Relationship-based practice and the creation of therapeutic change in long-term work: social work as a holding relationship

ABSTRACT Relationship-based practice has become an influential theory through which social work practice is understood. However, much more critical attention needs to be given to the kinds of relationships involved. This paper is based on an ethnographic study of long-term social work that spent 15 months observing practice with service users and organisational life to find out how social workers establish and sustain long-term relationships with children and parents in child protection cases, or don’t. The paper introduces into the literature the concept of a ‘holding relationship’, which was present in several of the cases we studied, especially where therapeutic change occurred. The aims of the paper are to document the nature of a holding relationship and to develop it as a concept. A ‘holding relationship’ involved social workers being reliable, immersing themselves in the service user’s day-to-day existence and developing their life-skills, getting physically and emotionally close to them, and practicing critically by taking account of power and inequalities. The concept of a ‘holding relationship’ draws on psycho-dynamic and sociological theory to provide new ways of thinking that can help make sense of the practical and emotional relating involved in social work and promote the development of such helpful relationships.

social workers establish and sustain long term relationships with children and parents in child protection cases, and what occurs when they don't, and how practice is influenced by organizational cultures, office designs and forms of staff support.
'Relationship-based practice' has become a highly influential perspective through which what goes on between social workers and service users is theorised and analysed Trevithick, 2012). But the presence of a relationship in itself explains a limited amount, we need to go much further by drawing out the kinds of relationships social workers and service users have and their effects. The research involved 15 months of fieldwork during which a sample of 30 cases were shadowed for as long as they were open for up to a year . The findings show that relationships between social workers and families in long-term casework took several forms. Some parents were opposed to statutory child protection involvement and the relationships were invariably tense and sometimes hostile (see, Ferguson et al., 2020a). Others went through cycles of meaningful engagement and detachment as social workers as well as family members disengaged at times. Some were reciprocal and cooperative, or became that way after a beginning where the service users were reluctant to be involved. Some such relationships were therapeutic, in the sense that service users were seen regularly on meaningful encounters that had the effect of helping them to change. It is the latter kind of relational work on which this paper focuses and introduces into the social work literature the concept of a 'holding relationship'.
A holding relationship was present in several of the cases we studied, especially where therapeutic change occurred. The paper has two key aims: to document the nature and dynamics of such a relationship; and secondly, to develop the concept of a 'holding relationship' to provide a theoretical and analytical perspective that can help make sense of the interactions, practices, feelings, the relating involved in social work. The analysis of the data draws upon psycho-dynamic theory (Kanter, 2004;C. Winnicott, 1963) and sociological perspectives on power (Featherstone et al., 2018) to show how a holding relationship involves social workers incorporating four key components in their practice: being reliable; immersing themselves in the day-to-day existence of service users and helping them develop life-skills; getting physically as well as emotionally close to service users; and practicing critically in ethical ways that take account of power relations and structural inequalities, using 'good authority' (Ferguson, 2011).
After outlining the methodology of the research, the paper focuses in detail on a small number of cases that bring the nature of practice and the relating that went on to life. These case studies illustrate findings that were apparent across all of the cases where meaningful therapeutic change occurred through a holding relationship that was skilfully crafted by social care practitioners and service users over the course of a year.

Researching practice and long-term relationships in social work
The fifteen months of fieldwork were conducted simultaneously in two Local Authorities in England, in long-term social work teams whose primary role was child protection. We observed and audio-recorded encounters between service users and practitioners and also observed how social workers and managers worked and interacted in their organisations (Leigh et al., 2020). Social workers were shadowed everywhere they went, around offices, walking in the community, in cars, and into and around service user's homes. This mobile, sensory ethnographic approach enabled us to get close enough to practice to not only see and hear what was going on, but to feel what social work encounters and relationships are like Ferguson, 2016a;Pink, 2015). Research team members spent 402 days in the field and a total of 271 practice encounters between social care staff and service users were observed, of which 146 were home visits. Fifty-four staff supervisions were also observed and 54 interviews took place with families. We also extracted data from social work case files about the total work that was done over the year.
The first three months of the 15 months that were spent alongside social workers were used to identify a sample of 30 cases that were then shadowed for as long as they were open for up to a year, which enabled us to draw out how relationships were made (or not) and the processes of 'change in the making' (Neale, 2019), or its absence. We shadowed 15 child protection cases at each site, a sample that provided a broad range of cases and enabled the depth of inquiry needed. We knew that cases classified as high-risk were likely to have significant involvement but we had no way of knowing at the outset how long the casework would last. Twelve of the 30 sample cases stayed open and were shadowed for the full 12 months, two for 11, three for 10 and all 30 were open for at least five months. In the 12 month case studies the lowest number of encounters between social care workers and children and families we observed was 10 and the highest 21, with many other interviews with professionals, service users and observations of organizational practices on top . The data was inputted into QSR NVivo 11 software, coded, and thematic analysis and standard techniques of constant comparison were used to produce findings (Bryman, 2012).
