Introduction

Motherhood and infancy were once considered normal phases of human life. As medicine expanded and specialized, however, motherhood and infancy were increasingly reframed as pathological. (Eyer 1992:129)

The perceived importance of addressing mental health issues as early as possible became evident in the first weeks of ethnographic fieldwork in ‘Oliveto,’ an Australian inpatient psychiatric facility for mothers with postnatal mental illness and their infants. While Oliveto admissions were primarily facilitated through a mother’s diagnosis of mental illness, it was the relationship between the mother and child that determined the progress for the dyad’s discharge. Staff approached patient admissions as an opportunity to treat and stabilise the mother’s mental health condition, but also to prevent any possible future damage to the child’s social and emotional wellbeing. This thinking was based primarily on the assumptions produced through John Bowlby’s attachment theory (1952, 1958), a long standing psychological construct, promoting the idea that the relationship between a child and its primary caregiver (always the mother in Oliveto) set the template for the child’s relationships for their entire life. As in Eyer’s quotation above, motherhood and infancy in Western cultural contexts came to be seen through attachment theorising as the potential site for pathology where relationships did not conform to cultural assumptions of normalcy.

This paper explores how attachment theory was understood and applied in this institutional and therapeutic setting. The paper begins with a summary of the key principles of attachment theory, as well as critiques of the assumptions that underpin it, particularly from a cross-cultural perspective. The methodology and context of this study is then described. The paper moves on to demonstrate how attachment theory was understood and applied in Oliveto, particularly through the use of the Circle of Security concept. Staff of Oliveto used attachment theory to assess the quality of mother–infant relationships. Attachment issues were understood to place infants at risk, and mother–infant therapy was used to mitigate this risk—a therapeutic process which we explore in depth here. In particular, we analyse the practice of mother–infant therapy through extensive case analysis of Helen and Javier, a dyad admitted to treat Helen’s mental illness. Through analysis we argue that attachment theory in this context has a gendered and dyadic focus on mother–infant relationships, to the exclusion of other relationships. Staff apply attachment theory uncritically, with little concern for socio-economic, historical or cultural specificities, in order to determine therapeutic options, to correct a woman’s behaviour, and then to assess whether the mother showed appropriate improvement and ‘insight’ in order to allow her discharge from the institution with her infant.

Attachment Theory, Uses and Critiques

Attachment (Bowlby 1958) is a psychological construct first theorised by British psychiatrist and psychoanalyst John Bowlby to define ‘lasting psychological connectedness between human beings’ (Bowlby 1971:194). His theory of attachment arose out of studies of the impact of a lack of maternal care published in the 1930s and 1940s (for example, see Goldfarb 1943, 1945; Levy 1947; Rheingold 1943) suggesting that infants in institutional care almost always had inferior development (Eyer 1992:48). Bowlby was the first theorist to draw these studies from across Europe and the USA together to form a coherent argument about the negative effects that maternal deprivation could have on infant mental health. Moreover, he believed that maternal deprivation could occur in seemingly ‘normal’ home environments, and that ‘[d]eprived children, whether in their own homes or out of them, are a source of social infection as real or serious as are the carriers of diphtheria or typhoid’ (Bowlby 1952:157). Bowlby began positing the idea that the relationship between a mother and infant could be pathological, suggesting that a child without maternal love could develop an illness synonymous with a disease, treatable by medical (psychological) intervention. Consistent with the ideals of the social hygiene movement, mothers were placed in a position of responsibility for their child’s, and by extension, societal wellbeing. Although Bowlby was clear in his writing that mothers were not the only important influence in a child’s life, the term ‘maternal deprivation’ suggests the role of the mother as paramount for ensuring a healthy psychological future. The role of fathers was given very little recognition in his theorising, however, nannies and permanent surrogate carers were mentioned.

Bowlby understood attachment and the primacy of the mother to be universal. He contended that an individual’s first significant life relationship formed a template whereby all other love relationships could be understood. Inherent in this thinking was the concept of transgenerational transmission of these relationship formations. A child’s relationship to its mother (or the person who mothers him or her) was fundamental to the development of a healthy ‘internal working model’ (Bowlby 1971). This model was the basis by which the child learns to understand the world and guides them in their emotional and social behaviour in the future. Bowlby believed that an individual’s effectiveness when interacting with others, their ability to see others as trustworthy, and their ability to see themselves as valuable, all rested upon the dynamics that occurred in their first significant relationship.

