The scientific approach to the study of ‘child abuse’ has continued both in parallel and largely separate from gendered analyses of the problem. It is as if the politics of a gender based approach is not the proper business of science: the facts should speak for themselves. The empirical evidence base now consists of several prevalence studies across the world on ‘child abuse and neglect’ (WHO 2000) which should be improving understanding and informing responses to the problem. This paper builds on existing feminist arguments that gender plays a significant role in child maltreatment. It proposes that the way in which gender is categorised in prevalence research is insufficient to enable policy and practice to mainstream gender as a key issue informing responses which otherwise continue to reinforce the gender divisions of the countries in which they are based (see for example, Scourfield 2003).
Whilst there are some gender based critiques of the evidence base that highlight the fact that statistics on child sexual abuse in particular show clear gender differences that cannot be ignored, there is less on other aspects of maltreatment. Findings that women physically assault their children in equal numbers to men are reviewed in the context that women spend more time caring for children (Straus, Hamby, Finkelhor, Moore and Runyan 1998; Pringle 1995; Featherstone 1997). These remain partial but not adequate explanations that do not address the meaning of violence. Within the evidence base, ‘violence’ is treated as a global category against which gender (another global category) is manipulated resulting in a somewhat flat analysis of both dimensions.
These interpretations fail to fully acknowledge the significance of gendered social relations and violence in their situated contexts, rather than ‘gender’ and ‘abuse’ as uni-dimensional variables of childhood violence. In addition, the dominant orthodox scientific paradigm and the responses it generates create social relations of difference between ‘victims’, ‘perpetrators’, ‘abusers’and ‘protectors’. In doing so a one sided interpretation of each of these categories is perpetuated that fails to offer appropriate help (see, for example, Milner 2004). There is, thus, a need to develop normatively accepted methods that adequately represent these social relations of violence in their situated contexts through the authority of scientific research.
Sexual abuse is the most researched area amongst child maltreatment prevalence. Whilst there has been considerable debate about methodological variation in relation to such issues as sampling, definition, method of administration, age restrictions and questionnaire design (Bolen 2001; Gorey & Leslie 1997; Finkelhor 1986; Leventhal 1998), the research has contributed to what is considered to be an increasingly robust evidence base (Bolen 2001; Macdonald 2001). Definition is given particular importance in the orthodox approach to enable comparison between studies and to allow for choice about what should be included or excluded. For example, a study conducted in the UK found that one in two girls and one in five boys had an unwanted sexual experience before the age of 18 (Kelly, Regan, Burton 1991) which in broad terms could be defined as ‘sexual abuse’. Using a narrower definition (including a minimum five year age difference between the people involved and narrowing the definition to include penetration or coerced masturbation) the prevalence rate was 5% of girls and 2% of boys. Taking account of varying definitions, two meta analyses in the US found prevalence rates in national surveys ranging from between 12%-17% for girls and 5-8% for males (Gorey and Leslie 1997) and from 2% to 16% for males and from 8% to 30% for females with a mean prevalence of sexual abuse across studies of 9% for males and 19% for females (Bolen 2001). In a review of studies conducted outside the US, an estimated prevalence rate of 20% for females and 10% for males was considered realistic (Finkelhor 1994). The only gender relevant conclusion to be drawn from child sexual abuse (CSA) prevalence research is that girls are at 2 to 3 times more likely to be sexually abused than boys.
In most cases over the last 10 years, CSA prevalence research reports fail to give specific information that is aimed at probing gendered sexual relations and assaults in any detail. This is despite early research in the 1980’s and feminist texts over the last decade that clearly mark out connections between gender, power relations and sexual abuse (Russell 1984; Driver and Droisen 1989; Nava 1992; Kitzinger 1990; Itzen 2000). It is therefore necessary to revisit prevalence data, where such information is available, in order to begin to explore the relationship between gender and child maltreatment. Bolen (2001) has done this in relation to CSA in the US. Much of the discussion that follows is based on that secondary analysis of data from Bolen’s work and a national random probability prevalence study by Cawson et. al. (2000) conducted in the UK, on which a similar secondary analysis has been undertaken. This enables comparison with Bolen’s US study, bringing it into a European context.
Bolen’s secondary analysis of data from Russell’s (1984) research specifically explores evidence showing large proportions of CSA for both males and females are extra familial. Cawson et. al. (2000) similarly find that the majority of CSA is extra familial (see Table 1).
|
|
Intra Familial |
Extra Familial |
||
|
|
Male |
Female |
Male |
Female |
|
Russell 1986 |
N/A |
16 |
N/A |
31 |
|
LATP 1985 |
2 |
8 |
12-13 |
17-20 |
|
Cawson et al 2000 |
1 |
3 |
11 |
20 |
Figure 1: Gender and relationship (Bolen, 2000 and Cawson et. al. 2000)
The most common definition of CSA in prevalence studies is that the sexual experience was unwanted (not consented to) or the other person involved was five or more years older. Some studies present a global figure for CSA, without breaking down the age, sex or relationship of the other person involved in the unwanted act. Where information is available, this definition identifies a high proportion of peer sexual assault, what might be referred to as sexual ‘dating’ violence (Kelly et. al. 1991). For example, in the UK study by Cawson et. al. (2000) between 58% (touching) and 70% (penetrative sex) of respondents described the person involved in the unwanted sexual experience as a boyfriend or girlfriend. Thus, the evidence points to CSA as a gendered problem, inherent in gendered social, sexual, relations – although this is rarely made explicit.
