Stuck in separation: Liminality, graffiti arts and the forensic institution as a failed rite of passage

Forensic psychiatric institutions are tasked with both containment and transformation, with securely policing the border between institution and society and readying patients for return to the community. Forensic institutions can thus be theorised as a form of ‘rite of passage’, engaged in a process of transformation which both navigates and demarcates social limits. This article contributes to literature on risk and control in clinical institutions by offering a novel theoretical synthesis of features of rites of passage and liminality, as facilitated by an art project in a forensic setting. Through the prism of the Graffiti and Wellbeing Project (GWP), an arts initiative, we explore the ways in which forensic institutions thus offer or impede opportunities for transformation. The project engendered a space for the transformation of difficult emotions and histories through the medium of art creating a liminal space of transformation within the confines of a secure institution. Drawing on Douglas, Kristeva


INTRODUCTION
Forensic mental health services are complex institutions, sitting between the psychiatric and criminal justice systems, linked to both and yet fully located in neither. In the UK, patients can be referred to forensic institutions from both the criminal justice and mental health systems. Whilst most patients will have an 'index offence', meaning they have been convicted, some patients are instead referred directly from mental health services without a conviction, having been deemed too risky or complex for generic mental health services. Following the move to short-term crisis care in inpatient psychiatric services, forensic services have been a site of 'reinstitutionalistion' of those with long-term mental health needs (Kaliski, 2013;Priebe et al., 2005). This complexity of context is matched by a tension in purpose. Forensic units are tasked both with containment and transformation; they must both securely police the border between patients and society as well as ready patients for return to the community. This latter function is more central for forensic services than either prison or psychiatric inpatient facilities. In forensic services, discharge is on the basis of risk assessment, not length of sentence. Whilst prison can aim to be rehabilitative, in the majority of cases it is not necessary to demonstrate change in order to exit the institution. In addition, the offending history of most of the patients in forensic services adds additional layers to discharge, in comparison with generic psychiatric services. For some patients, discharge involves multiple agencies and must be sanctioned both by psychiatrists and the justice system, attracting gatekeeping by both systems. The pressure on forensic services to demonstrate and facilitate change is therefore more intensive than either prison or psychiatric services.
The tension between these different purposes within forensic services has often been characterised in terms of 'care' and 'control' (Mullen, 1993) familiar from other disciplinary institutions (see Philo & Parr, 2019), and also reworked as 'security vs therapy' (Curtis et al., 2013;Tucker et al., 2019) and 'risk vs recovery' (Drennan & Alred, 2013). A common critique of forensic services is that control/ security/risk tends to preside over care/therapy/recovery, with forensic services offering few genuine opportunities for the latter (Curtis et al., 2013;Drennan & Alred, 2013). This article contributes to literature on risk and control in clinical institutions by offering a novel theoretical synthesis of features of rites of passage and liminality, as facilitated by an art project in a forensic setting. We add to the critical literature formulating alternatives to a calculable risk approach to institutions for people who have been marked as threatening or dangerous. Risk has been argued to be the opposite of trust (Mayer, Davis & Schoorman, 1995), taking an orientation to the future which tries to control and minimise possibilities for action. A danger of this mindset is that minimising possible negative outcomes also limits possibilities for change and growth. Positive risk taking, for instance, is a growing approach within clinical practice which aims to instead begin with the understanding that some risk is needed for any change to take place (Drennan & Alred, 2013).
Whilst control and security are primarily static concepts, risk, therapy and recovery contain the possibility of transformation-through risk reduction, therapeutic change or recovery from mental | 3 STUCK IN SEPARATION distress-and are thus the concepts through which the transformational purpose of forensic services is filtered and realised. In this article, we look afresh at how the forensic institution manages, facilitates and impedes transformation and transition. There are perhaps three distinct and overlapping notions of 'transformation' which require some clarification. Firstly, transformation could be seen as a conservative process of progression through a stage or ladder of pre-defined goals, to be able to emerge back into society as no longer a 'threat' to self and others. Secondly, transformation could also be used in a therapeutic sense, to mean an individual's movement towards awareness; a subjective process of healing from trauma, acknowledging that to transform pain and distress does not delete it (Herman, 1992). Transformation is not erasure. Finally, transformation is used within a liberatory framework, conceptualised as conscientization (Friere, 1970) towards collective emancipation and liberation from the forces of oppression. Forensic services can be seen to be attempting to engage in the first two of these forms of transformation through the process of readying patients for discharge into the community. We argue in this article that these are both severely limited and blocked by the practices of the institution. The final sense of transformation, as liberation, we hold is impossible within a carceral forensic setting which removes liberty and impedes collective consciousness raising. For this reason, we use 'transformation' in the sense of change, from one state to another, and thus interchangeably with 'transition', throughout this article.
