The criminal sexual activities of offenders against underage people are, of course, extremely varied, including murder with rape through to exhibitionism and other non-contact offences, While there is a remarkable range, this does not mean that either extreme is typical of offending.
Indeed, the extent to which non-criminal sexual activities are more typical of those with a sexual interest in children has not been quantified. For example, masturbation might be the typical response to arousal towards children, even among those with criminal convictions. The number of people whose whose sexual actions towards children consist entirely of secret, private fantasy and masturbation is unknown.
The formal evidence of what offenders do comes from two main sources -- victims and offenders. It can make tedious reading for a number of reasons. There is no standard way of categorizing abuse, for example, due to variations in legal definitions, which vary widely between different countries an regions. The inconsistency of definition makes it difficult to summarize the data. The settings in which information is collected, as well as the sampling techniques used, have a big influence on the outcome. This produces somewhat eccentric swings in the trends found from study to study. One can opt to "fudge" the issue by over-generalizing or by presenting something of the variability found in research reports. Whichever, the topic is an object lesson in the problematic nature of studies of the incidence of offending.
While the definition of sexual abuse can be problematic (Howitt, 1992), it would seem likely that the bulk of offences committed against underage girls are non-contact acts such as indecent exposure, public masturbation an solicitations for sex, or legally less serious contact offences such as fondling or rubbing one's genitals against the body of the victim.
In a stratified sample of young women in Los Angeles, more than 40% of the childhood incidents recalled were non-contact (Wyatt, 1985). Genital fondling accounted for over half of the contact incidents, intercourse and attempted intercourse about a quarter, and oral sex was relatively rare.
Another study of adult women, again in Los Angeles, produced somewhat different findings (Russell, 1983). Very serious abuse
occurred in about a quarter or less of the incidents involving family members, but in three-quarters, approximately, of the recorded acts of strangers, acquaintances, friends and friends of the family. It seems somewhat counterintuitive that such high proportions of extra-familial abuse were serious.
This impression is reinforced by surveys in Britain. Nash and West (1985) studied a sample of women patients registered with a doctors' practice as well as a sample of female students. In general, the students reported substantially more experiences of the "less serious" forms of abuse in their childhoods
The variability between samples might be to do with how they construe abuse. Students, for example, might be more aware of the feminist view of non-contact offences.
One survey had nurses and doctors evaluate abuse cases treated in the emergency room of a Florida hospital. Not surprisingly, given the context in which the data was collected, most of the assaults involved penetrative sex (Cupoli and Sewell, 1988). In the case of boy victims, buggery accounted for most of the cases, followed by the composite category of fellatio on the child, the masturbation of the adult and fondling. For girls, penile penetration of the vagina dominated, followed by a mixed category of cunnilingus, masturbating the adult and fondling, and the penetration of the vagina with fingers. Because girls were more frequently victims than boys, the most frequent victims of anal sex were girls. At a Canadian hospital, Mian et al. (1986) found fondling the most common (40% of cases), although oral sex occurred in 15% of cases and "dry" intercourse
in 14%. The entire sample consisted of six-year-olds and under, which may have affected the characteristics of the sexual activities involved.
Cases of the sexual exploitation of children up to the age of 12 reported by the
Netherlands State Police suggest that the commonest forms of sexual contact involved
the perpetrator fondling
Each of these institutionally-based sources is likely to be biased by being an extreme situation, likely to attract the worst cases. Relatively trivial episodes are probably less likely to be reported.
There exist claims that, on average, child molesters offend against 150 male or 20 female or victims according to their paedophiliac orientation (Abel et al. , 1987), although this may be a selected sample biasing the trends.
One of the most thorough of the studies of offenders reports data obtained from sex offenders at the Minnesota Security Hospital (Erickson, Walbek and Seely, 1988). Consecutive cases over a 10-year period involving offences against children under 14 years of age were examined.
There were few age differences between the male and female victims:
The broad trends in the acts committed were:
1. For females under 10 years
2. For females between 11 and 13 years
3. For males under 10 years
4. For males between 11 and 13 years
These do not exhaust the things done to children by offenders:
Bribery was the most frequent way of obtaining sex; threat was not so common. Size and age discrepancies between adults and children make the question of coercion rather problematic, of course. In terms of what the perpetrators said, physical injury was comparatively rare. Usually offenders remained clothed during the offences. Nudity, when it occurred, was more common in domestic offences. Mostly the victims were individual although offenders might have several contemporaneous victims. Group activity involving several victims in the same episode occurred only with adolescent males.
Methods that inmates at an Ohio correctional institute had employed in their abusing were reported by Budin and Johnson (1989). The median numbers of boy victims was 3.5, and that of girl victims 2. The men tended to prefer a particular type of child, although the attractive characteristics varied substantially from offender to offender. Nearly half of those who answered preferred their own children and/or ones they described as passive, quiet, troubled, lonely children from "broken" homes; two-fifths mentioned children from single-parent homes.
A sort of "empathy" is demonstrated by the fact that 45% mentioned targeting children who reminded them of themselves. Victims were typically found in the close vicinity; over half of them near the offender's home, a third in the neighbourhood, a fifth at family get-togethers and an eighth in playgrounds. Only a fifth molested victims recruited far from the offender's own home.
