Gender-and power sensitivity, securitisation and social peace: rethinking protection for children exposed to post-separation violence

This article offers a rethinking of protection based on synthesised data from Finland and Sweden on children’s and mothers’ experiences of post-separation stalking, and social workers’ case reports on children risking exposure to gender-based violence after separation. Drawing on critical childhood studies and a feminist approach to violence and security, we ask how children’s everyday lives can be incorporated in a rethinking of protection for children in post-separation contexts. Departing from identified limitations in protective solutions for children, we propose three ways of rethinking the issue of protection: (1) protection as gender-and power sensitivity, (2) protection as securitising the here and now, and (3) protection as social peace. Our findings call for some changes in professional practices, social policy and legislation.


Introduction
It is generally agreed that children who hear, see and in other ways experience violence against their mothers are, in principle, victims in their own right, meaning that gender-based violence is also violence against children (Øverlien, 2010;Eriksson, 2011;Callaghan et al, 2018).However, in this article we argue that while children's exposure to violence has been a significant concern in child welfare, recognising violence at the symbolic level -that is at the level of policy and discourse -does not imply that protective solutions will be provided (for example, Pringle, 1998;Humphreys and Absler, 2011;Stanley and Humphreys, 2015;Knezevic, 2020a).The obligation to protect does not imply that the state and its representatives, including child welfare services, are always an effective option or the only source of safety (for example, Pringle, 1998;Stanley and Humphreys, 2015).
In this article, we focus on post-separation gender-based violence, here defined as a form of men's violence against women and children that occurs after separation.Post-separation violence consists of repeated, threatening and intrusive acts that may encompass physical, psychological, emotional, verbal and other forms of violence, such as stalking.Victims perceive these acts as coercive and controlling, causing feelings of loss of dignity and safety, powerlessness and entrapment (Stark, 2007;DeKeseredy et al, 2017;Thiara and Humphreys, 2017).Although post-separation violence often remains statistically unrecorded, the phases before, during and after separation have been documented as dangerous for women and children, and increasing risks of abuse, feminicide and homicide (DeKeseredy and Schwartz, 2009;Hester, 2011;DeKeseredy et al, 2017;Bruno, 2018).A report from Finland indicates that half of the men who had perpetrated violence during a relationship continued violent, threatening, or harassing behaviour after separation (Piispa, 2006).Statistics from Sweden relate 75 per cent of the lethal violence by a current or former partner to the separation phase (Socialstyrelsen, 2018).
Our analysis on the limitation of protection for children is based on data from Finland and Sweden concerning children exposed to post-separation violence and authorities' protective responses.Inspired by critical childhood studies and by adopting a feminist lens on violence and security, we ask how children's everyday lives can be incorporated in a rethinking of protection for children in post-separation contexts.
In the sections that follow, we discuss previous contributions to the field and outline our theoretical points of departure, our data and methods used.Our analysis presents three overarching themes of protection, each addressing limitations but also ways forward: (1) gender-and power sensitivity, (2) securitising the here and now, and (3) social peace.Finally, we discuss new directions for research, professional practices, social policy and legislation.

Protective measures for children in post separation: symbolic and practical limitations
The idea that men's violence against women and children are two distinct phenomena continues to endanger children's lives.The role of children and parenting are the central factors constituting the dynamics of post-separation violence (for example, DeKeseredy et al, 2017).The abusive parent seeks contact with the ex-partner and children through tactics such as contact arrangements, making false allegations of child abuse, and using children as an indirect means of contact with the ex-partner.
