Violence Regimes: A Useful Concept for Social Politics, Social Analysis, and Social Theory

This paper critically interrogates the usefulness of the concept of violence regimes for social politics, social analysis, and social theory. In the first case, violence regimes address and inform politics and policy, that is, social politics, both around various forms of violence, such as gender-based violence, violence against women, anti-lesbian, gay and transgender violence, intimate partner violence, and more widely in terms of social and related policies and practices on violence and anti-violence. In the second case, violence regimes assist social analysis of the interconnections of different forms and aspects of violence, and relative autonomy from welfare regimes and gender regimes. Third, the violence regime concept engages a wider range of issues in social theory, including the exclusion of the knowledges of the violated, most obviously, but not only, when the voices and experiences of those killed are unheard. The concept directs attention to assumptions made in social theory as incorporating or neglecting violence. More specifically, it highlights the significance of: social effects beyond agency; autotelic ontology, that is, violence as a means and end in itself, and an inequality in itself; the relations of violence, sociality and social relations; violence and power, and the contested boundary between them; and materiality-discursivity in violence and what is to count as violence. These are key issues for both violence studies and social theory more generally.

Violence, and gendered and intersectional forms of violence, are an extensive global problem connected to power, inequalities, health, economy, crime and security, impacting all societies (Krug et al., 2002). Violence can also be a point of departure to examine and intervene in social life, whether the focus is on social politics, 1 social analysis or theoretical development (Hearn et al., 2016). Indeed, as a driving principle, analyzing and understanding violence is key for its reduction, as a fundamental link between theory and society.
The range and depth of both violence and research on violence are immense. Within this range, there is considerable research that examines the co-variation 2 of and interconnections both amongst and between forms of violence (for example, homicide, child abuse, civil conflict) and further aspects of violence (political organizing, policy development against violence, knowledge production, measurement, definitions/inclusions/exclusions). However, much research on violence remains fragmented in different disciplinary and with substantive focuses on different forms or aspects of violence.
This paper addresses two linked questions: the extent to which research literature reports on how different forms and aspects of violence co-vary and interconnect, or not; and the need to work towards a more inclusive, more comprehensive conceptualization of violence than more partial approaches. Specifically, the paper addresses two key research questions: • How can the extent to which specific forms and aspects of violence co-vary and interconnect, as reported in research literature, be usefully conceptualized and theorized in a more inclusive, more comprehensive way? • What difference would a more inclusive, more comprehensive conceptualization of violence make to social politics, social analysis, and social theorizing on and against violence?
We address these questions by arguing that the concept of violence regime (Hearn et al., 2020;Strid et al., 2021) -that is, the governance and production of forms and aspects of violence -constitutes a more inclusive and more comprehensive approach to such interconnections (or lack thereof), and a way of dealing with tensions and limitations in more partial frameworks. We also hypothesize that forms and aspects of violence co-vary with each other to constitute a violence regime. Furthermore, we hypothesize that such co-variance is likely to be greater than with other inequalities and measures that are often assumed to 'cause' violence: in this view, a violence regime is relatively autonomous rather than merely a subset of other inequalities.
In line with Davenport's (2021) recent call for "integral violence studies", this paper aims to contribute to overcoming the fragmentation of policy and politics, empirical research and theoretical development on violence, and its different forms and aspects, and to further interdisciplinary research on violence. These are important issues for social politics, social analysis, and social theory. Accordingly, we see violence regimes as a policy concept, an empirical concept, and a theoretical concept.
But, there is an immediate problem: namely and simply, what is violence? There are multiple contestations of what violences are -physical violence, assault, sexual violence, coercive control, homicide, genocide, as well as less directly physical violences, such as cultural, symbolic, epistemic and systemic violence (Bourdieu, 1998;Žižek, 2008). Violence includes, but is not limited to, state violence, economic violence, terrorism, interpersonal violence, gender-based violence (GBV), violence against women, anti-lesbian, gay and transgender violence, intimate partner violence, gang violence, hate crime, cyberviolence, and stalking. So, is violence a set of material bodily actions and effects? A range of discursive constructions? Is violence more structural in character, as, for example, through institutions or structural inequalities? Or all of these, intersectionally gendered? In this paper, we use illustrative examples from GBV, whilst not restricting discussion to gendered violence regimes. 3 In this text, violence and violence regimes are understood as material-discursive, including how violence is defined, what is included and excluded as violence, in knowledge production on violence, and as operating across micro/interpersonal, meso/institutional, and macro/(trans)societal scales. Theoretically, we locate ourselves in debates on the close relations of materiality and discourse, such that recognitions and contestations of violence are part of the problem of violence. Indeed, the problem of what violence 'is', and how violence is named, defined, limited, understood and explained, is a key question across many realms: everyday life, social politics, social analysis, and social theory more broadly, albeit unevenly. Knowledge production on violence, who gets to define violence, contestations on definitions, are part of violence regimes, thus engaging with wider theoretical debates, such as on regimes of truth (Foucault, 1980). Violence as an open-ended, but not relativist, concept that refers to multiple ways in which humans 4 can be attacked, transgressed, harmed, destroyed, materially-discursively. Violence certainly involves direct physicality, but it is more than that, as with physical harms without direct physical violence.
The violence regime approach brings a focus on violence, and re-centres violence -that it is not only concerned with specific 'types' of violence but sees violence as both cause and consequence of social realities. As such, this paper draws on and develops the very long tradition of feminist theory, practice and research on violence (for example, Brownmiller, 1975;Hanmer and Maynard, 1997;Hester et al., 1996;Kelly, 1988;MacKinnon, 1982;Moran and Sharpe, 2004;Gordon and Meyer, 2007). We consider uses of violence regimes as a heuristic analytical tool for social politics, social analysis, and social theory. Building on earlier empirical, policy and conceptual work (Hearn et al., 2020;Humbert et al., 2021;Strid et al., 2021), the concept of violence regime forces consideration of different understandings of what is to count as violence, from direct killing to forms of violence often not recognized as violence at all, such as violence to non-humans, colonial violence, slow and environmental violence.
Following this introduction, we continue by examining how studies on violence are subject to disciplinary fragmentation, as a precursor to examining five inspirations and ways of moving towards a broadened, more inclusive conceptualization: violence regime. The following section focuses on the notion of regime more directly, building on earlier work, and spells out the violence regime concept. The latter parts of the paper consider three applications of violence regimes -in social politics, social analysis, and social theory. In the first case, violence regimes address and inform politics, policy and policy development, labeled here as social politics, around violence, such as gender-based violence, violence against women (VAW), intimate partner violence (IPV), antilesbian, gay and transgender violence. Extending understandings of violence is important for policy formulation, development and implementation, whether working at the UN scale, regionally as in, say, the Council of Europe, nationally or locally. More comprehensive analysis of the complexities of violence is likely to demand transcending both disciplinary and policy area boundaries, with policy responses often depending on where within state (or parastate) machinery violence policy is located, for example, war and militarism in ministries of defence, violent crime in ministries of justice.
Second, violence regimes, and indeed intersectional gender violence regimes, are part and parcel of social analysis, including the analysis of the place and relation of violence to societal contexts, broadly based structures of inequality, governance, welfare state regimes, gender regimes, and social movements, and empirical data thereon.
Third, violence regimes engage with social theory. One may ask: what assumptions on the construction of the subject, for example, 'rational' or individualist, pertain in different social theories, in relation to violence, and the experience of being violated? Specifically, social theory is partly constructed by, through and in relation to violence, including the frequent exclusion of the knowledges of the violated, sometimes but not only through killing. The concluding section discusses further implications of the conceptualization of violence regimes.
Different disciplines have tended to focus on different forms of violence, ranging from, for example, psychological studies of offenders to international relations, as well as varying in the extent to which the gendered nature of violence is highlighted (cf. O'Toole & Schiffman, 1997;Ray, 2011Ray, /2018Pease, 2019). 5 Psychological-orientated disciplines tend to focus on individual experience of inflicting or receiving violence and more individualized forms of violence. In contrast, more societally-orientated disciplines tend to address broader societal and comparative patterns of violence, collective violence, and more structural explanations thereof (Aya 1979;Ray, 2011Ray, /2018, previously often framed by class more than gender or intersectionality. The relative fragmentation of research on violence is seen in how different forms of violence are often studied separately in different disciplines: inter-state, i.e. violence between states (e.g. war, militarism) predominantly in political science (Ní Aoláin et al., 2018); intra-state, i.e. violence within states (e.g. state-citizen violence) in sociology; and interpersonal violence (e.g. assault, homicide, sexual violence) in criminology, psychology, and gender studies (Walby, 2013;Lombard, 2018). Some studies seek to bridge two, and occasionally three, of these broad forms. Some texts bring together the first and second approach (True, 2019;Gentry et al., 2018), while all three forms are addressed within The Palgrave handbook of criminology and war (McGarry & Walklate, 2016). Even so, disciplinary hegemony remains strong.
Some approaches are consciously multi-disciplinary. For example, Collins' (2009) micro-sociological theory of violence bridges sociology, psychology, anthropology and physiology, deriving from situational interactions on a face-to-face level. He emphasizes that violence is "hard to perform", typically after preliminary interaction ritual of emotional escalation, and depends on overcoming human physiological and social inhibitors-regardless of the underlying conditions or motivations. A virtue of this approach is that violence is seen as embedded in everyday social life, not a separate 'domain' alongside 'economy', 'polity', 'civil society' (Walby, 2009). However, this approach focuses on direct physical violence on the body and by use of weapons, sometimes at a distance, rather than other forms and aspects of violence, which we address. Others employ multi-/interdisciplinarity on violent situations, for example, in analyzing GBV before, during and after cyclones, Rezwana and Pain (2020) span social geography, violence, trauma, militarism and war, international relations, and disaster studies.
Within this disciplinary context, we aim to "de-fractionalize" the study of violence by bringing together research from various academic fields (such as political science, sociology, criminology, law) to examine what it means to think of the organization of violence as producing distinct kinds of violence regime. By merging cross-disciplinary collaborations and multi-disciplinary studies, data and indicators, we argue for substantial policy, empirical and theoretical analyses of violence and violence regimes across disciplines.

