewasters, and those who see another order of significance in their behaviour. He argues that, because most patients who cut do so whilst alone, the cutting therefore must serve internal needs, rather than communicative ones. This is a clear intervention against certain reactions to the behaviour and also seems to preclude – for Waldenberg– any attempt to link the behaviour to communicative overdoses. However, his reasoning does not quite hold, as communicative overdoses would also – in the majority of cases – be performed alone and later discovered, much like the cutting incidents. However, the internal tension-fuelled motivation becomes a powerful counter of legitimacy for the discrete nature of the behaviour pattern, as well as countering perceived negativity from other staff members. Indeed, when recounting all the other deviant behaviours, from truancy to incest to overdosing, Waldenberg relates the difficulty he has in isolating a single psychological motive or explanation for these various behaviours.⁷⁸ Thus he chooses to narrow the focus to explaining just one – the self-cutting – rather than taking a more general sociological, deviance-based approach that might attempt to make sense of all the behaviours as a group.

He acknowledges that patients vary in the reasons they give for their cutting, but that they normally allude to tension in some form. He further notes that despite the varying responses, tension and anger (both with oneself and others) feature prominently.⁷⁹ There is some ambivalence here: he acknowledges variation in explanations, and that interpersonal anger is a factor. He continually oscillates between awareness of social factors and emphasis on internal ones; he consistently promotes the internal rather than interpersonal causes. He addresses the issue of communication directly, acknowledging that even though any intent to send a message might be denied by patients, they also have obvious expectations about the reaction of staff to their behaviour. However, his final judgement is stubbornly internally focused. He claims that even those with such ‘social’ expectations report relief and satisfaction from the sight of their own blood, the experience of which outstrips any pleasure they might get from a doctor’s reaction.⁸⁰

Waldenberg discusses the views that his control group of ‘non-cutters’ have on the subject of cutting, which is reported as grudging approval at the discipline or ‘nerve’ required to cut oneself. This approval is presented by Waldenberg as evidence, but not for the way in which cutting becomes a socially acceptable, valorised and aspirational pattern of behaviour. Instead, it is deployed as evidence of the internal needs serviced by cutting – the control patients are presumed to have slightly less-powerful urges. He also claims that pleasure from bleeding is a ‘simpler’ explanation than anticipation of the therapist’s reaction – a clearly loaded assessment.⁸¹

He does not deny that his group of ‘wrist-cutters’ receive much attention after cutting, and that they may indeed derive satisfaction from this attention, but he calls the gains from the act of cutting, itself, as ‘primary’, and that they outweigh the secondary, interpersonal effects. These primary gains are internal, emotional, and heavily psychoanalytic.⁸² The motives of these patients are clearly multifaceted (and this is acknowledged), but the consistent emphasis is on the internal, affective regulation of cutting or of seeing blood. The number of times this oscillation is played out indicates how hard Waldenberg has to push against the socially embedded analyses of self-harm, especially in institutions. There is considerable room for a mixture of both causes, but one is emphasised, and he reproduces the conclusions of the influential American literature.

On a practical, methodological level, Waldenberg is in part aware of how the research methods employed might influence the findings in this way. He explicitly admits that whilst the patients’ feelings and thoughts that immediately precede the cutting episode are subject to close examination and questioning, less attention is afforded to possible motivating factors in the patient’s social circle. Such methodological candour and awareness is striking and shows how the focus of the dissertation, influenced by the North American literature and the local staff conflicts, emphasises the internal, psychic motivations over the social setting. It is acknowledged that an interviewer’s questions can be ‘leading’ and that the information sought by the questioner is often furnished by the interviewee.⁸³ However, at the end of all these oscillations, Waldenberg’s judgement call specifically emphasises the internal over the external.

Angela and Alan Gardner publish a study in 1975 from Long Grove Hospital in Epsom and the London Hospital in Whitechapel. They investigate a group of 22 female inpatients (8 from a psychiatric ward of the London Hospital, and 14 from Long Grove, a traditional mental hospital), who are admitted over the course of one year, from July 1972. They use the Middlesex Hospital Questionnaire and the Obsessive-Compulsive section of the Tavistock Inventory: both are psychiatric rating scales. The former is developed by Arthur Crisp and Sidney Crown during the mid-1960s as a rapid, self-report diagnostic tool for neurotic patients. The latter has its roots in the psychoanalytically oriented Tavistock Clinic.⁸⁴ These 22 patients are compared with a control group.

