Criminology, Gender and Risk: The Dilemmas of Northern Theorising for Southern Responses to Intimate Partner Violence

Criminology’s unitary, unifying and gendered embrace of risk is rooted in Northern theorising. This understanding of risk not only takes its toll on the discipline; it also has consequences for the practices associated with risk: risk assessment. Such practices reflect a range of different assumptions that silence women’s everyday experiences of violence, silence culture, and contribute to the construction of all women as fearing and vulnerable subjects. In particular the policies and practices of risk and risk assessment as responses to violence against women continue to travel across the globe with scant regard for the shaky foundations on which they are based and, by implication, their relevance for other settings. This paper explores the nature of the shaky conceptual foundations of risk assessment to reflect on the problems and possibilities for more locally nuanced and culturally sensitive responses to violence against women demanded by the agenda of Southern criminology.


Introduction
There is now a well-established literature pointing to criminology's unitary and unifying embrace of risk (O'Malley 2004).It has also been established that this embrace is a gendered one (Chan and Rigakos 2002;Walklate 1997; see also Hannah-Moffat and O'Malley 2007).This gendered understanding of risk constructs women as risk avoiders rather than risk seekers and embeds a range of different assumptions in both criminology and victimology.Such assumptions offer a comfortable and comforting vision of women as uniform, fearful, vulnerable subjects (Walklate forthcoming).Moreover, they smooth out and silence the everydayness of women's experiences of violence (Kelly 2011;Shalhoub-Kevorkian 2016a), and their situated experiences of culture (Machado, Dias, and Coehlo 2010).Arguably, all of these assumptions are rooted within Northern theoretical presumptions about the nature of fear, risk, vulnerability and gender and how these concepts might be related to one another.Similar assumptions also inform policy, particularly the policies and practices of risk assessment.These policies and practices, especially in relation to intimate partner violence, have travelled the globe with scant regard for the shaky conceptual and gendered Northern foundations on which they are based (see, inter alia, Walklate and Fitz-Gibbon 2018;Goodmark 2015).Moreover, as policies they simultaneously reflect a tendency to erase important differences in legal codes.At the same time, these interventions and practices notwithstanding, rates of femicide as a result of intimate partner violence have remained stubbornly consistent across the globe (see, for example, Cussen and Bryant 2015 in Australia; Smith et al. 2014 in the United States (US)).Taken together, these observations raise a number of questions.Of particular concern for this paper is the relevance of risk assessment practices generally for cases of intimate partner violence but particularly their application in cultural contexts in which they were not generated.This concern arises for at least two reasons.First, there is confusion around what these practices actually do: are they about prediction or prevention?Second, and perhaps more fundamental for the purposes of this discussion, they emanate from assumptions rooted in Northern theorising, the relevance of which, in other parts of the globe and for those whose voices are not reflected in such theorising, is open to critical scrutiny.This paper will explore the shaky conceptual foundations of risk and risk assessment, reflect on the problems generated by these foundations, and examine the possibilities for more locally nuanced and culturally sensitive understandings of risk that might better inform such practices in relation to intimate partner violence.
The paper falls into four parts.The first offers a brief overview of the concept of risk and its deployment within criminology with a view to highlighting some of its conceptual blind spots.The second part considers the possibilities of thinking differently about risk informed by the challenge posed for Northern theorising in the work of Connell (2007) and others.This challenge requires recognising the imperialism endemic in criminology's liberal nomothetic impulse (Morrison 2015;Young 2011) and the presence of that imperialist impulse in risk assessment practices.The third part discusses the transference of the lacunae of Northern theorising into the practices of risk assessment and the consequences this has for responses to intimate partner violence.The conclusion suggests that the colonising assumptions of risk entrenched within risk assessment tools constitute a gendered 'occupation of the senses' (Shalhoub-Kevorkian 2016b) and are thus doomed to failure for those whose lives are othered as a result of both gender and settler colonialism.Mythen (2014) has pointed to the general limitations associated with the social science embrace of risk.He highlights three in particular: the limited visibility of power and power relations within risk theory; the partial view of human agency embedded within it; and the tendency towards catachresis (misapplying or overstretching the use to which the concept of risk has been put).All of these limitations can be found within the criminological and victimological embrace of risk.Moreover, whilst these areas of investigation have centred concerns about risk, what this actually means is often poorly articulated and is frequently partial and eclectic.As O'Malley ( 2004)

Criminology and risk
observed, risk is structured: it is neither uniform nor unifying.Who is deemed at risk, and who is deemed risky, is a multi-facetted phenomenon, mediated, for example, by global geo-political positions on the one hand (Aas 2012) and locality on the other (Evans, Fraser and Walklate 1996).
