Developing a Theoretical Framework to Discuss Mothers Experiencing Domestic Violence and Being Subject to Interventions: A Cross-National Perspective

The discourse about domestic violence has developed in patriarchal societies, and so we position our understanding of ‘mother’ within a patriarchal framework. We explore the ways in which ‘mothering’ and ‘mother blame’ have been constructed within that framework and how this becomes relevant in the context of domestic violence and child welfare social work. We review literature from Australia, Canada, England and Wales, and the United States of America that has focused on child welfare responses to mothers experiencing domestic violence and abuse. On the basis of that review, we argue that mothers are responsibilised for violence and abuse they do not perpetrate. We show that the way legislation operates in some jurisdictions facilitates hegemonic, patriarchal constructions. We call for a review of current child welfare social work policy and practice in which domestic violence is present.


Introduction
In the area of domestic violence and abuse (DVA), applied research and practice have frequently focused on legal and social policy changes. There are also numerous studies that consider the incidence and outcomes of DVA (e.g., femicide), and others focus on the effect on children. The violent person, who is most frequently a man, is commonly absent from the academic-, policy-and practice-based discussions (Coy and Kelly 2019;Walby et al. 2017;Gracia 2014;World Health Organization 2013), and limited attention has been paid to the role of the abusive man within the discourse of child welfare (Bancroft and Silverman 2002;Douglas and Walsh 2010;Featherstone and Peckover 2007). With the focus 'kept away from men's violence and its' impacts on both women and children' (Lapierre 2008: 458), discussion has concentrated on 'women's actions' (Lapierre 2008: 456;see also Scourfield 2003). It is the behaviour of women as mothers that has become the frequent subject of study and intervention (see also Humphreys and Absler 2011).
We discuss the absence of violent men in the discourse of child welfare social work (hereafter referred to as 'child welfare'; note that in the United Kingdom [UK], this is referred to as 'child protection') by exploring the literature on mothers experiencing child welfare interventions in a DVA context. We also suggest scholarship in this area 'has been characterised by a general move away from an analysis that centralises issues of gender and power' (Lapierre 2008: 455).
In our review of research and practice within child welfare, we focus our scholarship on the central question of gender and power. We show how the ontological framework that underpins child welfare in a DVA situation contributes to a material effect on mothers. Currently, child welfare is practised in a way that is ontologically gendered, although this is unacknowledged. This means that the gendered practices that underpin the law, policy and practice of child welfare in a DVA context in countries across the world pass without widespread critical examination. Our paper seeks to develop theorising about child welfare policy and practice in a DVA context and argues for a review of social work policy and practice and the underpinning legal frameworks.

Background: Ontology and Child Welfare Interventions
Critical to understanding practice within a DVA context within the family are the conflicting ontological assumptions within the pertinent legislative frameworks. Within criminal law, there is an ontological assumption that a person is responsible for their own actions. Prosecution pivots on the notion that an action is undertaken with intent or with disregard for the outcome 1 (Marchuk 2014: 8-10). Further, the law in the UK, Australia, Canada and the United States of America (USA) is expected to treat all people subject to its jurisdiction equally and to be 'blind' to characteristics such as sex, gender, race and ethnicity. 2 Violence is a defined series of behaviours delineated within and subject to criminal law, intervention and sanction. Following criminal law's ontological premise, an actor is responsible for their violent behaviour; a person using violence within the home (perpetrator) is responsible for the violence they inflict on another (victim). If others are present (adult or child) when the perpetrator behaves violently, the perpetrator is responsible for the outcomes of the violent behaviour and it is the perpetrator who should be held to account. However, within the UK, for example, the Domestic Violence, Crime and Victims Act 2004 (UK 3 ) (DVCV Act 2004) affects families in which DVA occurs and with which a child or vulnerable adult lives. The Act states that a child who is related either to the person being abused or to the perpetrator and who sees or hears or experiences the effects of domestic abuse is also to be regarded as a victim of DVA. Further, the Act states that those who live in the same household as and have frequent contact with a child or vulnerable adult have a duty to protect them from harm and might reasonably be expected to act even if they are the person at whom the violence and abuse are aimed. They might not therefore actively contribute to the harm of the child or vulnerable adult, but under the Act, they may be held accountable under certain conditions. The Act therefore moved the premise of responsibility for a violent action from lying only with the actor. An offence might be judged to have been committed if a person could be considered to have allowed or caused the death of a vulnerable adult or child, and which applied would not need to be proven (DVCV Act 2004). The maximum penalty is 14 years imprisonment.