Unauthenticated | Downloaded 09/14/23 03:41 PM UTC Children can also be direct targets of violence (for example, Nikupeteri and Laitinen, 2015;DeKeseredy et al, 2017;Callaghan et al, 2018).Furthermore, the lack of common understanding about what constitutes separation/divorce can hinder children from getting protection.Post-separation violence is often assumed to occur after physical or legal separation.However, many women make the decision to leave the abusive relationship long before they are actually able to exit it physically or legally (DeKeseredy and Schwartz, 2009).Previous research on protection of children also highlights the discrepancies between policy and practice by shedding light on the limitations of policy implementation and the pitfalls of gender-neutral constructions of violence (Breines and Gordon, 1983;Eriksson, 2011;Bruno, 2018).Other challenges point to contradictions between different approaches to violence in the area of domestic violence work, child contact and child protection (Hester, 2011).Limitations to the protection of children from post-separation violence have also been linked to the tension between the criminal law, targeting adults and family law focusing on children (Nordborg, 2005).
These and other contributions discussing how children view, experience and reconceptualise post-separation violence and protection compart critical responses to adultcentric protection and disbelief in children's concerns as victims (Nordborg, 2005;Øverlien, 2013;Åkerlund and Gottzén, 2017;Moore and McArthur, 2017;Bergman and Eriksson, 2018;Callaghan et al, 2018).While previous research acknowledges both adultcentrism and gender neutrality of authorities' handling of violence, less attention is paid to the practical solutions and the limitation of their protective dimensions.However, researchers advocate children's involvement in developing services and some commentators describe contemporary services as 'a "bolt on" to existing domestic violence services' (Mullender et al, 2003;Callaghan et al, 2018Callaghan et al, : 1554)).Thus, the support provided may not be adequate or effective from the perspective of children (Pringle, 1998;Hester, 2011;Stanley and Humphreys, 2015;Callaghan et al, 2018), and could pose a risk for the child involved (Bergman and Eriksson, 2018).Also noteworthy is the role of legislative frameworks, and, in particular, the fact that the requirement of the guardian's consent may deprive younger children of support (Heimer and Palme, 2016), or may grant abusive parents the right to decide on the support children receive (Bruno, 2018).Although these contributions are crucial to understanding the barriers to accessing support, there is a tendency in the field to equte support with therapy or pay less attention to protective measures for children other than therapy.
Aligned with critical research on how individual-, therapy-and symptom-oriented responses depoliticise violence and are inadequate as responses to societal problems (Knezevic, 2020a(Knezevic, , 2020b;;Vera-Gray, 2020).In this study we move beyond problemframing inquiries (see Bacchi, 1999) to an examination of solutions and security constraints.We take as our starting point child welfare responses and children's and mothers' accounts of their daily lives, and adopt a processual approach to violence and security, as elaborated on next.

Securitisation and social peace
Inspired by feminist violence and security studies and critical childhood studies, we start with the idea that analysing the everyday life of children and women is central to efforts to conceptualise security from their perspective.Adopting a lens of critical childhood studies, we acknowledge children and childhoods as worthy of study in their own right rather than acknowledging them as the adults they are to become (James et al, 1998;Moore and McArthur, 2017).Furthermore, we recognise children as knowledgeable of security concerns.This does not mean treating children as isolated beings with authentic voices disconnected from societal norms and discourses (Nordborg, 2005;Åkerlund and Gottzén, 2017), nor does it mean adopting a neoliberal discourse of victim blame that assigns responsibility to children (or women) for their own safety (Raby, 2014).
On the contrary, we use the notion of securitisation from feminist security theory as a tool for analysing how securitising (state) actors construct concerns into security issues (Hansen, 2000;Mackenzie, 2009).Feminist scholars share a long history of discussing security issues for women and girls, including the unequal distribution of securitisation.The crucial questions asked are what constitutes a security issue, who or what is constituted a threat and whose and what forms of security issues become possible in the process (Hansen, 2000;for example, Wendt Höjer, 2002;MacKenzie, 2009).