While it was Bowlby who first brought attention to the innate cues that infants use to attract proximity to their caregiver, Ainsworth’s work was responsible for classifying the ‘quality’ of the mother–infant attachment. Ainsworth believed that it was a mother’s responsiveness that determined the type of attachment pattern of the relationship. According to Ainsworth and her colleagues: ‘securely attached’ infants rarely cried and were happy exploring when their mothers were present; babies who cried frequently, even when being held by their mother and did little exploration were categorised as ‘insecurely attached’; and infants who displayed no differential behaviour to their mothers had ‘no’ or ‘poor attachment’ (Ainsworth, Blehar and Waters 1978).

Ainsworth’s work not only focused on attachment, but also on exploratory behaviours of infants. She introduced the concept of the secure base to attachment theory, yet also carefully pointed out that strong attachment promoted relative independence rather than precluding it: security and autonomy developed at the same time (Ainsworth 1967; Ainsworth and Wittig 1969).

Bowlby’s initial theorising in the 1940s has been, and continues to be, expanded on and developed in numerous ways. Ainsworth’s work in Uganda (1967) was understood as the test for its universal applicability, though this has been critiqued. Attachment theory more recently has not only been applied to mother (parent or caregiver)—infant interactions, but is also used in therapeutic practice for the treatment of disrupted relationships throughout the human lifecycle. Attachment theory is used extensively in contemporary psychology, for example, to address bereavement and loss (Bowlby and Parkes 1970; Stroebe, Schut and Stroebe 2005), couple’s therapy (Davila and Kashy 2009; Johnson and Whiffen 2003), family therapy (Cassidy and Shaver 2008:xv; Liddle 1999; Liddle and Schwartz 2002), and depression in adolescence (Diamond and Lebow 2005). The extensive application of attachment theory demonstrates that it has not disappeared or become outdated. The theory has evolved, from its early beginnings related to ethology (the study of the biology of behaviour), to the latest developments in neurobiology (see for example, Landers and Sullivan 2012; Sullivan 2012).

The conceptualisation of attachment behaviours as universal has been challenged by anthropologists and others who ethnographically observe childrearing and attachment models that do not correspond with the theory (Quinn and Mageo 2013:4–5). As such, anthropological critiques argue that assessing early infant–caregiver relationships through attachment theory is ethnocentric and therefore inadequate for capturing all cross-cultural variations in childrearing (Quinn and Mageo 2013). This is particularly so in relation to the measurement of attachment relationships through Ainsworth’s Strange Situation Procedure, which due to its prolific use over time has become synonymous with the theory (Bretherton 1993:25; see also Quinn and Mageo 2013:4). The Strange Situation Procedure focuses primarily on reunion behaviours and not on the natural context where interactions take place (LeVine and Miller 1990:76–77). As such, attachment relationships are ‘decontextualised’ and fail to take into account important information in the cultural structuring of childrearing practices (Quinn and Mageo 2013:17).

While traditional attachment assessment takes the view that attachment security is dyadic, cross-cultural evidence disputes the centrality of one secure base for young children; rather, children develop multiple bases as sources of security (Meehan and Hawks 2013), and attachment security can vary across their network (Meehan and Hawks 2014). Anthropological writing critically questions the ‘concept of the exclusive biological mother that forms the basis of much Western thinking and psychological theorizing’ (Barlow and Chapin 2010:326). Ethnographic research has clearly demonstrated that child-caregiver patterns can be communal, cooperative and diffuse. Infants and young children have been shown to exhibit a range of secure base relationships that respond to: cultural values (Chapin 2010, 2013; Gaskins 2013; Gottlieb 2004; Seymour 2004, 2013; Tronik, Morelli and Winn 1987); ecological conditions (Barlow 2013; Crittenden and Marlowe 2013; Meehan and Hawks 2013; Scheper-Hughes 1985, 1993; Tronik et al. 1987; Weisner 2005; Weisner and Gallimore 1977), and maternal availability through competing workloads (LeVine and Miller 1990; Quinn and Mageo 2013; Rottger-Rossler 2014; Seymour 2013). Evolutionary anthropologist Hrdy (2009:84–85) points out that the exclusive dyadic mother–infant relationship viewed as the hallmark of attachment theory is not a universal phenomenon. The prevalence of a wide range of secure base caregiver arrangements in the cross-cultural literature prompted Gottlieb (2004:137) to refer to sole caregiving as a ‘statistical anomaly’.