A further gender relevant finding from the evidence base is that males are more likely to be responsible for the assault. Bolen notes that the first random studies in the US suggested that 95% or more of offenders were male, however, like Russell’s 1984 research, many of these early studies only focused on females. Finkelhor, Hotaling, Lewis, and Smith (1990) found that 17% of male victims, as compared to 1% of female victims, were sexually abused by females. Cawson et. al. (2000) find that, for girls, between 95-99% of those believed responsible are male but for boys this is lower (30%-78%). Their study confirms that females are more likely to sexually assault males, but also shows a distinction in the type of sexual acts they perpetrate. For female respondents, women were significantly more likely to be involved in voyeurism or pornography with other males, rather than in other sexual acts. Only 2% of unwanted penetrative/oral sex acts were perpetrated by females on females in contrast to 70% by females on males (see Table 2).
|
|
Penetrative/oral acts |
Attempted penetrative/oral acts |
Touching |
Voyeurism/ pornography |
Exposure |
|
Female resp. |
6% |
6% |
10% |
1% |
7% |
|
Male involved |
97% |
99% |
95% |
95% |
99% |
|
Female involved |
2% |
2% |
5% |
33% |
3% |
|
Male resp. |
1% |
2% |
3% |
2% |
2% |
|
Male involved |
30% |
43% |
27% |
78% |
55% |
|
Female involved |
70% |
67% |
74% |
27% |
45% |
Figure 2: Gender of respondent by gender of perpetrator (Cawson et. al. 2000)
These findings should be interpreted in the context of a higher prevalence of CSA on females in that rates varied from 1% (voyeurism/pornography) to 10% (unwanted sexual touching) for girls in comparison to between 1% (penetrative/oral acts) to 3% (unwanted sexual touching) for boys. Thus while the proportions of female assailants of boys are higher they apply to a smaller percentage of the general male population (see Table 2).
When asked to what degree force was involved the findings are more mixed. Cawson et. al. (2000) found that females are more likely to be physically forced into sexual acts. They are more than twice as likely to experience the use of force for penetrative or oral sex acts and over 3 times as likely to experience physical force in touching or fondling (see Table 3). Thus, males are more likely to coerce their victims than females.
|
|
Penetrative/oral acts |
Attempted penetrative/oral acts |
Touching |
Voyeurism/ pornography |
Exposure |
|
Female resp. |
6% |
6% |
10% |
1% |
7% |
|
Force used |
37% |
40% |
35% |
17% |
30% |
|
Male resp. |
1% |
2% |
3% |
2% |
2% |
|
Force used |
14% |
32% |
11% |
0% |
21% |
Figure 3: Gender of respondent by physical force used (Cawson et. al. 2000)
Bolen argues from the existing evidence base that (US) society’s responses to CSA are failing because assumptions have far more to do with the historical conceptualization of child sexual abuse than with the empirical knowledge base (Bolen 2001). Like many others (see for example, Herman and Hirschman 1981) she places the historical roots in Freud’s initial discovery and then analysis of incest as fantasy which resulted in denial of the problem. Once incest was acknowledged as a reality, the Freudian legacy and the introduction of family systems theory (see for example, MacLeod and Saraga 1986) encouraged a ‘victim blaming’ stance. It was not until the end of the last century that CSA was recognised as something that did happen both inside and outside the family, that it was not the child’s ‘fault’, and something that the state should respond to in most of the developed world. Despite the overwhelming evidence, however, the primary focus of existing responses, and in much of the literature, has remained on intra-familial abuse.
The evidence overwhelmingly points to CSA as a problem affecting a significant minority of the population, both male and female. Over 90% of all CSA is perpetrated by males and approximately one third of this involves physical force. Approximately two thirds of CSA is extra-familial and appears to be rooted in gendered sexual relations, particularly around dating relations. Bolen’s analysis of the empirical evidence results in her describing CSA as an ‘epidemic’ and concluding that the most effective way forward is to recognise CSA as a problem of gendered socialization.
‘To effectively target an epidemic often requires changes on the part of all members of society, who must buy in to the need for addressing the epidemic. Yet, today’s social milieu is not one in which most Americans will willingly commit to some of the changes in socialization patterns of the power structure that will be necessary for substantially reducing the problem of child sexual abuse” (Bolen 2001, 274).