Conceptualizing violence: working towards violence regimes
As noted, there are many ways of conceptualizing violence. Differences often centre around contrasts between physical violence and non-physical violence, or between different forms of violence, on one hand, and power and domination, on the other. This is a start, but by no means is satisfactory as a solution to the bounding of violence, especially in moving away from more immediate agent-driven violence. Accordingly, we now consider some ways of broadening the conceptualization of violence. Appropriately, we do this through a multi-perspectival approach, following several lines of inspiration. Examination of established approaches to the conceptualization of violence is useful and necessary in two ways: first, as indicative of more partial approaches to violence that need to be built upon; and, second, as examples of forms and aspects of violence to be critically interrogated and included in developing the concept of violence regime. Building on and bringing together established approaches is a necessary step in specifying a more inclusive, more comprehensive violence regime approach.

Beyond (certain kinds of) illegal physical violence
First, let us take a point of departure from the The concept and measurement of violence against women and men (Walby et al., 2017) book, which focuses on and limits violence to physical violence and the illegality of violence: "The meaning of the concept 'violence' is contested. It has been stretched beyond physicality so that it encompasses many forms of power and harm, losing its distinctiveness, becoming submerged within notions of 'abuse' and 'coercion'. For the purposes of a theory of change -in order to potentially make visible the relationship between violence and other forms of power and to identify the levers of transformation -it is better to restrict the concept of 'violence' to a specific and precise definition connected to intended physical acts that cause harm. Yet, many of those who use a precise definition of violence underestimate the extent of violence against women, leaving this dimension invisible." (p. 4) (our emphasis) Here, the authors prioritize "intended physical acts that cause harm" in addressing violence as what appears to be a pragmatic approach to the measurement of specific forms of violence, with "harmonized counting rules", and as a way forward for political and policy development. However, that narrowing of violence stands in opposition to much political debate generated from the women's shelter, LGBTIQA+ politics and related feminist movements. One of the problems here is that violence is typically framed as certain kinds of identifiable, intended 'physical acts'. This approach excludes several possible versions of violence, for example: intended non-physical acts that cause harm; physical acts that do not cause harm (however claimed or defined by some); unintended physical acts that cause harm. Narrowing violence can make for simpler argumentation, and perhaps greater policy leverage in the short term. However, such an approach does not tackle some of the complexities of broader notions of violence, abuse and violation. We aim to go beyond that approach.
Definitions thus range across context, intention, harm, form and extent of damage, short term/long term forms, experience and subsequent interpretation of experience, as well as physical, sexual, emotional, representational, organizational, and so on. In addition, there are clear variations in the extent of recognition of violence, by states, the law, social movements, and so on, and how violences and potential violences are recognized as illegal or legal, illegitimate or legitimate. Thus, to summarize so far, violence can be understood, in terms of: • intention to harm and do damage, and the extent and explicitness of that intention to kill, to harm, or to cause damage and destruction in other ways; • use of force or other means as the medium for inflicting harm and damage, including direct violent force, forceful use of parts of the body to hit and attack, indirect violence; use of weapons or prosthetics, sometimes at a distance; • extent of harm and damage, and the experience of that as damaging and/or violating, including physical damage, psychological damage, further damage, control, violation; • extent, whether absolute or relative, to which violence is legal/illegal and/or constructed as legitimate/illegitimate; • recognition and extent of recognition as violence, and by whom: states, law, professions, social movements, public opinion, social media.
While these five elements are clearly important, they do not always coincide or relate neatly to each other; moreover, they are not exhaustive, as we discuss below. In these and other ways violence becomes less restricted, more open to a variety of forms, processes and recognitions. For example, violence, specifically physical violence, may be (understood variously as): the enactment of dominance; resistance to felt, potential or actual loss of dominance; compensatory; enacting resistance to dominance; or even as unnecessary to maintain control because of previous violence or threat thereof (Hearn, 1998b;Muro-Ruiz, 2002).