Gardner and Gardner claim that self-mutilation has been around for centuries, but only recently brought into focus by Offer and Barglow, whose work has ‘focussed interest on the patients, usually female, who repeatedly cut their wrists’. They acknowledge that ‘since then ‘a number of reports have appeared, mainly from the U.S.A’ and they are rather dismissive of British literature, stating that there have been ‘[o]nly three studies of consequence’: McEvedy (1963), McKerracher et al. (1967) and Waldenberg (1972).⁸⁵ The North American provenance of most of the analysis of self-cutting is again implied.

Gardner and Gardner argue that both Offer’s and Barglow’s (1960) and McEvedy’s (1963) studies show that ‘repeated self-cutting appears to have an “infectious” quality, leading to outbreaks involving several patients. This suggests that factors in the ward milieu play their part’. They address this by selecting matched controls from the same wards as the cutters. However, they state explicitly that when interviewing the patients, ‘[s]pecial attention was paid to the patient’s mental state during self-cutting’. This again shows the emphasis on an internal, psychological perspective over a socially focused enquiry. They report that the ‘initially private nature of the act is well emphasized’, which contains both an emphasis on being alone, but a concurrent awareness that the consequences of the cutting may be displayed later. Key is their contention that ‘[b]y far the commonest experience leading to self-cutting was the onset of an unpleasant feeling of tension, this increased in intensity until the patient cut her skin, which brought an immediate lessening of tension and a feeling of relief’. It is noted that this might have to do with the patient’s social circle, but this is downplayed: ‘Sometimes the feeling of tension was related to angry feelings towards self or others, but more often than not there was no apparent precipitating factor’.⁸⁶

Despite this partial acknowledgement of the social setting, their entire therapeutic strategy is based around feelings of tension. They claim that for any treatment that attempts to halt self-cutting behaviour, it is ‘logical to seek some other superior tension-relieving reward’. Gardner and Gardner do acknowledge that tension relief and communication might exist in the same action, as they characterise another article’s findings ‘regard[ing] the self-cutters’ method of tension relief as a preverbal message’. They fail to establish any secure differences between cutters and controls, but claim that ‘it remains possible, even probable, that differences do exist but are found perhaps in the quality of child/parent relationships and other areas difficult to assess with certainty in retrospect’.⁸⁷ Thus, they remain committed to the psychological discreteness of this population of ‘self-mutilators’. This article again attempts to differentiate self-cutters from other kinds of psychiatric inpatient, on psychological grounds that are increasingly tension-focused.

Stability and comparison with self-poisoning

One of the final steps in this process that isolates ‘self-cutters’ as a distinctive object of psychiatric research and treatment in Britain (especially as the study of self-poisoning is so well established) is to compare these ‘self-cutters’ explicitly with a population of self-poisoners. This is done in 1975 by Michael A. Simpson, a clinician who publishes on the topic of medical education and later on borderline personality disorder (BPD), and who trains at Guy’s Hospital in London. He conducts an interview survey on 24 self-cutting patients brought to his attention by a ‘Psychiatric Emergency Services Unit, dealing with all requests for a psychiatric opinion from the general medical and surgical wards and also a busy Emergency Room’ based at Guy’s. He also produces the first comprehensive literature review on self-mutilation, which published the following year in a collection edited by the eminent North American suicidologist Edwin Schneidman.⁸⁸

Simpson estimates the lethality of all the self-damaging acts he includes in his clinical study and concludes that all wrist-cutters fall into the bracket of lowest lethality, whilst the self-poisoners are more variable . What is striking about Simpson’s ratings – at least for those familiar with the studies of self-poisoning produced during the 1960s and early 1970s – is that Simpson’s ratings do not appear to include any assessment of the social setting. In contrast to Stengel’s assessments (Chapter 2), which adjusted lethality ratings according to precautions taken to avoid or ensure discovery, Simpson’s focus is decidedly on the mental state and experience of the patient, rather than on the social environment. As Simpson puts it, patients were ‘interviewed with special reference to the phenomenology of the act of cutting’. (In this sense, ‘phenomenology’ indicates a focus on the subjective experience.) Accordingly, the entire enterprise is based upon patients’ self-report of their feelings and motivations. They are ‘asked to state their first and second most serious or troubling symptoms. Of the cutters, nine complained of depression as the first or second most serious symptom, twenty-one complained predominantly of “emptiness”, and eighteen of tension’.⁸⁹ These statements match up with some of the previous studies, but what is most interesting here is how far the kinds of questions asked correspond to the quality of the answers. Simpson makes no mention of the social setting, patient relationships, or possible communication.

