Who Owns Child Abuse?

Abstract

Expectations of contemporary child protection apparatuses are strongly influenced by beliefs inherited from the nineteenth century child rescue movement. In particular, the belief that child abuse determination is obvious. However, this assumption fails to make a distinction between nineteenth century’s emphasis on impoverished environments and the twentieth century introduction of the pathological child abuser. Moreover, the proliferation of kinds of child abuse, and the need to distinguish child abusers from non-abusers, means knowledge is now spread across an array of disciplines and professions, which necessarily destabilizes the definition of child abuse. The increasing exposure of alternate care systems as potentially abusive has similarly destabilized the old common sense solution to neglected children—namely removal. Finally, as uncertainty increases, and definitions become more divergent, the question of what child abuse is, and what should be done about it, becomes increasingly politicized.

1. Introduction

Perhaps the most curious thing about child abuse is the degree to which its practical definition is usually treated as obvious even though its precise composition and boundaries are subject to chronic and intense disagreement. The problem is not so simple as whether one person’s discipline is another person’s child abuse, nor whether children’s exposure to adult sexuality is child abuse or bad manners, nor even what the appropriate balance between “coddling” and “neglect” might be. To frame the problem in these binaries obscures the degree to which the understanding of child abuse is fundamentally epistemological. In other words, this view inevitably presumes child abuse has a reality independent of the varied discursive practices that constitute it. The presumption implies a consensus of what child abuse really is, what its effects are known to be, and what practical solutions should be applied even where there are deeply divided opinions.

Suggesting that knowledge of what is and what is not child abuse is problematic seems out-of-step with modern professional and lay knowledge about child abuse. Public perception, inevitably fueled by inquiries into child deaths and stoked by lurid and occasionally hysterical press coverage, tends to view the presence or absence of child abuse as obvious to anyone with eyes to see and ears to hear. Meanwhile, the copious studies linking certain parental behaviors and practices with poor intellectual and emotional outcomes for their children add the imprimatur of science to this impression. Law, too, especially as expressed in the reports of legally trained inquirers, tends to view individual cases of child abuse—and especially the child deaths that are inquiries’ frequent objects—as uniformly avoidable with the proper application of widely available professional knowledge.

The view that there is a single and universal phenomenon designated by the term ‘child abuse’ is fundamentally a modernist conception. It is a means of placing child abuse outside the confusions and complexities of history by positing a true description of a fixed reality 1. To paraphrase the sociologist Norbert Elias, this manner of describing child abuse is a “retreat into the present” whereby current practices and discourses are projected as epistemological absolutes applicable at any time and in any place [4]. From this perspective, as civilization develops child abuse is increasingly discovered or revealed as our knowledge expands. However, historians of childhood—and others utilizing historical sociology—have repeatedly drawn attention to the developing status of childhood and the place of children within society more generally [4,5,6]. This development is often seen as a form of teleological progress, but is more accurately described as an aspect of a culturally specific moment; a pattern of conduct necessitated within historically specific social and cultural configurations. For example, a concept such as “fostering” has dramatically different meanings for the medieval period in contrast to our contemporary age. More significantly, behind this difference lay radically different notions of what is best for children, and what kinds of attachments children require.

This paper will make several relatively simple points: First, the “obviousness” of child abuse is not a reflection of the current state of knowledge, but rather a legacy inherited from the 19th century child rescue movement and the over-confidence of the “scientific charity” ideology that absorbed it. Second, new knowledge in the field of child abuse is more or less incoherent and therefore it is to be expected that child protection action will reflect that incoherency. Finally, child protective action is fundamentally political because of this broad range of definitions and specialist epistemologies. These observations are significant for anyone engaged in the child protection project, but most especially for the review mechanisms that have proliferated over the past several decades. If child abuse is a fixed property engaged in by a specific and pathological kind of human, then it is reasonable to believe child abuse can be eliminated by deploying a fixed set of techniques. However, if child abuse is not a fixed property but variable, not only through time but across knowledge domains, then it follows that proposed solutions are likely to vary according to the epistemological rules of each knowledge domain. In other words, there is no reason to suppose a consensus of what constitutes child abuse is possible, no matter how desirable it might be. In other words, calling for “oined-up” services, or single “child-centered” practices will always be thwarted by the incommensurability of varied knowledge constructions and their resultant practices.