Anthropological studies of attachment have tended to focus cross-culturally and on developing countries, with very little ethnographic attention paid to attachment in countries like Australia, Europe or the USA. LeVine and Norman (2001) have provided a reanalysis of findings on attachment in Germany to clearly question the normative assumptions that secure attachment leads to mental health, while insecure or avoidant attachment leads to pathological outcomes. Moreover, LeVine (2014) links attachment theory to the culture and values of middle to late twentieth century middle class Anglo-America, arguing it cannot be seen as having universal reach. Yet there needs to be more ethnographic focus and research on Western countries. There is also a dearth of ethnographic studies of attachment in institutional or therapeutic settings, despite the common application of this theory to assess relational patterns. This paper aims to contribute a critical ethnographic perspective to the literature and critique attachment theory through a discussion of how attachment principles shape therapeutic practices in an Australian institutional setting.

Methodology and Context

This paper is based on 12 months’ ethnographic fieldwork carried out by the first author in 2008–2009 in Oliveto, an acute public psychiatric unit, admitting up to six women and their babies at one time.Footnote 1 For the most part, women were referred and admitted to Oliveto with a formal diagnosis from their General Practitioner of a mood disorder with postpartum onset, commonly referred to as postnatal depression (PND). Women were also admitted with major depression and anxiety. Many of these women were diagnosed after admission with borderline personality disorder (BPD) in conjunction with their mood disorder. There were also a small number of women diagnosed with psychotic disorders—diagnoses that most of them had received many years earlier but since the birth of their child their illness behaviors had raised concerns about their ability to parent their child(ren) in accordance with acceptable cultural norms.

On rare occasions,Footnote 2 fathers also stayed with the dyad throughout the treatment process. Fathers with a psychiatric diagnosis were never admitted into this mental health facility: the relationship between the mother and her child was given primacy. What set their psychiatric treatment apart from other women diagnosed with a mental illness was that these women were mothers to young children (birth to 3 years), and the woman’s mental illness was viewed in the context of her mother–infant relationship.

The participant observation conducted in Oliveto largely consisted of two parts; participant observation in the public areas of the unit, involving women/mothers who were patients in the unit, their babies, family, friends and staff (predominantly nursing); and observation in the areas of the unit where members of Oliveto’s multidisciplinary team interacted to discuss patients in formal and informal contexts. Approximately 70% of time in the field was spent in participant observation with patients and nursing staff in the public area of the psychiatric space and both informal and formal interviews with patients. The remaining ethnographic data was collected in areas of the fieldsite that were reserved for therapeutic staff interactions and included observations of staff meetings, formal and informal interviews with staff, and review of written materials and documents.

Ethics approval was granted by the relevant bodies, and all participants provided informed consent. Twenty-six of Oliveto’s staff members agreed to be involved in the research, which was almost all of the staff. Most staff were women, with only 3 men on staff. All treating staff were university/college graduates and by virtue of this, of middle socio-economic backgrounds. Staff were from a variety of health professions including psychiatry, psychiatric and general nursing, and allied health professions.

One hundred and four women were admitted into Oliveto during the period of ethnographic research. Most women admitted were from low to middle socio-economic backgrounds, with the majority from white Anglo-Saxon or Western European backgrounds, born and raised in Australia. Despite Australia’s multicultural population, this was mostly not reflected in Oliveto’s patient population. Very rarely women were admitted who required interpreters, and none during the research period. The cultural context of Oliveto, as an institution in Australia mostly staffed by middle-class Anglo-Australians treating a similar group, largely replicated the context in which attachment theory was conceptualised.

Twenty four of the women admitted into Oliveto during the research period consented to participate as informants in the ethnographic research, including Helen and her 6-month-old son Javier.Footnote 3 Helen had been diagnosed with anxiety and depression, and referred to Oliveto by her General Practitioner when her symptoms continued to worsen despite medication. During discussions with Helen she described herself as ‘tired’ and ‘sad’, and had concerns about her son, particularly his lack of interest in food. She was receiving psychiatric treatment by Oliveto’s multidisciplinary team including assessment, medication and mother–infant therapy as focused on in this paper. Her case was chosen for this paper as it was one where access and data was most complete, including multiple observations and interviews with Helen, with the mother–infant therapist, and observations of interactions with her son on video. Some details of the case have been changed to protect anonymity.

