Abstract
Sexual homicide typically implies a crime involving two people: perpetrator and victim. Thus, multiple-perpetrator and multiple concurrent victim sexual murderers are unusual, empirically invisible due to their exclusion from (or burying within) research samples. The present study examines 21 such cases of male sexual murderers having perpetrated at least one sexual homicide either together with a co-offender or alone but killing two victims at once. The aim was to investigate just how unusual, or not, these cases may be in relation to current scientific understanding of sexual murderers and their crimes. A descriptive analysis of offenses (co-offender and offender–victim dynamics, modus operandi) and offender characteristics is presented. Main findings, that multiple-perpetrator and multiple concurrent victim sexual murderers are not so unusual in that they are well conceptualized through application of the established sexualized, grievance, and rape murderer typology, are discussed in relation to clinical and empirical implications.
Introduction
Easy as it is to find popular literature biographies of serial killers and “couples who kill”; in reality, most sexual homicide involves one victim and a lone perpetrator who does not go on to commit further homicidal crimes (James & Proulx, 2014). Thus, sexual murderers are not necessarily serial killers, and to avoid lexical ambiguity the opposite is worth noting: Serial killers are not necessarily sexual murderers. Also, although reference to murder carries legal connotations, the term sexual murderer is used in relation to perpetrators of sexual killing(s) regardless of criminal disposition. Empirical research describing sexual murderers has suffered numerous methodological shortcomings, relating to issues that are inevitable for studies of a rare phenomenon (i.e., low base rates), practical difficulties such as the availability of reliable information pertaining to this particular population, as well as limitations introduced through unclear sampling criteria and sample selection bias (Carter & Hollin, 2010; Meloy, 2000; Proulx, Cusson, & Beauregard, 2007). The evidence base once relied heavily upon criminal investigator and clinical observations (e.g., Brittain, 1970; Holmes & Holmes, 2009; Malmquist, 2006; Schlesinger, 2007), providing valuable insights, but also generating a number of generalizability questions in the absence of empirical data. The influential Federal Bureau of Investigation’s (FBI) organized/disorganized dichotomy (Burgess, Hartman, Ressler, Douglas, & McCormack, 1986; Ressler, Burgess, & Douglas, 1988) represents seminal work, but the research suffered difficulties in scientific rigor and the model has subsequently not held up to external validation (Canter, Alison, Alison, & Wentink, 2004).
However, there have been fairly large strides toward overcoming some of the methodological limitations common to this field of research. In the late 1990s, a highly representative sample (comprised using a sampling strategy that minimized selection bias) of 40 sexual murderers (of women) incarcerated in the Canadian province of Quebec was examined in a series of studies published with complementary research from a sample of 58 sexual murderers (majority adult-female victims) detained in prisons in England, UK (Proulx, Beauregard, Cusson, & Nicole, 2007). Proulx, Beauregard, et al. (2007) demonstrated a greater degree of heterogeneity among sexual murderers than had previously been empirically observed (the sample was not typified by the popular image of rampage-killing sexual predators), and at the same time greater homogeneity than may have been expected in terms of criminological, psychological, and sociological factors by comparison with non-homicidal sexual aggression. In Germany, a forensic psychiatric evaluation of 166 sexual murderers (mostly nonserial, majority adult-female victims) highlighted sexual sadism and schizoid personality traits as characteristic features (Hill, Berner, & Briken, 2018), while in Australia a sample of 85 sexual murderers (Kocsis, Cooksey, & Irwin, 2002) and in Belgium a sample of 33 sexual murderers (Gerard, Mormont, & Kocsis, 2007) contributed to the evidence base describing heterogeneity among these offenders and their crimes. Most recently, there has been research drawing upon considerably larger samples emerging from Canada (Beauregard & Martineau, 2016) and England and Wales, UK (Stefanska, Higgs, Carter, & Beech, 2017). Beauregard and Martineau (2016) compiled police data on 350 sexual homicides (including 100 unsolved cases), providing rich criminal event information. Meanwhile, the U.K. researchers constructed a database of 350 sexual murderers applying stringent inclusion criteria, focusing exclusively on adult-male, nonserial sexual murderers with female victims aged 14 and above. The data support the presence of differing pathways to sexual homicide (Stefanska, Carter, Higgs, Bishopp, & Beech, 2015) and differing psychosocial characteristics in subtypes of sexual murderers (Higgs, Stefanska, Carter, & Browne, 2017).