Broadening harm, and the experience of harm
The question of the nature of harm deserves further elaboration. A harm-based orientation to violence is distinct from the measurement of violence in terms of intentional physical acts (Myhill, 2017;Myhill & Kelly, 2019;Bjørnholt & Hjemdal, 2018). Harm -physically, psychologically, existentially -is far from being a simple effect of immediate violence. Moreover, the extent to which harm is experienced or consciously recognized is different from harm as effects. There is considerable empirical research on how harm can be caused to people by negative treatment without the conscious knowledge of those subject thereto. For example, Landrine and Klonoff (1997) argued that it is the presence and exposure to sexist acts, rather than women's subjective appraisals thereof, which is the best predictor of women's negative health symptoms. Krieger and Sidney (1996), surveying 4,000 black and white young adults in the US, found that blood pressure was highest for working-class black adults who accepted discrimination as "a fact of life" or who denied they experienced discrimination, and lower for those who challenged unfair treatment. There are well-documented processes of extended, repeated, delayed and reactivated trauma following violence (Herman, 1997). More broadly, in interrogating harm, there is accumulated evidence for trauma at individual and group levels, but also (historical) collective levels, as examined in analyses of slavery, colonialism, indigeneity, and feminist, queer and associated commentaries (Pain, 2020;Swanton, 2021).
Furthermore, the experience of the violated persons in receipt of violence, whether direct or less direct, is here more pressing than the focus on perpetrators' intentions to harm. Violence can also be harmful and seen as transgressing bodily integrity without the direct use of physical force, as in certain forms of child abuse, sexual assault, sexual grooming and online violence. With digital and online violations, harm, that is, experiences of harm, can delayed, even for many years (Hall & Hearn, 2017;Hall et al., 2022). Thus, use of direct force, intention to harm, and even immediate experience of harm, are all not sufficient in knowledge production on violence.

Multi-scalar continua of violence
Another powerful inspiration impacting on the development of our work is the idea of a continuum of violence, from more direct physical violence to less direct "nonphysical" violence, even while these have physical, bodily effects. Such ideas developed in the 1970s and 1980s through Women's Movement activism, policy development and research on violence against women, and codified in the continuum of sexual violence (Kelly, 1988; also see Boyle, 2019;Graaff, 2021). The continuum metaphor has been elaborated in terms of a temporal and spatial continuum of violence (Cockburn, 2004(Cockburn, , 2014, spanning from personal to international, including structural violence and economic distress, militarization and arming, discursive shifts in ideology, war, political terror, mobilization, everyday life disruptions, brutalization of the body, sexual violence, peace processes -across pre-conflict, conflict, peace-making and reconstruction. Continua thinking has also been applied to interrogate and partially transcend constructed divides of private/ public, and war/ peace (Scheper-Hughes & Bourgois, 2004;Yadav & Horn, 2021). Similarly, works in feminist political economy have addressed violence and harms across multiple spheres of social reproduction (e.g. True, 2012;Meger, 2016;Gentry et al., 2018), for example, using a depletion (e.g. Rai et al., 2014;Elias & Rai, 2018;Chilmeran & Pratt, 2019) Continua approaches link with attempts to examine violence across disciplines and scales, as, for example, violence as a health burden within a multidisciplinary ecological framework (Heise et al. 1994;Heise 1998); and violence seen as the combination of structural violence, cultural violence and physical violence, with an inclusive frame of peace (Galtung, 1969(Galtung, , 1990. Such research bridges some, but not all, of the fragmentations noted. They still tend to focus on continua of forms of violence, rather than aspects of or knowledge production on violence.

Varieties of structural violence
These continua imperatives suggest a fifth inspiration, namely debates on structural violence, which deserve further elaboration. Here, the emphasis is on both different broadened forms of violence and different analytical abstractions of and around violence. The theoretical move is from direct violence, with limited time and space between act and impact, such as killing and interpersonal physical violence, and not only to non-physical violence (with physical or non-physical harm), onto questions of violent institutions, and thereby to violent structures, that lead to collective and societal harms, premature deaths and avoidable disabilities (Galtung, 1969(Galtung, , 1975Pogge, 2008;Scheper-Hughes & Bourgois, 2004). Structural violence can include: • structural patterns of individual and interpersonal violence; • systemmatic oppression of a specific social group; • violent acts and effects of social institutions, for example, the state; • violent (nature and effects of) inequalities and poverty on a world scale, including gender inequalities in power, health, education, paid and unpaid work; • violent effects of war and collective violence; • structural relations of institutions historically violent or underwritten by violence, for example, capitalism, the state; 6 • the social conditions (re)producing violence; and • violence as social relations.
These are all important forms of violence. However, a remaining gap is how structural violences, along with multi-scalar continua approaches, engage with epistemology.

