When the clinic becomes home: on the limits of kinship care in an eating disorder treatment centre in Italy

Abstract Drawing on fieldwork in a public residential facility for eating disorders in central Italy, the paper examines the relational temporalities of therapeutics by looking at how time affects treatment at the intersection of professional and family care practices. In arguing that ‘chronic cases’ put into question the specific kind of kinship care that is at the basis of treatment, the paper contributes to the anthropological literature on eating disorders by bringing time under the analytical lens, and to the literature on ‘chronicity’ by complicating simplified assumptions about structural care problems. In addition, the paper draws on and goes beyond anthropological works that have highlighted the potentially harmful side of kinship – including those that have explored how kinship can be framed as a source of mental distress and at the same time as a therapeutic tool. Kinship as a therapeutic tool here becomes risky because professionals need to borrow from kinship practices in their own work with patients, balancing those with the necessary clinical detachment. The paper shows that the time chronic patients need in residential treatment generates a particularly complex mix between what is seen as ‘functional’ and what is seen as ‘dysfunctional’ in kinship care, because the ‘efficacy’ of the kinship work that is at the basis of treatment rests on that being partial and temporary. Long term care in the facility complicates what otherwise allows clinical detachment: the treatment team ends up literally substituting the patient’s family, with professional and family care mixing ‘too much’ with one another.


Introduction
One morning in January 2019 I found myself in a hospital ward in central Italy, where I had been told I would find Maia.Maia was a 33-year-old woman diagnosed with anorexia nervosa, from a small town nearby.She had been recently transferred to the hospital from Il Giardino, 1 the residential facility for eating disorders where I had been conducting fieldwork and working as a volunteer for the previous seven months.The treatment team there had asked me if I could visit her to make her feel supported by 'familiar faces' .Maia had received a diagnosis when she was 15, and had been in and out of Il Giardino (the time 'out' spent mostly in hospitals and other residential facilities) for the last ten years.She had been admitted to hospital because her continuous weight loss over the previous weeks had put her in a situation deemed to be at immediate risk for life.The drip for intravenous nutrition that her arm was attached to was a visible reminder of that.'Ohhh Giulietta, you found me!' , Maia greeted me with her characteristic enthusiasm.I smiled and asked her how she was doing.She told me that things were 'going well' -that she was regaining weight and managing to avoid 'hyperactivity' (iperattività), the technical term used in the world of eating disorder treatment to refer to patients' tendency to constantly move with the aim of 'consuming calories' and therefore 'losing weight' .The hardest part, Maia remarked, was 'staying away from Il Giardino and from you all' .She could not wait for the moment she would be allowed to return to the facility, and yet she said she was worried that once returned she might also go back to her 'hyperactivity' .This comment was striking in that it strongly resonated with the way most other patients I had met talked of their stay in the facility: the hardest part of residential treatment, many had been telling me, was being away from home and their families; they could not wait to return home, and yet were worried that once they found themselves in familiar spaces they might easily slip back into 'dysfunctional habits' (abitudini disfunzionali): restricting food, purging, 'hyperactivity' , using laxatives, and so on.In Maia's words, instead, there was no desire to go 'home' .She wanted to go back to Il Giardino.How could this make sense?By looking at Maia's long history with her condition, it will become clear how the many years that she spent in treatment had turned Il Giardino into 'home' for her.Yet, the very relationships with the professionals that made it home also engendered what put her life at risk.
Patients diagnosed with anorexia and bulimia nervosa frequently experience severe health complications.Longitudinal studies show that they have the highest mortality rate of all mental illnesses, with a crude mortality rate between 4% and 5% (Smink, van Hoeken, and Hoek 2012).One of the reasons behind this figure, the scientific literature suggests, is that rates of relapse after treatment are very high, and between 10% and 23% of cases develop into 'chronic conditions' (Berkman, Lohr, and Bulik 2007).
Here it is important to pause on what 'chronic' means when one talks of eating disorders.In their introduction to a special issue on 'new chronicities' , Greco and Graber (2022, 3) have noted how, although recent anthropological work has already criticised the dichotomy between acute and chronic as failing to consider blurred contexts in which limited access to treatments affects whether a condition can turn from acute to chronic, further unpacking the nebulous notion of chronicity can generate new research directions on 'the transformations of diseases and illnesses, the materiality of care, the temporalities of illness, and the work patients and families do to access treatments in order to achieve the status of new chronic patients ' (ibid., 10).In the cases they consider, chronicity is mostly something that is hoped and desired by patients and their families, as it amounts to 'live with' rather than 'die of ' diseases that used to be quickly fatal. 2 In her own contribution, Greco (2022, 110) reported that many patients and professionals contested the recent shift towards calling metastatic breast cancer a 'chronic condition' because of the misleading association with manageable, non-life-threatening diseases (ibid.:112).Some, instead, considered it as a 'recurrent condition' , as 'a series of acute episodes linked together by areas of more chronic stabilization ' (ibid., 114).As we will see, eating disorders could be considered recurrent conditions too, but here the association with 'chronicity' , if present, is all but positive: since the premise is that eating disorders are curable with swift access to proper treatment, once a case is considered 'chronic' it means that something went wrong -that a person who could be freed from a very debilitating and life threatening condition will, instead, probably have to not only 'live with it' for the rest of her life, but also, at some point, probably die prematurely because of it.Families tremble at the idea that their children might become 'chronic cases' , and professionals experience futility when they find themselves dealing with them.Chronicity here is not a hope tied to access to biomedical innovations; it is an anxiety seen as the result of lack of proper treatment.