Unlike child sexual abuse, ‘physical abuse’ is defined through the obligation to care, a duty that usually rests with the parent. For example, the World Health Organisation (2000) states that ‘physical abuse of a child is defined as those acts of commission by a caregiver that cause actual physical harm or have the potential for harm’ (WHO 2000, 60). In most prevalence studies physical abuse is measured by the Conflict Tactics Scale (Straus 1979; Straus et. al. 1998). The scale was developed on the basis of conflict theory which proposes that all human relationship will contain conflict. This conflict can be resolved in many ways ranging from discussion to murder. The Conflict Tactics Scale rates maltreatment items as violent acts by parents or carers that have a high probability of causing injury. These items include kicking, biting, punching, hitting (or trying to hit) with something, beating up, threatening with a knife or gun or using a knife or gun. A comparative study across five countries using this scale gave an incidence rate of between 4% and 25% of severe physical punishment reported as used by mothers over a 6 month period (WHO 2000). A survey conducted in the US using a modified version of the CTS (Conflict Tactics Scale Parent to Child, CTSPC) found a prevalence rate of 49 per 1000 children (Straus et. al. 1998).
As with CSA ‘sex’ of victim and assailant is treated as a variable but gender differences are rarely discussed, and neither is the distinction between violence inside and outside the family. The findings on victim ‘sex’ vary. For example, two US studies find that males are more likely to be victims but in the UK there do not appear to be any significant differences (see Table 4).
|
|
Males |
Females |
|
MacMillan et. al. 1997 |
31% |
21% |
|
Scher et al. 2004 |
21% |
17% |
|
Cawson et al. 2000 |
21% |
20% |
Figure 4: Examples of Prevalence of ‘Physical Abuse’ by Gender
Most studies find that mothers or female carers are assailants in equal or greater proportion to fathers or father figures. When asked which parent or carer was responsible for the violence, Cawson et al. found 51% of female respondents identified their mothers or step-mothers and 47% identified their fathers or step-fathers. Forty eight percent of male respondents identified their mothers/step-mothers and 50% identified their fathers/step-fathers as responsible. Thus, in intra-familial child violence, females appear to be responsible for the violence and also victims in more or less equal numbers to males.
Following Bolen’s analysis of the differences between intra-familial and extra-familial sexual abuse it may be helpful to review findings on extra-familial physical violence. Whilst it is undeniable that some children experience violence within the home it is also the case (as with child sexual assault) that they experience more violence outside of it. Youth violence research tends to be seen as separate and other than ‘physical abuse’ (see, for example, the organisation of the World Report on Violence and Health, WHO 2000). A youth victimisation study conducted in the US found that young people are among the most highly victimized groups with overall violent crime victimization rates and injury rates for youth 12 to 17 years of age both 2.7 times higher than the rate for adults (Hashima and Finkelhor 1999). Yet many European countries (including the UK) do not include young people in their national crime victimisation surveys. This prevents analysis of victimisation and violent crime amongst children and adolescents. In the US study, boys were approximately three times as likely as male adults to be victims of aggravated assault. In contrast, girls were four times as likely as adult females to be victims of sexual assault and they were three times as likely as female adults to be victims of verbal threat of assault (Hashima and Finkelhor 1999). The failure to recognise interpersonal violence as a significant problem for young people may be to do with the fragmentation of childhood violence into categories such as ‘bullying’, ‘child abuse’ and juvenile crime which prevents an integrated analysis of gendered relations in the context of interpersonal violence in childhood. Hashima and Finkelhor (1999) found that young people were more likely to know their assailant than adults (64% of juvenile victims compared to 51% of adults). The two groups of assailants identified most frequently by young people were ‘schoolmate’ (29%) and ‘friend’ (15%); two categories often associated with ‘bullying’. As with all forms of violence there was a low reporting rate and 36% of young people stated they dealt with the violence in another way. Further, Hashima and Finkelhor (1999) suggest that causal relationships may become blurred as violence correlates with the risk of delinquency. Within the evidence base children are presented either as victims or as ‘villains’ and, as with adult intra-familial violence, the complexities produced by participants not fitting neatly into categories are lost. The US study identifies the fixed spatial location of children as a contextual factor:
‘When children live in families that mistreat them, they are not free or able to leave. When they live in dangerous neighborhoods, they cannot choose on their own to move. If they attend a school with many hostile and delinquent peers, they cannot simply change schools or quit. They cannot drive around in private cars, live alone, or work in limited-access offices and factories as can adults. This absence of choice over people and environments may affect juveniles’ vulnerability to intimate victimization and street crime’ (Hashima and Finkelhor 1999, 816).
Thus, important connecting characteristics between intra and extra familial childhood physical violence have more to do with the social construction of childhood and its limitations than parental behaviour per se, and childhood relations are known to be gendered. Integrating data on physical and emotional intra and extra familial violence in childhood may enable the field to develop further understanding of the gender, power and age relations that span interpersonal violence in childhood.