DVA, threats, intimidation and fear are considered mitigating factors in the legislation (DVCV Act 2004), and it must be proven that someone either failed to take reasonable steps to shield the victim from a serious risk of harm or caused the death of the victim (DVCV Act 2004). 'Reasonable steps' are not defined. Within current child welfare, the primacy of the rights of the child drives the nature of the engagement with the family. In the context of DVA, a social worker will become involved to consider the rights of the child not to be harmed by living in a family in which a perpetrator is behaving violently towards an adult victim. The family is not approached from the ontological assumption of criminal law. The focus is not on holding the violent person responsible for their own behaviour and its effect. Instead, each interaction with a family is framed within a different ontological framework of the primacy of the rights of the child and therefore the right of the child not to witness or be harmed by violence being used against another family member. We suggest that the DVCV Act 2004 is influenced by this ontological assumption of the primacy of the rights of the child. This causes unacknowledged ontological slippage that has led to child welfare practices evidenced in the research discussed later in this paper.
Although the legal example above focuses on the UK, child welfare practices in Australia, Canada and the USA follow this trajectory. Child welfare practices place responsibility for safeguarding the children in the home on the mother, not the perpetrator of the violence (Douglas and Walsh 2010). The effect of the violence on the mother is minimised or ignored (Hughes, Chau and Vokkri 2015). Discrimination is embedded in legislation such as the DVCV Act 2004 in England and Wales and 'failure to protect' legislation in Canada and the USA (Goodmark 2010;Strega and Janzen 2013).
When a child welfare investigation is initiated in a family in which DVA is used against the mother, current child welfare practice in Australia, Canada, England and Wales, and the USA rarely engages with the violent perpetrator (man) and instead focuses on the behaviour of the mother and what it considers her ability to 'protect' her child from witnessing DVA. 4 We propose that the focus on the mother is critically linked to our patriarchal, historical, legal and cultural understanding of 'motherhood' and 'mother'.

Domestic Violence and Abuse
DVA is a common experience in all countries (Gracia 2014), and although feminist organisations have countered the normalisation of domestic violence (Dobash and Dobash 1980;Mooney 2000), a study of attitudes across Europe demonstrated that normalising attitudes were commonly held within varying degrees (Gracia 2014). Studies in Australia have also found normalising or excusing attitudes, especially when the perpetrator expresses remorse (e.g., Mitchell 2011).
In England and Wales, the Office of National Statistics (ONS 2017) demonstrated that while both women and men may occasionally experience violence within the home from an intimate partner, domestic violence is most experienced by women, and the use of violence leading to death and murder is highly gendered. Of those murdered in a DVA incident in 2016, 70% were women killed by male partners, and men accounted for the overwhelming majority of violent perpetration leading to death (ONS 2017). Further, women from 16 to 75 and over are most likely to experience a violent assault domestically (ONS 2017); this is not true for men. And in Australia, Mitchell (2011) noted that the homicide of a woman is most commonly at home and perpetrated by an intimate partner. In England and Wales 95% of referrals to multi-agency panels where serious risk of harm is anticipated, are harm perpetrated by men against women (ONS 2017).
Globally, the pattern of violence against women within their home has been repeated and exacerbated during the recent COVID-19 pandemic. UN Women (2020) has referred to a shadow pandemic, with a worrying rise in violence against women and DVA in countries around the world.
The role of violence within the family and this relationship to gender is therefore a critical factor, and it is important to consider why child welfare has paid limited attention to it. Further, it signifies the importance of addressing the scant attention paid to the abusive man within the discourse of child welfare (Bancroft and Silverman 2002;Featherstone and Peckover 2007).

Legal and Sociohistorical Construction of 'Wife' and 'Mother'
Family is central to the social policy and the economic and social care frameworks that have been constructed within Western democratic states. However, the family is rarely theorised in a way that informs the practice of social work with families in which DVA is used (Humphreys and Diemer 2018;Lapierre 2008), and, while much feminist research has theorised the family, this debate has been largely separate from the discourse of child welfare.
