Endemic Sexual Violence and Abuse: Contexts and Dispositions

Endemic sexual violence and abuse has been observed in a number of specific circumstances, most notably conflict zones, remote and marginalised communities, and religious and state institutions. In this article we examine several documented examples and argue that a similar set of causal processes are at work in all of these otherwise apparently disparate circumstances. Rather than construing the problem as ‘organised’ sexual abuse, we present the problem in terms of the breakdown (or disorganisation) of usual individual, situational and ecological constraints.


Introduction
Dramatic increases in the incidence and prevalence of sexual violence and abuse (SVA) have been observed in some specific circumstances: in particular places and often over particular periods of time. Understanding the causes of these 'outbreaks' of SVA has been made more difficult by the apparently disparate circumstances in which they are observed to occur. Is it possible that the same mechanisms are at play in such otherwise diverse settings as conflict zones, remote and marginalised communities, and religious institutions? In this article we conjecture that this may indeed be the case.
Briefly, our hypotheses are as follows: 1. the potential to engage in SVA is virtually universal, is biologically based, and is particularly strong among adolescent and adult males; 2. this biologically-based potential to engage in SVA is widely inhibited by individual and social controls.
IJCJ&SD 112 Online version via www.crimejusticejournal.com © 2015 4(2) It follows from these two propositions that: 3. SVA is more likely to occur in contexts in which individual and social controls are weak or absent; and 4. endemic SVA will occur in social and situational conditions that induce emotional arousal, create social pressure, and weaken moral restraints.
We begin by outlining our theoretical understanding of SVA. We then examine several documented examples of endemic SVA and, finally, illustrate how the same mechanisms may apply in these otherwise apparently disparate settings.

Explaining sexual violence and abuse: An individual-situational-ecological formulation
We have set out our theory of SVA in detail elsewhere (Smallbone 2006;Smallbone and Cale in press;Smallbone, Marshall and Wortley 2008), and deal with only a few key points here. Our theory builds on Marshall and Barbaree's (1990) original integrated theory, whereby sexual offending of all kinds was understood to involve the interaction of individual (biological and developmental), situational, and sociocultural factors. In short, Marshall and Barbaree argued that positive socialisation is necessary to inhibit a universal biologically-based potential for adolescent and adult males to engage in self-serving and aggressive sexual behaviour. Males whose socialisation experiences have not embedded in them a deep inhibition against such behaviours are therefore particularly susceptible to sociocultural environments that permit or excuse SVA (for example, negative attitudes toward women), and to transitory disinhibiting factors such as anger, loneliness, intoxication, and anonymity. Sexual offences per se occur when the offender's proclivities or vulnerabilities interact with opportunities (for example, presence of a 'suitable' victim) and conducive situational conditions (for example, low perceived risk of detection).
In concert with Marshall and Barbaree's theory, and notwithstanding individual biological differences (for example, genetic variations; head injuries; drug dependence), our theoretical position assumes a more or less universal biologically-based potential for SVA among adolescent and adult males. Generally, males are more biologically predisposed to employ aggression in the service of social, and especially sexual, goals (Archer 2004;Panksepp 1998). Important gender differences are also apparent in the biological organisation of emotionmotivation systems associated with nurturing behaviour. Certainly males can learn to become very effective caregivers, but nurturing motivations for males are more instrumental (Panksepp 1998) and are therefore more susceptible to competing motivations, including self-serving and sexual motivations. We argue that these universal biological foundations of male social and sexual behaviour, particularly their stronger predisposition to sexual aggression and weaker predisposition to nurturing, help to explain why males are more likely than females to exploit vulnerable others for sexual purposes, and why sexual motivations are more likely to arise for males in everyday social interactions with vulnerable others, including children.