Rather than taking the lens offered by the language of services-of risk, recovery, therapy or security-we instead theorise forensic institutions as being engaged in a specific kind of process, a rite of passage (Van Gennep, 1960). In taking this processual approach, we aim to move beyond thinking of forensic institutions engaged in either care or control, but instead consider the ways in which the institution offers or impedes opportunities for transformation. We explore the conditions which help promote transformation through the concept of liminality (Stenner, 2018;Turner, 1969;Van Gennep, 1960), whilst also holding in mind the role that forensic institutions play in maintaining and policing social and ethical boundaries, in acting as a limit. We explore these issues through the prism of The Graffiti and Wellbeing Project (GWP), an arts project which, we argue, was successful in creating a liminal space of transformation within a forensic service. In doing so, the project highlighted the ways in which the usual practices of the forensic unit impeded transformative practice, leading to the theoretical reflections outlined below. We first explore the concepts of liminality and rites of passage in more detail, before outlining the GWP, then exploring three ways in which this project disrupted the usual space of the forensic unit.

LIMINALIT Y AND RITES OF PASSAGE
Rites of passage are specific ritual and social forms which accompany and facilitate transition between different states. The term 'rites of passage' was coined by Van Gennep (1960) in his classic anthropological study. He explored rituals associated with passing thresholds, boundaries and the transition between states, such as that between childhood and adulthood. Van Gennep identified three stages to transitional rituals: I propose to call the rites of separation from a previous world preliminal rites, those executed in the transitional stage liminal (or threshold) rites, and the ceremonies of incorporation into the new world postliminal rites. (p. 21).
These three stages can be exemplified by the processes of traditional marriage. An engagement marks the separation of the couple from the 'previous world' of their families; betrothal is a liminal period in which the couple are in an in-between state, neither married nor single, and finally, marriage marks the 'incorporation' of the pair into the community as a married couple. Van Gennep noted that in many cultures, people in liminal states are physically excluded from the community, before being reincorporated. Of these three stages, the liminal has garnered the most attention. Turner (1969) expanded greatly upon Van Gennep's original ideas to argue that liminal situations are a form of 'anti-structure': It is as though there are here two major "models" for human interrelatedness, juxtaposed and alternating. The first is of society as a structured, differentiated, and often hierarchical system of politico-legal economic positions with many types of evaluation, separating men in terms of "more" or "less." The second, which emerges recognizably in the liminal period, is of society as an unstructured or rudimentarily structured and relatively undifferentiated comitatus, community, or even communion of equal individuals. (p. 96) Turner, therefore, saw the liminal period as being characterised by non-hierarchical relationships and a dissolution of normative roles. In this view, the liminal phase, 'betwixt and between' (p. 95) exit and re-entry, dissolves social structures and norms whilst rites of separation and rites of incorporation act to reaffirm existing social structures and order. Performing rites of separation marks boundaries as a precursor to crossing them; blessing a door before entering a room, for example, is a classic purification rite marking the threshold before overstepping it. Rites of incorporation, meanwhile, renew the structures to which the person returns from liminal space. A marriage ceremony thus marks the transformed state of the couple as well as acting to reinforce the institution of marriage. Turner saw the relationship between these two states as dialectic; he did not argue that we should move to permanent liminality, but saw structure and liminality as co-dependent forms of social process between which societies fluctuate.
The idea of liminality as that which enables the emergence of novelty, and ultimately transformation, has been further developed by Stenner (2018), who uses an extended sense of the word to mean 'transformations within any given form of process, or to any transition between given forms of process'. (p. 262) Here, liminality is seen as a processual form which enables 'pattern shift'. He distinguishes between 'devised' and 'spontaneous' liminality, both of which can operate on individual and social levels. 'Devised' liminality might include art, fiction or performance (Stenner & Zittoun, 2020), which can create the same conditions of 'non-structure' identified by Turner and thus provide a kind of 'liminal affective technology' (Stenner, 2018), a cultural form facilitating transformation.
Liminality, therefore, is used in three senses, all of which are of interest in our exploration of forensic services, firstly, that of being between two states; secondly, of being beyond a threshold or boundary; and finally, of being a transformative process. In the rituals of transition explored by Van Gennep and Turner, liminality is generated in situations outside of the usual bounds, roles or norms of society. It is within these in between and beyond situations, that transformation is able to take place, a movement from one way of being to another. A threshold, whether literal or symbolic must be first passed and then returned back over, to mark the end of the liminal period. Liminality, thought about in spatial terms, therefore, is often used to mean 'on the edge', as well as 'in between'. Liminal situations are both 'in between' two thresholds, in temporal terms, and 'beyond' a boundary of some kind, in spatial terms.
Turning back to the forensic institution, liminality understood in these three ways is both relevant and yet also somewhat alien to the practices of forensic units. These are certainly liminal institutions in the sense of being on the edge, or beyond the threshold: forensic institutions are-in the prison and the asylum-descended from Foucault's (1965Foucault's ( , 1977 disciplinary institutions, formed to demarcate the edges of post-Enlightenment society. Secure hospitals house people who have transgressed two social boundaries: crime and madness. The ostensible purpose of the institution is to manage patients' | 5 STUCK IN SEPARATION transition from transgressors to citizens, enabling 'reincorporation' back into the community. The institution is thus a liminal space, beyond the edges of the community, and set up to facilitate liminal processes of transformation and transition. It is not, however, easy to find the features of a liminal space in the day-to-day practices of forensic units. Rather than 'communitas', we instead find strictly defined roles and hierarchies (Norton & Dolan, 1995), a tendency for order and structure over 'non-structure' (Mann et al., 2014;Simonsen & Duff, 2019). Forensic units do not perhaps regularly offer a liminal space, in the sense of 'nonstructure' which facilitates transformation. Below we explore a liminal space which was successfully generated in forensic services: the GWP.