Cultural settings, historical discourse and legislation provide the framework for the understanding of and response to DVA, and this includes how the roles and responsibilities of 'mother', 'father' and 'child' have developed over time. Commonly, research across a variety of Western democracies demonstrates that our understanding of the role and responsibilities of 'mother' has significant similarities to the cultural, historical and legislative frameworks in the dominant Judaeo-Christian traditions of the countries considered in this paper.
Relevant to the concept of 'wife' and 'mother' are the frameworks of 'family' and 'home', and acts of violence in the family and of an intimate partner continue to be both gendered and placed within a heteronormative framework (Gracia 2014;Walker 2017). This has been culturally explicit, if frequently ignored, in many contexts of social policy and welfare intervention. It is, however, important to surface and recognise that within many jurisdictions, including England and Wales, the law has historically legitimated certain forms of violence that could be used by a husband (male) to a wife (female) (Kennedy 2018).
As Hagemann-White and Meysen (2019: 204) note, most countries in Europe include a legal scoping of what can be defined as 'domestic violence' and violence within the family. It has historically been treated differently from violence outside the family, and until quite recently (in a historical sense), the use of violence and sexual assault against women and children within the family has been legitimated in law. This historical division in our cultural response to violence and abuse within the family has continued to influence current child welfare practice. The limited use of theory to examine the family and violence within this context has allowed for ontological 'slippage' regarding the criminality of violent behaviour within a family. And facilitated the lack of proper attribution of responsibility for perpetration and harm. Wives and their property historically belonged to their husband whom they were meant to 'obey', and husbands had the legal right to chastise them for failing to do so (Abbott and Meerabeau 1998; Anderson 2016; Kennedy 2018). Legislative prohibition curtailed wives from exercising verbal power characterised as 'scolding' and 'nagging', and in the 17th century, prosecution for 'scolds' in England included widely used inhumane treatments such as 'ducking stools', which frequently led to death ('Ducking stool' 2018). Although no longer a legal sanction, the premise of the husband's right to curtail 'scolding' wives continues to be used as a defence for femicide and leads to a mitigated sentence (Elliott 2008).
In England and Wales, women gained limited rights to separate from their violent husbands under the Matrimonial Causes Act of 1878, with legal protections against acts of violence within a family context not put into place until the 1890 Protection of Children Bill; sexual violence (e.g., incest) was not prohibited until 1908. Violence within marriage became illegal in England and Wales under the Domestic Violence and Matrimonial Proceedings Act 1976 (UK), and it was not until 1991 that a legal judgment established the crime of rape within marriage; however, this was not enshrined in legislation until the Sexual Offences Act 2003 (UK).
Children in England and Wales were the sole property of the husband until the 19th century when mothers gained some rights to petition for access to their children in the Custody of Infants Act 1839 (UK). Nonetheless, the concept of the 'rights' of a husband has continued even when he is violent towards the mother. Madeleine Moon member of the UK Parliament raised this matter in 2016 in the Houses of Parliament, citing an example in her constituency in which a Family Court social worker told children "who were afraid of their father and did not want to visit him that if they did not go, their mother would be in deep trouble" (House of Commons Debate 15 September 2016, 1.20pm, no page number, emphasis added). Humphreys and Diemer (2018: 127) have argued that this 'ubiquitous and destructive' practice of privileging the rights of fathers has been common throughout parts 'of the … intervention system', such as child welfare where the father's use of violence is overlooked and the mother investigated for 'her ability to protect the children from violence'.
