Introduction

By the second session of a 4-week support and psychoeducation course for mental health carers, the small group of mostly middle aged women have shed their initial reticence. One mother weeps as she explains that over the weekend, she called a Crisis Assessment and Treatment (CAT) Team after her heavily inebriated son threatened to commit suicide before speeding off in his car. Another says that if she wasn’t so consumed by the needs of her mentally unwell son and young granddaughter, then she would immediately admit herself to a psychiatric ward, such is the level of her own distress. The facilitator of this group is Danny: a spirited, leather-clad, self-styled outsider. He listens carefully and nods. He speaks of the importance of developing a “safety net” to support mental health recovery. He writes “carer/person being cared for” in the middle of whiteboard and draws a circle around it. “Who or what might be involved in this safety net?” he asks. Initially the group members name people: extended family, friends, neighbours, psychologists, psychiatrists, case managers, pharmacists, and so on. Danny writes these down, circling each name and connecting it back to the centre circle. Then the group begins to name physical environments and material conditions: appropriate housing, financial support, workplaces, neighbourhoods, schools. The whiteboard starts to look more like a web or network. Finally, the group identifies institutions (child protection departments, welfare agencies, legal aid services, state trustees) as well as the encoded norms and objects they produce and enforce (e.g. Advanced Care Directives, Compulsory Treatment Orders, Financial Management Orders). “This safety net isn’t static,” Danny says, looking at the busy whiteboard, “It needs to be maintained. How do we keep it alive?” The members of the group talk about sustaining connections with the various people and institutions involved. Yet the suggestions for how to achieve this maintenance centre on communication technologies and material objects: calls, texts, emails, facebook messages, Christmas cards, home baked cakes, and flowers.

This vignette serves to introduce the main themes of this article. First, the messy constellation of people, institutions, environments and objects on the whiteboard suggests that mental health recovery is hardly the “deeply personal” (Anthony 1993:17) process that it is often presented as. Peoples’ lives are enmeshed in complex and heterogeneous networks of interdependence, and subjective experiences of recovery can be seen as arising from and reflecting these human and nonhuman networks (Duff 2016). The second theme is the importance of family to the recovery process. All of the members of the carers’ support group were there to help their own adult children (as well as grandchildren, in some cases), but families very often play a role in recovery, positive or negative, regardless of the focal point. Over 50% of Australians living with severe mental ill-health are estimated to have daily contact with family member(s) (Morgan et al. 2012) and roughly 20% live with dependent children (Maybery et al. 2009). Many people are simply not able to neatly distinguish their own recovery process from their family relationships or their responsibilities as parents (Nicholson 2014). Third, while people and their connections with each other are important, this vignette points to a way of thinking about families that challenges the anthropocentricity of many studies of family life, mental health and recovery. A view of families as simply comprising their human members may be necessary for many purposes (e.g., for legal reasons), but this approach can also limit the scope of social enquiry. While studies of family life and recovery tend to focus on parenting practices and identities, the quality of family relationships, and communication between family members, the lived experience of family life is not confined to, or defined by, these relationships, roles and dynamics, as the messy web of relations on the whiteboard affirms. Being a part of a family can profoundly shape the material, social, discursive and affective organization of peoples’ lives, which in turn has implications for the ways in which their mental ill health manifests, is understood, and managed.

In this article, we use ethnographic data, collected as part of a project focused on families in which the father experiences mental ill health, to explore the impact of family life on mental health recovery. There is a small but growing research and practice literature on “family recovery” (e.g., Nicholson 2014), which like the current article tends to centre on the experiences of families in which one or both parents experience mental ill health. While sympathetic to this literature, especially for the way it challenges the individualistic focus of most recovery models and frameworks, we argue that it remains hampered by its anthropocentricity. In order to open up our analysis to the “more-than-human” (Pyyhtinen 2016:63) constituents of family life, we employ the emerging concept of the “family assemblage” (Price and Epp 2015), which conceives of families as assemblages of human and nonhuman entities and forces. As we demonstrate below, viewing families as assemblages allows for novel insights into some of the most basic aspects of the recovery process.

Recovery and Its Limitations

In much of the English-speaking world, recovery (or the “recovery approach” or “recovery model”) is now among the most influential paradigms shaping mental health policy and practice (Slade et al. 2014; Braslow 2013). Emerging from the 1960s and 1970s anti-psychiatry and psychiatric survivor movements, over the last decade recovery has become an “unquestioned organizing principle of public mental health services, as sacrosanct as was the belief in the healing powers of the community a generation earlier” (Braslow 2013:783). Although there has been little consensus among the numerous definitions and models of recovery used internationally (for a review, see Andresen, Oades and Caputi 2011), the largest and most systematic effort to date to synthesize the recovery literature resulted in the influential CHIME model (“CHIME” being the acronym for the five recovery “processes” around which the model is structured: connectedness, hope, identity, meaningfulness and empowerment) (Leamy et al. 2011). Regardless of the particular model or framework used, recovery is positioned as an alternative to the “medical model” of mental health (Braslow 2013). This medical model promotes an explicitly biological view of mental ill health in which conditions such as bipolar disorder and schizophrenia are regarded as chronic, incurable deviations from normal brain functioning. Like the biopsychosocial approach in psychiatry (Engel 1977), recovery promises a broader and more hopeful alternative to the medical model, a framework with which to critique and move beyond the deficit-based interventions and practices associated with the latter. If the medical model aims at clinical recovery, as indicated by the remission of mental health symptoms, recovery promotes the notion of personal recovery, which entails a meaningful, satisfying, empowered and hopeful life, even if the symptoms and limitations of mental ill health persist (Slade 2013; Anthony 1993).