GRAFFITI AND WELLBEING PROJECT
The GWP in North London Forensic Service was a project commissioned by the Occupational Therapy service and led by Ben Wakeling, artist and founder of the Outsider Gallery London. The empirical material for this paper came from an evaluation of the project (see McGrath et al., 2019). This was a mixed-methods process evaluation, consisting of 'active participation' (Spradley, 2016, p. 61) by one of the evaluation team to gain an understanding of the practice and dynamics of the sessions. The observer made field notes responding to prompt questions to create a consistent record of the sessions. They attended seven sessions altogether. There were also staff interviews (n = 3), and a survey of the wider staff community (n = 14). The interviews included questions related to staff experiences of the art project, and the final artwork, any changes they observed in terms of relationships, the atmosphere of the ward, comparisons with other therapies taking place in the ward, and their thoughts on the role of art in forensic settings. There was therefore a range of views gained from staff closely involved in the project, to those who observed from a distance.
Each iteration of the project lasted 12 weeks, located on a different ward. The first six weeks took place in a closed room, where the facilitator worked with participants to build trust and confidence in working with visual materials. The facilitator named this stage of the project 'expressive roaming': participants were encouraged to draw from imagination, exploring feelings and memories, and finding ways to express these visually. This stage provided the relational and artistic base for the second stage: spray-painting an external wall located in the centre of the unit, repeatedly painting over the imagery produced. Whilst the wall was never a 'finished' work, photographs were taken of the imagery produced and some were displayed back in the ward. In addition, participants had the opportunity to exhibit in the Outsider Gallery London, such as with this example (Figure 1): Participants in the evaluation noted the marked difference between the project and the practices of the rest of the unit. As one staff member commented: There isn't really a comparison with other therapies as these are often focussed on their illness, offending and education [...] having an external facilitator who isn't part of the core clinical team means that patients interact differently and can develop relationships that allow expression of another part to themselves.
(Survey response) Highlighted here is the perception by staff in the unit that the GWP took a fundamentally different approach to the rest of the service. The facilitator's position as an outsider, a non-clinical expert, enabled a space of freshness or liminality, in which participants' histories and diagnoses were no longer simply limiting factors but rather that the space was one of potential: [The facilitator] was an outsider, that made a huge difference in the dynamic and the environment of the room. Uhm, and it was a bit more relaxed or something. It was kind of nice.
(Staff interview) The environment is less clinical and it's great to see people's individual creativity and motivation whilst working on the project (Survey response) There was a sense that the project was a disruption of 'business-as-usual' what patients 'normally' do, confined by the individualism of ward life , versus what the 'less clinical environment' of the GWP, defined by 'creativity and motivation' rather than 'illness', made possible: The patients don't normally associate with one another. [In the GWP] we got a big piece of paper and we got them to work on the same piece of paper which was really nice to see them doing that.
(Staff interview) [I] think some patients found a new sense of appreciation for one another. People became surprised about how expressive and talented they were becoming. It's not easy to work in a group, complex emotions and past history can easily get in the way (Staff interview) The liminal qualities of the GWP are evident in these descriptions. We can see here the ways in which the project created a liminal space of 'communitas', which operated differently to the usual hierarchies and diagnostic lens of the forensic institution.

STUCK IN SEPARATION
Within this liminal space, staff noted that the project allowed rather than suppressed possibilities, in contrast to the usual risk-driven practices of the service: I suppose somewhere that usually we are telling people not to do things and we spend 90% of our day saying you can't do that, you can't do this and then we can go well actually you can do this and we encourage you to do that so that's quite nice.
(Survey response) The medium of graffiti in itself seemed to offer a permission to regain an 'adolescent' liminality linked to the experimentation with and transgression of boundaries: [the facilitator] explained the theory behind patients using the wall outside to draw -he expressed that it took people back to their childhood/adolescence almost as if you were getting in trouble for doing so. He believed that this was a way to express any feelings that could not be expressed verbally.
(Participant Observation) This transgressive quality was also evident in comments reported from higher management in the trust: [we have had] some negative comments from higher Trust management staff not understanding the importance of the project and stating it looked "messy". (Survey response) On the graffiti wall itself, it was noticeable that through the ongoing process of expression, wiping and renewal, what was produced on the wall moved from early written expressions mainly in text, including swearing and violent language, to more conceptual visual imagery, often poignant and powerful ( Figure  2): In this succession of images, the early, more chaotic expression of text, lines and shapes is slowly transposed into more defined and symbolic use of imagery marking a visual record of the development of the participants' 'visual voice'. As the facilitator said: there is a development that occurs when we keep painting.
