‘Gold Standard' Legislation for Adults Only: Reconceptualising Children as ‘Adjoined Victims' Under the Domestic Abuse (Scotland) Act 2018

In this article we argue that the Domestic Abuse (Scotland) Act 2018 should not be regarded as ‘gold standard’ in the way in which it seeks to recognise the harms caused to children who experience intimate partner coercive control in their living environment. We argue that children should be reconceptualised children as ‘adjoined victims’ of intimate partner domestic abuse and that the 2018 Act should be amended to include a parallel section 1 offence of ‘abusive behaviour towards partner or ex-partner and adjoined child’. By offering the first academic analysis of why and how the criminal law should seek to capture children’s experiences of coercive control, this article contributes to broader discussions about criminalising coercive control and the scope of such offences. It highlights key lessons that can be learnt from the Scottish story so far and sounds a note of caution against simply ‘rolling out’ the Scottish approach elsewhere.


Introduction
Although discussions about criminalising domestic abuse are decades old (Tadros, 2004(Tadros, -2005Tuerkheimer, 2004), over the last five years these have become less academic and abstract. Numerous legislatures, ranging from those in Northern Ireland to New York to New South Wales, have engaged with the question of whether to create a distinct offence of domestic abuse/coercive control, and accompanying questions about the scope of new legislation. 1 Scotland features heavily in these ongoing debates, having recently introduced what has been termed the 'gold standard' offence (Scott, 2020: 177) of 'abusive behaviour towards partner or ex-partner' via section 1 of the Domestic Abuse (Scotland) Act 2018. While we are firmly convinced by the arguments for a distinct offence of domestic abuse, in Scotland and elsewhere, the overarching aim of this article is to demonstrate that the Scottish legislation should not be regarded as 'gold standard' in one significant respect: the way in which it seeks to recognise the harms caused to children by coercive control in their living environment. Through our analysis of the Scottish legislation, we will show that reconceptualising children as 'adjoined victims' 2 of intimate partner domestic abuse under the substantive criminal law is symbolically important and recognises far more accurately children's experiences of coercive control than the aggravation model currently in place in Scotland. We will argue that the 2018 Act should be amended to include a new and parallel section 1 offence of 'abusive behaviour towards partner or ex-partner and adjoined child'.
Although the term 'coercive control' does not appear anywhere in the 2018 Act, there is no doubt that Evan Stark's work (Stark, 2007) on coercive control inspired the drafting of the section 1 offence, 3 which seeks to capture ongoing abuse characterised by micro-regulation, manipulation and control tactics. The introduction of a distinct offence of domestic abuse was necessary because existing Scots law offences did not adequately capture this wide range of goal-orientated behaviour and the cumulative and ongoing harms that result from it. The section 1 offence, in contrast to other more episodic Scots law offences, is designed to allow the whole context or 'story' of an abusive relationship between intimate partners to be libelled together under a singular offence, and it is this unique feature that closes a gap in the law and has been said to make the offence 'gold-standard'. Overall, then, the offence represents a direct challenge to the issue of 'fragmentation' that tends to the occur in the criminal law and has been acutely problematic in the context of domestic abuse (Tolmie, 2018: 51-53;Hanna, 2009). Our analysis will show not only that a gap persists when it comes to how Scots law recognises the harms that coercive control causes children, but that the depth and consequences of this gap have been obscured by late amendments to the 2018 Act and misguided promises of future reform of child abuse and neglect legislation. Recognising children as adjoined victims via the introduction of a parallel offence, as we propose, matters not only because it validates children's experiences via the expressive and symbolic power of criminal law, but because it would allow a court to consider the full range and reach of harms caused by coercive control together, and sentence accordingly. The section 1 offence can only tell the whole story, and be truly deserving of its gold standard reputation, if children's experiences are recognised in this way.
Our focus in this article is on the way in which children aged under 18 4 experience intimate partner coercive control in their living environment, and the inadequacies of the Scots law response to this. While there is an ever-growing literature on how children are harmed by coercive control, and an extensive body of legal research exploring the complexities of criminalising coercive control, to date these two strands of academic research have not been pulled together, in Scotland or elsewhere. To our knowledge, this article is therefore the first academic analysis of why and how the substantive criminal law should seek to capture the harms children experience due to coercive control. Although focussed on Scots law, much of our analysis is rooted in international research and will be helpful to other jurisdictions where conversations about the merits of criminalising coercive control, who any new offence should apply to, and how the experiences of children should be recognised, are still ongoing. Indeed, our analysis should send a strong note of caution to jurisdictions which may be considering replicating or 'rolling out' (Bettinson, 2020: 197) the current Scottish model. Our analysis is developed over four main parts. In Part 1, we explore what existing research tells us about children's experiences of domestic abuse (specifically coercive control) and propose that children should be reconceptualised as 'adjoined victims'. In Part 2, we develop our arguments about harm and expose the inadequacies of the 2018 Act, focussing on the section 5 'aggravation in relation to a child'. In Part 3, we argue that attempting to capture the harm caused to children by coercive control in an offence elsewhere in the law (e.g. in child abuse and neglect legislation) would be misguided, and reveal the dangers of a fragmented approach in this context. In Part 4, we set out the details of our proposal to introduce a parallel offence of 'abusive behaviour towards partner or ex-partner and adjoined child' and discuss the advantages and possible challenges of doing so. The article concludes by highlighting wider lessons that can be gleaned from the Scottish story so far.
Part 1: Children and the Harms of Coercive Control: Reconceptualising Children as 'Adjoined Victims' a) Children and Domestic Abuse: What the Research and Data Tells Us Domestic abuse commonly begins or intensifies during pregnancy, and can impact children before they are even born (Cook and Bewley, 2008). After birth and into their childhood, children can experience domestic abuse in a multitude of ways and both actively and passively (Øverlien, 2013: 277), whether through being used as part of the abuse itself, trying to intervene to protect a parent, dealing with the aftermath of a particular incident, calling out for help, or being isolated from friends and family (Scottish Children's Reporter Administration, 2016). The problem is widespread, with the UK charity SafeLives estimating in 2017 that 12,480 children in Scotland were living with the highest risk of domestic abuse (SafeLives, 2017). A small-scale Scottish study in 2014/15 revealed that, in the 12 months prior, 63.7% of adults who reported partner abuse while children were living in the home said that the child was present during the most recent incident (Scottish Government, 2016: 33). 75.3% of these children saw or heard what happened and 18.8% became involved in the incident (Scottish Government, 2016: 39). While these figures are useful in terms of illustrating the scale of the problem, the focus in the Scottish study on 'incidents' of abuse is reflective of a more traditional approach in research in this area, which has tended to centre upon children's experiences of physical domestic abuse within the home, been incident-focussed, and overlooked the equally detrimental impacts of non-physical coercive and controlling behaviour (Katz, 2016).