The public/private divide and other processes of (de)politicisation are important to consider in relation to the issues and individuals that fall outside security discourses, hence are desecuritised and rendered private issues (for example, Breines and Gordon, 1983;Hansen, 2000;Wendt Höjer, 2002;MacKenzie, 2009;Hester, 2011).The women's movement has been pivotal for framing violence as a public and gender-based issue but, in the absence of a corresponding movement for and by children (Breines and Gordon, 1983) the problem continues to be framed from a predominantly adult perspective, thus effectively excluding children from discourses of gender and gender-based violence (Callaghan et al, 2018;Knezevic, 2020b).However, there are also differences in securitisation of children's violence exposure.For example, state authorities protect children from 'radicalisation' or the patriarchal 'other' but use legal and other measures to maintain the nuclear family ideal and biological fathers' rights in Nordic and white contexts, which are generally viewed as less patriarchal (Nordborg, 2005;Bruno, 2018;Kirton, 2018;for example, Knezevic, 2020a).Thus, securitisation works in tandem with racialisation and other gendered processes of othering, targeting groups of people such as migrants or Muslims, across societal contexts, including that of securitisation of gender-based violence (Aharoni and Féron, 2020).
In an attempt to analyse securitisation through the solutions provided and in relation to children's everyday life, we use the notion of social peace for children, also referred to as 'children's peace' (Nordborg, 2005).In feminist violence research, social peace connotes Kvinnofrid (women's peace), a liberty law referring to women's inviolability, safety and social peace (Wendt Höjer, 2002;Nordborg, 2005;Bruno, 2018).The term Kvinnofrid re-emerged in the 1990s in Sweden in the context of The Protection of Women's Integrity Reform that addresses women's freedom from gross violation of integrity.Informed by a feminist approach to violence, the reform implied a shift in focus from singular criminal events to violence as gender-and power-related, systematic, repeated, and as a process (for example, Breines and Gordon, 1983;Wendt Höjer, 2002;Nordborg, 2005).Aligned with this, we approach both violence and protection from it as a process that can permeate children's everyday lives, across space and time.

Methods
This article has resulted from an international collaboration of a Finnish project (Children's Knowing Agency in Private, Multiprofessional and Societal Settings -the Case of Parental Stalking).We combine two larger projects -one from Finland, and one from Sweden -both focusing on children's exposure to men's violence against women.The Finnish dataset comprises children's and mothers' accounts of postseparation stalking.The data from Sweden consist of social workers' written child welfare assessments, here delimited to children at risk of being exposed to violence after separation.While interview excerpts allow us to examine children's and mothers' experiences in situations in which they feel unsafe and their attempts to create a sense of safety (Nikupeteri and Laitinen, 2015), assessment reports provide documented information on recommended services in corresponding situations.By combining two different data sets, we aimed to gain an enhanced understanding of protection for children exposed to post-separation violence.

Interviews with children and mothers
The data contain semi-structured, teller-focused (Hydén, 2014) interviews with 14 children (eleven girls and three boys, aged 4-21), who had experienced their father's/stepfather's stalking of their mother, and interviews with 14 women who had experienced stalking from their male ex-partners.All the children and most of the mothers were recruited through two local chapters of the Federation of Mother and Child Homes and Shelters in Finland.This organisation provides nation-wide services to the victims and perpetrators of violence.Professionals at the local chapters identified potential interviewees who met the eligibility criteria for this study in terms of the stalking situation and the children's age and asked about their interest in participating.Five mothers were recruited either through the researcher's contacts with other academics in the field, or the mothers contacted the researcher after having heard about the study.Most of the children and mothers were unfamiliar with one another.
The children were interviewed at the offices of the local chapters.Some of the mothers were met additionally in a quiet location conducive to discussing their experiences, if they wished so.The mothers were interviewed in the presence of a familiar professional (with an educational background in social and therapeutic work with victims of violence), who could provide emotional support during the interview, if the interviewee wanted this.The professionals interviewed the children without the presence of the researchers to help them feel safe in the interview situation.The professionals and researchers planned the interview themes.Some children chose being interviewed individually and others with their siblings.All interviews were recorded and transcribed and confidentiality was carefully maintained.
The research project was approved by the Research Ethics Committee of the University of Lapland and the interviewees and children's guardians gave their informed consents.Throughout the project, the researchers prioritised meeting the ethical demands of researching a sensitive topic and working with vulnerable people.