[…] Sometimes the aggressive, initial thoughts, excitement, immediate voice comes through then later those challenging raw thoughts start to develop into more understandable visual language. […] F I G U R E 2 Progression from text to imagery on the Graffiti Wall Time is always needed and becomes a fertile ground for raw expressive emotions to be understood.
(Facilitator, Personal communication) Taking three sets of comments from the evaluation as our starting points, in the remainder of this article we explore the ways in which the GWP project exposed how the usual practices of forensic institutions act to refuse and dispel liminality. Rather than an analysis of the evaluation data, we offer a set of theoretical reflections and arguments inspired by these features of the GWP. The first set of comments focus on the 'messiness' of the project, which exposed the usual order of the space, drawing on Douglas ' (1966) classic work on purity and danger. The second set of comments centre on 'expression' of emotions and histories, which prompted reflection on what was usually sublimated in the institutional space, drawing on both Kristeva's (1982) concept of the abject and Bakhtin's (1984) carnival. Finally, we explore comments which highlighted the 'outsider' nature of the project, considering the ways in which the project was able to maintain liminality within inhospitable structures of the unit.

MESSINESS: PURIT Y AND ORDER IN THE FORENSIC UNIT
[we have had] some negative comments from higher Trust management staff not understanding the importance of the project and stating it looked "messy".
(Survey response) In this quote, a staff member reports feedback from the senior management of the Trust which positions the graffiti wall, and by extension the GWP, as 'messy'. This reported comment encapsulates a central tension which the project exposed between the qualities of transformative liminality embodied in the GWP, and the highly ordered nature of forensic spaces. The 'management' figures here seem to represent both authority and disconnection, taking a dismissive approach to the visible 'difference' that the graffiti wall, and by extension the GWP, represents.
Change requires, as Turner (1982) put it, 'the subjunctive mood': possibility, uncertainty and contingency, an openness to mistakes. Forensic institutions are instead governed by risk which aims to shut down, rather than open up such possibilities (Drennan & Alred, 2013). A concern with limiting risk is often translated into seeking certainty, gearing services and practitioners to restrain and limit possibilities for patients and activities that could entail risk to self and others (Barker, 2012). Services have thus been argued to produce static, limited spaces, empty of risk but also of the very materials needed for meaningful 'recovery' . These processes of ordering and tidying are deeply ingrained in forensic practice. In an ethnography of a new generation forensic mental health service in Denmark, for instance, Simonsen (2020) observed that whilst the building was designed with an open plan area to facilitate greater agency and more fluid interactions between staff and patients, staff engaged in intensified tidying and ordering activities to demarcate the space, through clearing objects and moving furniture. As with the senior management in the quote above, 'messiness' and ambiguity were not tolerated and prompted a drive to tidy and order the space.
We have moved here between literal and conceptual examples of 'mess' and 'order', scuttling from concrete objects-walls, chairs-to abstract policy. We are influenced by Douglas' (1966) framework for understanding order, tidying and cleaning as operating simultaneously on material, relational and symbolic levels. Particularly relevant is her contention that 'dirt', that which needs cleaning or tidying, is best understood as 'matter out of place', something which is not properly ordered, is ambiguous, or | 9 STUCK IN SEPARATION crosses a boundary. What is counted as dirt, therefore, is something left over from the construction of order, a 'residual category, rejected from our normal scheme of classifications' (p. 10). Douglas highlights the example of shoes, which are 'dirty' when they are in a house, and even more 'dirty' if they are on a kitchen table, but not at all 'dirty' when they are outside; 'dirt', therefore, 'exists in the eye of the beholder' and can be best understood as 'essentially disorder' (p. 10). She argues that purity practicescleaning, tidying and sanctifying-function to create and maintain order in the face of disorder: Ideas about separating, purifying, demarcating, and punishing transgressions have as their main function to impose systems on an inherently untidy experience. It is only by exaggerating the difference between within and without, about and below, male and female, with and against, that a semblance of order is created (p. 10).
Similar ideas govern symbolic dirt or pollution, that which is seen as attacking the social order or the 'body politic' (p. 133). These arguments, therefore, apply as strongly to people who disrupt, trouble or transgress the social order, as they do to polluting objects. Transgression, incursion, marginality and ambiguity are all identified by Douglas as relations to the social order which can invoke ideas of social 'dirt', 'danger' and 'pollution'. These are all relations which trouble or complicate existing boundaries or patterns, which blur clean lines or ignore them completely. Douglas further claimed:'...if uncleanness is matter out of place, we must approach it through order. Uncleanness or dirt is that which must not be included if a pattern is to be maintained' (p. 40).
Those living in forensic services, having transgressed two major social boundaries, can thus be seen as occupying a doubly precarious, symbolically 'dirty' position, seen as doubly dangerous and potentially polluting to the rest of society. In addition to the social and symbolic transgressions of madness and crime, forensic services also gather higher numbers of people from socially denigrated groups, including those marginalised by class and race (West et al., 2014). 'Purity' and order are, indeed, concepts also laden with racist and colonial connotations (Duschinsky et al., 2017). Considering the multiple layers of the ways in which forensic patients are positioned as social 'dirt', as being people 'out of place', it is perhaps unsurprising that forensic services also 'approach [their work] through order' (p. 40). Order and purity can indeed be seen to characterise both security/risk and therapy/ recovery practices in the forensic system.
