Teaching ‘small and helpless’ women how to live: Dialectical Behaviour Therapy in Sweden,ca 1995–2005

Introduction

Dialectical Behaviour Therapy (DBT) was developed by US psychologist Marsha Linehan in the 1980s to treat a specific patient group: young women diagnosed with Borderline Personality Disorder (BPD) who exhibited strong suicidal or self-harming tendencies. DBT grew out of the more popular Cognitive Behavioural Therapy (CBT) and shares many of its premises. In particular, within DBT, emotion is perceived as automated and involuntary, but nevertheless an event, or process, that can be regulated. DBT relies upon a model of emotional distress that prejudices the internal and the individual over the social. While distress is perceived to result from a combination of environmental triggers and an individual’s particular vulnerability, the focus is solely on teaching the patient strategies to function effectively in the triggering environment. This idea of an individual as a self-regulating unit resonates with the wider socio-economic context in which this model prevails. Economic (neo)liberalism, the political programme of choice in the West since the 1980s, promotes neo-Victorian values such as self-help and individual responsibility as the route to wellbeing, and can be contrasted with the collective social programmes of the post-war period.

This article maps the history of DBT in Sweden from its introduction in 1995 up until the publication of the initial findings of the first major domestic DBT studies. In doing so, it addresses a small but important chapter of post-war psychiatry in Scandinavia, which commenced with a sweeping reform of Swedish psychiatric care in 1995. Drawing on existing historical accounts of political and economic reform in Sweden, as well as key policy documents of the time period considered, the article places the introduction of DBT in the context of the profound political shift taking place in Sweden at the turn of the 21st century. It begins by sketching out this shift and the events preceding it, and then goes on to situate the first DBT study against the backdrop of previous research on BPD in Swedish psychiatry, highlighting an increased focus on female patients with the advent of DBT. The introduction of DBT is then charted through the publications arising from the first domestic study and its theoretical framework, which includes Linehan’s original work and its translation and adaptation by Swedish clinicians. Finally, the implications of the changing ontology of the Borderline diagnosis within DBT is considered in relation to the wider shift in economic policy and political discourse that facilitated the psychiatry reform. This shift was ideologically framed as a necessary liberal reform of an overly bureaucratic system responsible for economic stagnation and the stifling of personal growth. In this context, DBT was presented as a way to teach Borderline patients the skills necessary to function as independent individuals responsible for their own wellbeing. Patients perceived their participation in therapy as a form of ‘work’ requiring diligence and commitment, prompting questions about the role and meaning of work as therapy/therapy as work in 21st-century neoliberal society.

1995 was a watershed year in Swedish psychiatric care. Across the country, local counties began the closure of psychiatric hospitals in a sweeping reform that saw the number of beds dramatically reduced by the end of the decade. The same year as the psychiatry reform began, the first Swedish study of DBT was commissioned. Known as the SKIP study, its chief aims were to ‘investigate the prevalence of Borderline Personality Disorder (BPD) among women who make repeated suicide attempts’ (Karolinska Institutet, 1999: 21), and to compare DBT to other forms of treatment for Borderline patients. Key investigators in the early Swedish research on the efficacy of DBT included psychiatrists Marie Åsberg and Åsa Nilsonne, as well as Anna Kåver, a clinical psychologist. Following Linehan’s model of DBT (Linehan, 1993), the SKIP study focussed on female Borderline patients with strong suicidal tendencies, an approach that was reiterated and theorised in Kåver and Nilsonne’s clinical manual on DBT published in 2002.

The 1995 psychiatry reform significantly reduced the number of hospital beds, seeking instead to integrate the severely mentally ill into the community. The SKIP study explicitly framed DBT as a cost-effective treatment for a category of patients who required significant institutional support, teaching them the skills needed to survive in the post-institutional world. Much of the language of DBT corresponded to the increasingly popular policy language of the period emphasising ‘responsibility’, ‘independence’ and ‘work’. Moreover, within the SKIP study, the typical Borderline patient was constituted as a ‘helpless’, almost childlike woman unable to control her own strong emotions or to cope with life. DBT was framed as a strategy to teach this Borderline patient the skills she needed in order to function in a society that increasingly valued autonomy and independence. Initial assessments of patients’ experiences of DBT reveal a perception of therapy as a form of ‘work’ that the patient had to undertake in order to get better. Work as therapy has been central to modern psychiatry since the early 19th century, where it went hand in hand with a worldview that promoted industriousness as a virtue, and played a key role as a restorative and corrective practice in asylum treatment (see e.g. Freebody, 2016; Scull, 1993). In the therapeutic context of DBT in Sweden at the turn of the millennium, therapy itself took on the role of restorative labour. Following from this, it is argued here that the therapeutic structure of DBT, the ways in which it was deployed in Sweden, and the language used to talk about its effects, cannot be separated from a wider socio-economic context that increasingly promoted work over welfare and individual responsibility over social support.

A rich history exists charting the development of psychotherapies in Britain and the USA, as well as the early history of psychiatric treatment in Western Europe (see for instance Burnham, 2012; Caplan, 1998; Engstrom, 2003; Marks, 2012). More recently, an emerging body of work addresses the history of psychiatric care and treatment in the former communist states (Savelli and Marks, 2015) and the history of psychotherapy in Western and Central Europe (Chaney, 2017; Cummings, 2017; Freis, 2017). However, the history of psychiatry and psychotherapy in the Scandinavian countries remains underexplored, and existing works have largely taken a descriptive approach (e.g. Gieser, 2009). While diagnostic tools and treatment approaches used in Scandinavian psychiatry and psychology have for the most part been borrowed from the Anglo-American context, post-war socio-economic development in these countries was markedly different from both Britain and the USA. The health systems emerging in Scandinavia in the second half of the 20th century shared many features with, for instance, the British National Health Service but were, in other ways, unique. Local variations in policy language and attitudes to health and illness meant that diagnostic language and treatment approaches were not simply transferred wholesale; thus, the transfer of knowledge can only be understood with reference to the wider cultures into which such knowledge was translated and used.

