Speaking bodies – silenced voices: Child protection and the knowledge culture of ‘evidencing’

Using the metaphors body and voice and drawing on critical contributions on biopolitics, this article interrogates children’s participation rights in a knowledge culture of ‘evidencing’. With child welfare and protection practice as an empirical example, I analyse written assessment reports from a Swedish child welfare agency, all exemplifying how social workers evidence needs for protection and reasons for removing children from the home. I discuss how ‘evidencing’ equals a knowledge culture of seeing-believing and predicting-believing and the search for visibly damaged bodies and underdeveloped minds. I furthermore problematise how such conceptualisation of evidencing foregrounds children’s ‘speaking’ bodies while silencing their voices. By showing these manifestations of evidencing, this critical contribution discusses some wider epistemic concerns for fields influenced by the knowledge cultures of ‘the evidence-based’.


Introduction
In the field of child welfare and protection, evidence may refer to the citation of testimonies, police reports and other documentations. Perceived difficulty in providing proof, or what I here call evidencing children's exposure to violence and other problems, has been presented as a possible reason for the cautious approach of Swedish child welfare and protection services, and for authorities not intervening (Linell, 2017;Ponnert, 2007). This applies to the context addressed in this study, namely the Care of Young Persons Act (sv. Lag med särskilda bestämmelser om vård av unga 1990: 52) and younger children who are in need of protection due to issues in their family and environment. Child welfare scholarship describes interventions as a 'reactive enterprise' and as taking place in cases where child abuse is severe, has been long ongoing, and the children are already visibly harmed (Östberg, 2010: 209; see also Leviner, 2014;Linell, 2017).
By going beyond some previous inquiries that problematise evidencing in terms of easier/harder to prove, in this article I adopt a broader approach. I instead argue that the knowledge culture (Knorr Cetina, 2007) and its conceptualisation of proof determines what may be seen as harder or easier to prove or to disclose in the first place. The notion of evidencing not only captures what accounts and whose evidence counts as trustworthy, but is also a practice, or a performative act (Butler, 1990) grounded in a distinct epistemology. This, I argue, not only includes the legal discourse on evidence but also a development that has been prominent in Sweden in the field of social work since the 1990s. With this development, the term evidence also figures in what has been described as a social movement, a paradigm, and an idea -evidence-based practice, hereafter EBP -that associates evidence with 'the best available scientific knowledge' (Commission of inquiry 2018/18:10; see also Bergmark and Lundström, 2011;Holmes et al., 2006;NBHW [National Board of Health and Welfare], 2012;Svanevie, 2011).
While having many meanings in diverse professional fields such as psychiatry, psychotherapy, nursing, education and social work etc., EBP nevertheless refers back to evidence-based medicine and Professor Archie Cochrane, who advocated systematic reviews and randomised controlled trials (RCT) as methods for showing reliable evidence for best medical care (Holmes et al., 2006;Svanevie, 2011). EBP may be discussed in relation to interventions, methods and treatments but in the context in focus of this study, EBP associates with knowledge more broadly and what is widely called knowledge-based social work (NBHW, 2012(NBHW, , 2018. This applies also to its meaning in the framework for assessment, BBIC (sv. abbreviation of 'Children's Needs in Focus'). BBIC is, like EBP, a 'travelling' idea. BBIC is originally based on an assessment framework in the United Kingdom and is now used by the Social Services in almost all municipalities in Sweden. Several other child welfare and protection systems use the assessment framework in some modified form (NBHW, 2018). The framework for assessment is based on universalist theoretical legacies and discourses that are consistent with developmental psychology (see Knezevic, 2017, for an overview;cf. Woodhead, 1999). In addition, BBIC draws on research that by no means can be restricted to national borders (NBHW, 2018).
In BBIC, EBP primarily has to do with professionals' inclusion of multiple 'sources' of knowledge. A division is made between three asymmetrical 'sources': scientific knowledge, professional expertise, and service user perspective. The perspective of the service user is subjective, in contrast to the 'objective perspective' of science and professionals (NBHW, 2018: 14). Besides the child service user, also parent/s and siblings may constitute the fragmented service user perspective. Thus, when the Swedish child welfare policy advocates children's right to participate in assessments and have a voice and influence in matters concerning them, this advocacy is articulated within the context of a subjective-objective dualism and epistemic asymmetry, that is, asymmetry in knowledge status between (child) service users and professionals/science.