Firstly, security architecture is designed to maintain and protect boundaries, to separate patients from the world and from each other, to demarcate the boundaries of the unit, to police and maintain the spatial and social order. Whilst forensic spaces are generally located on hospital sites, in the UK they are also surrounded by high, prison-like fences, and operate with high levels of security. Risk is related to security but is more nebulous; it permeates through all policies and practices in the forensic system (Drennan & Alred, 2013). As discussed above, risk is the primary lens through which progress is measured in forensic services, and sufficient reduction in risk must be demonstrated before discharge. 'Recovery' in the forensic service is primarily measured by movement along the care pathway and is thus inextricably linked with a patient's assessed level of risk. A patient who is seen as more recovered, would be granted more leave, have more freedom, whilst non-compliance with the rules, such as drugs or violence, results in a withdrawal of these privileges . By demonstrating an ability to conform to order inside the institution, patients are slowly granted the right to comply with the larger social order outside the bounds of the institution. In this way, both security/risk and therapy/recovery can be seen as attempts at ordering the 'disordered'. To be seen as 'recovered' a patient must demonstrate that they have been reordered by the institution to conform to the social pattern they previously transgressed. To be kept 'secure' the unit itself must have visible, strong barriers which reinforce the boundaries between the 'ordered' world and the 'disordered' people within.
Engulfing the whole forensic system, therefore, is a drive for order. We can view forensic institutions as being engaged in an attempt to reorder the 'disordered' back into the social pattern. As Douglas (1966) points out, this is ultimately a process of pattern maintenance, of ensuring that the 'matter out of place' does not continue to exist and thus threaten and trouble the stability of the social pattern. Part of the function of the forensic institution, therefore, is to maintain the very limits which patients have transgressed. The limits being enforced by the institution are not those which we can easily argue should be dissolved, as they include serious violence and harm. In a process of transition; however, the affirmation of order and boundaries should appear at either end of the process-in rites of separation and incorporation-but they should be less visible during the middle, liminal phase.
The forensic institution can hence be seen as being as stuck at the beginning of a process of passage, re-performing the rites of separation which affirm the boundary between society and a space of transition. There is little room here for the liminal to emerge; a dissolving of roles or rules in this context would be viewed as a threat to the boundaries being reaffirmed and reinforced through the institution. Without liminality though, it is difficult to see where transformation can emerge, as 'purity is the enemy of change, of ambiguity and compromise' (Douglas, 1966, p. 164). The GWP, therefore, can be seen to be 'messy' in several senses. Painting on the walls disrupts the clinical cleanliness of the institution, producing an ambiguous, never finished space for patients to write their own stories onto the walls of the institution. Relationally, the 'communitas' of the project, characterised by equal and trusting relationships, also 'messes' the usual hierarchical order. As an outlet and container for the expression of difficult emotions, experiences and histories, the GWP also disrupts another norm in the institution, as we explore below.

EXPRESSION: TRANSFORMING THE ABJECT
I feel as if this will tell their story of the past and what they want for their future also a way where they can pour out their feelings even if it takes 10 minutes.
using the wall outside to draw […] it took people back to their childhood/adolescence almost as if you were getting in trouble for doing so. He believed that this was a way to express any feelings that could not be expressed verbally.
(Participant observations) Another feature of the GWP highlighted by staff was the way the project enabled patients to openly explore and express their experiences of distress, madness and violence. At first glance, this does not perhaps seem like a revolutionary act; distress and difficult histories are, after all, what bring people into the forensic institution in the first place. The GWP project here reveals another tension in the processes of the forensic institution, that whilst distress, violence and madness are the very reasons for the existence of the institution, and for patients being placed there, these histories are rarely directly addressed (Brown & Reavey, 2016;Drennan & Alred, 2013). Direct expressions of distress or violence are, as we have previously discussed, indeed often understood primarily through the prism of risk, and can lead to greater restriction. Violence and distress, are, after all, expressions of the transgressions which the institution is engaged in trying to reorder. As well as the forensic unit, and the people who live there, being positioned as social 'dirt' which needs to be properly ordered before being allowed back into the main pattern of society, there is again 'dirt' within the system of the unit itself. To sustain order in this most 'disordered'

| 11
STUCK IN SEPARATION of places, with people who have been categorised as themselves 'disordered', means that much must be disallowed, rejected and sublimated. Whilst many, if not most, forensic patients have violent histories, Drennan and Alred (2013) argue that these histories are only marginally engaged with in secure services. Violence is mainly addressed through anger management (Walker & Bright, 2009), with little exploration of the linkages between trauma, distress and violence which often feature in the histories of forensic patients. We have here a paradox, that violence is often a reason for the transgressions which have placed forensic patients in the service, and yet this violence is rarely explored in depth. Instead, these violent histories are tidied away. Sexuality is similarly treated primarily as a threat, something to be suppressed and controlled, rather than acknowledged or explored (Dein et al., 2016). Finally, mental distress itself is often tidied away, rather than openly explored; the priority is to 'stabilise' patients in the present with little attention paid to histories of distress and trauma (Brown & Reavey, 2016). It is notable that distress, sexuality, violence and emotion itself are all elements of life which trouble and disrupt rationality. Again, the very transgression which is the reason for the institution-distress, or 'unreason' (Foucault, 1965)-is also that which is most eagerly suppressed. We can see the various ordering attempts in the unit, whether limiting through risk or shaping through recovery, as attempts to further eradicate, or purify, these lingering irrationalities.