For example, Cawson et. al. (2000) found that 31% of respondents had been bullied during their childhood by their peers and 3% had been bullied by adults. There were no significant sex differences in terms of overall bullying which included verbal and psychological violence as well as physical violence. There were, however, significant differences between the experiences of males and females in relation to physical assaults by peers with boys being twice as likely to have been physically bullied (see Table 5) adding to the evidence on gender differences found in the US victimization study and confirming that certain types of violence and masculinity are related during childhood.
|
|
Males |
Females |
|
Serious Intra-familial |
6% |
8% |
|
Intermediate Intra-familial |
15% |
12% |
|
Bullying (peers) |
20% |
10% |
|
Bullying (adults) |
1% |
1% |
Figure 5: Prevalence of Physical Assaults in Childhood (Cawson et. al. 2000)
When asked why they thought they had been bullied respondents give a mixed set of replies. The most frequently given reason was ‘size’, followed by ‘class’ and ‘intelligence’. The connection between all the categories is difference; boys and girls perceive themselves to be physically, verbally and psychologically punished because of a characteristic that marks them out from others (see Table 6).
|
|
Male % N=530 |
Female % N=669 |
Total N=1199 |
|
Size |
29 |
23 |
26 |
|
‘Class’ |
20 |
22 |
21 |
|
Intelligence |
23 |
15 |
19 |
|
Interests |
14 |
6 |
10 |
|
‘Race’ |
9 |
7 |
8 |
|
The place you lived |
7 |
7 |
7 |
|
Appearance |
4 |
6 |
5 |
|
Disability |
4 |
2 |
3 |
|
Sexuality |
3 |
1 |
2 |
|
Don’t know |
19 |
23 |
22 |
Figure 6: Reasons for Bullying (Cawson et. al. 2000)
Similar patterns can be found in relation to ‘emotional abuse’. Intra-familial childhood violence studies find that emotional or psychological violence is more prevalent amongst girls but that if extra familial emotional violence is included the rates are high for both boys and girls, who seem to be affected in almost equal proportions (see Table 7).
|
|
Males |
Females |
|
Scher et al 2004 |
9.6 |
14.3 |
|
Cawson et al 2000 (parents) |
4 |
8 |
|
Bullying verbal |
37 |
36 |
|
Bullying Humiliation |
18 |
18 |
Figure 7: Prevalence of Emotional Violence
It is not possible for neglect to be reviewed across the intra/extra familial divide because, by definition, it refers to a duty of care by parents. Several commentators have noted how the expectation of duty of care most frequently falls on mothers and how it is mothers and not fathers who are generally the focus of intervention in child protection services (Farmer and Owen 1995; Featherstone 1997). There have been very few studies done on the prevalence of neglect. However, two show that prevalence rates reveal differences between males and females, with boys reporting neglect by their carers more frequently than girls (see Table 8).
|
|
Males |
Females |
|
Scher et al 2004 |
22% |
14% |
|
Cawson et al 2000 |
17% |
14% |
Figure 8: Prevalence of Neglect
In addition, a national incidence study found that in the US boys were significantly more likely to be reported as emotionally neglected and that rates for boy’s physical neglect increased greater than for girls over a 6 year period (Sedlak and Broadhurst 1996). Cawson et. al. (2000) hypothesise that this may be because boys are allowed more freedom than girls and are subject to less supervision, again reflecting cultural norms in gendered socialization (see also Straus et. al. 1998).
Salient findings from reviewing the empirical evidence base for all forms of child maltreatment are that; girls are more likely to experience sexual violence than boys; men are more likely to perpetrate sexual violence with force than women; boys and girls are equally likely to experience physical violence within their households; women and men are equally likely to perpetrate physical violence on boys and girls at equivalent rates in their households; boys are more likely to experience physical assault outside the home; girls are more likely to experience sexual and verbal assault both inside and outside the home, and more likely to experience this violence outside it; boys are more likely to experience neglect (most of which is supervisory neglect); women are more likely than men to be held responsible for protecting children, both inside and outside the home.
According to Winch (1989) science provides for a paramount reality which appears to be in some way superior to other realities (such as a reality which makes oracles meaningful, or acts of God). The methods through which science knows reality leads to a partial and somewhat distorted view of the social world and within that, social problems such as violence to children. Measurement is an artifact of the lifeworld and is only one way of knowing it. As Cicourel pointed out many years ago:
‘Measurement pertains to properties of objects, and not to the objects themselves. Thus, a stick is not measurable in our use of the term, although its length, weight, diameter, and hardness might well be...’ (Cicourel 1964).
To create a variable such as ‘sex’ or ‘abuse’ an initial imagery of the concept is required which immediately imposes one version of reality. It is then necessary to specify the dimensions of the concept and select observable indicators. This describes the definitional activity that goes on in relation to ‘child abuse’ and ‘gender’ when located within a scientific paradigm. The construction of a variable requires:
‘an initial imagery of the concept, the specification of dimensions, the selection of observable indicators, and the combination of indicators into indices’ (Lazarsfeld 1959, cited in Cicourel 1964, 15).
Within a measurement epistemology there must first be an initial imagery of the concept. For ‘child abuse’ this is generally specified along such dimensions as physical, sexual, neglectful, emotional. In the case of physical and emotional abuse and neglect, indicators of these dimensions are aspects of the physical condition of the body e.g. healing fractures, bruising, bites, burns, weight (but sometimes service level data), or actions likely to cause injury (such as hitting with a hard implement, shaking) combined with identification of a person responsible from within the family. In the case of CSA indicators are the withholding of consent or an age difference between those engaged in the sexual act of five or more years. The selected indicators for gender are physical characteristics that reveal a body’s ‘sex’ or self-identification of a sex category by respondents.