It is also gendered and the implicit acceptance of risk as a gender neutral concept (within criminology and victimology) not only hides the risk seeking behaviour of women but also hides the vulnerabilities of men (Walklate 1997).Thus, as Chan and Rigakos (2002: 756) state: A recognition of risk as gendered relies on acknowledging that there can be no essential notion of risk; that risk is variable; risk itself is more than one type.… Risk is gendered on a continuum both in the sense of empirical potential harm and the recognition and the definition of that harm.Women, it may be argued, are required to engage in instrumental risk in order to interact socially, work, cohabitate with a man etc.However, this does not signal women's victimhood but rather their agency in flouting potential dangers in the general pursuit of material subsistence.
For example, Sanders' (2005) articulation of a 'continuum of risk' reflects a highly subtle and active process of making choices about where to work and how to work, alongside managing questions of emotions, identity, health and relationships, all of which are entailed in the material life of being a prostitute.Thus, whilst 'risk burns in many different degrees' (Walklate and Mythen 2008: 215), 'risk is not about modernity and the ontological insecurity people experience: for women it is about misogyny and the continued perpetuation of women's oppression through fear of crime and blame for their situation' (Stanko 1997: 492).Stanko's (1997) assertion of the ongoing powerful influence of misogyny notwithstanding, it is important to note that the criminological and victimological embrace of risk reflects a categorical denial of women's agency, presumes their inherent vulnerability, is deeply gendered and strains towards uniform understandings in order to construct a unified individualised liberal (and fearful) subject (Walklate forthcoming).This conceptual blind spot, expressed by Mythen (2014) as a partial view of agency, also translates into the policies and practices associated with risk, particularly those of risk assessment.This is of especial importance for the discussion here.However, first of all it will be of value to consider the global resonance of this conceptual vision.

Northern theorising and risk
The dominance of Northern theorising for Southern (criminological) agendas in the context of criminology is being subjected to ever increasing critical scrutiny (see, inter alia, Carrington and Hogg 2017;Carrington et al. 2016).Much of this scrutiny has been informed by the work of Connell (2007), amongst others.Her initial intervention focused attention on the ways in which theoretical assumptions about the nature and impact of globalisation formulated under the rubric of Northern theorising led to the conclusion that such processes take their toll across the globe in the same way and to the same extent.The consequence of this is that Westo-centric bias has become reified, with the 'systematic violence of the metropole' (Connell 2007: 378) being overlooked.Embedded in this theorising is an assumed linear progression from a pre-modern, to an industrial, to a world risk society with three resultant effects: other voices and visions of social change are excluded; non-metropolitan experiences are erased; and the gathering of data from the 'periphery' becomes framed and informed by Northern concepts and methods (Connell 2007: 380).De Sousa Santo (2014: 56) makes similar observations.Drawing on the values on which the Nuestra America operated as a way of 'living in transit and transitorianess' long before the 'invention of the "risk society" [Beck 1992]' as his example, he makes a compelling case for a 'sociology of absences'.This demands both a democratic and an epistemological imagination.As has been argued elsewhere, these principles pose some very fundamental questions for criminology generally expressed succinctly by Aas (2012: 14) in the following way: The global does therefore not present itself as a smooth, unified surface, a plane of immanence accessible through a zoom function, but rather as a dynamic multiplicity of surfaces and tectonic boundaries.It is in these meeting points and frictions between the global north and south, between licit and illicit worlds, that criminology has an opportunity to gain (and provide other social sciences with) invaluable insight into the nature of the contemporary world order.
Given the ongoing governmental temptation for policies to travel (particularly in relation to intimate partner violence) and for that direction of travel to be from north to south (see Walklate and Fitz-Gibbon 2018), it is imperative that criminology and victimology address the conceptual frictions between the global north and the global south.As Carrington and Hogg (2017) point out, whilst the distinction between north and south is blurred and contested, it is, nevertheless, the case that the impact of an uncritical embrace of Northern theorising across the globe values some ways of knowing and devalues others.This is evident in a number of different ways but no more so than in the way in which policies and practices relating to violence against women have travelled the globe embedding particular assumptions about 'women' and 'risk'.