In a recent special issue of this journal, Sarah Marks importantly notes that historians ‘have, to a large degree, not attended to the intellectual and cultural development of many therapeutic approaches’, which ‘has the potential consequence of implying that therapies have emerged as value-free techniques, outside of a social, economic and political context’ (Marks, 2017: 4). Here I follow Marks in arguing for the importance of explaining therapeutic developments with reference to the societies that produced them. Moreover, I suggest that such an approach to the history of psychotherapy and the psy disciplines more generally is – or should be – a political act. In the present article, I aim to show that the history of DBT in Sweden is not simply a history of psychiatric treatment, it is a history of the relationship between psychiatric knowledge and political economy. At any particular moment in time, prevailing models of mental illness and its treatment do not merely reflect, but are produced by and reinforce the existing socio-economic order. While I do not wish to suggest a grand narrative of ‘social control’, I implore historians of psychiatry to recall Foucault’s observations about the pervasive role of psychiatric power in modern society (see e.g. Foucault, 2006: 189–90) and his insistence that we analyse and critique the role played by apparently benevolent institutions such as psychiatry in the reproduction of hierarchical societies.

From egalitarian bureaucracy to ‘choice revolution’: The rise and decline of the Swedish welfare state

At the end of World War II, when much of Europe was faced with the momentous task of rebuilding infrastructure and creating new social structures, the construction of the Swedish welfare state was already underway. Because of the country’s official position of neutrality during the world wars, from the 1870s economic growth continued largely uninterrupted for a century (Henrekson, Jonung and Stymne, 1996). The Social Democratic Worker’s Party (Socialdemokratiska Arbetarpartiet) governed Sweden without interruption from 1932 until 1976. During this time, the party attempted to build the Swedish folkhem (‘people’s home’), a society founded on Keynesian economics and social egalitarianism. While such ideas were equally popular in many other West-European states, two things are worth noting about the Swedish context. First of all, state ownership of public goods and regulation of the market went further than in most democratic countries in the West. In addition to schools, hospitals, public transport, telecommunication, and energy, the Swedish state also held monopoly on pharmaceutical trading and alcohol sales. Moreover, in stark contrast to many other West-European states, private alternatives to state-provided services were virtually non-existent in both the education and health care sectors. By the 1970s the private health care industry had been significantly marginalised following a 1969 law that made all medical practitioners employees of the state and placed strict limitations on their ability to run private practices on the side (Borgenhammar, 1984: 472).

Secondly, it was not primarily the level of public spending and ownership that distinguished Sweden from other European democracies with similar systems, ‘but rather the way in which it…institutionalized the values of universalism and social egalitarianism’ (Blomqvist, 2004: 140). There was an explicit aim on the part of the Social Democratic government to create a classless society, not just in an economic sense of the word, but also in the sense of social equality and shared values. In large part this was a top-down, state-driven project, reducing income inequality through heavy taxation of the wealthy and creating a social security system that ensured a comparatively comfortable standard of living for all citizens (Björklund and Freeman, 1997: 33–6). Gender inequality was tackled primarily through a heavily subsidised child care and preschool system and a universal and comparatively generous allowance for parental leave, which both parents were encouraged to share. The overall aim of the Social Democratic project of building the Swedish welfare state was to create a society where not only the most visible socio-economic inequalities were erased, but where every person could, in effect, feel like one of the people in the ‘people’s home’.

Until the mid-1970s, the welfare state had the support of a significant majority of the population on both sides of the political divide. Swedish citizens did not feel the effects of the 1973 oil crisis and subsequent economic downturn quite as strongly as, for instance, citizens in the Anglo-American countries. This was, however, in part thanks to continued high public spending, which soon generated a substantial deficit. Right-of-centre opposition to the government’s fiscal policies increasingly gained ground, and in the 1976 general election the Social Democrats were unseated by a centre-right coalition. When the former returned to power in 1982, their previously unequivocal commitment to public ownership had grown ambivalent. Soon after the 1982 general election, the new Social Democratic government began a process of deregulation and privatisation. Deregulation of the mortgage market removed the previously quite strict limits on household borrowing and produced a housing bubble that eventually led to the country’s worst economic downturn since the early 1930s. In 1991, at the height of the recession, a new right-of-centre coalition came to power. The conservative-led government lasted only three years, but by this point faith in the Keynesian model had declined on both sides of the political divide, in favour of deregulation of the market and privatisation of public goods (Blomqvist, 2004: 144–6; Lindbom, 2001: 171–2; Sejersted, 2011: 333–7, 388–92; Svanborg-Sjövall, 2014: 182–3).

As a result, the health care industry was gradually opened up to private actors in a reform modelled on the British system. ¹ It is important to note that this was not just a shift in policy, but also in the general attitude to public goods. The previous idea (or ideal) of a universal health care system that aimed to eradicate health inequality by ensuring that service quality was equally high across the country, was replaced by what became known as a ‘choice revolution’, ² driven by the contrary belief that equality and access to high quality care could only be guaranteed by a system that allowed each citizen to freely choose their preferred health care provider from a wide range of public and private service providers.

From the mid-1970s to the mid-1990s a notable shift in attitudes occurred, in the first instance among the political establishment, but eventually also across wide sections of the public – what one observer has aptly termed a ‘paradigmatic drift’ (Carson, 2005: 193). The Social Democratic ethos was gradually abandoned, and with it the language of egalitarianism, solidarity, and social cohesion that had underpinned and reinforced the folkhem concept. This was replaced by liberal language centred around choice and freedom, coupled with a stronger focus on individual responsibility for one’s socio-economic situation.

The move toward neoliberal economics in Sweden occurred against the backdrop of a wider shift in the West beginning in the 1970s. Neoliberalism is, however, used here to refer not only to an economic idea centred upon privatisation, deregulation, and, more broadly, free market capitalism. Following Nikolas Rose, we can understand it as an ideology (here meaning a set of ideas about how a society should be organised) in which ‘the well-being of both political and social existence is to be ensured…through the “enterprising” activities and choices of autonomous entities – businesses, organizations, persons – each striving to maximize its own advantage’ (Rose, 1999: 153). The fiscal approach championed by Reagan in the US and Thatcher in the UK was closely wedded to a belief in individualism and human freedom. Indeed, Austrian economist F.A. Hayek, on whose work Thatcher based much of her approach, had originally argued for economic liberalism as the only way to defend political freedom against the threat of totalitarianism (Hayek, 2001[1944]). However, while a new politics of privatisation, deregulation, and austerity introduced in Sweden in the 1980s and 90s largely mirrored that of the Anglo-American context, there was a notable difference in the language used to sell neoliberalism to the public. Merging the language of liberalism with the language of the traditional left, the new economic policy was marketed as a ‘choice revolution’ by the right-of-centre coalition that governed between 1991 and 1993.