Also along with Marshall and Barbaree, we argue that positive socialisation is required to counter this biologically-based potential to engage in SVA. Thus in our view SVA does not need to be learned, because its potential is hard-wired. At some point in the past a disposition to rape or to select young sexual partners has presumably brought evolutionary benefits in that it has increased the probability that perpetrators' genes will be reproduced (Landolt, Lalumiere, and Quinsey 1995;Laws and Marshall 1990). Some have even argued that sexual engagement with young males may have served important social functions such as group cohesion, enculturation and protection (see Rind and Yuill (2012) for a comprehensive anthropological and evolutionary analysis of heterosexual and homosexual 'hebephilia'). What has changed is not human nature, but the social rules and norms concerning sexual behaviour.
IJCJ&SD 113 Online version via www.crimejusticejournal.com © 2015 4(2) Offenders almost certainly learn from committing such offences, of course, and persistent offenders may accordingly acquire more or less exclusive preferences for specific deviant sexual behaviours. Such offenders may also develop preferred methods of targeting victims and avoiding detection. Nonetheless, we argue that SVA-related motivations per se do not need to be originally learned. Rather, what needs to be learned is (a) the capacity for self-restraint; and (b) responsiveness to social controls, so as to constrain sexual behaviour in accordance with social rules and laws concerning age-appropriateness, mutuality and consent. Key mechanisms of selfrestraint are emotional self-regulation, empathic concern and moral reasoning. Key mechanisms of social control are positive interpersonal and social attachments.
In most circumstances these individual and social constraints are remarkably effective: proportionally very few individuals in modern Australian society, for example, engage in SVA, at least not beyond the legally defined threshold. Even then, those who are caught and prosecuted for sexual offences are a very diverse population, ranging from the psychologically ordinary to the dangerously psychopathic. Once caught, most sexual offenders are not re-arrested for further sexual offences, although a small proportion -possibly 10-15 per cent -are persistent, serial offenders. The greatest risk arises when highly disposed offenders encounter highly conducive situations. However, precisely because sexual behaviour is so sensitive to social and situational cues, within highly conducive situations SVA-related motivations may also be elicited in otherwise ordinary individuals. The most conducive social situations are those that not only present opportunities to engage in SVA and avoid detection and punishment, but that also induce emotional arousal, exert social pressure, and weaken moral restraints. Figure 1 summarises the theoretical framework used here. It shows that SVA dispositions vary, although we assume that few people lack any disposition. These dispositions will be modified by rules that are internalised during early socialisation, in most cases inhibiting the expression of any disposition to committing SVA. The socialised individual goes on to meet specific situations that provide immediate stimuli relating both to their underlying dispositions and to the rules governing behaviour in that situation and, in some cases, these will provoke SVA dispositions and/or set rules that permit or encourage it. The behavioural consequences of immediate dispositions depend on opportunities for their expression, which may or may not be present; for example suitable victims of SVA may or may not be readily available. Whether the opportunities available to those with the disposition to exploit them are taken depends on the balance of expected positive or negative consequences. The perpetrator's continued participation in SVA depends on the actual consequences.
For the most part the processes shown in the main body of Figure 1, which can both filter out and in some cases incite SVA, are linear, moving from disposition through to actual consequences, as indicated by the direction of the upper horizontal arrow. It is important, however, to recognise also that there are feedback processes, as indicated by the lower arrow; for example where actual consequences inform expected consequences, or where opportunities produce rationalisations legitimating SVA and undermining learned rules that would otherwise prohibit it.
Additionally, Figure 1 distinguishes more 'distal' causes from more 'proximal' ones. Distal causes are temporally removed from the situations in which behaviour takes place and proximal ones comprise features of the immediate situations in which actions occur. Both distal and proximal causes operate within and are in turn shaped by broader physical, social and cultural conditions, which also embrace economic, organisational, community, structural and systemic circumstances that operate on individuals and groups.