The concept and practice of recovery has been challenged from a number of perspectives. Some authors have identified a lack of conceptual clarity in the recovery literature, noting that at different times the term “recovery” has been used to denote a philosophy, a paradigm, an approach, a model, or a social movement (Roberts and Wolfson 2004; Thornton 2012). Others have argued that in the hands of professionals such as psychologists, social workers and policymakers, the radical vision of mental health consumers that catalysed the development of recovery has been subverted (Rose 2014; Hunt and Resnick 2015). As Rose (2014:217) puts it, “what began as a liberatory discourse has become instrumentalised and mainstreamed such that it is aligned perfectly with our neoliberal present.” Rose here points to what is perhaps the most persistent criticism of recovery, and the one that is most relevant to our concerns in this article; the contention that recovery is underpinned by a narrow and highly individualistic worldview (Adeponle, Whitley and Kirmayer 2012; Rose 2014; Harper and Speed 2013). Adeponle, Whitley and Kirmayer (2012) argue that the “recovery approach builds on Anglo-American individualism and on an egocentric concept of the person as a self-sufficient, self-determining, independent entity” (p. 116). It is certainly true that the aims of recovery are almost always defined and operationalised at the individual level, and the central actor in the recovery journey is always the person with mental ill health (Topor et al. 2011). Thus the onus of recovery rests on the individual, while the social, material and political contexts of recovery are largely obscured (Adeponle, Whitley and Kirmayer 2012; Harper and Speed 2013; Duff 2016). This is despite the fact that social and structural determinants, such as health care service provision, family support, education attainment and/or progressive workplace arrangements, are consistently identified as among the strongest predictors of positive mental health outcomes (Allen et al. 2014; Furlong 2015).

Some scholars have attempted to overcome the limiting individualism of the recovery paradigm by emphasising the interpersonal aspects of recovery processes (e.g., Marino 2015; Mezzina et al. 2006; Topor et al. 2011). For example, a series of qualitative studies in Europe explored the decisive roles played by family members, friends, professionals and community members in the recovery process (Topor et al. 2006; Schön, Denhov and Topor 2009; Mezzina et al. 2006), leading Schön, Denhov and Topor (2009:345) to conclude that “(a)n individual’s recovery takes place within a social context and the respondents in this study attached central importance to the relationships in their lives.” For such authors, recovery is seen as an “inherently social process” (Marino 2015:68) rather than the deeply personal journey presented in much recovery literature. In a similar vein, the notion of “family recovery” has recently gained currency among both researchers and mental health practitioners (Maybery et al. 2015; Nicholson 2014; Solantaus and Toikka 2006). Proponents of family-oriented conceptualisations of recovery argue that many people live in close relationships with others—relations marked by deep care and commitment—and the mental health of one family member often impacts on many others; for many people, perhaps especially those who live with dependent children, the family provides the most salient social context for recovery (Maybery et al. 2015). For example, drawing on systems and ecological theories, Nicholson (2014) developed a model of family recovery that encouraged practitioners working with parents with mental ill health to identify “the relationships between parent and child characteristics, the family and the environment, and the interactions and transactions among them to suggest targets and pathways for recovery” (p. 7). In a complementary approach, Wyder and Bland (2014) modified the CHIME framework so that each of its recovery processes (i.e., connectedness, hope, identity, meaningfulness and empowerment) is conceptualised as taking a different form depending on whether it is considered from the perspective of the consumer, the caregiving relationship, or the family. This tripartite division promotes a view of recovery processes as dispersed across an interpersonal network, rather than resting solely with the individual consumer.

Other attempts to overcome the individualism of recovery have taken a more radically relational position by drawing on the ideas and resources of the recent “posthuman turn” (Braidotti 2013) in philosophy and the social sciences (see Andrews, Chen and Myers 2014; Duff 2014; Rudnick 2012; and/or Nichterlein and Morss 2017 for a review). This emerging literature typically concludes that while contemporary models of recovery provide insights into some common characteristics of recovery processes (hope, empowerment, and so on), they struggle to account for the actual causes and conditions from which these experiences emerge. Going beyond the reified, individual subject of recovery, these scholars have aimed to incorporate a range of human and nonhuman bodies, spaces and forces into their accounts of recovery. A common goal has been to reveal more of the work of recovery, and the varied resources, relationships, bodies, spaces and objects involved in this work.

From this “posthuman” perspective on recovery, the significance of the family recovery literature is still limited. Although attending to the family rather than the individual alone, models of family recovery still conceptualise families and family life as institutions, systems or processes that largely involve human beings. Non-human environments, objects and forces are often considered, but their significance is quickly elided as they are reified as a “macrosystem”, “context”, or “society” that serves as a backdrop for, or outside influence on, the family proper. This logic positions nonhuman objects and forces as external to the family—which always remains the central unit of analysis—rather than co-constitutive elements of the family. As a result, family recovery models often draw attention to the importance of parents’ experience of hope, for example, to the wellbeing and functioning of the other members of their family, but with limited attention to the various spaces, forces and objects through which family life unfolds, and through which this hope is developed and maintained (see also Anderson 2006). If the intention is to explore the complex relationships between the experiences of family life and recovery, it makes little sense to draw sharp a priori divisions between the “social” and “natural” realms of either. Like recovery itself (see Duff 2016), family life, understood as a process, is sustained by broad networks of humans and nonhuman entities (Price and Epp 2015). In the following section, we introduce assemblage theory, and specifically the notion of “family assemblages”, as a way of advancing this non-anthropocentric understanding of both the family and family recovery.

Family Assemblages

Originally conceived by Deleuze and Guattari (1988), the concept of the assemblage has been developed by authors such as DeLanda (2006), Bennett (2010) and Latour (2005) into a systematic theory of social life. Assemblage theory contends that all knowable entities, human and nonhuman, social and natural, should be regarded as “wholes whose properties emerge from the interactions between parts” (DeLanda 2006:5). Entities from individuals, to transient conversations or meetings, to longstanding organisations and nation states can all be viewed as assemblages. Assemblages are composed of both material and non-material components—places, bodies, affects, discourses, norms, habits, material objects—which come together for a time, exert an influence on the world that is greater than the sum of their parts, and eventually transform or break apart (DeLanda 2006). A defining feature of all theories of the assemblage is the refusal to reify the forms and structures of social life and to focus instead on the specific means by which these forms emerge, are made stable over time, are contested and rearranged, even dismantled and replaced (Duff 2014).