Central to the new economic policy was a focus on ‘work’ as an end in itself. In terms of welfare policy, this sentiment generated increasing support for what became known as arbetslinjen (‘the work line’), a policy that was eventually endorsed by political parties on either side of the centre. According to this view, work is an end in itself, as well as the antidote to most of society’s challenges, such as segregation and youth crime, but also mental illness (Piippola, 2010: 258). This approach came to gradually influence the public medical insurance system, where illness was increasingly defined in terms of (in)ability to work rather than according to medical criteria. This had a particular impact on perceptions of mental illness, where the individual’s ability to carry out some form of employment became central in terms of defining and treating pathological states (see e.g. Söderberg, 2009: 498–9 for an outline of the new medical insurance system). This view of work as a moral good and a measure of a person’s usefulness in society has roots going back to the 19th century, where work was deployed as a corrective strategy in prisons and as a therapeutic practice in lunatic asylums, as well as a way to ensure that the poor were being useful to society through the introduction of workhouses (Driver, 1993; Ernst, 2016; Wiener, 1990: 117–22). As we will see below, the introduction of DBT in Sweden added a new dimension to the idea of work as restorative, where rather than work being used as a form of therapy, therapy itself was constituted as work.

The 1995 psychiatry reform

Against the backdrop of the shift described above, on the first of January 1995 a law came into force that produced a sweeping reform of psychiatric care. The anti-psychiatry movement that had emerged in the UK in the 1960s was also influential in Sweden, but had not lead to any significant changes in the care and treatment of psychiatric patients. While there was a modest and gradual reduction in inpatient places over the next two decades, sweeping deinstitutionalisation only occurred following the 1995 reform (Arvidsson and Ericson, 2005: 186–7; Stefansson and Hansson, 2001: 82). The closing of psychiatric hospitals in Sweden was, importantly, not primarily motivated by a critique of institutionalisation of the mentally ill, but rather a product of the neoliberal ‘choice revolution’, driven in part by a belief that psychiatric patients should learn to live independently and be integrated into the community, and in part by concerns to make psychiatric treatment more cost-effective during a time of recession. ³

In 1992, a parliamentary commission was tasked with investigating access to services by the long-term, severely mentally ill, and its findings formed the basis for the subsequent 1995 reform (Arvidsson and Ericson, 2005: 187; Brinck, 1994: 265; Stefansson and Hansson, 2001: 83). ⁴ The commission concluded, among other things, that greater measures should be taken to rehabilitate the mentally ill and reintegrating them into society and the job market. Following from this, the Social Insurance Agency (Försäkringskassan) and local municipalities were given greater responsibilities than previously in terms of preparing mentally ill patients for a return to work, through various job-training and social rehabilitation programmes (Brinck, 1994: 265–6; Stefansson and Hansson, 2001: 82–4). Once begun in earnest, deinstitutionalisation was rapid and sweeping. The number of days of psychiatric hospital care was reduced by 75% by the end of the decade, reflecting the downsizing or closing of a number of inpatient institutions. In the late 1960s, there had been approximately 33,000 psychiatric beds in Sweden, servicing a population of 7.5 million. At the start of the 21st century there were around 5000 beds left, in a population of just under 9 million (Arvidsson and Ericson, 2005: 186–7; Brinck, 1994: 258–61).

The advent of DBT in Sweden can only be adequately understood within this context. DBT was explicitly framed as a response to the need for a cost-effective outpatient treatment for a group of patients who required long-term psychiatric care and support. As we will see below, DBT placed individual responsibility in focus; the patient was to learn the skills needed to function independently in society. This emphasis on learned autonomy and independence was, I argue, both moral and economic, informed by a neoliberal approach to social economics (and, to some extent, by a neo-Victorian view of humanity). However, the harsh political climate present in Britain and the US during the Thatcher–Reagan era was, at the time, largely absent from the Swedish context (for a poignant critique of mental health policy in the hands of Anglo-American neoconservatives, see Scull, 1989). It is important to note that when neoliberal ideas were introduced to the wider Swedish public in the lead-up to the 1991 election, reform (of the welfare state in general as well as of psychiatry) was framed almost exclusively as a positive strategy aimed at improved quality and more flexible services. The keywords were choice and independence: freedom to choose one’s school, doctor, and energy supplier (and so on), and independence from a controlling and bureaucratic state apparatus that sought to dictate to people how to live their lives. In a 1990 policy document entitled Ideas for Our Future, the largest right-of-centre party, Moderaterna (the Moderates) asserted that ‘every intervention in citizens’ lives, every regulation, every tax, makes planning more difficult and kills life projects. We need more freedom. Much more’ ( Moderata samlingspartiet, 1990: 65). This sentiment was reiterated in their 1994 election manifesto, which held that ‘power over one’s situation and everyday life should belong to individual human beings, not authorities’ ( Moderata samlingspartiet, 1994: 8).

Similar language of autonomy versus dependence would frame the introduction of DBT into the Swedish mental health system. DBT was formally introduced in Sweden the same year as the psychiatry reform commenced, as part of an attempt to find an effective treatment for people (or women, rather, as will be clear momentarily) diagnosed with BPD, and who exhibited ‘suicidal’ or ‘self-harming’ behaviour. These patients were seen as unable to cope with regular life, demanding a level of stability and permanence from the structures and people around them that could, I argue, not be provided in the post-institutional, post-Social-Democratic world of infinite choice and individual freedom. DBT aimed to teach these patients the skills necessary to exist in this brave new world.

Borderline personality disorder, suicidality, and the creation of DBT

In the decade leading up to the 1995 Swedish psychiatry reform, the number of psychiatric inpatients diagnosed with personality disorders grew by 30% (Brinck, 1994: 261). This development can perhaps in part be attributed to the publication of the third edition of the American Psychiatric Association’s (APA) Diagnostic and Statistical Manual of Mental Disorders (DSM-III) in 1980, which gave personality disorders a much more prominent place than in previous editions, and introduced BPD as a formal diagnosis for the first time. The DSM-III framework for personality disorders was taken up into clinical practice and research in Sweden, and the usefulness of the APA’s criteria for diagnosing BPD was assessed in a number of domestic studies in the 1980s and 90s (e.g. Kullgren, 1984; Ståhlenheim and Von Knorring, 1996).