Conflict zones
Dramatic increases in the incidence and prevalence of SVA have been observed in recent conflict zones around the world. Spikes in SVA have almost certainly occurred in these types of circumstances over the course of human history and indeed are probably much less prevalent now than in previous eras (Pinker 2011). Nevertheless the issue has attracted significant attention in recent times, particularly since the mid-1990s following the Rwandan genocide and the ethnic conflict in Bosnia-Herzegovina (Mullins 2009;Snyder, Gabbard, May and Zulcic 2006).
It has been estimated that over 350,000 Rwandan (mainly Tutsi) women were raped during the genocide or later in refugee camps (Bijleveld, Morssinkhof and Smeulers 2009), and that up to 50,000 mostly Bosnian Muslim women were raped by primarily Serbian military personnel during the Bosnian conflict (Snyder et al. 2006). A high prevalence of rape and other forms of SVA has also been reported in armed conflicts in the Democratic Republic of Congo, Sierra Leone, Nigeria, Sudan and Ethiopia (Arieff 2009), and in other regional conflicts. As part of the Arab Spring revolutions of this current decade in the Middle East, female protestors in Egypt's Tahrir Square were subjected to forced 'virginity tests' and other forms of sexual harassment and assaults. While media and academic attention has cast the problem mainly in terms of sexual violence directed towards adult women, very young females and children have also been victimised. Rapes against children by Gaddafi soldiers were regularly reported in Libya (Marcus 2011), for example.
In conflict zones, normal rules for conduct, individual moral self-censuring, and formal and informal social controls often alter or cease to apply. Armed combatants are exposed to extreme violence, stress and trauma, and are often brought into contact with vulnerable, sometimes displaced, women and children with limited means to protect themselves. External scrutiny is often limited due to safety concerns for United Nations or other potential international monitors.
Armed conflict alone does not explain these upsurges in SVA: indeed, significant variations in SVA have been noted both across different conflicts and between groups within the same conflict (Human Security Report Project 2012). Wood (2006) reported that, in comparison to Rwanda and Bosnia-Herzegovina, for example, conflicts in Israel/Palestine and Peru involved considerably lower levels of SVA, and the prevalence of SVA during the conflict in El Salvador IJCJ&SD 115 Online version via www.crimejusticejournal.com © 2015 4(2) varied over time, decreasing over the period of the war. Using a cross-national dataset of 86 civil wars from 1980, Cohen (2013 identified 18 conflicts with widespread rape, 35 with elevated sexual violence, 18 with only isolated reports, and 15 with no reported sexual violence. Wood (2008Wood ( , 2009 proposed four social processes that affect the likelihood of escalating SVA in armed conflicts: (1) strategies used by military hierarchies and leaders (for example, expectations of behaviour standards, prohibition or reinforcement of behaviours, recruit training); (2) the socialisation processes adopted by combatants (for example, culture of obedience, diffusion of responsibility, depersonalisation of the victim group); (3) social norms and beliefs that existed prior to the conflict, or that develop and change as the conflict unfolds; and (4) group dynamics. Even in otherwise violent and chaotic war contexts, it remains possible for positive social controls to be maintained, and therefore for non-combat-related violence such as SVA to be kept in check. Wood (2009) provides an example of this, arguing that, despite the frequent use of other serious forms of violence including mass civilian killings, the apparent absence of SVA perpetrated by Liberation Tigers of Tamil Ealam (LLTE) combatants in Sri Lanka may be explained by the maintenance of pre-existing Tamil social norms, including the prohibition of sexual relationships between unmarried persons. These norms were upheld and strictly enforced by LLTE leadership through a strict code of conduct, consistent and firm discipline, and the systematic training of new recruits.