Thus, the emergence of 'Southern criminology' compels both criminology and victimology to cast aside their current love affair with risk since this not only entraps both areas of investigation within the risk of politics and the politics of risk (Mythen and Walklate 2008) but also deflects attention from the deep-rooted problems with risk and risk assessment practices themselves.These deep-rooted problems range from the conflation of prevention with prediction, to the transference of such practices from offenders to victims, to the conceptual failure inherent in these practices themselves.As these practices travel the globe, such deep-rooted problems become further embedded in policies and practices as governments and policy makers strive to be seen to be doing something about intimate partner violence, sometimes in the absence of validated evidence that such practices do anything at all.

Risk, risk assessment and intimate partner violence
Some time ago, O'Malley (2006: 49) pointed out that 'crime prevention has succeeded in marrying risk with a more traditional social and behavioural form of criminology by translating the old causes of crime into risk factors'.This pre-occupation with risk factors is evident from the local to the global.For example, following the lead of the World Health Organization (WHO), various countries have embedded the 'ecological model' of violence into national violence prevention strategies.
The assumption underpinning this model and its 'risk factors' is that violence can be prevented by reducing the violent characteristics of individuals.Thus individuals are 'sorted' (Feeley and Simon 1994) according to their respective risk factors for violence.Note at this juncture the emphasis on prevention within this ecological model.A wide range of risk assessment tools has developed in the wake of this model and its associated assumptions about risk, tools which have become a constituent part of managing crime (Garland 2001).Yet the development of these tools is marked by a conflation of prevention with prediction.Indeed, it is unclear in the slippage from one to the other what the purpose of particular tools might be (Day et al. 2014).Moreover this slippage takes a particular toll in the increasingly widespread use of such tools for victims at risk of intimate partner violence as developed below.
In general terms, risk assessment tools, whether clinical or actuarial, focus attention on predicting future behaviour, the assumptions of the WHO model notwithstanding.Logically, of course, such tools can never provide anything more than hypotheses concerning what might happen in the future.Indeed, their capacity for weak and/or modest prediction has been reported by Medina et al. (2016).However, such weak findings stand in precise opposition to the purpose of the ethos of risk management which is to minimise, in Bernstein's (1996: 334) words, 'when wildness will break out'.Thus the endeavour of risk management, designed to deny the possibility of 'wildness', simultaneously denies human agency (qua Mythen 2014).This denial is captured in the conflation of prevention with prediction and can be traced in two ways: what it is that the professionals actually 'do' when assessing risk; and what it is those deemed 'at risk' might do in the light of having been assessed as risky.These issues become particularly problematic when used to inform risk practices in cases of intimate partner violence and there are (at least) four reasons for suggesting this.The first three, when taken together, might be termed illustrative of implementation failure in the process of assessing risk.However the fourth, in relation to the subject matter of this paper, is arguably the most profound.This raises the spectre of conceptual failure (Lewis and Greene 1978) insofar as the tools themselves both misapply risk and misunderstand the nature of the risks being faced: in Mythen's (2014) terms, they suffer from catachresis.Each of these problems will be discussed in turn; first, the problems associated with implementation failure.These return us to the difficulties inherent in identifying and measuring 'risk factors' for those subjected to intimate partner violence.Westmarland (2011: 300-301) observing that the relationship between such risk assessments and subsequent femicide is somewhat arbitrary and potentially problematic (see Day et al. 2014 andJohnson et al. 2017).Nevertheless, there seems to be some agreement on the 'risk factors' for interpersonal homicide in the context of intimate partner violence.These are: prior interpersonal violence; age difference; cohabiting; estrangement; and the presence of a child not biologically related to the abuser.Other factors include homes where there is mental illness, drug abuse and the presence of weapons (Campbell et al. 2009).