Thus, he reports that ‘[t]he non-cutters complained primarily of depression, each included it as one of the two principal symptoms, and nine cited tension as their second most-troublesome complaint’. It is important to remember that the people designated as ‘non-cutters’ are in fact self-poisoning patients. Simpson has little time for psychiatric diagnoses, arguing that it is ‘not helpful with regard to wrist-cutters and they are best regarded as a separate category in planning management’. This shows the (still relevant) ancestry of the behaviour pattern as a sociologically influenced management problem. He reports: ‘Nine of the present series of cutters absconded from hospital on numerous occasions, a pattern of behaviour which was not seen in any of the non-cutters’.⁹⁰ Again, this is a management issue after patients are admitted to psychiatric wards. As much as he focuses upon the subjective experience of the patients, there are many echoes of the social field and of sociological studies that focus upon the particularly intractable management problems presented by these patients, problems that are increasingly reduced to ‘cutting’.

Despite these management issues, Simpson becomes increasingly confident that the behaviour is essentially a response to tension, even if also learned or contagious. He argues that ‘[t]his form of a response to tension can be learned and propagated in a hospital or institution and is often sustained by the widespread conflict and guilt such acts tend to arouse in the staff’. Any focus on the social or administrative setting is secondary to the essence of the behaviour, which is characterised as a ‘response to tension’. The focus on internal, emotional states is combined with a desire for solitude, leaving any wider social or communicative significance out of the reckoning: ‘The patient feels depressed, angry and tense, and wants to express the extent of her feelings, but feels unable to do so in words. Tension becomes the predominant affect ... she will seek solitude if she is not already alone’.⁹¹ Simpson’s study illustrates the lack of concern for the social setting with his assessments of lethality, the focus upon the subjective experience (phenomenology) of the cutting, and it shows the management issues that persist in assessments of self-cutting. Again, the focus is upon individual, intolerable tension as motivating the cutting incidents.

Simpson’s literature review of self-mutilation, published in a collection about suicide, reports a ‘very clear composite picture of the typical cutter’ as being a young, attractive, intelligent woman. He mentions Offer’s and Barglow’s analysis, which ranges from the interpersonal and social settings to internal motivations. According to Simpson, they propose ‘several motives such as gaining attention, the need to be loved and cared for, attempts to control aggression, tension reduction, and gaining prestige among the social group in the ward’. Such an explicit mention of the social setting merits significant disagreement: Simpson claims that ‘[e]lements of such motivations may well play a part in the dynamics of self-mutilation, but they are inadequate explanations – Why choose to gain attention or express the need for love by cutting one’s wrist?’⁹² Note that it is the social, communicative motivations singled out for their inadequacy, rather than (for example) the observations about aggression or tension.

However, he also mentions that many authors have focused upon issues of loss and abandonment as precipitants for cutting incidents. The social focus of this behaviour comes through most clearly in a passage where in Simpson discusses how ‘cutting behavior can be learned and propagated in a hospital, clinic or institution’ and how patients may compete for the title of ‘chief cutter’ through the number of stitches that they have received. In a startling description of the social significance of these acts, he writes: ‘While patients may claim afterwards that they do not want others to know of their act, they often manage to flaunt the wound or their bandage like a newly engaged girl wearing her diamond ring for the first time’. There is clear ambivalence here, but he returns again to internal psychopathology in a passage referenced as central a decade later by influential scholar of self-mutilation, Armando Favazza. Simpson argues that ‘self-mutilators commit what amounts to anti-suicide, employing the wrist-cutting as a means of gaining reintegration, repersonalisation, and an emphatic return to reality and life from the state of dead unreality’. This reintegration (drawing on the work of Karl Menninger) is conceived of as far more psychic than social. He claims that ‘there exists a clearly identifiable condition of self-mutilation, usually involving wrist-cutting, which exhibits much of the stability of a syndrome’. He also asserts that ‘[w]hile self-mutilators represent a significant problem group within the territory of suicide and para-suicide, they can be clearly distinguished from other similar presentations with significantly higher lethality, and thus warrant different treatment’.⁹³ The comparison with parasuicide shows how cutting and poisoning are increasingly seen as different phenomena.