Assemblage theory has been utilized extensively in a number of fields, including human geography (Anderson and McFarlane 2011), sociology (Marcus and Saka 2006), political science (Buchanan and Thoburn 2008), and international relations (Acuto and Curtis 2014). More recently, scholars have also begun applying assemblage theories to the study of mental health and recovery, mainly in an attempt to overcome the individualism and anthropocentrism of existing accounts of recovery, as we noted in the last section. For example, Andrews, Chen and Myers (2014), Duff (2014, 2016), Fox (2016), Nichterlein and Morss (2017) and Tucker (2012) have drawn on the assemblage thinking of Deleuze and Guattari (1988), DeLanda (2006) and Latour (2005) to account for the more-than-human characteristics of recovery processes. These scholars have focused on the range of human and nonhuman bodies, forces, signs, spaces and events that may be shown to mediate the social, affective and material modulations of recovery, the heterogeneous ebb and flow of what Duff (2014:117–120) has called “becoming-well”. From this perspective it is the assemblage and all of its constituent parts, human and nonhuman, that recovers (or fails to recover), rather than any one particular element within it. A key concern of this approach has been the matter of affect and “affective life” (see Tucker 2012) and the ways recovery may be empirically analysed as a kind of affective modulation of becoming well and becoming ill (see also Duff 2016). Focus on the affective modulations of recovery foregrounds the specific encounters, events, processes and relations by which assemblages gain (and lose) capacities to affect, and be affected by, other entities (human and nonhuman) in the everyday experience of becoming well, getting better, achieving higher quality of life, greater function and so on. Again, the focus does not rest with the individual subject of mental illness, but rather this literature emphasizes the entire assemblage in which this subject emerges.

It is only recently, in the field of consumption studies, that assemblage theory has been applied to studies of family life (Price and Epp 2015; Huff and Cotte 2016). Price and Epp (2015:60) were the first to offer a detailed conceptualization of “family assemblages”, defining the family as an “assemblage of heterogeneous components (e.g., bodies, practices, objects, stories and everyday interactions) that form contingent relations across time to produce an emergent whole with a collective identity”. The concept of the family assemblage shifts the focus from individual actors or organisms within the family, to the arrangements, relations and capacities involved in “the open-ended project of assembling family” (Price and Epp 2015:60). While family systems and ecological theories also focus on processes rather than individuals, the notion of family assemblage is distinguished by its emphasis on heterogeneity, openness and flux. As Price and Epp (2015) put it, “Rather than thinking of families as having boundaries (static and fixed) we think of them in terms of territories that are blurred, dynamic and dependent on family movements within and without”. This means that family assemblages include far more than simply human aspects such as parents and children, and the relational dynamics between them, encompassing a host of additional objects, technologies, norms, affects and agencies. To use an example from empirical work exploring the practices of families with ageing members (Huff and Cotte 2016:11), even institutions, such as elder care facilities that are used to look after an ageing parent, can be seen as a part of the family assemblage:

This is not to say that the family considers a nursing home to be a member of the family; rather, that the nursing home, by nature of its function in providing residence and care for the family member, becomes integral to the way the assemblage does family, and, hence, becomes a component of the assemblage.

As Latour (2005:72) would put it, anything (human or nonhuman) that “participates in the action” of family life can be considered a part of the assemblage.

The concept of family assemblage has been productive in marketing and consumer research with families, where the focus is on new products or technologies and the ways in which they can organise, disrupt or change family practices. Yet to date the concept has not been applied in studies of health or mental health. We have argued that assemblage theory offers a way to think outside of the limiting individualism of much of the recovery literature, and, further, to challenge the anthropocentrism of the family recovery literature. Beyond these arguments, the remainder of this paper begins to explore what creative paths and possibilities the notion of family assemblage may open up in studies of mental health recovery.

Method

The data for this article are drawn from ethnographic research and in-depth interviews conducted by the first author in Melbourne, Australia, between July 2014 and August 2015. The study was approved by the Monash University Human Research Ethics Committee. No services specifically for fathers with mental ill health operated in Melbourne at the time of the study. However, one hundred and fifty-five hours of participant observation were conducted at sites at which the authors believed the issue of fathers’ mental health would be salient, including a non-clinical mental health service operating through a community garden, a mental health carers support group, a monthly support group for struggling fathers, and the Melbourne Children’s Court. Issues and discussions surrounding fathers’ mental health were indeed present at each site, and interview participants were recruited at these sites.

Multiple recruitment strategies were employed to produce an interview participant sample that was socio-demographically, ethnically and geographically diverse. Thirty-one participants were recruited through targeted social media advertisements (12), community-based mental health organisations (10), participant observation activities (5), and snowball sampling (4). In order to expand the number of participants, recruitment occurred in various locations throughout Australia. Most participants from Melbourne were interviewed in person at a public location, while participants from other states were interviewed on the telephone. Interviews lasted between one and two hours and were semi-structured, allowing participants to discuss topics of importance to them.

Participants in the study included fathers with mental ill health (n = 10), partners/former partners of fathers with mental ill health (n = 10) and service providers working with families affected by paternal mental ill health (n = 11). In three cases, both father and partner were (separately) interviewed. Inclusion criteria for men included: (1) currently receiving some form of professional mental health treatment, (2) having become a father (biological, step- or adoptive) in the past 18 years, and (3) having experienced the symptoms of mental ill health before becoming a father. Partners/former partners were eligible to participate if the associated father was eligible to participate in the study, while service providers were eligible to participate if through their professional activities they had direct contact with families affected by paternal mental ill health. Across the father and partner/former partner participants, self-reported paternal mental health diagnoses included depression and/or anxiety (7), bipolar disorder (5), schizophrenia (4), personality disorders (2), and post-traumatic stress disorder (2). All names used in this article are pseudonyms, and in some cases details have been changed to ensure participant confidentiality.

Analysis involved an iterative dialogue between empirical data, relevant literature and assemblage theory. The empirical data were analysed by each author using the iterative approach characteristic of ethnographic research (Bernard and Ryan 2010; Glaser and Strauss 1967) and informed by Clarke’s (2005) “situational analysis”. Building on previous analysis (i.e., Duff 2012, 2016), data were analysed using a priori theoretical concepts, particularly the notion that assemblages can be seen as comprising material, social, discursive and affective components. Analysis of interview transcripts involved open, axial and selective coding to explore how the various components of participants’ family assemblages related to commonly identified aspects and processes of the recovery model, and further how these recovery processes may be reconceptualised in light of assemblage theory.