As challenges have mounted against the notion that domestic abuse is episodic, research on domestic abuse and children has evolved. There is an ever-growing interest in the impact of coercive and controlling behaviours on children (Callaghan et al., 2018;Katz, 2016), coinciding with calls to reject the conceptualisation of children as 'passive' observers who 'witness' or are 'exposed to' domestic abuse (Callaghan et al., 2018(Callaghan et al., : 1555. There remain challenges, however, with measuring the numbers of children living with coercive control. In the UK, this is largely due to a lack of robust official data distinguishing between forms of domestic abuse (e.g. coercive control/intimate terrorism vs situational couple violence/individual incidentson this distinction see Johnson (1995)) and recent changes to legal definitions in this area (McLeod, 2018: 7-8). Although there is a lack of specific data on the prevalence of coercive control, it has been estimated to 'characterise the strategy' used by up to 80% of abusive men (Stark, 2017: 20). In terms of the extent to which coercive control impacts children, Stark and Hester (2019: 96) have suggested that the crossover between control-based domestic violence and child maltreatment should lead us to anticipate that coercive control is experienced by children in a high proportion of cases (see also Jouriles et al., 2008). Moreover, 2017 SafeLives data indicates that emotional abuse 'as a result of abuse' is the most common experience identified by children who have been living with domestic abuse (cited in McLeod, 2018: 10). As emotional or psychological abuse is a common feature of coercive control, this is a strong indicator of its presence. A vast body of literature demonstrates that the effects of domestic abuse on children are diverse and far-reaching, including effects on emotional and psychological well-being, relationships with family, friends and peers, experience of education, and behaviour (including risk-taking and criminality) (Molina and Levell, 2020). Some of these effects are potentially long-lasting: as Callaghan et al., have highlighted, evidence is growing that domestic abuse can have enduring neurological effects on children (2018: 1553).
Although there is, therefore, a wealth of evidence demonstrating the detrimental impact of domestic abuse on children, it is important both to avoid the assumption that children are a monolithic group who are impacted in identical ways, and to acknowledge that there has been some push back in the literature against the narrative that domestic abuse is 'inevitably and permanently damaging' to all children (Mullender et al., 2002: 121). By the same token, the severity and frequency of the abuse that children live with is likely to influence how and to what extent they are impacted and harmed. In their recent examination of the heterogeneity within children's 'exposure experiences', Haselschwerdt et al., found that 'exposure' to coercive and controlling violence, as opposed to situational couple violence, is associated with 'an increased risk of exposure to severe and frequent physical violence and child abuse that may result in serious injury ' (2019: 1534). An additional challenge with assessing the impact of different forms of domestic abuse is that the most vulnerable children (in particular those in shelters) may be the hardest for researchers to gain access to (Øverlien, 2013: 285). In light of this, it has been claimed that 'not distinguishing between the nature of the violence is to render the children most seriously exposed to violence invisible' (Øverlien, 2013: 285). While we are mindful of, and sensitive to, these nuances, our starting position in this article is that all children in living environments with coercive control are either at risk of harm or are harmed to some extent (Stark and Hester, 2019: 96).

b) Children and the Harms of Coercive Control
It is well-established both that coercive control causes distinct cumulative harms such as erosion of self-confidence, agency and freedom (Tadros, 2004(Tadros, -2005 and that traditional criminal law responses have been incapable of appropriately capturing and remedying these harms. What has been overlooked in the legal literature and debate over the criminalisation of coercive control, however, is that children often experience these harms together with their abused parent or caregiver who, given the highly gendered nature of both domestic abuse and caregiving, is usually the mother (83% of Scottish domestic abuse incidents in 2019/20 Scotland had a female victim (Scottish Government, 2021: 14)). In order to understand these interlinked harms, it is crucial to begin with an appreciation of the close link between parenting, or more specifically attacks upon mothering, and coercive control (Heward-Belle, 2017). The concept of 'maternal alienation', which involves perpetrators exerting control through systematic efforts to physically and emotionally isolate children from their mothers (McLeod, 2018: 20-22), is particularly important here. In seeking to disrupt the relationship between mother and child, perpetrators of domestic abuse may use a number of specific tactics, ranging from directly including the child in the abuse (often through manipulative means e.g. bribery) to attacking parenting abilities in front of the child (McLeod, 2018: 20-22). More generally, control of the mother-child relationship, and control of financial resources that mothers require for their children's needs, are extremely common tactics used by perpetrators to undermine and subordinate (Lapierre, 2010). Combined with other tactics of coercive control, these strategies can have a significant impact on the victim's parenting, including the behaviour or actions they take to protect their child/children (Lapierre, 2010). These protective behaviours are highly context specific and may not appear 'reasonable' or 'expected' to professionals or other outside observers (Sidebotham et al., 2016: 81). In some cases where a mother's capacity to protect is limited or reduced by coercive control, abused women can find themselves blamed, or even criminalised, for any harm experienced by their children at the hands of the perpetrator (Singh, 2017(Singh, , 2021 Although maternal alienation is a common tactic, it is not successful or present in all cases. Indeed, in many cases a child will be highly dependent on the non-abusive parent and/or develop a closer relationship with them as way of coping with the abuse (Katz, 2015). Either way, research shows that, even if a child is not the primary target of abuse, the experience of the effects and harms of coercive control is shared. This shared experience has been documented extensively by Katz (2016), whose research has revealed that children in living environments with coercive control directly experience control of time, movement and activities within the home, narrowed space for action (i.e. freedom), isolation from sources of support, and reduced social interaction (e.g. from the perpetrator controlling family attendance at parties or visiting family). Katz's research (2016) has also demonstrated that perpetrator control of how mothers spend time outside of the home has a severe and direct impact on children's social lives, and that the impact of the perpetrator's behaviour on children, even if not directed towards them, is cumulative rather than incident-specific.
The consequences of coercive control on children also mirror those of their parent/ caregiver in the post-separation context, with fear continuing to permeate their every-day lives. A study drawing on interviews with 29 children in the UK and Finland, concluded: 'the omnipresent father occurred in children as a mental and emotional state; a state brought on by children's real experiences of fathers' dangerous and 'admirable' behaviours. The omnipresent father created continual fear in children, and this restricted children's liberty and freedom' (Katz, Nikupeteri and Laitinen, 2020: 310).
The idea of permeating fear, and its severe impacts, is borne out in other empirical studies. Øverlien, for example, has described the fear experienced by children living with patriarchal terrorism as 'embodied' and 'integrated' into their everyday lives (2013: 284). The consequences of living with this fear, and its knock-on effects on freedom of action, can be so severe that they can result in a sense of 'loss or absence of childhood' (Swanston, Bowyer and Vetere, 2014: 197). In the Scottish context, Houghton's work has emphasised the 'shared' experience of coercive control for adults and children (Houghton, 2015: 237), for whom 'treading on eggshells becomes a family trait' (Houghton, 2017: 3, quote from IMPACT project participant). In addition to experiencing the constraint associated with coercive control, children often have conscious awareness of the coercive and controlling behaviour, can articulate how it impacts the family unit, and actively develop strategies to deal with the abuse and try to ensure their safety (Callaghan et al., 2018;Katz, 2016). These strategies can range from adapting behaviour and speech, making changes to social interactions, and avoiding particular spaces at certain times (Callaghan et al., 2018(Callaghan et al., : 1564(Callaghan et al., -1567. The emotional and psychological toll that living in a constant state of fear gives rise to will, for many children (and their non-abusive parent/caregiver), have been more pronounced due to conditions created by the COVID-19 pandemic (Burns, 2021;Pereda and Díaz-Faes, 2020). During this time, children have had fewer social interactions that have a positive psychological impact, and have had reduced access to 'safe spaces' such as schools or nurseries (Scottish Women's Aid, 2020: 5). The pandemic and accompanying measures (including periods of national 'lockdown') have also widened the potential for perpetrators to exercise maternal alienation and control how the non-abusive parent spends time with others, including children, in their living environment (Scottish Women's Aid, 2020: 5; Scottish Women's Aid, 2017b). In families where the perpetrator has spent increased time at home due to home-working and periods of lockdown, there will also have been fewer occasions for other members of the household to spend quality time together, rebel against the perpetrator's wishes, or capitalise on opportunities to exercise 'control in the context of no control' (Stark, 2007: 216) in order to restore a sense of normality and autonomy (Katz, 2016: 54).