Several anthropologists (e.g.Lester 2019;Myers 2015;Pinto 2014) have indeed shown how, for conditions that are biomedically classified as pertaining to 'mental health' , even when people are treated, treatment itself can contribute to this undesirable chronicisation.Such scholarship depicts inpatient treatment as riddled with ambiguous, if not outrightly harmful 'care' .This is also the case for eating disorders: previous ethnographic works on treatment (e.g.Gremillion 2003;Lester 2019;Warin 2005) have importantly drawn attention to how inpatient facilities in North America, the UK and Australia are characterized by structural settings of care, dynamics of surveillance and treatment practices that unwittingly reproduce the very dynamics that sustain eating disorders, making treatment fail and contributing to their becoming 'chronic' .In these accounts, what appears to be required is mostly more resources and better-attuned care.Warin (2005) for instance has shown how hospital inpatient services feature the very configurations of space that support eating disordered practices, calling for research on residential programmes that position themselves in contrast to hospital-based treatment.
This paper, therefore, contributes to the medical anthropology literature on 'chronicity' by complicating simplified assumptions about therapeutic time and structural care problems.It does so by zooming in on the temporalities of therapeutics, and especially on the relational temporalities of care.Following both Warin's (2005) call, and Carsten's (2019) call for looking at kinship as a practical realm of action through time, the paper examines how the hazards of relationality that are entangled in the specific kind of kinship care enacted at Il Giardino emerge in all their potency when the time a patient spends in residential treatment lacks a foreseeable end.It shows how care transforms over time in unpredictable and complicated ways (Cook and Trundle 2020), and how the time of treatment and the time of illness intersect to have unsettling effects on care practices that might otherwise work in a snapshot.In doing this, the paper also draws on and goes beyond anthropological literature that, by looking at kinship in practice, has usefully highlighted the potentially harmful side of kinship (e.g.Carsten 2019; Pinto 2014) -including those works that have explored how kinship can be framed as a source of mental distress and at the same time as a vehicle for remedy (e.g.Buchbinder 2015; Chua 2011).In addition to all of these aspects, kinship here becomes a therapeutic tool in the sense that professionals employ kinship care in their own reparenting work with patients, by borrowing from family care practices but cleaning them of the 'dysfunctionality' that is seen to characterise patients' families.We will see that this therapeutic tool is a risky one to use, not only because of the constant need of balancing it with the necessary clinical detachment (as in Gremillion 2003), but because the specific kind of detachment that is seen as necessary alongside reparenting rests on, and is made possible by, the fact that this kinship work is both partial and temporary.Accordingly, the 'chronic' time that certain patients need in residential treatment generates a harmful mix between what is seen as 'functional' and what is seen as 'dysfunctional' in kinship care.Following Bourdieu's (1990Bourdieu's ( [1980]], 82ff) critique of 'detemporalizing science' , the paper thus pays attention to the 'properties [practice] owes to the fact that it is constructed in time ' (ibid., 98).In focusing on time, the paper also further contributes to the anthropological literature on eating disorder treatment, which has been paying much more attention to space (e.g.Eli and Lavis 2022;Warin 2005), even if in lived worlds, and in those ethnographies too, spatial dimensions can hardly be disentangled from temporal ones.Professionals, we will see, have to deal with the limits of their own care.

Methods
The paper draws on ethnographic research that took place between July 2018 and August 2019 in a public residential facility in central Italy for people diagnosed with anorexia and bulimia nervosa.The research sought to explore some of the assumptions and practices of care involved in treatment, where the patients' , the relatives' and the professionals' different constructions and experiences of the 'eating disorder' need to be reconciled.It sought to highlight how the willingness to receive care shapes the extent to which, and the ways in which, carers can give care in clinical landscapes, and how complex moralities circulate around care and expectations for 'recovery' .
The centre employs a multidisciplinary approach that is based on the coordination of about thirty professionals -the vast majority being women in their thirties and forties.The staff include two psychiatrists -one of whom is the Director -one endocrinologist, six psychotherapists, five nutritionists, a dance-movement therapist, two philosophical counsellors, five educators, a social worker who acts as Coordinator, two nurses, and three healthcare assistants.Their work is supported by a vast number of young interns and volunteers.For simplicity of expression, here I group them together as simply 'the professionals' or 'the treatment team' .About 16-18 patients -virtually all girls and women 3 -inhabit the residential facility for three to five months on average.During my stay, patients were aged 10 to 58 years old, with the vast majority between 15 and 30.Patients' socioeconomic situation ranged from those whose families owned multiple properties and spent holidays in fancy hotels around the world, to those whose families could not afford the costs of a threehour trip by car to visit during weekends.
