Escalation from Fetish Burglaries to Sexual Violence:
A Retrospective Case Study of Former Col., D. Russell Williams
ANDREW E. BRANKLEY*, ALASDAIR M. GOODWILL and KYLIE S. REALE Department of Psychology, Ryerson University, Toronto, Canada
Abstract Criminal history narrative studies reveal an escalation in sexual offender behaviour from non-contact to contact offending, with an ever-increasing likelihood of sexual violence and homicide. In particular, researchers have found that sexual offenders often have a history of committing burglaries prior to contact offences. Accordingly, researchers have suggested that burglaries may be associated with an increased likelihood of future sexual offending, particularly when they have a sexual element to them. However, to date, there has been little quantitative research focusing on the mechanisms of escalation in sexual of- fences. This paper seeks to study factors associated with sexual offence escalation in terms of changes in offence seriousness and frequency. Speci?cally, case evidence was gleaned from a series of fetish burglaries and subsequent sexual assaults and murders committed by the former Canadian Colonel David Russell Williams (RW). Cluster analysis, chi- square, ANOVA, and regression analyses were conducted on the crime scene information of RW’s 82 cases of fetish burglary. Analyses revealed a signi?cant escalation in the fre- quency and seriousness of RW’s fetish burglary offences prior to committing acts of sexual violence and ultimately sexual homicide. Recommendations for future research predicting escalation of sexual offending by frequency and seriousness of offending behaviour are discussed. Copyright © 2013 John Wiley & Sons, Ltd. Key words: escalation; sexual offenders; fetishism; burglary; non-contact offending
The current scienti?c framework to psychological research provides for both nomothetic and idiographic approaches, but the latter is sadly underutilised, particularly in criminal, forensic, and investigative psychology. In clinical forensic case studies, idiographic approaches are thought to be useful in identifying variations in assessment and treatment ef?cacy at the individual level (Smith, 2012). Within criminal and investigative contexts, it is argued that, although police investigations tend to work in a reverse manner, where the
*Correspondence to: Andrew E. Brankley, Department of Psychology, Ryerson University, 350 Victoria Street, Toronto M5B 2K3, Canada. E-mail: andrew.brankley@psych.ryerson.ca The authors would like to dedicate this paper to the lives and memories of the victims discussed herein and all individuals affected by these offences. We hope that future crimes may be prevented through further understanding of sexual offending behaviour.
Journal of Investigative Psychology and Offender Pro?ling J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) Published online 5 November 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jip.1406
Copyright © 2013 John Wiley & Sons, Ltd.
intent is to identify the offender from behavioural and crime scene details (Rainbow & Greg- ory, 2011), the reverse, idiographic approach can still be fruitful. One clear example is pro- vided by Dan Fishman (1999), who suggests the systematic study of individual cases can be used to build a pragmatic nomothetic evidence base. Indeed, case studies of offenders have been used in the past to test and optimise current investigative methodology (Aggrawal, 2011;Cooper,Swaminath,Baxter,&Poulin,1990;McCann,2000;Schlesinger,2002;White, 2007;Wilcox,Foss,&Donathy, 2005).Therefore,itbecomesincreasingly importantto eval- uateaggregatednomothetic-basedunderstandingofsexualoffendingbehaviourfromtheindi- vidual idiographic level for congruent and conceptual validity. This transaction between the experimental,deductiveapproachandtheclinical,inductiveapproachisthecoreofpragmatic psychology and provides the substantive framework for this manuscript. In the current study, the details from the case of former Canadian Air Force Colonel David Russell Williams (RW) will be analysed in hopes of gleaning a deeper understand- ing of offending behaviour for both behavioural research and criminal investigations. RW was convicted in 2010 of 88 counts, including two counts of ?rst-degree murder, two counts of sexual assault, two of forcible con?nement, and 82 break and enter1 charges, dat- ing back to 2007. The majority of RW’s offending behaviour focused on the acquisition of fetish items (usually undergarments) and the extensive photographing of these items; typically neatly laid-out on a ?at surface (e.g. ?oor, beds, and tables) and/or on his person in various states of undress. The investigation and conviction of RW provide a useful investigative case study, as a large amount of forensic evidence (e.g. photographs and un- dergarments) was collected from both the crime scenes and from RW’s homes, providing much insight into his offending behaviour. As such, the current paper will provide a quan- titative analysis of escalation in frequency and seriousness in his offending behaviour over time and contextual life events, based on available data (R v. Williams, 2010).
CASE HISTORY OF RUSSELL WILLIAMS
Russel Williams was named the Canadian Press Newsmaker of 2010 in a poll of radio sta- tion news directors and newspaper editors for his string of fetish burglaries, sexual assaults, and murders (Morrison & Liberrto, 2011). RW’s notoriety was clearly due not only to the nature of his crimes but also to the fact that, when arrested, he was an acting colonel in the Canadian Air Force, as well as being the commander of the country’s largest military air?eld, Canadian Forces Base (CFB) Trenton, in eastern Ontario, Canada.
Background Although born in England, RW immigrated to Ontario and was educated in one of Canada’s ?nest independent high schools before proceeding to complete a degree in poli- tics and economics from the University of Toronto (Appleby, 2011). At the age of 23, RW travelled to British Columbia for basic training in the Canadian Forces with aspirations of becoming a pilot. He married at the age of 28 and continued his career in the Canadian Forces, moving across Canada frequently before settling in Orleans, a suburb of Ottawa,
1Breaking and entering—prohibited in Canada by s.348(1) of the Criminal Code (1985)—involves trespassing along with intent to commit an indictable offence or actually committing an indictable offence once inside. To make the paper as accessible as possible to an international audience, we will use the terminology ‘fetish burglary’ to refer to break and enters with a supposed sexual paraphilic intent.
116 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Ontario in 1995 (Appleby, 2011). RW and his wife also kept a modest cottage home in the town of Tweed, roughly 160km west from Orleans, which they visited regularly. RW excelled in his duty and services to the Canadian Forces, culminating in his appointment of commander of CFB Trenton on 15 July 2009.
Fetish burglaries Russel Williams plead guilty to 82 counts of breaking and entering that occurred over a period of 2years. At the time of his ?rst offence (9 September 2009), RW was 44years old, held the rank of lieutenant-colonel, and worked at the Directorate of Air Requirements at the National Defence Headquarters in Ottawa. All of the fetish burglaries took place in close proximity (less than 2km) to either his home in Orleans or his cottage in Tweed. Although the agreed upon statement of facts from RW’s trial is partially incomplete in regard to victim age, it is a matter of fact that RW stole items (e.g. undergarments) indis- criminately from both children and adults (R v. Williams, 2010). RW would ‘skilfully’ (many victims did not even know they had been victimised) break into homes (often unoc- cupied) and procure panties, bras, girdles, thongs, swimsuits, and occasionally bathrobes and shoes. Interestingly, although items of objective worth (e.g. money, jewellery, and electronics) were readily available to steal, RW never stole any such item on any occasion (Appleby, 2011; R v. Williams, 2010). Further, he meticulously catalogued the number and type of each undergarment stolen, invariably taking photographs of the clothing carefully displayed by type and/or colour. After, and later during, the offences, RW would also take pictures of himself dressed in the female undergarments (e.g. see pictures below, R v. Williams, 2010)—many of which showed him masturbating. In later offences within this series, he began taking pictures of objects and items from within the homes he entered, such as pieces of identi?cation, a screenshot of a victim’s Facebook page, and screenshots of police websites reporting the crimes. As he continued to commit fetish burglaries, he became even more brazen, masturbating in the victim’s home, entering occupied homes, stealing sex toys, and leaving notes for his victims, written in an increasingly aggressive manner. All fetish burglaries were committed during the hours of darkness, typically around midnight.
Sample of pictures taken by RW including self-portraits and stolen undergarments (R v. Williams, 2010).
Escalation from fetish burglaries to sexual violence 117
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Sexual assaults Russel Williams committed his ?rst contact offence on 17 September 2009, against a woman in her twenties who had recently given birth. At the time of the offence, the vic- tim’s boyfriend was out of town and she was home alone with her baby. RW broke into the home through a window by cutting a screen (Appleby, 2011, p. 130). The victim awoke to RW pressing down upon her, at which time a brief struggle ensued. RW continued to hold her down and, for a period of roughly 30minutes, a conversation took place in which RW told her that he did not plan to kill her but rather take pictures of her. With his hands, RW then struck the woman three times in the head and threatened her into submission (Appleby, 2011, p. 130). RW then blindfolded her with a pillowcase, tied her up, posed her, and took pictures of her for approximately 2hours. Before leaving, RW went to the bedroom and took six pieces of her underwear (R v. Williams, 2010). Thirteen days later, RW committed his second sexual assault. Again, RW targeted a sleeping woman home alone, just as he had done in his ?rst sexual assault. However, instead of waking the victim with a verbal warning as he did previously, RW jumped on top of the sleeping woman, placing a blanket over her head, and beating her savagely, warning her not to look at him (Appleby, 2011, p. 12). As with the ?rst victim, he then blindfolded her and proceeded to take photos of her for approximately 3hours, forcing her to display herself in various sexual poses. RW eventually left, taking no articles of clothing (R v. Williams, 2010). Although both crimes are sexual assaults, no penetration occurred (Appleby, 2011; R v. Williams, 2010). Instead, RW forced his victims to pose in various stages of undress whilst taking pictures of them, probably in an attempt to procure more (deviant) fetish- istic material. RW did fondle each woman’s breasts but stopped short of further assault, notably when the second victim protested his advances (Appleby, 2011). In both cases, RW entered the victim’s home and used a surprise (or blitz) attack to initially gain control of the victim.
Murders Over the winter of 2009–2010, RW escalated from sexual assault to commit sexually motivated murders of two young women. In these offences, there is clear behavioural escalation in fetishistic and violent behaviour from the previous two sets of offences, as RW now engaged in sexual penetration, torture, and ultimately murder. Moreover, RW showed signi?cant escalation in his pursuit to procure an even more deviant and arousing fetishistic and paraphilic material by extensively videotaping both offences. The ?rst victim was a 37-year-old ?ight attendant at CFB Trenton; it is possible RW may have met her there (Appleby, 2011), but, perhaps more notably, he would have had access to her work schedule. RW horri?cally spent over 5hours violently sexually assaulting this victim. She was found dead in her home, just west of CFB Trenton, on 25 November 2009. RW had tied her up so tightly that the ropes left burn marks on her forearms and wrists (Appleby, 2011, p. 151). On 29 January 2010, RW broke into his second sexual murder victim’s home and lay in wait for her to return from work. RW had previously observed this victim while driving on his regular route home from CFB Trenton to his cottage in Tweed (Appleby, 2011; Gibb, 2011). RW spent a considerably longer time with the second victim (close to a day) and conducted several separate photo sessions lasting over 3hours each. In a video of the
118 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
offence, RW can be seen taunting the victim with hope of freedom (Appleby, 2011, pp. 180–183). He showed a diverse range of behaviour from choking her with a zip tie while she performed oral sex to feeding her to maintain her strength. In both cases, the women were ultimately killed through asphyxiation.