The intensity with which the harms of coercive control will have been experienced by mothers and children together during the pandemic has brought an already strong case for recognising children as adjoined victims into sharp focus.

c) Reconceptualising Children as 'Adjoined Victims' of Domestic Abuse
When it comes to capturing children's status as victims terminologically (and then reflecting this in lawsee Part 4), the challenge is adopting a term that simultaneously respects children's autonomy and captures the connectedness of the harms experienced. One of the terms used in the literature is 'secondary victims' which, it has been suggested, is acceptable because the adult victim is the primary and direct target of abuse that would likely be occurring irrespective of the child's existence or presence (Stark and Hester, 2019: 96). We, however, take issue with this term because it risks implying that the harm children experience is secondary and lesser when, as demonstrated earlier, the harms experienced by children can in no way be described as 'secondary'. Similarly, we take the view that the 'direct' adult victim v 'indirect' child victim dichotomy is problematic insofar as the term 'indirect' can be readily associated with passivity, and the notion that children are 'collateral' victims who are 'exposed' to 'adult' coercive control. Moreover, focussing on children's status as 'indirect' victims can be used to downplay and obscure the severe impacts of coercive control on children, and risks obfuscating the responsibility of the perpetrator for the harm caused to the child (Humphreys et al., 2019: 322). Indeed, the traditional dichotomy between direct (adult) and indirect (child) victims of domestic abuse contributes to what Heward-Belle describes as the 'problematic fragmentation' of male abusers into 'binaries of partner and father', and has resulted in a false perception that perpetrators of domestic abuse can be 'poor partners' but 'good fathers' simultaneously (2016: 324).
In short, our position is that distinctions between primary and secondary, and direct and indirect, victims are hierarchical in nature and rooted in traditional and narrow understandings of how children experience coercive control. In contrast, our proposal to reconceptualise children as 'adjoined victims' is driven by the understanding that coercive control gives rise to 'a spectrum of interrelated harms stemming from a single source' (Stark and Hester, 2019: 96) and is a regime that traps children together with their parent or caregiver in an environment that feels like a 'battlefield' (Øverlien, 2013: 283). Compared with other terms, the concept of 'adjoined victims' (and indeed our arguments in general) directly challenges the traditional adult minimisation of the impact of domestic abuse on children (Callaghan et al., 2017) and is in line with calls to make children's agency and perspectives more visible in existing conceptualisations of domestic abuse (Houghton, 2017). In adopting the term 'adjoined victim', we are not implying that a child's victimhood is 'collateral' or an 'add-on', but rather that the harms experienced by the child and adult victim are inextricably connected. While the terms 'distinct victim' or 'victim in their own right' also have merit and appear, at first sight, to better recognise children's autonomy as victims, they risk implying that a child's victim status can arise even if the adult victim did not experience abuse. Adoption of these terms would therefore potentially misrepresent our main arguments, which are, first, that the cumulative harms that result from intimate partner coercive control are 'triadic' as opposed to 'dyadic' (McLeod, 2018: 25) when a child lives in that environment and, second, that these harms, which can be experienced by both adult and child, should not be fragmented in law.
With a solid understanding of how children are harmed by coercive control established, we are now well-positioned to turn to the question of how well the criminal law captures, and is equipped to deal with, the harms that children experience as adjoined victims. While this article is the first to engage with this question in any depth, those who have studied the narratives and experiences of children are clear that the criminal law should recognise children as victims who experience harm together with their parent/ caregiver (Callaghan et al., 2018(Callaghan et al., : 1555Katz, Nikupeteri and Laitinen, 2020: 320). As we will demonstrate in the remainder of this article, if the criminal law does not do this then it not only fails to engage with children's lived experiences (Callaghan et al., 2018(Callaghan et al., : 1555 but, given its symbolic and expressive force, may reinforce their invisibility. This would perpetuate rather than challenge the criminal law's ambivalence to children as legitimate legal subjects (Smart, 1999), and is also counter to the purpose of the criminal law, which should recognise distinctive harms rather than obscure them further. With this in mind, we now move on to demonstrate that there is a disconnect between what various empirical studies tell us about how coercive control harms, and is experienced by, children, and how these harms are dealt with under the Scottish 2018 Act.
Part 2: Children and the Domestic Abuse (Scotland) Act 2018 (a) The 'Gold Standard' 2018 Act As noted in the introduction to this article, a distinct offence of domestic abuse was necessary because existing Scots criminal law offences did not adequately capture the wide range of goal-orientated behaviour used by perpetrators of coercive control, including psychological and emotional abuse often, but not necessarily, intertwined with physical abuse (Stark, 2007). The offence of 'abusive behaviour towards partner or ex-partner' therefore seeks to give statutory recognition to the context in which abuse occurs and to capture the cumulative and ongoing harms of coercive control. The section 1 offence is committed if A (the perpetrator) engages in an abusive course of behaviour (meaning behaviour on at least two occasions: section 10(4)) that a reasonable person would consider to be likely to cause B (the partner/ex-partner) to suffer physical or psychological harm (section 1(2)(a)), defined as including fear, alarm and distress (section 1(3)).
Abusive behaviour is defined in section 2 and includes behaviour directed at B that is 'violent, threatening or intimidating' (section 2(2)(a)) or behaviour directed at B, at a child of B or at another person that either has as its purpose the bringing about of 'relevant' effects on B or that a reasonable person would consider likely to bring about 'relevant' effects on B (section 2(2)(b)). The relevant effects are defined in section 2(3) and include isolating B from friends, relatives or other sources of support; controlling, regulating or monitoring B's day-to-day activities; and frightening, humiliating, degrading or punishing B (sections 2(3)(b),(c) and (e) respectively). All the effects draw on the concept of coercive control and are designed to capture the micro-regulation and power tactics that characterise it. As individual incidents and patterns of violent, threatening or intimidating behaviour could already be prosecuted under Scots law prior to the 2018 Act, it is the criminalisation of effects-based abuse, and the fact that it can be libelled together with violent, threatening or intimidating behaviour, that are the stand-out features of the new offence. Significantly, the mens rea of the section 1 offence is intention or recklessness (section 1(2)(b)), meaning that a perpetrator's lack of intention to cause physical or psychological harm does not preclude prosecution. The combination of these unique features has led to the 2018 Act being praised for its effective translation of the research on coercive control in adult intimate partner relationships into legislation. Indeed, the section 1 offence has been placed on something of a pedestal internationally, with Bettinson claiming 'that the world should be watching how Scotland implements its domestic abuse offence ' (2020: 198).
While children are not currently recognised as victims of the offence in section 1, there are several ways in which the 2018 Act seeks to recognise the effects of domestic abuse on children and strengthen the legal protection available to them. For example, the Act explicitly provides that a non-harassment order imposed in favour of a victim of a domestic abuse offence can also make provision to protect any child usually residing with the victim or perpetrator (or both), or any child named in a section 5 aggravation (section 234AZA Criminal Procedure (Scotland) Act 1995, as inserted by section 12 and schedule of the 2018 Act). Moreover, as mentioned earlier, the Act explicitly states in section 2(2)(b) that behaviour directed at a child can be regarded as abusive behaviour for the purposes of section 1 if it has, or is likely to have, particular effects. Although this recognises that perpetrators can and do involve children in their abuse, the provision only stretches to circumstances where children are used to cause harm to the adult victim (B) and fails to validate the harm that coercive control does to children. Section 5 of the Act, however, goes further and creates an 'aggravation in relation to a child', which, if established, will be recorded upon conviction and taken into account during sentencing (section 5 (7)). While the aggravation attempts to capture a variety of ways in which children can experience domestic abuse, our position is that the aggravation model is fundamentally flawed. Before turning to demonstrate this, analysis of why children were not recognised as victims under the 2018 Act is necessary.