Sexualized, Grievance, and Rape Murder
In conjunction with these developments in the evidence base, recent systematic reviews of the extant literature have found that three types of offense are captured by the overarching term “sexual homicide”: sexualized killing, which involves a functional (often, but not necessarily, sadistic) relationship between the act of killing and sexual arousal; grievance killing, which is driven by angry schema and an excessively aggressive response style; and rape killing, in which there is only an indirect association between the sexual element of the offense and killing such as when homicide is instrumental, to silence the victim (Higgs, Carter, Tully, & Browne, 2017). While these types were derived mostly from nonserial sexual murderer samples, comparisons of serial and nonserial sexual murderers show that serial sexual murderers offend in ways that are consistent with the sexualized killing subtype, whereas nonserial sexual murderers are more typically characterized in line with the grievance killing subtype (James & Proulx, 2014, 2016).
Sexual Murder as a Unique Form of Sexual Aggression
Attempts to understand sexual homicide within the wider context of sexual aggression have tended to find more similarities than differences between sexual murderers and non-homicide sexual aggressors (Stefanska, Beech, & Carter, 2016). However, almost all of the comparison studies operationalize sexual homicide according to its broad definition (i.e., amalgamating sexualized killing, grievance killing, and rape killing), rather than attending to the argument that there appears to exist a subgroup demonstrating lethal intent in their offending behavior (Healey, Beauregard, Beech, & Vettor, 2016). Indeed, this concords with the conceptualization of the sexualized murderer (for whom the act of killing is necessary for, or is itself the source of, sexual excitation) and implies a type of sexual murder constituting a unique form of sexual aggression.
One study differentiating between subtypes of sexual murderers in comparison with non-homicide sexual aggressors found that while childhood adversity is common across the different groups, their psychosocial and criminal trajectories differ (Higgs, Stefanska, et al., 2017). Higgs and Stefanska (2018) present a multi-trajectory model of sexual aggression, suggesting that sexual murderers who most likely correspond to the sexualized killing subtype respond to early adversity by internalizing difficulties and becoming increasingly socially disconnected, experiencing difficulties in adulthood such as a lack of intimate relationships, eventually acting upon deviant, often sadistic, sexual drives. According to the model, non-homicide sexual aggressors and rape murderers respond to early adversity through a more antisocial trajectory characterized by a sexually and nonsexually violent criminal career and sexual promiscuity. Grievance murderers share some psychosocial and criminal career characteristics with both sexualized murderers and rape murderers, but likely more so the latter as their crimes do not typically involve sexually sadistic acts (Higgs & Stefanska, 2018).
The Unusual Suspects
A problem arises in the practical application of research and theory in forensic clinical practice when confronted with a case that might be seen as a “square peg” failing to fit the “round holes” in the evidence base. That is, the pursuit of large-scale studies interested in generalizable results demands predetermined sampling criteria necessarily precluding certain cases, such as those involving multiple perpetrators or more than one concurrent victim (meaning multiple victims within a single sexual homicide event, as opposed to “spree” or “mass” killings with no sexual element). These cases are either excluded from research altogether as unusual cases inconsistent with the sample inclusion criteria or simply subsumed within samples described according to the characteristics of the majority. In such instances, the presence of a co-offender or multiple concurrent victims may be treated as a modus operandi variable, or worse, omitted in the reporting of either the sample characteristics or results. This difficulty in how to adequately include unusual cases and their consequential invisibility in research leaves us without an empirical basis when tasked with clinical decisions such as intervention planning and other aspects of case management.