Investigation and conviction The police eventually identi?ed RW by matching tyre tracks left at the second murder victim’s house and those of his Nissan Path?nder through the Royal Canadian Mounted Police’s Canadian Police Service Identi?cation Centre (Chouinard, 2011). After an intensive police interview, RW confessed, leading the police to the body of the second murder victim. Although RW plead guilty to all charges discussed previously, he refused to address a hidden cache of child pornography found on his computer—as part of his plea bargain and full confession, no charges were brought forth on this matter (Appleby, 2011).
ESCALATION IN SEXUAL OFFENDING: FREQUENCY AND SERIOUSNESS
Despite its importance within the criminal career and offending trajectory (Ward et al., 2010; Warren et al., 1999), there is a paucity of research regarding the mechanism(s) of offence escalation in comparison with other aspects of sexual offending (Liu, Francis, & Soothill, 2011). Clarke and Cornish’s (1985) continuance model for general offenders pre- dicts a temporary escalation in offence frequency, through the mechanism of positive rein- forcement, until the offender plateaus at the optimal rate of offending (Lussier, Leclerc, Healey, & Proulx, 2012). Liu and colleagues (2011) acknowledge the importance of esca- lation in terms of frequency of offending but also emphasise that, for violent offenders, one must consider changes in offence seriousness. However, only a handful of researchers have studied both seriousness and frequency in escalation (Armstrong & Britt, 2004; Cochran, Sellers, Wiesbrock, & Palacios, 2011; de Boer, Caramaschi, Natarajan, & Koolhaas, 2009), with most de?ning escalation either in terms of frequency (Cale & Lussier, 2011; Nieuwbeerta,Blokland,Piquero, &Sweeten,2011; Sherman et al.,1991; Wallace, Mullen, & Burgess, 2004) or seriousness (Blumstein, Cohen, Roth, & Visher, 1986; Carrington, 2013; Hoeve, McReynolds, Wasserman, & McMillan, 2013; Soothill, Francis, & Liu, 2008). This lack of de?nitional consistency in the literature may help explain the lack of consensus surrounding the mechanisms of escalation. In order to provide a theory to bridge these often disparate interpretations of escalation, evidence for the importance of paraphilia and life events in considering patterns of escalation in sexual offences will be discussed.
The relationship between paraphilia and escalation Presentresearchhassuggestedthatthepresenceofaparaphiliahasimportantimplicationsfor escalation(Harris,2011).Inthe?ftheditionoftheDiagnosticandStatisticalManualofMen- tal Disorders,paraphiliasarede?nedasdisordersthatinvolveintense,sexuallyarousingfan- tasies, sexual urges, or behaviours along with at least one of the following: (1) non-human objects; (2) the suffering or humiliation of oneself or one’s partner; or (3) children or other non-consenting persons that occur over a period of at least 6months (American Psychiatric
Escalation from fetish burglaries to sexual violence 119
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Association [APA], 2013, pp. 522–523). Paraphilias are more common in sexual offenders who commit homicide than sexual offenders who do not (Koch, Berner, Hill, & Briken, 2011). A potential explanation for this ?nding is that despite efforts to conceal the fantasies, overtime,offenders’ paraphiliastendtobecomeincreasinglyeroticandpotentiallyviolentin nature (Arrigo & Purcell, 2001). Paraphilic behaviour is enacted through a cyclical, reinforcing process, which consists of a relationship between the paraphilic stimuli, fantasy, and orgasm, often enhanced through facilitators such as drugs or pornography (Arrigo & Purcell, 2001). The repeated exposure to a particular stimulus will decrease the size of the response through the process of habituation; causing an ever-increasing desire for greater and increasingly novel stimulus. If the individual’s fantasies and imagery increase in violence and sexuality, then the actual paraphilia tends to escalate in intensity and frequency (Arrigo & Purcell, 2001). Progressively violent behaviours that develop from the paraphilia can manifest in criminal activity, including burglary, assault, rape, and murder, even developing into sadistic sexual activity involving torture, mutilation, and necrophilia (Burgess, Hartman, Ressler, Douglas, & McCormack, 1986)—all of which are behaviours RW engaged in. Two paraphilias particularly relevant to the present case are fetishistic and transvestic dis- orders. Fetishistic disorder (known commonly as fetishism) is characterised by the repetitive use of, or dependence on, non-living objects as the primary elements associated with sexual arousal (APA, 2013, pp. 700–701), the most common of which are female undergarments (Schlesinger&Revitch,1999;Wiederman,2003).Transvesticdisorderreferstotherecurrent and intense sexual arousal from cross-dressing, as manifested by fantasies, urges, or behav- iours(APA,2013,pp.702–703)andfoundexclusivelyinmen(Wiederman,2003).Menwith transvestic fetishism typically portray traditional male gender roles and sexual identity, but cross-dress to varying degrees for sexual arousal (Wiederman, 2003). Regardless, there is typically recognition by the individual that their sexual behaviours are not within the norm and they may make immense efforts to conceal them (Wiederman, 2003). In an early study by Revitch (1978), transvestism was found to be associated with fetish burglaries. Later, in a sample of 52 serial sexual homicides, Schlesinger and Revitch (1999) found that over a third of the sexual murderers had prior convictions for burglary. In those cases, the theft of fetishistic objects was found to be directly related to escalation to obtain (more of) the object and then, although less frequently, to assault or murder the female that occupied the home. Interestingly, the study also found that peeping behaviour commonly preceded burglaries and, in a small amount of cases, also led to assault and murder (Schlesinger & Revitch, 1999)—a tactic many speculated that RW used in order to plan his offences (Appleby, 2011; Gibb, 2011). Further, Langstrom and Zucker (2005), in a large study of 5250 men from 18–74years of age, found that individuals who engaged in transvestic fetishism were signi?cantly more probable to spy on what others are doing sexually and be sexually aroused by the use of pain—again, behaviours in which RW knowingly engaged in (Appleby, 2011; Gibb, 2011; R v. Williams, 2010).
The impact of life events on escalation Although Clarke and Cornish’s (1985) continuance model acknowledges the impor- tance of situational variables associated with an offender’s likelihood of committing a crime, little effort has been made to formally account for life events as a mechanism related to escalation. One notable exception is Pino (2005) who suggested that a
120 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
comprehensive approach to the study of escalation must focus on the various contex- tual and situational life events surrounding the behavioural escalation, including detailed analysis of the offender, the victim, and the police response. Although not directly related to escalation, Hanson and Harris (2000) advocated a similar examina- tion of the life events of offenders, ?nding that acute stressful events typically preceded sexual offending. Interestingly, research conducted on life events and paraphilia indicate that periods of stress are often associated with engagement in transvestic behaviour (Wiederman, 2003). A cognitive-behavioural explanation for positing paraphilia and life events as mecha- nisms in the escalation patterns of sexual offending can be made by modelling sexual deviancy as addictive behaviour (Carnes, 1985; Herman, 1988; Irvine, 1995). Some social learning theorists have proposed a ‘buffering hypothesis’ based on a dynamical systems model of addiction. Proponents of this theory (Gorman et al., 2004) hold that the rate at which an individual engages in the maladaptive target behaviour is moderated by social support that reduces their perceived stress and increased abstinence. For example, Ames and Roitzsch (2000) found that in a sample of individuals diagnosed with substance depen- dence, perceived stress predicted cravings but the level of social support moderated this relationship. As the operant conditioning model of paraphilia (discussed previously by Arrigo & Purcell, 2001) is consistent with addiction models (Brown, 1986; Olsen & Winder, 2010; Van Ree, Gerrits, & Vanderschuren, 1999), it stands to reason that life events, potentially those related to stress and social support, may be related to escalation in sexual offending.
THE CURRENT STUDY
The current study presents a comprehensive case study analysis of the life events and criminal activity of RW with the aim of eliciting better understanding of esca- lation within an offender’s criminal trajectory. Speci?cally, four hypotheses will be investigated. It is expected that increases in escalation will be detected in terms of (1) frequency at which offences are committed and that (2) life events will in?u- ence this relationship. Based upon the research discussed previously, it stands to rea- son that (3) a relationship between paraphilic behaviour and offence escalation exists (Arrigo & Purcell, 2001; Burgess et al., 1986; McCann, 2000; Schlesinger & Revitch, 1999). Lastly, (4) life events will be related to increases and decreases in offence seriousness.
METHODS
Russel Williams’s case history provides a unique opportunity to study the link between paraphilia and escalation in offence frequency and seriousness, as RW not only collected an immensely large set of fetish items, he was also extremely fastidious in documenting the items taken. RW kept detailed records of his offences through digital photography (of items and premises) and through his personal journal. Each photograph was time stamped, allowing us to know when he started taking pictures of the house (usually documenting his arrival to the house) and when he approximately left and/or the cessation of the offence. This allowed conservative estimates of the time spent in the house to be
Escalation from fetish burglaries to sexual violence 121
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
made (Appleby, 2011; R v. Williams, 2010). RW was so conscientious in his record keeping that on one occasion when he realised that he had failed to reset his digital camera for day- light savings, he re-opened saved an additional corrected set of photographs, documenting the error in his journal (Gibb, 2011, p. 373). Most research on crime scene behaviour relies largely on police records (Santtila, Hakkanen, Canter, & Elfgren, 2003) or self-report offender interviews (Beauregard & Leclerc, 2007). The former are limited by the variables collected by the investigators and the latter may be unreliable due to deliberate biassed recall by the offender or simply through memory decay. By utilising the Agreed Statement of Facts presented in the trial (R v. Williams, 2010), which included information based upon the digital photos taken by RW, a more objective and reliable account of the offences is possible. Further, due to the fact that RW was a public ?gure with a prominent public record, including appearances, promotions, and other professional and social engagements, it is also possible to identify some of the possible signi?cant life events and stressors that occurred prior to and during his offending.
Data collection The Access to Information Act (1985) provides Canadian citizens or permanent residents the right to request and be given access to any record under the control of a government institution (s.4[1]). The agreed statement of facts is a list of the objective facts (e.g. offence details) relevant to sentencing a convicted offender. In the case of RW, the agreed statement of facts (R v. Williams, 2010) was based upon a combination of RW’s own confession of the crimes and evidence garnered by police at crime scenes and RW’s residences. The agreed statement of facts was read during his ‘guilty plea’ trial in 2010 and provides a detailed documentation of the 88 offences committed by RW in chronolog- ical order (R v. Williams, 2010). The agreed statement of facts was analysed to create a database detailing various offence parameters and behaviours, coded as categorical or continuous variables. If the presence or absence of a variable could not be determined by the agreed statement of facts, then the behaviour was considered as missing. Time of entry was based upon the timestamps of the digital photographs. Detailed information regarding RW’s method of entry into homes was categorised into skilled (e.g. picked the lock), force (e.g. break a window), easy (e.g. door was unlocked), and unsuccessful (e.g. intent was present but did not enter). Details of the numbers and types of clothing items taken in each offence were provided in the agreed statement of facts. This allowed for the total amount stolen during each offence to be recorded and then broken down into separate variables depending on the articles of clothing (for example, undergarments, bathing suits, and house robes). In the majority of offences the number of pictures taken by RW of fetish objects (e.g. undergarments, RW dressed in undergarments, RW masturbating, and sex toys found in the home) were also known. There was also information regarding each offence on whether or not the victims were aware that someone had broken into their home. A crime scene variable glossary was constructed for the current study (Appendix A). An inter-rater reliability analysis performed on a subset of the cases (n=10) to determine con- sistency between two raters by using the Kappa statistic for nominal variables and inter- class correlation for continuous variables. While Kappa statistics ranged from 0.75 to 1.00 (M=0.98, SD=0.07), all inter-class correlation statistics were 1.00 demonstrating high reliability.