2. Child Saving

The term “child saving” is generally associated with philanthropic movements of the nineteenth century whose principle interest was saving children from the depredations and vices of the newly industrialized cities of Europe and North America. In general, the objective of child savers was removal; ideally children would be removed from squalid urban conditions to healthful and wholesome rural surroundings where the endemic physical and moral diseases of city life would be cleansed in fresh country air and the salutary company of rural yeomanry (see for example: [7]).

The emphasis on removal is significant because it marks a break with prior childcare practices. Residential placement of impoverished children in homes specifically designed for children was unusual in Western Europe until the eighteenth and nineteenth centuries. By contrast, in the Byzantine Empire state-sponsored and private orphanages were a common occurrence beginning as early as the fifth century A.D. [8]. The Ospedeli degli Innocenti founded at Florence in the early fifteenth century was the first—or at least one of the first—purpose built orphanage in Western Europe [9]. London had an orphanage by the mid sixteenth century (notable because it was not directly tied to a specific parish) [10], and Paris opened an orphanage in the early seventeenth century [11]. The later appearance of orphanages in Western Europe may have been a consequence of monasteries’ ability to receive (male) children as oblates, or perhaps because of the widespread and traditional use of fostering and abandonment. In any case, it seems likely that most foundling children died, were more or less indentured, or found themselves in the usual resorts of the destitute; almshouses, hospitals, and the like [2].

In contrast, by the beginning in the mid-eighteenth century there is a discernible shift in adult attitudes toward dependent children. First, child welfare institutions were increasingly child specific. Second, early orphanages such as Captain Coram’s London Foundling Hospital became larger than their predecessors because, in principle, they sought to service all children in need [12]. Even so, by the mid nineteenth century orphanage populations were so large children were exported from Britain to the colonies by the tens of thousands. Second, orphanages were increasingly linked to the health of the nation. Destitute and neglected children were not simply the devil’s workshop, but were indicative of the decline of the race, and in the case of Britain, a threat to Empire. Thus, for example, Coram’s boys were set to work making rope and sails in anticipation of their future employment in Britain’s vast imperial merchant marine.

At least for England, the need for universal orphanages was linked to the traditional Poor Laws, which were organized on the localism of the parish. Consequently, as parishes merged into urban areas, and the population became more mobile, ensuring the support of the poor—including children—by their home parish became increasingly difficult to enforce [13]. Certainly, by the mid-nineteenth century the combination of mass migration to the cities and uncertain employment had created an army of street urchins that overwhelmed traditional forms of charity. The result was the child-savers and other reformers of the urban poor.

These nineteenth century child savers were both similar to and different from those who had gone before. They were similar insofar as both iterations of philanthropy knew full well that “orphan” meant poor; not “without parent”. Both understood that infants who entered orphanages usually died, and both reckoned a child’s stay in an orphanage was limited to middle childhood because children were routinely apprenticed or placed in service around the age of ten or twelve years. As well, for the most part those who financially and administratively supported orphanages were motivated by religious and social status factors. The nineteenth century child savers, however, were markedly more zealous in their activity than their forebears. The Victorian philanthropist was not content to wait until approached for favor by destitute parents, but actively sought out the poor in their hovels. Further, these reformers saw themselves as buttressing the very foundations of society. They were not simply alleviating poverty through charity, but reforming and reorganizing society upon scientific principles. The anticipated result in the British Empire was the furthering of Imperial glory and the advancement of civilization, while in the United States child welfare was believed necessary for the survival of republican democracy.

Many explanations have been advanced as to why the nineteenth century evinced such deep concern for children. These include: a new muscular Christianity; the spread and dominance of a bourgeois family ethic; a sacralisation of childhood; the universalization of schools; the demographic transition; the rise of romanticism; and more. No doubt all of these are true to a degree, but from the point of view of historical sociology it seems this refiguring of childhood and adult responsibilities toward children is reflective of a vector of development stretching across centuries. In the classic work of Ariés [5] and Elias [14] our attention is drawn not only to differences in the structures and functions of historical social formations, but also the corresponding shifts in sentiment accompanying these changes. Of course, one should not overly romanticize general Victorian sentiment. No doubt those who went on “slumming” tours of London’s East End were as likely drawn by the same impulse one might have for going to the zoo [15], but there is no denying many others were deeply moved. The novels of Dickens, the reporting of Jacob Riis, and the publicity photographs of Barnardo, were not mere cries in the wilderness. People were emotionally moved, and some were moved enough to become tirelessly active.