A case study method provided a way of capturing what happened over time by assembling all the data that had been gathered on each family/case into a chronological narrative. The 30 case studies of long-term casework show in enormous detail the amount and type of work that was done with the family, the nature of the relationships and what was going on in the organization at the same time. The sheer size of each casestudy and the insights they give into the detail of practice is a key justification for focusing this paper around a small number of cases, while drawing from the wider pool of insights generated by the entire sample. Doing so raises questions about the level of rigour in presentation of the findings. Donald Winnicott (1989, p. 369) claimed that 'one case proves nothing, but it may illustrate much'. Wengraff (2001) argues that rigour can be achieved by analysing the data in ways that enable the identification of 'focal' or 'gold-star cases' within qualitative research samples that illustrate the general research findings particularly well and that is our approach here.
The research was ethically approved by the social work agencies who took part and the research ethics committees of the participating universities. Professionals and families were only shadowed and interviewed if they gave informed consent. While the case examples used in what follows reflect actual events and findings, details have been changed to protect the anonymity of the families, professionals and the research sites.

Relationship based practice: towards a holding relationship
The concept of a holding relationship that we are developing from the data is informed by two kinds of research and theory: psychotherapeutic and sociological. Our approach draws from the expanding body of relationship-based and psycho-dynamically informed work that has enriched social work in recent years. This has arisen in part because of how since the 1970s and 80s the growth of mangerialism has pushed social work towards becoming 'a technical/rational rather than an ethical and relational endeavour' (Ingram & Smith, 2018). Social work scholarship has sought to foreground the emotional, unconscious and relational components of people and practices (Ingram, 2013). Ruch (2018) shows how the concept of relationship-based practice has developed by integrating theory from psychoanalysis, attachment theory, systems theory, politics and sociology and now offers 'a joined-up way of thinking about relationships that acknowledge the visible and invisible, conscious and unconscious components that comprise all relationships and recognises the important connections between the intrapsychic, interpersonal and broader social contexts in which they are embedded' (Ruch, 2018, p. 28). In this vein, Parkinson (2018) shows the challenges involved in sustaining relationships where hopelessness, depression and despair are present. Smith (2018) has analysed the dynamics of sustaining relationships where there is anger, aggression and hostility, and Turney (2018) considers relationships where love and positive feelings are the strong emotions involved. Turney (2012, p. 153) argues that the concept of relationship-based practice seeks to avoid 'psychologizing' the lives of service users and social workers, at the expense of leaving out the wider social and political context within which relationships go on (see also, Ruch et al. (2018), while Coulter et al. (2020) suggest that greater theoretical coherence can be achieved by locating such relational work within systems theory.
While conceptually valuable, this work shares with most social work research the limitation that what happens in actual social work relationships is not shown because of the use of methods that have kept a distance from practice. While a body of work has emerged based on observations of one-off encounters between practitioners and service users as there are going on (Ferguson, 2016a(Ferguson, , 2016bD. Forrester et al., 2019;Henderson, 2018;Noyes, 2018;Ruch et al., 2017;Winter et al., 2017), our research broke new ground by being the first ethnographic study of long-term practice and relationships.
Understandings of relationship-based practice must be grounded in what is known about the nature of the problems that affect the lives of those who come to the attention of social workers. Studies of child and family social work and child protection have consistently shown a relatively low level of actual serious injuries to children (Lonne et al., 2009) and many more families with multiple and often enduring difficulties (Morris, 2013). Children and families referred to the social work teams we studied fitted such a profile. Some (a minority) concerned physical and/sexual abuse of children and sexual exploitation of young people. Most involved children and women struggling to cope with the impact of domestic abuse from men, addictions, poverty and fractured relations with their wider families. They were classified as either emotional abuse and/or neglect, reflecting national trends in how those categories of problems dominate in child protection plans (Department for Education (DfE), 2019). Several of the women had already had children removed from their care and some had experienced repeat removals of babies. Broadhurst et al. (2015) show that in recent years a quarter of mothers whose children were removed into care have experienced repeat removals. Some specialist services are emerging to try to prevent this, but despite statutory social work services having a central role, what they do and their contribution to preventing children being removed has barely been researched. This paper helps to fill that gap.
In doing so, we want to expand the concept of relationship-based practice by drawing upon the rich literature on working with vulnerable children and families that goes back to the 1950s and in particular the concept of 'holding' developed in the influential work of Donald Winnicott, a celebrated paediatrician and psychoanalyst, and Clare Winnicott, an eminent social worker and academic. They began their collaboration during the Second World War working with evacuated children with special needs in Oxfordshire and later married. Clare worked as a social work academic at the London School of Economics (LSE), trained as a psychoanalyst and was appointed as Director of Child Care Studies in the Home Office in 1964 where she initiated programs to train child care staff across the UK. Meanwhile, Donald, while pursuing an active life in the psychoanalytic community, taught social workers at the LSE from 1947 to 1970, published in social work journals, and often spoke to social work audiences (Kanter, 2000(Kanter, , 2004. For Donald Winnicott (1957), the prototype for therapeutic relationships between therapists and clients was the mother-infant relationship. Although usually expressed in terms of mothers, unlike most of his contemporaries Winnicott gave some attention to fathers (J. Forrester, 2017). He described the way a mother becomes sensitised to her baby by feeling herself into her infant's place and referred to a mother's unconditional love and deep empathy with the baby's predicament as creating a 'setting' for the infant, a 'facilitating environment', 'in which her infant may be able to be and to grow' (Abram, 1996, p. 261, original emphasis). A social and psychic environment is created that is crucial to the emotional health of the infant. Winnicott discovered the difference between a good environment and a not so good one and applied these ideas to the setting therapists create for their clients and called it the 'holding environment'. Within a 'holding environment' service users who have experienced harm may have a chance to heal through the therapeutic relationship and setting (Abram, 1996, p. 166).