During fieldwork, I had the double role of researcher and volunteer.I spent most of my days helping staff members with supervising meals and other daily tasks, and participating in patients' group activities.These included 'psychological' and 'nutritional' group sessions, drama therapy, dance-movement therapy, meditation and creative writing, which, together with the five meals per day and with weekly individual sessions with a psychotherapist and with a nutritionist, constituted treatment.I also attended the weekly staff meetings, the weekly support meetings for families 4 and the monthly supervision meetings for the treatment team.My role as volunteer, with its 'extra tasks' of providing support in patients' homes and hospitals, gave me access to dynamics that went beyond structured clinical encounters.
For this paper I focus on the stories of patients that particularly highlight the double-edged aspects of 'family care' through time.Parents were the main family carers in most cases, including for patients who, like Maia, were already in their thirties.This was due both to the specificity of the Italian context -with its renowned long-term cohabitation in the parental house and economic dependence on parents (see Rusconi 2004;Saraceno 2015) -and to patients' social life having shrunk after years of suffering from a condition that substantially limits social interactions, let alone intimate relationships.After situating the context of the issues addressed, the paper proceeds by showing how the specific kind of kinship care that the treatment team enacts at Il Giardino becomes a therapeutic tool that rests on that care being 'partial' and 'temporary' .It then zooms in on the story of Maia to show how her chronicity is framed as the result, and at the same time as the cause, of 'failure' in both these aspects of kinship care -with professional care and family care mixing 'too much' with one another.

Eating disorder treatment in Italy
The current landscape of eating disorder services in Italy is all but rosy.In 1998 the Health Ministry published specific guidelines for the treatment of these conditions (Ministero della Sanità 1998), calling for four integrated levels of treatment: outpatient services as the preferred option; day-service facilities as a more intensive intervention, with assisted meals; admission to hospital if the patient's condition is critical; and rehabilitative admission to a non-hospitalized residential facility for three-to-five months.Several surveys carried out by the Health Ministry, however, show that for the more than 3 million people who have received a diagnosis there are few dedicated public facilities, causing a continuous migration towards those available and long waiting lists.In the context of a shrinking Italian welfare State, characterised by cuts to healthcare expenditures (see Minelli and Redini 2015), the burden of care falls mostly on the family, and the State is seen as the main actor responsible for the numerous patients' deaths.A mother I met put this starkly: We live in Sicily and my daughter was refused admission on the ground that she was 'not undernourished enough' -she was 32 Kilos.I had to attach my daughter to a drip of forced nutrition at home, with the help of a nurse paid by me.My daughter is doing okay now, thanks to us parents and to those who treated her privately.But who would have been responsible if something went wrong?
Similarly, professionals working at Il Giardino would often highlight the 'ridiculously short time' they had to treat what were seen as extremely complex conditions requiring months of continuous care.These limited time frames are imposed by the 'authorization' that the patient's Regione 5 of residence would grant for treatment in another Regione.Since it is the patient's Regione which pays for treatment, professionals have to routinely plead for at least twelve weeks of treatment, and for the necessary extensions.Once discharged, most patients go back to their Regione, where outpatient services often lack specially trained staff.Unless patients have the means to pay for private outpatient treatment, they start meeting a therapist once every two weeks.This, professionals agree, 'blows up all the work done' , often causing what is known in psychiatry as the phenomenon of the 'revolving door'whereby patients are re-admitted in a cyclical pattern because of severe relapses.
These systemic issues play an important role in what comes to constitute 'treatment failure' and the consequent 'chronicisation' of eating disorders, and have recently been at the forefront of discussions in the national press (e.g.Palazzo 2020).Focusing exclusively on systemic factors, however, risks overlooking that something important can be missed of the patient-professional interaction, especially in cases of patients who, like Maia, receive residential treatment within their Regione of residence, and therefore are not subject to the same time limits that the healthcare system imposes on others. 6We will see that at Il Giardino the reasons professionals posit for 'failure' do not simply relate to wider structural impossibilities of doing good (as in Lester 2019), but also put into question one of the pillars of their treatment model: family-based care.Although treatment failure is seen to materialise in patients' bodies, we will see that it is mostly non-material things that are felt to be failing when 'treatment fails': relationships, and especially relationships of kinship care.
This framing of kinship as potentially dangerous to patients has a history.Since early clinical descriptions of anorexia nervosa in the late nineteenth century, family members have been described as 'the worst attendants' (Gull 1874, 26), as having a 'particularly pernicious' influence on treatment (Charcot 1889(Charcot , 2010)).Until the 1970s, the psy-sciences framed 'the family' not only as responsible for the development of the illness, but also as a 'toxic' presence detrimental to recovery.Patients were therefore seen as requiring 'parentectomy'-i.e.exclusion of the parents from treatment and work with patients to 'extricate' them from their families (Murray and Anderson 2015, 279).Starting from the 1970s, what were subsequently considered 'family therapists' challenged this view: since they saw anorexia nervosa as developing within 'dysfunctional' family contexts (see Minuchin, Rosman, and Baker 1978), then rather than separating the child from their parents, treatment should be directed at families to address their 'dysfunctions' .In Italy, the psychiatrist Mara Selvini Palazzoli (1978) founded what came to be known as 'the Milan family systems approach' .These theories laid the foundation of 'family-based therapy' for eating disorders: families became the main 'resource' to be mobilised for a successful recovery (see Russell 1987).