Borderline first emerged as a psychiatric concept within post-war North-American psychoanalysis, to describe a category of patient who was perceived to oscillate between neurosis and psychosis, and who was not susceptible to analysis (see e.g. Knight, 1953). The Borderline diagnosis was not explicitly gendered in the early decades, nor were suicidality and self-harm distinct features, but this changed gradually in the lead-up to DSM-III. ⁵ During this period, a number of different diagnostic models and criteria for Borderline were offered (see esp. Kernberg, 1967; Grinker, Werble and Drye, 1968; Gunderson and Singer, 1975). The language used to talk about Borderline patients changed, with adjectives such as ‘narcissistic’, ‘aggressive’, and ‘manipulative’ gaining popularity, in large part thanks to the influence of Otto Kernberg’s (1967, 1971 and 1975) work. This was subtly reflected in the DSM criteria introduced in 1980, where impulsive and self-destructive behaviours were the key symptoms around which the diagnosis was constituted. ⁶

The DSM criteria can be broadly summarised as made up of three key factors: dysregulated emotionality, self-destructive behaviours, and stereotypically female personality traits. This definition of BPD has not changed significantly since 1980. Borderline is presently defined as a ‘pervasive pattern of instability of interpersonal relationships, self-image, and affects, and marked impulsivity’. Key criteria include ‘chronic feelings of emptiness’, ‘suicidal behaviour, gestures, or threats’, ‘frantic efforts to avoid real or imagined abandonment’, ‘affective instability’, and ‘impulsivity’ in areas such as ‘sex’, ‘spending’, and ‘binge eating’ (APA, 2013). Between 70 and 80% of those who are diagnosed with BPD are women (APA, 2013; Lieb et al., 2004), and the diagnosis has received criticism and scrutiny for its overt gender bias. Nehls (1998) has argued that female Borderline patients are stigmatised by clinical staff, who often perceive them as manipulative attention seekers. Skodol and Bender (2003) have interrogated the overrepresentation of women among those who receive the diagnosis, concluding that it is largely the result of ‘sampling bias’, and that the actual gender rate is unknown. As we will see below, this ties in with early BPD studies in Sweden prior to the advent of DBT, where the male–female ratio varied significantly between studies.

While historians have begun to trace the history of Borderline and of personality disorders more generally, the clinical and theoretical processes that produced BPD as a typically female diagnosis have received scant attention. Becker (1997) has placed Borderline on an historical trajectory following medieval ‘witches’ and 19th-century ‘hysteria’; however, BPD is a modern diagnosis and must be explained with reference to the psychiatric culture that produced and moulded it over the last 60 years. More recently, Lunbeck (2006: 153) has offered the beginnings of an archaeology of the Borderline concept, aptly suggesting that its emergence is best understood as ‘the fitful coalesence of a new category – assembled from new concepts and new terms – that has enabled clinicians to take note of and diagnose dimensions of personhood they had not formerly been able or inclined to register’.

Early post-1980 studies of BPD (the majority of which occurred in the USA) tended to emphasise the perceived ‘attention-seeking’ and ‘aggressive’ quality of patients’ self-destructive behaviour, meaning that patients’ self-injurious acts were primarily read as ‘manipulative’ rather than genuinely suicidal (e.g. Koeningsberg, 1982; Zanarini et al., 1990). ⁷ The overrepresentation of women and the apparently high rates of suicidal behaviour among BPD patients attracted the attention of US psychologist Marsha Linehan, who started researching gender and suicidality in the 1970s (Linehan, 1971, 1973). ⁸ Linehan (1993: 42) went on to develop a ‘biosocial’ model for BPD, which held that the disorder is ‘primarily a dysfunction of the emotion regulation system’. She suggested that individuals suffering from this condition have a genetic vulnerability making them more sensitive to emotional pain and distress, less able to tolerate negative emotions, and more dependent on social connections and support for their wellbeing. This genetic vulnerability would act in conjunction with environmental factors. For instance, Linehan noted that people diagnosed with BPD had a high prevalence of having experienced sexual abuse in childhood (see also e.g. Shaw and Proctor, 2005). In her view, then, the two factors, genetics and environment, together produced the disorder. ⁹

Linehan rejected the view that Borderline patients were manipulative, selfish, and attention seeking, suggesting instead that they were experiencing intolerable pain, and that this pain was the source of their impulsive and self-destructive behaviour. Moreover, in line with her argument about environmental factors, Linehan offered a ‘feminist’ perspective on the aetiology of BPD, suggesting that that the diagnosis must be placed in the context of a society that values individual autonomy over a ‘social self’. She suggested that ‘the importance of a relational or social self among women has been highlighted by many feminist writers’, and that, following from this, ‘the problems encountered by the Borderline individual may result in part from the collision of a relational self with a society that recognizes and rewards only the individuated self’ (Linehan, 1993: 31–2). ¹⁰ Linehan developed DBT to teach strategies for dealing with the pain arising from this collision between the Borderline patient and their environment, and with the suicidal behaviour that she saw as resulting from it. In this way, Linehan also sought to overturn the stigma attached to the diagnosis and challenge the negative perception of Borderline patients (ibid.: 42–65).