Findings

Rather than analyse the recovery accounts provided by our participants in terms of a particular recovery model (or models), we have elected to analyse these accounts with more specific reference to the assemblage thinking introduced above. In so doing, our goal is to avoid the anthropocentrism of existing recovery models in favour of renewed attention to the social, affective and material aspects of recovery processes as they unfold within assemblages of human and nonhuman bodies, spaces, forces and signs. As we have indicated, assemblage thinking is profoundly challenging to individualistic notions of recovery, which are generally grounded in notions of human emotion, identity and agency. Each aspect, emotion, identity and agency, must arguably be rethought in order to advance a non-anthropocentric account of recovery, more attentive to the actual processes, events and relations in which recovery advances. Our approach then is to explore how the materiality of everyday life, and the social, economic and political structures that shape the environment, are central to the organisation of family life, rather than being contextual or circumstantial. We develop this analysis by way of three case studies that indicate how the family assemblage links the individual, the household, and wider connections of family and friends, with varied nonhuman elements including spaces, organisations, objects and technologies. These case studies are not intended as comprehensive treatments of the different lines of flight the concept of family assemblage opens up in studies of mental health recovery. Rather we offer this analysis as indications of the kinds of insights that the concept of family assemblage can offer the study of mental health recovery.

Family Affects

It is a dark and wet day at North Park Community Garden. The garden is run by one of the largest non-clinical mental health services in the region, and offers people with mental ill health the chance to socialise, learn new skills, and interact with the broader community. On a normal Wednesday, I (first author, Rhys) would be working out in the garden, but today I am staying warm in the kitchen, cleaning up, making tea and chatting with the few participants who turned up despite the weather. Helen, the manager of the garden, comes into the kitchen with a man she introduces as Troy. She has been telling Troy about my research and he is interested in talking to me. I suggest that the workshop would be a more private place to talk. Brushing away some wood shavings, I sit on a high stool at the workbench. Troy stands, in his baggy clothes and baseball cap, smoking cigarettes as we talk.

Troy tells me that he was 19 years old when his daughter was born. It was around that time that he was first incarcerated and soon after diagnosed with schizophrenia. He has spent most of the 16 years since then living in a psychiatric prison. For the last few years he has been allowed unsupervised into the community during the day, and more recently to live in an apartment managed by the prison, but he has been restricted from travelling from Melbourne to Queensland, where his daughter lives. She has only visited a handful of times since she moved there when she was three years old. He tells me how he sends her money and gifts:

I send her… like, on her 16th birthday, I sold four artworks that I’d done and got about just over 700 bucks and I gave that to her. I sent her pictures of the artwork. I sent her a pewter jewellery box that I got off eBay with 47 butterfly pendants… because I don’t know what little girls like, you know.

Most of all, he loves talking to his daughter on the phone:

I talk about everything. I ask her what she’s doing. I have my own opinions, my own stories and all that sort of shit. I will talk to her for about half an hour, maybe once or twice a week, depending. If I ring her up during the middle of the week, it’s because I’ve got a bit of money that I can give to her and I ask her if she needs anything. And if she says, “No,” then I say, “Well, you’re getting 30 dollars, so don’t worry about it”. Yeah.

Troy’s face and voice soften as he talks about his daughter. My body relaxes in response. I want to learn more about him as a father. But our conversation keeps returning to his frustration and rage at the prison authorities. He tightens as he tells me how they block him from achieving what he wants in his life: “If they just would fuck off and die I would be happy, you know. I’m not the type of person who holds a grudge, but if they’d just fucking leave me alone.” I catch glimpses of the man who used to get “in heaps of fights with people” in prison. The stark movement between emotional softening and contraction that I notice in Troy is not confined to our conversation, as he indicates:

If I can latch onto something that feels positive, I always go with that. If I’m not, if I start regressing, I just get angry. A lot of the time when I’m on my own, I get quite angry. …I feel that when I’m talking to my daughter and she’s telling me about school and work and all that sort of thing, she’s great. I forget about everything, you know, and I just want to do what’s right for her, and what’s right for her is what’s best for me. And, you know, I don’t want anyone standing in the way, but there’s a lot of people that are, I don’t know, obstructing that progress.

One way of interpreting Troy’s statements about his relationship with his daughter, common to the existing recovery literature, would be to say that his identity as a father, and the specific ways he adopts a paternal role, help him experience the positive emotions like hope and empowerment that promote recovery. However a focus on affect, as opposed to emotion per se, suggests alternative analytical directions with a stronger focus on relational capacities rather than subjective feeling states (Massumi 2002; Gregg and Seigworth 2010). Drawing on Deleuze and Guattari (1988), Massumi (2002) makes a distinction between emotions, which are subjective and personal, and affects, which are pre-subjective, “pre-personal intensities” (Massumi 2002:29). Affects do not reside in one particular body or subject, but obtain in encounters between bodies (human and nonhuman). Affects mark a transition in a body’s capacities to affect (direct, shape the conduct of, please or annoy, inspire or restrict another’s actions, shape a scope of activity) or be affected by other bodies in an encounter. Affect describes a capacity or preparedness to act within a specific scope of activity. It records an increase or reduction in an individual body’s power to act as this power is inflected or modulated in an encounter (see Duff 2014:41–45). This modulation typically bears an emotional valence—joy or sadness or some combination of feeling states—but only insofar as emotion is the intelligible trace of a prior affective transition. In other words, emotion is the coming into subjective awareness of a pre-subjective, pre-cognitive affective modulation in a body’s capacity to act (see Duff 2014; Brown and Stenner 2001). If the arrangement of a given assemblage limits an individual’s power to act in the world, the associated emotional experience is likely to be experienced as negative—anger, sadness, anxiety, and so on. Conversely, an expansion of an individual’s power to act will normally give rise to positively experienced emotions, such as happiness, excitement and hope.