Epistemologies of violence
A final inspiration for our task is epistemological in character. One example here concerns the relation of materiality/physicality/bodily effects and discourse/representation/perception in studying violence (Hearn, 2012(Hearn, , 2013. The issue of the relation of the enactment of different forms of violence and the social construction of such violence has been raised previously in different ways in the journal. For example, Ghoshal (2013) has addressed the question of collective memory, or more precisely mnemonic opportunity structures, in relation to violence; Skotnicki (2019) examined the phenomenological structure of suffering from slow environmental violence; and Hearn (1998a) highlighted differential discursive positionings in researching violence. Such debates complement more structuralist accounts of violence by not just including the voices of those concerned, but by showing how those discourses, voices, perceptions, memories and so on impact back on the production and reproduction of violence. The material physical doing of violence, and people's difficult and diverse relationships to that doing, in turn affects the construction and very recognition of violence, and what counts as violence. We return to these issues later in discussing applications to social theory.
In summary, these multiple meanings and interpretations of violence -violence against women, gender-based violence, intimate partner violence, public violence, indirect violence, institutional violence, continua of violence, structural violence, epistemologies of violence, and much more -here become an inspiration for policy, analysis and theory, rather than an awkward problem for empirical measurement (cf. Walby et al., 2017). The literature reviewed is relevant for our task, but to contribute to policy, analysis and theory a different approach is needed.

Violence regimes
Having reviewed the above inspirational debates on violence, we now return to our key research questions: how can the extent to which specific forms and aspects of violence co-vary and interconnect be usefully conceptualized in a more inclusive, more comprehensive way? And what difference would a more inclusive, more comprehensive conceptualization of violence make to policy, studies and theorizing on and against violence? We consider the first question by spelling out the concepts of regime and violence regime, as a more comprehensive frame to examine co-variations and interconnections between forms and aspects of violence. After setting out the concept of violence regime, we consider the second research question by outlining applications of the violence regime concept in social politics, social analysis, and social theory, before concluding discussion on its wider implications.
In addressing the question -how can the extent to which specific forms and aspects of violence co-vary and interconnect be usefully conceptualized in a more inclusive, more comprehensive way? -we turn to the concept of regime and then violence regime. This regime concept has been taken up, particularly in sociology and political science, as an alternative, more open-ended concept compared with the loaded concepts of (social) system and (social) order, with their different connotations and reference points. The regime concept suggests greater flexibility and accommodation of the macro-systemic (Walby, 2009), meso-institutional (Connell, 1987 and everyday relations of relations of ruling (Smith, 1993) than system or order.
The concept of regime has been applied in a range of further contexts, perhaps most famously welfare regimes (Esping-Andersen, 1990;Goodin et al., 1999). Feminist critiques have gendered welfare state regime typologies, recognizing the contribution of family, unpaid work and care to economy and welfare (Lewis, 1992;Sainsbury, 1999), women's welfare dependency, and the gendered state-market-family nexus. Yet, such critiques have often, rather surprisingly, excluded consideration of violence (see Humbert et al., 2021;Strid et al., 2021), overlooking deep-rooted causes and consequences of gender inequality, as well as questions of race, intersectionality and bodily integrity (Pringle, 2011).
Another example of regime thinking is that of gender regimes, within which polity, economy, civil society and violence operate as institutional domains (Walby, 2009). In Walby's gender regime theory, which draws on critical realism and complexity theory, the institutional gender regime domain of violence takes the other identified institutional domains (economy, polity, civil society) as its environment, with the whole making up the gender regime. This allows analytical separation of 'violence' from the other domains shaping it -a way of conceptualizing violence whose limitations we have previously examined (Hearn et al., 2020).
The welfare regime and gender regime approaches are useful and powerful conceptualizations, but are distinct from how we develop the violence regime as a framework for comparative, policy, analytical and theoretical work. There is certainly a variety of evidence to suggest significant interconnections (but also disjunctions) between welfare regimes, gender regimes, and violence regimes. Interconnections have been demonstrated between various socio-economic inequalities and violence, for example: • economic inequality, gender inequality and (gender) violence (Walby, 2009); • rape and conditions of food scarcity, migration, and conflict (Sanday, 1981); • greater gender equality and less violence (Holter, 2014); 7 • men's domination of labour force participation, and women's exclusion therefrom, along with greater likelihood of internal societal conflict (Caprioli, 2005).
Such studies might cast doubt on the relative autonomy of violence regimes. However, in contrast to welfare regimes and gender regimes, we move from an approach that sees violence as only an institution or a domain (cf. Walby, 2009) to violence as an overall regime, with its own violence domains. A violence regime comprises the governance (in a broad sense) 8 and production of violence, including perpetration, victimhood/survivor-hood, responses, policies and knowledgemaking. While different identifiable forms of violence, such as murder, may seem self-evident and speak for themselves as violence (unlike, say, violence in violent sports, corporal punishment or even rape in marriage), forms of violence are also constructed through the governance of violence.
Violence regimes is a relatively new concept, theoretical framework, and policy tool. Schinkel (2013) wrote in 2013 on the relation between interpersonal, structural and state forms of violence, claiming he "introduces the idea of a regime of violence" (2013: 313); however, this was not entirely correct. Kössler (2003) had earlier used "regimes of violence" to discuss state monopoly on the legitimate use of violence and changing relations between states after 9/11. Schinkel's later use of the concept describes the relation between forms of violence, arguing that regimes of violence constitute a way of governing conduct via the medium of violence.
We go further than Schinkel and his focus on interpersonal, structural and state forms of violence. In our usage, violence regimes refer to the governance and production of violence, that includes both more forms 9 of violence and also aspects of violence within violence regimes as politics, policy responses, extent of criminalization, attitudes to violence, organized resistance to violence, and knowledge production on defining, dealing with understanding, explaining and framing violence, and what is and is not considered violence. The violence regime concept interrogates if and how the production and institutionalization of violence co-vary to constitute distinct violence regimes. Importantly, many forms of violence -interpersonal (e.g. crime, GBV), inter-state (e.g. war), state-citizen (e.g. use of death penalty) and group-state (e.g. terrorism) -are connected, such that an increase in one form likely leads to increases in other forms. Connections have been shown between gender unequal and homophobic societies and those with the highest level of societal violence, and most at risk of armed conflict in their own territory (Ekvall, 2019).
In handling the diversity of violences, violence regimes can be seen as relatively autonomous, sometimes contradictory, across different scopes of violence -concerning direct and indirect violence, across several domains of comprehensiveness. The concept of violence regime (Hearn et al., 2020;Strid et al., 2021) which we develop here deploys the metaphor of four pillars 10 or rather domains 11 of violence (see Appendix) that vary in both (material) forms or manifestation and (discursive) aspects, knowledges, understandings and framings of violence: • deadly: manifestations of violence with potential to kill, usually direct and directed towards someone (has a 'victim' or 'object'), as in deadly violence; • damaging: (more precisely, directly damaging, but not immediately deadly) manifestations of violence/violations with potential to harm or injury, usually direct and directed towards someone (as 'victim' or 'object'); • diffuse: underpinnings to manifestations of violence, usually less direct, and directed towards a group, usually with an identifiable 'victim' or 'object'; and • dispersed: other manifestations not necessarily understood as violence, usually indirect, sometimes towards a group but with a less easily identifiable perpetrator(s), 'victims' or 'objects' (Hearn et al., 2020).
Moreover, these domains move from deadly homicide, femicide, death penalty, and militarism, to broader conceptions of damaging violence, such as recorded violent crime, violence against the person, IPV and stalking. Beyond those direct violences lie indirect, diffuse violences, including legitimizations and regulations of violence, and dispersed violences that are not usually or widely recognized as violence. The first two domains involve direct physicality; the latter two may do so too but can include forms and aspects of violence with harmful, including fatal, physical effects, without direct, immediate physicality by specific perpetrators.
The capacity to harm, and indeed to kill, varies across the domains, though not in a strictly linear way. The left-hand domain, deadly violence, is explicit in terms of killing, as constructed by the state: homicide, suicide, and "legal" violence of the death penalty and war. The domains further to the right of that may appear less explicit in that respect, but in fact both often provide the ground for harm, and also lead sometimes indirectly, to greater harm, killing and deaths. For example, deadly interpersonal violence, as with murder as an individual, physical direct act of violence (at the top left) can be compared to dispersed violence and multitudinous collective deaths at the macro-level, say, through colonialism (bottom right). The domains also vary in terms of the increasing time/space between the immediate, instantaneous violent act and the consequent impact of actions, and the broadening of impact with the greater distance in time/space, from physical and temporal proximity to geopolitical distance and temporal delay.
The domains assist in structuring the examination of what it means to think of the organization of violence as producing distinct kinds of violence regime, across some or all domains. The concept of violence regime can be utilized in empirical studies of a particular society or locale, for example, in considering historical change in how violence is manifested, governed and understood, and in comparative studies between societies and between locales. For example, forms of violence unacknowledged in one historical period, specific societal, cultural, legal or indeed social science framing, such as chastising/hitting children, anti-transgender violence, marital rape or violence to animals, may become acknowledged subsequently.
Thus, different and particular versions of violence regimes are to be specified, in terms of the extent to which there are consistent patterns of co-variation and interconnection among and between forms and aspects of violence, such as extent of violence manifested, punishment of violence or opposition to violence, or alternatively as showing uneven patterns, that is, lack of or non-co-variation in relation to violence. The autotelic nature of violence, in simplistic terms, can be understood as a set of positive correlations between different forms of violence. As such, we would seek to establish reliability in measurement scales of different forms of violence. Where we find uneven patterns, then we simply fail to establish this, but this does not mean that other kinds of relationships do not exist -either between some but not all violence domains, or in terms of other relationships such as curvilinear ones (u-shaped) or even negative ones. Violence regimes can also be used as a theoretical concept, in terms of how violence is defined, understood, explained, framed and theorized. Applications -policy-wise, empirically, and/or theoretically -can be made from all or some of the four domains, not least because of the variability of data. Even considering the first two domains can, from our previous comparative studies, be highly productive Strid et al. 2020). On the other hand, a theoretical, and sometimes empirical, development is to consider violence regimes across all domains, including non-recognition of certain forms and aspects of violence.
Considering violence in this broadened way links with debates on several major issues: long-term feminist work on multiple forms of violence, problematization of the supposed private/public divide, impacts of technology on violence, transnational violence, interconnectedness of multiple violences, as well as historical and ongoing contestations on definitions, understandings and boundaries of violence.

