To summarize, consideration of sexual fantasies is important in understanding the offending behaviours of child molesters (Abel and Blanchard, 1974). Despite the acknowledgement of the importance of fantasies, little controlled research has been conducted in this area. Research that has been conducted on the sexual fantasies of child molesters has not examined content or actual frequencies (e.g., Marshall et al., 1991), or has not compared groups on the content of the fantasies (Rokach, 1990). In addition, these studies have not examined the conditions under which offenders are likely to engage in deviant fantasies, which may be important for the development of relapse prevention treatment approaches (Russell, Sturgeon, Miner & Nelson, 1989). Many of the arousal reconditioning studies have addressed content or frequency issues, but the studies to date are poorly controlled and with samples too small to allow firm conclusions to be drawn (see Laws and Marshall, 1991 for a review of the masturbatory reconditioning literature).

THEORETICAL IMPORTANCE OF FANTASIES IN SEXUAL OFFENDER POPULATIONS

Finkelhor and Araji (1986), suggested four motivating factors in sexual offending against children: (a) emotional congruence, the offender seeks to have emotional needs met by engaging in sexual activity with the child; (b) sexual arousal, the offender finds the child sexually arousing; (c) blockage, appropriate means of fulfilling needs are unavailable or less attractive; and (d) disinhibition, the usual inhibitions regarding sex with children are overcome. These authors proposed that the offender commits sexual assaults against children due to an interaction of two or more of these factors.

It is hypothesized here that the process of fantasizing by pedophiles may also be explained by these preconditions. First, it is generally agreed that sexual fantasies about children are related to sexual arousal to children (e.g., Abel and Blanchard, 1974).

A second and less obvious feature of sexual fantasies is related to the emotional congruence component from Finkelhor and Araji's (1986) model. Fantasies not only serve a sexual purpose, they also have a strong emotional component (Singer, 1975). It follows that masturbatory fantasies not only serve to produce arousal, but that they also will satisfy some sort of emotional need for the individual.

Disinhibition may also be a factor as an antecedent to inappropriate fantasies. It seems that pedophiles' sexual offenses are more likely to occur when the pedophile is exposed to extreme stress; for example, after arguments with his wife, getting fired from a job, and so on (Pithers, Beal, Armstrong & Petty, 1989). It can be hypothesized, therefore, that pedophiles may also be more likely to fantasize deviantly when under stress, and appropriately when things are going well in their lives. The results of Wilson and Lang (1981) provide some support for this last hypothesis. They reported that frequency of fantasies with deviant themes (sadism, masochism) was related to dissatisfaction in relationships among non-offender males.

The present study was designed to examine the following hypotheses: 1) Child molesters will report more fantasies about prepubertal children than rapists and nonsexual offenders; 2) In light of Finkelhor and Araji's model regarding emotional congruence and disinhibition factors, child molesters will tend to fantasize about children when in a negative emotional state (e.g., under stress or when angry) and about adults when in a positive emotional state.

METHOD

Subjects

Three groups of subjects from two different medium security prisons participated in the study. One group consisted of men who had been convicted of offenses against female children 12 years of age or younger (child molesters). The second group consisted of men convicted of sexual offenses against females 16 years of age and older (rapists). Only men who had female victims were used in order to facilitate matching the two sex offender groups. As well, the men were chosen from currently running treatment groups, or from a list of men accepted for treatment, and who were admitting responsibility for the offense(s) for which they had been convicted. The third group consisted of men convicted of nonsexual offenses, who reported a heterosexual preference. These men served as a "normal" control group and were volunteers chosen at random from the inmate list of their institution.

One possible source of bias in this study is related to demand characteristics of the prison setting. It is possible that sex offender subjects would report information regarding their fantasies in a manner which they believe would help their case in terms of treatment reports and early release. In order to reduce the possibility of this bias affecting results, subjects were informed in writing that participation was voluntary and confidential, and that the information they provided the researcher would in no way be shared with their therapist. They were also informed that the study was in no way related to their evaluation in terms of the program.