(b) The 2018 Act's Focus on Intimate Partners: Questioning the Exclusion of Children as Victims
As reflected in the title of the offence, section 1 is restricted to partners and ex-partners, which includes spouses, civil partners, cohabitants 'living together as if spouses of each other' and those in an 'intimate personal relationship' (section 11). Even if all the other conditions of the offence are met, other forms of abuse within the home or family, for example, sibling abuse, elder abuse or adolescent to parent violence, are excluded by virtue of the relationship between the parties. The restriction of the offence to intimate partner abuse was discussed by the Justice Committee and was justified on three interconnected grounds: the need to maintain consistency with other Scottish law and policy in this area; the benefits of a narrowly framed offence in terms of ensuring focussed and effective police responses and coordination between different agencies; and the fact that there is a very particular and gendered 'dynamic' to abuse that occurs between intimate partners (Scottish Parliament Justice Committee, 2017: 23). Although there is a difficult question over where exactly the line of relationship applicability should have been drawn (see Cairns, 2017), answering this is beyond the scope of this particular article. It is important to acknowledge, however, that recognising children as victims could have problematised the narrative put forward to justify limiting section 1 to partners and ex-partners. If children had been recognised as victims on the basis that they also directly experience the harms and effects of domestic abuse, then some could have perceived this to undermine the argument that there is a particular dynamic to intimate partner relationships that justifies excluding other relationships, characterised by trust and interdependence, from the scope of the offence. Stark and Hester's recent acknowledgment that studies on children and coercive control, 'lay the foundation for reconceptualising coercive control as a strategy for establishing dominance across a spectrum of relationships that include children' (2019: 98, emphasis added) seems to support this claim.
At first sight, therefore, not recognising children as victims under the section 1 offence appears to be in line with the policy and legislative equation of victimhood with the 'intimate adult dyad' (Callaghan et al., 2018(Callaghan et al., : 1555. However, our view is that a shift away from an exclusive focus on intimate partner abuse, towards one that positions children as adjoined victims (see Part 4), would not require the relationship parameters of the offence to be stretched so far as to capture a broad range of family relationships. There is, in our opinion, an equally strong argument for framing an offence focussed on intimate partner adult abuse as there is for framing an offence that recognises children as victims within the context of that intimate partner abuse. As discussed in Part 1, a range of tactics (including maternal alienation) are used by perpetrators to directly target the relationship between the child and the adult victim, and there is an irrefutable research base demonstrating the profound and potentially long-lasting negative impacts of coercive control on the adult and child together. The scale of this social problem, considered together with children's inherent vulnerability (including to the negative developmental impacts of domestic abuse) and their level of dependence upon -and inseparability from -their caregiver, further justifies the inclusion of children as victims under section 1, even if the focus of the Act remains on intimate-partner abuse. In contrast to, for example, elder abuse (Bows, 2020) or adolescent to parent violence (Bettinson and Quinlan, 2020), coercive control involving children is intrinsically triadic in nature, with the harms experienced by the adult and child/children stemming from a single source.
This position is in line with that taken during the passage of the 2018 Act by groups with extensive knowledge and understanding of how children experience domestic abuse, who took the view that it would be both appropriate and beneficial for children to be recognised under the new legislation as victims. Perhaps the most powerful response to the Bill came from the IMPACT Project, a partnership between Dr Claire Houghton and 8 young adults who experienced domestic abuse as children, which took particular issue with that fact that the section 2 effects would not apply to children (Houghton, 2017). The implication of this, according to their response, is that the perpetrator is not held fully accountable for the harm caused to the child/children and that 'for women with children half the context is missing' (Houghton, 2017: 4). The Scottish Government, however, sidestepped these concerns by arguing that the provisions of the Bill were not readily adaptable for children and that 'psychologically abusive behaviour directed towards a child' would be best addressed in separate legislation, pointing to the planned modernisation of section 12 of the Children and Young Persons (Scotland) Act 1937 (Scottish Government, 2017b: 20-22). In other words, the imminent reform of this child abuse and neglect offence was provided as a reason for not considering the complexities surrounding children's experiences of domestic abuse in any more detail. However, as we demonstrate in Part 3, reform of section 12 raises a distinct set of issues from the question of whether children should be recognised as victims of intimate partner coercive control under the core domestic abuse offence. At the same time as kicking the legislative reform can down the road, the Government also emphasised the value of section 5, claiming that the aggravation strikes the right balance between recognising how domestic abuse harms children and avoiding significant amendment to sections 1 and 2, which were 'crafted with partners and ex-partners in mind' (Scottish Government, 2017a: 11).

(c) The Section 5 Aggravation
Section 5(2) provides that the section 1 offence is aggravated where, at any time in the commission of the offence, the perpetrator directs behaviour at a child or makes use of a child in directing behaviour at the other partner. This would apply, for example, in circumstances when the perpetrator seeks to control or frighten their partner by directing verbal abuse at a child, or when a child is directed to 'spy' on an ex-partner. Under section 5(3), the offence is also aggravated if a child sees, hears or is present during an incident that forms part of the abusive course of behaviour and, under section 5(4), if a reasonable person would consider the abusive course of behaviour (or an incident that forms part of it) to be likely to adversely affect a child usually residing with either the perpetrator or the other partner (or both). Evidence that a child has, or has ever had, awareness or understanding of the perpetrator's behaviour, or has been adversely affected by it, is not required in order for the aggravation to apply (section 5(5)). Overall, the section 5 aggravation is intended to reflect 'the harm that can be caused to children who see or hear abuse taking place or are otherwise present, irrespective of whether the perpetrator directly involves them in the commission of the abuse' (Scottish Government, 2017b: 19). While this is intended to send a strong message that domestic abuse involving children is particularly serious and harmful, the extent to which it properly reflects the harm caused to children can be challenged.

(d) The Problems with the Section 5 Aggravation
In this section, we argue that there are several fundamental issues with the aggravation, and that as a model it serves to perpetuate the notion that children's experiences of coercive control are secondary to, and separate from, the adult victim. Although there is no requirement for harm to be established, the section 5 aggravation is intended to recognise the harm caused to children by domestic abuse. However, we would argue that the logic of an aggravation model is flawed and unprincipled in this context because, instead of positioning the child as a victim of the offence, it risks implying that the harm (or potential harm) experienced by one individual (the child) serves only to 'aggravate' the harm caused to someone else (the adult victim of the offence). Reducing a child's harm to the aggravation model in this way, rather than capturing it through the assignation of victim status under the baseline offence, diminishes their experience and undermines their autonomy. If the harm caused to children by domestic abuse is regarded as sufficiently serious to warrant the introduction of an aggravation which can lead to the perpetrator being sentenced more gravely, then it should, logically, be regarded as sufficiently acute to warrant their recognition as victims under the baseline offence. The section 5 aggravation is atypical when compared with other aggravations in Scots law (see e.g. section 1 of the Abusive Behaviour and Sexual Harm (Scotland) Act 2016; Hate Crime and Public Order (Scotland) Act 2021) insofar as no others are used as a mechanism to recognise harm to another identifiable person in the way that section 5 is. In response to this observation, it could be argued that section 5 serves a broader purpose, marking out domestic abuse involving children as a particularly objectionable societal harm. In our view, however, seeking to recognise a wider societal harm through an aggravation, without explicitly acknowledging the child experiencing the domestic abuse as a victim of the core offence, is an approach which fails to fully respect their personhood.