Yet these are not experiences and histories which are easily tidied away. Despite the drive to order in forensic institutions, these are not calm and collected places. Whilst patients often describe secure facilitates as 'boring', devoid of activity and interaction (Farnworth et al., 2004), they are also places which are often turbulent and tense, in which staff grow accustomed to monitoring minor changes in the atmosphere of the wards to try and pre-empt and manage disruption and conflict . This combination, of boredom and tension, of nothingness and turbulence lends many secure settings a peculiar intensity, rather than an ordered calmness. Emotion, distress and violence are all still present in the space, but not welcome. These qualities of being present and absent, both quiet and loud, recall Kristeva's (1982) concept of the 'abject': 'something rejected from which one does not part, from which one does not protect oneself as from an object'. In similar terms to Douglas, Kristeva (1982) characterises the abject as 'what disturbs identity, system, order. What does not respect borders, positions, rules. The in-between, the ambiguous, the composite (p. 4)'. Distress, violence and emotion can all be seen to be abject within the forensic system, dismissed and disavowed but still ever present .
In providing a space to express emotion and aggression, and explore difficult memories, the GWP can thus be seen to engage with those parts of patients' experiences which are made abject in the context of the service. The medium of art in general, and graffiti in particular, is important here in enabling the transformation of the abject. Kristeva argues that art is often 'rooted in the abject it utters and by the same token purifies' (p. 17). Art is here positioned as a form of catharsis, a way of transforming the abject into object, bringing that which has been rejected back into visibility, able to be reincorporated. Art, she argues, both 'arouse[s] the impure, the other of mind, the passionatecorporeal-sexual-virile' whilst also being a way to 'harmonize it, arrange it differently', offering a catharsis which can 'soothe frenzied outbursts […] by contributing an external rule, a poetic one, which fills the gap […] between body and soul'. (p. 28). Through the process of the GWP project, participants are provided with the space to acknowledge those parts of their experience which are suppressed in the service, and also to express, translate and make visible their 'abject' feelings and histories into cultural objects. In this more sanctioned form, these difficult feelings and histories can then be seen and heard. Being violent would carry penalties, but painted violence can be hung on the wall of the ward, or in a gallery. This transposition from the realm of action to the realm of the symbolic enables those elements of participants' lives which are most threatening to the space of the forensic unit, to be visible, expressed and incorporated. They can be ordered, without having to be ignored.
The GWP, therefore, can be seen to provide a space to transform that which is made abject by the institution. This is achieved partly due to the artistic medium used; as such experiences can be rendered visible whilst not being fully present. The form of liminal space created by the GWP can thus be understood to be carnivalesque (Bakhtin, 1984). Bakhtin identifies in the carnival tradition-from the medieval Feast of Fools to the Roman festival of Saturnalia-a set, bounded period of time in the year when, temporarily, the usual hierarchies were dissolved and the norms of the elite upturned. In Saturnalia, for instance, slaves would be served by their masters, and in the medieval Feast of the Fools, a 'Lord of Misrule' was appointed from among the peasant class to rule in place of the feudal lord for the period of the feast. Carnival, Bakhtin argued, went further than upturning hierarchy and contained an overt celebration of all that was usually suppressed or marginalised: 'Carnival celebrated temporary liberation from the prevailing truth and from the established order; it marked the suspension of all hierarchical rank, privileges, norms and prohibitions. Carnival was the true feast of time, the feast of becoming, change and renewal. It was hostile to all that was immortalised and completed' (p. 10).
An upturning of usual rules was made possible, Bakhtin argued, through the aesthetic forms used in carnival: the comic and the grotesque. These two forms pierce the main modes of elite power in European medieval society: religion and social hierarchy. Comic forms challenge hierarchy, mocking rather than respecting those in power. The grotesque, meanwhile, brings the sacred 'down to earth', through celebrating what is usually debased in Judeo-Christian religion: the material, the leaking or uncontrolled body, as well fertility, reproduction and birth. By evoking such a different aesthetic from the more elite forms of cultural representation, the carnival was able to create a new space outside of-and nonthreatening to-official cultural practices, in order to air and celebrate that which was usually suppressed and denigrated. Carnival can thus be seen as the celebration and bringing into the light of all the elements of life which Douglas (1966) identified as attracting purity practices and pollution beliefs. Whilst one response to these elements of life-the material, the ambiguous, the unordered, is to try and order them away, another is to provide a space and time-like a carnival-in which these parts of life can be aired for a limited period.