As we are developing it for social work, a 'holding relationship' goes on within the holding environment. The notion of 'holding' has similarities to Bion's (1962) concept of 'containment' and they are sometimes used interchangeably to refer to how a parent accepts and takes in their infant's anxious feelings and returns them in a digestible form, providing emotional reassurance, comfort and security. It is also relevant to when adults are 'held in mind' by being helped by trustworthy people to process their emotions and gain clarity in their thinking-such as what supervisors can provide for social workers (Ruch, 2007). 'Holding' as we are developing the concept here is a more expansive experience than containment in that it refers to both individual acts of care and a network of care and control that envelopes the service user. As John Forrester (2017) writes: 'Holding for Winnicott is about "care-cure". By being held the patient can (re)experience and repair the past; a maturational process occurs.' The language of 'care-cure' and 'maturational processes' may sound old-fashioned, medical, and even oppressive, but it is similar in spirit and method to aspects of what in social work today is called relationship based practice. Similar forms of holding work are also practiced and theorised by leading psychotherapists (Music, 2019).
What the concept of the holding relationship being developed here incorporates with greater clarity than the theory and practice of 'care-cure' is a sociological perspective on power that takes account of inequalities as well as the complex psychological and social dynamics of helping encounters (Bywaters et al., 2016;Featherstone et al., 2018). While social workers in our study were limited in what they could do to change such inequalities, what they could do-and many did-is take account of their impact and when this was done it constituted what we are calling 'ethical holding'.
Clare Winnicott expressed what was involved in creating reliable, ethical, effective helping relationships through a holding environment as follows: I think it ['the caseworker's basic technique'] lies in the provision of a reliable medium within which people can find themselves or that bit of themselves which they are uncertain about. We become, so to speak, a reliable environment, which is what they so much need: reliable in time and place -and we take great trouble to be where we said we would be, at the right time. . . . We are not only reliable in time and place but in the consistent attitudes which we maintain towards people. They know how they will find us. Here again we take deliberate trouble to remember all the details about a client's life and not to confuse him (sic) with other cases. We can "hold" the idea of him in our relationship so that when he sees us, he can find that bit of himself which he has given us. This is conveyed by the way in which we remember details and know exactly where we left him in the last interview. And not only do we hold a consistent idea of people, but we hold the difficult situation which brought the client to us by tolerating it until he either finds a way through it or tolerates it himself. If we can hold the painful experience, recognizing its importance and not turning aside from it as the client re-lives it with us in talking about it, we can help him to have the courage to feel its full impact . . . I have deliberately used the word "hold" in what I have been saying, because while it obviously includes "acceptance" of the client and what he gives us, it also includes what we do with what we accept. To sum up, the professional relationship is the technique whereby we provide a limited and enclosed environmental setting which is personal because it contains all that the client has put into it himself, and which is reliable because it is accepting and holding.' (C. Winnicott,in Kanter,p.152, original emphasis).
Our methodology of shadowing encounters between social workers and service users from week-to-week, month-to-month over the course of a year enabled us to explore the presence or absence of this kind of reliability and holding. We were able to find out how service users and social workers behaved and felt over time, what difficult feelings and situations practitioners accept and hold on service users' behalf; in effect, the kinds of relationships they create and how change happens.

The practice of a holding relationship
A striking feature of the cases in the study where there were holding relationships was the regularity with which families were visited by social workers and in some by other services like family support. Social work in the UK is often preoccupied with what social workers cannot do due to high caseloads and bureaucratic demands limiting the time available to spend with service users (Munro, 2011). In some cases we studied holding relationships were not established because either the family did not wish to be held (see Ferguson et al., 2020a), or social workers could not be reliable due to high caseloads and staff turnover and the emotional complexity of the work. While social workers often felt exhausted, our findings show how social work some of the time has a significant amount of involvement with some service users and the delivery of a reliable service where staff feel held and supported by the organisational culture and supervision is fundamental to holding and helping them (Ferguson et al., 2020b).
