Organizational Turbulence,Trouble and Trauma: Theorizing the Collapse of a Mental Health Setting

Introduction

The concept of turbulence has long been discussed as a feature of organizations’ external environments (Emery & Trist, 1965). Accordingly, the turbulence literature is predominantly oriented outwards, exploring how turbulence, current and future, affects organizational relations (Child, Faulkner & Talman, 2005; Tsoukas & Shepherd, 2004).

Turbulence, however, is not merely a condition of external environments. Organizations’ ‘inner lives’ significantly affect their interactions with societal contexts. As Beck (1992) asserts, organizations demonstrate an increasingly reflexive relationship to their environments. External turbulence may induce inner turbulence, but inner turbulence may lead to a construction of an environment as turbulent and threatening. Below the surface of organizational rationality, many scholars have come to view a micro-politics of inner relations as fundamentally shaping contemporary institutions (Clegg, Courpasson & Phillips, 2006; Fleming & Spicer, 2007; McGivern & Ferlie, 2007).

If turbulence describes environmental conditions, the concept of trouble is often used in connection with an organization’s inner life and predicaments faced by its members. The dynamics of organizational ‘trouble’ have attracted some scholarly interest, from micro-political studies of deviance to organizational misbehaviour (Ackroyd & Thompson, 1999) and collective grievance (Earl, 2009). Originating in the social problems literature, the concept of trouble is helpful in researching the relationship between organizations’ formal and informal dimensions. Studies have analysed the development of trouble, its attribution to ‘troublemakers’ and its management by ‘troubleshooters’ (Emerson & Messinger, 1991; Rubington & Weinberg, 2003). Yet despite interest in its capacity to mobilize resources, the prevailing image of trouble is as something to be controlled and prevented.

Some studies have sketched a contrasting picture of trouble, not as an exceptional situation calling for emergency action, but a currency in which organizational members deal. Behind the scenes of the formal organization, participants from college roommates to corporate executives produce, trade and manage trouble in personal and emotional ways (Courpasson & Thoenig, 2010; Gabriel, 1995; Morrill, 1995). Beyond official control, yet inherent in an organization’s inner life, trouble occupies an ambiguous, liminal space. Even if ostensibly private, it may interact with the social setting, becoming heated and volatile, to produce dynamics of turbulence.

In this article, I argue that interactions between the formal and informal management of trouble produce turbulence that may mobilize resources and collective action, or conversely lead to dysfunction and crisis. Drawing on social constructionist insights on trouble, and linking these to relational psychoanalytic theory, I develop the concept of relational turbulence across what M. Stein (2007) terms ‘phenomenological’ or experiential boundaries, at individual, group and organizational levels.

Linking constructionist and psychoanalytic perspectives reveals, of course, certain lacunae in each; but more importantly, their combined lenses focus attention on organizational dynamics through which both trouble and its repercussions are readily produced. Social constructionism, in particular, emphasizes the role of ‘reaction processes’ (Gibbs, 1966) through which interested parties and official ‘troubleshooters’ actively construct certain matters as problematic.

Whereas social constructionism ties trouble analytically to the concept of deviance (Holstein & Miller, 2003; Schneider, 1985), relational psychoanalytic theory illuminates intersubjective dynamics of trouble and its potential for turbulence as an exchange intrinsic to all social relations. Contemporary relational psychoanalysis regards both the self and the social world as intersubjective, contingent and shaped by political and cultural dynamics (Mitchell, 2000; Quinn & Strauss, 2006). Through interactions at the interstices of social relations, liminal ‘third’ spaces may develop in which experiences, emotions and fantasies are exchanged across these boundaries (Crapanzano, 2006; Diamond & Allcorn, 2009).

The psychoanalytic literature has long regarded ‘third’ or transitional spaces central to personal and cultural development (Winnicott, 1953). Occupying an intermediate area between social ‘reality’ and inner experience, they cultivate individual and collective exploration and creativity (Anzieu, 1984). Yet their involvement in trouble suggests they also have turbulent and destructive potential, hitherto overlooked. I here link organizational perspectives on transitional space (Diamond & Allcorn, 2009; H. Stein, 2007) with recent scholarship on the perverse organization (Armstrong, 2005; Long, 2002, 2008) to explore how turbulent and potentially destructive forms of transitional space develop.