Despite this fixation on the family in the history of therapeutics for eating disorders, the role of the family tends to be overlooked in anthropological accounts of treatment (Eli and Lavis 2022, Lester 2019, Warin 2005).When the patient's family is present, it features mostly as an additional complication to 'good care' (e.g.Lester 2019).In Gremillion's (2003) notable exception, which examines the production of normalised bodies, subjectivities and families in a hospital-based inpatient programme in the US., the family features as the realm that offers a template for anorexigenic power relationships that can be observed within the treatment team and between patients and professionals.Gremillion's analysis of the treatment team as a 'substitute family' focuses on the gendered cultural dynamics that make the mother-child relationship the locus where clinicians locate 'the heart of the problem' , and that ultimately make treatment fail (ibid., 104).In this paper I do not discuss the gendered norms that emerge in treatment -the professionals' expectations on patients' family members and their own gendered roles and power differentials.This is not because these issues are unimportant, but because they are not central to the key question I am asking here -namely, what happens when we bring time under our analytical lens.While I address gender elsewhere (Sciolli 2021), 7 here I am interested in how this gendered work of care also has a specific temporality that can stretch and blur kin and clinical care.

Residential treatment: a partial and temporary delegation of kinship care
We can substitute your feeding role, but there are no substitutes for parents.You entrust us with their care, but there is no way of delegating your being their parents.Your parenting role should be made even more explicit while your daughters are here.It's just a different way of parenting from the one you are used to.
Family counsellor Carola, support meeting for families A key principle at Il Giardino is that patients' parents are 'too emotionally involved' to be able to effectively help their children.The only way for parents to help, I was told by the Clinical Director, is to 'temporarily delegate parts of their role as parents to the treatment team' (my emphasis).On the contrary, professionals are supposed to be able to 'stay detached' from the personal dynamics underlying patients' 'dysfunctional behaviours' , and hence to act as 'a more skilled second family' by forcing them to eat 'normal meals' and by preventing them from purging and over-exercising, while working to understand the reasons why they developed an eating disorder.
Being able to entrust professionals with the immediate care of their children -i.e.feeding them, monitoring their weight and health, being there in moments of crises -becomes a fundamental part of being 'a good parent' .A designated 'family counsellor' has the role of updating parents and addressing their worries, but parents are not told the details of what happens in treatment -what patients are fed, how much weight they gradually gain.In her ethnography, Gremillion (2003, 101) describes how parents are completely excluded from patients' treatment and are not given advice about how to later help them at home, on the assumption that 'a natural maternal feeling will perform invisible healing work ' (ibid., 110).At Il Giardino, instead, parents are very much involved in treatment, so that they can gather all the tools necessary for when patients will return home.Moreover, there is no assumption of, let alone confidence in, any kind of 'natural' parental healing.Here 'proper' healing work is the result of effort and training by professionals.After admission, parents are expected to make patients feel that they are 'present' by visiting during weekends and by 'working on themselves and their ways of parenting' .This includes learning different ways of managing arguments with their children and how to manage food-related issues at home 'in a way that can be supportive' .Parents are required and expected to have a very specific behaviour during mealtimes when patients are at home for weekends: they should never make comments on what patients eat, limiting themselves to following the instructions of professionals.Parents are supposed to learn 'how to behave' by coming to the weekly support meetings and to the appointments with the assigned family counsellor.At Il Giardino, the delegation of 'immediate care' to professionals thus comes with the expectation of other forms of care.
Lack of this 'family work' is one of the greatest causes of frustration for the treatment team, who often joke about wanting to set up 'a parallel residential facility for parents' .Unlike in clinical contexts such as hospitals for incurable conditions in children (e.g.Mattingly 2010), here recovery is seen as inherently possible.This aspect sets up the possibility of blame: if treatment seems to be failing, it is 'people' that are to blame, not 'biology' or 'the disease' .One complication is that this care work is supposed to last for years -Mattingly (2010, 27) calls it 'chronic homework': for those who are lucky, two-to-five years are considered necessary for 'complete recovery' , but relapses are common and many people end up living with an eating disorder through the rest of their lives.
If the 'family work' done by parents is supposed to last, the one done by the treatment team is supposed to be temporary.Il Giardino was born in response to hospital-based inpatient services, and explicitly aims to reproduce, as the Clinical Director put it, a 'temporary but ideal second home' for patients.Located in a sixteenth century villa whose walls enclose a garden surrounded by 300-year-old trees, its rooms feature high ceilings, terracotta polished floors and wooden furniture, and are decorated with paintings and pictures.I have often heard professionals, patients and parents alike say that Il Giardino 'sembra una casa e sa di casa' ('looks like and tastes like a house').In this 'house' a lot of effort is put by the whole treatment team into recreating 'un ambiente materno e famigliare' ('a maternal and family-like environment'): professionals wear casual clothes and have a dedicated bedroom for the nightshift; after dinner, they spend time in the living room chatting, doing puzzles, watching movies or playing board games with patients.