In some instances the holding relationship went on through relatively little contact between the family and social care. The 'Nicholls' family had just eight social work visits over the eight months we shadowed it and no family support involvement. The mother, Maria, had been involved with social work on and off since a child. She had her children taken into care some years previously due to drug addiction and a very abusive partner. She became pregnant again with a new partner and social work did a pre-birth assessment, which was positive and they kept their baby girl. The child was very well cared for and by the end of the research period she had come off the child protection plan and the case was closed. The pattern of casework over the eight months it was open after the baby was born was monthly home visits. The holding environment also included a drugs counsellor, health visitor and doctor. On all of the (seven) encounters we observed the social worker gave quality attention to the parents and to the baby, on one visit holding her for 12 minutes. This involved the social worker holding the baby in front of her and making eye contact, smiling and talking to her, bouncing and jiggling her and allowing the baby to nestle calmly in her arms. Both Maria and the social worker explained in research interviews how, due to very negative past experiences with some workers, Maria mistrusted the social worker when they first met. She insisted that everything the social worker wrote down she must show it to her and the worker always did this and trust developed. This typifies our finding of how in all cases the holding relationship had to be made, and often had to be fashioned out of an initial deep suspicion of social work. This speaks to another important dimension of 'holding' recognised by Clare Winnicott, where the social worker 'creates and maintains a professional relationship for [the service user] which, among other things, eliminates fear of retaliation' (C. Winnicott, in Kanter, 2004, p. 175). Practitioners would respond empathetically to service users expressions of anger, rejection and other emotions and keep going back.
The 'Clarke' family are typical of those cases where there was much more frequent social care involvement and within our sample it is a 'gold-star case' for illustrating the nature of a holding relationship (Wengraff, 2001). Some 69 encounters took place between the family and professionals during the 12 months we observed it, 17 of which were home visits by social work and 41 by family support services-34 by an intensive family support service that was external to the social work office, and seven by a family support worker who was based with social workers. Fourteen of these encounters between the family and social care were observed and audio-recorded by the research, 10 of which were on home visits, one in court, two at meetings in the social work office and one case conference. We also observed one staff supervision at which the case was discussed, and interviewed the parents once. At the heart of the casework was the holding relationship between the social worker, here called 'Miriam', and the family.

Phase 1: in care
At the outset of the research the baby, 'Louis', had just been born and was removed at birth. His mother 'Samantha' and father 'Oliver' were both White British. The first phase of casework involved parenting assessments and decision making about the child's future. It was less than a year since Miriam took Samantha's older child into care due to drug and alcohol abuse and the serious domestic abuse she suffered which the child was traumatised by. At that time Miriam said she 'disliked Samantha intensely': 'She was vile when I met her, aggressive, vile, wouldn't let us in the house, the most difficult person we had to work with.' The outlook for the baby Louis being returned to the family was regarded as hopeless.
By month two Miriam told the social work team meeting that Samantha was cooperating with the parenting assessment, seemed motivated to change and that she was now contemplating that Samantha could have the baby back. This was the first sign of the holding relationship becoming established and beginning to take effect. Miriam was now feeling that by keeping the baby in care 'we are being so punitive' and challenged the view of some social work colleagues that Samantha is engaging in 'disguised compliance', meaning a perception that the changes she was showing were not genuine. The tensions were such for the social worker that it had 'been keeping me awake at night'. By month three Miriam was even more convinced that the baby should be allowed to go home and this was 'one of the most difficult parenting assessments I've ever done. Everyone's against me.' Miriam was then observed in supervision making the case for the baby to be returned home and her manager fully supported it. The next challenge was to get the court to support the social work plan. At a meeting in the office at this point in month five the social worker and parents discussed the plan for court and Miriam was observed offering lots of praise and reassurance. The court social worker rejected the proposal that the baby should go home. However the judge was happy for the reunification to go ahead: 'Samantha and Oliver are obviously delighted-they are both smiling so much and I can't help but smile as well' (field notes). Samantha told the social worker that 'no one has ever stood up for me in the way you have', and Miriam said in an interview after the court hearing that she sensed how the parents 'felt empowered being beside me in court.' Opposing the court worker's position in order to stand up for the relatively powerless family was a crucial act of 'ethical holding' by the social worker.

Phase 2: reunification month six
The baby was returned to his parents at the age of six months and the home then became the site where almost all of the social care practice went on over the second half of the year. An extensive visiting plan was now being implemented, with the intensive family support service going in daily and the social worker visiting most days. The first home visit that was observed in the research occurred the day after the baby had spent his second full night at home. The social worker forgot to take the researcher on the visit the day the baby spent his first night at home, because, she said, of the sheer difficulty of being able to think clearly due to how busy she was: 'it's been one of the hardest weeks the [team] has had for a long time, 17 new cases.' This typifies the weight of work that was competing for Miriam's attention and the scale of the challenge involved in being able to see families regularly and think clearly about them.