Based on a longitudinal interorganizational ethnography in the field of mental healthcare, this article examines how trouble associated with people with severe personality disorders generates turbulence in organizations that manage them. Through the case of a democratic therapeutic community (DTC), the study explores the production of turbulence, its escalation and organizational repercussions. Whereas a restorative mode of turbulence has developmental and self-regulating properties, a contrasting mode of perverse turbulence has destructive and escalating effects, ultimately leading to the DTC’s collapse.

These findings are theorized by extending the psychoanalytic concept of transitional space. In contrast to the notion of transitional space as a safe, protective area, I argue it can take the form of a turbulent, formative space in which authentic, mutual relations develop, useful for organizational productivity. Conversely, when perverse forms of turbulence develop, an intractable perverse space may take hold, from which participants’ abilities to extricate themselves are uncertain.

This article contributes to the turbulence literature by developing the concept of relational turbulence, and through developing a model of contrasting formative and perverse spaces created by this turbulence in organizations.

The Sociology of Trouble

In exploring the relation between trouble and turbulence, an important foundation is C. Wright Mills’ (1959) distinction between ‘personal troubles of milieu’ and ‘public issues of social structure’. According to Mills, trouble is a private matter of an individual’s experience of self and immediate relations, concerning personal values felt to be under threat. When private trouble reflects wider societal structures, though, trouble needs to be understood in terms of public issues. Mills maintained that public issues such as divorce, unemployment and large-scale social problems connect personal troubles with societal and institutional dynamics. He advocated paying attention to relations between personal biography and social history, studying what each produces in and reveals about the other.

A number of approaches have been adopted to study this relationship between individuals and society, particularly within the social problems literature. Some emphasize structural contradictions that produce problems through conflicting discourses and values (Blumer, 1961; Spector & Kitsuse, 1977), while others elucidate the development of trouble at a micro-level (Holstein & Miller, 2003). Whereas early literature focused on substantive social conditions, from perspectives of social pathology, disorganization, value conflict, deviance, labelling and critical theory (Rubington & Weinberg, 2003), social constructionism introduced a shift in focus. Instead of studying putative social conditions, scholars turned their attention to how interested parties come to define conditions as problematic (Best, 2002; Schneider, 1985).

On the one hand, according to constructionist arguments, social problems are established through assembling moral claims about undesirable conditions (Spector & Kitsuse, 1977). Through grass-roots agitation and referral to official parties, personal circumstances become entangled with public structure and rhetoric (Emerson & Messinger, 1991). Everyday ‘social problems work’ brings public discourses into play at a micro-level (Gubrium, 1991), ‘reassert(ing) the importance of practice … while context and interpretive structure promote pattern’ (Holstein & Miller, 2003, p. 78). Overall, constructionist perspectives offer a framework with which to explore ‘reaction processes’ (Gibbs, 1966), albeit emphasizing formally orientated control.

On the other hand, the concept of trouble addresses local disruptions to social norms which may be perceived and handled in informal ways. Analytically connected to the notion of deviance, trouble refers to discontent or upset arising from interpersonal interaction (Emerson, 2008). Participants’ activities and interpretive schemes can produce all forms of interpersonal trouble, often beginning amorphously with a tenuous sense of ‘something wrong’, only some of which are constructed as problematic (Emerson & Messinger, 1991). At this informal level, trouble is relational in character and tends to be interpreted in intra- and interpersonal rather than official terms. What might eventually become a social problem often hinges upon restorative responses as parties attempt to influence trouble’s nascent dynamics (Emerson & Messinger, 1991).

Some authors hold that informal trouble is inherently part of organizational life, occupying ‘niches’ in unmanaged areas of the organization (Gabriel, 1995, 1998). In studies of trouble among executives, trouble is predominantly handled as ‘an affair set behind closed doors’, pursued privately through conciliatory attempts, secretive complaining or sabotaging careers (Courpasson & Thoenig, 2010; Morrill, 1995). Similarly, Emerson (2011) describes informal responses aimed at preserving civil relationships through circumventing trouble or negotiating remedial solutions.