In response to these arguments, it could be argued that the aggravation offers the additional benefit of formally recording that a child has been involved in the commission of the section 1 offence (and having this reflected on the perpetrator's criminal record). While we would not dispute that this is a benefit, it is of course only a requirement where the aggravation is specified in the libel (and there is prosecutorial discretion as to whether to do this). In addition, we would suggest that the current requirement in section 5(7)(c) that the court 'take the aggravation into account in determining the appropriate sentence' is somewhat ambiguous and means that only those cases that are originally prosecuted with an aggravation may be sentenced more harshly. While this requirement is in line with other aggravations in Scots law, this ambiguity is less defensible when it comes to section 5 because this aggravation relates to another identifiable person. The fact that the requirement to take the aggravation 'into account' is standard for Scots law aggravations also means that it is unlikely that the section 5 aggravation could simply be amended to, for example, make it mandatory for a court to sentence more harshly when the aggravation is established.
In response to our concerns, some might also suggest that the section 5 aggravation could simply be amended to refer to children as 'victims'. However, such an amendment would only be tokenistic and would not fix the problem of children's harms and experiences being positioned outwith the core offence. In addition, it would not solve the fundamental issue that the aggravation model amounts to a legislative separation, or fragmentation, of the harm experienced by child and adult victims. The addition of the aggravation as an 'extra' provision reinforces the idea, which is not only unsupported but contradicted by research, that children are collateral damage whose harm is detachable and secondary to that of the adult victim. Some may point to the fact that the harm would be occurring but for the child (and is not primarily directed towards them) to argue that separation is justifiable, however, such an argument focusses too much on the intent underlying the abuse and not enough on the result, namely that similar harms can be experienced by child and adult victims (see Part 4).
In addition to arbitrarily separating out the intertwined harms experienced by adults and children, yet another issue with the aggravation model is that it does not capture the cumulative harms experienced by children living with domestic abuse. The aggravation's episodic focus was most evident prior to the late addition of section 5(4), which was added as an amendment during the passage of the Bill. Prior to this amendment, the aggravation only covered situations where the perpetrator either directed behaviour at a child, involved a child in directing behaviour at the adult victim or when a child saw, heard or was present during an incident. Although we do not dispute that the aggravation might be useful for one-off incidents (including, for example, if a child witnessed a friend's father assaulting the friend's mother while visiting their house), the implication that harmful behaviour is directed at a specific moment reduces a child's experience to only this moment. When children live in an abusive environment, however, such an episodic focus is clearly at odds with the 2018 Act's objective of capturing coercive control with a cumulative impact. The language of 'seeing', 'hearing' or being 'present during' also continues to position children as witnesses of coercive control rather than as agents within a triadic relationship, perpetuating a narrow conception of coercive control as only occurring within the intimate partner dyad.
Concerns of this nature led to the addition of section 5(4), under which an offence can also be aggravated if an incident or course of conduct might have an adverse effect on the child. This is a deliberately broad provision and, as its application does not depend on a child's witnessing or presence, one that brings the legislation more into line with what research tells us about how children experience domestic abuse (Scottish Women's Aid, 2017c). While section 5(4) at least allows for possible consideration of the various ways in which children experience coercive control, there is no guarantee that consideration will happen in practice, especially as no guidance is provided in the legislation on what factors should be taken into account in making an objective assessment of adverse effect. One might question both what a reasonable person is expected to know about the effects of coercive control on children's well-being and development, and whether the language of 'adverse effect' continues to diminish children's experiences by framing them as 'affected by' abuse between adults (Callaghan et al., 2018(Callaghan et al., : 1566. The vagueness of section 5(4) also stands in stark contrast to section 2, which provides an exhaustive list of the effects that intimate partners may experience, and seeks to capture the cumulative harm caused by living in a state of perpetual fear and readiness. The aggravation fails to encapsulate that children also live with this permeating fear, and suffer the resultant harms in terms of isolation and restriction of freedom. The consequences of this are not just academic. Powerful quotes from IMPACT project participants make clear that the failure to recognise children as victims has been received negatively by young people themselves. While Declan questioned 'so these effects, that I have experienced, do not apply to me?', Lola stated that the Bill's treatment of children made her feel 'like a non-person' (Houghton, 2017: 1). In light of such statements, the demotion of children's experiences to an aggravation model can be seen as part of the broader trend of adult minimisation of children's agency and experiences in the domestic abuse context. While the 2018 Act meets its objective of moving Scots domestic abuse law beyond an incident-focussed model for intimate partners, it further engrains an outdated understanding of domestic abuse into law when it comes to children's experiences. The next two parts of this article turn to demonstrate how this can be remedied.

Part 3: Recognising Children as Victims of Intimate Partner Abuse Elsewhere in the Law a) A Fragmented Approach
Given the problems with the aggravation scheme in place in Scotland, the question arises as to whether children might be recognised as victims of intimate partner coercive control in an offence elsewhere in the law. In the Scottish context, it is likely that this recognition will be sought via reform of section 12 of the Children and Young Persons (Scotland) Act 1937, which has been flagged for reform to inter alia criminalise more effectively the 'emotional' abuse of children (Scottish Government, 2018a). Although this is a broad offence concerned with abuse directed towards children, some of the acts and behaviour stakeholders have suggested for inclusion as criminalised conduct, including 'subjecting a child to witness abusive acts' (Scottish Government, 2019: 26), are tied clearly to the impact of intimate partner abuse in children's homes. This is perhaps unsurprising given that, as discussed earlier, the imminent modernisation of this offence was used by the Scottish Government to sidestep the question of whether children should have been recognised as victims under the 2018 Act. Although the Government's position remains that the aggravation scheme addresses sufficiently the harm to a child in the context of abuse primarily directed at their parent/caregiver (Scottish Government, 2018a: 29), others are likely to be of the view that section 12 provides a further opportunity to attempt to recognise children as victims of intimate partner abuse somewhere.
Our intention in this article is not to engage in detail with the planned reform of the section 12 offence, but rather to question the broader notion that children can be recognised as victims of intimate partner abuse elsewhere in the law, such as in child abuse legislation. In our view, using a separate legislative framework to recognise the harm to the child caused by intimate partner coercive control would be misguided. Such an approach would reincarnate the very fragmentation that the 2018 Act sought to challenge and is out of line with contemporary research which stresses the importance of recognising and managing coercive control as a 'spectrum of interrelated harms stemming from a single source' (Stark and Hester, 2019: 96). In the next section we analyse some of the problems that might be associated with a fragmented approach. Although our focus remains on Scots law, these arguments are of broader relevance to other jurisdictions which might seek to separate out the impact of intimate partner abuse across 'adult domestic abuse' and 'child abuse' frameworks.

b) The Problems with a Fragmented Approach
Use of a separate legislative framework to recognise the harm to a child caused by intimate partner coercive control would continue to exclude children from the core domestic abuse offence. A first issue with this 'fragmented' approach to recognising the impact of coercive control is therefore that it would perpetuate the problematic notion that domestic abuse occurs only within the intimate adult dyad. Subsuming children's experiences within a different legislative framework (and not the core offence) would weaken recognition of the triadic nature of intimate partner abuse and obscure their victimhood in that context. Furthermore, use of a different framework would continue to separate artificially the interrelated harms experienced by adults and children in a way that is likely to conceal the full context of coercive control and minimise children's experiences. Use of a separate provision to capture a child's experience suggests, but also relies upon, their experience being detachable and separate. This harmfully reduces the wrong they have experienced to that of being 'exposed' to adult domestic abuse, rather than recognising that the wrong lies in directly sharing in and experiencing the effects and harms of coercive control. As already discussed in the context of the aggravation scheme, this is out of line with what the research tells us. While an aggravation scheme (despite its flaws) at least exists alongside the core domestic offence in the same legislation, a separate offence, contained in a different legal framework altogether, is severed from the whole context and pattern of abuse, only captures part of the story, and is even more episodic in character. A separate offence, to an even greater degree, lends itself to criminalising incidents where a child's 'witnessing' of, or 'exposure' to, intimate partner abuse can be neatly (but superficially) isolated or compartmentalised, risking obscuration of the direct and ongoing experience, and totality of harm. A consequence of a fragmented approach is that the symbolic importance of recognising children as victims would be diluted or lost entirely.