Children's participation rights have been in focus in research on child welfare policy, practice and legislation (Heimer and Palme, 2016;Iversen, 2013;Knezevic, 2017;Leviner, 2014;McLeod, 2006;Pölkki et al., 2012;Sundhall, 2008). Yet, the status of child service users' own evidencing and participation remain unexamined when social workers make assessments and gather 'evidence' based on the three knowledge sources and prior to recommending out-of-home placements (see Bergmark and Lundström, 2011;Iversen and Heggen, 2016, for status of scientific knowledge and professional expertise). This makes it important to consider children's participation status in a context that, apart from emphasising children's participation, also makes distinct sciences and technologies of evidencing central.
For instance, the etymology of the word 'evident' refers to the Latin word vidēre, 'to see'. The term hence suggests a 'visual bias' (Holmes et al., 2006: 182). Thus, the previous critique of the inherent primacy of vision and its reliance on positivist logics of distancing, presumed impartiality, objectivity and gender neutrality is applicable to both the legal discourse on evidence and what in recent times goes under the term evidence-based knowledge/practice (Burman, 2008;Haraway, 1988;Holmes et al., 2006;Sevenhuijsen, 1998). Using the metaphors body and voice and drawing on a wide range of critical contributions on biopolitics, including from feminist, postcolonial and critical childhood studies, this article interrogates the limitations and possibilities of the emerging focus on 'the child' and children's voices as a metaphor for children's participation rights in child welfare and protection as a knowledge culture.

Body, voice and biopolitics
The biological body has been in focus in much scholarship in the aftermath of Foucault's works on biopower and 'the politics of life' (Foucault, 2008;Rose, 1998Rose, , 2001Wells, 2011). Biopolitics has generated responses to biological bodies of 'biological sufferers' or 'psychosomatic sufferers' whose life conditions are responded to with medical, therapeutic and symptom-oriented treatments (Knezevic, 2020a;Rabaia et al., 2014;Sweis, 2017). Also foregrounded are particular techniques of 'evidencing'. For instance, critical anthropologists use concepts such as 'biometrics' and 'biological evidence' to denote some contemporary biopolitical phenomena of inclusion and exclusion (Fassin, 2011;Fassin and D'Halluin, 2005;Ticktin, 2011: 140). Biological evidencing is rooted in a seeing-believing epistemology -the visible and tangible -but also one that emphasises professional experts and scientific knowledge as truth makers (Fassin and D'Halluin, 2005). In such a conception, the biologically evincible is simultaneously measurable, observable and possible to categorise. For instance, in the context of immigration politics, that which is acknowledged as truth is not what the immigrants are expressing with their voice but rather what their bodies perform and display, when seen through the biomedical gaze (Fassin, 2011;Ticktin, 2011). Together, these accounts suggest how biopolitics foregrounds the population as a body, rather than the commonly used metaphor for citizens' participation, voice, as the channel through which citizenship rights claims and conditions for participation are being made (Chatterjee, 2004;Fassin, 2011;Knezevic, 2020a;Shier, 2001;Wall, 2011).
However, much research on children's rights emphasises voice as a metaphor for participation (Eriksson, 2012;Heimer and Palme, 2016;Shier, 2001;Spyrou, 2011;Sundhall, 2008;Wall, 2011). For children, participation is usually linked to the UN Rights of the Child and it is claimed to play an indicative role for their status in various contexts.
As critical childhood researchers have noted, the metaphor of voice may evoke biological connotations and biological reductionism of children in similar ways as does a focus on the biological body. It is widely discussed that children's participation rights are adjusted to their age and maturity, hence bear biological and developmentalist connotations. There is an underlying assumption that someone else is supposed to assess children's age and maturity in order for them to have a say (Lee, 1999;Wells, 2011). According to Wells (2011), biopolitics as a mode of governance laid the cornerstone for children as rights holders. Children's rights to health, well-being, and life are explicit examples of this. Simultaneously, rights associated with children's autonomy and their right to voice and participation are detached from structural issues and made with a monolithic depoliticised subject in mind -a socially disembodied body (Burman, 2008;Knezevic, 2020a;Raby, 2014;Wells, 2011;Woodhead, 1999). Thus, although the metaphors of body and voice link to diverse regimes of power and governance (Chatterjee, 2004), the developmentalist lens on the body of the child does not necessarily rule out participation rights but is rather intrinsic to the conceptualisation of children's rights as such (Burman, 2008;Wells, 2011). This makes body and voice difficult to separate (Lee and Motzkau, 2011).