Abu Ghraib: A conflict zone within a conflict zone
In 2003, US soldiers tortured and sexual abused Iraqi prisoners held in Abu Ghraib Prison in Baghdad. This received worldwide attention in 2004 following media reports and the publication of explicit photographs documenting the abuse (Hersh 2004). Reported abuses included threatening prisoners with rape, actual rape of prisoners, digitally recording naked prisoners, arranging prisoners in sexually explicit positions, forcing them to masturbate publicly and digitally recording this, and at least in one case attaching electrical wires to a prisoner's penis to simulate and threaten torture (Taguba 2004). It is suggested that thirty children or more aged 10 to 17 years were housed among the many prisoners at Abu Ghraib (Zimbardo 2007), though it is unknown whether any were included in the kinds of abuses depicted in the photographs.
These and other abuses were primarily committed by Army Reserve soldiers of the 372nd Military Police Company, during the night shift on Tier 1A of the prison. Seven of these reservists were prosecuted for their actions, with most serving prison sentences. Other military personnel were reprimanded and disciplined in relation to their misconduct associated with these events (Zimbardo 2007). The official government response to the atrocity was to attribute blame to individual soldiers, directing the focus particularly to the seven publicly implicated soldiers and away from broader systemic factors. In what became known as the 'bad apple' versus 'bad barrel' debate, independent commentators argued instead that the situation itself was to blame for the behaviour of these ordinary individual soldiers (Lankford 2009;Zimbardo 2007).
The individuals involved were all low-ranking reservists who were overworked, stressed, and had limited or no training for the job they were given at Abu Ghraib (Zimbardo 2007). They endured numerous pressures and hardships, including rationed water, no sewage, few cleaning supplies, inconsistent electricity supply, an overwhelming workload, regular mortar attacks, and threats from detainees. Against this backdrop existed a military culture that condoned the abuse of prisoners. Tier 1A contained the most dangerous prisoners, often the subject of interrogations by military and civilian intelligence teams who used a range of tactics to prepare prisoners for interrogation, including keeping them naked, making them wear female underwear, and using dogs to intimidate them.
IJCJ&SD 116 Online version via www.crimejusticejournal.com © 2015 4(2) The army reservists responsible for guarding these prisoners were specifically directed to 'set the physical and mental conditions for favourable interrogation of witnesses' (Taguba 2004: 18). They were advised that different rules applied to Abu Ghraib prisoners, consistent with the brutality towards other prisoners actioned at Guantanamo Bay and in Afghanistan (Human Rights Watch 2005;Lankford 2009). Bystander inaction at Abu Ghraib reinforced these norms, with numerous soldiers reportedly witnessing abuse but taking no action to address or formally report it. Reports also suggest a lack of action when these abuses were questioned (Zimbardo 2007).
As in conflict zones more generally, the involvement of multiple perpetrators at Abu Ghraib cannot be explained by individual psychopathology. It is simply not tenable that the armed combatants in Rwanda, Bosnia-Herzegovina and many other places, or the army reservists at Abu Ghraib, brought with them to those situations a peculiar predisposition to sexually assault enemy combatants and civilians. Indeed in the case of Abu Ghraib, prior military psychological testing detected no special psychopathology or other personality traits that would foreshadow the abuses that were later to occur (Zimbardo 2007). The only viable explanation is that there was something about the situations in these places and at these times that enabled or precipitated SVA in otherwise ordinary men (and, in the case of Abu Ghraib, also women). As we discuss below, it seems that other, quite different kinds of situations have a similar effect.