At the same time, however, professionals say they need to 'be firm' , to 'contain patients' irrational thoughts' and, most importantly, to 'avoid taking their difficulties with food personally and overreacting, which is what their parents tend to do' .The detachment these professionals need and acquire, therefore, is not a lack of emotional display (cf. McDonald 2015), but more a neutral stance, as the psychotherapist Carlotta explains: It's as if during a football match a player throws the ball to me -and I am the referee -asking me to start playing for his team.But I know that, in my role, I can't.At this point the player might get angry, and this happens here.But knowing that I have this role and that I cannot take sides allows me to bear that person's anger and the other negative emotions that might follow much more than I could in a personal situation.This holds for virtually everything, including the assistance with meals, when patients ask you to let them leave the table without having finished their food.
Carlotta's statement became especially clear to me after I met Katia.Katia was a 19-yearold girl from a large city of a Regione nearby, who had been diagnosed with anorexia nervosa when she was 13.On admission, Katia was accompanied by her mother, a woman in her fifties whose face carried the signs of years of difficult care work.During the weekly staff meeting that day, the Clinical Director reported that Katia was an extremely difficult case, that nobody else had apparently 'managed to treat' and that clearly tended 'towards chronicisation': after several failed attempts with outpatient treatment, countless hospital admissions, and three residential treatments in other parts of Italy, professionals located Katia's main problem in a 'highly symbiotic relationship with her mother' .This 'unhealthy relationship' , they held, had developed during years of illness, but was traced back to her 'disastrous family history' .Katia had lost her father -who used to be violent with her motherdue to sudden death when she was 9 years old, and had since then been raised by her single mother.Two days after her admission Katia tried to escape, and then became so resistant to treatment that the only way the Clinical Director convinced her not to self-discharge was to allow her to stay in the residential facility from 11 am to 3 pm only, when she would go home with her mother.At 4.30 pm, a volunteer from Il Giardino would go to her house and stay with her until 9 pm, making sure that she ate her afternoon snack and dinner and that she did not spend time walking around town.
I was among the 'volunteers' recruited.Katia's mother had been asked to leave the house while volunteers were there for the assisted meals, but she almost never did because Katia would start crying.When my turn came, with difficulty I managed to send the mother away for a walk at the time of the afternoon snack.I was chatting with Katia after she had finally finished her snack, when her mobile phone rang.It was her mother telling her that she was in the square two minutes away from the house.Predictably, Katia wanted to join her mother, and there was no way of keeping her at home.That evening, for dinner, Katia only ate the salad sent by the facility, substituting the first course with some chickpeas and the second course with a few slices of turkey breast.The mother ended up eating the meal sent by the facility, and I could only make sure that Katia ate at least all the things she herself had chosen.
Eventually, the difficulty for insufficiently trained volunteers to assist meals in Katia's house led the Director to decide that Katia should be left at home with her mother every day from 5 pm.Professionals told me that this whole arrangement was unprecedented and beyond any regulation: 'It runs against the principle of making parents delegate their feeding role to us.It's clear that the mother can't do it anymore, that the two together don't work' .The reasons for things not working was ascribed by professionals to the fact that Katia's mother let her do whatever she wanted, 'because she has a neurotic interlock with her: she is completely dedicated to her ill daughter.Once Katia recovers, she will lose her aim in life' .The implication here was that this was the reason why -'unconsciously' -the mother was counteracting treatment with her presence.Professionals used the psychological term 'enmeshment' (invischiamento) to denote Katia's and her mother's 'unhealthy' interdependence, suggesting that it is possible to 'care too much' .
But what happens if professionals start caring 'too much' , too? Staying 'neutral' or 'detached' was not depicted as an easy or 'natural' task, as the educator Marta points out: It's difficult because we get attached too.During nightshifts, we see patients sleeping in their beds, brushing their teeth, straightening their hair… it's an intimate thing.But we must remain distant.As professionals, we have the duty to do it.
These invocations of clinical detachment are constant refrains.Professionals strive to be caring in a way that mimics the family, but they are aware that they should not become 'too emotionally involved' .As Marta's words suggest, this is the ideal -what the treatment team strives for -not necessarily what always happens in practice.Here one can see in a nutshell a key problem of the treatment apparatus: since one of the main causes contributing to the development of eating disorders is seen to lie in 'dysfunctional' kinship relationships, therapeutic relationships are seen as key to recovery.They are believed to be the tool through which patients can experiment with what they could not experience in their families.The problem is seen to lie in the fact that, as Myers (2015) has noted in her ethnography of a mental health clinic in the US, maintaining boundaries between patients and professionals thus becomes difficult.This holds for any patient in the treatment centre, but specifically so for chronic patients.To illustrate why this is the case, let us now go back to Maia's story.

Maia: when kinship care and professional care mix too much
You need to keep in mind that they are not your daughters: they are guests, patients of this treatment centre.