The legislation informing the Swedish child welfare and protection services exemplifies the aforementioned body-voice tensions. The Social Services Act supports the child's right to be heard, and the principle of the best interest of the child (sv. Socialtjänstlag 2001: 453 § § 1:2, 11:10). The prerequisites for out-of-home care, however, suggest a biopolitical focus on children's health and development. Thus, for a child to be protected with an out-of-home placement, there must be a risk to the health and development of the child (Care of Young Persons Act, sv. Lag om särskilda bestämmelser om vård av unga 1990: 52 § § 2-3).
In addition, psy disciplines (Rose, 1998) and life sciences that have occupied a great role in studies and practices related to children, including child welfare theory and practice, tend to emphasise health and development (Burman, 2008;Knezevic, 2017Knezevic, , 2020aLee and Motzkau, 2011;Pettersson, 2001;Wells, 2011;Woodhead, 1999). This also applies to the Swedish child welfare and protection assessment framework, BBIC, for which developmental psychology constitutes a central theoretical point of departure, one that remained untouched although the framework itself is a 'travelling' idea (Knezevic, 2017;NBHW, 2018).

Child protection assessment reports
In order for an out-of-home placement to take place in Sweden, a social worker must first recommend the intervention in a written case report. While Administrative Courts are the decisive bodies in this process, the municipal Social Services gather and evaluate information from various sources and in this manner provide evidence for their claim.
In this article, I refer to the Swedish child welfare and protection system and case reports as the child protection assessment reports that recommend out-of-home placements. Such placements may be provided on voluntary (Social Services Act, sv. Socialtjänstlag 2001: 453) or compulsory ground (Care of Young Persons Act, sv. Lag om särskilda bestämmelser om vård av unga, 1990: 52). Both grounds are mentioned in the case reports below. Thirteen child protection cases from a Swedish middle-sized municipality, all closed in 2015, are used as examples of evidencing needs for child protection. They address ten children. Of the children assessed, eight were consulted (sv. barnsamtal).
In the other cases, reasons cited were young age, anxiety and impermanent. These children were observed. The case reports are guided by the framework for assessment BBIC and the 'BBIC triangle'. The triangle illustrates the three main areas that ought to be assessed by practitioners: child development, parenting capacity, and family and environment.
The case reports are selected from a larger sample of 283 assessment reports. The choice fell on children taken into care and assessment reports that address children's exposure to violence (child maltreatment and abuse, intimate partner violence and child neglect), and that together exemplify variation regarding information sources used in contexts of violence. This, however, does not include all possible evidence used in child protection cases in the large sample, such as for instance drug test results. The sources identified were: consultations with (pre-)school teachers, health care professionals (including psychiatry), women's shelter workers; medical expertise (i.e. forensic medicine), police and legal authorities, social workers that have been in contact with the family, scientific research, and conversations with children, parents and other people in the child's life. While the focus is on evidencing, it is important to note that the word 'evidence' is rarely used in these documents. However, words such as 'facts' are used and the assessments are written in a detached manner, and seemingly from the child's perspective, for instance when referring to parents as 'mum' or 'dad', although children are seldom quoted (see also Knezevic, 2020a). The quotations in the analysis below are taken from the more 'argumentative' sections of the reports where the most relevant information is outlined and summarised, that is, 'Analysis and Assessment' and 'Decision'.
When identifying modes of evidencing, I have focused on evidencing as a performative practice and the gathering and using of 'evidential' information as a 'doing' (Butler, 1990;Holmes et al., 2006). I identified scientific evidencing in those cases where social workers refer to research. I interpreted professionals' observation of families as another mode of evidencing. Biological evidencing links more broadly to biological bodies, such as development, health, medical conditions and diagnoses. Finally, legal evidencing refers to the evidence addressing legal issues or stemming from police and legal authorities. These forms are distinguished analytically, yet may overlap. For instance, teachers' and other referents' observations and expert opinions may simultaneously be biological evidencing. These modes, furthermore, are considered in relation to the included accounts from (child) service users.