Remote Indigenous communities
In recent decades a series of National, State and Territory inquiries has repeatedly drawn attention to high rates of SVA in regional and remote Australian Indigenous communities ( While these average findings are disturbing, it is important to remind ourselves that they mask significant variations: some Indigenous communities appear to suffer considerably higher than average prevalence of SVA, and some presumably experience the problem at or below rates in the broader community. Thus it is not 'Indigenous communities' where endemic SVA is to be found, but specific communities, and then perhaps only at particular times. Unfortunately, because so much of the research and commentary on this issue is problem-focused, little attention has been given to safe communities and consequently much less is known about aspects of Indigenous community life that protect against high levels of SVA and other violence. Similarly, much less is known about variations within communities themselves, and why it may be that some individuals, families, groups, or places are at greater or lesser risk than others. As with the dramatic increases in SVA in some combat zones, endemic SVA in some Indigenous communities cannot be explained on the basis that large numbers of male community members possess a specific sexual psychopathology that predisposes them to such behaviour. Instead, most inquiries and reports have pointed to contextual factors: breakdowns of Aboriginal culture and customary Lore; exposure to, and normalisation and tolerance of, various forms of violence; general social disadvantage (for example, poverty, poor education, high unemployment, inadequate housing, poor health, high mortality rates); early exposure of children to sexual activity; community isolation; significant barriers to reporting abuse; failed statutory and professional responses when abuse is reported; and so on. More immediate situational factors IJCJ&SD 117 Online version via www.crimejusticejournal.com © 2015 4(2) proposed to contribute to SVA in these communities include intoxication, the presence of unsupervised, vulnerable children, and reluctance by other community members to intervene or report the behaviour ( Sutton (2001Sutton ( , 2009 argues that some originally functional aspects of traditional Indigenous culture and society are no longer functional in the context of the present day environment. Examples may include some pre-colonial child-rearing practices (for example, those that emphasise the autonomy of young children), and kinship obligations that present barriers to intervening informally or reporting problems to authorities. Desensitisation of whole communities to endemic SVA may create a vicious cycle of further diminishing community guardianship and action in response to specific concerns.
Whatever the contribution of specific local cultural factors to the problem of SVA in some Indigenous communities, it is clear that these are neither a necessary nor sufficient condition for endemic SVA to occur. We have already seen this in our examples of conflict zones and in the extraordinary situation at Abu Ghraib. We now turn to another example of endemic SVA in a very remote community, but one with no connection at all to Australian Indigenous culture and society.

Pitcairn Island
Pitcairn Island, a small, remote island in the South Pacific Ocean, is a British Overseas Territory administered from New Zealand. In 2000, a New Zealand police investigation code-named Operation Unique uncovered child sexual abuse on Pitcairn Island spanning at least 40 years and involving almost the entire population (47 at the time of the trials in 2004). These abuses included violent rapes, gang rapes and assaults on children as young as five, committed by adult and adolescent male residents. As a result of the police investigation, 13 men, representing a large proportion of the adult male population and including seven still living on the island at the time, were arrested and charged with a total of 96 sexual offences under British law. Most of the accused were found guilty and sentenced to prison terms or community orders (Marks 2008;Power 2007;Vaughan 2006).
Pitcairn Island is situated midway between New Zealand and Chile, virtually in the middle of the Pacific Ocean. It has been described as 'possibly the world's most remote inhabited spot' (Marks 2008). The Island is accessible only by sea but, without a safe anchorage, local longboat drivers must escort all visitors through difficult waters to reach the Island. Neither Britain nor New Zealand appears to have played significant operational or monitoring roles, with only sporadic and time-limited official visits to the Island, minimal oversight of statutory responsibilities, and little to no involvement in daily management and administration. Pitcairn's geographic isolation and distant governance structure effectively shielded the community from external influence and scrutiny.
In response to Operation Unique information emerged of an ineffective local justice system (local police and court personnel were unqualified and untrained) and lack of statutory or legal responses when concerns were raised about suspected sexual abuse. Community leaders, including men in powerful local political and official positions, were among the accused. General desensitisation to sexual abuse and lack of guardianship were also reported, with some parents normalising or minimising the behaviour. Some women argued that sexual activity between local men and 12-13 year-old girls was natural, and reflected their Polynesian heritage. Some IJCJ&SD 118 Online version via www.crimejusticejournal.com © 2015 4(2) placed pressure on children to withdraw complaints. Many of these women had experienced similar behaviours in their own childhood and had come to perceive this as a normal part of life. Even some external professionals, including a school teacher and church pastor, colluded with the alleged offenders and/or ignored suspicions and explicit reports of sexual abuse. The children themselves reported feeling powerless to speak out, with accounts of inaction by local officials and fear of retribution and punishment when they did (Marks 2008(Marks , 2012Power 2007).