For these child savers, the problem was obvious. Neglected children either lived ragged and alone on the street, or inhabited dark and filthy tenements and hovels where violence and vice were rampant. Dr. Barnardo and his lamp, the New York Children’s Aid Society agent, the lady “friendly visitor” from the Settlement House sought out children living in these conditions and cajoled, bribed, or outright abducted them to places of safety. With the exception of sexually abused girls and criminal boys the particular experiences of particular children were not of much interest to their saviors [7,16]. Child savers did not save children from categorized forms of abuse, but from a life of urban degradation and squalor. Nor did the child savers vary much in their remedial recommendation; children were to be removed to the country. By the turn of the twentieth century, even those few voices speaking in favor of institutional care through cottage style orphanages agreed such orphanages should be located in the country (See for example [17]).

Even as the twentieth century created the social worker along with her social casework skills, the self-evident-ness of children’s plight persisted. Some children continued to be removed due to the effects of poverty, but poverty faded from the causation horizon. Rather, the social casework formula propagandized by Charlotte Whitten in Canada [18] and Mary Richmond [19] in the United States spoke to the need to moralize the poor rather than engaging in the indiscriminate charity they believed encouraged dependency and sloth. This is not to denigrate the enormous and widespread scientific project that accompanied the Child Study movement, the Juvenile Courts, the influence of psychoanalysis, and the promotion and dissemination of research-based knowledge by such organizations as the Children’s Bureau in the United States. Children, it seemed, were suddenly at the center of a scientific industry. Yet for all that, social casework and its allied professions remained remarkably confident in its own moralizing activities and its ability to simply “know” when children were in need of removal. The new knowledge, no matter how carefully researched and publicized, was not so much incorporated as grafted onto an existing moral assuredness predicated upon the inherent superiority of the bourgeois family.

Such was the state of child protection for the first two-thirds of the twentieth century. Through radical improvements in infant and maternal care, infant mortality dropped sharply. New nutrition strategies such as the Canada Food Guide emphasized the importance of healthy practices underwritten by healthy baby contests at local fairs, and universalized public health and sanitation practices meant children were healthier throughout their lives [20,21]. Eventually, a point was reached where children’s health was simply assumed and both parents and experts could turn their attention to whether children were happy and achieving their full potential. There were, however, some nagging worries. In the late 1940s radiographers began to notice that children’s bones were being broken and healed in secret. While psychology and psychiatry worried about toilet training and attachment interruption, a small group of doctors began to wonder if even middle class parents were deliberately physically harming their children [22]. Finally, in 1962 Kempe and his colleagues published The Battered-Child Syndrome [23] and the world of childhood was revolutionized.

3. Child Abuse as Science

If the early child savers needed little besides the evidence of their own eyes to identify neglected children, their judgments with respect to the children’s parents and other adults were equally morally straightforward. Children’s ill treatment was caused by poverty, ignorance, and in the often-brazen anti-Catholic sentiment of Protestant child savers, by the ignorance of the priesthood [16]. In the eyes of some of the more broad-minded child savers poverty and ignorance may have excused neglect and vice, but generally speaking assessments of parents were framed within moral parameters. With the exception of Tardieu’s mid nineteenth century work at Paris, there is no figure of child abuser as a pathological threat [24]. Whether in the diaries of CAS workers in New York, the textbooks written by Mary Richmond, or the deliberations of the White House conference on dependent children [15,19,25], parents are never described as pathologically disordered. Thus, when Hacking [26] describes the displacement of cruelty, neglect, and incest with the general term “child abuse” he is perhaps only half right. It wasn’t that “child abuse” was a twentieth century discovery, but the child abuser in all his or her iterations, became a new explanation for the maltreatment of children.

Thus, when Kempe and his colleagues described the “battered-child syndrome” in 1962 (almost twenty years after radiologists had first described evidence of it), it was not the existence of the “syndrome” that created the ensuing moral panic, but the realization that it was parents from all social classes who were inflicting the battering. Kempe specifically described these parents as in need of medical supervision and psychiatric counseling. Those who maltreated children were no longer morally depraved or hopelessly impoverished; they were mentally ill and extremely dangerous.