The study was approved by the regional Ethical Review Board in Uppsala (dnr. 2014-350). The cases are translated and anonymised through pseudonyms. Some details are removed, and some alterations are made to age, gender and family constellation (where it was not the object of analysis).

Seeing-believing speaking bodies
Michel gets a lot of attention in the assessment of his younger sister. The assessments of Michel date years back in time, with referrals from various places: school, friends of the family, the police and health care professionals. Information is included about the child's exposure to several forms of violence and risks to health and development. At the time an out-of-home placement is recommended, Michel manifests behavioural issues and developmental delays. The Social Services express 'a profound concern for [the child's] health and development' and refer to alarms raised in previous assessments: The school has expressed serious concerns about care deficiencies stating that Michel is feeling tired, has poor hygiene, smelled of [cigarette] smoke, urine, dirt, had been skinny and perceived as apathetic. He has been considered to be at a low level of maturity and has had major difficulties in social interactions.
While the rich description includes the sense of smell, the description in the quotation above nevertheless aligns with the visualising practice of observation. The assessment of Michel follows the logic of seeing-believing (Burman, 2008) that figures strongly in what seems to be evidencing based on previous observations by social workers and teachers. There is an emphasis on physical and mental health and well-being, which is visibly inscribed onto Michel's observed body ('tired', 'poor hygiene', 'skinny') and observed behaviour ('apathetic', 'difficulties in social interactions'). In this sense, it is the body that 'speaks' through biological evidencing (Fassin, 2011;Fassin and D'Halluin, 2005;Ticktin, 2011), 'biological suffering' (Sweis, 2017), or psychobiologism and 'psychosomatic suffering' (Knezevic, 2020a).
Similar biological and developmental approaches to children's bodies and minds inform other cases. In the case of Theo, there is information about witnessing intimate partner violence but there is also a previous report to the police for parents' failure to ensure that Theo attends school. This violates the child's right to education and to compulsory school attendance. However, the final assessment of Theo and his siblings downplays these dimensions. Instead, the focus is on the school that the child is attending, which is described as inappropriate due to the 'cognitive difficulties' of the child. This, in turn, is assessed as parents' lacking insight into the 'special needs' of the child. The assessment of these two children as vulnerable and in need of protection links primarily to biological age and deficiencies in cognition.
Unlike voicing, the objectified 'speaking' body in the case reports is seldom linked to any specific age. Instead, 'low level of maturity', 'deficiencies in cognition', etc. are ranked against what is understood to be the normal developmental stage for a certain age (Woodhead, 1999). However, there are nuances also to the age aspect. This can be seen in relation to the excerpt above describing 'low age' as indicating vulnerability. White (1998) discusses how social workers need to find out during the assessment whether psychiatric issues and developmental problems of the child are intrinsic innate issues or issues that are an effect of the child's milieu. I interpret the case of Michel and his sister exemplifying this exploration. It suggests that when there are concerns for more than one child in the family, these concerns are more easily linked to the home and family situation than innate (neuro-)biology of the individual child. This may be why Michel and Theo get a lot attention in the assessment reports about their younger siblings. The developmentalist idea of the generic child suggests precisely this kind of generalisation where the older sibling becomes the prototype for what may face the younger ones if they continue to live in the same home. While the older sibling may undergo multiple assessments, for a younger sibling the intervention may be immediate or happen at an earlier 'stage'. Hence, although developmentalist assessments tend to foreground stages, age and stage cannot be entirely separated (Knezevic, 2020b;White, 1998). In this sense, the body of Michel or Theo is also 'speaking' for their younger sibling, or at least suggests the kinds of 'risk politics' (Rose, 2001) that the investigators need to consider.

Biocartographies
The collaboration between professionals in Swedish child welfare and protection practice enables, to borrow Ong's terminology, a well-documented 'biocartography' (Ong, 2006: 195). These may involve consultations from regular examinations and documentations by child health care, (pre-) school but also psychologists (Knezevic, 2020a). In this sense, the generic and universal developing child body 'speaks' through specific documents, such as reports based on professionals' observations and medical records, but also through the absences thereof.