Whilst the GWP does not engage with comic or grotesque forms, we argue that there was a carnivalesque element to the kind of liminal space created by the project. It was contained, in space and time. It actively engaged with and celebrated the elements of life which are made abject in forensic services, such as violence and distress. Finally, it achieved the transformation of the abject through a non-elite aesthetic form: graffiti. Graffiti is inherently transgressive and subversive, a form of art which can be seen as itself 'matter out of place', positioned as 'dirty' in many contexts and often cleaned away. As well as often being literally illegal, transgressing the official rules of urban space and carrying a risk of police involvement, graffiti is also an underground rather than elite form of art, which originates in marginalised communities (White, 2001). Within the highly ordered space of the forensic unit, it was therefore a powerful tool in signalling the suspension of usual rules. The non-elite nature of graffiti as an aesthetic form also has this quality of giving licence to air and challenge the abject, the messy, in ways which are not usually sanctioned in the ordered and risk-saturated spaces of the forensic unit. Part of this is facilitated by the ambiguous status of graffiti; whilst being 'underground', graffiti is also increasingly sanctioned by the mainstream art world, especially through figures such as Banksy, whilst still retaining some 'outsider' status and practices. This dual status was mobilised through the GWP, as this wider cultural status of graffiti arguably enabled the wall itself and the photographs of the graffiti work to be displayed in the ward and the gallery.
We are aware that the idea of a 'carnival of the abject' might not sound appealing to those who work in forensic services. Certainly, the idea of actively 'celebrating' the transgressions which the | 13 STUCK IN SEPARATION institution is engaged with controlling and containing, is likely to be seen at the least as strange and at the most, actively dangerous. These are, after all, the parts of experience which the institution is aiming to be rid of, so that patients can be safely discharged into the community, to be 'reincorporated'. Faced with evidence of these transgressions, the urge to suppress, to reaffirm the boundary of unacceptable and acceptable behaviour, or madness and sanity, is of course strongly embedded in the logic of the institution. Yet suppression is not the same as transformation, and there is a risk that suppressing and sublimating distress and violence means they will reappear again in the future. The GWP instead made possible a 'safe' and sanctionable expression of these experiences.

TACTICAL INDETERMINACY
There isn't really a comparison with other therapies as these are often focussed on their illness, offending and education, however I think that it is fantastic balance to these type of core pieces of work that give people a culturally specific creative outlet which is invaluable. Similar to key changes, having an external facilitator who isn't part of the core clinical team means that patients interact differently and can develop relationships that allow expression of another part to themselves.
(Survey response) This comment from a staff member working in the unit, highlighted some of the qualities which marked the GWP apart. The facilitator's 'outsider' status coming into the unit from the Gallery rather than as a mental health professional is credited here in relaxing the unit's usual hierarchical dynamics. The project was also set apart from the usual care pathway of the institution. Indeed, the project tended to be discussed in terms of what it was not, rather than what it was. The facilitator was positioned as 'not a therapist'; the project was 'not Occupational Therapy' and also 'not Art Therapy', the wall was 'not a mural'. This indeterminate quality seemed to be part of what enabled Wakeling to create the liminal, carnivalesque space of the project, and also what enabled this space to survive within the highly regulated bounds of the institution.
In Stenner's (2018) expanded ontology of liminality, he explores the occasions which precipitate liminal experiences, those which can engender transformation and 'pattern shift', including experiences outside of formal or 'devised' liminal situations. He argues that 'during liminal experiences our usual 'representations' fail us, and new ways of going on are required' (p. 49). Liminality then, in this understanding, is a form of process where the new can emerge because the old has failed, or is found to be insufficient. As Stenner argues, 'liminal experiences that can be construed as events that begin with an uh oh! (expressing negation) and end with an ah ha! (expressing affirmation)' (p 71). We can think of liminal experiences as emerging on the edges of our current experience, understanding, when we reach a limit, and have to reformulate, or generate something new, to traverse the 'uh oh'. Stenner labels these 'this is not experiences' (p. 71).
The GWP, held in the context of the forensic institution, can be seen as creating many 'this is not' experiences; indeed, overall it was something of a 'this is not' project. As explored in the previous section, the fact that graffiti is an underground and subversive form of art clearly signalled to both participants and the wider institution that the GWP was set apart from the norms of the institution. The act of being able to paint on the walls (rather than on paper), symbolically and literally going beyond the usual confines, also communicated that this was a space with different permissions. The initial stage of the project where the facilitator developed the 'communitas' relationships with patients was also in a closed room, and participants could only be admitted if they arrived on time, which staff commented as unusual in the unit. For patients attending the sessions, therefore, the early stages of the project were filled with 'this is not' experiences, signalling ways in which the space of the project stood apart from the usual practices of the institution. These 'this is not' experiences can be seen as the process through which Wakeling was able to build a liminal space with patients to enable the expression of the abject.