As Nelson [27] has famously documented, the impact of Kempe’s paper was both wide and deep. That is, Kempe didn’t just put child abuse on the map; he arguably created an entire industry that rapidly spread from Denver to encompass the globe. By the end of the 1960s, the figure of the child abuser was as feared and endemic as “Reds under Beds” in the 1950s. And, just like the Reds, child abusers came in increasingly diverse and secretive forms. In the 1980s when sexual abuse was recognized not as “incest”, “white slavery”, or peculiar rural marriage practices, but as a specific form of sexual deviancy practiced by child abusers who were usually known to their victims, child protection workers were confronted with a whole new set of challenges. Largely ignorant of children’s sexuality, and ill-equipped to manage the coordination of family and criminal matters, the child protection system was overwhelmed by allegations of present and past sexual abuse by parents, friends, institutions, and, finally, sadistic cults practicing “ritual abuse” [28].

With all of this attention to saving children from child abusers, the scientific world was equally overwhelmed. For example, the hard science of memory found itself questioned as therapists made claims about the mind’s ability to recover unadulterated repressed memories [3]. But while the Recovered Memory movement fought desperate battles with their “False Memory Syndrome” opponents, other less emotive questions arose. Suddenly everything from baby’s sleeping habits and parental smoking [29], the mysterious Munchausen’s by Proxy Syndrome [30], and the problem of SIDS were drawn into the ambit of child abuse [31,32]. The term child abuse came to absorb virtually every possible way a child might be harmed, and in keeping with the characteristic drift of neoliberal regimes toward responsibilization, parents were increasingly scrutinized for error [33].

Thus, child abuse research continually produces correlative studies purporting to show connections between particular parental behaviors and poor later outcomes for children. This research, combined with a seemingly endless parade of scandals in child protection systems across the globe led to the incorporation of “risk” into child protection practice (See for example [34]). Specifically, the 1990s saw the widespread adoption of risk assessment tools into child protection practices. As I have discussed elsewhere, the adoption of risk assessments was not so much a utilization of scientific insight as a political strategy designed to blunt attacks on child protection systems and their functionaries [35]. The central characteristics of child protection risk assessments were their assumption that child abuse occurred within the family, and their emphasis on the characteristics of the potential adult abuser 2. That said, one could also argue that the shift to risk thinking in child protection was away from specific persons in a specific situation and towards the search for risks in and of themselves. Hence, child protection began to slide into a world of almost pure abstraction wherein risks became more real than actual parents and children. The numbers generated by risk assessments purported to provide a scientific basis for decision-making and, moreover, the claim was always that these risk values were “research-based”.

Precisely how “scientific” risk assessments truly are is open to debate. However, it is very clear that once child protection systems begin to think in terms of risks the universe of potential sources of child abuse becomes potentially endless—especially where risk of child abuse, and risk of harm, comes to mean more or less the same thing. In the kind of responsibilizing environment produced by neoliberal regimes there are no accidents; everyone is potentially at fault for their risk miscalculations [36]. Under such circumstances, determining child protection action is no longer a matter of common sense. The combination of ever multiplying sources of risk, and the ever more complex ways in which these risks are calculated and evaluated, suggest limitless possibilities for forms of child abuse.

4. Alternative Care as Iatrogenic Abuse

In the back alleys and tenements of the great Victorian cities the condition of children was so obviously abysmal that the only practical permanent solution appeared to be removal to places of safety. There were, of course, other forms of assistance available in the form of ragged schools, newsboy’s hostels, and so forth, but the enormous growth of orphanages combined with large-scale child migration testified to the unflagging belief that removal, and ultimately country air, was the universal solution to the problem of urban child neglect [7,16].

Children removed to orphanages had notably mixed experiences. Unsurprisingly, by the beginning of the twentieth century a general consensus emerged claiming institutional care—especially congregate care—was generally harmful. Institutional care was said to produce listless, unimaginative, and docile children who would grow to be adults incapable of earning their own living or caring for themselves. Orphanages reacted by adopting cottage style living quarters that attempted to approximate normal family life, but even this style was considered a poor second choice to loving family care. There were dissenting voices—particularly from Catholic charities—but they were generally overruled.