These [medical] records show signs that [father] failed to give the child the care she needed. For example, from the age of three, examinations at the child health clinic have been difficult to perform, due to [child's] lack of concentration and immaturity.
As seen in the assessment of Patricia, described as 'not sufficiently mature for school', the failure to perform regular developmentally stage-based medical examinations evidences delayed development. Constraints to examinations, or gaps in the biocartography are viewed as signs of developmental damage, and consequently of neglect itself.
Neglect, furthermore, is a recurrent theme in the case reports. Although suspicions of exposure to emotional, sexual, psychological and/or physical forms of violence are raised in relation to at least one child in each family, legal evidencing, such as previous police reports and interrogations, is downplayed. Also, testimonies of the service users are inconsistent when several service users are consulted or when a service user changes the narrative. Thus, the social workers find themselves in situations where visible physiological injuries as direct signs of violence are altogether absent and they cannot determine what version is true if the child's version of an event is at all known and differs from that of parent(s) or siblings.
However, it could also be interpreted as meaning that what children say, witness or report on seems irrelevant in comparison to what their bodies are evidencing (Fassin and D'Halluin, 2005;Sweis, 2017;Ticktin, 2011). As noted in the excerpts above, 'speaking' bodies and behaviours, as observed by teachers and health-care professionals, figure strongly in child welfare evidencing. To borrow Fassin's expressions, voices are silenced and suffering is doubted until the bodies 'speak' (Fassin, 2011: 288). In this context, the bodies may speak of psychosomatic suffering (Knezevic, 2020a) or biological or developmental harm, that is, neglect. As noted previously, this 'speaking' may be possible in the aftermath of some forms of violence, in this case long-term neglect, and less so as a result of violence that does not leave physical or visible psychological marks (Fassin, 2011;Fassin and D'Halluin, 2005;Knezevic, 2020a).
This analysis is aligned with previous research indicating that severe physical violence and neglect is often required in order for children to gain protection (Leviner, 2014;Linell, 2017;Östberg 2010). What the excerpts above also suggest is that symptoms of such violence need to be seen and validated by professionals to be believed or documented in medical or school records -in biocartographies and the biopolitical administration of life, health, well-being and development (Foucault, 2008;Wells, 2011).

Predicting-believing
Besides observing bodies and behaviours in situ, social workers also take into account previous assessments and knowledge about a family. This previous knowledge becomes that which is 'known' over time, and helps to assess the future (Knezevic, 2017(Knezevic, , 2020b. In assessments of risks, seeing-believing shifts into, or is combined with, predicting-believing. The predictions figuring in assessments draw primarily on scientific research. Scientific evidencing occurs for instance in a case where the problem of (intimate partner) violence figures in previous criminal records and police reports, and what could be interpreted as a form of legal evidence of crime. In the assessment of 7-year-old Sara, exposure to violence is described as a serious issue. Yet, legal evidencing is downplayed even though sentencing of the violent parent had taken place. The Social Services instead foreground future biological and developmental problems and issues in 'social adjustment': Research shows that children who witness domestic violence are affected in five main areas; physical or biological functions, behavioural, emotional, cognitive development, and social adjustment [reference to research]. For example, bodily stress reactions are common, such as asthma, eczema, eating disorder, stomach ache, headache and sleep difficulty among children living with violence in the family [reference to research]. A child who witnessed violence in the family can also have long-term mental health problems such as depression, anxiety, self-destructiveness, aggressiveness, and low selfesteem [reference to research]. Research also indicates long-term impacts on personality such as lack of trust in other people and a pessimistic view of the ability to influence one's own life situation [reference to research].