In this paper, we explore this assertion against the existing literature. We review the evidence from specific child welfare studies and consider upon whom responsibility is placed during a child welfare investigation in Australia, Canada, England and Wales, and the USA. 5

Methodology
A review of the international literature available in English was conducted using two searches in two separate periods. The searches focused on two key areas: the construction of mother blame and child welfare interventions with mothers experiencing DVA. The search terms that were used include 'mother blame' and 'domestic abuse' and 'domestic violence' and 'intimate partner violence'; 'failure to protect' and 'domestic abuse' and 'domestic violence' and 'intimate partner violence'; 'mother' and 'child welfare social work' and 'child protection' and 'domestic abuse' and 'domestic violence' and 'intimate partner violence'. The review was undertaken by searching relevant databases, including EBSCOhost, Applied Social Sciences Index and Abstracts, ProQuest Central, JSTOR, PyschInfo, Social Services Abstracts, Scopus and Wiley Online. Additionally, Google Scholar was searched to identify any potential additional studies that had not been obtained; this included a citation search of the main articles listed below. We also searched reference lists of the main identified texts. We did not apply an exclusion date to the literature.
We looked first at studies in which mother blame was the specific aim, and we discuss this literature first. Studies regarding DVA that did not include a discussion of work with the family in a child welfare context were excluded, as were studies in which violence was perpetrated by the mother. The articles that are included are those that discussed work with mothers and their children where the mother was not responsible for violence or abuse and where mothers were subject to child welfare interventions because they were experiencing DVA. This allowed us to consider the operationalisation of mother blame in a social work context leading to responsibilisation for violence perpetrated by the husband or father.
The literature on mothers and mothering is largely presented in a homogeneous way that is heteronormative and white-centric, and class, disability, sexuality and other factors are largely unexplored. Additionally, current practice is frequently de-gendered by using apparently gender-neutral language that masks the role and relevance of sex and gender. Our current understanding of the ways in which social and criminal justice policy are now legislated and enacted against mothers and their children in a child welfare context as the result of domestic violence is therefore largely heteronormative and whitecentric with little detailed understanding of how it is affected by inter-sectional factors. 6 With the assistance of reviewers we have sought to address the white-centric gap in the literature.

Mother Blame
'Mother blame' is a term used by academics to examine the origin of the behaviour it seeks to describe (Millet 1969). Academics suggest that following World War II, Western women sought more recognition for their work in the home because they were made to return to the home to accommodate men returning to the workplace (Garey and Arendell 2001;Sommerfield 1989). The role and importance of mothering became the subject of scientific research. This discourse magnified the role of mothering and created an idealised image. Although mother blame was seen to exist before this period, it is argued that the historical and cultural circumstances of the period magnified this concept (Sommerfield 1989), with 'mothers being held responsible for the actions, behavior, health and well-being of their (even adult) children' (Jackson and Mannix 2004: 150). Further, Phares and Compas (1992) observed a sexist bias that focused on a mother's contributions to her child's welfare and outcomes while ignoring the father's contribution (see also Davies and Krane 2006).
This important period plays a role in establishing the expectations that inform current social work practice. The confluence of social, historical, legal and cultural factors created the structural relationships and discourse surrounding family when social work and social work interventions expanded and became influenced by psychosocial explanations (e.g., Lymbery 2005).

Constructing How Mothers Are to Blame
A review of the literature, past and present, pertaining to mother blame demonstrates a perspective of 'mother' in the West as a woman who is giving, selfless, caring and nurturing. Mothers who have these qualities are said to be good mothers, but those who work, are single parents or cannot provide for their children (for whatever reason) will be judged as bad mothers and blamed for any presenting issues. Additionally, Ladd-Taylor (2004) explained that a good mother is willing to lay down her life and sacrifice herself for her child (ideals underpinned by religious notions).
There is always the potential for someone to fall short of an ideal, and Ladd-Taylor (2004) explored how modern-day notions have elevated mothering and its importance in children's lives. Concerns to protect maternal welfare in the 1920s and 1930s (often led by women) changed to a focus on protecting society from bad mothers and their bad children.
In addition, numerous studies outlined a consistent trajectory leading to the responsibilisation of mothers for various matters. In the 1930s and 1940s mothers could be cast as domineering, controlling and in need of psychotherapy to understand their place in the family (Ladd-Taylor 2004). Other examples of the range of outcomes attributed to mothers included autism, schizophrenia, homosexuality, the emotional breakdown of soldiers (Garey and Arendell 2001), and raising transgender children (Johnson and Benson 2014).
After World War II, mothers who stayed at home were considered at risk of smothering their children and raising homosexual boys; those who worked were considered at risk of neglecting their children and creating 'delinquents'. Research might also draw on racial or class-based stereotypes, for example leading to black mothers being represented as matriarchs responsible for black men's poverty and unemployment (Garey and Arendell 2001).