In these early iterations of mass fostering it is remarkable how little effort was expended on finding out what happened to removed children. The New York Children’s Aid Society did some follow up work in response to complaints from the Western States that New York’s incorrigible children became the Midwest’s juvenile and adult prison population [7,25], and inspections were carried out in Canada in response to complaints that orphans were being used as virtual slaves on Canadian farms, but by and large the efficacy of sending children West, and the superior care provided by family care, was taken for granted. Moreover, the movement of children was largely predicated upon a “clean break” philosophy. As ever, Victorian orphans were not without parents but permanent separation of children and parents was generally encouraged [37]. Despite this, children voluntarily placed by parents were never entirely out of the minds of their parents as correspondence between orphanages and parents attests [38]. For those children removed across continents and between continents, of course, such parental contact was effectively terminated.

In short, from the very beginning of the modernist child protection project the question of whether children were better off left in inadequate living circumstances, or starting life anew with foster or adoptive families was never entirely resolved. Model institutions such as the New York Orphanage on the Hudson suspected children in foster care were little more than cheap labor [17], while parents never ceased to see orphanages as stopgap measures who’s main purpose was to help families through rough economic times. Jane Addams, for example, used the 1909 White House Conference as a platform to demand worker’s compensation schemes to prevent children becoming dependent on charity or the state due to their fathers’ industrial injuries [25].

The problem of alternative care for children tended to fade during the early decades of the twentieth century. Across the English-speaking world, orphanages closed or were repurposed. A growing welfare state alleviated much of the brutality of nineteenth century capitalism and consequently the street urchin disappeared. Those orphanages that survived were increasingly repurposed to service children suffering some sort of pathological condition. Orphanages ceased to be normative and, instead, became therapeutic [39]. Alternatively, orphanages were absorbed into a network of municipal social services as with Vancouver’s first orphanage, which became—and remains to this day—a “Neighborhood House” [18]. The fostering and adoption concept remained, and at least in the U.S. became a major mechanism for families to survive the Great Depression, but orphanages were more or less terminated as a child welfare technology. The trend away from alternative care became a positive injunction with the post war acceptance of John Bowlby and Anna Freud’s widely publicized and influential work on childhood attachment. According to this new truth regime, any separation of children from their attachment figures—normally understood as mothers—was itself harmful and should be avoided except for exceptional instances of neglect or abandonment [40].

When two decades later the Battered-Child Syndrome created the figure of the pathological child abuser the child protection system was placed in a quandary. Should children remain in the care of psychiatrically disturbed parents, or should they be removed? In Canada, the problem was particularly stark because the closure of Residential Schools for aboriginal children created a child care crisis for First Nations’ settlements. The result was what became known as “the sixties scoop” when some First Nations lost virtually all their children to child protection authorities [41]. Without the now discredited mass institutional care provided by orphanages, child protection authorities sought to expand the fostering and adoptive systems to provide for the influx of children. However, by the end of the nineteen seventies this approach, too, had lost its appeal. It was increasingly obvious that children in foster care led a transient life drifting from one home to another, and that at least some foster parents were as abusive [42,43], if not more abusive, than the parents the children were supposedly being saved from.

In Canada, events such as the suicide of Richard Cardinal and other exposés in the press, called the entire concept of alternate care into disrepute [44]. As the growing body of inquiries and Court Judgements begun in the early nineteen seventies and continuing down to the present day demonstrate, not only were children being abused in their own families but they were also abused in the alternate care facilities and foster homes meant to protect them 3. Despite the obvious fact that no amount of exhortation to better judgment, training, or education on the part of social workers was likely to increase the potential pool of foster parents living in an increasingly urbanized environment of two-earner families, variations of long term fostering or adoption became the clarion call of 1980s child welfare.

It is clear child protection apparatuses developed over the past two centuries have generated their own moral hazards. Put at its bluntest, the practice of child protection has become a source of child abuse because child protection activities have iatrogenic properties. Faced with this dilemma, child protection authorities are generally exhorted to engage in “preventative practice”, which is usually understood as forms of increased surveillance and intervention