DBT shares a foundation with CBT (from which it derives), but its primary focus is not on breaking or interrupting disordered thought patterns and associated actions, but rather on learning to tolerate pain and regulate strong emotions. As such it can be viewed as a more comprehensive programme than CBT: rather than teaching patients strategies to deal with a certain phobia, or with panic attacks, obsessive-compulsive behaviours, or generalised anxiety, the aim is, in a more basic sense, to teach patients how to live. DBT is constituted around two core principles: acceptance and change. It is between these two that a dialectic is seen to occur. The patient has to learn to accept the present situation, however difficult or intolerable, while at the same time working toward changing it. The synthesis, then, becomes ‘a life worth living’ (Dimeff and Linehan, 2001: 2; Robins, Ivanoff and Linehan, 2001: 448). The core aim of DBT is to teach patients ‘emotion regulation skills’. Emotion regulation has become a key concept in 21st-century psychiatric treatment of affective and personality disorders. Linehan, Bohus and Lynch (2007: 583) describe emotion dysregulation as ‘the inability, even when one’s best efforts are applied, to change or regulate emotional cues, experiences, actions, verbal responses, and/or non-verbal expressions under normative conditions’. The idea that suicidality and self-harm occurs as a response to dysregulated affect can be traced back to institutional studies on self-harming female patients in the US in the 1960s and 70s (Millard, 2013). However, the idea of self-regulation of emotion as a desirable skill has an even longer history. Victorian physicians, keen to explain elusive mental functions in the language of modern science, cast emotion as an involuntary physiological process, suggesting that the individual had little control over its eruption. This model did not, however, sit comfortably with Victorian moral principles of self-help and respectability, thus, volition and control were restored through the concept of habit, which suggested that with persistent practice and ‘watchfulness’, individuals could learn to counter their natural impulses and manage their emotional life (see e.g. Maudsley, 1884: 93). While the late 20th- and early 21st-century science of emotion regulation is markedly different from its Victorian precursor, the moral connotations have persisted and become amplified in the context of individualism and economic liberalism. As we will see below, the ideal of emotional self-regulation as a valuable life skill fits neatly within the framework of neoliberal discourse on individual choice and responsibility.

Linehan’s textbook on DBT, first published in 1993, introduced the practice to a wider clinical audience, where it gained approval as a treatment for the notoriously ‘difficult’ Borderline patient. Since then, DBT has been implemented as a targeted treatment for BPD in a number of countries and multiple studies have been carried out. The vast majority of these hail DBT as a great success in treating Borderline patients. This has also been the case in Sweden. However, the trial, assessment, and measured outcome of DBT as a treatment for female Borderline patients cannot be considered independently from the socio-political context in which it occurred, a context that helped define both the problem to be addressed (the psychopathology of BPD) and what constituted a ‘successful’ outcome. The pathology to be treated: weakness, helplessness and lack of agency and control, and the desired result: an independent woman able to function in the post-welfare society without a strong support structure and thus being less of a financial burden on the system, were both intimately tied up with the new liberal project, as we will see below.

Borderline in Sweden before DBT

By the early 1990s, psychotherapy had become an important component of psychiatric care and treatment in Sweden. A government-regulated training programme in psychotherapy was instituted in the 1970s, and from 1986 onward all psychiatric staff were required to have undergone psychotherapy training. Cognitive therapy was first introduced in Uppsala in the early 1980s, but the national spread of CBT was slow and gradual prior to the 2000s (Gieser, 2009). Swedish uptake of the new DSM-III diagnoses, was, however, much more rapid. During the 1980s and 90s, Umeå psychiatrist Gunnar Kullgren headed a number of studies on the new Borderline diagnosis and suicidal behaviour as well as completed suicides (Kullgren, 1984; Kullgren, et al., 1986; Kullgren, 1988). One such study found the male/female ratio of BPD in their group of subjects to be 3:6 (Kullgren, 1984). However, in a different study Runeson and Beskow found the BPD diagnosis to be more common in their male subjects (Runeson and Beskow, 1991). Runeson wrote extensively on suicide and suicidality in young people during the late 1980s and early 1990s, and found that repeated suicide attempts occurred among both male and female psychiatric patients, and that the Borderline diagnosis corresponded to approximately 30% of the cases he reviewed, which were of both genders (Runeson, 1989, 1990; Runeson and Beskow, 1991; Rich and Runeson, 1992; Runeson, Beskow and Waern, 1996). In a study published in 1993, Bodlund, Ekselius and Lindström found that gender differences in the Borderline diagnosis were ‘much smaller than is usually reported’ (p. 432) and a 1996 study investigating psychopathy and criminal acts found that in a group of 61 male subjects, 17 fit the diagnostic criteria for BPD (Ståhlenheim and von Knorring, 1996).

These early studies are significant as they recruited both male and female subjects, and found that significant numbers of both genders exhibited suicidal behaviour as well as fit the Borderline diagnosis. “What is most striking about Bodlund, Ekselius and Lindström’s study is the authors' conclusion regarding the gendered nature of the BPD diagnosis.” When discerning Borderline symptoms using a self-screening tool, there was significantly less difference in the male–female ratio than when the same patients’ records were reviewed by clinicians. This led the authors to conclude that the gender bias in BPD was ‘not exclusively linked to the criteria per se, but also reflects a diagnostic bias on the part of the diagnostician’ (p. 432). The introduction of DBT in Sweden ensured a greater focus on women as the main sufferers of BPD. Within the context of DBT, the Borderline persona was transformed from a ‘manipulative’ and ‘aggressive’ person of either gender, to a ‘weak’ and ‘helpless’ woman unable to cope in a post-institutional society that favoured independence, freedom and choice.

‘Powerless, small, and helpless’: The 1995 SKIP study and the new Borderline patient

In 1995, the same year as the psychiatry reform commenced, a study was commissioned by Stockholm’s Primary Care Trust in conjunction with the Department of Clinical Neuroscience at Karolinska Institute (KI). The study, which became known by its acronym SKIP (Stockholm County Council and Karolinska Institute Psychotherapy Project for Suicidal Women with BPD) aimed at investigating the prevalence of BPD among self-harming female psychiatric patients and comparing the efficacy of three types of treatment for suicidal ideation and behaviour: DBT, Kernbergs’s psychodynamic psychotherapy, and standard psychiatric treatment. It should be noted that while Kernberg’s targeted treatment method for BPD was part of the study, his clinical description of the diagnosis, which included anger as a key feature, was not. The SKIP study was carried out in two regions: Stockholm and Uppsala, and sought to evaluate both patients’ and practitioners’ experience of DBT, as well as its cost-effectiveness. The principal investigator was Marie Åsberg, who had previously worked on suicidality and depression. Åsberg founded a Swedish DBT research group, which came to include Åsa Nilsonne, a psychiatrist linked to the Department of Neuroscience at KI, and Anna Kåver, a psychologist trained in CBT (Jimson and Lindahl, 2013: 7). While the final results of the SKIP study are at the time of writing yet to be published, it generated a number of research articles, and opened the door to other DBT studies in Sweden, including a study in Uppsala overseen by Kåver.