Applications of violence regimes
We now move on to the second research question: what difference would a more inclusive, more comprehensive conceptualization of violence make to policy, studies and theorizing on and against violence? Thus, we now turn to three primary applications of the violence regime approach: in social politics, social analysis, and social theory.

Violence regimes and social politics
The concept of violence regimes can be useful for both studying and developing policy responses to violenceas a policy concept, it is useful for both analyzing current regimes, politics and policies, and thinking beyond present conditions to enhanced policy, prevention, intervention and change. A violence regime approach to social politics makes visible the interconnectedness of multiple forms of violence. It re-centres violence and raises questions about what we can learn from approaching social politics through violence policy.
Policy responses to violence depend significantly on where in the state machinery violence is located, different forms of violence are treated separately. Indeed, the prevailing concept of violence in social politics fragments violence and policy, and in turn fragments the field of violence, with specific forms of violence dealt with in separate policy domains. In some countries policies on violence are spread across different parts of the state, for example: violence from the state towards other states, in terms of militarism, war and defence, located in ministry of defence; violence from the state towards its citizens and non-citizens, and violence from adult citizens to adult citizens and non-citizens in ministries of justice, the interior or home affairs; and violence from adults to children in ministry of health, family or social affairs.
Our previous research on the policy process, policy content and measurement of violence in multiple policy domains, has shown that while the policy process and policy content of violence may differ, measurements of violence across deadly and damaging domains of violence remain similar. For example, while the measurements of deadly and damaging violence are similar in France, Sweden, and the UK , there are clear differences in violence and anti-violence policy process and policy content between these three countries (Balkmar et al., 2021a), suggesting different violence regimes in terms of the governing of violence.
These findings connect to issues of measurement more generally, including what counts as violence in policy contexts, across different political contexts and temporalities. When used as a policy concept, violence regimes approach can assist in expand the notion of violence, and thus the range of policy. It can be applied to develop broad and inclusive policy responses to violence. For example, the United Nations' Beijing Platform for Action (UN 1995), agreed on by 189 states, defines violence against women as: "any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm or suffering to women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public or in private life" (article 1). It includes physical, sexual and psychological violence in the family, general community and/or perpetrated or condoned by the state. In consequence, for states to pursue "all appropriate means of eliminating violence against women", the violence policy needs to be expanded and policy makers urged to emphasize interconnections and inclusion of multiple forms of violence. This wider framing of violence suggests that all violence domains outlined (deadly, damaging, diffuse, dispersed) are important. While a policy focus on acts of physical force is clearly vitally important, this may neglect forms of violence less easily defined and measured, such as psychological, emotional and digital violence and abuse, as well as financial, health and social coercive controls, as emphasized especially in diffuse and dispersed domains (see Appendix). This can suggest not only ways of comparing violence policy across countries and locations, but importantly forms and aspects ripe for further activism, politics and policy development against violence.