Alec Roy’s 1978 study from the Maudsley compares 20 consecutively admitted self-mutilating inpatients with a control group and explicitly attempts to rectify the failure of Gardner and Gardner (1975) to establish difference between cutters and controls. He finds that nine self-mutilators reported anger at themselves as their predominant reason for cutting, whilst seven cited the relief of tension. He is unsure about the tension argument (even though Gardner and Gardner cite it as central) because ‘the non-current cutter groups [those who had not cut within the 14 days preceding the interview] had anxiety and depressive symptoms [, so] other variables may be important’. In formulating a general statement about self-mutilation, Roy considers ‘intrapsychic, personality, interpersonal and psychosocial factors’. He expands on this, hypothesising that ‘[t]heir hostility, introversion and neuroticism may lead to anger and depression at their difficulties in forming and maintaining relationships and to the initiation and maintenance of this behaviour’. This roots the behaviour in the intrapsychic and personality realms, and makes the interpersonal and psychosocial distinctly secondary.⁹⁴

Today, Keith Hawton is perhaps the best-known psychiatrist working on attempted suicide and self-harm in Britain. He is instrumental in establishing the Oxford Monitoring System for Attempted Suicide in 1976, and has written numerous papers on deliberate self-harm in a variety of ways, in a variety of settings – including studies at A&E departments and retrospective questionnaires in schools.⁹⁵ Research that he publishes in 1978, ‘Deliberate self-poisoning and self-injury in the psychiatric hospital’, is first mentioned in print in 1975, when clinicians from the Grayling well Hospital in Chichester refer to a paper given by Hawton in 1974. Hawton’s figures are reported as referring to ‘attempted suicide’ in inpatients and day patients.⁹⁶ Hawton writes to the British Medical Journal in order to clear up possible confusion arising from such a citation. He states of his study of psychiatric inpatients: ‘I would in any case be loath to use the term “attempted suicide” to describe the majority of these acts since many involved, for example, minimal cutting of the skin’.⁹⁷ Hawton is well aware that attempted suicide does not mean a genuine attempt at death in this context, but he is still strongly against calling ‘minimal cutting of the skin’ by the same name as a communicative act of self-harm that seeks help from the environment through the symbolism of suicide. For him, there is something different occurring, and it is no coincidence that this study is based in an inpatient institution – the Warneford Hospital in Oxford.

Hawton’s study forms a bridge between the profile and description of self-cutting that emerges from psychiatric inpatient facilities, and the studies that include self-cutting as a minority behaviour in self-poisoning-dominated studies. It is published in 1978 and, throughout, it compares the inpatient data with the literature focusing on A&E studies (predominantly concerned with self-poisoners). The inpatient behaviours (named ‘self-injury’) are then differentiated from self-poisoning in terms of psychological motivation that are familiar from a twenty-first-century standpoint: ‘the motivational factors leading to self-injury may be different from those underlying self-poisoning behaviour in the community ... self-cutting is often used as a method of tension reduction and may be associated with states of altered awareness. Although self-poisoning may have a similar effect by temporarily interrupting consciousness, clearly the act is qualitatively very different’.⁹⁸

What remains implicit in Simpson’s choice of a control group of self-poisoners is made explicit by Hawton, who is able to expand upon the qualitative differences between self-cutting and self-poisoning – differences that are still included together without much comment in A&E-based studies. Hawton goes on to state that one patient reported that the difference is between feelings of tension (cutting) and feelings of hopelessness (overdosing): ‘[S]he cut herself in response to feeling extremely tense, and took an overdose when she felt depressed and hopeless’.⁹⁹ This maps quite precisely onto one of Jan Sutton’s ‘respondents’ who is quoted as making the following differentiation nearly 30 years later:

There was always a clear distinction for me between the cutting and the overdosing. The cutting was far more frequent and was about survival, about coping with the intolerable feelings I was carrying inside. Overdosing meanwhile was about giving up for good.¹⁰⁰