The limitation of the dataset is that the interviewees were asked broadly about their experiences of post-separation stalking without explicitly focusing on the theme of protection.Thus, the conclusions derived from the passages are the adult researchers' interpretations of the interviewees' everyday life experiences of post-separation stalking (for example, Åkerlund and Gottzén, 2017).

Child welfare assessment reports
The data consist of 26 written assessment case reports addressing 26 children, up to the age of 12 years, from 23 families.The purposeful selection and reading were undertaken after readings of a larger sample of 283 case reports, all closed in 2015, and collected in one child welfare agency in Sweden.The reports involve history of violence in heterosexual relationships, primarily biological fathers' violence against the child's mother.While more cases address violence, the selected case reports recognise violence more explicitly, that is, it is mentioned in the case summary, in the outline of previous reports and/or in the investigated questions.Despite being limited regarding children's views, the data provide insights into the kinds of responses and changes made to the child's everyday life circumstances that professionals' recognition of violence generates for the child concerned (see Knezevic, 2020a).
The case reports exemplify voluntary services and court-ordered support in postseparation violence contexts.Unlike compulsory care (under the Care of Young Persons Act, SCS [Swedish Code of Statutes], 1990: 52), recommended services need to be accepted by custodians to become actual for the children in this age group (Social Service Act, SCS, 2001: 453).The suggestions (number of cases in parentheses) irrespective of acceptance/rejection can be categorised as follows: individual or group-counselling for child (6), court-ordered visitation support (5), support person for child (adult mentor) (1), residential care to enable child's visitations with father (1), future follow-up (1), counselling for mother (8), home therapist (2), counselling for father (6), programme for violent men (4), and familylaw counselling (3).Previous services are not always documented.However, two case reports mention previous protection in shelters for mothers and children and two reports discuss the issuing of restraining orders.
Although the services vary across municipalities, the case reports were guided by the framework for assessment, BBIC (Barns behov i centrum) (Children's Needs in Focus), which has a nation-wide usage in Sweden and supports a standardised approach, hence suggesting a similar assessment regardless of location (Knezevic, 2020b).
The Swedish study was approved by the Regional Ethical Review Board in Uppsala, Sweden.Confidentiality was applied in both data sets by providing pseudonyms and leaving out detailed information about the individual cases.This included omitting information about age, ethnicity and class, thus limiting an intersectional analysis in this study (see Knezevic, 2020a).

Analysis
The child protection system in Finland and in Sweden is described as family serviceoriented, but recent research also discusses increasing child-focus orientation (Gilbert et al, 2011).With this similarity in mind, we chose protective responses to children as a point of orientation.First, we analysed the two data sets separately by reading and thematising the experiences of the children and mothers, and professionals' practical responses, in the original languages.Then we shared the themes and discussed the problems of providing protection for children.We developed thematic clusters with short memos of children's and mothers' experiences of protection needed and received in post-separation stalking contexts and professional responses to post-separation violence.This helped in creating an integrated understanding of protection in the two data sets (for example, Miles and Huberman, 1994).
Second, we did an interpretative reading based on the clusters by looking at the limitations of protection and formulating alternative approaches in dialogue with the theoretical concepts used.Finally, we formed three overarching themes: (1) gender-and power sensitivity protection, (2) protection as securitising the here and now, and (3) protection as social peace.To reach this stage, several cycles to synergy and reflective discussions between us researchers were required to maintain the integrity of the original data sets.

