Zona et al. (1993) were the first to systematically study assault in stalking. They suggested a low risk of overt violence, with only two out of 74 stalkers physically assaulting their victims. In contrast, Harmon and colleagues (1998) reported that 46% of stalkers exhibited violent conduct. The majority of stalkers who threatened their victims subsequently acted upon their stated intentions. Kienlen and colleagues (1997) reported that 32% of stalkers in their study committed assaults, with assault higher among the nonpsychotic subjects with personality disorder or substance abuse. Mullen et al. (1999) reported over a third of the victims in their study were attacked by their stalker. In addition, 6% of stalkers assaulted third parties whom they believed were impeding their access to the target.
These studies are based on samples of stalkers. Victims, however, are in our view the most reliable source of information about intimidation, threats and violence. Hall (1998) reported that 41% of the 145 victims studied had been threatened, 43% had their property damaged, 38% were hit or beaten, and 22% were sexually assaulted. In addition, 11 subjects were kidnapped and two were victims of arson attacks. Pathý and Mullen (1997), in their sample of 100 victims, reported that 58 had been threatened, 36% were assaulted and 7% suffered sexual attacks. Threats preceded assault in 70% of cases. Assault was significantly more likely for victims who had had a former intimate relationship with the stalker. Meloy (1999, 1998) concluded that approximately half of all stalkers threaten the victim. The majority of those who threaten do not proceed to subsequent violence. Nonetheless, threats should be taken seriously, as those who proceed to assault have usually threatened previously. Violence occurs in approximately a third of the cases, yet rarely results in serious physical injury.
There have been several attempts to describe the different types of stalker (Harmon et al., 1995; Mullen et al., 1999; Zona et al., 1993). No generally accepted classification has yet emerged.
Mullen et al. (1999) proposed a multiaxial classification. The first axis was a typology derived primarily from the stalker's motivation, the second from the prior relationship to the victim, and the third a division into nonpsychotic and psychotic subjects. This attempted to capture the stalker's behavior in terms of both motivation and the needs and desires the stalking itself satisfies. They described five subtypes:
- The Rejected respond to an unwelcome end to a close relationship by actions intended to lead to reconciliation, an extraction of reparation from the victim or both. For the stalker, the behavior maintains some semblance of continued contact and relationship with the victim.
- The Intimacy Seekers pursue someone they have little, if any, relationship with in the mistaken belief that they are loved, or inevitably will be loved, by the victim. The stalking satisfies needs for contact and closeness while feeding fantasies of an eventual loving relationship.
- The Incompetent are would-be suitors seeking a partner. Given their ignorance or indifference to the usual courting rituals, they use methods that are, at best, counterproductive and, at worst, terrifying. The stalking provides an approximation of finding a partner.
- The Resentful respond to a perceived insult or injury by actions aimed not just at revenge but at vindication. The stalking is the act of vengeance.
- The Predatory pursue their desires for sexual gratification and control. The stalking is a rehearsal for the stalker's violent sexual fantasies and a partial satisfaction of voyeuristic and sadistic desires.
When the typology, relationship to the victim and psychotic/nonpsychotic dichotomy were combined, the result predicted the duration and nature of the stalking, the risks of threatening and violent behavior, and, to some extent, the response to management strategies (Mullen et al., 1999; Mullen et al., 2000).
The rejected used the widest range of behaviors, such as following, repeatedly approaching, telephoning, letter-writing and leaving notes. In contrast, the predatory stalkers concentrated almost exclusively on furtively following and maintaining surveillance. Intimacy seekers were the most prolific letter-writers, and they also sent the most unsolicited gifts and other materials. Duration was longest in the rejected and intimacy seekers and shortest in the predatory. The psychotic subjects were most likely to send unsolicited materials, and the nonpsychotic to follow and maintain surveillance.
The psychotic and nonpsychotic were equally likely to threaten, but the nonpsychotic were twice as likely to proceed to assault. The rejected were the most likely type to assault and the resentful, although often issuing threats, were the least likely to resort to overt violence.
The best predictor of stalking duration was typology. Also best predicted by typology were assaults. When assaults were combined with substance abuse and a history of prior convictions, they accounted for most of the explained variance.