Following, Dorothy Smith (1974) it may be helpful to describe the process of indicator selection and then definition of the categories of ‘child abuse’ and ‘gender’ which inform practice and theory knowledges as ‘theorising tricks’. Smith gives the ‘tricks’ as:
In the measurement of ‘physical, sexual, emotional abuse and neglect’ mother’s, father’s and children’s experiences are separated from their situated circumstances and contexts. The subjects selected become detached from the ‘original practical determination’ of their becoming countable in the first place. Aspects of experiences of child rearing and socialization are also separated out from their situated circumstances and contexts in order to establish variables. This results in uni-dimensional measurement of both childhood violence and gender. The subjects selected become detached from the original practical determination of their becoming countable in the first place. However, the coding categories come to stand for and represent ‘child abuse’ and ‘gender’. The selected behaviours and characteristics are arranged in an order that accounts for what is claimed to be observed (‘child abuse’ and relationships between ‘child abuse’ and other variables). They are then re-attributed to ‘reality’ by attributing them to actors who come to represent the ideas: for example, ‘maltreating mothers’, ‘abusing fathers’ and ‘abused children’ incorporating hegemonic versions of masculinity, femininity and ‘child abuse’. Yet coding categories and ‘substantiated cases’, are never seen as artefacts of the processes of their construction. The consequence is the collection of a vast array of human interaction and moral judgement into a global category (‘abuse’) which, in turn, leads to and authorises collectivised (intra-familial surveillance and risk assessment) responses. The global categories of ‘male’ and ‘female’ similarly collectivise a variety of masculinities and femininities but work to invoke a hegemonic masculinity and femininity when used as coding categories in research (Pringle 1995).
None of this would matter much if the topic under consideration was something morally indifferent. However, ‘child abuse’ has come to represent a particular, negatively constructed, immoral set of actions that link to blame, stigma, criminalisation and pain. It has created its own hegemonic discourse that constructs interpersonal violence in specific ways. ‘Gender’ as a variable is restricted to descriptive power and loses any explanatory force. It is not helpful either for understanding or prevention of childhood violence to use either ‘abuse’ or ‘sex’ as variables. There is a need for much greater specificity within the child violence field and, in addition, greater attention to context. It is clear, and unsurprising, that gender is implicated in these violent social relations (as it is in all social relations) but more sophisticated data collection and analyses are required that reflect the complexities of both violence and gender.
I have selected two studies that begin to highlight some of the complexities that the current empirical science based work seems to avoid. Firstly, Thorpe and Jackson (1997) attempted to go behind the statistics to find out why the prevalence of ‘physical abuse’ was so high among mothers, when women seem under represented in other forms of violence. They examined 147 ‘physical abuse’ referrals made in Western Australia between 1988/99. The first finding was that only a third of these (34%, N=50) were substantiated (Thorpe 1994; Gibbons et al. 1994). Similar levels of substantiation have been found in other studies and suggest that the over representation of women and single parent mothers in child protections systems have more to do with the surveillance of child rearing practices, motherhood and risk than actual violence to children (Thorpe 1994; May-Chahal et al. 2004). Levels of substantiation should indicate caution when basing any claims on service level data. In the analysis of file text carried out by Thorpe and Jackson it was found that the majority of children in substantiated cases were reported as a consequence of excessive corporal punishment (the excessive disciplining of children in order to socialize them).
They refer to the work of Everingham (1994) who researched mother’s methods of discipline in three playgroups and found that techniques were normative for the group. In one group she observed that the mothers used more extreme methods of discipline and expected other parents to apply harsh punishment if their child deserved it. She contrasts linear and cyclical discipline. Cyclical discipline is regulated through ‘intersubjective social relations (subject to subject) which operates when each of the actors is trying to understand the other’s perspective. Linear discipline is regulated through ‘a subject – object social relation’ where one actor attempts to impose his or her definition of a situation on another. Thus different disciplinary styles are the outcome of different social relations. Thorpe and Jackson find that actions of mothers and fathers in their sample could be explained within it. Of the 41 children where data was available it was found that the majority (N=33) fell into the ‘subject object’ relational category. The child was hit because they were perceived to have behaved wrongly. For example, in one case concerning two teenage girls it was stated that ‘their father often slapped them around the head and face or hit with a belt or banged their heads together. This happened if the were late or if the chores were not done properly’ (Thorpe & Jackson 1997). Where parents live together, both may consider their actions justified. For example, in the case of a young boy who was hit in the face and stomach by his father for stealing, the mother stated that ‘she considers father fair, and that he is the disciplinarian of the family…physical punishment is the only one that works’. However, different relations are found in different cases, for example, the ‘father involved in a case concerning a 2 year old boy who was kicked for screaming, stated that ‘he believes in discipline’, whilst the mother was described as ‘passive’. Children old enough to give their views appeared to have the attitude that they ‘deserved it’, or it was so accepted as part of family life that ‘physical abuse’ was not the greatest issue’ (Thorpe & Jackson 1997). Eight of the cases were described as subject – subject relations where women used physical methods of discipline as ways of reducing risks to their children. For example, three children were hit by their single female parent to confine them to the house as the family lived in accommodation next to a very busy highway. These children were not hit at other times. In another case a mother hit her 15 year old daughter in a ‘one-off’ incident because of the risks she felt her daughter faced by continuing a dating relationship. In all cases the punishments were secondary to other considerations about children’s welfare. Physical punishments were not ‘systematic or consistent’, unlike the ‘subject-object’ group. What begins to unfold in the stories behind the statistics are genuine attempts by both parents to socialize children where violence is used as part of that duty. This observation, along with the finding that excessive corporal punishment was the primary reason for harms and injuries reported to child protection services in 8 European countries, led the Co-ordinated Action on the Prevention of Child Abuse in Europe to reframe this type of ‘physical abuse’ as excessive care relations (May-Chahal et. al. 2004). Such an analysis also points to the complexities involved in defining ‘violence’ as wholly negative or ‘abusive’ where cultural expectations provide a context of normative disciplinary practices. The gendered dimension requires development in that mothers and fathers (or substitutes) were often reported to be in agreement.