In multiple-perpetrator rape, the available evidence speaks to the role of group dynamics and suggests the usual presence of a more behaviorally dominant individual leading the assault while the motives of the majority tend to be nonsexual (Da Silva, Woodhams, & Harkins, 2015). Empirical understanding of multiple-perpetrator rape rests, however, upon a relatively small evidence base (Da Silva et al., 2015) and does not typically include cases of homicidal sexual aggression. We know of no similar research of multiple-perpetrator sexual homicide, and as has been highlighted, sexual murderers with multiple concurrent victims are also an unstudied group.
Study Aims and Present Research Questions
The current research was concerned with how relevant the extant sexual homicide literature may be in relation to cases involving more than one perpetrator or more than one concurrent victim. Conversely, the study was designed to reveal factors particularly pertinent or previously undiscovered in relation to such cases. Specific research questions were as follows: In multiple-perpetrator sexual homicide, are existing models of sexual homicide useful in understanding the individuals involved, or is the impact of a co-offender similar to the group dynamics observed in multiple-perpetrator rape? Under what circumstances does a sexual murderer kill more than one victim during the perpetration of a single criminal event? What are the offender and victim characteristics in these cases? The present study aims to contribute to the scientific understanding of multiple-perpetrator and multiple concurrent victim sexual homicide, with the further aim of establishing the impact of such cases if included in the samples used for past and future research and theory development.
Method
Design
The study was exploratory and descriptive in design. Data were obtained from 46 courthouses in France by searching through case files (psychiatric, psychological, and medical reports, and social, school, work, military, and prison records), investigative reports (crime scene photographs, autopsy, toxicology, anatomy-pathology, genetics, and ballistic reports), psychosocial profiles of the victims, transcripts of witness interviews, suspect interrogations, and criminal trial court arraignment orders. Criminal records were accessed using the Service Central du Renseignement Criminel (national criminal database). The study aimed to describe the offense (multiple-perpetrator dynamics, modus operandi), victim characteristics, and offender characteristics of multiple-perpetrator and multiple concurrent victim sexual murders/sexual murderers.
Sample
The sample used in the present study comprised 21 males, all of whom had received a sentence for homicide. Participants were identified through a review of homicide convictions in France between 1975 and 2012, carried out by investigators and behavioral analysts of the Gendarmerie Nationale Behavioral Sciences Unit and by prosecutors and chief registrars from different parts of France. For inclusion, cases were required to meet the Ressler et al. (1988) specification that a homicide may be considered sexual in nature when evidence or observations include: victim attire or lack of attire; exposure of the sexual parts of the victim’s body; sexual positioning of the victim’s body; insertion of foreign objects into the victim’s body cavities; evidence of sexual intercourse (oral, anal, vaginal); and evidence of substitute sexual activity, interest, or sadistic fantasy. (From the preface, p. xiii)
For research purposes, using this description as operationalization criteria for sexual homicide is an empirically supported approach, as despite limitations associated with false negatives, it has been shown to distinguish sexual from nonsexual homicide (Carter, Hollin, Stefanska, Higgs, & Bloomfield, 2016). This sampling strategy yielded an initial 212 cases of sexual homicide, subsequently scrutinized applying the specific inclusion criterion for the present study: that the offense involved two or more perpetrators and/or two or more concurrent victims. Eighteen cases were identified involving two male perpetrators with up to four concurrent female victims, and in three cases, a lone male sexual murderer killed two female victims in one place at one time. The ethnicity of the sample was mostly White: 76.2% (n = 16). The remaining 23.8% (n = 5) were from visible minority ethnicities (persons who are non-Caucasian in race or non-White in color; Statistics Canada, 2008).