Psychotherapist Mauro 'Oh, you met Maia then, our little scab' , I was told by the psychotherapist Michele at the end of my first day of fieldwork.We were in the kitchen, and I was helping him to load the tray trolley for the patients' dinner.I had just come back from a 'guided tour' of the facility, arranged for me by the Clinical Director.I had soon realised that my guide was not a professional nor an intern, but a patient.Her blonde thin hair, puffed up in a way that made it the most voluminous part of her body, seemed to be leaving a trace of pale light behind while she walked.'Maia will tell you everything -she knows more of this place than we do!' , the Director had commented when Maia was called to give me the tour.This comment, I would later learn, related to the fact that Maia had been an inpatient at Il Giardino for almost four successive years.
Professionals often mentioned Maia as an example of both how 'treatment doesn't work if the family doesn't work, too' , and of how their own reparenting work had 'gone too far' .During my stay Maia's parents never attended the support meetings for families and I managed to see them only a couple of times, in situations of emergency.The fact that Maia had been ill for more than half of her lifetime made professionals 'make sense' of her parents' absence: they were 'understandably exhausted' after eighteen years of their daughter's illness, and their old age now meant they did not have 'the energy' to be much involved in treatment anymore.Still, most professionals depicted Maia's parents' behaviour as unacceptable.When Maia's conditions worsened to the point that the treatment team decided to admit her to hospital, they strongly advised Maia's parents to hire a professional that could have the role of 'meal controller' , because hospital staff are not trained nor paid to monitor patients' meals.Maia's parents refused to hire someone and her mother eventually took the role of 'meal controller' for the two weeks of hospital stay, with professionals claiming that they were sure Maia would be able to deceive her.The problem came up again when, upon discharge from the hospital, Maia's parents were told that she would have to stay at home for a while before being readmitted to residential treatment: her condition was still precarious enough to require the intravenous feeding that she had been receiving in the hospital on top of regular meals, and that was not possible in the residential facility -so the treatment team had arranged for a nurse to be present at home for her.The Coordinator of Il Giardino was the one who called Maia's parents to inform them of this decision.She reported to the rest of the team that the parents had reluctantly said: 'We will see what we can do.Then, it will be what the Lord wants' .She said she had felt she 'wanted to get her hands on them' when she heard them say that.She continued: 'But we can't abandon her.We have to keep trying to help her' (original emphasis).Professionals were very critical of the 'delegation of parenthood' that Maia's family was gradually pushing for, but it seemed that they themselves had already 'internalized' it in a way.
From the beginning of fieldwork, it was evident that Maia was a special patient.The first notable aspect was her absence from shared spaces.Patients were supposed to spend as little time as possible alone in their bedrooms, and were often called if they failed to show up in the living room during the 'relaxation time' between meals and group activities.Maia, however, seemed to be exempted.She also seemed to be exempted from participating in the otherwise compulsory group activities that would not normally start until every patient was in the therapy room.Why?The younger professionals told me this was just 'the way it is' -that it was impossible to keep running after her with all the other things to do, and pointless because she would always find a way 'to sneak away' .Those who had been working in the facility for longer would answer in a way that expressed apprehension and a sense of guilt: 'We don't know what to do with her anymore.We are failing her' .
What did 'failure' mean here?Gradually, I came to realise that the 'failure' professionals referred to was equated with 'giving up' on the fact that Maia would ever recover and have 'a life outside' .They felt they were now concerned with 'just keeping her alive' , and now even 'keeping her alive' seemed to have become difficult: Maia -who had been 'stable' at an already very low weight for a long time -had started losing more and more weight.Professionals reported finding 'traces' of some of the ways in which she was perhaps managing to do that: notwithstanding a personal assistant during meals, Maia somehow spread breadcrumbs on the floor and tomato sauce on the edge of her chair, unnoticed at the time.Some said they thought Maia was hiding food in her clothes.However, the team agreed that this could not possibly be enough for her to lose so much weight, considering the high calorie meal plan that she had.The main problem was instead located in her 'hyperactivity' -in the fact that whenever she was not watched, she would constantly move.Volunteers reported seeing her walking up and down the stairs located in the bedroom area.Those on night shifts suspected that she spent most of the night walking in her room.Frustratingly for those treating her, Maia did not deny any of this.She recognised that she could not control her 'urge to move' , and even asked the treatment team to place a gate at the entrance to the bedroom area to prevent her from sneaking out and spending hours moving up and down the stairs.Since this had already been attempted in the past with no results (Maia had started 'climbing' other stairs), some suggested putting Maia in a wheelchair to reduce her movement.Others, however, quickly reminded the team that they had already tried with a wheelchair too, and it had not worked: 'We would have to tie her to the wheelchair to prevent her from moving!' Besides the supposedly obvious risks to Maia's physical health, her behaviours were a source of professional concern because they seemed to be shrinking her relational world.During a staff meeting the Coordinator noticed how Maia's world was 'reduced to those little stairs behind the bedrooms, where she walks up and down.How do we shake her out of that?' .The treatment team eventually came up with what was seen by some as 'a crazy idea': 'promoting' Maia to the role of staff assistant.What came to be known as 'the crazy proposal' was discussed during a supervision meeting with an external psychotherapist.Maia's case had been chosen for supervision because the whole treatment team felt 'stuck' .This was a common reason for a case to be selected for supervision, but this time something in the way the case was presented seemed exceptional: this request came from a gradual realisation on the part of the team that they were 'unconsciously treating her differently from the other patients' (Clinical Director) and that it was important to understand how that was affecting her treatment.The Director was adamant that notwithstanding the role that Maia's family dynamics had played in the development and persistence of her condition, the influence of her family was now 'virtually inexistent' .After all, she saw her parents 'barely 30 min per week': What we should reflect on is how we interact with her after all the time she has spent here.It's as if we've lost hope.The resignation that we see in her parents, she probably sees it in us.She can probably perceive that we are keeping her here just to keep her alive.She, but also we, see this treatment as an end point, not as a transition towards something else.