By the 1960s, the growth in the use of recreational drugs and the development of a more liberal social structure meant mothers were blamed for their teenage children's interests and behaviour, including drug use, political protests and sexual activity (Garey and Arendell 2001).
Within the field of social and health care intervention, there have been, and continue to be, implications and repercussions for mothers that arise from this social discourse. Caplan and Hall-McCorquodale (1985) reviewed psychologists' files from the 1970s and 1980s and demonstrated that mothers were blamed for a total of 72 of their children's problems, including family dysfunction and the child's psychopathology. Davies and Krane (2006) argued that the insidious nature of mother blaming has resulted in its general acceptability within social work to the point that it has achieved commonsense status. Moulding et al. (2015) reported that because ideal mothering is elusive and unobtainable, it is easy for mothers to blame and criticise themselves, making them vulnerable to the blame of others. Caplan and Hall-McCorquodale (1985) demonstrated how culture and society affect research. And Winett et al. (2016) highlighted how scientific language is used to hold women accountable for children's issues with no regard to social determinants. Thus, if, as Corcoran (1998) argued, Western societies are permeated with the belief that the mother's behaviour exclusively and directly affects the child's behaviour and outcomes and this forms the basis of scientific study, it becomes problematic when the child has issues, for it is concluded that this is the result of the mother's poor functioning. The range of issues that children are said to experience includes, sleeplessness, poor school performance and low self-esteem (Moulding et al. 2015), chronic disease and risk of disease (Winett et al. 2016), and being sexually abused (Davies and Krane 1997). In addition, failure is said to project into adulthood, with poor or inadequate mothering leading to poverty, divorce and a failure to marry or work (Garey and Arendell 2001). The concentration on mothers is accompanied by an absence of a focus on fathers (Phares and Compas 1992). Davies and Krane (1997) used a case study to investigate how mother blame might be operationalised in certain circumstances. They chose to look at child welfare social workers' responses to allegations of sexual abuse and sought to understand how these practices reproduce mother blame (Davies and Krane 1997). They reported that mothers had a range of responses when child sexual abuse was uncovered, but despite this, the view that mothers were collusive and blameworthy permeated social work practice (Davies and Krane 1997).

Child Welfare and the Operationalisation of Mother Blame
Further, Davies and Krane (1997) argued that social work protection is accomplished through the mother. Outcomes include recourse to the criminal and civil legal systems, and in England and Wales, a mother may be imprisoned for her child's truancy from school (Bray 2016).

Mother Blame, Social Work, Child Welfare, and Domestic Violence and Abuse
Articles were included for review in this section if the research explored how child welfare social workers intervened with families in which mothers were experiencing DVA 7 or considered how social workers' behaviour affected mothers experiencing DVA. Nine papers met these criteria and were drawn from Australia, Canada, England and the USA.  (2008), the USA all used qualitative methods to research mothers' experiences of child welfare in Canada, the USA, England and Australia. All studies acknowledged some good practice but showed that this was not reflective of most mothers' experiences. Social work aims to be a caring profession, but most mothers reported being blamed and made responsible for abuse perpetrated against them, being required to leave a violent partner, experiencing forced child removal and being treated callously by social workers (Neale 2018;Witt and Diaz 2018;Keeling and Van Wormer 2012;Lapierre 2008Lapierre , 2010Douglas and Walsh 2010;Ramsay 2017;and Johnson and Sullivan 2008). Humphreys and Absler's (2011) paper reviewed child welfare documents from the UK, the USA and Canada from 1900 to 2011, and the 'Failure to Protect' Working Group (2000) reviewed documentation across numerous jurisdictions. Neale (2018) stated that staff in overstretched criminal justice, healthcare and social care agencies work with high prevalence rates of domestic violence, which places a significant strain on the financial and emotional resources available to them. Thus, for professionals to make their workloads more manageable, they ignored significant aspects of cases, including the perpetrator's violent behaviour (Neale 2018). They focused solely on the mother's response to the abuse and subsequently blamed her for her inability to manage it; the result was that perpetrators received the message that their behaviour could continue unchecked (Neale 2018). Mothers carried the emotional burden of abuse and continually felt blamed by professionals (Neale 2018). Witt and Diaz (2018) explored childcare social workers' attitudes towards female victims of domestic violence in England. Their study discussed the concept of mothering