5. Child Abuse: Event or Pattern?

The definition of child abuse within child protective practices tends to view child abuse as patterned and repetitive behavior on the part of a pathological child abuser. While it is true that child removals may occur as a consequence of acute and specific events, it is rare for those events to occur within an informational vacuum. By now given the constant call for professionals to be vigilant, the effect of mandatory reporting laws, and the general acceptance of high levels of surveillance and their accompanying risk calculi, it is unusual for an abused child to be completely unknown to authorities [45]. Indeed, a frequent complaint of inquiries into child protection scandals is that all the necessary information was known; it wasn’t acted upon. It seems that in the practical world of child protection the question of whether a child is in need of protection is not a simple one-to-one correlation with a specific action classed as abuse, but rather a matter of thresholds of tolerance. For example, a single drunken party won’t result in your child being removed, but a pattern of drunkenness eventually will. A single slap across the buttocks of a two-year old is acceptable (in Canadian law), but a pattern of spankings is not. By contrast, sexual abuse is generally considered a “zero tolerance” form of abuse, but even here the boundaries between abusive behavior and inappropriate affection or poor judgment, are not always as obvious as one would hope. When, for example, the La leche league counsels mothers to breastfeed to as old as seven years, and acknowledges part of the attraction of this practice is its erotic overtones, is this child abuse [46]? Finally, who is to say when a familial sex abuser is “cured” or, perhaps more accurately, under sufficient self-control, they can resume a normal parental role?

The role of patterns plays out in much psychological research into the effects of child abuse. This body of research almost never refers to a single traumatic event but rather to patterns of harmful conduct. Insofar as child abuse may be defined as anything that causes harm to a child, these described patterns become the psychologists’ definition of child abuse. More accurately, each of the many harmful patterns this literature describes is child abuse because they cause harm and therefore children ought to be protected from them. The immediate problem for child protection apparatuses is that no harmful pattern can be fully described by a psychological study. Harmful patterns must take place within a larger and often quite specific context, which at the very least will modulate the harmful pattern’s effects and may even neutralize them. This problem is the fundamental sociological problem of understanding the relation of individuals to society at large, and then attempting to make universal and predictive statements about that relationship within a specific context.

Moreover, it is possible that while abusive behavior may be patterned, the actual harm done is caused by a specific incident. This is most obvious in the case of pediatric homicide where a child’s death occurs as a result of a specific act of violence, and yet the larger context may be without positive evidence of previous child abuse. Such was the problem confronting the Canadian Goudge Inquiry in its consideration of wrongful convictions for child homicide [47]. Like the U.K. experience with the famous trio of wrongful convictions for child homicide, the Goudge Inquiry was prompted by the problem of determining cause of death in cases of apparent Shaken Baby Syndrome, short-fall injuries or Sudden Infant Death Syndrome [47]. In these instances, “cause of death” had significance in both the child protection and the criminal prosecution context. Further, both these contexts were reliant on knowledge and expert judgment generated outside their respective fields—namely medicine, and specifically pediatric forensic pathology. When the pathology turned out to be unreliable, both the child protection and criminal prosecution apparatuses were left without evidential foundation. For the prosecution apparatuses the result was wrongful convictions and incarcerations. For child protection, it meant child “wrongfully” removed and adopted because of those wrongful convictions. In the wake of these errors (and Lord Laming has similar advice for police officers with respect to doctors), it seems each knowledge regime has to operate according to its own standards and rules of establishing truth. Thus, while the child protection system is urged to find a singly language, its components are simultaneously advised not to trust truth generated outside their specialty. Significantly, the Goudge Inquiry specifically instructed coroners and pathologists to ignore patterned evidence suggesting child abuse and advised them to focus solely on the immediate circumstances when determining cause of death [48].

The instances highlighted by Goudge and the UK cases are not only about the specifics of diagnosing cause of death. Rather, they necessarily call into question approaches to child protection based upon contextual inference—that is to say, risk assessment. The Goudge Inquiry Report, and the mea culpa evidence given by leading international pathologists there, specifically criticizes cause of death diagnoses that apply general knowledge about child abuse causation to specific events, or apply prior prejudicial knowledge of the actors. It also implicitly denies diagnoses of exclusion; i.e., if there is no other obvious cause child abuse is assumed. Instead, the consensus now insists upon positive knowledge meeting the “beyond reasonable doubt” standard for criminal evidence.

The problems associated with diagnosing cause of death in instances of short-fall injuries, Sudden Infant Death Syndrome, and Shaken Baby Syndrome, are dramatic both in their subject (child mortality) but also in their ability to shift the figure of the child abuser into the figure of the wrongfully convicted. While these instances focus our attention, they are but examples of a much larger problem of certainty in the child protection world. Where, for example, do we stand on Munchausen’s by Proxy Syndrome? At one time, the medical discovery of this life-threatening psychiatric disease seemed a major victory for child protection, but in the wake of its main discoverers being disgraced—even if partially rehabilitated—the diagnosis is left in a curious kind of limbo [49]. The point is not whether these syndromes do or do not exist, but their sudden uncertainty clears the way for other claims under uncertain conditions. Do vaccinations cause autism? Does inscribing a swastika in ink on a child’s arm cause the child to become a fascist? (And is that abuse?) Do sports involving head collisions necessarily cause brain disease? Is Uncle Tom’s Cabin too racist for children to read?