Procedure and Measures
Offense summaries were reviewed by two of the authors (a forensic psychologist and a clinical psychologist) independently to determine the overall offense dynamics, with the aim of categorizing participants in the multiple-perpetrator cases as either dominant (leader) or passive (follower). Dominant was defined as follows: Perpetrator is more active in offense planning (where there is evidence of premeditation to sexually offend, whether or not intentionally homicidal); perpetrator is responsible for all (or the majority) of sexual or violent acts; and/or perpetrator verbally or physically manipulates the co-offender in the perpetration of sexual or violent acts. Passive was defined as follows: Perpetrator is minimally or not involved in any level of premeditation, or demonstrated opposition or reluctance to sexually offend; perpetrator does not engage in sexual or violent acts or only does so if influenced by the co-offender. The purpose of this was to examine whether offender characteristics might differ depending on whether a more dominant or passive role was taken in the offense.
Data were collated using the Sexual Murderers Multidimensional Inventory (SMMI; James & Proulx, 2015), which is an instrument designed to capture information on developmental, psychopathological, lifestyle, criminal career, predisposing, and precipitating factors, as well as modus operandi, crime scene, and victim characteristics from file sources. Evidence for each variable was coded dichotomously (present or absent), according to item operationalization provided by the SMMI. Inter-rater reliability for data coding using the SMMI has been found to be excellent (intraclass correlation coefficient [ICC2] = 0.82; James, Beauregard, & Proulx, 2019). Criminal history variables included both criminal convictions and charges not reaching conviction. The inclusion of non-official criminal antecedents is a recommended research strategy that aims to reduce bias associated with the underestimation of criminal behavior reflected in official criminal records (Thornberry & Krohn, 2000).
Psychological characteristics were evaluated in terms of implicit theories (e.g., “women as sex objects”), which are latent schemata that favor the emergence of cognitive distortions, such as “women are provocative.” Such latent schemata have previously been observed among sexual aggressors of women (Polaschek & Ward, 2002), children (Ward & Keenan, 1999) and in sexual murderers (Beech, Fisher, & Ward, 2005). In addition, each case was assessed for psychopathy because of the “psychopath hypothesis”—a common but poorly empirically supported theoretical supposition for understanding sexual homicide (Beauregard & DeLisi, 2018). Psychopathy was measured using the Psychopathy Checklist: Screening Version (PCL: SV; Hart, Cox, & Hare, 1995). Psychometric properties based on data from forensic samples support the use of this measure for research purposes (Higgs, Tully, & Browne, 2018). Sexual sadism was also measured, using the Sexual Sadism Scale (SeSaS; Nitschke, Osterheider, & Mokros, 2009). The SeSaS is appropriate in the assessment of forensically relevant sadism, given its focus on offense-related behavior, and has demonstrated validity for capturing moderate to severe levels of the latent trait of sexual sadism (Stefanska, Nitschke, Carter, & Mokros, 2019).
Results
Offender categorizations (dominant/passive) made by the two raters were compared, finding that in all but one case, neither perpetrator was identified by either rater as dominant or passive. Instead, offenders tended to share involvement in any premeditation, were both active in sexual and violent acts (even if in some cases one offender may have perpetrated more sexual acts than the other or may have been more strongly or uniquely implicated in inflicting fatal injury), and there was not typically evidence of one offender being coerced by the other in the lead-up to offending or during the offense. As such, data were subsequently treated at the group level for multiple-perpetrator sexual murderers (N = 18), and the three lone-perpetrator/multiple-victim cases were analyzed as single case studies.
Multiple-Perpetrator Sexual Murderers
Sociodemographic characteristics and variables describing the general and sexual lifestyles of the sexual murderers acting with a co-offender are presented in Table 1 (N = 18). The vast majority of these offenders were unemployed (88.9%) and had financial difficulties (88.9%). These men were commonly found to be socially isolated (88.9%), and in all cases co-offenders were either living together or had lived together prior to the offense. Substance misuse was frequently present, with 55.6% of the sample meeting criteria for alcoholism and 50% being drug users. General lifestyle data also frequently indicated irresponsibility (94.4%), patterns of impulsive behavior (72.2%), and the use of violence (72.2%). Sexual lifestyles often involved infidelity (66.7%) and sexual contact with sex workers (61.1%). Sexual deviancy was identified through evidence of deviant pornography consumption (33.3%) and sadistic sexual behaviors (27.8%).