122 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
RESULTS
Descriptive statistics for fetish burglaries Table 1reports the descriptivestatisticsfor the variables codedfrom RW’s fetish burglaries used in the present study. In total, during the fetish burglaries, RW stole 1366 articles of clothing (m=17.5,SD=28.5), the majority of which (59.2%) were bras and panties. The largest theft of clothing occurred on 19June 2009, when RWspent over 2hours inthe house of a 24-year-old woman, taking 186 pieces of clothing. In general, RW burgled homes when victims were not present (87.4%) and spent on average 22.5minutes (SD=47.2) in each house. Offences were evenly split between both ?rst time victims (55.4%, n=46) and repeat victims (44.6%, n=37) and between locations, Tweed (57.8%, n=48) and Orleans (42.2%, n=35). Interestingly, victims noti?ed the police in only 26.5% of the cases (n=22);
Table 1. Descriptive statistics of offence variables for fetish burglaries (n=83) Variable Count or total* % Mean SD Maximum
Series Tweed 48 57.8% Series Orleans 35 42.2% Cool-down 9.2 16.4 135 Previously victimised 37 44.6% Method of Entry: Easy 57 69.5% Used force 11 13.4% Skilled 6 7.3% Unsuccessful 8 9.8% Time in house (minutes) 1645 22.5 47.2 171 Interrupted during crime 11 13.3% Victim present during offence 8 9.6% Children lived at victimised home 31 37.3% Masturbation: None 58 70.7% After crime 7 8.5% During crime 15 18.3% Both 2 2.4% Naked during offence 20 24.4% Naked outside residence 2 2.4% Naked inside residence 19 23.2% Total clothing stolen 1366 17.5 28.5 186 Bras & panties 808 11.7 19.3 115 Slips & camisoles 24 0.3 1.2 10 Swimwear 10 0.1 0.3 1 Pajamas & housecoats 9 0.1 0.4 2 Dresses & tops 19 0.2 1.0 8 Total pictures taken 2937 37.2 48.2 284 Pictures of RW dressed in clothing 322 5.0 12.6 78 Pictures of clothing laid out 896 13.6 23.6 126 Pictures of RW masturbating 208 3.4 10.1 62 Pictures of RW erect 97 1.8 4.6 20 Pictures of victim’s personal items 31 0.5 1.3 6 Total of other objects stolen (e.g. personal items) 38 0.7 1.5 7 Victim noti?ed police 22 26.5%
*Counts and totals re?ect only known evidence/data
Escalation from fetish burglaries to sexual violence 123
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
potentially, suggesting that many did not know they had even been burglarised. This could attesttoRW’sskill,thevulnerabilityofthehomes,ornon-disclosureduetothepotentialem- barrassment for the victim in reporting the crime.
Escalation in frequency of fetish burglaries Regression Regression analyses were conducted to model the frequency of RW’s burglary offences, measured by the number of days between his offences in the burglary series, referred to here as ‘cool-down’. Figure 1 illustrates the increases, plateaus, and decreases in RW’s of- fences across both geographical series of offences combined. The number of days that RW cooled-down by the number of days elapsed from the beginning of the series was analysed by linear and cubic regression estimate functions, as indicated in Figure 1. Both the linear (ß=0.018, SE=0.008, p=0.028) and cubic (ß=-0.217, SE=0.074,p =0.004) regression models signi?cantly predicted cool-down days from days elapsed across the series. The linear model accounted for 5.7%, while the cubic model accounted for 15.1% of the variance; suggesting the cubic model was a better ?t for the data.
Offending phases Figure 1 indicates the temporal time points of major transitions in RW’s life based on a sampling of qualitative information gleaned from data sources (Appleby, 2011; Gibb, 2011; Morrison & Liberrto, 2011; R v. Williams, 2010). Because of the superior ?t of the cubic regression model, an attempt was made to search RW’s life history to associate signi?cant changes and/or events in RW’s working life with periods of increases and decreases in offending. Unfortunately, because of the archival nature of the current study, the authors limited the characterising of life phases to veri?able events (i.e. usually com- prising work or offence events), as it was not possible to con?dently make assertions regarding personal life events. Referring to both the subjective content of each transition
Exploratory Intensification Maintenance Deterioration
Figure 1. RW “cool-down” period between offences by days elapsed in offence series. Notable time points (number of days elapsed – by event): 053 – close call; 188 – resumes offending; 425 – trip to Kandahar; 430 –

Escalation from Fetish Burglaries to Sexual Violence:
A Retrospective Case Study of Former Col., D. Russell Williams
ANDREW E. BRANKLEY*, ALASDAIR M. GOODWILL and KYLIE S. REALE Department of Psychology, Ryerson University, Toronto, Canada
Abstract Criminal history narrative studies reveal an escalation in sexual offender behaviour from non-contact to contact offending, with an ever-increasing likelihood of sexual violence and homicide. In particular, researchers have found that sexual offenders often have a history of committing burglaries prior to contact offences. Accordingly, researchers have suggested that burglaries may be associated with an increased likelihood of future sexual offending, particularly when they have a sexual element to them. However, to date, there has been little quantitative research focusing on the mechanisms of escalation in sexual of- fences. This paper seeks to study factors associated with sexual offence escalation in terms of changes in offence seriousness and frequency. Speci?cally, case evidence was gleaned from a series of fetish burglaries and subsequent sexual assaults and murders committed by the former Canadian Colonel David Russell Williams (RW). Cluster analysis, chi- square, ANOVA, and regression analyses were conducted on the crime scene information of RW’s 82 cases of fetish burglary. Analyses revealed a signi?cant escalation in the fre- quency and seriousness of RW’s fetish burglary offences prior to committing acts of sexual violence and ultimately sexual homicide. Recommendations for future research predicting escalation of sexual offending by frequency and seriousness of offending behaviour are discussed. Copyright © 2013 John Wiley & Sons, Ltd. Key words: escalation; sexual offenders; fetishism; burglary; non-contact offending
The current scienti?c framework to psychological research provides for both nomothetic and idiographic approaches, but the latter is sadly underutilised, particularly in criminal, forensic, and investigative psychology. In clinical forensic case studies, idiographic approaches are thought to be useful in identifying variations in assessment and treatment ef?cacy at the individual level (Smith, 2012). Within criminal and investigative contexts, it is argued that, although police investigations tend to work in a reverse manner, where the
*Correspondence to: Andrew E. Brankley, Department of Psychology, Ryerson University, 350 Victoria Street, Toronto M5B 2K3, Canada. E-mail: andrew.brankley@psych.ryerson.ca The authors would like to dedicate this paper to the lives and memories of the victims discussed herein and all individuals affected by these offences. We hope that future crimes may be prevented through further understanding of sexual offending behaviour.
Journal of Investigative Psychology and Offender Pro?ling J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) Published online 5 November 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jip.1406
Copyright © 2013 John Wiley & Sons, Ltd.
intent is to identify the offender from behavioural and crime scene details (Rainbow & Greg- ory, 2011), the reverse, idiographic approach can still be fruitful. One clear example is pro- vided by Dan Fishman (1999), who suggests the systematic study of individual cases can be used to build a pragmatic nomothetic evidence base. Indeed, case studies of offenders have been used in the past to test and optimise current investigative methodology (Aggrawal, 2011;Cooper,Swaminath,Baxter,&Poulin,1990;McCann,2000;Schlesinger,2002;White, 2007;Wilcox,Foss,&Donathy, 2005).Therefore,itbecomesincreasingly importantto eval- uateaggregatednomothetic-basedunderstandingofsexualoffendingbehaviourfromtheindi- vidual idiographic level for congruent and conceptual validity. This transaction between the experimental,deductiveapproachandtheclinical,inductiveapproachisthecoreofpragmatic psychology and provides the substantive framework for this manuscript. In the current study, the details from the case of former Canadian Air Force Colonel David Russell Williams (RW) will be analysed in hopes of gleaning a deeper understand- ing of offending behaviour for both behavioural research and criminal investigations. RW was convicted in 2010 of 88 counts, including two counts of ?rst-degree murder, two counts of sexual assault, two of forcible con?nement, and 82 break and enter1 charges, dat- ing back to 2007. The majority of RW’s offending behaviour focused on the acquisition of fetish items (usually undergarments) and the extensive photographing of these items; typically neatly laid-out on a ?at surface (e.g. ?oor, beds, and tables) and/or on his person in various states of undress. The investigation and conviction of RW provide a useful investigative case study, as a large amount of forensic evidence (e.g. photographs and un- dergarments) was collected from both the crime scenes and from RW’s homes, providing much insight into his offending behaviour. As such, the current paper will provide a quan- titative analysis of escalation in frequency and seriousness in his offending behaviour over time and contextual life events, based on available data (R v. Williams, 2010).
CASE HISTORY OF RUSSELL WILLIAMS
Russel Williams was named the Canadian Press Newsmaker of 2010 in a poll of radio sta- tion news directors and newspaper editors for his string of fetish burglaries, sexual assaults, and murders (Morrison & Liberrto, 2011). RW’s notoriety was clearly due not only to the nature of his crimes but also to the fact that, when arrested, he was an acting colonel in the Canadian Air Force, as well as being the commander of the country’s largest military air?eld, Canadian Forces Base (CFB) Trenton, in eastern Ontario, Canada.
Background Although born in England, RW immigrated to Ontario and was educated in one of Canada’s ?nest independent high schools before proceeding to complete a degree in poli- tics and economics from the University of Toronto (Appleby, 2011). At the age of 23, RW travelled to British Columbia for basic training in the Canadian Forces with aspirations of becoming a pilot. He married at the age of 28 and continued his career in the Canadian Forces, moving across Canada frequently before settling in Orleans, a suburb of Ottawa,
1Breaking and entering—prohibited in Canada by s.348(1) of the Criminal Code (1985)—involves trespassing along with intent to commit an indictable offence or actually committing an indictable offence once inside. To make the paper as accessible as possible to an international audience, we will use the terminology ‘fetish burglary’ to refer to break and enters with a supposed sexual paraphilic intent.
116 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Ontario in 1995 (Appleby, 2011). RW and his wife also kept a modest cottage home in the town of Tweed, roughly 160km west from Orleans, which they visited regularly. RW excelled in his duty and services to the Canadian Forces, culminating in his appointment of commander of CFB Trenton on 15 July 2009.