But this again should not simply be viewed as a smooth progression with all researchers singing from the same song sheet. In 1979, the second edition of Uncommon Psychiatric Syndromes is published by David Enoch and W.H. Trethowan. Buried in the entry on Munchausen syndrome and related disorders (which involves the chronic fabrication or induction of illness in order to receive medical attention) is an intriguing passage and case study of one 21-year-old female. People who self-mutilate are said to ‘scarify themselves with pieces of glass or metal, or indulge in parasuicidal wrist-cutting attempts’. This is said to indicate similarities with Munchausen patients in its ‘tendency towards self-inflicted disability, together with a marked degree of attention-seeking behaviour and, perhaps, an unusual tolerance of pain and discomfort’. When the patient is questioned, she says, ‘I sometimes feel I have to let the poison out that is in me!’, and she also ‘admitted to being very angry with herself and to feeling as if she were sitting on a volcano’. These are heavily internally focused motivations, and Enoch and Trethowan do mention that her behaviour ‘undoubtedly reveals much of her basic emotional difficulties’. However, they also argue that ‘such self-destructive behaviour ... must be seen, if it is to be understood at all, as a method of communication – a cry for help as well as for attention’.¹⁰¹ There is nothing inevitably internally focussed about self-cutting behaviour, and it can be interpreted either as inward- or outward-looking. Once again, the meaning of behaviour is highly contingent.

Hugh Gethin Morgan’s book Death Wishes? published in 1980, is based upon extensive study in Bristol. It shows how the assessments of hospital presentations of self-harm are changing. It is already noted how his studies during the early–mid 1970s use the term ‘deliberate self-harm’ to describe all methods of self-harm. Despite this terminological discussion, as we see at the beginning of this chapter, lacerations are not thought to be meaningfully different behaviour. All this changes by 1979. Here, Morgan argues for strong differentiation between self-poisoning and self-cutting, even though both actions are considered to have only an ambiguous relation towards self-accomplished death. He claims:

In looking for causes of DSH [deliberate self-harm] it is important to consider self-laceration separately in order to discern psychopathological mechanisms which may be peculiar to it and which are not shared by those who take drug overdosage.¹⁰²

Morgan sees self-laceration as concerned with an altered state of consciousness, a need to obtain relief from tension and a high incidence of obsessional, phobic and narcissistic tendencies. Immediately after this discussion, as if to restore a sense of balance, he states that ‘DSH cannot be understood entirely in terms of intrapsychic pathology. There is a massive body of evidence testifying to its close relationship with interpersonal social events, and not merely as a blind reaction to them’.¹⁰⁴ Remembering that, for Morgan, DSH refers to both self-cutting and self-poisoning, it is clear that the behaviours are still linked, even if self-cutting requires a level of differentiation and discrete concern.

In 1982, the first edition of Keith Hawton’s and José Catalán’s Attempted Suicide is published. It contains a special chapter entitled ‘Self-injury’,¹⁰⁵ which details a tripartite division between: ‘superficial self-cutting’, which is ‘usually of the wrist or forearm, associated with little or no suicidal intent’; this is followed by serious self-injury, which involves deep cuts that endanger blood vessels or tendons, as well as shooting, hanging and jumping from buildings which, as they note, ‘are usually associated with serious suicidal intent’. Finally, there is the category of self-mutilation that ‘may result in disfigurement’ and is associated with psychosis; it may or may not be life-threatening. It is obvious that much nosological effort has been expended here. Differences are minutely examined and categorised in multiple ways. The authors mention that in Oxford, ‘particular care has been taken to try to identify all cases of self-injury coming to the general hospital, irrespective of whether they have been referred to the hospital psychiatric service’.¹⁰⁶

Hawton and Catalán note that wrist cutting has been treated as a distinct syndrome (referencing the North American literature), but they are unconvinced, adding that it is ‘doubtful whether this is a useful approach to the problem, especially for clinical purposes’. They rehearse the now-familiar picture that ‘[t]he predominant sensation is one of tension, which steadily mounts until it becomes unbearable ... Immediately before cutting, a sense of numbness or emptiness may be described’. Crucially, this differentiates the behaviour from self-poisoning. They argue that ‘clinical teams which manage attempted suicide patients should be familiar with the special problem of patients who deliberately injure themselves, and not just deal with them as if the behaviour was the same as self-poisoning’.¹⁰⁷ This is exceptionally clear. In the second edition of their guide, they further note: ‘Wrist-cutting, which is predominantly a behaviour of younger patients, is often repeated and in many cases appears to be a different phenomenon in psychopathological terms from self-poisoning’.¹⁰⁸ We are– so to speak – arrived at the present. And this context is that of the 1980s, where the relationship between the state and social life is being radically reimagined (rolled back), and where neo-liberal ideas of self-reliance and independence are dominant (see Conclusion).