On arrival at the home, father Oliver was there with the baby, Louis, Samantha was at therapy. The visit lasted 32 minutes during which Miriam interacted with the baby on a total of seven occasions. The social worker was open with Oliver about how she wanted to help him build his confidence as a carer, and in challenging him to be open and honest. The following scene conveys the essence of how the social worker practiced: Miriam touched the baby on the face three times while the father fed him in a high chair. He takes the baby out of the chair and onto his knee. The social worker touches the baby on the foot twice -she encourages the father to be more self-confident. He smiles and agrees he needs to bethe social worker gives her finger for the baby to hold and she says hello when the baby looks at her. The social worker suggests changing the baby and she takes the changing mat and puts it on the floor. The father lifts the baby off the mat and kisses him. The social worker is now sitting on the settee and asks for "a quick cuddle". Father hands Louis to her and the social worker stands the baby on her knee and holds one hand around his knee and bottom and the other under his arm -she is jogging Louis as he stands on her knee facing her. This typifies how from the outset the worker immersed herself in the family relationships, helped the parents to develop their parenting skills and confidence and how she also built a meaningful relationship directly with the baby. She explained in the car afterwards and on many subsequent occasions how for her touch was an important sense to use in relational work with babies and in a child protection role, in checking children's wellbeing and safety. By getting physically as well as emotionally close to the child such 'intimate child protection practice' (Ferguson, 2011) was central to the creation of a holding relationship.
The combined efforts of the intensive family support service and social work was vital during the early reunification period. They were observed doing joint home visits twice and the dynamic of these encounters is typified by a scene in month 6 in which Miriam and the intensive family support worker Susan are both present, three weeks after the child went home.
Samantha is sat on the floor and talking about her expectations of group work, Miriam affirms how painful it can be to dig deep. The doorbell rings, Samantha answers it, it's Susan and as soon as she comes in the positive energy increases further. Samantha admires Susan's appearance and sits on the floor again and is showing Miriam and Susan the book on her therapy. She speaks knowledgeably, including about "disassociation". Susan tells her "I feel really proud of you and I am rooting for you". Samantha is so open and genuine, there are real relationships here. Miriam says, "I'm really chuffed and proud with how the 3 weeks of reunification have gone". . . . Susan turns to me [the researcher] and is telling me that she has known Samantha for a long time and says out loud how wonderful it is to see how much she has changed and tells me about how angry Samantha has been. Samantha agrees and says in the past she hid things. Both Miriam and Susan are still sat on the settee leaning forward, a body position that signals their interest and positivity. Miriam stands up and coos at the baby. Susan gets up from the settee and stands beside Samantha and touches her on her arm, tenderly.
We see here the reach and power of the holding environment coming alive: the mother speaks of the therapeutic effects of the group psychotherapy, while experiencing the strengths based, holding work of the social work and family support service. This scene also shows how good Samantha herself now was at co-creating a holding environment.
By month seven the social worker regarded Samantha as seeming 'much more relaxed now.' One effect of the development of trust and their close relationship was that from early on Samantha phoned Miriam nearly every day. But however much the relationship was based upon care, worries about power and control were ever present for Samantha. In month seven she expressed fear to Miriam that while social care are involved 'then he's only half mine. . . . If I was having a bad day, I'd worry that you'd come and take him away from me.'

Going deeper
By month eight the intensive family support was complete and the social worker and social care based family support worker settled into a routine of fortnightly visits. The 10 home visits that were observed in the six months after the child returned home lasted an average of 45 minutes (the longest was 60 minutes, the shortest 22). On all of these the child was awake and interacted with by the worker. The interactions mostly took place in the same space-the sitting-room-and always on the same settee, although the placing of it in the room changed half way through the year. Most of the time was spent seated on the settee, although Miriam always had periods on her feet and sometimes on the floor with the child. The social worker's mobility and willingness to move, in order to play, touch, and physically hold were central to her relational practice and effectiveness (Ferguson, 2018). On all the nine visits she was observed on, the social worker was tactile with and held the child, typifying how the holding relationship was built on physical and emotional closeness and certain actions being repeated time and time again.
The baby being settled at home opened the way to a new phase where attention to issues such as the couples' relationship and Samantha's past and identity as a mother of a child in care deepened. These topics and attention to the baby's health, eating and sleeping patterns would be gone over again and again on visit after visit. Mini crises were dealt with, one week it was a Facebook argument, on another visit in month nine Samantha asked Oliver to take the child out of the room so she 'can have a rant'. She showed Miriam a letter about benefits payments, explaining they are 'financially stretched' and she is stressed. Social work managers had the discretion to enable social workers to provide small payments when families were in dire need and at times, like following this visit, Miriam did this.
The mood of visits fluctuated according to how the parents were doing and how social care worked with them. On a visit later in month nine Samantha was 'very happy and playful throughout. She expresses her happiness and that she is "doing OK"' (field notes). The visit, which was by the family support worker who was part of the social work team, was a mixture of investigative work-the bedrooms and kitchen were checked-talking about support groups and the drugs service and friendly casual talk about the baby and both parents' joy in him. Field notes recorded how the worker's 'communication is attuned, they respond thoughtfully, with open-ended questions.' In month 10, Samantha was upset by the family support worker making an unannounced visit, fearing this meant an increase in concern about Louis's welfare. Miriam calmly reassured her that it did not represent any change in how she is regarded by social care and reminded the parents that unannounced visits were part of the visiting plan so as to be able to see the family as they naturally are and check for signs of drug use, for instance.