With this distinction between affect and emotion in mind, we can add that Troy spends much of his time in physical environments marked by constraint. His ability to move freely in the world is dependent on the decisions of others. Prison rules structure his daily activities and Troy experiences these environments as preventing him from achieving what he wants in his life. Troy’s emotional response to this obstruction is anger, which he identifies as a regression in terms of his ongoing recovery. This is one of the assemblages in which Troy’s recovery advances and retreats, a prison/incarceration assemblage. Yet Troy’s recovery also advances, less falteringly, in his family assemblage, structured around his teenage daughter, which extends beyond the confines of his incarceration. This second assemblage organises heterogeneous people (e.g., his daughter, her mother and stepfather), technologies (e.g., mobile phones, computers, internet banking services), institutional norms and practices (e.g., of the social welfare department, which provides Troy with a disability pension that he is able to pass on to his daughter) and discourses (e.g., the “breadwinner” fathering ideal, where men express love and care through financial provision) that come together to provide an affective counterpoint to Troy’s daily sense of obstruction. This family assemblage permits Troy to adopt and express capacities to act in ways that would otherwise remain restricted in the prison environment. He expresses love and care with his attention, gifts and money. He exercises a sense of authority and empowerment by, for example, sending his daughter money even when she says she does not need any. And he speaks of his hope that he will be involved in his daughter’s life into the future (“I want to be a granddad, yeah. That would be awesome, you know”). Troy’s ability to even imagine a family that extends across generations is dependent on the diverse structures that constitute his family assemblage.

The affective organisation of every family assemblage differs as it changes across time. For some men in the current study, family assemblages provided conditions that were highly destabilising of their moods and mental ill health, at least for a time. For other men, such as Troy, family assemblages provided a source of peace and stability, offering a way forward for their ongoing recovery journey. Our focus on affect suggests that the emotional experiences that comprise recovery are best seen as outcomes of affective, relational processes rather than subjective feeling states. Emotions are subjective; they belong to the humans that experience them. Affects are relational and pre-individual; they are the modulated outcomes of encounters between the different components of an assemblage, and the differing capacities that these encounters engender. Put simply, experiences such as hope and empowerment emerge from the contact between individuals and the assemblages of which they are a part. Such a view in no way invalidates the experiences of people with mental ill health, but simply suggests that the genesis of these experiences is more complex and relationally situated than individualistic conceptualizations of recovery allow.

Emergent Identities

The concept of identity is central to many models of recovery (e.g., Slade 2013; Leamy et al. 2011). For example, Slade’s (2013:8) influential practice guide, 100 ways to support recovery: A guide for mental health professionals, states that:

The first task of recovery is developing a positive identity outside of being a person with a mental illness. Identity elements which are vitally important to one person may be far less significant to another, which underlines that only the person can decide what constitutes a personally valued identity for them.

Like much of the recovery literature, this practice guide affirms that developing valued social roles and engaging in positive interpersonal relationships can shape or facilitate the process of identity transformation. Yet this transformation is still seen as one that largely takes place within, and through the efforts of, the individual subject of mental illness (Topor et al. 2011; Duff 2016). The onus remains with the individual to change their identity in ways that may be more congruent with their recovery goals. Like recovery itself, identity is seen as a unique and subjective phenomenon. The main role for mental health professionals in the process of identity transformation is to encourage reflection and novel ways of thinking by using techniques such as strengths-based assessments and practices that focus on the person rather than their diagnosis or symptoms (Slade 2013).

The literature on family recovery has moved the idea of identity in a more interpersonal direction by acknowledging the often intimate connection between an individual’s sense of self and his or her family relationships, particularly parent–child relationships in which the parent experiences mental ill health. This literature focuses on “parental identity”, which emphasises that for many parents it can be difficult or impossible to neatly separate their own sense of identity from their roles as caregivers. As many parents with mental ill health report that their parental identity is ignored by mental health professionals (Reupert and Maybery 2007), the aim of much of the family recovery literature is to encourage practitioners and service systems to recognise the central role that parenting plays in many consumers’ lives. However, when it comes to practical techniques for promoting a positive sense of parental identity, the advice given to professionals often mirrors the individualism of the broader recovery literature. For example, in Creating options for family recovery: A provider’s guide to promoting mental health, the first book-length family recovery practice guide, Nicholson and colleagues (Nicholson et al. 2014:34) encourage practitioners to “help parents create a positive identity” with the following advice:

Parents are likely to need your help remembering who they are and what they are capable of. Here are some sample questions to help parents think about what is important and meaningful to them, as well as to help them begin to define themselves on their own terms. Be sure to add other questions you think might be helpful.

  • How would you describe yourself?

  • What do you like about yourself?

  • What helps you feel better?

  • What kind of food do you like? Music?

  • What’s your favorite memory?

  • When are you happiest? Why?

These injunctions treat parental identity as a subjective, intrapsychic phenomenon that is best modified by personal reflection. By considering their history, their strengths, and their desires, it is thought that parents may begin to “define themselves on their own terms”. While not repudiating the value of self-reflection, the findings of the current study offer a more complex vision of the experience of parenting and the ways it might promote (or impede) recovery. For many fathers in this study, family life did indeed engender profound changes in their identities, but the processes and conditions through which these changes emerged spread well beyond the parent–child relationship. The story of Felix and his family illustrates how positive changes in identity often emerge as a part of a shifting assemblage of people, discourses, environments and things.

Felix and I (Rhys) meet for lunch in a hip, grungy pub, and even in this environment he stands out. Over six feet tall and broad, with long red dreadlocks, numerous tattoos and a bright psychedelic t-shirt, he looks like anything but a suburban dad. And that is how he likes it. For a long time, Felix’s sense of self was powerfully shaped by attitudes of anti-authoritarianism and non-conformity. He embraced his “crazy identity” in ways that worked against his recovery. Becoming a member of a family assemblage, however, destabilized the foundations of his defiant sense of identity. Although he continues to resist the persona of the “suburban family man” in favour of his own “wild and feral” image, his family assemblage has become the ground for a sense of identity that holds these images in creative tension.