Risk, risk assessment and implementation failure
Risk assessment tools incorporate these kinds of risk factors and thereby implicitly accept that risk can be measured, causes and offenders identified, and those deemed risky (whether potential offenders or potential victims) subjected to surveillance and managed.It is a moot point as to whether or not such management practices imply prediction or prevention.However, importantly, what has become embedded here is remarkably similar to what has become entrenched within risk theorising (Mythen 2014); that which de Sousa Santos (2014) might refer to as the ghostly relationship between theory and practice.Thus, in this context, not only is risk treated as a forensic concept, it is also treated as uniform and unifying (see also Robinson and Rowlands 2009)  Experiential knowledge can make its presence felt at a number of points in the process of risk assessment (see Walklate and Mythen 2011).Two points of intervention are worthy of note here: how professionals actually do the work of risk assessment; and the ways in which knowledge available from those deemed 'at risk' is, or is not, incorporated into the level of risk judged to be present.In relation to the first of these, Ansbro (2010), Broadhurst et al. (2010), Kemshall (2010) and Robinson (2010), in their different ways, cast some light on the processes of 'doing' risk assessment.This work draws attention to the ways in which professionals draw on all kinds of 'knowledge' to accomplish this practical task: some of it expert knowledge; some of it derived from 'risk factors'; and some of it from what they 'know' from working in the field.Thus, as Walklate and Mythen (2011: 109) suggest, it is possible that '… risk, and those deemed at risk, are not forensically measured at all: they are constructed within a logic of norms and values that are felt'.How this happens is still somewhat a 'black box' (Robinson 2010) but ultimately it seems fair to suggest that professionals do not rely on their knowledge of 'risk factors' alone: they make judgements based on their feelings and intuition, in the context of the case with which they are dealing.Indeed, Robinson et al. (2016) report that a small constellation of evidence-based risk factors seems to inform police officers' perceptions in the UK and the US of the level of risk present in cases of intimate partner violence, which veer towards evidence of physical violence rather than anything else that might be reported to them.
The risk assessment tools which endeavour to incorporate the victim's voice into the decisionmaking process is a further area of potential implementation failure.Women's experiential knowledge of when the next act of violence is likely to occur is open to what it is they 'know'.Indeed as Smith et al. (2010: 27) intimate, this form of knowing may be deeply embedded in their strategies for coping with 'battering' and 'include[s] family history of abuse, gender role socialisation, the attitude toward violence of the immediate and extended social network, and various characteristics of the abuse and abusive partner'.There is much within the feminist informed literature in support of this (see, inter alia, Kirkwood 1993), illustrated some time ago by Genn (1988) in her analysis of a woman's testimony concerning her routine, day-to-day experience of violence in her life in 'Bleak House'.In addition Day et al. (2014: 581)  There is some evidence to suggest that partner estimates of risk can also consistently predict future victimization, with approximately two thirds of victims correctly identifying their assessed level of risk.
However, the extent to which women's voices are actually heard in the process of risk assessment is moot.Recognition of this as an arena for silencing raises a number of fundamental questions.
For example, what counts as legitimate knowledge in the risk assessment process, what does that knowledge counts for, who can have it, and how is it applied?There are multiple examples of cases in which the failure to listen to women's own voices has resulted in their subsequent demise, from the Thurman Case in the US in the early 1980s that led to the introduction of mandatory arrest policies in cases of domestic violence, to that of Kelly  2014,2015).It is at this juncture that the fundamental flaw of conceptual failure endemic in risk assessment tools as applied in this arena of victimisation comes to the fore.This conceptual failure draws attention to how the concept of risk itself is understood.

Risk, risk assessment and conceptual failure
The question of conceptual failure has recently been subjected to critical examination in Canada.
The case of Ewert v Canada 2 (Hart 2016) exposes some of the problems in assuming that risk assessment tools, and their inherent understanding of risk, are valid across different knowledge systems.The questions posed by this case are explored in substantial detail by Hart (2016).He points out that, whilst this particular case has a narrow focus, the implications of it are broad.The possibility that risk assessments suffer from inherent bias has implications for anyone working with such tools (not just in the prison system or just with offenders), but is particularly relevant to indigenous peoples across the globe who are subjected to assessments presumed to be culturally neutral.For Hart (2016: 90), this raises questions about the use and deployment of all such tools claiming to operationalise 'risk factors'.Indeed, this point is developed by Cunneen and Tauri (2016) who argue that treating the patterns of crime and victimisation within indigenous communities as separate and separable from their specific historical and social context, as the focus on risk factors derived from the 'risk paradigm' (Cunneen and Tauri 2016: 159) does, renders indigenous people's rights secondary to their problematic status as a risk group (see also Blagg 2016).Indeed, Price, Langton and Cashman (2016) add some weight to this by pointing to the problems posed for Indigenous women in Australia whose experiences of violence were minimised in their view, because of a hegemonic cultural reluctance to challenge the violence(s) of Indigenous Australian men.For these commentators, this reluctance afforded the space in which many such violence(s) could continue, and were thereby legitimated.Challenging such hegemonic cultural presumptions (about indigenous communities and how they work) would mean addressing the problem of settler colonialism.