No matter how tendentious these sources of child abuse may appear, they do at least make causal claims. That is, they claim exposure to a given circumstance or context is harmful. There is, however, still the difficulty of what is meant by potential. During the latter part of the twentieth century child abuse began to be couched in terms of an ideal childhood in which parents and/or the state were said to have responsibility for ensuring children reach their full potential. On the one hand, the goal was laudable insofar as it placed children’s abilities and ambitions center stage. However, what was to be done about children who lived in situations where their significant adults actively discouraged children achieving their full potential, or didn’t much care about their children’s potential but were nonetheless adequate parents, or fundamentally disagreed with the child’s belief in their own potential. Does, for example, a child’s potential to be a great ice hockey player mean his or her parents should allow their child to risk his or her brain’s future health?

This multiplication of potential forms and sources of harm to children has combined with exhortations—often contained within government inquiries—reminding professionals, quasi-professionals and the general public that child abuse is everyone’s responsibility means the circle of experts has also expanded exponentially. As I have noted elsewhere [35], it is striking how inquiries into child fatalities in the 1970s saw social responsibility as resting exclusively with social work child protection apparatuses. It is equally striking that this is no longer the case. Inquiry reports such as the Gove Inquiry or the Laming Inquiry describe themselves as promoting a ‘child centered’ philosophy, by which they mean the entire structure of government and civil society should pivot around the discovery and amelioration of child abuse 4.

In short, unlike the nineteenth century child savers who found children in need of protection in precisely the places and forms they expected—and indeed could not ignore—now the potential for harm can come from almost anywhere. Indeed, concern about the proliferation of sources of risk to children has found its way into both academic and popular literature. Complaints about coddled young adults who experience “failure to launch” are heard next to the hard sell of surveillance technologies for daycare and the home. The whole is exacerbated by the increasing tendency to responsibilize individual parents for their children’s success or failure in an increasingly competitive world [50].

6. Conclusions: The Vanishing Point of Child Protection?

The most significant difference between nineteenth and twenty-first century child protection is that the sources and forms of children’s vulnerability are no longer obvious. In addition, it is not obvious what the solution to cases of child abuse ought to be. The confidence of nineteenth century child savers has been badly shaken over the past several decades. In the mid-nineteenth century the victims of child neglect and the forms that neglect took were glaringly obvious to anyone who visited London’s Eastside, New York’s Five Corners, or their counterparts in any major city. Further, removal to orphanages and eventually the country was the universally sanctioned solution.

There were, of course, disputes amongst child rescuers about the particular children to be assisted, and the particular forms that assistance ought to take. Orphanages and charitable support typically distinguished between full orphans and half-orphans, religious affiliation, and the depth of moral depravity children were likely to express. However, these distinctions were not made on the basis of types or severity of child abuse—that was generally seen as generic—but rather by the moral standing of parents or the children themselves.

Disagreements over congregant care as against a cottage model, continuing contact between parents and “clean break” philosophy, and the nagging suspicion that children placed on farms might be little more than exploited labor flared up from time to time, but the basic faith in the technology of removal seldom wavered. The late twentieth century recognition that alternate care systems, either institutional or fostering based were also sources of harm to children (not to mention sources of liability) has brought removal as a solution under suspicion.

The crucial contribution of Kempe and his colleagues was to introduce the figure of the child abuser. This new figure was deliberately separated from a specific environment and therefore child protection workers could no longer assume children in need of help were confined to specific locations associated with poverty. Now, abusers were potentially everywhere including within alternate care systems. The result was a turn to risk technologies in the hope that scientific knowledge, appropriately ordered, would reveal child abusers through a global analysis of conduct and desire. Necessarily, no one discipline or profession could possibly know or predict all potential sources of harm and therefore child protection was increasingly cast in terms of global social effort. Knowledge of what child abuse is, and what should be done about it, was supposed to cross these “silos” of intellectual and professional effort.