Sociodemographic Characteristics and the General and Sexual Lifestyles of Sexual Murderers Involved in Multiple-Perpetrator Sexual Homicide (N = 18).
Average age at the time of crime: M = 26.67, SD = 8.75.
The most common cognitive distortions identified were “women are sexual objects” (77.8%), “women are provocative” (72.2%), and “uncontrollable sexual drive” (61.1%). In addition, “dangerous world” cognitions were frequently found (44.4%), as well as “women deserve to be punished” (38.9%), “women are dangerous” (27.8%), and “women are untrustworthy” (27.8%). The results of PCL: SV assessments (M = 16.89, SD = 3.80) indicated that at least 77.8% of the sample scored above the threshold associated with psychopathic personality functioning based on available normative data (psychopathy is indicated at scores of 14-16 in European samples; Dolan & Blackburn, 2006; Urbaniok, Endrass, Rossegger, & Noll, 2007). Scores were high for interpersonal/affective traits (Part 1; M = 8.22, SD = 2.90), as well as for lifestyle/antisocial traits (Part 2; M = 8.67, SD = 2.45). Scores on the SeSaS (M = 4.39, SD = 3.18) showed severe sexual sadism in a third of the sample (33.3%) based on a threshold score of ≥7 (Mokros, Osterheider, Hucker, & Nitschke, 2011).
As shown in Table 2, evidence of prior criminal behaviors was found for all but one offender. Criminal histories tended to involve various nonsexual, nonviolent crimes, particularly acquisitive offenses such as theft (82.4%). In addition, the majority of the sample had a history of violence, frequently including intimate partner violence, sexual violence both in and outside of intimate relationships, and nonsexual violence. In the year preceding the homicide offense, almost half of the sample (44.4%) faced legal difficulties (Table 3). Problems at work and financial difficulties were common (83.3%), and nine of the 18 men described feelings of anxiety, failure, and sadness in the 48 hr leading up to the offense. Seven men were found to be happy or generally in a state of well-being in the lead-up to the offense. The most prevalent precrime emotional state was sexual excitation (88.9%). Sexual gratification was the primary motivation identified in almost all cases (16 out of 18), with sadistic motivation being present in around a third and homicidal fantasies present in almost a third.
Official and Non-Official Prior Criminal Behaviors of Sexual Murderers Involved in Multiple-Perpetrator Sexual Homicide (N = 18).
M = 2.82, SD = 2.07.
M = 3.94, SD = 2.14.
Antecedent Factors for Sexual Murderers Involved in Multiple-Perpetrator Sexual Homicide (N = 18).
Precrime refers to the 48 hr prior to offending.
Victimology is presented in Table 4. All victims were female, most frequently post-pubescent. Almost half of the victims were known to one or both aggressors. None were involved in sex work, nor did they have criminal histories. Table 5 shows modus operandi factors identified in the nine criminal events carried out by the 18 perpetrators in the present sample. Most victims were killed either in their own home or in a secluded area such as in the woods or in a water area (such as by a river), although they were often taken to one of these places after the perpetrators initially approached them elsewhere (using a con strategy more than half of the time). Regardless of the initial nature of the interaction between perpetrators and victims, verbal threats and aggressive or abusive language were always used during the offense, and all of the victims were coerced to comply with demands aimed at physical control or to perform sexual acts. When victims resisted, half of the time perpetrators reacted with increased violent intensity. Penile penetration was forced upon all of the victims, vaginally, anally, and/or orally. Postmortem sexual interference occurred in around a third of cases. Victims were most frequently killed by strangulation (more than one cause of death may have been reported in coroner/autopsy reports when multiple injuries were inflicted; however, manual strangulation or use of a ligature was reported as the primary fatal injury in 44.4% and 22.2% of cases, respectively). Victim’s bodies were usually moved (unless left in their own home) and left either in a wooded or water area, or in the street (Table 6). Otherwise, the majority of offenders tended not to take elaborate precautions to avoid being detected. They most frequently reported positive postcrime affect.