Kåver and Nilsonne both received their DBT training from Marsha Linehan, and the pair co-authored the first Swedish language DBT manual, published in 2002. Both the SKIP study and Kåver and Nilsonne’s subsequent manual were largely faithful to Linehan’s model for DBT as well as her ontological description of Borderline. However, where Linehan had offered a nuanced and detailed discussion of the gendered features of the Borderline persona as a social construct drawing on key feminist literature, these arguments were simplified in Kåver and Nilsonne’s condensed translation, which did not cite any sources for the argument that women are more reliant on social networks for their mental wellbeing than men, nor made any reference to feminist theory. Readers of the manual are presented with a one-dimensional view of gender and of feminism, one that fails to address the social production of typically ‘female’ traits, suggesting only that these may be at odds with a society that values autonomy. The implication is that ‘woman’ is a fairly stable category, but that ‘female’ characteristics are more or less accepted in different cultures (Kåver and Nilsonne, 2002: 61–2).

The SKIP study recruited only female participants. This was not coincidental but rather part of the stated remit of the study (Persius, 2006: 33). The decision to recruit exclusively female subjects set SKIP clearly apart from earlier studies on BPD in Sweden, where recruitment was not done on the basis of gender, and where consequently both male and female subjects participated. SKIP framed DBT as a strategy seeking to redress patients’ difficulties by on the one hand validating their suffering and recognising their helplessness, while at the same time demanding that patients work toward changing their behaviour and taking responsibility for their own lives. Following Linehan’s manual for DBT, patients participating in the SKIP trial had to sign a contract committing to 12 months of therapy, which consisted in one-on-one therapy as well as group-based ‘skills training’. They were also given the opportunity of utilising phone consultations with their therapist when required. The contract between the patient and the therapist in DBT is perceived as an integral part of the therapeutic relationship and process, and is one of the ways in which patients are taught to be responsible for their own lives. Patients who sign a DBT contract have to commit to attending all sessions within reason (if 4 consecutive sessions are missed, patients are ‘out of the programme’ without exception); agree to work toward reducing self-harm and suicidal behaviour; and agree to reduce therapy-disrupting behaviour. Continued therapy after the contract expires is made contingent upon the patient making progress during the initial contracted term (Linehan, 1993: 112–15).

The first results of the Lund study were published in 2003, in the form of a qualitative evaluation of participating patients’ and practitioners’ perceptions of DBT. The authors introduced the study by suggesting that ‘[t]o a very great extent it is younger women who are suffering from’ BPD (Persius, et al., 2003: 218). In the same paragraph an earlier Swedish study on BPD and suicidal behaviour (Runeson and Beskow, 1991) was cited as evidence that suicidality is common in BPD, but without any mention of the fact that this study had found Borderline to be more prevalent among the male subjects who participated. Ten women between the ages of 25 and 49 took part in the Lund study (as well as two male and two female therapists). The authors reported that all patients claimed that DBT was ‘lifesaving’. Patients adopted the language of DBT when talking about their transformations; for instance, one patient held that ‘I can now regulate myself not to go as far as a suicide attempt’. The authors suggested that the strong focus on teaching Borderline patients responsibility for their progress in therapy as well as for their lives more generally was an integral part of what made DBT so successful. Patients’ statements were quoted as evidence for this. One participant claimed that ‘It’s been a hell of a struggle all the time, but it’s been worth it’, with another one stating that ‘I have got a responsibility, it’s me that has to change things with support from my therapist’, while a third declared therapy to be ‘a full time job’ (Persius et al., 2003).

The perception of ‘therapy as work’ offers a striking parallel with ‘work as therapy’, which Scull (1989: 90) has referred to as a ‘cornerstone’ of psychological treatment in the 19th century. In the latter context, productive employment was seen as restorative, a view anchored in Protestant ethics. The 19th-century language around work and productivity saw a revival with the advent of the new liberalism in the late 20th century. Thus, the patients’ language mirrors the wider political discourse in Sweden at the time, where ‘work’ and ‘responsibility’ had become keywords in the neoliberal choice revolution.

Throughout the 1990s and early 2000s, the right-of-centre parties, particularly the Moderates, repeatedly portrayed the old Swedish welfare state as stifling of personal growth and restrictive of individual freedom, arguing that it encouraged citizens to live off government support rather than take responsibility for their own financial situation and find ‘real work’. ¹¹ In a policy statement adopted in 2001 (with the fetching subtitle ‘on the threshold of the best era of humanity’), the Moderates held that ‘the new society that is now unfolding improves the conditions for more people to become masters of their own fate’ (Moderata samlingspartiet, 2001: 12). Freedom, however, came with greater individual responsibility. ‘The free society’, it continued, ‘also places demands on people. It is first and foremost up to each individual to take responsibility for their own life and the life choices they make. The consequences of the personal life choice must be borne by each and every one’ ( Moderata samlingspartiet, 2001: 12). Upon regaining power in the 2006 general election, the Moderates transformed this sentiment into a welfare reform similar to the British ‘workfare’ system, where financial support was made reliant upon participation in various occupational schemes (Liljeberg, Martinsson and Thelander, 2013).

The idea of DBT as a full time job teaching patients responsibility was reiterated in the second article to come out of the Lund study (Persius et al., 2005), where the authors explored participating patients’ perceptions of their own suffering before and after DBT. Patients’ experiences were organised into three categories: ‘Life on the edge’; ‘A balancing act on a slack wire over a volcano’; and ‘The good and the bad act of psychiatric care in the drama of suffering’. These categories, which can be read as an attempt to decode as well as homogenise patients’ individual experiences, together present a very distinct image of the Borderline patient within DBT. She is a relatively young woman who is living a difficult life filled with self-destructive acts (life on the edge), a life characterised by strong and painful emotions which are beyond the patient’s control and which can erupt at any moment (the volcano), and a life where she has been mostly misunderstood by society and especially psychiatry, treated as manipulative when in fact she is helpless and vulnerable (psychiatric care in the drama of suffering). Patients are described by the authors as feeling ‘powerless, small, and helpless’. It is, however, unclear whether these are words used by the patients to speak of themselves or whether it signifies the interpretation of the authors. The patients who are quoted as evidence for this view speak of lacking and longing for support in their personal lives, for instance ‘It’s hard climbing this narrow path with precipices on either side. I wish there was someone to hold my hand supporting me over the obstacles’ (Persius et al., 2005: 164–5). Conclusively, DBT is framed as teaching the Borderline patient to regulate her painful and strong emotions as well as the self-destructive acts. This in turn transforms her from a weak and vulnerable person in need of stable and static structures and support, into an autonomous and responsible individual able to function in a society that requires independence and flexibility.