The ways in which trouble is informally handled, then, significantly shape its course and transformation. Yet institutions also influence how trouble is interpreted and responded to through interacting with its development. Official trouble-shooters shape ‘indigenous’ approaches and may exacerbate trouble to produce relational turbulence. In the following section, I introduce relational psychoanalytic perspectives that regard trouble not merely as an interpretive construction, but an intersubjective exchange; what begins as local trouble can thus escalate and become contagious, producing relational turbulence across boundaries between individuals, groups and organizations.

Relational Dynamics of Turbulence

In contrast to social constructionist perspectives, psychoanalytic theory puts trouble at the heart of social relations. Psychoanalysis argues that central to the human condition are conflicting impulses, aggressive-destructive as well as libidinal-constructive, producing significant affects, as well as defences that mitigate them. A major strand of psychoanalysis has concentrated on clinical dimensions of trouble. However, beginning with Freud’s ‘Group Psychology’ (1921) and ‘Civilization and its Discontents’ (1930), psychoanalytic concepts have been used to study social phenomena.

The psychoanalytic study of social phenomena was developed particularly by the Frankfurt School sociologists who sought to integrate psychoanalytic insights with critical theory. Their study of political, cultural and social aspects of institutions developed a metatheoretical, hermeneutic science, strongly represented in cultural criticism and organizational studies (Alvesson & Sköldberg, 2000; Gabriel & Carr, 2002; Habermas, 1972 [1987]). Much of this work is orientated to institutional power relations that, though not always directly observable, may be tangibly experienced.

While critical psychoanalytic perspectives have examined all kinds of social problems, contemporary relational idioms have renewed the relevance of psychoanalytic concepts to social theory (Elliott, 2000). Organizations are not merely rational, dispassionate entities, but contain unconscious motivations, emotions and fantasies, linking subjective experience with wider political and societal forces. Fotaki’s (2006, 2009) studies of welfare policy reveal dynamics of social projects as ‘utopian dreamlands’ originating in attempts to overcome psychic tensions and fragmentation. Institutions are thus imbued with symbolic meaning and ‘cultural signifiers’ (Long, 1991), revealed most dramatically in societal conflict (Volkan, 2004).

According to relational psychoanalytic theorists, societal structures provide resources to colonize members’ worlds, but trouble is locally fashioned in intersubjective exchanges (Mitchell, 2000). Crapanzano’s (1992, 2006) anthropological studies reveal intersubjective scenes, reducible neither to subjective experience nor based in objective reality, but involving ‘interlocutory dramas’. Real and imagined encounters between participants and their inner worlds mediate experiences of oneself and others (Quinn & Strauss, 2006). Yet ‘there is … nothing irrational, nothing even fictive about the scene … Both the scene and … objective reality are subjectively experienced’ (Crapanzano, 2006 p. 398).

Trouble is thus inherently relational, ‘located’ in-between persons (Foulkes, 1948 [1991]), at the interstices of social relations. Following Winnicott’s (1953) original theory of transitional space, such intermediate areas are usually associated with positive relations (see McGivern & Fischer, 2012). Involving deep participation in intersubjective relations, transitional space transcends individual subjectivities (Mitchell, 2000), creating a mutually constructed ‘third’ space when participants make genuine contact with other individuals, groups and organizations (Diamond & Allcorn, 2009; Fischer, 2008).

Winnicott (1953) conceived of this as a potential space in which certain boundary or ‘transitional objects’ are emotionally invested with psychological attachment and meaning. Originating in his observation of children and their families, he argued that psychologically safe settings cultivate boundary areas in-between the psyche and its social environment that are important in social and psychological adaptation, and maturation. These areas provide conditions for self-comfort, play and creativity; indeed, all cultural experience may be located in these transitional spaces (Winnicott, 1967).

Yet liminal areas are not always associated with positive dynamics. M. Stein (2007) describes permeable relational boundaries as sites of trouble between front-line employees and customers. When exchanges are excessively hostile, employees experience themselves being poisoned by interactions, producing emotional and sensate experiences that can spill over to contaminate other parts of the organization, producing ‘toxic environments’. Similarly, H. Stein’s (2007) anthropological studies find that certain traumatic experiences can be understood only through embodied sensitivity to organizations’ interior dynamics.