Fetish burglaries Russel Williams plead guilty to 82 counts of breaking and entering that occurred over a period of 2years. At the time of his ?rst offence (9 September 2009), RW was 44years old, held the rank of lieutenant-colonel, and worked at the Directorate of Air Requirements at the National Defence Headquarters in Ottawa. All of the fetish burglaries took place in close proximity (less than 2km) to either his home in Orleans or his cottage in Tweed. Although the agreed upon statement of facts from RW’s trial is partially incomplete in regard to victim age, it is a matter of fact that RW stole items (e.g. undergarments) indis- criminately from both children and adults (R v. Williams, 2010). RW would ‘skilfully’ (many victims did not even know they had been victimised) break into homes (often unoc- cupied) and procure panties, bras, girdles, thongs, swimsuits, and occasionally bathrobes and shoes. Interestingly, although items of objective worth (e.g. money, jewellery, and electronics) were readily available to steal, RW never stole any such item on any occasion (Appleby, 2011; R v. Williams, 2010). Further, he meticulously catalogued the number and type of each undergarment stolen, invariably taking photographs of the clothing carefully displayed by type and/or colour. After, and later during, the offences, RW would also take pictures of himself dressed in the female undergarments (e.g. see pictures below, R v. Williams, 2010)—many of which showed him masturbating. In later offences within this series, he began taking pictures of objects and items from within the homes he entered, such as pieces of identi?cation, a screenshot of a victim’s Facebook page, and screenshots of police websites reporting the crimes. As he continued to commit fetish burglaries, he became even more brazen, masturbating in the victim’s home, entering occupied homes, stealing sex toys, and leaving notes for his victims, written in an increasingly aggressive manner. All fetish burglaries were committed during the hours of darkness, typically around midnight.
Sample of pictures taken by RW including self-portraits and stolen undergarments (R v. Williams, 2010).
Escalation from fetish burglaries to sexual violence 117
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Sexual assaults Russel Williams committed his ?rst contact offence on 17 September 2009, against a woman in her twenties who had recently given birth. At the time of the offence, the vic- tim’s boyfriend was out of town and she was home alone with her baby. RW broke into the home through a window by cutting a screen (Appleby, 2011, p. 130). The victim awoke to RW pressing down upon her, at which time a brief struggle ensued. RW continued to hold her down and, for a period of roughly 30minutes, a conversation took place in which RW told her that he did not plan to kill her but rather take pictures of her. With his hands, RW then struck the woman three times in the head and threatened her into submission (Appleby, 2011, p. 130). RW then blindfolded her with a pillowcase, tied her up, posed her, and took pictures of her for approximately 2hours. Before leaving, RW went to the bedroom and took six pieces of her underwear (R v. Williams, 2010). Thirteen days later, RW committed his second sexual assault. Again, RW targeted a sleeping woman home alone, just as he had done in his ?rst sexual assault. However, instead of waking the victim with a verbal warning as he did previously, RW jumped on top of the sleeping woman, placing a blanket over her head, and beating her savagely, warning her not to look at him (Appleby, 2011, p. 12). As with the ?rst victim, he then blindfolded her and proceeded to take photos of her for approximately 3hours, forcing her to display herself in various sexual poses. RW eventually left, taking no articles of clothing (R v. Williams, 2010). Although both crimes are sexual assaults, no penetration occurred (Appleby, 2011; R v. Williams, 2010). Instead, RW forced his victims to pose in various stages of undress whilst taking pictures of them, probably in an attempt to procure more (deviant) fetish- istic material. RW did fondle each woman’s breasts but stopped short of further assault, notably when the second victim protested his advances (Appleby, 2011). In both cases, RW entered the victim’s home and used a surprise (or blitz) attack to initially gain control of the victim.
Murders Over the winter of 2009–2010, RW escalated from sexual assault to commit sexually motivated murders of two young women. In these offences, there is clear behavioural escalation in fetishistic and violent behaviour from the previous two sets of offences, as RW now engaged in sexual penetration, torture, and ultimately murder. Moreover, RW showed signi?cant escalation in his pursuit to procure an even more deviant and arousing fetishistic and paraphilic material by extensively videotaping both offences. The ?rst victim was a 37-year-old ?ight attendant at CFB Trenton; it is possible RW may have met her there (Appleby, 2011), but, perhaps more notably, he would have had access to her work schedule. RW horri?cally spent over 5hours violently sexually assaulting this victim. She was found dead in her home, just west of CFB Trenton, on 25 November 2009. RW had tied her up so tightly that the ropes left burn marks on her forearms and wrists (Appleby, 2011, p. 151). On 29 January 2010, RW broke into his second sexual murder victim’s home and lay in wait for her to return from work. RW had previously observed this victim while driving on his regular route home from CFB Trenton to his cottage in Tweed (Appleby, 2011; Gibb, 2011). RW spent a considerably longer time with the second victim (close to a day) and conducted several separate photo sessions lasting over 3hours each. In a video of the
118 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
offence, RW can be seen taunting the victim with hope of freedom (Appleby, 2011, pp. 180–183). He showed a diverse range of behaviour from choking her with a zip tie while she performed oral sex to feeding her to maintain her strength. In both cases, the women were ultimately killed through asphyxiation.
Investigation and conviction The police eventually identi?ed RW by matching tyre tracks left at the second murder victim’s house and those of his Nissan Path?nder through the Royal Canadian Mounted Police’s Canadian Police Service Identi?cation Centre (Chouinard, 2011). After an intensive police interview, RW confessed, leading the police to the body of the second murder victim. Although RW plead guilty to all charges discussed previously, he refused to address a hidden cache of child pornography found on his computer—as part of his plea bargain and full confession, no charges were brought forth on this matter (Appleby, 2011).
ESCALATION IN SEXUAL OFFENDING: FREQUENCY AND SERIOUSNESS
Despite its importance within the criminal career and offending trajectory (Ward et al., 2010; Warren et al., 1999), there is a paucity of research regarding the mechanism(s) of offence escalation in comparison with other aspects of sexual offending (Liu, Francis, & Soothill, 2011). Clarke and Cornish’s (1985) continuance model for general offenders pre- dicts a temporary escalation in offence frequency, through the mechanism of positive rein- forcement, until the offender plateaus at the optimal rate of offending (Lussier, Leclerc, Healey, & Proulx, 2012). Liu and colleagues (2011) acknowledge the importance of esca- lation in terms of frequency of offending but also emphasise that, for violent offenders, one must consider changes in offence seriousness. However, only a handful of researchers have studied both seriousness and frequency in escalation (Armstrong & Britt, 2004; Cochran, Sellers, Wiesbrock, & Palacios, 2011; de Boer, Caramaschi, Natarajan, & Koolhaas, 2009), with most de?ning escalation either in terms of frequency (Cale & Lussier, 2011; Nieuwbeerta,Blokland,Piquero, &Sweeten,2011; Sherman et al.,1991; Wallace, Mullen, & Burgess, 2004) or seriousness (Blumstein, Cohen, Roth, & Visher, 1986; Carrington, 2013; Hoeve, McReynolds, Wasserman, & McMillan, 2013; Soothill, Francis, & Liu, 2008). This lack of de?nitional consistency in the literature may help explain the lack of consensus surrounding the mechanisms of escalation. In order to provide a theory to bridge these often disparate interpretations of escalation, evidence for the importance of paraphilia and life events in considering patterns of escalation in sexual offences will be discussed.
The relationship between paraphilia and escalation Presentresearchhassuggestedthatthepresenceofaparaphiliahasimportantimplicationsfor escalation(Harris,2011).Inthe?ftheditionoftheDiagnosticandStatisticalManualofMen- tal Disorders,paraphiliasarede?nedasdisordersthatinvolveintense,sexuallyarousingfan- tasies, sexual urges, or behaviours along with at least one of the following: (1) non-human objects; (2) the suffering or humiliation of oneself or one’s partner; or (3) children or other non-consenting persons that occur over a period of at least 6months (American Psychiatric
Escalation from fetish burglaries to sexual violence 119
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Association [APA], 2013, pp. 522–523). Paraphilias are more common in sexual offenders who commit homicide than sexual offenders who do not (Koch, Berner, Hill, & Briken, 2011). A potential explanation for this ?nding is that despite efforts to conceal the fantasies, overtime,offenders’ paraphiliastendtobecomeincreasinglyeroticandpotentiallyviolentin nature (Arrigo & Purcell, 2001). Paraphilic behaviour is enacted through a cyclical, reinforcing process, which consists of a relationship between the paraphilic stimuli, fantasy, and orgasm, often enhanced through facilitators such as drugs or pornography (Arrigo & Purcell, 2001). The repeated exposure to a particular stimulus will decrease the size of the response through the process of habituation; causing an ever-increasing desire for greater and increasingly novel stimulus. If the individual’s fantasies and imagery increase in violence and sexuality, then the actual paraphilia tends to escalate in intensity and frequency (Arrigo & Purcell, 2001). Progressively violent behaviours that develop from the paraphilia can manifest in criminal activity, including burglary, assault, rape, and murder, even developing into sadistic sexual activity involving torture, mutilation, and necrophilia (Burgess, Hartman, Ressler, Douglas, & McCormack, 1986)—all of which are behaviours RW engaged in. Two paraphilias particularly relevant to the present case are fetishistic and transvestic dis- orders. Fetishistic disorder (known commonly as fetishism) is characterised by the repetitive use of, or dependence on, non-living objects as the primary elements associated with sexual arousal (APA, 2013, pp. 700–701), the most common of which are female undergarments (Schlesinger&Revitch,1999;Wiederman,2003).Transvesticdisorderreferstotherecurrent and intense sexual arousal from cross-dressing, as manifested by fantasies, urges, or behav- iours(APA,2013,pp.702–703)andfoundexclusivelyinmen(Wiederman,2003).Menwith transvestic fetishism typically portray traditional male gender roles and sexual identity, but cross-dress to varying degrees for sexual arousal (Wiederman, 2003). Regardless, there is typically recognition by the individual that their sexual behaviours are not within the norm and they may make immense efforts to conceal them (Wiederman, 2003). In an early study by Revitch (1978), transvestism was found to be associated with fetish burglaries. Later, in a sample of 52 serial sexual homicides, Schlesinger and Revitch (1999) found that over a third of the sexual murderers had prior convictions for burglary. In those cases, the theft of fetishistic objects was found to be directly related to escalation to obtain (more of) the object and then, although less frequently, to assault or murder the female that occupied the home. Interestingly, the study also found that peeping behaviour commonly preceded burglaries and, in a small amount of cases, also led to assault and murder (Schlesinger & Revitch, 1999)—a tactic many speculated that RW used in order to plan his offences (Appleby, 2011; Gibb, 2011). Further, Langstrom and Zucker (2005), in a large study of 5250 men from 18–74years of age, found that individuals who engaged in transvestic fetishism were signi?cantly more probable to spy on what others are doing sexually and be sexually aroused by the use of pain—again, behaviours in which RW knowingly engaged in (Appleby, 2011; Gibb, 2011; R v. Williams, 2010).