On one of the visits in month 12 the mood was celebratory as Louis had just had his first birthday. He was now crawling at speed-including over to the researcher and climbing up them! On a further visit in month 12 Samantha was feeling sad and again wanted to talk in depth about her feelings about her other child who was still in care. She said she and the child could not understand why they cannot live with her. Miriam reminded Samantha that there was an agreement that the return of her child would not be considered until Louis was well settled with her and it was clear that she was able to care for him and that having two children to care for would not put that at risk. Miriam reassured her: No one thinks you have given up on the child, no one thinks that, we know that you are there for them 100%. I think it is just very tragic and really sad that that was the situation that we were in and now you know we are in this situation, and you need to hold on to the fact that you have got Louis with you and you do have contact with [the child in care]. They are really resilient and they are doing really well in school.
After a pause Samantha says she understands that it is still too soon but she does want her child home eventually. She reflects on how: last year before [child's name] was taken it was heart-breaking to be fair, it wasn't fun for me or for them. . . . Sometimes when you think you are protecting your child the most you are not, you are shutting them away from the reality of what you're feeling which can potentially worry them more . . . You couldn't imagine me saying this a year and a half ago could you?
The worker is empathetic, including recognising the responsibility of the men who abused Samantha, while not protecting her from the reality of the harm to her child. This is sensitive, skilled work. The worker's challenge is not to invalidate past problems and this mother's part in them while holding her as a good enough parent in the present. She was able to hold the difficult painful situation of the domestic abuse and removal of her child by tolerating it until Samantha found a way through it and was becoming able to tolerate it herself-as shown by her comment, 'You couldn't imagine me saying this a year and a half ago could you?'. How the social worker holds Samantha and her pain from the past and guilt about her older child helps her to tolerate those feelings and this enables her to create the mental space to care for her baby, and to do it well.

Parental perspectives
Samantha and Oliver were interviewed in month eight. Oliver said Miriam asked for him to be involved, but he would have liked this to have happened sooner when Louis was first removed. He has gained a better understanding of Miriam and now sees her as supportive. He felt the family support workers and Miriam have helped him to see what being a father involves, including the need to play, which has promoted his bond with his son.
[Louis] means everything to me, everything to me . . . That is just really why I have always just wanted to be, really just be a father and just do everything that my Dad didn't really do for me.
Samantha recounted a long history of abuse as a child, periods in care and involvement with social work. She expressed many criticisms of social work, how vulnerable she has been and described herself as having been difficult to work with: 'I didn't have any manners. . . . I just needed help-I don't like their systems and I don't trust them'. When Louis was removed at birth social workers told her she probably wouldn't get him back. When he was returned to her she found the involvement of so many professionals and daily visitors while coping with a new baby 'overwhelming, stressful'. To prove herself she would clean the house more. She feels she has changed-'the social workers showed me how bad my relationships were.' This has helped her to become a better person, and she has changed so she can fight for her other child in care. Samantha felt her relationship with Miriam had changed, improved, and how 'I haven't got to be so panicky'. I didn't feel obliged to have any manners or respect for them [social workers] to be fair. I am just thankful now that they have finally seen that I am a good mum and it is not the fact that I can't care for my children, it is more the fact of the situations that I have been in. They use a lot of my history but then instead of looking into why my history is the way it is, and trying to help me do that, you take away the only bit of love and unconditional kind of affection that I have, and that is my kids, so I have always been a bit yes . . . I have always kind of held a grudge against the Local Authority because of that to be fair. I couldn't learn to love my baby, every time he kicked I wanted to enjoy it but then I was too scared to attach myself to a baby that they are telling me they are going to take away as soon as he is born and then I would have been criticised if I ended up not bonding with my child. I am still cautious, of course I am. I am still worried about "right you come in my house you will say something, how do I know you're not going to go back to the office and write something totally different". But I think being able to communicate with her [the social worker] is a little bit more easier. Since [month four] my opinion of Miriam has changed. She's not the bitch I thought she was, [the relationship now] is a lot better, not so tense, in a way a little bit more friendly as well, it is not so formal. It is more casual when she comes round now and she is a lot more open with me about herself, I didn't even know she had children -she has told me about how one of her children struggled so that was quite reassuring.
This captures with agonising clarity the deeply conflicted experiences of a mother carrying a baby she knows will be removed from her at birth and how she dares to hope that it might be different, that by showing she is bonding with her unborn baby her capacity to care and love may be recognised by social workers and viewed in a positive light. Because of her long troubled history with social work, her mistrust in them ran deep. But Samantha had recently changed her opinion of Miriam, who is 'not the bitch I thought she was' and earned her trust. This occurred at exactly the point at which the social worker was observed changing her mind about Samantha and recognising her potential to care and love-that is, at precisely that point when the holding relationship became established.