Attachment in Oliveto

In Oliveto attachment was viewed as the embodiment of a mother’s love for her infant, and the foundation for all psychological, emotional and cognitive progress—referred to in psychological terms as ‘development’ (Piaget 1964). The centrality of this concept was demonstrated through the ubiquitous appearance and use of a diagram entitled Circle of Security: Parent Attending to the Child’s Needs (Cooper, Hoffman and Powell 1998). The diagram made explicit that the child’s needs were significant in this context: prior to beginning fieldwork it had been expected that the treatment focus in Oliveto would be centred on the mother’s mental illness with children present in the unit but not part of the therapeutic process. However, the overt display of the Circle of Security (COS)Footnote 4 diagram around the unit suggested that the needs of children were central in this context.

The significance of this diagram lay predominantly in that the COS described in pictorial form the basic principles of attachment theory. As a therapeutic tool, the COS was designed to be used by professionals working in a therapeutic capacity, to assist with explaining the core elements of attachment theory in relation to clients or patients. The diagram shows two hands on the left hand side of a circle where the top part is depicting the child’s exploration from a ‘safe base’, and the bottom part shows the child’s return to the parent, as a ‘safe haven’. The hands shown in the diagram refer to a child’s primary caregiver (assumed to be the mother). In a healthy attachment relationship, mothers were viewed as a ‘secure base’ from which to explore the world. A child who has a secure attachment to its mother should be happy to leave their secure base and explore the world around them. In response to this, a mother who is understood to display a healthy attachment with her child will watch over her child by keeping them in their eye’s view (or for an infant, maintaining eye contact), keep them from physical danger, show delight in their exploration through facial expressions and responsiveness, help their child by assisting them (only) when they are not able to do something on their own, and show enjoyment in them by joining in their play. The bottom half of the circle referred to the caregiver’s ability to be a ‘safe haven’ for their child—a place to return to when they have finished exploring the world, or are needing comfort or protection. It is clear that the COS encompasses both attachment and exploratory aspects of parent–child relations as conceptualised by Ainsworth.

According to the COS, the concept of a ‘safe haven’ encompassed the idea of protecting and comforting the child when they were hurt or upset, and being able to help their child manage their emotional distress. This is done through naming the emotion for the child, or staying with them while they are experiencing an emotional state. According to the COS roadmap, the role of the mother is to be ‘bigger, stronger, wiser and kind’, with the intention of following the child’s lead as much as possible, but taking charge when necessary. One of Oliveto’s staff members described the secure base concept as follows:

I’m your secure base and I care for you and I love you. Then you go off into the world and then when that’s a bit much or something goes wrong I, you know, help you sort it out, or at least be with you when you are sad and reassure you.

Another staff member referred to the COS as being synonymous with the movement of an elastic band. Just as an elastic band can stretch and allow for movement, it will always return to its original form. Similarly, a child displaying a healthy attachment will always return to their ‘secure base’ or look to their mother as a ‘safe haven’ when feeling unsafe or unsure. No matter how far the elastic band is stretched, it will, by its very nature, always return to its original form and remain intact as a circle.

The COS highlighted the commonly held view in biomedical cultural practices that the child’s physical and emotional needs were a domain to be managed by one primary caregiver—generally the biological mother. It became apparent early on that the role of fathers was largely excluded or downplayed, as were any other caregivers or family members. The treatment focus was on the dyad of the mother and infant, not broader parent–child or family relationships.

The COS template is a diagram, or ‘roadmap’ as it was referred to by its creators, designed for use in a psychotherapeutic context. While the COS was indeed used for this purpose, observations in the field saw reference to the COS, and the language inspired by it, in numerous contexts both formal and informal. It was also used by a variety of Oliveto staff members from different disciplinary backgrounds including psychiatry, psychology, nursing and occupational therapy, in talking about both attachment and exploration. The COS roadmap was approached in Oliveto as a tool depicting universal principles of attachment, and it was applied to all patients regardless of socio-economic or cultural circumstances.

In addition to the COS, the categorisations of attachment styles identified by Ainsworth were of particular significance. Staff in Oliveto would assess attachment as secure, avoidant or ambivalent after observing mother–infant interactions, and these assessments formed part of the determination of appropriate treatments for the dyad. The Strange Situation Procedure was not observed to be used in this institutional context for assessment. It was clear, however, that secure attachment was viewed as the ideal relationship between mother and infant, and the desire to promote secure attachment between mother and child was, in fact, part of the rationale for the establishment of this mother–baby unit and its policy of dual admissions.