The project also sat in a strangely indeterminate position in terms of the relationship to the wider institution. It was ultimately sanctioned by the institution and commissioned as part of the Occupational Therapy service. Yet, as made explicit in the quote above, the project was still seen as sitting apart from the 'core work' of the team and the institution-it was categorised as neither security nor therapy, and not explicitly tied to the care pathway. The project was within, but not of the institution. This indeterminacy extended to the art process itself. Unlike many street art projects in health or secure environments the graffiti wall in this project was not an end product, but a continuous and ongoing process (cf, Walker & Hanson, 2017). There was no final mural being worked towards, which would then exist in a static state. The project thus remained true to the processual and ephemeral nature of graffiti art practice, which can always be erased, written over, or in other ways altered by official or unofficial means. The fact that the wall was always in process rather than becoming a final object can also be seen as an effective tactic to keep the institution at bay. A mural requires permissions, sign off, risk assessment, including, presumably, approval of the 'final image' for inclusion in the unit. By keeping the wall in process, these permissions were sidestepped, enabling the expression of the usually disallowed. Photographs were taken of some images and displayed, but these extracted images, transformed to objects, were not 'the wall' itself which was kept instead in a state of process.
Within the context of the unit, indeterminacy was a successful strategy in enabling the project to create and maintain a liminal space of transformation. Slipping between established staff categories enabled the facilitator to build authentic trusting relationships with patients which were less compromised by custodial concerns, which are an ongoing struggle for forensic psychiatric staff (Marshall & Adams, 2018;Martin & Street, 2003). The usually abject could be explored and aired precisely because it was not formally concretised as part of participants' pathway through the institution. Returning to our three stages of passage, this indeterminate quality of the project perhaps poses a problem for the final stage: incorporation. There were elements of a rite of incorporation in the images from the project captured and displayed in both the gallery and the ward. These are ways in which the demonstrations of patients' newly developed 'visual voice', objects which demonstrated their transformed self, were able to be incorporated both into the space of the ward itself and the wider community. The necessity of keeping the institution at bay in order to make space for liminality, however, meant that these opportunities for incorporation were limited and partial.

CONCLUSIONS: MAKING SPACE FOR THE LIMINAL IN FORENSIC SETTINGS
In this article, we have explored the ways in which forensic settings can be considered as a failed rite of passage, one which is overly engaged in the rites of separation-order and purity-at the expense of facilitating liminal spaces for transformation and providing a route towards reincorporation. Through the work of Douglas, Kristeva and Bakhtin, we have explored the idea that there are many potential responses to transgression and marginality, of which order, which dominates in forensic services, is only one. We do not argue here that forensic institutions can abandon their function in maintaining legal and ethical limits of society. Nevertheless, we have argued that for meaningful transformation to occur there does need to be a broadening of the register in which forensic institutions engage with the experiences of patients. This project illuminates how one cultural form, graffiti, can be harnessed as an aesthetic means to create carnivalesque space for exploring and transforming the abject. The dilemmas involved in creating such a space within an institution which simultaneously must hold boundaries were explored here and need careful consideration. Resisting ways in which the 'pure process' of liminal space becomes concretised into the ordering practices of the institution is the tactic which was successful here. It is worth considering whether this is optimal, however, as without integration into the practices of the institution, there is a danger that the 'gains' in the liminal space of the project evaporate rather than being taken forward to the 'rites of incorporation'-a return to the community.
The arguments in this paper could be potentially useful in moving beyond an understanding of forensic practice as locked into a dilemma between security vs therapy. Understanding these as both being engaged in ordering practice means that we can start to unpick instead which activities and relationships can provide liminality, possibility and openness. In the example of the GWP, the carnivalesque and abject were powerfully engaged to create a liminal transformative space. Whilst an ongoing carnival of the abject may not be sustainable or welcome in terms of the everyday life of the ward, units could potentially build on recovery principles to create a more 'subjunctive mood' (Turner, 1982) on the wards, open to possibility and positive risk (Drennan & Alred, 2013). Therapeutic community models (Clarke et al., 2018;Rapoport, 1960) provide a ready-made way to think about how to create an institutional environment which is more equal and active, and more tolerant of disorder than the diagnostic and risk-driven approaches within the forensic institution. We hope that this article adds to alternative ways of seeing patients as transcending their label, diagnosis or offending history, with liminality offering a more human and expansive set of possibilities.
In many ways, the drive to order in forensic institutions is understandable, considering the substantial pressure placed on the forensic system to maintain order and boundaries. Risk management can be seen as an effective tactic to manage accountability for the maintenance of such boundaries from both the criminal justice and psychiatric perspective. It may help to ease nerves to disentangle the different functions of the institution, and to think of these processually. Whilst the edges of the institutions may need to be reinforced with signs of order, this is only part of the process and space of the institution. Both entry and exit may need ordering practices, to demarcate the border between inside and outside. But within, and in between, in liminality, there is room for, and the need to, create spaces which can facilitate transformation. These two functions-order and change-need different kinds of spaces, practices, relationships and roles. At the moment, one (order) overwhelms the other. Understanding when (before, after, in the middle) and where (inside, outside, on the edge) these different functions are needed would perhaps help to disentangle these seemingly contradictory functions.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.