and the processes through which mothers were potentially denigrated rather than empowered. They identified 10 main themes, including the recognition of domestic violence as a prominent issue within child welfare with a lack of training about how to assess parenting capacity, empower mothers or work effectively with multiple agencies, combined with poor attitudes towards mothers, a lack of trust between mothers and social workers and a need to raise social workers' awareness (Witt and Diaz 2018). They found that social workers' attitudes towards mothers experiencing domestic violence were often clouded by frustration. Social workers reported that mothers were adults who had a choice, whereas they argued that children did not (Witt and Diaz 2018). Witt and Diaz (2018) concluded that social workers did not minimise the difficulties that women with violent partners faced; instead, they found weaknesses with mothers, had difficulty balancing protecting the child and empowering the mother, and used the 'leave ultimatum' against the mother to force her to either leave her violent partner or lose her children (Witt and Diaz 2018). Keeling and Van Wormer's (2012) study found that mothers reported that social workers made them feel defensive, and they were unwilling to disclose any information and were fearful of the intervention. Mothers felt they would be punished if they did not comply with the social worker's demands to separate from the perpetrator or undertake the interventions the social worker outlined (Keeling and Van Wormer 2012). Mothers' said social workers used coercion and threats that they would forcibly remove the child if the mother did not leave, or undertake the actions required by the social worker (Keeling and Van Wormer 2012).
Humphreys and Absler (2011) reviewed 13 child welfare guidance documents from the UK, the USA and Canada from 1900 to 2011. Using content analysis, they identified problems such as invisible perpetrators, domestic violence being treated as a private rather than public problem and women being held responsible for abuse, with the repetitive patterns highlighting the structural nature of the problem (Humphreys and Absler 2011). They argued the problems they identified went beyond the intentions and practice of individual workers and called for a review of policy responses alongside profound cultural change (Humphreys and Absler 2011). Lapierre (2010) reported that mothers in his study who had been subjected to domestic violence felt they had been left to protect and care for their children without support. The mothers felt responsible for their children and that they should be caring and protective. Therefore, when social workers blamed mothers and argued they were bad mothers who were guilty of neglect and had failed to protect their children because they had allowed them to be in situations of violence, the mothers were confused and hurt, and internalised this blame (Lapierre 2010). Lapierre (2008) described the dominant discourse regarding the detrimental nature of children's exposure to domestic violence as informed by a deficit model of mothering. This included perceiving mothers as failing, considering DVA as marital conflict and fighting, and failing to focus on men's violence. Mothers experienced interventions as being punitive, causing them to distrust social workers; this discouraged mothers from seeking help (Lapierre 2008).
Mothers within Johnson and Sullivan's (2008) study demonstrated that they were blamed for the abuse they experienced, had their own concerns dismissed and had extensive requirements placed on them. The 'Failure to Protect' Working Group (2000) found that research on domestic violence highlighted the effect that exposure to domestic abuse had on children, which resulted in focusing on the harm a child faces. In some jurisdictions, such as the USA, mothers who were victims of violence were charged with failure to protect their child on the basis that by being the victim of violence, they had exposed their child to incidents of domestic abuse/violence. Lapierre (2010) showed how practices and policies were amended to state that witnessing or living in a family home in which domestic violence is used is harmful to a child and therefore a form of maltreatment. By this policy and practice, the victim of the violence was made visible and held accountable, and the perpetrator was made invisible. This policy and practice are highly gendered but masked by genderneutral language. It is this policy of failure to protect that has now been built into English and Welsh legislation in the DVCV Act 2004 and the Adoption and Children Act 2002. Utilising the language of failure to protect can lead to prosecuting mothers for behaviour that is not their own. A concern with 'exposure' to domestic abuse is also included in the definition of significant harm in the main pieces of legislation and guidance for child welfare in England and Wales (Adoption and Children Act 2002 [UK]; Department for Children, Schools and Families 2010; Department for Education 2018). The effect of these changes has been to involve social workers more in cases of domestic violence, and evidence demonstrated mothers have a lack of trust in social workers regarding domestic violence (Robbins and Cook 2017).