Whilst the Ewert case was appealed (an appeal which was allowed and subsequently found in favour of the appellant), the questions it raises remain pertinent to the issues under discussion here.Indeed Hart (2016) goes on to suggest different ways in which professionals might be enabled to move on from the implications of this case.For the purposes of this paper, the Ewert vs Canada case offers some insight into the profound implications that conceptual failure has when risk is deployed in a totalising as well as in a uniform and unifying fashion, as seems to be IJCJ&SD 8 Online version via www.crimejusticejournal.com © 2018 7(1) the case with risk assessment tools.In relation to intimate partner violence, such totalising presumptions carry particular consequences and it will be of value to explore these a little further.

Thinking differently about risk and risk assessment
In 2003, Shelhoub-Kevorkian asked the following questions about the ways in which femicide is understood and responded to: What is the alternative if her male adult 'protector' abuses here (sexually, emotionally, physically), and how can she speak about her abuse if she has never learned that it is possible to voice personal matters?How can she speak out when she knows that customs and cultural codes may be used to cause her death?How can she ask for help when her protectors might also be her enemies?How can she trust her family when their first reactions may be to kill her?Where can she go for help when the informal agents of social control tend to blame her and question her acts?How can she ask the help of the legal system when most agents of social control are men?What happens if the legal system supports her femicide?(Shelhoub-Kevorkian 2003: 603) We might add, following on from the observations of Cunneen and Tauri (2016) and Hart ( 2016), what would risk and risk assessment actually mean under these kinds of cultural-structural conditions?Shelhoub-Kevorkian ( 2003) situates her understanding of women's lives as rendered invisible by a series of veils-culture, law and political necessity-all of which are present under the conditions of settler colonialism (Shelhoub-Kevorkian 2016a) in which fear, security and risk take on very particular meanings for the everyday nature of women's lives, from the clothes they wear to their birthing practices (see also Listerborn 2015).This focus on the everydayness of life under settler colonialism is taken further in her analysis of the 'occupation of the senses' (Shelhoub-Kevorkian 2016b: 18) in which: The colonial regime works to inculcate a sense of control among the colonisers, while instilling discipline and obedience among the colonized … that ultimately aims to render the colonized senseless.
Importantly, in making these observations, she is hinting at the experiential connections between women's lives under settler colonialism in Israel with the indigenous colonialised lives of those elsewhere in the world, as well as women's lives more generally.As Pain (2012: 6) reminds us, '[k]eeping another person in a state of chronic fear does not require physical violence to be used all of the time, or at all'.Intimate knowledge of another person is sufficient.This, too, is an 'occupation of the senses', in which what is risky to do or not do occupies a woman's sense of her everyday life, albeit in different social contexts.Making these connections points to two further issues.
First, the everyday experience of security, fear and risk need to be appreciated as they are actually lived and experienced by women not as they might be imagined by the rationale of risk assessment tools.This experience is messy.For example, as Gill et al. (2017: 4) report in the context of the UK: Police need to understand that a visit by them can be seen as bringing 'trouble' to the family in a very public way-potentially invoking shame in the eyes of the community-and that this shaming will often have ramifications for the victim.
Indeed, understanding and formulating a (policing) response to these difficulties simply in terms of 'culture' misses the mark, particularly from the victim's point of view (Gill and Harrison 2016).Furthermore, as Medina et al. (2016: 2)  … officers and citizens involved in these interactions are often encountering each other from very different gender, ethnic, and professional vantage points.An endless combination of misunderstandings, judgement errors, and procedural mistakes can occur in the policing of domestic abuse at the frontline.