Between 2002 and 2004, another DBT study was carried out in at the Psychiatry Centre at Akademiska Sjukhuset in Uppsala. The study was overseen by Kåver, who cited the 1995 psychiatry reform as a key factor contributing to the need for a comprehensive DBT programme in Sweden, because one consequence of the reform was that ‘difficult’ patients were assigned to outpatient clinics badly equipped to deal with them. She noted that in addition to DBT giving Borderline patients the chance of a meaningful life, it had also been shown to reduce the financial burden of treating this patient group (Kåver, 2004: 2). The initial findings of the Uppsala study were published in 2005 as a dissertation by Erik Hjalmarsson (2005), under the supervision of Kåver. Citing Linehan, Hjalmarsson suggested that ‘within DBT it is assumed that people who have a strong desire to die do not possess the skills needed to create a life worth living’. The Uppsala study also recruited only female subjects, and had as one of its explicit aims to evaluate the cost effectiveness of DBT compared to other forms of treatment for Borderline patients. Hjalmarsson reported that DBT had resulted in improved psychological wellbeing and a reduction in self-destructive acts, remarking that where the latter had disappeared entirely, it had ‘probably been replaced by other more functional behaviours’ (Hjalmarsson, 2005). He also noted a significant reduction in hospital visits and in overall cost of care. In sum, DBT was presented as successful both in terms of alleviating human suffering and reducing the financial burden of caring for Borderline patients.

The cultural significance of Borderline and DBT

How, then, were Borderline patients to be taught the skills to create a life worth living? Kåver and Nilsonne explain the method in their 2002 clinical manual on DBT, where they also address the question of gender bias in the Borderline diagnosis. They opt to use the ICD-10 label Emotionally Unstable Personality Disorder (abbreviated in Swedish as IPS) instead of Borderline, partly because, they argue, it better reflects the core problematic of this patient group: an inability to effectively regulate their emotions. They also reject the BPD label due to the stigma attached to it, as it is used to describe a patient who is ‘usually female’ and ‘who is perceived to be difficult, provocative, and/or manipulative’, and where the prognosis is often negative to the point where ‘the patient is considered almost impossible to treat’ (Kåver and Nilsonne, 2002: 33). Kåver had previously expressed this sentiment in an article entitled ‘DBT for suicidal women’ (Kåver, 1997), suggesting of the BPD patient that her

surroundings misjudge the patient due to her ostensible and superficial competence. In actual fact, the patient has severe difficulties regulating her emotional life and to handle relationships and everyday problems in a good way. Linehan adds a feminist perspective to this problematic. Society finds it hard to accept dramatic and “wild” girls and the patient’s surroundings react in an invalidating way. (Kåver, 1997: 133)

In this way, DBT facilitates a transformation of the Borderline persona from a ‘manipulative’ woman to a ‘helpless’ one, that is, from a person ‘acting out’ with agency and autonomy, to one who is reacting to strong emotions over which she has no control. This transformation is particularly striking in light of Kåver and Nilsonne’s nod to ‘feminist psychotherapy’ in their book, suggesting that this offers a way of framing the emotional distress of Borderline patients as an ‘appropriate’ response to oppression rather than as ‘an expression of personal pathology’ (Kåver and Nilsonne, 2002: 31). While Kåver and Nilsonne acknowledge the potential role of patriarchal ‘power structures’ and ‘social values’ in producing the Borderline pathology, they fail to address the question of language as productive – of gendered features and of psychiatric symptoms, but also of the lived experience of both of these. Following Scott (1991), we might suggest that DBT and its language around female suffering offered new ways for Borderline patients to experience themselves as Borderline patients, new ways to make sense of their distress and how to address it. Through qualitative studies of that experience, patients’ perceptions were fed back into the diagnosis (see Hacking, 1995 on ‘looping effects’), further cementing the Borderline persona as a ‘weak’ woman, but one who could learn – through hard work – how to function as an autonomous and stable individual in a society that demanded that she did so. What is important to note in regard to the present story is that the ‘helpless’ woman better fits her contemporary wider socio-political narrative. That is, the infantile and dependent woman constituted by DBT was better suited to being transformed by the neoliberal ‘choice revolution’ into a responsible and independent adult than her wild and (by society) uncontrollable sister who was very much in charge of her own actions.

In their manual, Kåver and Nilsonne go on to outline various strategies that a DBT practitioner can use in conversation with their patients in order to facilitate this transformation. They suggest, following Linehan, that the aim of these strategies is in part to reduce the power imbalance between patient and therapist, creating a more equal and respectful relationship. Therapists are encouraged to use a mix of positive and negative approaches; specifically, they are advised to switch between ‘validation’ and ‘warmth’ on the one hand, and ‘irreverence’ on the other. This is presented as a particularly effective strategy when dealing with patients engaging in self-harm or in ‘therapy-disrupting behaviour’. In one example offered by the authors, a patient admits in session to having cut herself with a razor blade. The appropriate therapist response is given as validating the feeling that led to the cutting: ‘You must have been feeling terrible’ while not accepting cutting as an acceptable coping strategy: ‘you can’t carry on with this behaviour! Did you not use any of the skills you’ve learnt?’ When the patient explains what led to the cutting, in this case the therapist failing to answer her phone when the patient rang, the authors advise that an ‘irreverent’ response is appropriate: ‘well the world is full of careless people, you can’t go around killing yourself over that!’

Nilsonne and Kåver emphasise that while DBT uses both positive and negative validation, it never relies on punishment. At the same time, however, the strategies presented in their manual amount to a carrot and stick approach, where patients are rewarded when behaving appropriately and reprimanded when engaging in inappropriate behaviour, such as self-cutting. This is illustrated by another example, where the patient (‘Sofia’) has self-harmed in response to a difficult situation and failed to contact the therapist when she was experiencing severe distress (this constitutes therapy-disruptive behaviour). Kåver and Nilsonne present the following dialogue between therapist and patient:

In this way, the Borderline patient is taught to respond appropriately and responsibly to difficult life events, as inappropriate reactions will have negative consequences in terms of the patient’s therapeutic experience. Moreover, the terms of the contract ensure that if the patient wants to continue to receive help (therapy) she must adhere to the contract and conduct herself in a manner that is considered responsible in the context of DBT.