Even with the tacit (or open) approval of others on the Island, sexual abuse incidents still tended to occur away from others' direct scrutiny. From the limited detailed information on specific incidents, it appears that abuse typically occurred in private or hidden locations on the island (for example, banana plantations, gardens, in the bushes, in an old hut) where observation by others was less likely. Whilst some surprise attacks were reported, it also appears that young girls were invited to isolated locations, sometimes with the knowledge of their parents, to complete chores (for example, collecting firewood), ride bikes or engage in other games as a pretext for planned sexual abuse.
As is usually the case with sexual abuse, official charges are unlikely to represent the true extent of the problem on Pitcairn Island. Indeed many of the women who defended the accused men by normalising the behaviour did so by acknowledging their own experiences of SVA, indicating that the problem had been occurring for at least two generations. In some respects this resembles the community-wide accommodation of SVA -or at least the special dilemmas of reporting it -seen in some remote Australian Indigenous communities. The common thread so far in our examples, however, has been a breakdown in external influence, albeit for somewhat different reasons. We now turn to our final example -sexual abuse in the Catholic Church -and find here, again for somewhat different reasons, a breakdown in external scrutiny that enabled SVA to occur without fear of personal (and in this case organisational) exposure and punishment.

Sexual abuse in the Catholic Church
Public attention to sexual abuse in the Catholic Church grew as increasing numbers of affected people came forward with complaints in the 1980s and 1990s. The issue has come to be recognised internationally, with investigations in Australia, Austria, Belgium, Canada, France, Germany, Ireland, Mexico, Norway, Poland, the UK and the US. Numerous public inquiries have revealed extensive sexual abuse and exploitation by Catholic priests, deacons and other officials, and have called into question the often secretive and obstructive responses of the Church to abuse allegations. There is no reason to believe that the behaviours have arisen only recently. Rather they have surfaced, been highly publicised, and led to investigations by police services, and all this has created conditions in which large numbers of victims feel willing and able to come forward and expect their reports to be taken seriously.
An empirical study of sexual abuse allegations involving US Catholic priests and deacons, conducted by researchers at John Jay College (Terry 2008), provides the most comprehensive analysis. With the support of the US Council of Bishops Terry and her colleagues obtained information from almost all American dioceses relating to alleged abusers, victims, and abuse incidents, covering the period 1950 to 2002. Results showed that, over this 50 year period, allegations of abuse had been made by 10,667 individual victims. Over 3,000 additional potential victims who did not make an official report were also identified. The allegations involved 4,392 priests, four per cent of all priests in active ministry in the US in that period.
Two details in the John Jay study are instructive for our present purposes. First, records indicated that offending priests had on average begun abusing approximately 11 years after their ordination (Terry 2008). This timeframe generally coincided with moves into parish IJCJ&SD 119 Online version via www.crimejusticejournal.com © 2015 4(2) residences, with less day-to-day involvement with peers and less direct supervision from superiors, and with new responsibilities to engage with children and their families for routine religious, teaching, pastoral and other purposes (Terry and Ackerman 2008). In other words, sexual abuse generally commenced with fortuitous opportunity, rather than through the elaborate manipulation of circumstances.
Second, after a steady increase in the 1950s and 1960s, alleged abuse incidents peaked in the 1970s and then steadily declined. The peak year was 1970 when there were more than 400 new reports of abuse. By contrast, in each year from 1994 to 2000 there were fewer than 50, with half this again in the last two years of the study period. This steady decline in reported cases is inversely related to the number of persons making formal complaints: in the 1970s there were 266 reports; in the 1980s there were 1146 reports; in the 1990s, 4022; and, in just the three years 2000-2002, there were 4533 reports. These trends show that, as more people complained about abuse, the less abuse occurred. One interpretation of these data is that, over time, the prospect of abusing presented a much greater risk of exposure and punishment than had previously been the case.