In England and Wales, almost half of all cases dealt with by Children's Services on a statutory basis involve homes in which there is also domestic violence (ONS 2016). In the year ending June 2017, there were 228,385 child welfare referrals in England and Wales in which domestic violence was present (ONS 2017). How these cases are dealt with therefore affects a large number of mothers and their children. Douglas and Walsh (2010) stated that many mothers are blamed and experience a higher level of scrutiny than do the perpetrators of the violence. They demonstrated how social work interventions in Australia focus not on helping the perpetrator cease their abusive behaviour but on the mother's ability to protect her child(ren) (Douglas and Walsh 2010). As Hartley (2004) concluded, refocusing the judgement onto the mother means she can be judged as failing as a mother because she has not 'prevented' her child's exposure to domestic abuse; legally she has failed to protect her child. This apparently simple step removes responsibility and visibility from the perpetrator and ensures the focus remains on the mother and her mothering (Hartley 2004). Ramsay (2017) demonstrated how concepts of mothering and child protection are imbued with racist and cultural stereotypes and biases. The effect imposed on mothers by investigating social workers who have different racial and ethnic backgrounds from those of the mother, can lead to the forced removal of children. Ramsay's (2017) work examining child protection in a context of domestic violence demonstrated the utilisation by social workers of the concept of deservingness and brought to the fore the contested concept of mother. She highlighted the proven invisibility of male perpetrators in child welfare investigations that result from the perpetrators behaviour and demonstrated how this led to the subsequent investigation of mothers and their mothering (Ramsay 2017). Goodmark (2010) demonstrated how the leave ultimatum is frequently used within social work practice in the USA, with mothers charged with failure to protect if it cannot be evidenced that they have made reasonable efforts to leave the perpetrator and thereby prevent their children from being abused or neglected. Douglas and Walsh (2010) stated that mothers' fears of social workers and the legal system were well founded. Their findings showed that in some circumstances, men who were violent were deemed to be satisfactory fathers, and this justification was used to remove the children from the mother and place them with the father (Douglas and Walsh 2010).

Concluding Thoughts
We conclude that mother blame has been built into theorising about families and that these theories have impacted social work as it has developed. Utilising mother blame within social work practice is functional for a patriarchal society. These practices continue because an analysis of gender and power is largely absent from discussions. This allows for an ontological slippage that has meant mothers can be responsibilised for violence used against them; and mothers are held responsible for the protection of children in any circumstance.
We propose that this historical division in our cultural response to violence and abuse within the family has continued to influence current child welfare practice. The limited theorisation of the family and violence within social work practice has allowed for ontological slippage regarding the criminality of violent behaviour and the proper attribution of responsibility for perpetration and harm.
Reflecting on earlier historical and cultural expectations, we know, for example, that until 1976 in England and Wales, it was expected that mothers remained in abusive relationships. Responsibility was placed on the mother to ameliorate the violence, not on the husband to cease being violent. One of the 'new traps' (Walby 1990: 201) for women that arose as the patriarchy evolved to respond to the successes of women's movements was that a mother experiencing DVA was expected to leave a violent husband with little access to support and resources (Witt and Diaz 2018). She was also deemed responsible for controlling, ameliorating or mitigating violence within the home until, or even after, she was able to leave. Legally she could be held accountable for any apparent failure to control, mitigate or ameliorate violence used against her, if it was judged that children also living in the home were adversely affected by the violence used against her (Douglas and  Domestic violence may have become illegal in Australia, Canada, England and Wales and the USA jurisdictions, but mothers can be held to account for harm they do not commit. They can be told by social workers in the child welfare system that they must leave their home, or risk either forced child removal, or prosecution, purely because they are the victims of violence perpetrated by their male partner. Evidence demonstrates the negative outcomes associated with mothers leaving their homes as a result of DVA. These outcomes include poverty, multiple moves, disrupted schooling for their children and an increased risk of being killed, harassed and stalked by the original perpetrator of the violence (ONS 2017, UK). This article brings together qualitative studies drawn from Australia, Canada, England and Wales and the USA that evidence how mothers who are experiencing DVA are treated by child welfare practice. Examining international similarities of social work practice and DVA legislation allows us to step outside of each particular jurisdiction (emic) and theorise about their similarities (etic).