Victimology in Multiple-Perpetrator Sexual Homicides (n = 9 Criminal Events).
Victims (n = 13) were aged 10 to 52 years (M = 25.54, SD = 14.75).
Modus Operandi in Multiple-Perpetrator Sexual Homicides (n = 9 Criminal Events).
Range in minutes = 20 to 1,200 (M = 266.11, SD = 401.98).
Crime Scene Factors and Postcrime Behavior in Multiple-Perpetrator Sexual Homicides (n = 9 Criminal Events).
Number of detection avoidance strategies used: M = 3.56, SD = 1.24.
Lone-Perpetrator/Multiple-Victim Sexual Murderers
Case A
A was a White male, aged 46 years at the time of the offense. He lived alone, socially isolated, and experienced difficulties with women that would result in his harassment and aggression of them following perceived rejection. He was responsible for three rape offenses prior to the homicide, each in similar circumstances: known victims, toward whom he was romantically attracted or feeling vengeful having had his sexual advances rebuffed. He raped these victims in their homes, using a knife to threaten them into compliance. After a period of imprisonment for these offenses, A was able to secure employment. Misinterpreting social cues as had been problematic for him in the past (leading to the previous rape offenses), A developed a sexual interest in a female colleague and began stalking her. Following a stressful event at work (believing that his history of sexual violence may become indiscriminately known), A went to the home of the female colleague he had been stalking (targeted victim), knowing that she lived with her mother and that neither were home at the time. The eventual victims were both the targeted victim and her mother (mother had arrived home first; there was sufficient time to overpower her and await the arrival of the targeted victim). Both victims were found naked, tied to their beds, and killed by knife wounds (multiple wounds to the targeted victim, including to her face, neck, and torso, while her mother was killed by a single wound without other injuries). The perpetrator attempted to destroy the bodies and the crime scene by setting several fires and lighting the gas in the kitchen. He was apprehended almost a year later, during the commission of another sexual assault in similar circumstances.
Case B
B was a White male, aged 45 years at the time of the offense. He suffered depression, low self-esteem, was divorced, and had recently been dismissed from an executive-level job. He had a history of contact sexual offenses against children and was consuming child sexual exploitation material in the lead-up to the homicide offense. This was a premeditated offense: B went out in his car with children’s toys, books, and treats; a dog leash to be used to deceive his victims that he was looking for a lost dog; and a knife, rifle, and rope. The victims were two female children both aged 10 years. They suffered multiple sexual assaults and sadistic acts over a period of 2 hr. They were killed by strangulation and their bodies were thrown into a ravine. B was subsequently found unconscious in a hotel room, having made a serious suicide attempt.
Case C
C was a White male, aged 33 years at the time of the offense. He was employed, lived in a shared residence, but was socially withdrawn. His criminal history indicated sexually sadistic interests, based on previous convictions for nine rape offenses against young adult victims who were threatened with a knife and strangled, as well as possession of a photograph of a perceptibly terrified, naked woman. The victim in the photograph was never identified. The circumstances of the double homicide were that C happened upon his victims in an animal cemetery, where he had gone with the intention of destroying gravestones in reaction to irritation with a neighbor’s dog. The victims were two adult women. They were engaged in a dispute, to which the perpetrator became enraged because their argument triggered hostile memories of his mother. He attacked the women with a piece of wood, rendering them unconscious, and put them into the boot of his car. The victim’s bodies were found at the edge of a field, naked and burned. They had been strangled, and also one victim was cut to the throat, while the other had additional injuries from being beaten with the piece of wood. The perpetrator claimed that he had killed the victims to avoid that they report him to the police.