Such intersubjective involvement is central to the idea of transitional space. Group theorists such as Bion (1961 [1994]) and Foulkes (1948 [1991]) applied psychoanalytic perspectives to groups and organizations, revealing deep connections between external ‘objects’ in the social world and their inner, psychological representations (Brown & Zinkin, 1994; Trist & Murray, 1990). Indeed, group-based settings can elicit particularly strong psychological engagement in shared transitional space (Anzieu, 1984).

Relational psychoanalytic theory conceptualizes these relations as involving a dialectical tension between two types of object relations. In a paranoid-schizoid form, powerful ‘split’ connections develop between the inner psyche and part-aspects of the social world. Positive attachments are made to ‘good objects’ experienced as nurturing, but ‘bad objects’ associated with frustration and aggression are rejected as persecutory. This rudimentary form of object relations provides some protective psychological functions but may be highly dysfunctional in social relations, stirring dynamics of hostility, persecution and terror (Volkan, 2004). Psychological and social integration require a so-called depressive form that mediates and restores relations between inner experience and the social world. Both external ‘objects’ and their internalized representations thereby become more balanced, fostering relationships with ‘whole’ persons (Greenberg & Mitchell, 1983).

Transitional space is important in this process; its ambiguous boundary between inner and outer experience provides an area for integrating inner and outer ‘reality’. But might there be a darker side to this positive image? M. Stein’s (2007) toxic contamination involves a scene of excessively permeable relational boundaries, rather than authentic relations. Yet all intermediate space shares its means of exchange across relational boundaries. Intersubjectivity develops through projective identification, a complex mechanism in which a participant unconsciously projects impulses or mental representations into others. Through social interaction, ‘recipients’ are coerced into behaving and feeling in ways congruent with these projections. Participants are thereby ‘made to feel’ emotions and fantasies experienced as partly belonging to the other (Ogden, 1992).

Not all liminal areas have the emotionally invested quality of transitional space, but I question whether this space is necessarily positive. Transitional space has long been regarded an area of creativity and formation. However, this overlooks the possibility that it may involve more turbulent and destructive dynamics, capable of restricting development. As Crapanzano (2006, p. 401) argues, ‘intersubjective captivation’ can be so intense that subjectivity collapses into ‘entanglements of longstanding anger and bitterness’ that so enmesh participants together that they are unable to disentangle themselves.

To summarize, in exploring the idea of trouble, relational psychoanalytic theory conceptualizes it as intersubjective scenes that link inner and outer dimensions of the social world. Whereas trouble situates and localizes relational tensions, relational turbulence escalates these dynamics beyond the original scene, producing contagious ‘ripple effects’ (Barsade, 2002) across experiential boundaries of individuals, groups and organizations.

In the following section, I introduce a democratic therapeutic community as an atypical mental healthcare organization. Its methods of working with people with severe personality disorders in a participative, emotionally invested setting provide an ‘extreme case’ (Eisenhardt, 1989) in which trouble and relational turbulence feature prominently.

Managing Trouble in a Democratic Therapeutic Community

Democratic therapeutic communities (DTCs) may be especially suited for empirically studying trouble and its turbulent repercussions in organizations. Their unorthodox approach is based on strong participation and democratic decision-making, in which the whole community is responsible for managing forms of trouble (Kennard, 1998).

DTCs were first developed in two military hospitals in England, during the Second World War. Pioneering group psychoanalysts such as Bion (1961 [1994]) and Foulkes (1948 [1991]) sought to address institutional dynamics and their effects on soldiers’ neuroses by bringing the whole hospital to function as a large therapeutic group (Harrison, 2000). By focusing on social dynamics rather than individual conditions, clinical ‘trouble’ was treated as symptomatic of social tensions and thus capable of being handled by the ‘community as doctor’ (Rapoport, 1960).

DTCs have been found clinically effective in mental health hospitals, child and adolescent units, and even prisons. As ‘cultures of inquiry’ they develop organizational learning and resilience (Lees, Manning, Menzies & Morant, 2004). Yet their democratic-therapeutic methods challenge professional and managerial orthodoxies (Campling & Haigh, 1999). Baron’s (1987) account of a DTC beset by internal trouble exemplifies how professional power can disrupt therapeutic functioning. They are not readily governed through managerial techniques, but require shared engagement and negotiation (Whiteley & Gordon, 1979).