The difference between inpatient and outpatient objects of self-harm

Having demonstrated that self-cutting emerges in certain (inpatient) places in British psychiatry and then is able to migrate and to transform analyses in other (general hospital) arenas, the final task is to ask why the inpatient and A&E objects are so different, despite people lacerating themselves presenting at A&E and many self-cutters also having taken overdoses. A significant part of the answer is to be found in the ways in which different therapeutic environments bring different behaviours to prominence. As noted above, self-cutting behaviour rarely becomes the object of intensive psychiatric scrutiny outside of inpatient institutions. Cutting only becomes scrutinised when inside the high-surveillance environment of a psychiatric inpatient ward.

When describing some implications of their Bristol study in 1975, Morgan et al. found that those patients who fell into the ‘not interviewed’ category were ‘more likely to have lacerated themselves’.¹⁰⁹ In 1977 Norman Kreitman observes that ‘there is little doubt that self-injury is under-represented’ in the Edinburgh statistics.¹¹⁰ Richard Turner and Hugh Gethin Morgan note in 1979 that casualty-department-based samples cannot be regarded as representative of all self-harmers, because it has been shown that ‘20% of all those who present to Accident and Emergency Departments were discharged home without being admitted to hospital, and these [so discharged] were younger and more likely to have lacerated themselves than those admitted to medical wards’.¹¹¹ The method of self-harm has practical consequences, as one escapes psychiatric scrutiny with greater regularity at a general hospital if cutting rather than poisoning.

Similarly, Hawton notes of his 1978 inpatient study that ‘patients with minor scratches and cuts reported in this study might not have been referred to the general hospital and thereby identified if they had done this in the community’.¹¹² At A&E, however, Morgan reports that self-laceration might ‘appear trivial when seen in hospital Accident and Emergency Departments’.¹¹³ Conversely, self-poisoning figures regularly in the symptomatology of ‘self-cutters’ but is rarely emphasised. Waldenberg notes that many of his group of ‘wrist-cutters’ also took overdoses without intending to die.¹¹⁴ Gardner and Gardner relate: ‘We also had the impression that individual cutters took overdoses of drugs more often than the controls, but the actual number of cutters who had taken one or more overdoses was not significantly different from the control group’.¹¹⁵ It is clear that the kinds of behaviour that come under psychiatric scrutiny in psychiatric hospitals and in community studies are very different.

The information that constitutes the inpatient studies relies heavily upon the levels of psychological and biographical scrutiny that only the inpatient setting can provide. McEvedy’s dissertation provides a good example of this. The appendix of case histories he provides for each of his ‘cutters’ reveals the initial reason for admission of these patients. These reasons for admission include: ‘hysterical paresis of the right leg’, a suicidal attempt involving barbiturates, ‘abnormal eating habits and loss of weight’, ‘difficulty getting along with people and depression’, a referral from an Approved School because of depression and an incident involving severely ‘slashed her wrists [and] an overdose of aspirin’. Another patient is referred to the Maudsley’s forensic unit on a charge of shoplifting and was considered to be in need of admission because of an ‘apparently sincere attempt at suicide by taking an overdose of tablets’. Another patient is admitted due to supposed temper tantrums and spiteful behaviour towards other children. The only patient admitted for self-cutting is Kay R., who came to be at St Francis because of the severely slashed wrists and an aspirin overdose – but the mention of the severity seems to preclude the kind of cutting in which McEvedy is interested.

However, once these people are inpatients, other behaviours are discovered retrospectively: Penelope E. is described as pulling the emergency cord on a train then presenting herself to train staff with cuts and scratches, which the police think self-inflicted. This case is apparently not referred for any kind of psychological attention. Similarly, after Kay R.’s admission doctors learn of a past surgical procedure to remove a needle fragment from her leg – allegedly the result of a fall over her sewing basket – an explanation the staff considers ‘extremely unlikely’ in retrospect. Whilst in the hospital she puts scissors in her mouth in a way that ‘alarmed the nurses’ and is referred to the psychiatrist, but does not end up seeing one.¹¹⁶ This all points to how rarely certain self-damaging behaviours come under psychiatric scrutiny if performed outside of psychiatric inpatient settings.