The impact of life events on escalation Although Clarke and Cornish’s (1985) continuance model acknowledges the impor- tance of situational variables associated with an offender’s likelihood of committing a crime, little effort has been made to formally account for life events as a mechanism related to escalation. One notable exception is Pino (2005) who suggested that a
120 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
comprehensive approach to the study of escalation must focus on the various contex- tual and situational life events surrounding the behavioural escalation, including detailed analysis of the offender, the victim, and the police response. Although not directly related to escalation, Hanson and Harris (2000) advocated a similar examina- tion of the life events of offenders, ?nding that acute stressful events typically preceded sexual offending. Interestingly, research conducted on life events and paraphilia indicate that periods of stress are often associated with engagement in transvestic behaviour (Wiederman, 2003). A cognitive-behavioural explanation for positing paraphilia and life events as mecha- nisms in the escalation patterns of sexual offending can be made by modelling sexual deviancy as addictive behaviour (Carnes, 1985; Herman, 1988; Irvine, 1995). Some social learning theorists have proposed a ‘buffering hypothesis’ based on a dynamical systems model of addiction. Proponents of this theory (Gorman et al., 2004) hold that the rate at which an individual engages in the maladaptive target behaviour is moderated by social support that reduces their perceived stress and increased abstinence. For example, Ames and Roitzsch (2000) found that in a sample of individuals diagnosed with substance depen- dence, perceived stress predicted cravings but the level of social support moderated this relationship. As the operant conditioning model of paraphilia (discussed previously by Arrigo & Purcell, 2001) is consistent with addiction models (Brown, 1986; Olsen & Winder, 2010; Van Ree, Gerrits, & Vanderschuren, 1999), it stands to reason that life events, potentially those related to stress and social support, may be related to escalation in sexual offending.
THE CURRENT STUDY
The current study presents a comprehensive case study analysis of the life events and criminal activity of RW with the aim of eliciting better understanding of esca- lation within an offender’s criminal trajectory. Speci?cally, four hypotheses will be investigated. It is expected that increases in escalation will be detected in terms of (1) frequency at which offences are committed and that (2) life events will in?u- ence this relationship. Based upon the research discussed previously, it stands to rea- son that (3) a relationship between paraphilic behaviour and offence escalation exists (Arrigo & Purcell, 2001; Burgess et al., 1986; McCann, 2000; Schlesinger & Revitch, 1999). Lastly, (4) life events will be related to increases and decreases in offence seriousness.
METHODS
Russel Williams’s case history provides a unique opportunity to study the link between paraphilia and escalation in offence frequency and seriousness, as RW not only collected an immensely large set of fetish items, he was also extremely fastidious in documenting the items taken. RW kept detailed records of his offences through digital photography (of items and premises) and through his personal journal. Each photograph was time stamped, allowing us to know when he started taking pictures of the house (usually documenting his arrival to the house) and when he approximately left and/or the cessation of the offence. This allowed conservative estimates of the time spent in the house to be
Escalation from fetish burglaries to sexual violence 121
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
made (Appleby, 2011; R v. Williams, 2010). RW was so conscientious in his record keeping that on one occasion when he realised that he had failed to reset his digital camera for day- light savings, he re-opened saved an additional corrected set of photographs, documenting the error in his journal (Gibb, 2011, p. 373). Most research on crime scene behaviour relies largely on police records (Santtila, Hakkanen, Canter, & Elfgren, 2003) or self-report offender interviews (Beauregard & Leclerc, 2007). The former are limited by the variables collected by the investigators and the latter may be unreliable due to deliberate biassed recall by the offender or simply through memory decay. By utilising the Agreed Statement of Facts presented in the trial (R v. Williams, 2010), which included information based upon the digital photos taken by RW, a more objective and reliable account of the offences is possible. Further, due to the fact that RW was a public ?gure with a prominent public record, including appearances, promotions, and other professional and social engagements, it is also possible to identify some of the possible signi?cant life events and stressors that occurred prior to and during his offending.
Data collection The Access to Information Act (1985) provides Canadian citizens or permanent residents the right to request and be given access to any record under the control of a government institution (s.4[1]). The agreed statement of facts is a list of the objective facts (e.g. offence details) relevant to sentencing a convicted offender. In the case of RW, the agreed statement of facts (R v. Williams, 2010) was based upon a combination of RW’s own confession of the crimes and evidence garnered by police at crime scenes and RW’s residences. The agreed statement of facts was read during his ‘guilty plea’ trial in 2010 and provides a detailed documentation of the 88 offences committed by RW in chronolog- ical order (R v. Williams, 2010). The agreed statement of facts was analysed to create a database detailing various offence parameters and behaviours, coded as categorical or continuous variables. If the presence or absence of a variable could not be determined by the agreed statement of facts, then the behaviour was considered as missing. Time of entry was based upon the timestamps of the digital photographs. Detailed information regarding RW’s method of entry into homes was categorised into skilled (e.g. picked the lock), force (e.g. break a window), easy (e.g. door was unlocked), and unsuccessful (e.g. intent was present but did not enter). Details of the numbers and types of clothing items taken in each offence were provided in the agreed statement of facts. This allowed for the total amount stolen during each offence to be recorded and then broken down into separate variables depending on the articles of clothing (for example, undergarments, bathing suits, and house robes). In the majority of offences the number of pictures taken by RW of fetish objects (e.g. undergarments, RW dressed in undergarments, RW masturbating, and sex toys found in the home) were also known. There was also information regarding each offence on whether or not the victims were aware that someone had broken into their home. A crime scene variable glossary was constructed for the current study (Appendix A). An inter-rater reliability analysis performed on a subset of the cases (n=10) to determine con- sistency between two raters by using the Kappa statistic for nominal variables and inter- class correlation for continuous variables. While Kappa statistics ranged from 0.75 to 1.00 (M=0.98, SD=0.07), all inter-class correlation statistics were 1.00 demonstrating high reliability.
122 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
RESULTS
Descriptive statistics for fetish burglaries Table 1reports the descriptivestatisticsfor the variables codedfrom RW’s fetish burglaries used in the present study. In total, during the fetish burglaries, RW stole 1366 articles of clothing (m=17.5,SD=28.5), the majority of which (59.2%) were bras and panties. The largest theft of clothing occurred on 19June 2009, when RWspent over 2hours inthe house of a 24-year-old woman, taking 186 pieces of clothing. In general, RW burgled homes when victims were not present (87.4%) and spent on average 22.5minutes (SD=47.2) in each house. Offences were evenly split between both ?rst time victims (55.4%, n=46) and repeat victims (44.6%, n=37) and between locations, Tweed (57.8%, n=48) and Orleans (42.2%, n=35). Interestingly, victims noti?ed the police in only 26.5% of the cases (n=22);
Table 1. Descriptive statistics of offence variables for fetish burglaries (n=83) Variable Count or total* % Mean SD Maximum
Series Tweed 48 57.8% Series Orleans 35 42.2% Cool-down 9.2 16.4 135 Previously victimised 37 44.6% Method of Entry: Easy 57 69.5% Used force 11 13.4% Skilled 6 7.3% Unsuccessful 8 9.8% Time in house (minutes) 1645 22.5 47.2 171 Interrupted during crime 11 13.3% Victim present during offence 8 9.6% Children lived at victimised home 31 37.3% Masturbation: None 58 70.7% After crime 7 8.5% During crime 15 18.3% Both 2 2.4% Naked during offence 20 24.4% Naked outside residence 2 2.4% Naked inside residence 19 23.2% Total clothing stolen 1366 17.5 28.5 186 Bras & panties 808 11.7 19.3 115 Slips & camisoles 24 0.3 1.2 10 Swimwear 10 0.1 0.3 1 Pajamas & housecoats 9 0.1 0.4 2 Dresses & tops 19 0.2 1.0 8 Total pictures taken 2937 37.2 48.2 284 Pictures of RW dressed in clothing 322 5.0 12.6 78 Pictures of clothing laid out 896 13.6 23.6 126 Pictures of RW masturbating 208 3.4 10.1 62 Pictures of RW erect 97 1.8 4.6 20 Pictures of victim’s personal items 31 0.5 1.3 6 Total of other objects stolen (e.g. personal items) 38 0.7 1.5 7 Victim noti?ed police 22 26.5%
*Counts and totals re?ect only known evidence/data
Escalation from fetish burglaries to sexual violence 123
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
potentially, suggesting that many did not know they had even been burglarised. This could attesttoRW’sskill,thevulnerabilityofthehomes,ornon-disclosureduetothepotentialem- barrassment for the victim in reporting the crime.
Escalation in frequency of fetish burglaries Regression Regression analyses were conducted to model the frequency of RW’s burglary offences, measured by the number of days between his offences in the burglary series, referred to here as ‘cool-down’. Figure 1 illustrates the increases, plateaus, and decreases in RW’s of- fences across both geographical series of offences combined. The number of days that RW cooled-down by the number of days elapsed from the beginning of the series was analysed by linear and cubic regression estimate functions, as indicated in Figure 1. Both the linear (ß=0.018, SE=0.008, p=0.028) and cubic (ß=-0.217, SE=0.074,p =0.004) regression models signi?cantly predicted cool-down days from days elapsed across the series. The linear model accounted for 5.7%, while the cubic model accounted for 15.1% of the variance; suggesting the cubic model was a better ?t for the data.
Offending phases Figure 1 indicates the temporal time points of major transitions in RW’s life based on a sampling of qualitative information gleaned from data sources (Appleby, 2011; Gibb, 2011; Morrison & Liberrto, 2011; R v. Williams, 2010). Because of the superior ?t of the cubic regression model, an attempt was made to search RW’s life history to associate signi?cant changes and/or events in RW’s working life with periods of increases and decreases in offending. Unfortunately, because of the archival nature of the current study, the authors limited the characterising of life phases to veri?able events (i.e. usually com- prising work or offence events), as it was not possible to con?dently make assertions regarding personal life events. Referring to both the subjective content of each transition
Exploratory Intensification Maintenance Deterioration
Figure 1. RW “cool-down” period between offences by days elapsed in offence series. Notable time points (number of days elapsed – by event): 053 – close call; 188 – resumes offending; 425 – trip to Kandahar; 430 – grooming for colonel begins; 632 – promoted to colonel; 650 – letter (appendix a); 675 – sworn in as commander; 739 – sexual assault #1; 740 – misses of?cial event; 747 – dissociative episode; 776 – sexual assault #2; 807 – homicide #1; 873 – homicide #2.