Their collective meaning and purpose is underpinned by an emotionally invested community ideology (Ormrod, Ferlie, Warren & Norton, 2007). Through the principle of permissiveness, a wide range of ‘acting out’ and interpersonal trouble is tolerated. This is constrained by obligations of communalism and reality confrontation. More especially, democratization reduces power and role differences through sharing equally therapeutic and administrative decisions (Rapoport, 1960). Wider institutional support for their methods is considered important (Hinshelwood, 1987), and may be crucial when handling tensions between individual, community and authorities’ objectives (Spandler, 2006).

A Longitudinal Ethnographic Study

The empirical data for this paper are drawn from a four-year ethnographic study of a DTC and its interorganizational relations with other health, social care and criminal justice services. Ethnography was used to explore participants’ activities, beliefs, meanings, values and motivations – seeking to understand and interpret the social world as its members did (Hammersley & Atkinson, 1995).

As is conventional in ethnographic research, the study’s design was shaped progressively in early fieldwork (Hammersley & Atkinson, 1995). In a first phase, I studied the DTC’s relations with organizational stakeholders across three city conurbations and one rural area. I conducted in-depth interviews (1½ to 2 hours duration) with practitioners, managers and service users, observed meetings, participated in interorganizational (outreach) projects, and followed individual ‘cases’ as residents were referred to and discharged from the DTC. A second phase focused on the residential DTC following a critical incident, exploring its effects on external and internal relations.

In order to be immersed in the field yet retain freedom of movement and thought, I developed an observation-orientated fieldwork role. ‘Observing in order to write’ allowed close attention to dialogue during meetings (Emerson, 2001), yet permitted spontaneous participation outside meetings and in informal activities, which I recorded later. At the time of this study I worked as a healthcare consultant and psychotherapist elsewhere in the region, and I was generally treated as an insider. While this opened considerable access to backstage regions, it presented additional demands on managing the fieldwork role.

As Gans (1999) argues, the ethnographer’s emotional participation may be more important than the role notionally adopted. Especially in settings under stress, ethnographic insights arise through ‘subjecting the self – body, belief, personality, emotions, cognitions’ to the same situation and experiences of participants (Van Maanen, 2011, p. 219). I treated this involvement as important data and sought to increase sensitivity to diverse perspectives through what H. Stein (2007) terms ‘deep listening’. This treats emotional responses and private associations not as insight but as intersubjective material requiring further investigation and reflexive questions about relations with the setting and within oneself (Hinshelwood & Skogstad, 2000; Hunt, 1989). Such material remained a ‘live’ element of fieldwork that continued long after leaving the field, informing analysis and ethnographic writing.

Formal fieldwork involved 76 formal interviews, 195 hours of participant-observation and analysis of written records (see Table 1). As a consultant in the region, I conducted numerous informal interviews with wider stakeholders. These informed my understanding of the field, but are not included in the formal data.

OPEN IN VIEWER

Table 1.

Summary of fieldwork

	Participant-observation		Formal interviews
	Hours	Episodes
Phase 1
DTC staff	42	9	15
Interorganizational outreach projects	51	15	14
Former residents	7	1	4
Organizational stakeholders			32
Phase 2
Residential community	95	10	3
External officials			8
Total, phases 1 and 2	195	35	76

Fieldnotes and interviews were transcribed, and NVivo software assisted data management and analysis. A modified grounded theory method was used for ethnographic analysis, making particular use of key incidents and memo-writing to develop analytic categories (see Charmaz & Mitchell, 2001; Emerson, Fretz & Shaw, 1995). I first used open coding for content analysis and exploration of themes across the whole dataset; I next selectively coded key incidents and rich anecdotes to compare incidents, deepen analysis and explore alternative meanings. At this stage, analytic categories such as ‘trouble’, ‘turbulence’ and other variants emerged inductively as strong themes, which shaped a preliminary analytic framework. I then re-analysed the dataset using focused coding, following Katz’s (1983) strategy of constantly comparing deviant cases, both testing and revising analytic categories. Social constructionist and psychoanalytic perspectives were used to examine the emerging analysis, and also ensure interpretations were strongly over-determined by the empirical data (Gabriel, 1999). An empirically grounded framework was thus gradually developed, moving iteratively between data and theory (see Eisenhardt, 1989).