Secondly, one of the first feminist studies of mother’s physical aggression found that anger and frustration were the norm for mothers caring for infant children (Graham 1980). The 111 mothers, who were all interviewed when their babies were approximately one month old, found different strategies for dealing with their anger, from sticking their head in the airing cupboard, leaving the room or the house, to hitting the cot and smacking. The research found some interesting class differences which remain under researched in subsequent studies. Namely, that women who had previously been in professional occupations (Social Class AB) and those who had been employed in semi and unskilled manual jobs (Social Class DE) were more prone to feel guilt, anger and aggression. Those who had been in semi-skilled non-manual occupations (Social Class C) appeared to cope better. Most physical abuse prevalence studies find much higher rates in the lower socio economic groups. For example, significant differences were found in relation to socio-economic status in a national family violence survey in the US, with families earning less than $20,000 per year having the highest rates of severe violence (Gelles and Straus 1987). Cawson et al. (2000) similarly found a trend associated with respondent’s social grade for severe violence (4% of Abs, 5% of C1s, 7% of C2s and 12% of DEs). However, these studies rely on retrospective self-reports and are influenced by memory, which is unlikely to recall incidents experienced under 3 years of age. In Graham’s study, the majority of the women said they could understand why mothers ‘batter’ their babies after their own experiences of child rearing. The study ends by suggesting that a pressing question is why more women do not assault their children, rather than why a minority do. Both these studies highlight how narrow definitions of ‘sex’ and ‘abuse’ do not allow for exploration of child rearing and aggression in normative contexts and indicate the need for development of contextualised explanatory frameworks for violence.
Prevalence study categorisations of ‘sex’ and ‘abuse’ perpetuate narrow constructions of gender and violence in childhood. This prevents development of the childhood violence knowledge base and impedes the development of gender specific solutions and responses. In all countries where there are services, these are primarily dedicated to responding to child abuse in its narrow construction and the focus of policies and organisation is primarily on children in the domestic sphere, maternal and to a limited extent paternal responsibility. This is turn results in:
At the same time prevalence research reveals that experiences of childhood violence are likely not to be reported. As a consequence, girls and boys, women (as mothers) and men (as fathers) are left to deal with these tensions, contradictions and violences privately and individually in the context of public discourse; they are enabled to distance themselves or be distanced (Wise 1991; Leudar, Marsland and Nekvapil 2004) and thus fail to seek appropriate help, feel stigmatised (abnormal) and social relations of difference are created and made possible between ‘victims’, ‘perpetrators’, ‘abusers’, ‘protectors’, ‘providers of services’.
A collectivising scientific approach respecifies ‘the sheer circumstantiality of ordinary activities so that order can be exhibited analytically’ (Garfinkel 1996). The vast range of details that are presented by individual experience present analytic problems. The problem is a ‘normal natural trouble’ founded on the enterprise of abstracting from daily life what is articulated or formulated for other purposes. There are many different ways in which people who claim to have experienced violence express their experience, its relevance and what is relevant to them, which cannot always be bound by formal analytic claims. Whilst some of the details can be collated into similar types of categories there are many elements of accounts, and of the situations in which they are located, occasioned and achieved, that are not available in this way. A challenge to the empirical evidence base is to find a way of attending to context, to examine normative interpersonal violence in more detail (as Graham recommended in the early 80’s) and to conduct research that examines violence in the context of gendered social relations. There is, perhaps, a need for some preliminary work on childhood violence that broadens thinking on what could be measurable in any collective sense to enhance understanding of the meaning and context of the experience.