124 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
phase and thenature of theescalation (e.g. whether it was increasing, decreasing, or stable), each offending phase was named in order to ease interpretation and discussion. The ?rst phase, labelled exploratory, represents thebeginningof RW’s offending history at the age of 44. This stage extended from his ?rst offence on 9 September 2007 to a ‘close call’ incident on 1 November 2007. During the 1 November offence, the residents (who were next door for a barbeque) returned home brie?y to?nd an assailant (RW) inside their home (Appleby, 2011). RW ?ed the house immediately, and although RW was actually acquainted with the family residing there, they were not able to identify him as the intruder. Following that incident, RW did not offend for 135days, marking the longest gap in his offending history. The second phase, labelled intensi?cation, includes offences occurring between 15 March 2008 and 8 November 2008. This phase begins RW’s return to offending after almost being caught during the exploratory phase and continues until he becomes aware that he may be, imminently, promoted to colonel and is tasked to begin activities relevant to promotion (e.g. visiting a Canadian Forces base in Kandahar [Appleby, 2011]) and transfer to a new position. Upon visual inspection of Figure 1, this phase appears to be marked by a clear increase in offence frequency associated with shorter cool-down periods. The maintenance phase begins 12 November 2008, when RW is posted to the Canadian Forces Language School at the Asticou Centre in Gatineau to learn French (an essential step in attaining the promotion to colonel) (Appleby, 2011). This phase sees RW stabilise in offending frequency with relatively longer cool-down periods between offences. It seems at this increasingly busy phase of his work-life, RW offends approximately once a month, with some regularity, suggesting a possible mediating effect of the social changes in this offence phase. The ?nal phase, deterioration (2 June 2009–29 January 2010), occurs prior to the sexual murders and is suggestive of a broad and signif- icant escalation in frequency and deviancy in his offending. This phase, beginning with RW’s transfer to commander of CFB Trenton, is clearly marked by a number of incidents of unusual and uncharacteristic behaviour beginning in September 2009. RW missed several compulsory events formally attended by the base commander as well as reportedly ‘blanking-out’ (potentially an acute dissociative episode) for several minutes during a speech given at the Canadian Air Force ?ying school in Portage La Prairie, near Winnipeg (Appleby, 2011). According to attendees, RW stood blankly staring into the crowd for several minutes before simply continuing as if nothing had occurred. During this period, it was also brought into evidence that RW wrote, but did not actually send, a letter to the fetish burglary victim (described previously) from whom he stole 186 items of clothing (Appendix B, R v. Williams, 2010). In the letter, RW writes in a child-like manner, attempting to divert the police investigation to a young anonymous male. The letter is vul- gar, abusive, and ?lled with spelling and grammatical errors. Regardless of whether RW purposely wrote the letter in this fashion or not, it is wholly immature and probably pro- vides some insight into his mental state during this period.
ANOVA In order to determine if cool-down rate differed signi?cantly between offending phases, an ANOVA was conducted. The cool-down periods signi?cantly differed between offending phases, F(3,83)=5.69, p<0.001. Speci?cally, the exploratory phase was signi?cantly different from the intensi?cation phase (mean difference=30.88days, p<0.001), the maintenance phase (mean difference=25.72days, p=0.013), and the deterioration phase (mean difference=29.29days, p=0.002).
Escalation from fetish burglaries to sexual violence 125
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Table2.Clusteranalysisoffetishburglaryoffencevariables Deviance(Clusters)UnsuccessfulLowMediumHigh
?2orF value VariableCountormean%withinCountormean%withinCountormean%withinCountormean%within
Nakedduringoffence001a4.3%19b86.4%67.5*** Masturbation Aftercrime00730.4%0 Duringcrime001a4.3%14b63.6%67.8*** Both00029.1% Methodofentry(Easy)1a9.1%15b57.7%20b,c87.0%21c95.5%76.1*** Childlivedathome3a27.3%018b78.3%10a43.5%32.8*** Interruptedduringcrime5a45.5%1b3.8%4a,b17.4%1b4.3%13.9** Previouslyvictimised4a36.4%8a30.8%11a47.8%14a60.9%4.90 Totalclothingstolen(mean)016.5a24.2a18.0a Totalpicturestaken(mean)0a21.8a,b55.6b54.1b,c Timeinhouse(mean)0a19.3a,b12.1a,b50.8b N1113.3%2631.3%2327.7%2327.7%
Valuesinthesamerownotsharingthesamesubscriptaresigni?cantlydifferentatp<0.05.TestsareadjustedforallpairwisecomparisonsusingtheBonferronicorrection. *=p<0.05,**=p<0.01,***=p<0.001
126 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Escalation in offence seriousness Cluster analysis of fetish burglaries Four clusters of offence type were identi?ed, as summarised in Table 2, with variables presented in order of their ability to discriminate fetish burglary offences. Cluster 1, labelled unsuccessful, contained (13.3%) offences in which RW did not successfully enter the victim’s home, but where he attempted to do so. Clusters 2, 3, and 4 are described in relative terms of increasing from low to medium to high deviancy, respectively. The low deviance cluster (31.3% of offences) identi?ed burglaries in which RW was not naked nor did he masturbate during the offence. In comparison with medium and high deviancy clusters, in low deviancy offences RW stole the least amount of clothing, took the least number of pictures, and stayed less than 20minutes on average inside the house. Interest- ingly, the houses in which his behaviour was of low deviancy were also houses devoid of children under the age of 13years. Medium deviance offences (27.7%) were characterised by RW being naked and masturbating only once (4.3%) during the crime but nearly a third (30.4%) of the time after the offence. Burglaries in homes where a child under the age of 13years lived occurred signi?cantly more often (78.3%) than not in medium deviancy offences compared with others (?2(3)=32.8, p<0.001). High deviance offences (27.7%) were the most serious of all fetish burglaries. RW was naked inside the victim’s home (86.4%) signi?cantly more often than not in these offences, representing a signi?cant difference to other deviancy clusters (?2(3)=67.5, p<0.001). High deviance offences were also characterised by RW masturbating signi?cantly more often during (63.6%) these offences compared with other clusters (?2(9)=67.8, p<0.001), and in 9.1% of offences, he would also masturbate after the crime. Moreover, RW spent signi?cantly more time (mean=50.8minutes, p<0.01) in the house than in other offence clusters (F(3,69)=3.6, p<0.001) and took signi?cantly more pictures (m=62.3, p<0.05) of fetishistic items during these offences F(3,75)=5.9, p<0.001).
Clusters and offence phase Offence type (as de?ned by the cluster analysis) signi?cantly differed by offence phase, ?2(9)=22.4, p=0.008. Signi?cantly, RW’s exploratory phase (n=5) was characterised by a few medium (40% of offences in this phase) and high (40%) deviance offences. The
Table 3. Frequency of offence (deviance) cluster by offence phase
Deviance (Clusters)
Offence phase*
Total
Exploratory Intensi?cation Maintenance Deterioration Unsuccessful Count 0a,b 2b 6a 3a,b 11 Within phase % 0 5.6% 35.3% 12.0% 13.3% Low Count 1a 16a 3a 6a 26 Within phase % 20.0% 44.4% 17.6% 24.0% 31.3% Medium Count 2a 12a 4a 5a 23 Within phase % 40.0% 33.3% 23.5% 20.0% 27.7% High Count 2a 6a 4a 11a 23 Within phase % 40.0% 16.7% 23.5% 44.0% 27.7% Total 5 36 17 25 83
*Values in the same row not sharing the same subscript are signi?cantly different at p<0.05.
Escalation from fetish burglaries to sexual violence 127
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
intensi?cation phase was characterised by low deviance offences (42.5%) but with greater frequency and number (n=36). The maintenance phase (n=17) was characterised mainly by unsuccessful offences (46.2%). Lastly, the deterioration phase (n=25)wascharacterised mainly by high deviance offences (41.4%). Refer to Table 3 for further details.
Escalation in the sexual assaults and murders Near the middle of the last phase, the deterioration phase, while still committing fetish bur- glaries, RW commits his ?rst sexual assault. This ?rst sexual assault (739days elapsed from the ?rst fetish burglary), as described earlier, signi?es a marked increase in deviancy (e.g. contact versus non-contact offending) and an increase in the risk and brazenness of RW’s offending. RW’s second sexual assault occurring near the end of the deterioration phase (776days elapsed) indicates a substantial increase in violence and callousness towards the victim. However, in neither of these assaults does RW sexually penetrate the victim; instead, he subjects them to increasing physical and mental assaults, posing, binding, fondling, and taking pictures of them naked and in undergarments. The ?rst sexual murder occurs during the deterioration phase (807days elapsed) and is of a female military of?cer who may have been acquainted with RW (Appleby, 2011). This offence shows clear escalation from the sexual assaults and burglaries in the amount of violence used, time spent with the victim, increased callousness, sexual penetration and, ultimately, murder. Interestingly, given the fact that this victim could potentially recognise RW, there are also signs of increases in risk taking behaviour. It is posited that RW, realising the increased chance of being identi?ed, may have actually planned to kill this victim before driving to her home the night of 23 November 2009. The second sexual murder (873days elapsed), also reveals an increasing level of vio- lence (e.g. batters the victim to unconsciousness on several occasions), sexual deviance (e.g. uses sex toys, takes photos, and makes her perform sexual acts on him), and callous- ness (e.g. videotaping the entire offence). RW brutalises this victim, even abducting her and taking her back to his house, repeatedly raping and torturing her, and eventually killing her by strangulation and asphyxiation. Moreover, there is video evidence that sexual acts occurred while the victim was semiconscious and unconscious as well as post mortem; although in the latter case, the evidence obtained indicated the body was posed and manip- ulated, and it is unclear if penetration actually occurred (Gibb, 2011, p. 496). Notably, a sexual act with a semiconscious or unconscious individual is typical of Aggrawal’s (2011) Class I Necrophilia: Role Players, which may be indicative of a risk for more serious necrophilia behaviours, a further escalation in sexual deviance.
DISCUSSION
With the ability of hindsight, we are able to extract information about RW’s behaviour that generates new understanding about sexual offending. Not only was escalation observed in terms of both offence frequency and seriousness, but also, in contrast to escalation models generated by nomothetic research (Armstrong & Britt, 2004; Hoeve et al., 2013; Soothill et al., 2008), non-linear patterns of both types of escalation were observed. Therefore, this study has both academic importance and practical investigative utility in the understanding of escalation in sexually motivated offences.
128 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Overall, statistical analyses con?rmed all four hypotheses. Regression analyses con- ?rmed the ?rst hypothesis of a signi?cant escalation in the frequency of RW’s fetish bur- glaries. However, perhaps the most interesting aspect revealed in this analysis was that cubic non-linear model ?ts the data signi?cantly better than a simpler linear model, suggesting that the overall escalation was not unequivocally linear. This ?nding guided our attempts to explore the available evidence in RW’s life for speci?c events correspond- ing to observable phases of escalation and de-escalation of offending behaviour, as suggested by Pino (2005). While future research should strive to gather primary data from the offender and individuals close tothem, thecurrent study was based on archival dataand thus life events considered were based primarily upon RW’s professional history. Three major life events (very nearly being caught, preparing for command of CFB Trenton, and taking command of CFB Trenton) were identi?ed as corresponding to four distinct phases of escalation and de-escalation in offending. Figure 1 illustrates an initial escalation in frequency during the exploratory phase up to the point where he is almost caught (days elapsed 53days). After a relatively long break from offending, RW resumes offending with a marked increase in frequency corresponding to a phase of intensi?cation (days elapsed 188days). The ?rst cubic asymptote occurs nearing the end of the intensi?cation phase as RW begins to be groomed and prepared by high-ranking military of?cials for promotion to colonel, marking a decrease in his offending frequency over the maintenance phase. During the maintenance phase, a time of stable (e.g. non-increasing) escalation, there was a signi?cant increase in social support and immediacy in RW’s professional life as he was groomed for command of CFB Trenton. Accordingly, the second, downward- spiralling asymptote occurs at the conclusion of the grooming professional and social support when RW actually takes command of CFB Trenton. This transition is marked by a clear deterioration in RW’s psyche coupled with a rapid increase in offence frequency and deviancy. While originating in addiction literature, the buffering hypothesis provides a powerful means to illuminate the non-linear offending patterns evident across the four identi?ed phases. Regardless, it is postulated that the social support (e.g. ‘grooming’) RW received when he was tipped for the position of colonel in the Canadian Air force (e.g. maintenance phase) satiated and/or decreased his motivation to offend. Although this motivational ex- planation has considerable merit as a theoretical explanation, it is acknowledged that more practical aspects may also be relevant to the plateau of offence frequency at this point. For example, RW was considerably busier during this grooming phase and would have had a more restrictive schedule, potentially limiting his ability to offend. Further, RW was prob- ably more ‘underthemicroscope’ during this period with others moreaware ofhis comings and goings, at least, during his daily working activities. Both of these factors may have im- pinged on his ability to offend, even though his motivation to offend may well have been the same or even escalating. Nonetheless, the buffering hypothesis also facilitates a reason- able explanation for the second asymptote; speci?cally, the escalation seen in the deterio- ration phase—the stage at which he becomes colonel and commander of CFB Trenton. It is suggested that the increase in frequency and seriousness of offending in this phase were triggered by the increase in stress (e.g. taking over command of CFB Trenton) un-buffered by the now diminishing social support he had been receiving during the grooming period. Finally, it is important to again note that only life events that pertained to RW’s working life were considered as they were a matter of veri?able public record. Ultimately, it is still unclear whether personal life events may have relatively greater explanatory power in relation to the increases and decreases in RW’s frequency of offending.