Discussion
The research findings confirm that the quality of relationships is fundamental to effective social work and change. However, simply to refer broadly to relationship-based-practice is not sufficient. It is the kind of relationship involved that matters and this paper has drawn out of the research data an important strand of practice we are calling a 'holding relationship'. Our aims in the paper have been to document the nature of a holding relationship and to develop it as a concept. Having shown in detail the nature of such a relationship as it unfolded in casework practice over the course of a year, we will now draw out further how the findings show that a helpful, therapeutic relationship in social work is a holding relationship that incorporates several key components: reliability, immersion in service users' day-to-day existence and helping them develop life-skills, intimate practice and getting physically and emotionally close, and ethical holding in practicing critically by taking account of power and inequalities. We will now show further how these characteristics of practice were achieved across our sample and argue that it is possible for holding relationships to go on across all aspects of social work and for the concept to have generic value in social work education and practice.
All the casework that was practiced through holding relationships involved regular and punctual contact and workers maintaining consistent attitudes towards service users. What we have called 'ethical holding' was a key dimension in how the practitioners used critical reflection and brought an awareness of power and inequalities into their work. They saw service users holistically and in terms of their capacity to change and develop, trying hard not to reinforce the stigma of poverty, mental health problems, and having had children removed by the state. A striking example of this was how the social worker approached Samantha Clarke with an open mind and fought hard against colleagues who felt she shouldn't keep the baby because she was using 'disguised compliance'. As Miriam reflected back at the end of the year: I listened to her, she's never felt listened to. I invested a lot of time into it, reminded myself of the cycle of change, everyone was saying it's disguised compliance and I couldn't see that. . . . and then I was listening to her and I thought, this is not someone who is showing disguised compliance to me. This is someone who is actually, truly learning from the sessions they're going to and applying it.
This supports  argument that the notion of 'disguised compliance' is too often used in simplistic ways that prevent practitioners from learning more from families and believing them. In our study trust developed through the families feeling listened to, knowing that practitioners adjusted their approach in ways that respected their needs-like the social worker in the first case study we used above always showing Maria Nicholls what she had written down. The workers accepted families' criticisms of past workers and systems and they advocated on parents' behalf, in the manner Miriam did for the Clarkes in the courts, for instance. Social care did this while still placing clear boundaries around parental behaviour and limits on what they were prepared to accept, authoritatively checking home conditions and for such things as signs of drug or alcohol (mis)use. What social workers provide then is not simply 'care-cure' but an assemblage of 'care-control-cure', that represents the use of 'good authority' (Ferguson, 2011;D. Forrester et al., 2019).
This means that an important aspect of the complex position and identity of social workers involves them being 'reliable hate objects' (Ferguson et al., 2020a;Winnicott, 1949). By her own account, as we have seen, Samantha didn't 'have any manners or respect' for social workers. Part of her-and at times, no doubt, all of her-hated them. The reliability and ethical practice of workers was again crucial because it enabled such service users to come to know how they would find workers every time they had contact and that they would keep coming back even if the service users were angry, hurt, distressed and there was no fear of retaliation. Sometimes this did not just happen verbally but service users acted it out. In describing how 'vile' Samantha was when she first met her, Miriam elaborated: 'You just, you couldn't communicate with her. . . . You would have to walk away. Because she would be almost like an uncontained child, flailing arms and legs everywhere, you know, so really uncontrollable really.' But the social worker kept going back and a year on now regarded her as 'a likable person . . . charismatic, insightful, open.' Thus by month eight Samantha was able to tell the research that social care showed her how 'bad' her relationships were and that Miriam was 'not the bitch I thought she was'. The hate in terms of fear, suspicion, was still present, but so too to a far greater extent were respect, care, love. So the workers were 'reliable care objects' too. This holding is on one level an unconscious process in which service users know that the worker's mind is available to them and is open to their emotional experience. They feel held in mind by the key worker and the professional network even when the professionals are absent, just like a child does when cared for by a good enough parent/mother. The professional network provided individual acts of help and holding, but the role of social workers as case managers placed them at the centre of this larger holding environment. Forrester shows how for Winnicott case managing was the extension of holding: 'Psychotherapy, medicine and social work were like concentric circles of holding, centred on the fundamental metaphor of experience of a child being held by its mother' (J. Forrester, 2017, p. 101). A reliable holding environment and relationship created a lived experience of care and security which was pivotal to helping people for whom those experiences had been absent since childhood. It also worked by social workers forming an idea of the service user and their needs and them holding that idea in the relationship so that when they saw them the service user could find that part of themselves which they had given the practitioner. A compelling example of this was the way that Miriam held the shame, guilt, remorse, anger, the loss that Samantha Clarke regularly expressed about her child being in care.
Another crucial and easily missed dimension of the holding relationship is that while it has a psychological basis, it also has a literal meaning in the intimacy of encounters and the physical holding of children by professionals. This raises the peculiarly neglected question of what the use of touch means to parents and the children themselves (Green, 2017). The data suggests that the ways in which some social workers physically held infants was a vital way of having a holding relationship with young children but also seemed vicariously to help parents to feel held, symbolising the emotional holding and care that was being provided for the family. With older children such intimacy and emotional connection was achieved through play, encounters in the car, cafes, on computers and digital media. This is supported by Ruch et al.'s (2017) observational study of social worker's interactions with children which also found that practice needs to be understood in terms of how it incorporates the head, heart and hands.