We suggest patriarchy is a necessary ontological framework for developing our understanding of the 'depth, pervasiveness and interconnectedness' (Walby 1990: 2) of the legal, social policy and social work practices in a child welfare context that lead to the responsibilisation of mothers experiencing DVA.
Child welfare practices in Australia, Canada, England and Wales, and the USA in the arena of domestic violence have a profound daily effect. Legal frameworks serve to hide the gendered nature of violence within the family and ignore the cultural, legal and sociohistorical framing of the family. This facilitates an apparent child focus in child welfare practices that fails to engage with analyses of gender and power and in turn perpetuates and facilitates the structural victimisation of mothers experiencing DVA (Hooks 1989). In addition, it serves to obscure the responsibilisation of mothers for violence used against them. As Lerner (1986: 5) argued, women play a central and active role as mothers in creating society but become marginal to the meaning-giving process. In the context of mothers experiencing DVA leading to child welfare interventions, the central, active function of mothers is said to be to keep children safe; the solutions that mothers might offer about how best to achieve that appears principally absent from the current debate. The lack of engagement with families and their meaning-making in current social work practice in England and Wales has led Meysen and Kelly (2018:228) to suggest that on occasions it may seem as if the 'heart of social work had atrophied'.
Researchers do not want to be observed to blame social workers (Anglin 2002) and have argued that the stress and strain social workers experience and the financial constraints and legal frameworks under which they operate, may account for discrimination against mothers experiencing DVA. However, it has been repeatedly demonstrated that current child welfare practice in Australia, Canada, England and Wales, and the USA, blames mothers subjected to violence and abuse within their home. The violent father has become invisible, and his parenting is not subjected to social work scrutiny. For the father who is violent in the home, the parental, social, economic, criminal justice and practical benefits that result from invisibility are those you might expect to see within a system of private patriarchy (Walby 1990). Those men may therefore be unwilling to support changes to policy and practice or may actively oppose it.
Patriarchy has operated historically to circumscribe and delineate women's roles as wives, mothers and daughters in relation to men as husbands and fathers (Anderson 2016;de Beauvoir 1949;Hooks 1989;Kennedy 2018;Lerner 1986;Millet 1969;Walby 1990). While feminist discourse and women's rights moments have led to changes, these have been slow and incremental over two decades. As Hooks (1989) argued, patriarchy creates an ideological system used by women and men to oppress other women.
Mothers and fathers are judged by different standards, and child welfare concerns are rarely 'articulated as a protective mother / harmful father' (Douglas and Walsh 2010: 496). By placing blame on mothers for the violence they experience, we fail to address violent behaviour, and in some cases, we elevate the violent behaviour (ONS 2017). Failure to adequately place responsibility for the behaviour and violence where it belongs means we potentially widen social exposure to violence because those who use violence in their relationships have their positions legitimated, and they continue to behave this way in subsequent relationships, leading to further harms (see e.g., Wemrell et al. 2019). As Douglas and Walsh (2010:495) discussed, 'work with men who engage in domestic violence remains unusual'.
Structural issues, economic factors and society's concern to control and ameliorate risk (Anglin 2002) impact social work and these undoubtedly play a role in shaping responses to mothers experiencing domestic violence. We do not argue with this. We argue that by looking beyond country borders, we can see more clearly how responses are also embedded in ideological, ontological and patriarchal assumptions that shape how mothers experiencing domestic violence are perceived. Mother blame has been institutionalised within child welfare policy and practice, and legal frameworks and in five countries (Australia, Canada, England and Wales and the USA), children are forcibly removed from mothers who are the victims of abuse. For social workers, critical self-reflection and critically oriented supervision may be vital tools to disrupt these dominant discourses.
The literature demonstrates how an unacknowledged private and public patriarchal ontology is built into child welfare practice. As social workers and researchers, we should use this evidence to argue for policy and practice change (Humphreys and Absler 2011). We call for a review of policy and legal frameworks and for social workers and other key actors in this setting to argue and lobby for policy and practice change.