Escalation from fetish burglaries to sexual violence 129
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
The application of the buffering hypothesis in the context of sexual offending is an im- portant contribution that will aid future researchers and clinicians. In particular, it would be dif?cult to explain the non-linear patterns of offending displayed by RW without detailed examination of the particular life stressors surrounding his offending (Ward & Beech, 2006). Future research could apply the buffering hypothesis in nomothetic studies in order to investigate non-linear patterns of offending that are moderated by social support and perceived stress to generate generalizable instances of life event stressors. In addition, cli- nicians subscribing to the risk-need-responsivity principles of effective offender rehabilita- tion (Andrews, Bonta, & Hoge, 1990) could potentially utilise the buffering hypothesis to aid in case conceptualization, particularly in the identi?cation of dynamic risk factors. The one-way ANOVA con?rmed the second hypothesis, which the rate of offending differed signi?cantly by life phases. While the only signi?cant difference in offence rate was observed in comparing the exploration phase with the other phases, examining the mean differences reveals that RW offended most frequently in the intensi?cation and dete- rioration phase. In addition to evidence found con?rming escalation of offence frequency, there was also con?rming evidence of escalation in terms of offence seriousness (hypotheses 3 and 4). The cluster analysis of crime scene behaviours during the fetish burglary series produced four types of offences, varying in levels of sexual deviance and overall seriousness. The most discriminative variables between clusters were those related to paraphilic behaviour such as masturbation, being naked, and of course, the theft of erotic clothing (Table 2). Sexual behaviour in burglary is certainly rare and highlights just how unique these burglar- ies would be amongst an aggregate sample of burglaries (Goodwill & Alison, 2006). However, the collection of erotic clothing items, through burglary, suggests some level of excitement and thrill of the direct association of these female undergarments and the deviant way in which they are acquired (Prentky et al., 1989). The escalation of frequency and deviance of the offending can also be explained through particular behavioural mechanisms of reinforcement and compulsion (Coleman, Raymond, & McBean, 2003; Kafka, 1997). It is probable that RW became quickly (and perhaps in- creasingly quickly) sensitised to the undergarment(s) stolen soon after it was collected, worn, and masturbated to, motivating him to re-offend to collect novel stimulus. Similar to other fetishistic sexual offenders, RW’s masturbation enabled a process of behavioural sensitisation and desensitisation to the female undergarments, reinforcing a recurring cycli- cal pattern of compulsive behaviour on each occasion (Mick & Hollander, 2006). In fact, RW’s repeat victimisation of houses (one house in particular was re-victimised eight times) could besuggestive ofarecurring compulsive cycle; returning tothesame house toachieve the same thrill and sexual excitement as previously experienced. However, little is still known of how RW’s sexual paraphilias developed, and although they are seemingly necessary as explanatory factors in his sexual offence escalation, it is still unclear if they are absolutely suf?cient to do so (Donnelly & Fraser, 1998; Proulx, Beauregard, Cusson, & Nicole, 2007). It would be bene?cial for future researchers to expand our understanding of this connection between paraphilia and offence deviance in the context of predicting sexual and violent escalation. From the conceptualisation of RW’s offending as phases of intensi?cation and mainte- nance, it is perhaps unsurprising that offence deviance signi?cantly differed by offence phase. In particular, it is convergent with the buffering hypothesis that RW’s offences during the maintenance phase were largely unsuccessful or of low deviance. Indeed, if he was not as invested in offending during this period, as one might suspect with the
130 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
decrease in offence frequency, he may not be motivated to be as deviant as when he did offend. Contrarily, the larger proportion of high seriousness fetish burglaries during the deterioration phase, as well as the contact offences that were more deviant in nature and by de?nition more serious, indicates an escalation in RW’s deviancy during this phase. More- over,duringthedeteriorationphase,therewereanumberofmoredeviantbehaviourspresent that were rarely or not present in the other phases—such as writing the victim letters, masturbating frequently, and being naked during the offence (R v. Williams, 2010). Generally, it would be helpful in nomothetic studies to examine which crime scene behaviours are most indicative of deviant behaviour and which are most discriminating between offenders. As unusual as RW was in terms of his professional position and age of known criminal onset, his escalation in offending behaviour was not particularly unique in relation to re- search on fetish burglary and sexual violence (Harris, 2011; Liu et al., 2011; Schlesinger & Revitch, 1999). The current study ?ndings provide evidence for the importance of study- ing escalation of both offence frequency and offence deviance and seriousness, particularly in non-contact offences. Moreover, offence frequency and seriousness, evidenced through behavioural examination of multiple paraphilias, were meaningfully related to speci?c life events. By challenging the mainly linear models of escalation generated by nomothetic re- search, this case study provides new insight regarding the trajectory of criminal careers and narratives; in particular, we may have erred in assuming that escalation was a simple linear phenomenon. Moreover, clearly, as evidenced by RW’s gradual escalation from non- contact to contact offending, it is of the utmost importance for researchers and profes- sionals investigating non-contact offences, such as burglary, to attend to aspects and behaviours that are sexually deviant in nature. In light of this, it is strongly advocated that paraphilic non-contact offenders be considered as dangerous sexual offenders with substantive efforts to be made in tracking and capturing these offenders. Likewise, it is imperative that awareness is increased, particularly in the policing and investigative com- munities, that non-contact paraphilic offence(s) may be linked to other contact sexual offences. Investigators should give particular notice to burglaries in their jurisdiction, whicharesaturatedwithsexualmotivation(e.g.fetishitemstaken,evidenceofmasturbation) and bereft of the elements that comprise more common burglaries (e.g. loss of money and valuableitems).Overall,thenatureofparaphilicin?uenceoncriminalbehaviourhascon- siderable importance and utility for understanding the mechanisms of escalation from non-contact to contact sexual crimes.
APPENDIX A CRIME SCENE VARIABLE GLOSSARY
Clothing stolen: RWstole women’s clothesfromthe individuals he victimised.Clothing was divided into bras & panties, dresses & tops, pajamas & housecoats, slips & cami- soles, and swimwear. Cool-down: ~ period was measured by the number of days in-between offences. Method of entry: RW used a variety of methods to enter break into houses; these were grouped into skill. Easy: No skill or technique was required to enter the house (e.g. open door, open window).
Escalation from fetish burglaries to sexual violence 131
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Usedforce: No skill was used but RW used force to enter the house (e.g. broke window, patio door). Skilled: RWusedlockingpickingskillsorhisknowledgeaboutthehousetogainentry (e.g. hidden key) Unsuccessful: RW was not successful in gaining entry to the house. Interrupted during crime: RW was interrupted during the commission of the crime. Children lived at victimised home: Persons under the age of 18years lived in the house. Masturbation: RW masturbated during some of his crimes. Occurrences of masturbation were categorised by the timethey occurred, that is after crime, during crime, both, or none. Naked: RW was occasionally naked at some point during his crimes. Incidences of nudity werecategorisedbywherehewasnaked,thatis duringtheoffenceatsomepoint,insidethe residence, or outside the residence. Other objects stolen: While RW mainly stole women’s clothing, he occasionally stole other personal items. Picturestaken: RWtookseveralpicturesrelatedtothecrimeshecommitted.Thepictures found in his possession were categorised into the following groups RW dressed in clothing: Pictures RW wearing women’s clothing he stole. Clothing laid out: Pictures of women’s clothing laid out. RW masturbating: Pictures of RW engaged in masturbating. RW erect penis: Pictures of RW where his penis is visibly erect. Victim’s personal items: Pictures of victim’s personal items (not including clothing). Previously victimised: RW had victimised the house previously Series Orleans: The crime took place in Orleans, a suburb of Ottawa, Ontario, Canada. Series Tweed: The crime took place in Tweed, a municipality and village located in central-eastern Ontario, Canada. Timeinhouse(minutes): ThetimeRWspentineachhousewasconservativelyestimated, in minutes, by the time and date information imbedded within each digital picture. Victim noti?ed police: Whether a member of the house RW broke into notify the police. Victimpresentduringoffence: WhetheravictimwaspresentwhenRWbrokeintothehouse.
APPENDIX B LETTER TO VICTIM (R v. Williams, 2010)
‘[daughter’s name].doc’, which reads: ‘Beautiful [daughter]. I’m sorry I took these because I am sentamental [sic] to. Don’t worry because I didn’t mess with them. Also I am sure you know your beautiful but trust me your pussy smells fucking awesome! I should know because I been doing this for awhile. But I am going to stop because my moms will fucking kill me if I get caught. She is pretty sure I can be something. Besides your place was kinda like the motherload and I really like that I have a bunch of undies you put on just after you got fucked. I started this with a chick I knew from high school called … who lives down the road from you.Ithoughtitwouldbecooltohavesomeofherundies.ItseemsrightthatI?nishwitha specialchicklikeyou.IfyoudecidetocallthecopstellthemthatIamsorryforthetrouble andtheywon’t here from meagain.Now thatI know allaboutyouIthinkitmight becool tomeetyou.Maybeyoungerguysdon’tturnyouonbutIthinkwecouldbegoodtogether.