Gender-and power-sensitive protection
Although reforms have been made that frame violence as an issue of gender inequality and power imbalance, child welfare services in Finland and in Sweden continue to approach gender-based violence in gender-neutral terms as 'family violence'.The gender-neutral term downplays violence as a gender-inequality issue.Consequently, mothers are ascribed blame for violence and are held responsible for the wellbeing and security of their children despite being victims themselves (Breines and Gordon, 1983;Humphreys and Absler, 2011;Nikupeteri, 2017;DeKeseredy et al, 2017).If they cannot protect the children, they may have no choice but to let them be taken into care, as this seems the only way to protect them from the stalker: If we think about the child welfare services, what has irritated me is that I've asked for many kinds of help.I've tried many channels, but the response I got from the child welfare services was that here it's not the clients who make suggestions; these go top down.I'd suggested what kind of help would work in our case.Ultimately, the only thing they had to offer was that the children could be taken into care.This was the only way to protect the children.(Marjo, mother) If professionals equate protection with the child's sense of feeling safe, this feeling risks being diminished if children are removed from their safe parent.A removal contradicts recent research showing that mothers and children may support each other (Øverlien, 2013;Nikupeteri and Laitinen, 2015;Katz, 2015).This mutual support can be seen in the Finnish data when no other help is available for them: It might be that we tried to reassure her [mother]…that if we're there, father wouldn't do anything … I was mostly worried about my mother.(Rosa, child) We guard [the mother]; we watch over her.(Nelli, child) While the interviewed children perceive the mother's situation affecting their own, the data on child welfare assessments are, in many ways, indicative of a contrasting approach.For instance, children's experiences are kept separate from the experiences of the mothers (Eriksson, 2003), as the case of Max exemplifies: Max has told about an event when dad had physically hit and pushed him; but it is most of all dad's yelling and quarrelling that makes Max afraid of dad.Max is not too fond of visitations with dad.Max is perceived as an insecure boy.It is difficult to say how much the parents' different child rearing patterns and their ongoing conflict affect Max.
What could be interpreted as the father's continuous controlling behaviour is framed as a matter of the parents' conflict surrounding 'child rearing patterns'.A support person, that is an adult mentor, is recommended to strengthen Max's self-confidence, but Max remains unprotected from the father's yelling.The gender-and powerneutral assessment of the problem matches the equally gender-and power-neutral practical response to the child.However, it is not necessarily the assessment that guides the recommendation as the case of Alexander here suggests.Although professionals recognise the father's violence against the mother, this recognition does not seem to be affecting Alexander's situation.
The Social Services have granted treatment through counselling [programme for violent men] for Alexander's father, however, it is uncompleted.Support services in the form of family counselling have been discussed to improve communication between the parents since the investigation has shown that this may be a source of disagreement, irritation and quarrels between them.
The mention of a programme for violent men exemplifies a symbolic recognition of father's violence against the mother.However, the programme coexists with other recommendations and gender-and power-neutral framings of 'conflict' between equally responsible parties, hence a concern for 'family counselling' (Wendt Höjer, 2002;Hester, 2011;Bruno, 2018).In addition, the recommended voluntary programme does not guarantee improvements for Alexander because the father may not choose to participate in it.Thus, regardless of the father's (non-)participation in the programme, his interaction with Alexander remains unchanged.
Thus, when the professionals do recognise gender-based violence, the practical responses available for social workers and other professionals to protect children are limited.From this follows that recognition of gendered violence and recommendations for solutions seem disconnected from assessments of power and gender and how they affect children's everyday lives (Hester, 2011;Knezevic, 2020a).This resonates with research showing that the child's relationship and contact with the violent biological parent, typically the father who is coded as Nordic and white, may be taken for granted (Nordborg, 2005;Bergman and Eriksson, 2018;Bruno, 2018).Professionals' failure to recognise the gendered dimension of the problem may simultaneously result in removing children from their homes and thus separating them from their safe parent (Nikupeteri and Laitinen, 2015;Katz, 2015).

Securitising the here and now
Through its individualisation of children and its reliance on the idea of investing in childhoods to create healthy and productive future adults, child-focused child welfare foregrounds children in distinct and problematic ways (Gilbert et al, 2011).Our combined data point towards a need for a focus shift from protection from symptoms and future outcomes of violence to protection from violence per se.