The process described above is reminiscent of a parent raising an obstinate child. This type of therapeutic relationship has a history as long as psychiatry itself: it can be traced back to the ‘moral treatment’ of the early 19th century, where the fatherly alienist used a mixture of kindness and discipline to coax the ‘lunatic’ back to the world of responsible and rational human beings. Moral treatment also saw patients engaging in various forms of physical and mental activities (such as manual work, sewing, painting, singing, and going to church) designed to occupy their minds (and, in some instances, provide free labour). Barbara Taylor (2011) draws a parallel between moral treatment and the new psychological treatments of the post-war period in her autobiographical narrative of decarceration in Britain, suggesting that the latter was a kind of resurrection of the former. However, these two movements were anchored in fundamentally different worldviews and approaches to the human condition. The psychological treatments of the mid-20th-century were, like the dominant political projects of the time, profoundly social and relational – this was the era of couples’ and group therapies, as well as the period when psychoanalysis reached its popular peak in both Britain and the USA. In her discussion of moral treatment Taylor is, while critical, also largely forgiving of the intentions and wider social consequences of what has famously been described as a different, subtler yet sinister, form of control. However, while Foucault’s (2006[1961]) rendition of modern psychiatry was at times sensationalist and often generalising, it was so for a reason: his critique was ultimately a political polemic, and as such a tremendously important one. He saw the old shackles and chains replaced by something much more effective: an internalised desire – or compulsion even – to control oneself. Importantly, his critique was not just of psychiatry, but of modernity itself, that is, it was a critique of the Enlightenment project and its fruitful marriage to industrialisation and capitalism. Moral treatment, which sought to transform the lunatic into a rational, responsible, self-disciplining individual, was a fundamentally modern endeavour: a fusion of 18th-century moral philosophy and 19th-century liberal-capitalist (and, to an extent Protestant) values of ‘respectability’ and ‘self-help’. The focus was on the individual’s ability to function properly in society by way of becoming a rational person and internalising these values. In this way, asylum patients were assigned occupational tasks not just as a means to preoccupy their minds or provide free labour, but also to turn them into ‘useful members of society’ (Chaney, 2016: 277).

Thus, in the same way as the 21st-century neoliberal ethos draws upon a Victorian values system, moral treatment has far more in common with the behavioural therapies of the late 20th century, particularly DBT, than it does with the early post-war talking therapies, which can on the contrary be seen as part of the post-war Social Democratic project in Western Europe. Moreover, the occupational aspect of moral treatment can be compared to the 21st-century ‘workfare’ concept referred to above, a programme that encourages ‘labour on the self in order to achieve characteristics said to increase employability’ (Friedli and Stearn, 2015: 40). This is mirrored in DBT’s focus on ‘hard work’ in order to learn the skills necessary to improve one’s own quality of life. DBT, then, draws upon language and values central to turn-of-the-millennium Western society: freedom, choice, independence, and responsibility for one’s own life situation. At the same time, this relationship between DBT and its cultural context is a circular one, in that the former helps reinforce these values by teaching patients to internalise them. This process moreover applies to late-modern psychiatric conceptions of ‘self-harm’ more generally, whereby ‘neoliberalism’s stress on individual actors’ radical freedom to make choices for their own benefit fits well with a model of self-harm that emphasises the individualistic, private feelings of tension, and the self-regulation of these through cutting’ (Millard, 2015: 205). ¹²

Conclusion

Psychiatric diagnoses are malleable, reflecting the cultural contexts in which they operate (see e.g. Hacking, 1995; Jansson, 2013b; Young, 1995). The transformation of the Borderline persona within DBT occurred with ease, much in the same way as the shift from the Social Democratic welfare state to the new liberal society. The choice revolution was less a radical rupture than a seamless transition, which is not surprising when we consider that both of these socio-economic models are borne of the same modernity project. Writing shortly before DBT came to Sweden, Nikolas Rose (1993) suggested that the 20th-century welfare state and the market-oriented new liberalism are products of, and belong to, the same conceptual system of thought. Taking liberalism as a system of rule rather than a philosophical exercise, Rose argued that what he calls ‘advanced liberalism’ has produced a society where the state (and its institutions) and the civil and/or social spheres act in symbiosis rather than opposition, where people are governed by laws as well as by internalised norms embodied in the institutions and spaces that are run by ‘experts’ (e.g. psychiatrists) and which are not ‘public’ or governmental in the traditional sense. Within this system

relations are established between various centres of calculation and diverse projects of rule – more or less ‘rationalized’ as the case may be – such that events within the micro-spaces of bedroom, factory floor, schoolroom, medical consulting room might be aligned with aims, goals, objectives and principles established in political discourse or political programmes. (Rose, 1993: 287)

To conclude, then: as we have seen, DBT was introduced as a positive approach to a stigmatised and cost-heavy psychiatric condition, much like the choice revolution was introduced as a positive alternative to the stifling welfare state. Both emphasised choice and independence as desirable; as a marker of an individual’s wellbeing as well as their ability to function effectively in society. However, in a post-Enlightenment world of dichotomies, the flipside of emphasising choice and independence as desirable is what inevitably becomes undesirable as a result; in this case, an individual’s need for reliable, permanent structures and social support. Taylor’s (2011) astute assessment of mental health care in neoliberal, post-institutional Britain is an apt description of what is at stake for those who do not possess the ‘correct’ (i.e. normative) kind of independence and self-sufficiency, and is fitting also for the Swedish context:

Today the discourse of mental health ‘providers’ is all about autonomy and independence. The language of dependency is almost entirely negative. Its primary referents are to drug and alcohol addiction, but the pejorative connotations extend across most varieties of neediness, including for basic care and support. To need other people on a day-to-day basis (unless you are very young, very old or very disabled) is seen as inherently pathological; independence is a sina qua non of mental health. (Taylor, 2011: 198)