Severe Sexual Violence: Addressing the Relationship Between Sexual Violence, Sexual Sadism, and Sexual Homicide

The notion that sexual murderers constitute a unique type of offender can be traced back to Krafft-Ebing’s (1886/1965) seminal work Psychopathia Sexualis where he makes an explicit link between sexual sadism and sexual homicide. Thought to be driven by sexually sadistic fantasies, sexual homicide is often thought to be a behavioural manifestation of sexual sadism. Known as the unique offender hypothesis much of the empirical literature on sexual homicide posits the sexual murder as severely sexually violent and qualitatively different from other types of offenders. More recently, the differential outcome of a sexual assault hypothesis has challenged the long-standing assumption that the sexual murderer is unique. It suggests that sexual homicide may be the result of a series of situational factors present during a sexual assault, and not necessarily sexual sadism. These conflicting findings reflect the theoretical and methodical issues surrounding the scientific study of sexual violence, sexual sadism, and sexual homicide. At the present time there exist few models of sexual homicide and there have been even fewer attempts to test these models empirically. Further complicating matters are the measurement and operationalization issues associated with sexual sadism. This study has three overarching goals. First, to examine the convergent and predictive validity of a series of crime scene variables empirically associated with sexual sadism. Second, to concurrently inspect the utility of both the unique offender hypothesis and the differential outcome of a sexual assault hypothesis. Finally, this study will test the theoretical factors common to prominent sexual homicide models. Overall, the results of this study demonstrated that select crime scene indicators are valid measures of sexual sadism. Moreover, sexual murderers do not constitute one homogenous group of offenders. Instead, there was evidence suggesting the unique offender hypothesis and the differential outcome of a sexual assault hypothesis are both valid. Finally, the core features of existing sexual homicide models (i.e., low self-esteem and deviant sexual preferences) are important in the prediction of sexual homicide.