While the responses of various diocese varied -those that took direct action in response to allegations generally evidenced fewer allegations than others (Terry 2008) -a general culture of self-protection appears to have predominated, with now familiar stories of leaving accused priests in ministry positions involving ongoing contact with children, transferring them to new parishes ignorant to the risk, and dissuading victims from reporting abuse to police (Dale and Alpert, 2007). What seems to have been needed was to call the Church and other religious institutions to public account. The Nolan (2001) and Cumberlege Reports (2007)

Discussion
We originally developed our theoretical formulation to explain the more frequently detected forms of SVA in which a single offender assaults or abuses a single victim without the direct knowledge or collusion of third parties, though it also accommodates circumstances in which small groups of offenders act together. The theory deals with the questions of why SVA first occurs separately from why some abusers go on to become persistent and sometimes serial offenders (Smallbone and Cale in press). The present article is our first attempt to apply the theory to endemic SVA.
Our theory asserts, perhaps uncomfortably for some, that the potential for SVA motivations and behaviour are more or less universal, particularly among adolescent and adult males. The remarkable over-representation of males as perpetrators of SVA is, we have argued elsewhere (Smallbone and Cale in press;Smallbone et al. 2008), largely the result of biological gender differences (Panksepp 1998), though differential socialisation is also likely to play a role. Certainly positive social cognitive development is crucial to establishing the individual and social constraints necessary for inhibiting the expression of the potential for SVA and other antisocial behaviour. Because socialisation experiences vary widely, and probably also because of biologically-based variations, dispositions for SVA vary. Still, in conducive situations, many people are susceptible to SVA motivations and behaviour.
A common theme across all our case studies is that endemic SVA cannot be adequately explained by the coincidental presence of groups of individuals with unusually strong IJCJ&SD 120 Online version via www.crimejusticejournal.com © 2015 4(2) dispositions for rape and sexual abuse. Thousands of highly disposed rapists did not all at once come together in the Serbian army or involve themselves in the Rwandan conflict. And nor, we presume, did thousands of paedophiles join the US Catholic Church ministry in the 1960s and 70s, just as it is unlikely that the high levels of SVA in some remote Aboriginal communities are the result of some peculiar sexual psychopathology afflicting Aboriginal men and boys.
It seems clear that it is something about the situation that is at the root of the problem. Our integrated theory proposes two distinct but complementary ways in which situations affect SVA (Smallbone and Cale in press;Smallbone et al. 2008;Wortley and Smallbone 2006). First, situations can present opportunities that an already-motivated or criminally-inclined offender manipulates or exploits. In this respect the most suitable situation is one that provides the would-be offender easy access to a suitably vulnerable victim and where the risk of detection and punishment is low. Second, certain situations contain dynamic properties that can precipitate SVA-related motivations that would not otherwise occur, at least not at that time or place. This second role of situations follows the formulation set out by Wortley (2001), whereby situations can present cues that prompt a criminal response; they can exert social pressure to conform to situational norms or expectations; they can weaken moral restraints, and so permit a criminal response; and they can induce emotional arousal and thereby provoke a criminal response.
In the context of endemic SVA, specific incidents may be prompted by cues associated with the presence of vulnerable and unprotected women and children (or in the case of Abu Ghraib, male prisoners). Social pressure is likely to be created when 'everyone is doing it', or perhaps particularly when a friend, colleague, fellow combatant, or superior is involved. In conflict zones and some remote Aboriginal communities there may be a pervasive breakdown of usual rules and social controls that weaken moral restraints and signal permissibility of SVA and other violence. In other circumstances moral restraints may be weakened by collective justification or moral neutralisation, as may have occurred on Pitcairn Island and within some sections of the Catholic Church. Finally, beyond the emotional arousal associated with sexual desire itself, many of the circumstances in our case studies will have involved a range of other relevant emotions: stress, fear, anxiety, anger, hostility, vengeance, shame, guilt, loneliness, and so forth. Such strong emotions may provoke 'out of character' aggression, including sexual aggression, and in some cases perhaps even self-soothing sexual motivations.