Maia's enactment of 'old dysfunctional behaviours'-spreading breadcrumbs and tomato sauce around her seat, refusing to leave the bedroom area -were framed as possibly a way for her to tell the team that they were 'forgetting her' .Some professionals agreed that perhaps that was what was happening: 'Rules don't exist for her anymore.We are so exasperated that we stopped trying' .Even more interestingly, some of the team read Maia's 'regressing behaviours' as symptomatic instead of a fear of being potentially discharged: 'She sees us as her family now, we are her only meaningful relationships.She doesn't have anything nor anyone waiting for her outside' .
In her account of psychiatric clinics in India, Pinto (2014) argues that those clinics are not so much a zone of social abandonment in which the normal is reasserted by keeping the unhuman confined there, but a place where people with very human difficulties end up when kinship ties fall to pieces.At Il Giardino, what is reasserted is the ideal configuration of the kinship ties that are seen to have broken.When the reconfiguration fails, as in Maia's case, the patient is bound to stay in the clinic.However, unlike Pinto's interlocutors, Maia does not want to get out -there is not 'an out' for her anymore.This is what prompted the 'crazy idea' of the clinical staff.The plan was to ask her to help the Coordinator carry out administrative tasks.Maia would then feel reassured that she would not be 'abandoned' if her health improved but would always have a place in the facility.Arguing against this proposal, Maia's nutritionist observed that it was absurd to even imagine Maia doing any kind of work, considering her physical state: 'We tried to give her jobs in the past, when her conditions were much better than now.She always tends to get worse physically once she starts working, and what do you do then?She literally burned every idea we came up with' .She continued that Maia lacked 'the minimum motivation necessary for treatment' and that other patients with similar behaviours had been discharged because that kind of treatment was not seen as appropriate: 'She is putting us at the limits of our capacities.Is it possible that there is no alternative to keeping her here?' .When the external psychotherapist asked why this case was different from the many others the team had dealt with, the Director replied that it was because Maia was 'from here' (i.e. from the same Regione): whereas other patients had to be discharged at some point because their authorization expired, with her this was not the case.As anthropologist Davis (2012, 221-223) pointed out in her analysis of psychiatric treatment in Greece, the dilemmas generated by patients breaking the rules of treatment can turn into 'tedious, repetitive, inconclusive debates' that amplify professionals' responsibilities toward patients, as they implicate both patients' survival and the professionals' capacity to help, often producing 'dependencies that are nearly impossible to dismantle' .In Maia's case, the difficulty of dismantling dependencies had to do with professional care being entangled with kinship care.
As the reader knows from the opening of the paper, Maia was eventually admitted to hospital for intravenous nutrition.However, the hospital admission would only last for two weeks, after which Maia was supposed to return to a residential facility.Many team members at Il Giardino felt that Maia should be sent somewhere else, because they were now 'too involved' and therefore 'unable to help' .Others found the mere idea unthinkable for the same reason: 'We can't abandon her.We are her only family left!' (educator Flavia).The Clinical Director herself was very resistant to the idea of moving Maia somewhere else.At one point she proposed to send Maia to a different facility that had opened a few years earlier in a Regione nearby.Other professionals saw that as a clear sign that the Director wanted to do something for Maia but could not stand the idea of 'entrusting strangers with her care' .The team of that facility was in fact mostly composed of professionals who had previously worked or had been trained at Il Giardino.'But this is exactly why it wouldn't work.They have already been involved in Maia's care' , someone pointed out.What Maia needed, most professionals held, was to be admitted to a residential facility 'where nobody knows her' .Only this way, it was felt, could she be treated 'as any other patient' and have a chance of getting better.After weeks of discussions, Maia was eventually sent to a facility in Northern Italy.The Director of Il Giardino knew the treatment team there, as the two facilities had been in a partnership for a while.But the other team had not met Maia before, and that was seen to be enough.
Here we see how the reparenting model of treatment -which is supposed to be a key ingredient for recovery -becomes problematic when the time a patient spends in the residential facility lacks a foreseeable end.In Maia's case this lasted for years, with the result that the treatment team literally substituted for Maia's family in her care.The boundaries between professional care and family care had become blurred, and professional care had suddenly found itself immersed in the same issues that are typically ascribed by the treatment team to family care: excessive 'emotional involvement' , a subsequent inability to impose the rules necessary to tackle the symptoms of eating disorders, and exhaustion due to seeing that whatever one does to help, nothing seems to work.Once the treatment team realised this, they eventually resigned to the idea that they had to do what patients' families have to do when their children are admitted to Il Giardino: delegate their care and entrust professional strangers with it.