Although various forms of violence from corporeal punishment to humiliating treatment are recognised reasons for support in both countries, there may be a lack of protective response if a case does not indicate harm to health and development, as stated in compulsory care acts (Östberg, 2010;Husso et al, 2012;Knezevic, 2020aKnezevic, , 2020b)).Voluntary therapeutic service may be recommended for children who witness father's violence against mother or are the target of it themselves.The response is exemplified in the assessment of Saga, one among multiple case reports closing with counselling as a recommendation for the child.in telephone conversations with the mother, the social worker talks about [group counselling for children with separated parents] for Saga but mother says that even if Saga attends [the counselling] her father will still treat [Saga] the same way.
Violence is a common reason to open a case, hence recognised by professionals who make a referral and social workers who continue investigating (Östberg, 2010).In addition, both data sets include examples of professionals' recognition of mothers' need of protective judicial measures such as restraining orders.However, the practical output for the child in Saga's case here -counselling -does not constitute father's behaviour as a security issue.Thus, it does not protect Saga from the way she is treated during visitations, which makes her, as stated in the case report, repeatedly flee back home to her mother.Thus, protection targets primarily fathers' interaction with mothers and not with the children, as exemplified by Tobias' case: Both parents and grandparents describe Tobias as a healthy boy.There is no major concern on their part regarding his wellbeing at the moment.The investigator perceives the children as healthy and that they have a pleasant contact with [father] and grandparents.It is worrying that [father] does not accept the separation with [mother] and has been granted a restraining order.However, the children are currently protected from psychological violence directed at the mother because they have separated.In addition, the restraining order seems to offer protection from it [violence].
In the light of the above-mentioned health-and symptoms-oriented assessments of children, legal enhancement of protection for children would appear as a solution.The Finnish data in particular also exhibit other needs than just therapy and suggest that strong judicial interventions, such as prison sentences, may provide a sense of security, as Susanna describes: At least now [I feel] safe, it's the awareness that he [father] is kept in prison.(Susanna, child) The same sample also shows that the imprisoned father may use other persons as a means of stalking.In such situations, imprisonment does not safeguard the children and mothers.Similar limitations apply to the effectiveness of restraining orders, as stated by a mother: My ex-partner didn't accept this decision [the mother leaving with the child] and then the threatening continued, and coming to us or he's tried to come in to us, and there's been many litigations and I and kids have had restraining order against him for four years now, so the threat is targeted to all of us.[…] He blustered that he doesn't care about these [restraining orders].Nobody will stop him.Nobody arrests him.(Maria, mother) Some children emphasised the need for external protection, especially from the father's unexpected physical attacks and intrusions into the home or school.They brought forth the importance of physical protection where concrete devices, such as alarm systems and shelters, increased their sense of security.The mothers and children had organised their lives through minimising risks (see also Callaghan et al, 2018).They utilised relatives and dogs as guards, and organised self-protection through everyday routines.For example, some mothers and children decided to relocate to a place where they feel safe, as illustrated here: The flat felt safe because no one could climb to the balcony or enter the corridor.You couldn't go anywhere without a key.Then the mailbox, you couldn't drop anything in because the hatch was so small.(…) When you went out through the door, you had to check that no one was there.(Noora, child) Both therapeutic and legal responses to post-separation violence imply limitations.This is mainly because they are based on an adult-oriented (Nordborg, 2005;Hester, 2011;Callaghan et al, 2018) but also future-oriented protection.Previous contributions have problematised the latter in terms of a disengagement with a child's history of being exposed to violence (Nordborg, 2005;Hester, 2011;Bergman and Eriksson, 2018).Our analysis further problematises the future-orientedness of the existing approaches by revealing that children's fear of violence in their everyday lives is often overlooked.Using the notion of (de)securitisation and drawing on critical childhood studies, we propose a re-orientation from adultcentrism and symptom-oriented protection to an approach that actualises a child's legally defined position as a crime victim (Eriksson, 2011), hence making violence per se as the main security issue and childhood in the here and now as the focus of securitisation.We argue that our re-oriented approach would better match the security acts children themselves undertake in the face of the state representatives' failure to effectively respond and protect their rights.It also protects children's social peace, as will be discussed next.