Discussion
Multiple-perpetrator sexual homicide cases involve at least two individuals, each predisposed through socially disconnected, criminal, and sexually deviant lifestyles, to violence toward women. These are not “couples who kill,” but men whose social circumstances bring them together aggravating an extreme criminal event. In the present study, a number of psychosocial and criminal background factors characterized the sample. First, they were habitual offenders. Most had a history of acquisitive offending and other nonviolent, nonsexual crimes, and more than half had previously spent time in prison. A history of violence was also typical; the majority of these men had been nonsexually violent, physically as well as through verbal aggression, including death threats. Their tendency toward violence commonly involved the use of weapons, suggesting at least desensitization, or at the other extreme a proclivity toward weapon use. Subsequently, rape kits, which were often carried as part of the modus operandi of multiple-perpetrator sexual murderers, tended to comprise weapons as well as other items such as ligatures. In addition, with few exceptions these offenders had sexually violent behaviors and crimes in their background. They had coerced and assaulted intimate partners, and they were also responsible for sexual crimes outside of their intimate relationships.
Criminality among these men was one factor in a general lifestyle that may be described as socially disconnected, consistent with previous research observing social disconnection in psychometric data from sexual murderers (Higgs, Carter, Stefańska, & Glorney, 2017). In the present sample, almost all were “loners,” or socially isolated individuals, typically impulsive and unable to tolerate life’s frustrations. They did not manage responsibilities well and experienced difficulties in the workplace. Their problems were exacerbated by substance misuse and financial struggle. Commonly held self-other-world cognitive schemata pointed to perceived personal threat in day-to-day life, mistrust, and grievance toward women in particular. This pattern of social disconnection and egoistic traits/lack of empathic concern for others was reflected in generally high scores on the PCL: SV. Psychopathic traits were present in both the interpersonal/affective domain and the lifestyle/antisocial domain, which together with the overall profile emerging from the current data suggests that when sexual homicide involves multiple perpetrators it is likely that psychopathic personalities characterize each of them rather than one of them in particular. Offending may reflect sensation seeking and/or reckless disregard for the well-being of others in pursuit of satisfying personal needs.
In the context of a world experienced as overtly hostile, there was a certain inclination toward deviant sexual fantasies, indicated in the type of pornographic materials consumed. Furthermore, along with cognitions around the sexual objectification of women, rape fantasies were disclosed by almost three quarters of the individuals in the present sample. In a third of the offenses examined (a rate consistent with that observed in larger samples; Stefanska et al., 2015), victims were further degraded by sexual interference after their death. In some cases, there may also have been a sexual assault while the victim was still alive, whereas in other cases sexual acts are only carried out after the infliction of fatal injury. Explanations for such behaviors include the desire to humiliate the victim, opportunistic sexual gratification—importantly, not inhibited in the context of extreme violence—and the possibility that the act of killing forms part of a deviant sexual fantasy. Severe sadism was found at a prevalence of around one in three offenders in this sample, consistent with estimates derived from other samples of sexual murderers (Briken, Habermann, Berner, & Hill, 2005; Stefanska et al., 2019). However, in the present study, the SeSaS threshold used was the more conservative of those established by previous research (Mokros et al., 2011), whereas based on a lower threshold (≥4; Nitschke et al., 2009), criminal behaviors were indicative of sexual sadism in as much as 55.6% of this sample. Overall, the present results suggest that sexual deviancy, often sadistic, is an important factor to consider in forensic case formulation for multiple-perpetrator sexual homicide.