Another distinctive feature of the holding relationship was how it arose from social care practitioners' deep immersion in children and families' everyday lives and an existential focus on the present. The findings show that social workers are not therapists in the psychodynamic sense of delving into the unconscious or in drawing straightforwardly on Rogerian person-centred counselling to practice unconditional positive regard (Murphy et al., 2013). The paradox is that while social workers are not therapists, their work can have direct therapeutic effects. They connect in very practical, 'hands on', ways with the day-to-day needs of families, helping them to learn the skills of caring, develop confidence and self-esteem and to address their struggles with poverty, housing, relationships and so on, most of it done in the family home. Walsh refers to this as 'being alongside' service users and 'valuing the seemingly mundane' (Walsh, 2018, p. 219). Families spoke of how they wanted nothing more than to have an ordinary, settled life together and the pleasure they got from accomplishing day-to-day things, like going for walks and to the shops.
'Being alongside' also included social workers self-disclosing things about their own lives-like when Samantha said that her social worker sharing something of her own experience of being a mother helped her to trust her. As another mother explained about her social worker: 'She's talked about her husband before and it makes her more relatable. She knows I like magazines and they are expensive so she brings me round her magazines after she's read them. It all makes a difference.' As well as kindness, using boundaries in such a flexible way by giving something of themselves, ensuring it was for the service user's benefit, not their own, was part of how workers showed empathy and were relatable. This fits with the 'genuineness' that Rogers (1961) defined as a key aspect of an effective working alliance between worker and client.
Ending holding relationships is complicated and requires great sensitivity. Samantha regularly phoned Miriam and in month 10 when Miriam missed her call one day and returned it the next she explained to Samantha that 'you're not at the top of my list of visits anymore.' Samantha responded with disappointment, which the social worker interpreted as her feeling rivalry with other families for her attention and need for affirmation of how well she was doing. In another 12 month case-study when closing it was first raised the mother told the social worker: 'I don't think you should leave me!' and listed all the stressful events the family was facing. This again shows the ambivalence that can surround relationships with social workers, how liked/loved and mistrusted/hated they can be at the same time. Even some parents who were critical of aspects of how they had been worked with and still feared their children would be removed, would yearn for the social worker's attention and to be held by them.
Finally, the knowledge, skills, creativity and human capacities that are used to create holding relationships do not happen in a vacuum. In our study, both of the sites where the research took place were significantly affected by the bureaucratic demands of record keeping, report writing, audit culture, austerity measures and budget cuts. Managers and practitioners were under a lot of pressure to prioritise cases, making it difficult to provide the same level of support and holding relationships to all the families who needed them. Social workers and family support workers still helped many families, preventing children from being removed and, when they were, enabling some of them to be returned home. At the social work office that worked with Samantha Clarke, social workers were co-located with their team managers, family support workers and administrators in small team offices. This generated a supportive practice culture, as the families' needs and the workers' experiences of working with them were regularly discussed, generating a culture of reflection and on-going dialogue about the work (Ferguson et al., 2020b). This was no guarantee of consensus among colleagues, as shown by the initial differences of opinion within the team about whether Samantha Clarke should be reunified with her baby. But these differences did not fracture relationships or escalate into conflicts because of the mutual respect and emotional and practical support that was shared between Miriam, her colleagues and managers. This illustrates how vital it is for organisations and managers to create the conditions that enable practitioners to be reliable, by ensuring they have the time, the emotional support and the space to think clearly and critically about their work. This enables workers themselves to feel held and be consistently available to provide service users with the holding relationships that can help them change their lives.

Conclusion
The paper has outlined and analysed the research findings concerning aspects of how social worker's begin, develop and sustain relationships with service users by introducing the concept of a holding relationship. These are relationships that involve reliability, critical thinking and good authority, getting alongside service users, developing their lifeskills and connecting with them in intimate, emotionally meaningful and containing ways. Social work develops relationships that help people by connecting with their lives on two levels: the external-their quality of life, poverty, housing, practical child care skills and so on; and the internal-their 'self', caring capacities, sense of self-efficacy and emotional, unconscious relational experience. This is not a new insight, given that it was at the heart of the psycho-social approach of the Winnicotts and others and of social work going back to the 1950s (Stevenson, 2013). What is original about the contribution of this paper is how it updates those theories for 21 st century social work, based upon detailed ethnographic data and analysis of how long-term practice is carried out. This has provided original insights upon which the concept of a holding relationship has been built which, it is hoped, can be useful in social work education and in developing helpful relationship-based practice. culture in child protection both in the UK and in Belgium using a variety of different methods from ethnography to visual methodology.
Tarsem Singh Cooner is Senior Lecturer in Social Work, in the Department of Social Work and Social Care, University of Birmingham, UK. His specialism includes combining digital technologies with professional education, learning design and research. He is an Apple accredited film editor, app developer and registered social worker.
Liz Beddoe is Professor of Social Work in the School of Counselling, Human Services and Social Work, , University of Auckland, NZ. Her research interests are in critical perspectives on social work education, professional supervision in health and social care, the sociology of professions, the professionalization project of social work, and media framing of social issues and professions.