Restorative Turbulence and the Creation of Formative Space

Ruptured Transitional Space

The DTC’s ability to handle turbulence was transformed by a critical incident that traumatized the community. The incident involved two former residents, Mark and John, ¹ who had recently completed 12 months of DTC treatment and, unknown to DTC staff, moved together into a rented apartment. The couple formed an intimate relationship while in the DTC and, although they had experienced difficulties in their relationship, community members believed they were resolving them, with the help of a weekly ‘transition’ group for recently discharged residents.

Yet shortly after moving in together, the couple had a violent, drunken row – and Mark stabbed John to death in a frenzied attack. Both men were regarded as having been successfully treated by the DTC. Yet within just weeks of leaving the community, Mark was charged with and subsequently convicted of John’s murder.

The impact upon the DTC was profound. Residents and staff alike were horrified by the attack and shocked by Mark’s seemingly uncharacteristic violence. He had been regarded as a gentle and sociable member of the community, and he was felt to have been one of its least aggressive members. The incident overwhelmed the community with grief, remorse and intense anger, fuelled by recrimination. Resident and staff groups found themselves internally divided, split through torn emotions and loyalties to the two men.

The pressure has been immense … enormously traumatic. We just didn’t know who to be angry with … there has been a great deal of soul-searching – and such fallout in the staff team … a chain reaction of blame. (DTC therapist)

Relational trouble does not come in any stronger form than homicide. For a community whose identity and purpose was not merely containing trouble, but the successful treatment of personality disorders, John’s murder threw the community into crisis. The incident traumatized its members, producing a psychological injury that overwhelmed their capacity to cope with now heightened emotions and reactions. The death of a colleague can produce grief reactions analogous to death in the family (Hyde & Thomas, 2003), but the violence of homicide in therapeutic settings has more devastating psychological and social effects. By violating the social prohibition against extreme violence, the ‘elusive boundary between fantasy and reality, wish and action’ is eroded (Erlich-Ginor & Erlich, 1999, p. 203), rupturing the basic trust needed for therapeutic work (Erikson, 1985).

This idea of a ruptured boundary is important because the homicide altered DTC members’ sense of a shared, protective space in which the dynamics of trouble might be safely handled. It disrupted members’ belief in the community as an essentially safe space, protected from dangers of the outside world. As aggression was felt to lose symbolic aspects, coercion, scapegoating and threats of violence became frighteningly real. In its traumatized state, the community’s ability to manage ‘ordinary’ trouble was overwhelmed. DTC leaders suspended the therapy programme, resorting to around-the-clock emergency meetings in an effort to manage barely contained levels of turbulence.

The DTC’s internal dynamics were compounded, however, by the homicide’s repercussions within the policy environment. Government representatives and health authorities reacted with extreme anxiety and, as one official described, ‘panic’ as a high-profile national initiative threatened to become a policy embarrassment:

The homicide is telling: discharge planning was done by other punters … [yet] these are some of the most dangerous and manipulative patients that there are. It’s literally the case of ‘the lunatics running the asylum’. It’s scandalous … the project is unsafe. (Senior official)

Yet the homicide need not have been disastrous for the DTC. Homicides are not unusual in mental health services, and rarely lead to organizational closure. Although the DTC was in crisis, therapeutic communities are prone to extreme fluctuations in morale and functioning, and they may be remarkably resilient (Hinshelwood, 1987). Engaging with the painful reality of the homicide might therefore have led to integrative-reparative (depressive) grieving and recovery, leading to individual and organizational learning. Instead, in the homicide’s aftermath, a highly defensive mode developed in which turbulence took a perverse and destructive form.

Perverse Turbulence and the Development of Perverse Space

The DTC’s capacity for reflection and learning from experience was undermined by the homicide’s repercussions at local and national levels. Rising interorganizational conflict developed, focused on the community’s distinctive model of handling clinical trouble. Fearing political fallout, authorities attempted to avert blame, imposing externally prescribed risk management techniques.