Felix’s mental ill health first appeared following heavy substance use throughout his teens. After being arrested and detained in a psychiatric ward at age 19, he was diagnosed with bipolar disorder. His initial experiences in the mental health system were traumatic:

I had a very, very oppositional experience there with the staff and the psychiatrist and was given ECT [Electroconvulsive therapy] against my will. So that kind of laid the foundation for being very untrusting and aggressive towards the system, the mental health system, which carried on for over a decade actually. So from that point, when I got out, I was very angry and crazy. I was flat-out crazy.

Throughout his twenties, Felix cycled between periods of relative stability—the longest of which lasted for two years—and periods in which he “dropped off the radar” for months on end: manic, homeless, using intravenous drugs, and disconnected from family and friends. He estimates that he spent sixteen months of this decade in psychiatric hospitals. Inspired by his interests in spirituality and martial arts, he developed views about his mental health that kept him at odds with the professionals who were treating him, believing there was purpose and meaning in his manic symptoms. Each time he was discharged from hospital, he would stop taking the medications his psychiatrists had prescribed him. Increasingly, his identity cohered around defiance of the authorities in his life: “There’s almost a pride in it as well, that’s a part of it. …I learnt to embrace my crazy identity in such a way that it was like a kind of a frothing antagonistic, “Fuck the world, I’m crazy. You want to see crazy? I’ll show you crazy!”

In his early thirties, Felix met and fell in love with Belle. Before long the couple had two girls, born just over a year apart. Felix’s mother offered the young family cheap rent of her house while she lived in a bungalow in the backyard. In order to help support his family, Felix qualified as a mental health peer support worker and began working for a community organisation that based its interventions in the recovery approach. With more reasons than ever to keep his symptoms under control, Felix was for the first time inspired to comply with his psychiatrist’s treatment indications: “I have no intention of being on Lithium for life, but I do recognize it’s serving a purpose now in providing a safety zone of stability for me to work on my family and my career and kind of … yeah, it’s a safety, it’s assisting me in that phase of my life.”

While Lithium provided him with a sense of stability, his professional life encouraged reflection and restraint. He gives an example from the day before our meeting, when he returned to a hospital in which he was once an involuntary patient:

I was at St Mark’s hospital yesterday, to pick up one of our residents who’s relapsed, and I ended up telling off one of the nurses. “How come this guy’s… he’s voluntary, is he voluntary?” She goes, “yeah”. I go, “Well how come he doesn’t have leave?” And she goes, “Well, we just thought it would be best”. “So he can just leave the hospital whenever he wants, but he can’t go to the shops?!” I felt oppositional, like taking it personally, and I started to rise. I just wanted to go for the kill. But I had to back off and just like simmer down. So yeah, I’ve been there a few times now as a worker role and it’s been… it’s been… I’ve paused and thought about it, you know. It’s been thought provoking. And what it means, the kind of full-circle aspect of it, and working from within.

Some of the aggression and opposition that previously defined Felix’s relationship to the public mental health system is still present, but his place in this assemblage has changed. He now occupies a role of some authority, which encourages him to reflect on, and relate to, his feelings in different ways. Had he gone “for the kill”, he would have risked losing his job, which would have negative implications not only for himself but also for his partner and children.

When he is not at work, Felix is normally with his family, and here too he is now responsible for others’ wellbeing. His eyes widen as he speaks of how natural it has felt to assume the role of an authority figure with his children: “I’ve been shocked at how much more I’ve kind of… I care much more about things like screen time or treats or something like that, just consistency, than Belle does. And I’m constantly amazed when it happens.” Authority, stability, restraint: more than ever before, these qualities now inform Felix’s sense of who he is and how he should act in the world. His oppositional spirit is still alive and well, but it no longer sits alone at the centre of his identity:

I think [my sense of defiance]… it hasn’t quite gone, it’s kind of transitioned now. Like as a dad when I’m at the park or whatever, I find it comforting when I have a cider in my hand and I’m barefoot and I’m looking a little wild and rugged and I’ve got two beautiful girls and I’m aware of the incongruity of that and I kind of… I’m more comfortable being like a kind of wild and feral looking dad than the idea of just being a normal dad. So I’m still oppositional, definitely, to the idea of normality. … I don’t feel like I’m quite ready to just, I guess, conform to this notion of a suburban family man.

Having children has powerfully shaped Felix’s identity. Compared to the enthusiastic embrace of his “craziness” in his twenties, he has developed a more flexible sense of identity, one increasingly open to compromise, contradiction and change. This new sense of parental identity provides a firm foundation for his continuing recovery. It is important to state, however, that Felix’s parental identity has not arisen solely through conscious self-analysis or reflection, nor simply through interactions with his children and partner, but rather has emerged from the vast collective of human and nonhuman forces that comprise his family assemblage. His parental identity has been forged in specific environments: the family home, playgrounds, the various spaces and places he visits as part of his employment. It has developed among material objects, from the Lithium that helps stabilise his moods, to the televisions, sweet treats, and countless other household objects that participate in his parental boundary-setting. It has been informed by various cultural images and narratives: from narratives of paternal responsibilities and clichés of the “suburban family man”, to contrasting visions of the nature of mental ill health that he is exposed to in his professional life. In short, Felix’s parental identity, which promotes his ongoing recovery, simply cannot be separated from the diverse and heterogeneous territory of his family assemblage. As Macgregor Wise (2000:310) observes, “subjectivity is a product of territorializing, identity is territory”. Felix’s parental identity emerges as an effect of the particular family assemblage in which he is enmeshed.

Distributed Management

From its very beginnings, the recovery literature has placed a focus on personal agency and responsibility, positioning consumers, as opposed to health professionals or governments, as the central determinants of the recovery process (Sterling et al. 2010; Scott and Wilson 2011). As Scott and Wilson note (2011:40), “The recovery vision is one of people with mental illness stepping out of the limitations of a ‘passive’ patient role, and becoming active and autonomous authors of a self-actualised life”. A number of prominent recovery models emphasise “self-management” as crucial to the recovery process (e.g., Sterling et al. 2010; Leamy et al. 2011; Mueser et al. 2006), which entails activities such as maintaining a focused healthy lifestyle, ongoing self-surveillance and monitoring of one’s thoughts and emotions, and strategic planning for future events and episodes of ill health. Slade’s (2013:16) practice guide, referred in the section above, lists “Self-managing the mental illness” as the third of four “personal recovery tasks” that professionals should promote with clients. Professionals are advised that: “It is not the job of staff to fix people, or lead them to recovery. The primary job is to support people to develop and use self-management skills in their own life. The instinctive response of staff to any situation needs to be “You can do it, we can help”. Critics argue that the concept of self-management aligns neatly with neoliberal conceptions of personhood, where the ideal health consumer is an individual, autonomous, informed and rational decision-maker (Scott and Wilson 2011; Weiner 2011). For such critics, this ideal fails to capture the complex and situated ways in which consumers exercise agency and work towards managing their mental health in collaboration with others.