Violence regimes and social analysis
Second, violence regimes can be applied, as an empirical concept, to assist the analysis of violence in the relation to, for example, societal contexts, structures of inequality, governance, social movements, welfare regimes, and gender regimes. To illustrate the importance of such interconnections within a violence regime frame, we first return briefly, to the relation of violence to the earlier work on gender regimes (cf. Walby, 2009): violence regimes can be understood as based in the co-variation of different forms of violence with each other than their co-variation with their most relevant other institutional or gender domain, as well as lack of co-variation with other measures of gender (inequality) regimes and the other gender domains of civil society, economy and polity (Hearn et al., 2020). A key point here is that different forms of violence may co-vary more with each other than they co-vary with their most relevant other gender regime domain, that is, their environment (economy, polity or civil society). For example, forms of violence located mainly in one gender regime domain, such as IPV and child abuse (located in the gender regime civil society domain) may co-vary more with each other more than with other measures of gender inequality in civil society.
These different forms and connections of violence themselves often interconnect. For example, to elaborate on the last of these points listed above, and its linkages with some of the other aspects noted, previous studies on slaughterhouse communities have shown documented increase in violent crime, and the increases in domestic abuse and child abuse (Broadway, 1990(Broadway, , 2000Stull & Broadway, 2004;Fitzgerald et al., 2009). Similarly, cruelty to animals is bound up with violent behaviours towards other humans, as in the co-occurrence of animal abuse and intimate partner violence (Fitzgerald et al., 2019).
Conversely, violence may not co-vary with other measures of gender inequality and gender regimes. For example, IPV (in civil society), violence at work (in economy) and armed political conflict (in polity) may co-vary more with each other than with gender inequality in domains other than violence (see Ekvall, 2019). We have previously applied the violence regime approach to compare the extent to which violence regimes empirically map, or not, onto gender regimes . Empirically, neither welfare regimes nor gender regimes correspond, in any straightforward way, with violence regimes. Indeed, such relationships may well be uneven, and even counterintuitive. In the EU, there is a positive correlation  (FRA, 2014;Humbert et al., 2021), along with lack of evidence for a positive correlation between gender equality and homicide of men. 12 It is in this context of relative autonomy that Table 1 presents some empirical results on the relationship between different forms of violence. Drawing on publicly available European data, as an example, this table points to the co-varying, interconnected nature of the relationship between different forms of violence. The correlation between suicide rates and intentional homicide rates -both measures of direct deadly violence -is r = 0.67 for the total population across the EU. Interestingly, this extends to state-based violence, as evidenced by the high correlation between defence expenditure (as a percentage of total government expenditure) and intentional homicide rates (r = 0.52) (and even more so with intentional homicide rates by family members or relatives: re = 0.56). More gender-based, non-deadly forms of violence -damaging forms of violence -also correlate with each other: this concerns rape, sexual assault, physical violence, and sexual violence against women (ranging from r = 0.51 for sexual assault of women and physical violence to women by (ex-)partner since age 15; to r = 0.67 between sexual violence by a (ex-)partner since the age of 15 and psychological abuse which involved controlling behaviour by a (ex-)partner since the age of 15; and 0.96 for rape of women and sexual assault of women).
One way of understanding these positive interconnections is in terms of the autotelic nature of violence (Schinkel, 2010); we have previously summarized this co-variation as a societal Violence Regimes Index . This shows that, although 'deadly' forms of violence and gender-based damaging violence are related, as evidenced by the strong correlation that both domains have with the overall index, they nonetheless operate independently -as they correlate more to each other within the same domain than to other forms in the other violence domain. This might reflect issues in measuring GBV and prevalence surveys more specifically, such as response rates, modes of data collection (Walby & Olive, 2014), but it could reflect ontological difference in these forms of violence. In any case, the positive correlation between the two domains strengthens the case for an autotelic reading of such forms of violence. In this framing, violence, as in positive correlations between multiple forms of violence is not simply reducible to other inequalities, but can be understood as more autonomous, structured by autotelic logic. Specific violence regimes, of national or other social locations, involve variations in (a) the extent of violence (more violent or less violent regimes), (b) co-variation within individual violence domains, such as deadly or damaging (more homogenous or less homogenous violence domains), (c) co-variation across two or more violence domains (more consistent/even or more inconsistent/uneven violence regimes), and (d) relative autonomy from measures of gender inequality, gender regimes and welfare regimes.
The production of violence in different states constitutes regimes, at least in terms of the first two domains (deadly, damaging), analogous to welfare regimes and gender regimes. Measuring other forms and aspects of violence (diffused, dispersed) is more difficult, due to lack of comparable data, even across Europe. However, as noted, they are not the same, suggesting that the neglect of or even exclusion of violence in mainstream social research has produced empirical results that may not be valid. Violence, in its multiple and gendered forms, should thus be included in social analysis of, for example, welfare states and welfare regimes, not least in analyses that claim to address the women-friendliness of the state.

Violence regimes and social theory
Third, there are many questions in social theory that can usefully be rethought through a focus on violence, and specifically violence regimes, as a theoretical concept. The notion of historical and cultural constructions, the relationship of individual action, meaning and social structure, the place of experience in knowledge production, conceptualizations of power, and the (de)construction of the 'self' are all key issues in social theory that can be re-viewed through the lens of violence. However, in much 'general' (often gender-neutral) social theory violence is not understood as a characteristic or pervasive form of interpersonal or structural relations -interpersonal or structural violence. For example, interpersonal relations are easily assumed to involve 'rational individuals' with a relatively unified self, who conduct their affairs accordingly in a liberal, tolerant and mutually adjusting manner. In this scheme, when violence is used, it can be seen, in the terms of the actor themselves, as instrumental for certain ends. When a man is violent to a woman 'he loves' or 'is married to' or 'has a sexual relationship with', the violence is usually constructed, by men, as aberrant. Violence is widely portrayed as a relatively isolated, as incidents, even incidental, and isolatable exception to 'normal life', rather than in a broader way, as foregrounded here (Hearn, 1998b).
For the remainder of this paper, we discuss the significance of five key social theoretical issues that are highlighted by the violence regime approach: social effects beyond individual agency; autotelic ontology and violence as inequality; violence, sociality and social relations; violence and power, and their contested boundary; and materiality-discursivity in what is to count as violence. These are key issues for both violence studies and social theory more generally.