J.B. Watson’s patient is admitted for cutting – carried out in a particularly visible way. As Watson reports, the patient ‘became anxious and depressed and began to cut herself with glass or razor blades on her arms and face. She was admitted to a psychiatric unit’.¹¹⁷ It seems fair to assume that cuts on the face are much more noticeable (and perhaps more alarming) than those easily concealed on arms or legs. It is evident that very few self-cutters of the current literature cut themselves on the face, and facial self-mutilation is considered to be rather different to the kind of self-harm discussed here.¹¹⁸ In 1972 Waldenberg is specific that only one of his patients is admitted due to self-cutting behaviour.¹¹⁹ Again, this shows how behaviours such as cutting are much more likely to come to light once a patient is inside an inpatient institution, with all the opportunities for scrutiny (and perhaps desire for resistance of prescribed routines) that it entails. This could explain why this behaviour does not at first figure so prominently outside of these institutions. The environment is key in bringing to light certain forms of behaviour.

This may be (in part) because doctors do not believe cutting one’s wrists to be a particularly dangerous act, in the sense that it rarely endangers life. McEvedy states that it is ‘unlikely that death will result from a slash of the wrists’ even though popular opinion ‘continues to hold the belief that an injury [to the wrists] will prove rapidly fatal’.¹²⁰ In 1975, Simpson transforms this clinical view into numbers, as we have seen: ‘[A]ll acts of wrist-cutting were estimated at a lethality of 4 [lowest lethality]. There was a wider scatter of lethality scores for the self-poisoners, and an average score of 3.4’.¹²¹ Incidentally, this is something also noted by the North American studies of self-cutting. One influential study observes that wrist cutting ‘is an unusually difficult way to draw large amounts of blood’, whilst another claims that ‘ wrist slashing’ is ‘a notoriously poor method of suicide’.¹²² This means that people presenting at hospital are less likely to be admitted, and that cuts on the arms (but not the wrists) are unlikely to be discovered, let alone be the cause of a trip to A&E. However, once a clinical object is established in inpatient facilities, it can travel and become a psychologically distinct category, into which A&E patients might fall.

Concluding thoughts

Self-cutting emerges as an epidemic phenomenon and a management problem in psychiatric inpatient institutions, and it shifts from these sociologically informed perspectives towards an approach more focused upon an internal psychopathology which involves intolerable psychic tension. After this has become stable, it migrates to A&E departments, and informs analyses of the small numbers of self-cutters who present there. Thus, self-cutting and self-poisoning, treated as largely similarly motivated in A&E studies in the early 1970s, are strongly differentiated by the end of that decade.

This chapter charts the changes in explanations for an emergent mental health problem in Britain during the 1960s and 1970s. The behaviour, which is called self-cutting, wrist-cutting, self-harm and self-mutilation, does feature in the studies that focus upon self-poisoning, but is largely ignored as a methodological quirk, as is shown at the outset of this chapter. This contrasts with the emergence of cutting in psychiatric inpatient institutions – which first figures as a management and behavioural epidemic problem. The management problems (and the negative reactions of some staff members) feed into an emphasis of internal, psychopathological aspects of the phenomenon, rather than the communicative, imitative and competitive (potentially ‘manipulative’) aspects. When this has sufficiently stabilised, it is able to inform studies based at general hospitals. Explanations emphasising internal over possible external factors find traction when compared to the self-poisoning at A&E departments.

Self-cutting also receives a boost in visibility when it features in the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders for the first time in 1980 (DSM-III), as a possible symptom of borderline personality disorder. Hawton and Catalán mention ‘personality disorders’ in their analysis of self-mutilation in 1982 and, in the second edition of the same text (1987), they add the following sentence: ‘In the USA the DSM III diagnosis of “borderline personality disorder” would often be used for such individuals’.¹²³ This implies that the DSM and borderline are involved in the increasing prominence of such symptoms – as self-mutilation features in the DSM for the first time as a symptom of borderline in 1980. By 2014, it is afforded a diagnosis of its own: Non-Suicidal Self-Injury.

The behaviours of self-poisoning and self-cutting emerge in very different institutional settings, despite their common co-occurrence in the same patients. Self-poisoning is much more likely to bring a person under medical scrutiny, whereas superficial cuts to the arms are much more likely to be noticed when a person is already in an inpatient setting. This gap between inpatient and A&E studies has been forgotten in the transformation in the visibility of the two behaviours of self-cutting and self-poisoning.