At-Risk Patients & Mother–Infant Therapy

In formalised staff meetings, routine assessments of a patient’s risk were conducted, evaluating the patient’s risk of harm to self or suicidality, risk of harm to others, particularly the baby (infanticide or abuse), and risk of absconding from the treatment service. In Oliveto, the risk to the child was of particular significance, as it was believed that the age of children admitted into Oliveto was a time when a child’s brain was at its most formative developmental growth. Attachment templates formed at this time were believed to have lasting impacts on the child’s future; therefore addressing the mother–infant relationship during inpatient admissions was seen as a preventative measure for the potential risk to the child. Constructing the child ‘at-risk’ in this context generally referred to the risk posed to the child through its relationship with its mother; the risk caused by the father–child relationship was a cursory concern in this context. A large part of Oliveto’s therapeutic work was therefore to assess this risk, and make decisions about ways to prevent this maternal risk from occurring.

One of the key therapeutic processes to manage risk was mother–infant therapy. Many patient informants described mother–infant therapy as a highly useful therapeutic process they underwent while admitted into Oliveto. At the time of the ethnographic data collection, Kelly, an allied health practitioner, managed most of the mother–infant therapy work. Kelly viewed the role of mothering as an occupation that sometimes needed therapeutic work. Her expertise was in the area of mother and infant relationships, with a specific focus on the attachment relationship between the dyad.

Mother–infant therapy was chosen as a therapeutic option for women whom the multidisciplinary staff identified as having problems in the attachment relationship. The purpose-built therapeutic room had a video camera installed on the wall which was a significant feature of therapy. All sessions were video recorded by staff and played back to the mother in a later therapeutic session. The mother and her child, and sometimes other family members (fathers, grandparents, and other siblings)Footnote 5 would be asked by the therapist to enter the room and interact with the child in the way that they would normally interact, all the while knowing that their interactions were being recorded. While this took place, the therapist, and quite often a student (medical or allied) or other staff member, sat in a small room next to it watching the interactions on a screen. Once satisfied that enough of the ‘essence’ of the relationship had been captured on video, the recording was shown back to the mother and used a reference for psychological analysis. The frequency of these sessions was case dependent.

In Helen and Javier’s case, a series of mother–infant therapy sessions were recorded and used to facilitate ‘reflective functioning.’Footnote 6 This was a term used by the mother–infant therapist, to refer to the individual’s capacity to experience ‘one’s own and others’ behaviour in terms of mental state constructs’ (Fonagy and Target 1997:680). Ethics restrictions limited the researchers’ capacity to observe mother–infant therapy sessions first hand, however with Helen’s consent, the first author viewed the videos created during her mother–infant sessions. Then the first author interviewed both mother–infant therapist Kelly, and Helen, to explore their individual perspectives about how they understood those sessions, and the issues facing Helen and Javier.Footnote 7

Helen’s Perspective

Helen described her mother–infant sessions with Kelly as ‘like a conversation’ where she would meet with Kelly and ‘just talk.’ She described how every now and then the therapist would stop her and say ‘see; look at what you said then. See what you said to Javier!’ She would ‘catch’ her on things that she had not noticed herself doing. In Helen’s words, ‘because when you’re not well you’re not as logical, you can’t seem to process your thoughts through. You feel like you’re in zombie land!’ Helen’s perception of the mother–infant therapy sessions implied a relaxed tone, with a therapist with whom she felt comfortable talking. Yet these sessions were more than ‘just talking.’ Through observation and clinical logic, the therapist was gathering information that would be translated, using attachment theory as a guide for interpretation.

Helen explained that Kelly showed her how a few things kept ‘re-surfacing’ for her, referring to recurrent themes that Kelly had identified as areas worthy of therapeutic investigation in Helen’s recalling of her childhood. ‘She always tells me that I don’t give enough time to grieve or something.’ When the first author asked if she had had grief in her life, Helen dismissively responded ‘I guess so, in my family… If Kelly was here she would say “See you always just pull your socks up and move on!”’ Following this Helen was asked about the ‘losses’ in her family. She went on to explain that her entire family of origin lived overseas; her father had died when she was young and so she had mostly been raised by a single parent, and her mother was now ill.

She described learning of ‘family secrets’ as a young adult, which had been kept from her when she was growing up. While she had experienced challenges regarding her family life, which had implications for her identity as a family member, she considered immigration to Australia to present her with a ‘second chance’. She did not construct her childhood experiences in terms of grief or loss. She felt no need to analyse or return to her earlier experiences in life.