Case conceptualization consistent with the established typology of sexualized/grievance/rape murderer (Higgs, Carter, Tully et al., 2017) is entirely feasible based on the present results. In this sample, perpetrators tended to correspond either to the sexualized murderer subtype (killing was directly associated with sexual arousal) or to the rape murderer subtype (victims were killed following a sexual assault to prevent them from being able to report the offense). In addition, antecedents to sexual homicide as well as modus operandi factors mirrored the wider sexual homicide literature: These are socially disconnected individuals holding antisocial attitudes and beliefs, distrustful of women and/or seeing them as sexual objects, acting on deviant sexual fantasies which may include a sexually sadistic component. That the offenses were not well explained by leader/follower group dynamics further situates our understanding in this theoretical orientation, as opposed to research seeking to explain multiple-perpetrator rape where it appears that nonsexual motives and more passive or coerced participation occurs (Da Silva et al., 2015). In multiple-perpetrator sexual homicide, it appears that predisposing factors for sexual homicide are typically comparable for offenders individually. Life circumstances lead these otherwise loner-type offenders together: In all cases in the present sample, perpetrators previously or at the time of the offense lived together, usually in hostel or squat-type residence. With each having the individual propensity to perpetrate a sexual homicide, together they make a recipe for disaster. There may not be a leader coercing others in the planning or perpetration of the offense, but the presence of co-offenders may act as a disinhibiting factor both in terms of likelihood or trajectory toward sexual homicide and escalation of degree of harm caused during the offense culminating in fatality.
Although we may not attempt to generate hypotheses regarding multiple concurrent victim sexual homicide from the present description of three cases, some observations are useful concerning these cases in relation to the existing literature. Particularly, the fact of a second victim appears relatively inconsequential in understanding the offense and its perpetrator. At least in these cases, the offender characteristics and modus operandi were remarkably similar to those typical of single-victim sexual homicide. Only the circumstantially determined presence of a second victim not presenting an obstacle (conversely, presenting additional opportunity for sexual gratification) differentiated these offenses. Framed in routine activities theory (Cohen & Felson, 1979), the presence of two victims had no meaningful impact on the motivated offender or his ability to act on inclinations. The victim or target of the offense was affected only in the sense that the offender acted upon two victims rather than only one; the third person present was seen as an additional suitable victim rather than a capable guardian who may have prevented the crime from happening.
The offense summary in one instance (Case A) is consistent with Higgs, Carter, Tully et al.’s (2017) description of grievance murder, while Case C aligns to what they termed sexualized murder. However, Case B may be understood in relation to research comparing sexual murderers of women with sexual murderers of children. It has recently been demonstrated that while certain factors such as social isolation are common to all of these offenders, an important difference between those who kill adult women compared with those with child victims is that for the latter sexual deviancy is more predominantly pedophilic than sadistic, with fatal injury serving instrumental purposes (Proulx, James, Siwic, & Beauregard, 2018).
Study Limitations and Conclusions
Sexual homicide is a rare form of crime, meaning that research aiming to further reduce its prevalence is paradoxically hampered by small samples—even more so when unusual cases are considered. In the present study, some light has been shed on a small, but previously empirically invisible group of sexual murderers. While the sample size presents clear limitations for the study, some tentative answers to the research questions are possible: Multiple-perpetrator sexual homicides are explained equally well by the existing sexual homicide literature as the single-perpetrator sexual homicides it is primarily based upon. That is, forensic clinical case formulation with perpetrators of sexual homicide may draw upon established subtypes of sexual homicide (Higgs, Carter, Tully et al., 2017) whether offenders acted alone or with a co-offender. The factors driving offending may differ between the individuals involved (a sexualized murderer may offend with a grievance or rape murderer), but the presence of a co-offender seems to be a disinhibiting or enabling factor for each rather than appreciably altering the form of offending behavior. Offense dynamics are not likely to operate as seen in multiple-perpetrator rape, where there is typically a leader and more passive or coerced roles (Da Silva et al., 2015). When there are multiple concurrent victims, this does not necessarily meaningfully alter the modus operandi or other factors relevant to understanding the case. However, attention should be given to victimology given that sexual murderers of children differ from sexual murderers of women (Proulx et al., 2018).
These findings are relevant to clinical case management in forensic settings and also have implications for future research. Specifically, based on the present results, there appears to be no reason to exclude cases of multiple-perpetrator or multiple concurrent victim sexual homicide from larger research samples, subject to other inclusion criteria specifying age and gender of offenders and victims in the sample.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