Many fathers in the present study spoke in ways that appear to reflect the ideals and practices of self-management, although our analysis would suggest that the notion of distributed management is more fitting characterisation of these practices, as we shall explain. For example, Brett, a 34-year-old father diagnosed with depression and anxiety, and with a history of substance abuse, has in recent years developed the practice of “keeping a review of myself and managing myself on a day-by-day, an hourly, basis and changing the behaviour based on how you’re acting”. He gives the following example:

It’s like really keeping a third-person view of what I’m doing, I think is the way to say it. You know, if I’m having a conversation with someone and… they are getting, you know, that lean-back-sort-of-a-look or they’re watching my forehead rather than my eyes, then you know that “I’m not engaging with this person now, why is that? Let’s have a look, are you leaning into their space right now? Are you getting really too animated for the appropriateness of this conversation?” which is something that I do a lot. So learning to put your hands down and close your mouth and let them talk back to you at times. If I didn’t do that then it would get to the next thing of going, “well this person doesn’t like me, why don’t they like me?” and then you’re down a spiral within the next 20 min and then… all of a sudden that person’s your enemy, you know.

If we focus simply on Brett’s cognitive and emotional states as he describes them, then this statement may seem like a straightforward example of the effective self-management of certain behaviours and symptoms that contribute to his ill health. Yet if we broaden our view to see Brett as a member of a materially heterogeneous family assemblage, it becomes less clear where the locus of this management actually lies. For a start, contemporary discourses around engaged fathering have structured Brett’s desires to develop insight into, and maintain some control over, his thoughts and moods. In recent years he has taken primary responsibility for the care of his two boys, and fulfilling the role of an emotionally attuned father has become the central purpose of his life and recovery:

So in terms of being responsible and organizing and a being role-model… it’s just basically become that that’s everything. If I didn’t have the kids, I wouldn’t have a point to be sober, I wouldn’t have a point to be putting effort into, not being happy, but just not being miserable, you know.

Fulfilling the role of an engaged father is not something that Brett does alone. So that he can avoid the stresses of paid work and concentrate on fathering, Anita works full time, even though she would prefer not to. She also performs much of the housework so that Brett can spend more time with the boys. At those times when Brett’s mental health symptoms are most acute, she takes responsibility both for providing financially and the daily running of the family home, often relying on the practical support of her parents. The children, too, provide support in their own ways. Anita describes how Arlo, who is ten years old, does his best to help Brett when the symptoms of his ill health are pronounced:

Arlo just has this sense of responsibility and anything you ask him to do, he’ll do it for the most part. So yeah, he definitely changes the way he behaves when Brett’s not having a great day. Or if Brett just says, “Look dude, I’m not having a great time, can you just go do something else?” he will. Or “I need you to distract your brother”, he will. I think [the children] definitely help Brett manage his depression.

Being part of a family assemblage also strongly informs the spatial and material conditions of Brett’s experience. He now spends his time in markedly different environments than those of the past, such as the family home, as opposed to the bars where he used to work, or the raves that provided the context for much of his free time in his twenties. Environments such as raves mediate access to, and exchange of, various illicit substances, while also providing spaces in which rest and recovery following substance use is normalized (Dilkes-Frayne 2016). In such environments, Brett was a self-described “vice-consuming machine”, using alcohol and other drugs to “self-medicate” and avoid “having to think about stuff from earlier in [his] life”. The family home provides a very different context. There are no drug dealers on site. It is not possible to play loud music into the early hours of the morning. There is no bar with an endless supply of “after-work drinks” when he finishes putting the boys to bed. Although he still sometimes uses illicit substances to manage his moods, his capacity to contain his emotional life in this way is limited: “Over the last couple of months, I’ve been smoking pot intermittently again, because there’s been more stress than normal. But I’m not just sitting down and “bonging on”. I’m getting everyone into bed first and then having a little chill-out time to myself.” Without the sustained “release” (as he describes it) provided by substances such as marijuana, Brett is encouraged to adopt other strategies, such as monitoring his own thoughts and emotions and modifying his behaviours accordingly. While this is certainly redolent of the self-management strategies advocated in the recovery literature, the agencies involved in this management do not belong exclusively to Brett. A host of additional agencies are involved too, suggesting the need for a more distributed understanding of health management.

Agency is normally seen as a capacity of intentional human beings, who act in a world of passive objects. Yet in recent years, authors such as Latour (2005) and Bennett (2010) have stretched notions of agency far beyond their traditional anthropocentrism by asserting that objects, environments and discourses also act in the world—producing effects, affording actions, altering events—and so can be seen as having their own forms of agency. Bennett (2010:32) speaks of “distributed agency” as a way of acknowledging that the locus of agency is often a human-nonhuman assemblage. In a similar way, we can speak of the distributed management of the mental ill health of the fathers involved in this study. From the perspective of assemblage theory, the management of Brett’s ill health is dispersed across a network of discourses, people, environments and objects. If any of the elements of this network were modified or removed, then Brett’s situation may look very different. Remove cultural ideals about the paternal role model, and Brett, by his own admission, would see no reason to continually work at staying healthy. Take away the support of Anita and the couple’s extended family, and it seems unlikely that Brett would succeed in the fathering role as he does, which would again distance him from the sense of meaning that provides the impetus for his efforts at managing his ill health. Separate Brett from the salubrious environment of the family home, with its stabilizing duties, obligations and routines, and he might easily fall back on less sustainable methods of managing his suffering, such as heavy substance use.