Searching for Helen’s ‘Nursery Ghosts’

After watching Helen’s mother–infant therapy videos, Kelly was interviewed about her perspective on what was happening in the therapeutic sessions. In the discussion, Kelly began by describing the purpose of mother–infant therapy sessions and how they could be useful for someone like Helen who was ‘struggling in her relationship’ with her son Javier. Kelly introduced the concept of ‘ghosts in the nursery’, a term used to describe the psychoanalytic exploration of a mother’s childhood traumas to explain difficulties in her present mother–infant relationship (Fraiberg, Adelson and Shapiro 1975). The concept of ‘ghosts in the nursery’ adhered to a model of identifying psychopathology in the transgenerational transmission of attachment relationships. In Kelly’s description of ‘ghosts in the nursery’ she stated: ‘the reality of one’s childhood may be present without them being aware that they are present’. Using this model therefore meant exploring the mother’s past—her childhood—to understand her presenting difficulties with her child(ren).

As a mother–infant therapist, Kelly saw her role as encouraging Helen to ‘think’, ‘be reflexive’ and ‘become more awake’ to the ghosts from past relationships by helping her identify things from her upbringing that could be present in her relationship with Javier. She saw Helen as being ‘damaged’ by her past and this damage was disrupting her relationship with Javier and causing him to have ‘unusual’ behaviours that were not in line with biomedical and cultural expectations for an infant of his age.

Attachment and Child Development

Through analysis, it became apparent that mother–infant therapy was a process of exploration of the mother’s early attachment experiences which were then used to identify ways in which that experience impacted the mother–infant relationship in question. In many cases this was understood in terms of interruptions to the child’s development. Attachment was considered the most influential factor for a child’s future development, the first of a list of developmental achievements thought to measure physical, cognitive, social, emotional and communicative growth. These stages, commonly referred to as ‘milestones’, are argued to show universal similarities across cultures, and have been used for assessing a child’s behavioural changes according to their age (see Porter et al. 2002). However, as Super and Harkness assert, these scripts have evolved from the assumption of a ‘universal decontextualised child’ (1982:545). Anthropologists have argued that these milestones are judged in accordance with Western assumptions of childrearing, and disregard the cultural practices and beliefs about child rearing in other cultures (DeLoache and Gottlieb 2000; Gottlieb 2004; Whiting and Edwards 1988).

In Oliveto, developmental milestones were a guide for assessing a child’s developmental growth, and were used to question the quality of the mother–infant relationship. In Helen and Javier’s first video session, Kelly held a block up in front of Javier’s face; he looked straight past the block to Kelly’s smiling face behind it, cooing and babbling all the while. Kelly and Helen both recognised that he was much more interested in her face than he was in the object. A number of staff were concerned that Javier had developmental delays, and he was sent to another unit for a developmental review, which returned with the suggestion that the issue was relational and attachment based rather than idiopathic.

In a mother–infant session, Kelly explained to Helen that children are often more interested in faces than they are in objects. In the video, Helen exclaimed: ‘I thought it was just a good thing, that he was just like…a people person and then they start saying he’s not reaching for things or not doing the things that he should be doing, and I was thinking oh no, what’s going to happen?’ Javier’s behaviour had never been a concern for Helen until her admission into Oliveto. Helen had recognised that her son’s behaviour was different from her older daughter’s, but did not understand it to be problematic until it was brought to her attention in the unit. She had never really spent a lot of time or energy focused on encouraging her daughter to grab a toy and had not understood the importance of doing so with Javier. Helen had not had mental health issues when she had her daughter, and the daughter’s development had not raised any concerns. Staff provided no explanation of why Helen should suffer attachment issues with her second child but not her first. In Oliveto, the emphasis on Javier reaching the prescribed milestone of grabbing or reaching was given particular importance because of the belief that his lack of interest in doing so was indicative of an underlying concern arising within the mother–infant relationship.

For Helen, a far more important issue of concern was that Javier had begun to show a lack of interest in food. She had great difficulty in understanding why her son would not eat, and from her perspective, this was a source of great anxiety. This issue was also addressed in one of the video-recorded mother–infant therapy sessions when Helen was asked to feed her son in front of the therapist. In the recording, Helen persisted in trying to get food into her son’s mouth, but he kept turning his head and denying her advances. Reflecting back on this session, Helen explained how Kelly taught her that she was pushing her son too hard to try to eat the food, and that meal times had become ‘like a battle.’