These problems with a fragmented approach are relevant to any suggestion that the impact of intimate partner coercive control on children could be recognised as 'emotional abuse' in a reformed child abuse and cruelty offence in Scotland (see Scottish Government, 2019: 26). It appears that the section 12 offence is intended to continue (post-reform) as a stand-alone offence (Scottish Government, 2018a: 15) with an actus reus that can, inter alia, be constituted by acts of 'emotional abuse' (Scottish Government, 2018a: 17-19). It does not appear that a framework is being engineered to recognise the cumulative impact of an abusive course of conduct and the combination of tactics involved in the same way the 2018 Act offence was. Thus, in line with the concerns already raised about a fragmented approach, there is a substantial risk that if the section 12 offence tries to capture a child's 'experience' of intimate partner abuse, it will do so in an episodic fashion that does not fully capture the 'embeddedness' (Stark and Hester, 2019: 96) of coercive control. This stands in contradiction to the emerging research which stresses that we should 'expect most child outcomes in abuse cases to take shape gradually, over time, following the unfolding dynamics of coercive control, amid numerous social transactions' (Stark and Hester, 2019: 98). An offence labelling coercive control as 'emotional' or 'psychological' abuse would also be misguided as these terms are an 'inexact proxy' (Stark and Hester, 2019: 97) for coercive control. While acts of emotional or psychological abuse are aspects of coercive control, it is the combination of tactics 'into a pattern of domination that comprises the offense, not the acts themselves' (Stark, 2012: 15; see also Tolmie, 2018: 62). Use of these terms to label children's experience of living in an environment with coercive control is therefore reductive and misrepresents the wrong that they have experienced.
A second significant issue with a fragmented approach to managing coercive control is that it opens the door to possible unintended and problematic consequences in practice for victims of domestic abuse who are parents/caregivers. A consistent message in the research is that a shift is needed to hold perpetrators of domestic abuse fully accountable for the 'cascading harm' (De Simone and Heward-Belle, 2020: 4) they cause, against long-standing concern that both institutional and individual responses often transfer 'blame' to the non-abusing parent (usually mothers) for the harm caused to a child living in an abusive home (De Simone and Heward-Belle, 2020;Singh 2017Singh , 2021Wedlock and Molina, 2020). However, a legislative approach that does not recognise the abused parent/caregiver and child as victims together and simultaneously, but is instead engineered to separate their experiences, is more likely to lead to a decontextualisation of the coercive control. In doing so, it risks obfuscating the responsibility of the perpetrator and mutualising responsibility for the 'cascading harm' which results from domestic abuse, and may create more room for the abused parent to be held responsible, and potentially criminalised, for harm caused to a child living in an abusive home.
Concerns about legislation or accompanying policy 'backfiring' on victims of domestic abuse are not new. Indeed, the risk that coercive control legislation might end up being applied to 'primary victims' was a counterargument to its introduction in the first place (Tolmie, 2018: 60-62). Such concerns arise from experience in other domestic abuse contexts. An area in which unintended consequences for victims are particularly evident is in the application of pro-arrest, pro-charge and pro-prosecution policies, which have led to the re-victimisation of the abused party via criminalisation and dual arrests (Ryan et al., 2022). While avoiding 'backfiring' when it comes to legislation or policy relies on investigators recognising the context and identifying the 'primary perpetrator' in practice, experience suggests that investigators do not always get it right (Royal Commission into Family Violence, 2016: 17-21) and that the decontextualisation of specific incidents from the wider context of an abusive relationship can lead to the criminalisation of domestic abuse victims (Chesney-Lind, 2006). In this particular context, risks are attenuated given that forms of 'systems abuse', including false reports of child maltreatment made by the abuser, can be a common tactic of coercive control itself (Douglas and Fell, 2020). Decontextualisation can also occur in cases where a child suffers harm at the hands of the perpetrator and concerns arise over the adequacy of the mother's efforts to protect her child from violence. Singh, for example, has shown that evidence of past domestic violence paradoxically contributes to the criminalisation of women facing 'failure to protect' proceedings because it supports the prosecution case that harm to the child was foreseeable (Singh, 2017: 514). This process of criminalisation is enabled by gendered expectations which require that mothers be 'omnipresent' (Singh, 2017: 517) and protect their children at all costs, and by the legal system's inattentiveness to the abusive context that informs women's decisions and actions. Context and vulnerability are also likely to be overlooked in the application of broadened definitions of child neglect/abuse, again resulting in the criminalisation of women who are themselves victims (White et al., 2014).
These issues can be illustrated clearly in the Scottish context. At present, the recognition of, and assignation of responsibility for, cascading harm in the 2018 Act through the aggravation scheme is superficial. Though the aggravation could apply where, for example, a parent/caregiver's movements are controlled to the extent that it means they cannot take their child to a medical appointment (Scottish Government, 2017c: 7), the potential or actual harm to the child caused by the perpetrator's control of their parent is not independently and fully validated as the child is not a victim of the baseline offence. If the harm caused to the child through this set of facts can only be recognised fully elsewhere in the substantive law (through, for example, child abuse legislation) then that harm can only exist and be recognised in a vacuum. Through its artificial removal from that broader context of abuse, it becomes an independent harm rather than a cascading and interrelated harm. Because of the criminal law's tendency to obscure vulnerability, it is even possible that the parent /caregiver who did not take the child to a medical appointment could be recast as the author of this independent harm, rather than the perpetrator who controlled their movements. Legislative separation therefore leaves parents/ caregivers who are victims of abuse at the mercy of decision-makers and their sensitivity to context. Alternatively, those concerned with these risks may be left grappling with establishing affirmative defences for parents/caregivers who are victims of abuse (Scottish Government, 2019: 54-60). However, the possibility of a defence may not afford sufficient protection in practice, given problematic interpretative schemas deployed by decision-makers (see Tolmie, 2018: 57-58), the pervasiveness of highly gendered (Lapierre, 2010;Singh, 2017Singh, , 2021 and uninformed (McLeod, 2018: 22) expectations around parenting in the context of domestic abuse, and the general 'credibility discounting' that domestic abuse victims face (see Epstein and Goodman, 2018).