Social effects beyond individual agency
Agentic models of violence continue to inform much study of and theorization of individuals, families, groups, organizations and their relations to social structures. Thus, a first major theoretical question is how the concept of violence regimes helps to move beyond the realm of agent-driven violence, especially that by more individual agents, but also more collective agents. This is difficult, and a persistent challenge for social theory. It easily falls prey to metaphysics and polemic. But it is necessary and worth doing. 13 A conceptual framework of violence regimes certainly forces consideration of very different forms and aspects of violence -from direct killing to more wide-ranging understandings of the production, and indeed consumption (Das & Kleinman, 2000), of violence, in which violence and violation are not seen in narrow agentic terms. To do this means attending to the assumptions made in models of social effects, social action, social relations, and the human, and indeed posthuman, in social theory in terms of incorporating or neglecting violence. In particular, with dispersed violence there can sometimes be no clearly identifiable agentic perpetrator or dedicated perpetrator or direct victim: assessment as violence is dependent on material-discursive construction.
In addressing dispersed violence, we may use concepts such as environmental violence and slow violence. 14 Climate change is a form of slowly unfolding environmental damage, often dislocated from their original causes, with violent and dispersed impacts across micro, meso and macro levels. The knowledge claims by those living with the "slow suffering" from environmental pollution are often overlooked (Davies, 2019: 14). As Davies (2019: 14) argues, "[b]y interrogating the seemingly 'out of sight' (Nixon, 2011: 2) nature of slow violence, and instead asking 'out of sight to whom? we can become more attentive to alternative perspectives and knowledge claims in polluted spaces." Rather than seeing different forms of violence as "subjective", "objective" and "symbolic" (Žižek, 2008), or as located along a continuum from less severe to more severe, forms of violence are better characterized as domains of actions, rules, structures and agencies. Different forms of violence, for example, group bullying/ 13 As an illustration of the difficulty of writing coherently about violence beyond the individual or collective agent(s), let us consider Žižek's (2008) commentary on violence. He distinguishes: first, "subjective violence" that is visible and agent-performed, or agent-driven, "a perturbation of the 'normal' peaceful state of things" (p. 2); and, then, "objective violence", which is "invisible" (p. 2), "the violence inherent to this 'normal' state of things" (p. 2). He notes that objective violence is of two kinds: "'symbolic' violence embodied in language and its forms" (p. 1); and "'systemic' violence, or the often catastrophic consequences of the smooth functioning of our economic and political systems." (p. 1). He then asserts that three modes of violence are presented: subjective, objective, symbolic (p. 10). This is somewhat confusing. 14 Nixon (2011: 2) sees slow violence as "a violence that occurs gradually and out of sight, a violence of delayed destruction that is dispersed across time and space, an attritional violence that is typically not viewed as violence at all." Lee (2016: 106) specifies environmental violence as: "(a) the violence between people(s) over natural resources; (b) environmental policies that can be violent against people; (c) the secondary violence from the natural world (… excess earthquakes, tsunamis, heat waves, and hurricanes) as a result of human degradation of the earth; and (d) direct damage to the environment by humans that threatens their own survival." mobbing, rape and sexual assault, and criminal damage of property, have specific features and characteristics. At the same time, the violence domains outlined (see Appendix) overlay each other and to a large extent reinforce each other. Indeed, rather than being on a continuum, violences -the violence domains -are sedimented upon each other, in both directions: killing underlain by damaging, diffuse and dispersed violences, and dispersed violence underlain by deadly, damaging and diffuse violences.

Autotelic ontology and violence as inequality
Second, and more specifically, our theoretical approach suggests an engagement with an autotelic ontology of violence (Schinkel, 2010), whereby violence is done for itself, and violence begets further violence. Previous violence is a predictor of subsequent violence -rather than being only an expression of other inequalities. The regulation of violence cannot be reduced to individual psychological traits or dysfunctional families or institutions (Strid et al., 2013). Thus, we question whether violence is always to be explained by "something else". Violence can be a social and societal inequality in its own right (Hearn, 2013;Hearn et al., 2020).

Violence, sociality and social relations
Third, understanding violence as outlined forces a shift from seeing violence as individual, incidental, aberrant, isolated and exceptional, such that all four domains of violence outlined are important for social theory. This wider framing of violence highlights historical and structural forms and impacts of violence. Indeed, the fourth domain of dispersed violence, of what is (often) not yet seen as, accepted, measured or politicized as violence is especially important in terms of social theory. This raises questions of: what constitutes violence, including systemic violence, transnational violence, colonialist and capitalist violence, slow and environmental violence (including associated gender-based violence), violence to non-humans, and symbolic and epistemic violence. This fourth domain of violence regimes connects with those approaches to social theorizing where violence is a more fundamental constituent element of sociality and social life (Hanssen, 2000;Lawrence & Karim, 2007;Schinkel, 2010;see Evans & Carver, 2017, for inciteful critical summaries of key social theorists).
Violence regimes thus link closely with analysis of social relations, as in the question of the place of violence in social relations, such as capitalist, patriarchal or colonial social relations, as formative relations of society and across societies. In New reflections on the revolution of our time, Laclau (1990) refers to the characteristics of social relations: contingency, power, politics, and historicity. Within this mix, violence is implicit, and indeed social relations are often produced or underlain by violence. Social relations, violent social relations, can be cause and context of specific acts of violence, as well as an avenue to the expansion of specific acts of violence. Social relations can even be (seen as) violence.
More specifically, social relations are constituted in and maintained by contested ideology and hegemony, whereby social relations are sedimented into inflexible structures, practices and beliefs. Phelan and Dahlberg's (2011: 27) characterization that ideology "is present when a particular discursive system, such as neoliberalism, is seen as 'all there is', its hegemonic logics having become so naturalised and sedimented that the political … conditions of its initial constitution are no longer socially recognized." Seen thus, violence acts to produce and underly the economy, polity and civil society, their social organization, and social relations.