The situations in which specific SVA incidents occur do not arise randomly. According to our integrated theory, risky situations for SVA arise from various levels of the offender's and victim's social experience: their family, peer, organisational, neighbourhood and sociocultural environments. In this respect our theory aligns with public health approaches to SVA (Krug, Dahlberg, Mercy, Zwi and Lozano 2002). Such approaches situate individual offenders and victims within their natural ecological context, and locate risk and protective factors at various levels of the ecological systems in which individuals develop and live. Thus the causes of SVA exist not just within individuals, but also within the family, peer, organisational, neighbourhood and sociocultural systems within which they are socially embedded.
While all of these aspects are likely to influence individual behaviour, in our case studies the most relevant appear to be peer and organisational systems. By definition, endemic SVA involves groups of offenders and thus peer systems are key to creating social pressure and signalling permissibility of the behaviour. This appears to be the single common thread linking all our case studies. Organisational systems appear especially important in church abuse: on the one hand, the persistent failures of church officials to fully investigate and respond to complaints, and placing the value of the church's reputation above the value of protecting children, are now widely recognised as the products of dysfunctional organisational practices IJCJ&SD 121 Online version via www.crimejusticejournal.com © 2015 4(2) and culture. On the other hand, the persistence of many complainants and the increasing scrutiny and accountability of church organisations appear to have altered church practices.
Organisational systems are also crucial in conflict zones. Variations in the extent of SVA in different conflict zones appear linked to the ways in which military hierarchies have trained, acculturated and supervised their soldiers, and responded to problems. Abu Ghraib stands as a salutary lesson of failure in these respects. Organisational systems in remote communities may also furnish conditions producing a range of risk and protective factors. Schools, for example, may unwittingly accommodate relevant problems, or alternatively provide a crucial safety net for vulnerable children; human service systems may be disengaged, poorly targeted, and of poor quality, or provide effective responses to individual, family and community concerns. Table 1 summarises the case studies presented here along lines described in Figure 1. It shows that the examples are far from identical. It makes clear, however, that what is common across all the examples of endemic SVA is not so much a series of groups somehow peculiarly disposed to SVA, but rather a set of conditions in which the rules governing the normatively prescribed and proscribed behaviours into which many of us have been inducted are absent or can be suspended (in the Catholic Church by a small proportion but a large number or priests; in other cases by much higher proportions of the men involved where the SVA is conducted more collectively or collaboratively), where there are readily available victim populations, and where real and apparent risks to perpetrators have been low. Our hypothesis is that whenever these conditions occur there will be high levels of SVA. +…---/+ +…-+…-+…-Note: '+' refers to causal forces favouring sexual violence and abuse; '+/-' refers to ambiguity in causal forces favouring or inhibiting sexual violence and abuse; '-' refers to causal forces inhibiting sexual violence and abuse; '+…-' refers to population variability in causal forces favouring or inhibiting sexual violence and abuse.

Conclusions
Research, policy and practice concerning SVA are dominated by individual-level conceptions of the problem. This perspective tries to account for SVA by assuming the existence of a limited number of sexually deviant individuals, and ignores the more widespread potential of otherwise ordinary people to engage in these behaviours. Happily, only a relatively small minority express this potential, at least at the legally-defined threshold, because internalised rules into which most are socialised usually provide effective constraints. Occasionally these usual individual and social constraints break down in such a way as to facilitate, enable or provoke endemic SVA. In our view SVA of all kinds, including endemic SVA, is caused by interactions of individual dispositions and specific situations that arise in the context of the family, peer, organisational, neighbourhood and sociocultural systems within which offenders and victims are embedded.