) A Parallel Section One Offence
Our position is that children should be reconceptualised as 'adjoined victims' of intimate partner domestic abuse in their living environment, and that recognition of this should be integral to the core domestic abuse offence provision itself. To give legislative effect to children's status as adjoined victims within the context of intimate partner abuse, we propose that a new parallel section 1 offence should be added to the 2018 Act. This new offence of 'abusive behaviour towards partner or ex-partner and adjoined child' should exist alongside, and as an alternative to, the existing offence of 'abusive behaviour towards partner or ex-partner'. In our view, having two core offences in the 2018 Act is preferable to simply amending the terms of the existing offence to include an adjoined child because, from a fair labelling perspective (Chalmers and Leverick, 2008), it is necessary to be able to distinguish between those cases where there is an adjoined child, and those where there is not. If the child's involvement was brought into the terms of the existing section 1 offence, but the offence retained its shorthand name, then a conviction for 'abusive behaviour towards partner or ex-partner' would not reveal the involvement of the adjoined child, with the consequence that this would not be reflected on the perpetrator's criminal record. By the same token, if the shorthand title of the offence was amended to 'abusive behaviour towards partner or ex-partner and adjoined child' then the ability to differentiate between cases where a child was or was not involved would be lost. It would also be inappropriate for an offender to be convicted of such an offence where there was no child involved. In summary, then, our view is that two alternate offences are required to identify accurately the wrong and harms addressed by each offence and the parties involved. The use of the parallel offence, where the involvement of the child is reflected in the language and shorthand name of the offence itself, would allow for the involvement of a child to be reflected on the perpetrator's criminal record, and communicate clearly the nature of the offence to the public and those working in the justice system (on the communicative role of fair labelling, see Chalmers and Leverick (2008)).
Under our proposal, the 2018 Act would therefore include two alternate core domestic abuse offences. The existing section 1 offence would remain unaltered, and the aggravation could remain an available option for use alongside that because, as alluded to earlier, the aggravation may be an appropriate tool in limited circumstances. The new parallel section 1 offence of 'abusive behaviour towards partner or ex-partner and adjoined child' would then mirror the existing section 1 conditions to be met in respect of the partner / ex-partner. We would suggest that it should also include 'conditions' to be established in relation to the child as part of that offence. These conditions would not require the mens rea and actus reus requirements to be met separately in relation to the child (as this offence cannot be committed against them solely) but would exist instead to situate the child as an adjoined victim within the context of the course of abusive behaviour occurring in their living environment. As a starting point, we would suggest that the conditions to be met in relation to the child would be that they a) usually reside 5 with the adult victim (B) and/or b) are under the care of, or significantly dependent upon, B. Whatever the final precise terms of these conditions, the notions of dependency on, and connectedness to, the adult victim should be key to the drafting in order to capture the interconnectedness of the harms detailed in this article. Although we would stop short of recommending a legislative requirement that the parallel offence be used any time the conditions are met (on the basis that this would infringe upon prosecutorial discretion), we would welcome a legislative presumption in favour of its use, accompanied by a requirement that written reasons be recorded where it is not used.
Going forward, questions might also arise over the effect that our proposal would have on section 2 of the 2018 Act, which defines abusive behaviour. The Scottish Government previously resisted recognising children as victims in the 2018 Act on the basis that the section 2 list of effects of abusive behaviour had been designed with an intimate partner focus and were not readily adaptable for children (Scottish Government, 2017a: 11). In our view, concerns of this nature do not disqualify children from being recognised as adjoined victims in the manner we propose. These arguments would retain force only if one was concerned with establishing a child as a sole victim of the offence, because that would entail being able to read the child (C) directly into the place of the adult victim (B) in the legislation. By contrast, under our proposal, the child would be recognised as an adjoined victim, in acknowledgment that the child is likely to experience some, or all, of the same effects together with the adult victim. In light of this, we would suggest that a part (c) could be added to section 2(4) and provide that 'the effects of the perpetrator's behaviour may also be experienced by an adjoined child but do not require to be established separately'. This would ensure statutory recognition of the harm caused to children by intimate partner abuse without creating a separate burden on the prosecution to establish the effects of the perpetrator's behaviour on the adjoined child.
Some might argue that it is unsound to consider children as adjoined victims under the parallel offence on the basis that the perpetrator may not have intended to cause harm to the child when engaging in the abusive course of behaviour. However, this overlooks the fact that the 2018 Act is primarily concerned with effects rather than perpetrator intent. Including children as adjoined victims, even in circumstances where the perpetrator does not intend for them to be harmed, is consistent with that overarching logic. In addition, any claim that children should not be treated as adjoined victims on the basis that not all children will in fact suffer harm can be rebutted by pointing to the objective nature of the offence in general. The requirements in relation to the adult victim enable the offence to apply if a reasonable person would consider it likely that physical or psychological harm would result from the abusive behaviour. In cases where this is established and a child lives in the same environment, there is a clear and obvious risk, and indeed likelihood, of harm to them (as evidenced in Part 1). Thus, the combination of these factors, along with the magnitude of this social problem, in our view justifies children's inclusion as adjoined victims in a parallel section 1 offence. To put it another way, when one starts from the position that a core purpose of section 1, and indeed other criminal offences, is to mark out and provide redress for distinctive harms, rather than to criminalise abusive intent per se, it becomes perfectly appropriate to recognise children as adjoined victims under the alternative offence.

b) Challenges and Benefits of the Proposed Approach
Although evidence of actual harm to the child would not be required under the parallel offence, it might be suggested that the absence of direct proof of harm in practice may deter prosecutors from using the parallel offence and, more generally, that evidential challenges may undermine the progressive substantive reform that we propose. While it is beyond the scope of this article to explore issues of proof in depth, we would emphasise that our proposed approach does not create challenges or risks beyond what may already be the case in the Scottish context. Under the current framework, notwithstanding the statutory terms, it is possible that an absence of proof of actual harm to the child currently deters prosecutors from using the aggravation, which we have also shown to be unprincipled as a model. Furthermore, evidential concerns currently exist for adult victims due to Scotland's corroboration requirement and the fact that actual harm does not require to be established under section 1 (see Cairns, 2020). For the parallel offence, however, unique challenges arising from the corroboration requirement would be avoided, because the conditions of the parallel offence would not require the mens rea and actus reus of section 1 to be established separately in relation to the child.
Another potential challenge to our proposal to introduce a parallel offence of 'abusive behaviour towards partner or ex-partner and adjoined child' is that it might be perceived as over-complicating the legal landscape in respect of children's experiences of abuse within the home. This is because, given the cross over between domestic abuse and child maltreatment (see Part 1), it might mean, in certain factual circumstances, that a child could be both an adjoined victim under the 2018 Act and the victim of abuse directed towards them under the reformed child abuse offence. While we support the proposals to reform section 12 to ensure that Scotland has a robust child abuse offence, in our view, it is still necessary to recognise children as victims in both legislative frameworks to ensure that the full range of children's experiences in their living environments are captured in the substantive criminal law and in the most suitable place. While strict adherence to distinct 'adult domestic abuse' and 'child abuse' frameworks, focussed on dyadic relationships and interactions, might boast a more straightforward application, this is precisely because it rests upon an over-simplification that either fails to account for the triadic nature of intimate partner abuse or, as established in Part 3, would unhelpfully fragment it. Our proposal is intended as an additional, yet significant, substantive criminal law response, recognising that a child can be properly considered as a victim of an abusive course of conduct in their living environment even where it might be viewed as primarily or only directed at their parent/caregiver. While we accept that it might be inappropriate on the facts of specific cases to use both available options together -for example, if there would be a risk of 'double punishment' in the circumstances (Scottish Law Commission, 2009: 2) -this raises an issue of appropriate framework selection for the prosecutor, rather than militating against having both options available.