Indeed, such interactions more often than not occur at cross purposes.What women want from such interactions (perhaps to feel safe for a while) may be quite different to the 'risk-crazed' (Carlen 2008) directives of the police officer.More often than not these tangled interactions also occur in highly emotional circumstances for all parties in which feelings matter.They matter particularly in the context of intimate partner violence (Kuennan 2014).Again this is messy.Moreover, this messiness raises the importance of understanding context, whether that is a context framed by settler colonialism or one framed by more subtle but no less profound presumptions of normative heterosexuality and/or the role of violence in a particular relationship.Indeed, it is important to remember, of course, that violence can and does co-exist with all kinds of other features of that relationship, including love.Importantly an appreciation of this kind of messiness leads to the second issue here alluded to in the work of Shelhoub-Kevorkian (2003).This demands a different way of thinking about risk.
Cunneen and Rowe (2015: 27) express this issue in the following way: [Second] there is a need for a much deeper understanding of Indigenous ontologies and the way in which the 'self' is understood in connectivity to the social, physical and spiritual world.The centrality of interrelationality to Indigenous worldviews means that the understandings of particular situations and contexts, and the decisions which people make, are formed from within a worldview that is in strong contrast to colonising assumptions regarding individual decision making based on autonomous self-interest.
The questions posed here require more than a greater sensitivity to difference, and/or modifying existing (criminal justice) systems and practices to take account of the complex relationship indigenous people might have with such systems as implied by Marchetti and Daly (2016) with reference to Indigenous Australians.Blagg et al. (2017: 352) go so far as to suggest 'a pluralist alternative where settler law increasingly secedes sovereign power to Indigenous law and culture'.A similar principle might also apply to risk assessment tools in which such 'tools' might embrace an understanding of community, inter-relationships, power and shame in better informed ways (see Gill and Harrison 2016).(The questions posed here certainly require more than 'simply' inserting women's voices into the process and/or improved police training, though there is without doubt much that could be achieved on each of these fronts.).Fundamentally, however, these questions are epistemological: what kind of knowledge (about risk) counts and who can speak and be heard on that knowledge (about risk).
To summarise: the unitary and unifying embrace of risk reflects a partial sense of agency and stretches the practices of risk assessment into a wide range of domains, some of which may be inappropriate.In relation to risk assessment tools, this embrace is as likely to result in false negatives as well as false positives, as the individual cases cited above illustrate.Moreover, the uncritical embrace of risk in such practices not only fails to capture the reality of people's/women's lives but also embeds an understanding of risk largely emanating from the work of Ulrich Beck and others in which risk is seen as the 'master key through which the most pressing social problems of the age can be unlocked' (Mythen 2014: 33).However, the extent to which this key is master in a global sense is open to debate.It is at this juncture the criminal justice practice of risk assessment, aligned as it is with the wider criminological and victimological embrace of risk, raises the spectre of criminology's origin stories (Carrington and Hogg 2017)  offenders.Ultimately, once the presence of this spectre is made visible, our senses can become differently informed.

Conclusion: Northern theorising of risk as an 'occupation of the senses'
Shelhoub-Kervorkian (2016b) deploys the notion of the 'occupation of the senses' in a very specific context in which the practices associated with it have very powerful consequences, both literal and metaphorical.The adoption of this analytical frame here is in no way intended to downplay the import of her analysis.Indeed the intention is quite the reverse.In her work she: … refers to technologies that manage language, sight, sound, time and space in the colony: the administration of who acts, who speaks, who gives birth, and how, and who walks/moves/drives where and how; and what kind of language, music, smells, marches, colours, cultures and scenes are promoted and inscribed over the spaces and the lives of the colonized.(Shelhoub-Kervorkian 2016b: 1) These technologies are writ large and small in the everyday lives of those with whom she is concerned.In the context of intimate partner violence, such practices/technologies of control are no less dramatic and have no less of an impact on the lives of those living with violence.These lives, like those with whom Shelhoub-Kervorkian (2016b) is concerned, also often result in death.These deaths can be writ large as in the high rates of femicide in Central America's violent Northern Triangle (which are not all explicable by gang warfare and/or trafficking; Eguizábal et al. 2016) or writ small as in the individual tragedies of Clare Wood (after whom Clare's Law, or the Domestic Violence Disclosure Scheme, was introduced in England and Wales) or Kelly Thompson in Australia discussed above.Yet these deaths also allude to an 'occupation of the senses'.This occupation of the senses can be found in how criminologists, victimologists, and criminal justice practitioners theorise and make sense of the kinds of violence(s) under discussion here.This is a different kind of occupation of the senses without doubt but one that blinkers these areas of investigation and ensures that they, too, lose their vision in a different though no less profound way as those referred to in Shelhoub-Kevorkian's (2016b) work.This loss of vision can be traced in the conceptual failure of risk as applied in risk assessment practices and returns us to the imperative of excavating both in theory and practice the impact of Northern theorising.Failure to break free of this occupation of the senses will only further embed ways of thinking and doing, reflecting a blinkered vision of women's real lives with continuing fatal consequences. Correspondence Despite feminist interventions documenting the repeated nature of intimate partner violence, the 'discovery' of repeat victimisation by researchers mining data from criminal victimisation surveys contributed to different practices highlighting and informing responses to such patterns of victimisation.For example, initial responses in the United Kingdom (UK) 'flagged' such victims on police computer systems so that they could be dealt with appropriately.IndeedBoxall et al.