Of course the homicide caused a great furore. The chair of [the national commissioners] began to get anxious he was going to end up with a homicide inquiry … [and] panicked into commissioning a risk report of the entire national service … Did it make [local authorities] anxious? Oh my God, yes … [they] already had two homicide enquiries going on, both of which will severely criticise [them]. (Senior official)

Under what they perceived as threat of severe official sanction, and a ‘witch hunt’, community members experienced extreme, persecutory anxiety. Residents became preoccupied with the survival of the community and the security of their places within it, while staff members feared for their professional careers. Such anxieties disrupted the community’s capacity to think reflectively, concentrating a sense of failure on the DTC’s handling of clinical risk:

There is something about the unknown that unnerves them so the [officials] go to what they think is solid ground and demand information on patient behaviour. But residents should not be made to feel like subjects of risk, debris of pathology … slapped down to keep their feelings inside. (DTC therapist)

Critical external evaluations compounded feelings of guilt over John’s murder, and a sense of shame in the DTC’s failure to live up to its ideal. Whereas its democratic-therapeutic handling of trouble had previously been regarded a source of pride, it quickly became a cause of humiliation. Escalating recrimination and anger about the DTC’s responsibility for the homicide distracted the community from mourning its losses – even to speak of them:

Miranda [pointing to graffiti] says John had written that shortly before being murdered by his partner … She suddenly seems uncomfortable. I want to ask how it has affected the community, but stop myself. It feels intrusive to ask, as though it’s something very personal. They half-joke: ‘however, we don’t talk about that’. (Fieldnotes)

Unlike mourning, which involves internally directed feelings of guilt, important in psychological reparation, shame is externally focused, involving exposure to another’s negative evaluation. Failing to live up to one’s ideal self-image produces feelings of mortification and rage towards others, inhibiting the ability to grieve (Scheff, 1994). Avoiding talk of John’s murder may have temporarily reduced emotional distress, but the aftermath seemed to pervade the community: ‘going on and on, affecting everything we do’. For residents, it produced ‘months (of) hell … no protected time, no retreat … it’s like a year in Beirut’.

These experiences did not produce the relative cohesion of paranoid-schizoid splitting, however, which might have galvanized the community internally against an external adversary (Volkan, 2004). Instead, preoccupation with self-preservation produced a terrified and beleaguered community, ‘assaulted, undermined and torn apart (by) a very inflammatory and destructive engagement [with officials]’.

Distracted from its primary task, the community bypassed therapeutic ‘remembering, repeating and working-through’ (Freud, 1914) the events of the homicide. Yet its dynamics were re-enacted in a highly destructive mode of turbulence in which sado-masochistic patterns dominated. Central to this was residents’ deep resentment of an emerging betrayal by staff members who no longer seemed to believe in the DTC’s democratic ideals. Despite rhetorical protests against authorities’ efforts to steer the management of trouble, staff quietly adopted managerial perspectives. Confidential details of residents’ activities were covertly passed to officials and other agencies, ‘covering ourselves, in case the shit hit the fan’. DTC leaders pressurized the community to deal more rigorously with presumed ‘troublemakers’, threatening more challenging residents with expulsion. Residents were acutely sensitive to the emotional changes in community relations, perceiving a menacing shift in attitude:

There is not one community here – there are two. I really don’t trust staff – you can’t call it a community when you can’t talk with them about anything … you can’t call it democratic. (Night-time discussion between residents)

Instead of merely retreating from democratic participation, residents engaged in impassioned conflicts, furious at staff attempts to exert hierarchal authority. As democratic decision-making was undermined, a sense of shared, mutually constructed space was eroded, and the community became an area where staff experienced ‘no longer having moral authority’. Conflict between staff and residents escalated as one’s survival in the community seemed dependent on overcoming others’ aggression.