Although mindful of the potential challenges, we consider that the parallel offence would be of huge symbolic significance and have several benefits compared with the current approach. Recognising children as adjoined victims would validate their experiences of harm and assign them proper legal subjecthood. The need for this should be considered against the backdrop of the criminal law's traditional resistance to more expansive conceptions of harm when it comes to children and the resulting struggle of feminist campaigners to challenge 'the power of legal discourse to impose its internally generated set of meanings and understandings of the child' and harm (Smart, 1999: 394). This struggle is still ongoing in the context of domestic abuse, where law and policy continue to marginalise children's experiences. The direct and negative effect of this marginalisation is reflected in consultation submissions made during the passage of the 2018 Act. The IMPACT project in particular highlighted the importance of children being recognised and named as victims of domestic abuse, with proper recognition of children's experience being tied to their 'sense of worth', and their status as 'people with human rights' (Houghton, 2017: 1, 6-7). Our hope is that the explicit inclusion of children within a new parallel offence would speak directly to children who have experienced domestic abuse by clearly acknowledging that they can and do experience the distinct and cumulative harms of intimate partner coercive control in their living environment, and that they are victims of it. As the IMPACT submission emphasised, a major barrier for children has been that they have not traditionally had, but need, 'the language to name domestic abuse and to name it as something happening to them' (Houghton, 2017: 5, emphasis added). By explicitly including children as victims of domestic abuse via our proposed criminal offence, we also hope that children would be further empowered to speak out about domestic abuse in their living environments.
In our view, the introduction of the parallel offence would contribute to changing the way that society understands and responds to domestic abuse. This is because a parallel offence of 'abusive behaviour towards partner or ex-partner and adjoined child' would directly challenge public perceptions of domestic abuse as 'between adults' (Houghton, 2017: 5-6). While the adult-centric offence currently in place, and the accompanying aggravation, risks reinforcing children's invisibility and minimising their involvement, we would hope that children would become much more visible if our proposal was adopted. Given the expressive power of the criminal law, the reconceptualisation of children as adjoined victims of domestic abuse could also have a positive ripple effect across other legal contexts, by sending a clear message that children experience domestic abuse directly. It could, for example, have a material impact in Scottish child contact cases, where the consequences of domestic abuse being viewed as an 'adult concern', and the resulting failure to recognise how 'omnipresent' it is in children's lives, have been identified as particularly problematic (Morrison et al., 2020: 410; see also Morrison, 2015;Scottish Women's Aid, 2017a).
Furthermore, moving away from a strictly adult-centric offence provision to one that situates the child within it would provide important symbolic recognition of intimate partner coercive control as involving a shared experience of interlinked harms caused by a single perpetrator. Recognising children and their caregivers as victims together and simultaneously of an offence committed against them by the abuser would, in our view, send a clear message that perpetrators of intimate partner coercive control are responsible and blameworthy for the 'cascading harm' that they cause to children living in that environment. In comparison to the current approach, our proposed approach would be conducive to facilitating a societal shift towards holding perpetrators to account fully, and would prevent the transfer of blame to the non-abusing parent. Given the highly gendered nature of caregiving, and the centrality of attacks on mothering and the mother/ child relationship to the exercise of coercive control, in our view the parallel offence would further enhance, rather than compromise, the gendered understanding of domestic abuse which underpins the 2018 Act (Burman and Brooks-Hay, 2018).
At a practical level, the parallel offence would allow for a holistic approach to managing intimate partner coercive control where children are involved. In Part 2, we discussed the limitations of the current aggravation scheme in terms of sentencing, namely that it only requires a child's experience to be 'taken into account' when the prosecutor includes the aggravation on the original libel (which is, of course, not a requirement). By contrast, our proposed approach would afford the court the opportunity to hear the whole triadic story together and to consider the full range and reach of harms caused by coercive control together, and sentence accordingly. We anticipate that adoption of our proposal to introduce a parallel offence would have a more significant impact upon sentencing. By positioning the child as a victim of the core offence, the likelihood of a child's involvement being taken into account during sentencing is significantly greater because their full experience can be considered, as opposed to just the parts that have been established through the aggravation. Arguably, the prosecutorial incentive to recognise children's experiences, and to include these in the complaint or libel in the first place, would be greater with our proposal than with the current aggravation scheme.
Although the primary aim of this article is to establish the symbolic importance of recognising children as adjoined victims, it is also important to mention that there would likely be a range of further practical benefits which would flow from recognising children as victims under the 2018 Act as, in general, 'victim' status triggers a range of state obligations in terms of strategy and support (Scottish Government, 2018b). As was pointed out in a recent Victims' Commissioner report in England and Wales, an important reason for repositioning children as 'victims' of domestic abuse is that it leads to more targeted interventions and support for children, including better funding and access to victim entitlements and services (Wedlock and Molina, 2020: 7, 26-27). Although this can be achieved legislatively without making children 'victims' of a criminal offence of intimate partner coercive control, 6 in our view it is crucial that improved support for children stems from their inclusion as victims of the core domestic abuse offence, rather than from other legislative workarounds. This matters because of the symbolic importance of children's experiences being fully validated under the criminal law. It is crucial to ensure that perpetrators are held accountable for the effects their behaviour has on a child, through enabling criminal courts to consider the full range and reach of harms caused by coercive control.

Conclusion
As stated at the outset of this piece, this article is, to our knowledge, the first academic analysis of why and how the substantive criminal law should seek to reflect and validate children's experiences of intimate-partner coercive control. While the focus of this article has been on Scots law, its implications are further reaching. It contributes to broader discussions about criminalising coercive control and the scope of such offences, both in principle and in relation to specific legislative reform in other jurisdictions. Indeed, a core purpose of this article is to sound a note of caution against simply 'rolling out' the Scottish legislative approach elsewhere, because, despite its many strengths, it remains deeply flawed in its treatment of children. We therefore conclude by highlighting some key lessons that can be learnt from the Scottish story so far, and which should be considered by other jurisdictions.
The first is that, when seeking to introduce legislation to tackle intimate partner coercive control, discussions about its scope should start from a position whereby children living in that environment are recognised as victims who are trapped in a regime together with their parent/caregiver and who share in and experience directly the distinct and cumulative harms. As we argued in Part 1, reconceptualising children as adjoined victims within the context of intimate partner abuse is consistent with children's lived experience (as demonstrated by research) and the role of the criminal law in marking out distinctive harms. Given the expressive power of the criminal law, starting from this position is important in facilitating a societal shift away from the notion that coercive control is simply between adults.
A second lesson is that children's status as victims should not be captured via an aggravation scheme (or equivalent), because this problematically positions children as 'collateral damage' outside of the direct context of the core offence. Indeed, for those looking to the Scottish legislation as a 'model', the origins of the aggravation scheme, as well as its flaws, should be given due consideration. We would suggest that the aggravation is best viewed as something of a consolation prize, awarded during the consultation and legislative process due to stakeholder concern that children were not recognised as victims under the section 1 offence. The aggravation should, therefore, be properly viewed as a compromise that resulted from treating children as an afterthought, rather than as a model for validating children's experiences that other jurisdictions should aspire to and replicate. The third, and related, lesson is that legislation criminalising coercive control is the most appropriate place to recognise and acknowledge children as victims of intimate partner abuse. As we have illustrated, attempting to do this elsewhere in the substantive law will result in a fragmented approach that not only risks obscuring the full context of coercive control and minimising children's experiences, but potentially opens the door to unintended and problematic consequences for adult victims of domestic abuse.
We acknowledge that recognising children as adjoined victims of coercive control in the Scottish legislation (and indeed in other jurisdictions) would not be without challenges. However, just as drafting and evidential challenges did not preclude the introduction of the 2018 Act in Scotland in the first place, they should not preclude returning to this legislation to introduce a parallel section one offence of 'abusive behaviour towards partner or ex-partner and adjoined child'. Recognising children as adjoined victims via the introduction of a parallel offence, as we propose, matters not only because it validates children's experiences via the expressive and symbolic power of criminal law but because it would allow a court to consider the full range and reach of harms caused by coercive control together, and sentence accordingly. The section 1 offence can only tell the whole and triadic story, and be truly deserving of its gold standard reputation, if children's experiences are recognised in this way.