The broadening of knowledge apparatuses responsible for detecting child abuse, and the increasing sophistication of research into harms perpetrated by child abusers and experienced by children mean the general consensus about what constitutes child abuse and what ought to be done about it is increasingly spread across knowledge domains. Who would have dreamed in the 1970s that a crucial specialty in child abuse research would be bio-mechanics? For that matter, it wasn’t that long ago doctors refused to engage with child abuse at all, and if comments following the travails of Sir Roy Meadow are to be believed doctors may be withdrawing from child abuse research and treatment in alarming numbers.

While it may seem self-evident that drawing the entire array of social actors into the world of child protection is a good thing, such an ambition neglects the difficulties associated with multiple interpretations of what constitutes child abuse, and the multiple preferred methods for alleviating its consequences. Of course, complexity and multi-disciplinary interpretations are not, in themselves, problematic. Many contemporary scientific projects require intense collaboration across disciplines. Space flight, for example, would be unthinkable without such cooperation. However, a distinction between these projects and the demands placed on child protection authorities is precisely that nobody working on a complex aeronautics project would expect every discipline to know the same core knowledge, and have equally valid approaches to all possible issues arising. Rather, such multi-disciplinary projects have their various specialized knowledge bases coordinated by a project manager. No one would expect all of NASA’s scientists to take equal responsibilities for all the functions of space flight. Why then is it often assumed that everyone involved with children is equally skilled at detecting and understanding child abuse?

The central problem facing child protection in the twenty-first century is not necessarily too little knowledge—and certainly not too little information—but how to manage the avalanche of empirical claims and theories available. More to the point, how are those charged with authority to undertake the most draconian child protection actions (child removals) to evaluate the myriad sources of information inundating their practice? Moreover, as the alternative care system is increasingly understood as inherently harmful, what alternative technologies of residential care can child protection call upon? It seems as if the sources and types of child abuse increase in inverse proportion to the availability of safe alternatives.

Insofar as the child protection project continues, it does so because most child abuse is relatively easy to determine; not because of great scientific advances, but because child protection action remains rooted in the obvious just as it always has. In the final analysis, this is simply to say that most child protective action remains normative; children are removed from parents not because there is a demonstrable scientifically valid proof of harm, but because certain adult conduct is sufficiently outside the norm to draw the full force of moral policing. Some things obviously are abuse, some things obviously are not abuse, and many things are neither or both. As ever, it is the latter cases that preoccupy child protection practitioners because it is there the most dangerous hazards are to be found. Here danger is understood as the combination of potentially catastrophic consequences with high levels of uncertainty.

It is the problem of coping with uncertainty that defines child protection action as fundamentally political. When child abuse was obvious, child protection practitioners proceeded with the confidence that they represented the general “common sense” of larger public tolerances and were, therefore, outside politics. However, as the commonality of sense becomes increasingly divergent, and as expert knowledge divides along disciplinary and administrative lines, the practical definition of child abuse becomes more uncertain. Of course, the problem of complexity and diversity is not unique to child protection. The emergence of “risk” in the analytic work of Ulrich Beck, Richard Ericson, and others points to difficulties created by multiple experts analyzing uncertain conditions from varied epistemological locations. Divergence of opinion is to be expected amongst experts, but this divergence of opinion is not always tolerable. In particular, controversy over uncertain definitions fuels political action. Thus, insofar as child protection practice takes place in an uncertain domain influenced by multiple knowledge regimes deployed by a wide variety of experts, it must be understood as necessarily political. As Ian Hacking [26] pointed out, controversies over child abuse are really controversies over the meaning and practice of “normal”: and nothing is as superficially obvious or as political as defining a norm.

Nico Stehr [51] has remarked that social science scholars often erroneously use “complexity” and “unknowable” as synonyms. The present problem of knowledge regimes involved with child abuse is not one of complexity, but rather one of different epistemological domains grappling with evaluating human conduct. From this perspective, consensus on what constitutes child abuse, and what to do about it, can never be anything more than aspirational. Child abuse is not a coherent set of fixed or even relatively stable characteristics. Rather, it is an unstable and shifting discursive domain in which “harm” and “danger” to children are constantly negotiated and renegotiated within and across knowledge domains. This means there can be no “blueprint” or “master narrative” of child protection. And, this is not a bad thing. If child abuse had remained obvious, then there would be no need to change perceptions of what is, and is not, acceptable behavior toward children. Change, however, is always fraught—we do well to remember that none of us can predict what will count as the child abuse of the future.