Violence and power: a contested boundary
A fourth issue for social theory is that the violence regime approach engages with, and in some ways problematizes, is the question of distinctions of violence and power. Although violence is certainly part of some power relationships, power and its exercise are not necessarily violent/violence. As one example, Foucault (1983: 220) made a distinction between "a relationship of violence" -that "acts upon the body or upon things" that "forces … bends … breaks … destroys, or … closes the door on all possibilities" and "a power relationship" that "can only be articulated on the basis of two elements which are each indispensable …", whilst holding that violence can play a constitutive role in generating power (cf. Arendt, 1970, on power and violence are opposites; also see Menge, 2019). He goes on to speak of how in a power relationship "the other" needs to be recognized as a subject that acts, and that such a relationship may also open up and lead onto "a whole field of responses, reactions, results and possible interventions …". Foucault also argues that the opposite pole of violence "can only be passivity" -but passivity can also often be violent, as in neglect, condoning, collusion, complicity, delegation and derogation of responsibility, and thereby violence.
Within a violence regime approach, this (violent) passivity (or passive violence) can certainly sometimes be considered violence, making a clear distinction between violence and destructive power at the least contested; that contestation may itself be seen as a regime of truth. 15 Just as hegemony is not the same as domination (Buttigieg, 2005: 37), so violence is not the same as power, which can be facilitative and circuitous (Clegg, 1989), and potentially not violent. Without the wider reach of violence regime, the study of violence may remain at the level of the more immediately physical, as well as reproducing methodological nationalism, statism, and global-Northern-centred, and various forms of slow, environmental, and non-human violences. The challenges posed by blurring of violence into control, power and dominance should not inhibit analysis, and certainly not on political grounds.

Materiality-discursivity
Fifth, different violence regimes are likely to entail different understandings of violence, how broadly violence is understood, and what forms of violence are included and excluded -and what is to counts as violence -in the first place. As discussed, much violence is unseen and unrecognized. Historically, an ever-increasing range of forms of violence can be 'seen'. Activities not previously recognized as violence are now often seen as such. Neglecting the experiences of the victim/survivor/violated unseen limits (feminist) knowledge production on violence (Hänel, 2021). Thus, there are in effect that there are differential (violent) truths of violence, and regimes of truth. How truths are constructed by and through violence including the frequent exclusion of the truths of the violated, sometimes, but not only, through killing.
Violence is crushingly and fleshly material, transgressing bodily integrity, yet at the same time violence is discursively constructed, including the very questions of what is violence, and what violences are. The more that violence occurs, the greater the number of violent occasions or the greater the intensity of violence, the more that violence is likely to be taken-for-granted and normalized -by individuals, agencies, even whole societies. And of course, the greater the awareness of violence, the more that violence is likely to be identified. Concomitantly, the doing of violence reduces, sometimes to the point of obliteration, the voices (discourses) of the violated. As discussed by Hartman (1997) in relation to slavery, physical violence to slaves, even the self-sacrifice of slaves, leaves undocumented absences in and of the ontology of violence, compared with, say, recorded convictions for murder and violent crime. Destroyed lives, especially less valued or unvalued lives, leave less or no traces for the subsequent construction of social theory and other material discourses. Moreover, enactment or threat of violence can be the direct exertion of control and domination or such violence can mean that further direct physical violence is 'unnecessary' to maintain control and domination; this raises a major problem for measurement and empiricism. Diverse material-discursive constructions of violence work against empiricism, or at least any simple version thereof. 16

Conclusion
This paper has addressed two research questions: How can the extent to which specific forms and aspects of violence co-vary and interconnect be usefully conceptualized in a more inclusive, more comprehensive way? And what difference would a more inclusive, more comprehensive conceptualization of violence make to social 16 The relations of materiality and discursivity constitute a vast and contested theoretical terrain (Akrich and Latour 1992;Alaimo and Hekman 2008;Haraway 1988Haraway , 1992Hearn 1992Hearn , 2014van Eeden 2017). In this perspective, power-knowledge are multi-dimensional, spanning one-dimensional behavioural, two-dimensional non-decision-making, and three-dimensional structural (Lukes 1974) approaches, as well as those inspired by poststructuralism (Foucault, 1980;Lukes, 1986), post-poststructuralism (Johnson 1986) and post-constructionism (Lykke 2010). For focused discussion of the material-discursive in relation to violence, see Hearn (1998b) and Hirschmann (2003). politics, empirical studies, and social theorizing on and against violence? In responding to these questions, we have presented the concept of violence regime -as both a way of examining interconnections of violence, and an open-ended conceptualization of recognitions/non-recognitions of different violences. We have outlined the usefulness of the concept for social politics, social analysis, and social theory.
First, in terms of social politics, the concept of violence regimes is a way to describe and evaluate policy and politics around violence, and anti-violence, within a given socio-spatial territory, as well as pointing to gaps in policy, and ways forward for political intervention and the development and refinement of policy. Violence regimes also include attention to social movements and activism around violence, with their varying "strength", either way, promoting or opposing violence. Second, in social analysis, violence regimes can act as a framework for empirical study of different violences and activities around violence and anti-violence, relations between them, both within a given society or other cultural arena, and between societies for comparative or transnational study. Third, the concept of violence regimes assists theorizing of the place of violence and violation in social theory, by linking key issues in social theorizing to violence and dynamic material-discursive contestations over the meaning and recognition of violence. Additionally, the regime approach assists theorizing across online/offline, in that a regime approach does not require upholding a boundary between online/offline, as if they were two distinct "places" for violence, but rather recognizes the blurring of such boundaries. In these ways, we seek to contribute to enhanced, sustainable and substantial theoretical, empirical and cross-disciplinary analyses of violence and violence regimes.
Finally, a further implication of the approach to violence presented suggests a three-dimensional regimes framework -social politics, social analysis, social theory, with associated ontological, methodological and epistemological dimensions. Indeed, with the violence regime approach, violence shifts from being something to study as a specific separate field to being a point of departure to study other phenomena -such that violence becomes as an organizing principle that makes visible relations of forms and aspects of violence.