The Ability to be Attuned and Read Cues

During an interview with Kelly about Javier’s interest in faces over objects, the matching of behaviours observed during mother–infant therapy sessions to attachment theorising was made explicit. Kelly described Javier as being ‘hyper-vigilant’ to seeing what expressions his mother’s face would show. Throughout his young life he had many ‘intrusive’ experiences where his mother (and others) had tried persistently to feed him, or more recently attract his attention to toys, however he was more concerned to see the face behind it and the intention behind the intrusion. Kelly described Helen as ‘overbright’ in many of her interactions with Javier and ‘not sensitively attuned.’ She also described Javier’s rejection of his mother and her intrusive behaviour as being ‘caught between fight or flight,’Footnote 8 he was not sure whether he needed to put up his defences so was unsure how to manage interactions with his mother. Kelly referred to Javier as a ‘stressed’ baby.

In the therapist’s description of Helen’s interactions with her son, Kelly described Helen as having difficulty in ‘supporting his exploration’ with food, and instead was insistent on him eating food when he was clearly rejecting it by turning his head. She learnt that she needed to relax a bit more about her insistence for Javier to eat and let him lead their interactions. When she gave in to this battle and just let him eat when he wanted to, things started to improve, she reported. According to the COS model, when she was able to support his exploration, her son stopped rejecting her. Working with this model, the therapist identified the attachment style as avoidant, and hence problematic. Then consistent with the transgenerational argument of attachment theory, it was important for Kelly to locate Helen’s ‘ghosts’ from childhood and discover the origin of her familial attachment template, and hence the origin of Javier’s difficulties.

The therapeutic interview then projected back through time to Helen’s childhood. From this, Kelly identified food as having a significant role in Helen’s life. The therapist acknowledged food as a sign of caregiving, yet food was used to highlight pathology in attachment behaviour. Kelly described food as Helen’s way of connecting with Javier (and others). As a young girl, she was left at home alone for the majority of time: this was framed by Kelly as childhood neglect. The therapist’s interpretation left no room for recognition for alternative cultural practices or socioeconomic necessities, despite the fact that Helen was raised in a non-Western culture and her mother needed to work as a single parent with no welfare system. It is clear here that Kelly is assuming the universal applicability of attachment theory. She assesses both Helen’s mothering as intrusive and leading to avoidant attachment, and how Helen herself was mothered as neglectful. Kelly did not concur with Helen’s own view that she was mothered effectively and with care, but differently; nor does it appear that she sought out information on normal mothering and childrearing in non-Western families. Variations in parenting style in Oliveto are seen as the basis of pathology rather than viable cultural alternatives.

Searching for Insight

For the therapeutic process to be successful, Helen was expected to show ‘insight’ or agree with the model offered to her as an explanation for her symptoms of mental illness. In the psychiatric literature, the concept of insight in relation to mental health has been described as being ‘composed of three distinct, overlapping dimensions, namely, the recognition that one has a mental illness, compliance with treatment, and the ability to relabel unusual mental events…as pathological’ (David 1990:798). Throughout the therapeutic process, Kelly searched for Helen’s insight numerous times, repeatedly reframing Helen’s childhood recollections as ‘neglectful’ resulting in her grief or sadness (a euphemism for depression).

Can you look at that [childhood experience] now and see that that was a bit of a strange childhood to have?

The assumption that Helen could be reflexive was a contradiction in Oliveto’s practices. By definition, her mental illness was constructed as a barrier to reflexivity—yet in order to progress in the therapeutic process, she was expected to understand and agree with the psychological reframing of her childhood experiences.

Kelly felt that she had to be careful in her role as a therapist not to try to change the way families operated, but instead tried to encourage her clients to look and reflect on their family history and the roadmap this had imprinted in their current relationship struggles. She believed that Helen was very resistant to this idea in the beginning. Kelly claimed she was not trying to place blame on the individual’s history, but was trying to ‘wake her up to the way history has a way of impacting in the here and now.’ Kelly claimed that Helen had a lot of difficulty coming to terms with this at first, but came back to her a few sessions later and admitted she was able to see its importance.

How do you make sense that someone has altruistically brought you in and then neglects you? That’s a very complicated psychological thing to hold on to, but she’d never even contemplated the complexity of that as an adult looking back. So of course that’s going to be rumbling around in her relationship.

Kelly spoke about how Helen could not ‘make sense of her mother’s actions,’ and so ‘constructed a narrative’ as a way of managing her confusion. Kelly felt that ‘she was unable to see that she was severely neglected as a child.’ This disjunction between Kelly and Helen’s understandings was also reflected in Helen’s own words: although she could intellectualise what Kelly was trying to demonstrate to her, she claimed that she did not see things that way.