Whilst it is important to broaden definitions of sexual violence to encompass both intra and extra familial experience as Bolen suggests, this recommendation needs to be reviewed to take account of the narrowing focus that terms like ‘abuse’ currently encourage and to be applied to all forms of interpersonal violence; to not view the relationship simply as adult men perpetrating violence on girls or ‘mothers’ and ‘fathers’ ‘abusing’ their children. The consequences of these uni-dimensional variables are many but most importantly they enable the placing of people into categories and channel inappropriate responses (Milner 2004). Reality is rarely so neat. There is a need to view childhood violence in the context of reinforcing the gendered structure of patriarchal society, whether that be in the home, school or workplace, perpetrated by men and women, girls and boys on each other often in order to perpetuate gendered social relations and always in a gendered social context. Authoritative and legitimised methodologies must be developed that allow us to get to the dialectics of violence, both positive and negative. Statistical analysis is no substitute when it is informed by ‘flat’ variables. The question is, what is it about childhood violence and what is it about gender that matters for measurement to inform the development of effective responses? Milner points out that,
„As gender is not a fixed category or quality but a word which sets out to explain relationships between categories of men and women, we need to treat theory and research with great care, looking for what is most likely to persist and what to change. What persists historically is the complexity of male-female relationships” (Milner 2004, 95).
She identifies one of the most significant shifts since feminists first ‘drew the parameters of the discourse of domestic violence’ as the change in women’s economic power. It follows, therefore, that the relevant dimensions of this economic power must become essential variables in the evidence base. A consequence of changing economic relations is that of shifts in mothering and fathering. Thus, relevant dimensions of these roles also need to be more accurately reflected. Finally, as Hearn (1987) notes, it is no longer possible to understand patriarchy (and gender, and childhood) in an un-dialectical way. As childhood has expanded it has changed radically over the last century. One of the greatest shifts in recent childhood research has been recognition of the agency of children in late modern childhood (James and Prout 1990). Yet research into agency is usually confined to qualitative methodologies, often simplistically within the child maltreatment field, interpreted as an injunction to seek the views of children. Despite the argument being made for the quantitative reflection of childhood (Qvortrup 1990) a sense of dialectic agency has yet to be incorporated into quantitative methods. If violence and solutions to it are to be researched and responded to as gendered, ways of measuring the dialectic for both adults and children must be found.
Bolen, R. M. 2001: Child Sexual Abuse: Its Scope and Our Failure. New York: Kluwer Academic/Plenum Publishers.
Cawson, P., Wattam, C., Brooker, S., and Kelly, G. 2000: Child Maltreatment in the United Kingdom: A Study of the Prevalence of Child Abuse and Neglect. London: NSPCC.
Cicourel, A.V. 1964: Method and Measurement in Sociology. New York: The Free Press.
Driver, E. and Droisen, A. 1989: Child Sexual Abuse: Feminist Perspectives. Basingstoke: Macmillan
Everingham, C. 1994: Motherhood and Modernity: An investigation into the rational dimension of mothering. Buckingham: Open University Press.
Farmer, E. and Owen, M. 1995: Child Protection Practice: Private Risks and Public Remedies. London: HMSO.
Finkelhor, D. 1986: Sourcebook on Child Sexual Abuse. Newbury Park, CA:Sage.
Finkelhor, D., Hotaling, G., Lewis, I.S. and Smith, C. 1990: Sexual Abuse in a National Survey of Adult Men and Women: Prevalence, Characteristics and Risk Factors, in: Child Abuse and Neglect, 14, pp. 19-28.
Finkelhor, D. 1994:. The international epidemiology of child sexual abuse, in: Child Abuse and Neglect, 18, pp. 409-417.
Featherstone, B. 1997: What has gender got to do with it? Exploring physically abusive behaviour towards children, in: British Journal of Social Work, 3, pp. 419-433.
Garfinkel, H. 1996: Ethnomethodology’s Program, in: Social Psychology Quarterly, 1, pp. 5-21.
Gelles, R.J. and Straus, M.A. 1987: Is Violence Toward Children Increasing? A Comparison of 1975 and 1985 National Survey Rates, in: Journal of Interpersonal Violence, 2, pp. 212-222.
Gorey, K. M., and Leslie, D. R. 1997: The prevalence of child sexual abuse: Integrative review and adjustment for potential response and measurement biases, in: Child Abuse and Neglect, 21, pp. 391-398.
Gibbons, J., Conroy, S. and Bell, C. 1994: Operation of Child Protection Registers. London: The Stationary Office.
Graham, H. 1980: Mothers’ Accounts of Anger and Aggression Towards their Babies, in: Frude, N. (ed.): Psychological Approaches to Child Abuse. London: Batsford, pp. 39-53.
Hashima, P. and Finkelhor, D. 1999: Violent Victimization of Youth Versus Adults in the National Crime Victimization Survey, in: Journal of Interpersonal Violence, vol. 14, no. 8, pp. 799-820.
Hearn, J. 1987: The Gender of Oppression: Men, Masculinity and the Critique of Marxism. Brighton: Wheatsheaf.
Herman, J. and Hirschman, L. 1981: Father/Daughter Incest. Cambridge Mass: Harvard University Press.