Protection as social peace
The data from Finland show the importance of understanding how the father's stalking causes constant threat and sense of fear for children.The father's different forms of contact cause the children to fear that he could appear anywhere, anytime, to harass and possibly attack them.The father's physical, technological and virtual presence makes children feel intruded upon, irrespective of whether they are indoors or outdoors.Fathers' constant negative presence in the children's everyday lives disrupts children's social peace and sense of '"normal" everyday life' (Øverlien, 2013: 280;Thiara and Humphreys, 2017).
And then he harassed the children.When the children were outside, he might go ask them if their mother had a boyfriend.This happened repeatedly.The children often came inside and said that their father didn't let them play in peace and quiet and that they wanted to go inside to play.If they moved, he'd follow them right away.(Pauliina, mother) It was night time and someone using an unknown number called.I was wondering who it might be, but I answered.The voice said, "Hi, it's Daddy here", and I hung up immediately.It felt like . . .after that, my mom came to pick me up [from visiting a friend] because I was really in a panic.(Laura, child) From the perspective of social peace, post-separation violence resembles a process rather than an event (Breines and Gordon, 1983).Simultaneously, protection from postseparation violence may be reduced to single occasions, as the Swedish data indicate.In the Sanna's case, the child welfare agency provided Sanna and her mother a place in a shelter.In the meantime, it was decided that a support person, that is the presence of an external adult, would accompany Sanna during the regulated visitations with her father.This is the only service provided for children in the sample that can have a protective effect, though not intended as a protective measure per se (Bergman and Eriksson, 2018).The assistance given only covers regulated meetings, not unplanned ones that may happen on Sanna's way to school: Sanna has tried to go to school.But does not feel good at all, feels insecure and worried that she will meet her brother who also goes to [the same] school.Sanna's concern is that the brother might contact [her father].
Given the services provided, the idea of protection in Sanna's case is limited.It is assumed that children are protected either by being removed from the risky home (staying at a shelter), or by being accompanied during visitations with the abusive parent.Yet, children's daily lives involve far more activities and, as the Finnish data show, children describe also feeling unsafe when staying at the shelters, which are generally considered a safe haven: when we came to the shelter the first time, Dad wandered around the house and stalked.It was that kind of [thing], he stalked during the nights and in the end, he came even during the daytime, trying to come in, trying to get into the house where we were.(Lotta, child) Furthermore, visitation support may jeopardise protection from violence in postseparation (Eriksson, 2011;Bergman and Eriksson, 2018;Bruno, 2018).In one Finnish case, a support person was assigned to children to ensure safety of visitations but the professional whose duty was to supervise was so afraid of the children's father that a guard had to be present.A similar case in the Swedish sample describes how the support person had to be replaced because of father's threats.
[Liv] needs to have a safe relationship with her father.She has not met her father on a regular basis in recent years.Although the support person during visitation has been involved, some visitations have been unrestrained and [father] has been threatening towards both the mother and the support person.

[...]
Introducing a new support person [during visitation] is not considered sufficient.Additional support is needed and a residential care treatment is therefore granted.
As the case suggests, a support person during visitations is unable to protect Liv from psychological and verbal abuse.The case highlights the difference between 'feeling safe' and 'being safe' as it makes things appear safe rather than ensuring that the child feels safe at the same time (Moore and McArthur, 2017).By focusing on the child being safe during the visitation event, the Social Services do not consider that Liv, in the aftermath of her experiences with father, might feel unease and show resistance to the visitations (Bruno, 2018).Instead, the Social Services propose contact under more controlled forms, that is with the aid of institutional care.Thus, the support intended to create 'a safe relationship' between the child and the abusive parent desecuritises children's social peace.