132 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Tometeenagechicksareimpressedtoeasy.IguessIwouldliketobewithsomebodymore experienced.Youguysreallyneedtocleanoutthebathinthebasement.Itissomegnarly.I hopewhatIdidain’tpissedyouofftomuch.JTPsSinceIsortafeelguiltyaboutwastingthe cops time these are the places I hit, so they can close there books.’ [sic]
REFERENCES
Access to Information Act, RSC. (1985) c. A-1. Aggrawal, A. (2011). Necrophilia: Forensic and medico-legal aspects. Boca Raton: CRC Press. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author. Ames, S. C., & Roitzsch, J. C. (2000). The impact of minor stressful life events and social support on cravings: A study of inpatients receiving treatment for substance dependence. Addictive Behaviors, 25, 539–547. Andrews, D. A., Bonta, J., & Hoge, R. D. (1990). Classi?cation for effective rehabilitation – rediscovering psychology. Criminal Justice and Behavior, 17, 19 –52. Appleby, T. (2011). A new kind of monster: The secret life and chilling crimes of Colonel Russell Williams. Toronto: Random House Canada. Armstrong, T. A., & Britt, C. L. (2004). The effect of offender characteristics on offence specializa- tion and escalation. Justice Quarterly, 21, 843–876. Arrigo, B., & Purcell, C. (2001). Explaining paraphilia and lust murder: Toward an integrated model. International Journal of Offender Therapy and Comparative Criminology, 45,6 –31. Beauregard, E., & Leclerc, B. (2007). An application of the rational choice approach to the offending processofsexoffenders:Acloserlookatthedecision-making.SexualAbuse–AJournalofResearch and Treatment, 19, 115 –133. Blumstein, A., Cohen, J., Roth, J., & Visher, C. A. (Eds.) (1986) Criminal careers and ‘career crim- inals’ (Vol. 1). Washington, DC: National Academy Press. de Boer, S. F., Caramaschi, D., Natarajan, D., & Koolhaas, J. M. (2009). The vicious cycle towards violence:focusonthenegativefeedbackmechanismsofbrainserotoninneurotransmission.Frontiers in Behavioral Neuroscience, 3. doi:10.3389/neuro.08.052.2009 Brown, R.I.F. (1986).Arousal andsensation-seekingcomponentsin thegeneral explanationof gam- bling and gambling addictions. International Journal of the Addictions, 21, 1001–1016. Burgess, A. W., Hartman, C. R., Ressler, R. K., Douglas, J. E., & McCormack, A. (1986). Sexual Homicide: A Motivational Model. Journal of Interpersonal Violence, 1, 251–272. doi:10.1177/ 088626086001003001 Cale, J., & Lussier, P. (2011). Toward a developmental taxonomy of adult sexual aggressors of women: Antisocial trajectories in youth, mating effort, and sexual criminal activity in adulthood. Violence and Victims, 26, 16 –32. Carnes, P. (1985). Out of the shadows. Minneapolis, Minn: CompCare Publications. Carrington, P. J. (2013). Trends in the seriousness of youth crime in Canada, 1984-2011. Canadian Journal of Criminology and Criminal Justice, 55, 293–314. Chouinard, E. (2011, April). Tire tread database helps investigators catch a killer. The Gazette, 73(1). Retrieved from http://www.rcmp-grc.gc.ca/gazette/vol73n1/cover-reportage2-eng.htm Clarke, R. V., & Cornish, D. B. (1985). Modeling offenders’ decisions: A framework for research and policy. Crime and Justice, 6, 147–185. Cochran, J. K., Sellers, C. S., Wiesbrock, V., & Palacios, W. R. (2011). Repetitive intimate partner victimization: An exploratory application of social learning theory. Deviant Behavior, 32, 790–817. Coleman E., Raymond N., & McBean A. (2003). Assessment and treatment of compulsive sexual behavior. Minnesota Medicine, 86, 42 –47. Cooper, A. J., Swaminath, S., Baxter, D., & Poulin, C. (1990). A female sex offender with multiple paraphilia – a psychologic, physiological (laboratory sexual arousal) and endocrine case-study. Canadian Journal of Psychiatry-Revue Canadienne De Psychiatrie, 35, 334–337. Criminal Code, RSC. (1985), c C-46 s 348(1)
Escalation from fetish burglaries to sexual violence 133
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Donnelly, D., & Fraser, J. (1998). Gender differences in sado-masochistic preferences among college students. Sex Roles, 39, 391–407. Fishman, D. B. (1999). The case for pragmatic psychology. New York: New York University Press. Gibb, D. A. (2011). Camou?aged killer: The shocking double life of Canadian air force Colonel Russell Williams. New York: Berkley Books. Goodwill, A. M. & Alison, L. J. (2006). The development of a ?lter model for prioritising suspects in burglary offences. Psychology, Crime & Law, 12, 395–416. Gorman, D. M., Gruenewald, P. J., Hanlon, P. J., Mezic, I., Waller, L. A., Castillo-Chavez, C., & Mezic, J. (2004). Implications of systems dynamic models and control theory for environmental approaches to the prevention of alcohol- and other drug use-related problems. Substance Use & Misuse, 39, 1713–1750. Harris,A.J.R.(2011).Paraphilia:Incidenceandco-occurrenceinnormativeandsexoffendersamples. Research Report R-235. Ottawa (ON): Correctional Service Canada. Hanson, R. K., & Harris, A. J. R. (2000). Where should we intervene? Dynamic predictors of sexual offence recidivism. Criminal Justice and Behavior, 27,6 –35. Herman, J. L. (1988). Considering sex offenders – a model of addiction. Signs, 13, 695–724. Hoeve, M., McReynolds, L. S., Wasserman, G. A., & McMillan, C. (2013). The in?uence of mental health disorder on severity of reoffending in juveniles. Criminal Justice and Behavior, 40, 289 –301. Irvine, J. M. (1995). Reinventing perversion – sex addiction and cultural anxieties. Journal of the His- tory of Sexuality, 5, 429–450. Kafka, M. P. (1997). Hypersexual desire in males: an operational de?nition and clinical implications formaleswithparaphiliasandparaphilia-relateddisorders.ArchivesofSexualBehavior,26,505–526. Koch, J., Berner, W., Hill, A., & Briken, P. (2011). Sociodemographic and diagnostic characteristics of homicidal and nonhomicidal sexual offenders. Journal of Forensic Sciences, 56, 1626–1631. Langstrom, N., & Zucker, K. J. (2005). Transvestic fetishism in the general population: Prevalence and correlates. Journal of Sex & Marital Therapy, 31, 87 –95. Liu, J. Y., Francis, B., & Soothill, K. (2011). A longitudinal study of escalation in crime seriousness. Journal of Quantitative Criminology, 27, 175–196. Lussier, P., Leclerc, B., Healey, J., & Proulx, J. (2012). Generality of deviance and predation: Crime- switching and specialization patterns in persistent sexual offenders. In M. DeLisi & P. J. Conis (Eds.), Violent offenders: Theory, research, public policy, and practice (pp. 97–118). Sudbury, Massachusetts: Jones and Bartlett Publishers. Mick,T.M.&Hollander,E.(2006).Impulsive-compulsivesexualbehavior.CNSSpectrums,11,944–955. McCann, J. T. (2000). Borderline personality dynamics, fetishism and burglary in adolescence. American Journal of Forensic Psychology, 18,5 –15. Morrison, K. (Host), & Liberrto, J. (Director). (2011). Conduct unbecoming [Television series episode]. In D. Corvo (Executive producer), Dateline NBC. New York, NY: National Broadcasting Company. Nieuwbeerta, P., Blokland, A. A. J., Piquero, A. R., & Sweeten, G. (2011). A life-course analysis of offense specialization across age: Introducing a new method for studying individual specialization over the life course. Crime & Delinquency, 57,3 –28. Olsen, C. M.,& Winder, D. G.(2010). Operant sensation seeking in the mouse. Journal of Visualized Experiments, 10, 2292. Pino,N.W.(2005).Serialoffendingandthecriminaleventsperspective.HomicideStudies,9,109–148. Prentky,R.A.,Burgess,A.W.,Rokous,F.,Lee,A.,Hartman,C.,ResslerR.,&DouglasJ.(1989).The presumptiveroleoffantasyinserialsexualhomicide.AmericanJournalofPsychiatry,146,887–891. Proulx,J.,Beauregard,E.,Cusson,M.,&Nicole,A.(2007). Sexual murderers: A comparative analysis and new perspectives. Hoboken, NJ: Wiley. R v. Williams. (2010). ONSC CR-10-000033. Rainbow, L. & Gregory, A. (2011). What behavioural investigative advisors actually do. In L. Alison & L. Rainbow (Eds.), Professionalizing offender pro?ling: forensic and investigative psychology in practice (pp. 18–34). London: Routledge. Revitch,E.(1978). Sexually motivated burglaries. The Bulletin of the American Academy of Psychiatry and the Law, 6, 277 –283. Santtila, P., Hakkanen, H., Canter, D., & Elfgren, T. (2003). Classifying homicide offenders and predicting their characteristics from crime scene behavior. Scandinavian Journal of Psychology, 44, 107–118.
134 A. E. Brankley et al.
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Schlesinger, L. B. (2002). Stalking, homicide, and catathymic process: A case study. International Journal of Offender Therapy and Comparative Criminology, 46, 64 –74. Schlesinger, L. B., & Revitch, E. (1999). Sexual burglaries and sexual homicide: Clinical, forensic, and investigative considerations. Journal of the American Academy of Psychiatry and the Law, 27, 227–238. Sherman, L. W., Schmidt, J. D., Rogan, D. P., Gartin, P. R., Cohn, E. G., Collins, D. J., & Bacich, A. R. (1991). From initial deterrence to long-term escalation: Short-custody arrest for poverty ghetto domestic violence. Criminology, 29, 821–850. Smith, J. D. (2012). Single-case experimental designs: A systematic review of published research and current standards. Psychological Methods, 17, 510–550. Soothill, K., Francis, B., & Liu, J. (2008). Does serious offending lead to homicide? Exploring the interrelationships and sequencing of serious crime. British Journal of Criminology, 48, 522–537. Van Ree,J.M., Gerrits, M., & Vanderschuren, L. (1999). Opioids,reward and addiction: An encounter of biology, psychology, and medicine. Pharmacological Reviews, 51, 341 –396. Wallace, C., Mullen, P. E., & Burgess, P. (2004). Criminal offending in schizophrenia over a 25-year period marked by deinstitutionalization and increasing prevalence of comorbid substance use disorders. American Journal of Psychiatry, 161, 716–727. Ward, A. K., Day, D. M., Bevc, I., Sun, Y., Rosenthal, J. S., & Duchesne, T. (2010). Criminal trajec- tories and risk factors in a Canadian sample of offenders. Criminal Justice and Behavior, 37, 1278–1300. Ward, T. & Beech, A. R. (2006). An integrated theory of sexual offending. Aggression and Violent Behavior, 11, 44 –63. Warren, J., Reboussin, R., Hazelwood, R. R., Gibbs, N. A., Trumbetta, S. L., & Cummings, A. (1999). Crime scene analysis and the escalation of violence in serial rape. Forensic Science Inter- national, 100, 37 –56. White, J. H. (2007). Evidence of primary, secondary, and collateral paraphilia left at serial murder and sex offender crime scenes. Journal of Forensic Sciences, 52, 1194–1201. Wiederman, M. (2003). Paraphilia and fetishism. The Family Journal, 11, 315–321. Wilcox, D. T., Foss, C. M., & Donathy, M. L. (2005). A case study of a male sex offender with zoosexual interests and behaviours. Journal of Sexual Aggression, 11, 305–317.
Escalation from fetish burglaries to sexual violence 135
Copyright © 2013 John Wiley & Sons, Ltd. J. Investig. Psych. Offender Pro?l. 11: 115–135 (2014) DOI: 10.1002/jip
Copyright of Journal of Investigative Psychology & Offender Profiling is the property of John Wiley & Sons, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder’s express written permission. However, users may print, download, or email articles for individual use.