Itzin, C. 2000: Home truths about child sexual abuse, influencing policy and practice. London: Routledge.
James, A. and Prout, A. 1990: Constructing and Reconstructing Childhood: Contemporary Issues in the Sociological Study of Childhood. London: Falmer Press.
Kelly, L., Regan, L. and Burton, S. 1991: An Exploratory Study of the Prevalence of Sexual Abuse in a Sample of 16-21 Year Olds. London Metropolitan University: Child and Woman Abuse Studies Unit.
Kitzinger, J. 1990: Who Are You Kidding? Children, Power and the Struggle Against Sexual Abuse, in: James, A. and Prout, A. (eds.): Constructing and Reconstructing Childhood: Contemporary Issues in the Sociological Study of Childhood. London: Falmer Press, pp.157-183.
Leudar, I., Marsland, V. and Nekvapil, J. 2004: On Membership Categorization: ‘Us’, ‘Them’and ‘Doing Violence’ in Political Discourse, in: Discourse & Society, 2-3, pp. 243-266.
Leventhal, J. M. 1998: Epidemiology of sexual abuse of children: old problems, new directions, in: Child Abuse and Neglect, 22, pp. 481-491.
Macdonald, G. 2001: Effective Interventions for Child Abuse and Neglect. Chichester: Wiley.
Macmillan, H.L., Fleming, J.E., Trocme, N., Boyle, M.H., Wong, M., Racine, Y.A., Beardslee, W.R. and Offord, D.R. 1997: Prevalence of Child Physical and Sexual Abuse in the Community: Results from the Ontario Health Supplement, in: Journal of the American Medical Association, 278, pp. 131-135.
MacLeod, M. and Saraga, E. 1988: Challenging the Orthodoxy: Towards a Feminist Theory and Practice, in: Feminist Review, 28, pp. 16-55.
May-Chahal, C., Bertotti, T., Di Blasio, P., Cerezo, M.A., Gerard, M., Grevot, A., Lamers, F., McGrath, K., Thorpe, D.H. and Thyen, U. 2004: Child Maltreatment in the Family: A European Perspective, in: European Journal of Social Work, 1, pp. 3-20.
Milner, J. 2004: From ‘disappearing’ to ‘demonizing’: the effects on mean and women of professional interventions based on challenging men who are violent, in: Critical Social Policy, 1, pp. 79-101.
Nava, M. 1992: Changing cultures: Feminism, youth and consumerism. London: Sage.
Qvortrup, J. 1990: A voice for children in statistical and social accounting: a plea for Children’s right to be heard, in: James, A. and Prout, A. (eds.): Constructing and Reconstructing Childhood: Contemporary Issues in the Sociological Study of Childhood. London: Falmer Press, pp. 78-98.
Pringle, K. 1995: Men, Masculinities and Social Welfare. London, UCL Press.
Russell, D.E.H. 1984: Sexual Exploitation: Rape, Child Sexual Abuse and Workplace Harassment. Beverly Hills, CA: Sage.
Scher, C., Forde, D.R., McQuaid, J.R. and Stein, M.B. 2004: Prevalence and demographic correlates of childhood maltreatment in an adult community sample, in: Child Abuse and Neglect, 28, pp. 167-180.
Scourfield, J. 2003: Gender and Child Protection. Basingstoke: Palgrave Macmillan.
Sedlak, A.J., Broadhurst, D.B. 1996: Third National Incidence Study of Child Abuse and Neglect. Washington DC: US Government Printing Office.
Smith, D. 1974: Theorizing as Ideology, in: Turner, R., (ed.): Ethnomethodology. Harmondsworth: Penguin, pp. 41-44.
Straus, M.A. 1979: Measuring intrafamily conflict and violence: The Conflict Tactics (CT) Scales, in: Journal of Marriage and the Family, 41, pp. 75-88.
Straus, M., Hamby, S. L., Finkelhor, D., Moore, D. W. and Runyan, D. 1998: Identification of child maltreatment with the Parent-Child Conflict Tactics Scales: development and psychometric data for a national sample of American parents, in: Child Abuse & Neglect, 4, pp. 249–270.
Thorpe, D. 1994: Evaluating Child Protection. Buckingham: Open University Press.
Thorpe, D. and Jackson, M. 1997: Child Protection Services and Parental Reactions to Children’s Behaviour, in: Child and Family Social Work, 2, pp. 81-90.
World Health Organisation 2000: World Report on Violence and Health. Geneva, World Health Organisation. Accessed 26/6/06 < http://www.who.int/violence_injury_prevention/en/ >
Winch, P. 1989: The idea of a social science and its relation to philosophy. London: Routledge and Kegan Paul.
Wise, S. 1991: Child Abuse: The NSPCC Version. Feminist Praxis.
Author´s Adress:
Prof Corinne May-Chahal
Department of Applied Social Science
County South
Lancaster University
UK-LA1 4YL
United Kingdom
Tel: ++44 1524 594104
Fax: ++44 1524 592475
Email: c.may-chahal@lancaster.ac.uk
urn:nbn:de:0009-11-5399