The child's trust was established in the enticement process by being a friend in 90% of cases. Other aspects of enticement included
Tangible gifts such as toys, candy, cigarettes, beer and drugs were much less common. Perhaps "they would say this, wouldn't they" but less than a quarter claimed to have used threats to obtain the child's cooperation and silence. Of those who had used threats, the threat of hitting was
commonest, although less than 50%; also mentioned were threats to hurt loved objects and family members, and threats with knifes or guns.
At a specialist sex offender treatment programme in Seattle, therapeutically "successful" clients provided information about the characteristics of their offending (Conte, Wolf and Smith, 1989). The men were mostly experienced offenders with an average of seven victims, although the range was from one to forty.
They were asked to write a "manual" on how to sexually abuse a child. Among the examples they provided were:
Most of these offenders also expressed a preference for certain physical or behavioural characteristics in their victims -- generally smooth skin, long hair, dresses, or slim body, darker skinned, darker hair, acute face. Behaviours such as the victims being friendly and open were also mentioned.
After they had identified a potential victim, most of the offenders thought about getting caught and this fear caused them to modify when and how to abuse: "I selected victims that I thought wouldn't report me" (Conte, Wolf and Smith, 19"89, p. 296).
Although some claimed that there was no targeting involved (i.e. a process of becoming interpersonally engaged with the victim), this was unusual; an example would be a man who abused a sleeping victim. Most offenders described how they built up a relationship with the child prior to initiating sexual contact:
The use of adult authority, adult physical presence and isolation of the victim from others were among the methods used to gain mastery of the victim. Very few admitted to the use of physical violence or its direct threat, which suggests that alternative methods were available, such as warning the child to be careful not to tell anyone or withdrawing affection.
Lang and Langevin (1991) claim that about one in five sexually victimized children is subject to gratuitous violence as part of the offence on the basis of a review of the research. Nevertheless, Okami and Goldberg (1992) suggest that some writers substantially overstress the level of violence by ignoring
According to Bradfort, Bloomberg and Boulet (1988), the use of violence depends on the age group of the offender.
It is difficult to know precisely what is meant by violence, since the researchers offer no definitions of their categories.
This may not be typical of younger offenders since Saunders, Awad and White (1986) found that about half of paedophile offences by adolescents involved violence or the threat of violence. Perhaps it is relevant that about a third of recorded rapes are perpetrated against children (La Fontaine, 1990). Given that older children may be fully physically mature, we should not assume that these rapes are necessarily dissimilar in motivation from rapes of adults. That is, perhaps they better classified with rape than with sexual offences against children when considering their psychological implications rather than their legal ones.
Abuse take place at characteristic times. Victimization takes place most frequently on Fridays and Mondays during daylight hours or dusk (noon to 8.00 p.m.) (Peters, 1976). About a third of offences were between 8 p.m. and midnight, although these were almost all offences where the child was left in the care of the offender. Over two-thirds of the offences took less than 15 minutes according to the child.
Adolescents report rather longer periods -- half of offences taking more than an hour.
The offender was known to the child in 80% of cases and the offences took place in the victim's or offender's home in over half of cases. Nearly 80% of cases involved just one offender. This research was done on victims at a general hospital in Philadelphia and at a private psychoanalytic practice.
Maturity and Age Characteristics of Victims
Chronological age and physical maturity do not perfectly correlate with each other although most psychological approaches to offenders follow legal age classifications. There are problems in that the age ranges used to define paedophilia may have an upper limit varying from 12 to 18 years, say, if European or American legal criteria are employed. Taking puberty as the distinguishing factor would mean that the upper limit could be anything from 9 to 18 years.
There may be advantages in using an index of physical development such as Tanner Scores (Lang, Rouget and van Santen, 1988). The Tanner classification (Tanner, 1978) is a scale of 1 to 5:
Because these scores deal with actual physical maturity rather than age, they might have implications for understanding paedophiles' choices of victim. Some paedophiles, for example, say that hairless and smooth skin is what attracted them to a particular child.
A sample consisting mainly of girls in the age range of 1 to 18 years at the University of Alberta Hospital paediatric clinic was studied between 1980 and 1986 (Lang, Rouget and van Santen, 1988). All had been sexually abused, as confirmed by verbal disclosure or from police or social services investigations. A control sample of non-abused children attending the same clinic was used for comparisons. While there was a substantial correlation between Tanner
Classification and age in years, they shared only 50% of their variance in common. That is, age and physical development depart substantially from each other.
It is of considerable interest to find that relatively few (11%) of the victims of abuse were fully sexually developed -- that is, physically adults. Indeed, the great majority of the victims (66%) were at the lowest levels of sexual development (Tanner level 1). Victims tended to be small and light for their age. There was a bimodal age distribution of victims, with peaks at about 4 years and 11 years.
On average, victims were 23 years younger than their abusers. There was no relationship between the age of the offender and the age gap with his victim. In other words, younger offenders did not have a smaller age gap from their victims.
The sexual acts involved depended somewhat on the child's physical maturity.
In some instances, age predicted better than physical characteristics. For example, dry intercourse, anal intercourse and attempted intercourse rejected by the child were all better predicted by age.
Finally, Tanner scores better predicted whether ejaculation took place. It might be that the initiation of the abuse process might be influenced by the child's age characteristics, but when it becomes apparent that the child is physically immature the offender abandons his quest for ejaculation.