Brett puts personal effort into his recovery. He is motivated to act. He has desires and goals for his own and his family’s future. But his personal efforts are defined and structured by the interactions of the different components of his family assemblage. For many of the fathers in this research, having the sense that they were able to self-manage their mental health was important to their ongoing recovery. The intention of a concept such as distributed management is not to deny individuals of their feelings of competence or empowerment, but rather to emphasise that their own efforts are ultimately but one note in a larger composition of forces working towards managing their ill health. Bennett (2010:38) clarifies the relationship between personal effort and the work of assemblages by arguing that agency is:

distributed across a mosaic, but it is also possible to say something about the kind of striving that may be exercised by a human within the assemblage. This exertion is perhaps best understood on the model of riding a bicycle on a gravel road. One can throw one’s weight this way or that, inflect the bike in one direction or towards one trajectory of motion. But the rider is but one actant operative in the moving whole.

In much of the recovery literature, the focus on “self-management” tends towards reductionism, individualizing a relational phenomenon and ignoring the additional bodies and actants that characterise the assemblage in recovery. The concept of distributed management simply reminds us that to properly understand an individual’s efforts towards recovery, we must also keep in mind the “moving whole” to which these efforts contribute.

Conclusion

We will briefly conclude by returning to the example on the whiteboard with which this article began. Do the findings of this study add to or challenge what Danny was teaching the group of mental health carers? Danny’s rich and complicated diagram suggests that, like many social workers and mental health professionals, he already adopts a broad perspective on what is involved in efforts towards mental health recovery. Yet he still conceives of the various people, institutions, environments and objects in his diagram as a “safety net”; that is, as something that is distinct from the human relationships that sit in a circle at the centre of the whiteboard (i.e., the “carer/person being cared for”). Assemblage theory challenges us to rid ourselves of arbitrary distinctions between person and environment, private and public, proximate and distal, personal and social in our characterisations of recovery. Why must we locate the individual at the centre of our view of recovery, be it the “consumer” or the “carer/person being cared for”, thereby leaving them with primary responsibility for the work of recovery? Existing recovery models with their familiar distinctions are almost always drawn along individualistic and anthropocentric lines, maintaining the responsibility of individuals to manage their own recovery while ignoring the significance of a host of social, affective and material resources in this recovery.

Assemblage theory has the potential to sensitize both researchers and practitioners to a much broader sweep of family factors in the lives of individuals with mental ill health, as well as the ways in which purportedly private and personal phenomena, such as emotions and identity, emerge from diverse familial territories. Ethnographic methods provide the means with which to explore how subjects emerge from, and are enmeshed in, diverse and heterogeneous contexts. The assemblage perspective explored in this article is perhaps best seen as offering ethnographic studies of family life some unique emphases. Most obviously, assemblage theory populates ethnographic accounts with novel objects of enquiry. Many ethnographies assume the existence of certain entities and forces—agents, subjects, institutions, norms, power, and so on. In contrast, assemblage theory sees these standard entities and forces as the products of the coming together of a more fundamental set of relations. Thus, rather than taking concepts such as “parental identity” “family” or “recovery” as givens, the notion of assemblage encourages constant return to the “real experience” (Deleuze 1994) of family life. In this article, assemblage theory allowed us to reimagine phenomena that are normally considered deeply personal (e.g., emotions, identity, agency), as themselves emergent effects of assemblages. It also prompted us to view recovery as a radically relational process, where the fixed and static boundaries of familiar social entities (e.g., families) gave way to heterogeneous, unstable collocations of bodies, spaces, signs, norms and affects.

Assemblage theory also offers perspectives on agency and causality that can be deeply challenging to common sense. While many ethnographers assume that agency is a uniquely human capacity, assemblage theory questions this view by exploring how agency spreads throughout human-nonhuman constellations. Our exploration of “distributed management” is but one of the ways in which a “posthuman” perspective on agency may be used as a provocation to traditional understandings of behaviour within families. Similarly, assemblage theory provides a view of causality that begins to grapple with the complex, non-linear clusters of causes and conditions that underlie phenomena such as mental ill health and its recovery. Models of linear causality—where the same cause always leads to the same effect— quickly break down when the social world is understood as populated by entities that themselves are composed of numerous interacting components. In place of linear causality, DeLanda (2006:20) offers models of non-linear causality, where the “internal organization (of assemblages) may, for example, determine that an external cause of large intensity will produce a low-intensity effect (or no effect at all) and vice versa, that small causes may have large effects.” From this perspective, it makes little sense to attempt to abstract the causal strength of a particular member, practice or relationship within a given assemblage, because assemblages produce activity as an emergent effect of all associations immanent to them.

Yet of course, this kind of abstraction is common to clinical interactions in mental health services in Australia and elsewhere. When a man accesses mental health services in Australia, it is common for him to be treated as an isolated individual, with his role as a member of a family excluded from the clinical conversation (Berlyn, Wise and Soriano 2008). In some cases, particularly if his entry into the service system is through child protection or family services, his family life will enter the conversation, and he may be referred to a service or program aimed to develop his parenting skills, to help him change coercive or violent behaviours, or to increase the quality of family communication. This is certainly an improvement on being treated as an isolated individual, yet very often it will still fail to grasp the breadth of his family life. In this article we have explored how being a member of a family assemblage can have implications for every aspect of parent’s lives, beyond the specific times when they are engaged in parenting activities, and beyond their sense of parental identity. The family assemblage is a territory comprising environments, objects, people, discourses and affects, and for fathers experiencing mental ill health it is from within this territory that recovery advances or retreats. The concept of family assemblage prompts researchers and practitioners alike to ask questions such as: What does family mean in the life of this individual? What has it brought into their life? What has it excluded? Where do its borders lie? What kinds of associations, between what kinds of actors, objects and forces, are involved in the production or mitigation of this parent’s mental ill health and recovery? Which actors, objects and forces might best be mobilised in the design of recovery interventions? Although much work remains to be done in developing the concepts and problems introduced in this article, this article has opened up the family assemblage to a wider range of critical inquiries and clinical interventions by which the recovery of fathers experiencing mental ill-health, and the family assemblages in which these fathers emerge, may be supported more effectively.