We picked up the pieces every morning after a bloody bomb had gone off overnight. (DTC therapist)

In contrast to restorative turbulence, then, in which intersubjective relations have a moderating effect, in this perverse mode parties colluded to produce an enmeshed and intractable conflict. Members shared terrible images of being under siege in war-torn conflict, of being trapped in a death train, and of an impending Holocaust. These images were tangibly experienced and ‘made real’ through members’ intersubjective exchanges with each other. Persecutory anxiety created a shared experience of persecution at the hands of sadistic officials. Residents described horrific nightmares in which staff members appeared as their abusers. For their part, distressed staff members described in tears their experiences of such accusations as attacks upon their innermost selves: of having their ‘livers ripped out’.

It is so severe, all this battering. It’s so thin – the skin … it’s a crucifixion. I’m overflowing … with feelings. I just can’t seem to escape or forget the hurt that residents put me through. I’m constantly under attack. (DTC therapist)

With community relations so destructively engaged, the DTC’s capacity to restore therapeutic functioning collapsed. Through fragmented splitting, the DTC had degenerated from a perceived ‘jewel in the crown’ to a painful lost cause, yet one that members felt compelled to engage in ‘like a moth rushing into the flame’, amidst experiences of betrayal, sadistic persecution and fear of annihilation.

We tried to hang onto some semblance of authority. But the power the [authorities] were exercising over us diminished our ability to manage the attacks from residents. The community became disembowelled … Every day was like a nuclear reactor without the container – there was nothing around to hold the explosion. (DTC therapist)

Despite positive government-commissioned independent evaluations (Fiander, Burns, Langham & Normand, 2004), officials finally lost patience and closed the DTC, less than four years after its launch as a promising policy initiative.

Discussion

This empirical case reveals intersubjective ‘third’ space to be fundamental to the DTC’s handling of trouble and turbulence. Although turbulence is disruptive to transitional space, in a restorative mode the DTC’s methods were adaptive, resilient and capable of making productive use of turbulence through collaborative reflection and reparation. This restorative capacity was undermined following the homicide as conflict between the DTC’s and authorities’ management of trouble undermined the DTC’s methods, producing a perverse mode of turbulence.

A key question is why the homicide had such a destructive impact upon the DTC. Responsibility for the incident formally lay not with the DTC, but other agencies providing aftercare. Despite official concerns about risk management, independent evaluations had concluded the DTC was ‘clinically safe’ and effective in treating a problematic population (Fiander et al., 2004). So why was this incident such a turning point?

Conclusion

In linking social constructionist perspectives with contemporary relational psychoanalytic theory, this study has explored nuances of trouble and its potential for turbulence as part of an organizational, intersubjective ‘scene’. Studying how trouble is generated, exchanged and managed in organizations reveals dynamics of relational turbulence capable of escalating across experiential boundaries in individuals, groups and organizations.

This article contributes to the literature by introducing the concept of relational turbulence across boundaries, and through developing a model of formative and perverse spaces created by this turbulence in organizations. This model extends the psychoanalytic concept of transitional space, revealing turbulent and destructive dimensions, hitherto overlooked. Whereas restorative turbulence has a regulating function and may create formative spaces in which authentic, mutual relations flourish, perverse turbulence undermines organizational functioning and can lead to intractable perverse spaces in which aggressor–victim dynamics dominate.

As Eisenhardt (1989) argues, extreme cases such as this illuminate dynamics which, although present, may be harder to discern in other settings. In human service organizations such as healthcare, schools and prisons, trouble and its turbulent effects are inherent in organizational life. When handled informally, they can have important self-regulating dynamics, creating formative spaces that are important for productivity. Conversely, formal control may not merely attenuate transitional areas, but produce perverse spaces from which participants’ abilities to disentangle themselves are uncertain.

This study has implications for wider organizational research beyond such settings. Liminal areas may play a more vital part in organizations than is usually recognized. They can be a significant resource for creativity, as well as for dramatically ‘unleashed’ rebellion, forcing organizational transformation (Courpasson & Thoenig, 2010). This case study suggests that future research should pay greater attention to their dynamics as key relational spaces, where individuals, groups and organizations come together in interesting and often unexpected ways.

Bringing organizational ‘third’ space into focus has revealed an intersubjective scene in which turbulence plays a key role. This scene illuminates exchanges between ‘personal troubles of milieu’ and their handling as ‘public issues of social structure’ (Mills, 1959). Within ‘third’ space, relational turbulence may be important for productivity, though it is also potentially immensely destructive.