HELEN: When I look back now, I just think it’s not her fault ‘cause she didn’t have parents of her own and she just did the best she could.

KELLY: Every mother does the best they can… sometimes that isn’t necessarily enough. Sounds like your mother was in grief, hey? And she didn’t have parents of her own to look out for her.

Again, the COS therapeutic model of attachment was applied. Kelly engaged with the concept of transgenerational transmission, to ask Helen how her ‘ghost in the nursery’ may have originated and then transferred through her mother’s own ‘nursery ghosts’. Her mother’s neglect of Helen as a child was a result of her own mother’s childhood neglect. The therapeutic process crosses temporal boundaries again as Helen was asked to empathise with her mother as a child. Through this, a link of transgenerational neglect was made over several generations, reinforcing the need for Helen to address her mental health problem in the present for her son’s (and potentially grandchildren’s) future psychological well-being.

As Helen and Kelly were concluding their therapeutic session, Helen looked to the therapist for acceptance that the session had gone according to therapeutic expectations.

HELEN: Well what do you think of today then?

KELLY: What do you think?

HELEN: I don’t know, I guess I’m being analysed.

The therapist responded by seeking Helen’s insight about the session. Interestingly, Helen’s response failed to show that she had gained further insight into the roots of her anxiety and depression; she was more concerned about Kelly’s interpretation of events through her ‘analysis’ of her. It could be argued that the power of psychiatric practice and the knowledges produced by it that highlight the risks of mental illness, were a factor of her submission to the therapeutic process, particularly given that this institutional context has the ability to recommend temporary or permanent removal of a child from a family where women do not show adequate compliance with treatment and insight into their condition.

Discussion and Conclusions

Helen and Javier’s case was chosen as an example to illustrate the core values that infused Oliveto’s everyday assumptions and functioning. It became very clear that the dyadic relationship between mother and child was the key focus of therapy: in the mother–infant therapeutic sessions, Helen was held responsible for Javier’s difficulties, while Javier’s father was not, despite comments from staff about his mental health difficulties. Any cultural or socio-economic issues were sidelined in the process of constructing a narrative about the neglect Helen suffered as a child, and the way this now shaped her relationship with her child. Framing Helen’s mother’s behaviour as neglect, could be interpreted as the lack of engagement by staff with the socio-economic or cultural contexts of patients; and the cultural blindness of attachment theory as applied in Oliveto. It was up to Helen, with the assistance of staff in Oliveto, to have appropriate ‘insight’ into her condition in order to break this transgenerational pattern. Helen was not in a position where she could insist on alternative readings of her own or her mother’s behaviour. In this case, Helen demonstrated the required insight, and was able to be discharged with her son. Yet this was not always the case, with some women having children temporarily or permanently removed throughout the research period. Hence while it was the woman’s mental illness that prompted admission, it was the assessment of the quality of the dyadic attachment relationship by Oliveto staff that shaped the timing and nature of discharge. In this particular psychiatric context, mothers needed to demonstrate both presence and attachment to children, yet also balance this with allowing their exploration rather than being intrusive—in other words, the outcomes for the dyad depended upon staff judgements of appropriate mother love.

Attachment theory as understood and applied in Oliveto through the COS and mother–infant therapy to assess and then therapeutically manage mother–infant interactions and relationships rested on a series of assumptions partially drawn from the theory but also from the cultural context of the institution. As noted by LeVine (2014:51) attachment theory recasts ‘moral judgements specific to middle-class Anglo-American culture of the middle to late twentieth century into a psychiatric model of early development in which the infant’s need for “mother-love” leads to psychopathology when not satisfied by particular maternal behaviors’. The cultural context of Oliveto largely replicates that where attachment theory was developed: a context where biological mothering is given prominence in child rearing, and mothers are frequently held responsible for outcomes for children.

Cross-cultural evidence suggests a variety of parenting styles occur around the world which are not pathological, and which produce or emphasise particular valued characteristics for the society in question. Staff in Oliveto seemed unaware of this research, nor how unusual exclusive biological mothering is on a world scale. Attachment theory, and in particular the value placed on secure and sensitive mothering as mitigating mental health risks is accepted as universal reality in this context. Hence they applied their assessments of mothering and risk, and worked to therapeutically reshape the behaviour of the mothers in their care in order to more closely mimic the secure and sensitive parenting morally foregrounded in attachment theory.