The Swedish sample also indicates that a concrete change in the child's life may take place in the form of reduced visitation time (Knezevic, 2020a).However, this does not lead to social peace for those children who are afraid of their parent or are exposed to stalking.Particularly the latter situations raise needs for protection beyond single violent incidents, bound to a specific place (home) or time (visitation), and towards 'the continuity [of support that] doesn't vanish at the stage when it's considered the situation is over', as stressed by Minna, one of our adolescent interviewees.
Approaching protection as social peace recognises that violence and violations of children's liberty and rights (Nordborg, 2005;Stark, 2007;Bruno, 2018), and consequently protection thereof, should be seen as an ongoing situation in children's everyday lives.Inter-professional collaboration is essential from the viewpoint of continuing support.The processual approach requires that adults and professionals who spend time with children at school and other contexts are aware of the dynamics of post-separation violence.

Discussion
This study echoes previous research in that it shows the seriousness of post-separation violence and children's need for protection in these circumstances.Our proposed gender-and power-sensitive, securitising and social-peace approach highlights three dimensions of protection of children in the context of post-separation violence that require changes in child welfare services, professional practices, social policy and legislation.Although children's everyday lives and child services vary across childhoods and locations, our rethinking of protection contributes to the field by introducing concepts and analytical approaches applicable to research contexts beyond Finland and Sweden.
First, we argue that although gender-based violence may be recognised at a symbolic level, there may be obstacles to granting children protection in practice.Like previous research, our findings show that gender and power analyses may be absent once children are in the picture (Nordborg, 2005;Hester, 2011), this also applies in cases where professionals recognise gender-based violence.This suggests that patriarchal violence is not recognised in the full sense of the term because it falls short of accounting for adult male rule over children, and not only over women (Nordborg, 2005;Øverlien, 2013).Our main finding in this regard, however, is that there are no, or very few, services to recommend to children that actually protect them from this form of violence.We consider the solutions not as mere materialisations of discourses 'from above' (see Bacchi, 1999) but as influencing the kinds of problems that are assessed, and securitised, in the first place (Knezevic, 2020b).This calls for changes in policies considering the solutions on offer.By calling for more appropriate services for victims of post-separation violence, we pinpoint a way of altering the gender-neutral assessments that precede them.Without gender-and power-sensitive protective measures for children, assessments of the violence as gender-based, if such assessments exist at all, are hardly helpful 'in practice'.
We also argue that the experiences of the child and the solutions provided should neither be separated from, nor reduced to, those of the abused parent.Gender-and power-sensitive solutions that strengthen a child's relationship with their safe parent are important in protecting children as agentic subjects in their everyday lives (James et al, 1998;Øverlien, 2013;Katz, 2015;Callaghan et al, 2018).
Second, what we call 'securitising the here and now' relates to what previously has been addressed as criminalisation of children's exposure to violence (for example, Hester, 2011) and provisions of immediate safety (Øverlien, 2013).This is a critical response to two identified limitations: the limitation of therapeutisation and legal measures.Some children who are responded to with therapy simultaneously undergo regulated visitations with the abusive parent.This creates the paradox of therapeutic responses running in parallel with measures, which can be interpreted as additional desecuritisation and traumatisation of children.While the limitations of legal measures target primarily mothers (Hester, 2011), we argue that even the services that are supposed to protect children are oriented towards adults.Children not only rely on adults for access to the services (Heimer and Palme, 2016), but the services available are oriented towards their future adulthoods, rather than the lives they lead as children in the present.
This brings us to our third rethinking: protection as social peace.Protection is conceived of as processual, stretching beyond the 'service user' position and encompassing multiple parts of children's everyday lives.Protecting children in this way is not only the task of the Social Services but of all professionals that are in contact with the child.Recognising the strategies children are already employing is a step closer to an understanding of what makes them feel safe and secure (Stanley and Humphreys, 2015;Moore and McArthur, 2017).Together with the aforementioned approaches, this can form the base for more unified multi-professional practices (for example, Hester, 2011).Understanding protection as securitising of childhoods in the here and now, children's social peace and gender-and power sensitivity sensitivity -requires valuing