(2015)suggest more could be done in terms of such 'flagging' practices in sharing and communicating different kinds of information about particular incidents.Nonetheless, contemporarily quite sophisticated risk assessment tools are used to assist in decision making about and resource allocation to incidents of such violence.There is a range of such tools in operation, from the spousal risk appraisal guide (SARA), the Propensity for Abusiveness Scale (PAS); to the Partner Abuse Prognostic Scale (PAPS) (all quoted in Hoyle 2008: 327).The DASH (Domestic Abuse, Stalking and Harassment, and Honour Based Violence) model is favoured by most police forces in the UK.McCulloch et al. (2016) review a further nine tools (including DASH) designed to inform responses to intimate partner violence and assessing the levels of risk judged to be present for individual cases (high, medium, low).Robinson and Rowlands (2009: 191)comment that these developments are 'posited on a common understanding of domestic violence, in particular the likelihood of escalating risk by the offending partner'.Yet, as McCulloch et al.  (2016: 58)state, 'there is a paucity of empirical research evaluating the outcomes of [international] risk assessments …', with Walklate and Mythen (2011)9)s of information in respect of risk factors are hidden from view: structural variables and what might be termed 'experiential knowledge'(Walklate and Mythen 2011).The denial of structural variables as risk factors in incidents of intimate partner violence is inherent in the individualised focus of the ecological model underpinning many of the policy responses in this area.This focus, derived from this model for preventing individual violent behaviour, presumes that such behaviour escalates prior to the last violent attack (seeJohnson et al. 2017).Whilst there is evidence to support this presumption, it does not apply to all cases and structural factors are hidden by it.This carries very specific consequences.For example,Robinson and Rowlands (2009)point to the inherent hetero-sexism of risk assessment tools alongside the limited vision they offer of men as victims.This individualistic bias, and its associated blinkered vision, also translates into what professionals are enabled to 'see' in terms of doing risk assessment (see below).Further to this observation, asCunneen (2014)has commented, the risk factors that are included in such tools are given the status of 'facts' and as 'facts' pay little or no attention to the historical and social context underpinning the evidence base from which they have been derived.Moreover, as is discussed more fully shortly, the failure to appreciate the importance of historical, social and cultural factors takes a particular toll on those subjected to risk assessment under conditions of settler colonialism as both offenders and victims.So, what is included and, importantly, what is excluded in the operationalisation and measurement of 'risk factors' hints at another way in which the ghostly presence of (Northern) theory and policy meld together.In addition, these inclusions and exclusions have a compounding effect when the role of whatWalklate and Mythen (2011)have termed 'experiential knowledge' comes into view.
and is similarly applied to both offending behaviour and victimisation.As a IJCJ&SD 6 Online version via www.crimejusticejournal.com © 2018 7(1) Thompson in Melbourne, Australia in 2015, who was killed by her partner after 38 calls to police over a three-week period for breaches of intervention orders.The failure of risk assessment tools, particularly in assigning an appropriate level of risk of victimisation, is self-evident not only from coroners' reports (see Gray 2016) but also from thematic analysis of fifteen Independent Police Complaint Reports on individual cases of domestic homicide (Hopkins and Walklate 2017).Indeed, the ongoing inadequacies of policing responses to intimate partner violence are well documented in England and Wales (Her Majesty's Inspectorate Constabulary