In her work on time and care, psychosocial theorist Baraitser (2017) asks what it might mean for care to 'specifically and paradoxically be understood as itself a mode of change that requires time not passing ' (ibid., 11).Time for Maia, and for the treatment team caring for her, looked akin to what Baraitser calls 'unbecoming time' -'time that is lived as radically immoveable […], a "circle with no rim"' (ibid., 4).What should have been a 'transformational time' -admission to residential treatment -gradually became 'everyday time' , as the care Maia received did not generate any steady movement towards recovery.In Maia's story we see care taking the form of professionals waiting, delaying, enduring, returning, maintaining, preserving, remaining (ibid.).Although time is suspended, however, time is also passing, with devastating effects.Two temporal scales and two kinds of bodies are at play: on the one hand there is the biomedical body with its temporality of looming death; on the other there is the body-self acquired by Maia through her history with her family and with the treatment team -a history that rendered it unresponsive to care practices that would normally 'work' .When the materiality of Maia's body starts deviating from the 'circle with no rim' , but in the direction opposite to the one professionals are striving for, this care without ending needs to be interrupted because it is recognised as being leading to Maia's death.

Conclusion
One month after Maia was sent to the new facility, the Director, the Coordinator and four other members of the treatment team drove for more than six hours each way to see her for the allowed one-hour visit.One of them, when telling me about the trip, noticed that although Maia was 'full of joy' and said she 'thought she was dreaming' when she saw them appearing in her room, the visit was probably something that had made them 'feel better' .The fact that Maia's family had not visited her yet added to the feeling that they were 'more of a family to Maia than her parents' .
This paper has shown how the reparenting model of treatment becomes problematic when the time a patient spends in the facility lacks a foreseeable end: for Maia, the kinship care enacted by the professionals shifted from being a partial and temporary therapeutic tool to assuming the total and permanent aspects of 'real' family care.It became 'too much' like family care, unwittingly incorporating its 'dysfunctions' .This arguably generated failure for all those involved, but it also made the treatment team question their practices and find other ways to care.Signs of failure prompted professionals to question their work and realise that the relationship they had with Maia had changed over time, reaching a point that made it unhelpful.This, in turn, called for coming up with strategies 'out of the box' -the gate, the wheelchair, the 'crazy proposal' -that were not part of standard treatment, but a matter of continuously 'tinkering' with the resources that were available (Mol, Moser, and Pols 2010).When it became clear that none of the tinkering within the facility would work, a new kind of care took shape in the delegation of Maia's treatment to a different team.This was not a case of careless abandonment driven by exhaustion, although the professionals were exhausted.It was a difficult and long-resisted decision that required effort.
The chronicity of eating disorders is certainly co-produced by lack of resources and of proper treatment, including stays in residential facilities that are felt to be too short.But the ethnographic material analysed here also suggests that one key working mechanism of treatment -i.e.family-based care -is strongly tied to the very fact that residential treatment is short-term.One would need to see the extent to which the time patients spend in residential treatment could be helpfully extended before one starts observing the (unintentional) harmful dynamics that ended up characterising Maia's treatment path.More attention is needed to the forms that care could take, within and outside eating disorder clinics, for the many patients who end up living with the condition for the rest of their lives, and for whom clinics sometimes become homes.

Notes
1. Literally 'The Garden' , it is how I call the facility to maintain anonymity.2. For example, see Feudtner (2015) on Type 1 Diabetes.3. In fourteen months, two male patients were admitted.4. Held by two staff members, these meetings aim to provide support and training to patients' relatives.5.In Italy the central government controls the distribution of tax revenue for publicly financed health care.The Regioni -the first-level administrative divisions of the Italian Republic -are responsible for organising health services through local health units.6.Or in managed care contexts like the US (cf.Lester 2019).7.There I examine how the 'substitute family' takes a different shape at Il Giardino due to different gendered dynamics, and I go beyond focusing on the iatrogenic effects of the substitute family to highlight how kinship care in this treatment centre highlights something that is interesting to the anthropology of kinship and of medicine more generally -namely that kinship relations are seen to need remodelling in order to transform patients' eating behaviours, and, through those behaviours, their bodies.This suggests a co-substantiation of food and family but in a way that differs from existing anthropological accounts, as kinship and feeding are here aligned in both the production and treatment of disease.

Ethical approval
The research project on which this paper is based received ethical approval from the University of Cambridge's Social Anthropology Department and was approved by the Clinical Director of the treatment facility, as it was felt to be in line with its ethical guidelines.
having shared with me their stories and struggles.My sincere gratitude goes to all the members of the treatment team, for having welcomed me so warmly and for having shared with me so openly the many challenges of their difficult work and lives.Finally, many thanks go to my PhD supervisor Maryon McDonald, to my postdoctoral mentor Perveez Mody, to the editors of this Special Issue and to the anonymous reviewers for their dedicated feedback and support.