Although no biological evidence is found in this case, scientific evidencing in the passage above is still underpinned by biologism and developmentalism, though in terms of potential effects on the bodies and minds of children. Also 'social adjustment' is mentioned, which is primarily linked, again, to the child's mind and the prospect of a pessimistic outlook on others or herself, which here serves as a 'measure' of Sara's ability to change her life circumstances as an adult. This understanding of the 'social' simultaneously depicts a psychologised subject outside the 'societal' (Knezevic, 2020b). Biopolitics of childhood produces a neoliberal (child) subject whose life situation relies on the state of her mind (Foucault, 2008;Ong, 2006;Raby, 2014;Wells, 2011). Ponnert (2007: 185) discusses for instance that social workers, after reaching 'moral conviction' that out-of-home placement of a child is necessary, perceive that they have to present a 'legitimate conviction' to the court. Yet, scientific evidencing of future risks is, as Leviner (2014) has noted, not easily aligned with the legal evidencing in child protection contexts. A reason mentioned is that jurisprudence, that is, Administrative Courts tend to dismiss 'hypothetical risks': This might lead to a paradoxical situation when the courts, contrary to the knowledge base within medicine and behavioral science, but in line with the dismissal of 'hypothetical risk' as a ground for intervention, will only find risk of harm proven when children are already seriously harmed. As a result, the proactive ambitions and dimensions of the law are to a large extent limited to avoiding further harm in cases where harm has already occurred, i.e., a reactive outcome. (Leviner, 2014: 217) However, in the assessment of the toddler Malcolm the legal discourse co-exists with the medicalscientific, and with fewer tensions than those described by Leviner. This happens primarily through evidencing through forensic medicine ('medical expertise') and scientific research: Medical expertise has determined that these injuries have been caused to [Malcolm] by another person and that the damage may not have occurred in normal handling of a child. This states that parents have significantly failed in their care of [Malcolm] and in the ability to protect him. At the time of writing, [Malcolm's father] is currently suspected of probable cause and a police interrogation is under way. Whether or not the police interrogation leads to sentencing, the fact remains that [Malcolm] has had these injuries, and that someone has caused them by violence. The great concern of the Social Services for [Malcolm] is thus independent of prosecution. Children exposed to violence risk unfavourable development and the consequences of violence can extend over both short and long periods. In addition to causing physical damage, violence can affect both mental and emotional, cognitive and social development. One of the most serious consequences for a child who is exposed to violence is that the child suffers from what is called a failure to thrive. [. . ., reference to research]. [Malcolm] does not suffer from the failure to thrive but it could have been a consequence if the violence against him did not stop or recommence.
Unlike participatory citizenship in the traditional sovereign rule, biopolitics foregrounds the population as a body rather than voice (Chatterjee, 2004;Foucault, 2008). Following Chatterjee (2004) and Wells (2011), laws and conventions do not stand outside biopolitical governing, but are very much part of it. In the quotation above, the physical injuries evidenced by 'medical expertise' overshadow the lack of legal evidencing of crime. A child-focus, rather than a focus on the parent, makes it possible to refer to injury as both a medical and a legal 'fact', that is, injury caused by another. Even though the prosecution is not clear and the father only has the status of a suspect 'the fact remains that [Malcolm] has had these injuries'. Another point of reference used is scientific research on 'failure to thrive', as 'physical and psychological response to abuse', and as a risk to children's physiological, mental and social development. This case suggests how the Social Services navigate between modes of evidencing that, at a first glance, seem to be in tension with each other (Leviner, 2014), yet paradoxically, are mutually supportive.
I interpret the quotations above as being line with a predicting-believing of risks, a technique of evidencing futures. In terms of children's status in evidencing and their participation rights, predicting-believing becomes an epistemic challenge. Previous research about competing and asymmetrical epistemic statuses and in relation to children in institutional settings foregrounds different temporal aspects of children's participation status. For social workers, listening to children is an end in itself and is not necessarily linked to delivering services in line with children's wishes, according to Pölkki and colleagues (2012). Children's status as knowers, thus, is questioned in relation to the future, as knowledge claims about the future in terms of risk assessments and predictions are more available for professionals and scientists than children's wishes and suggestions (e.g. McLeod, 2006). Yet, as others note, not even experiences are taken seriously when expressed by children (Eriksson, 2012;Iversen, 2013;Sundhall, 2008). However, this study brings this discussion to another level, beyond subject/object and beyond voice. Regardless of whether Sara or Malcolm had the opportunity or capacity to voice, their voice is irrelevant for the main target of the knowledge culture and its mode of evidencing, that is, predicting future harm. In other words, predictions are reserved to specific scientific research in similar ways as observations of bodies and minds are restricted to certain professionals.
At the same time, scientific research enables an assessment of risks also when there are no obvious signs of symptoms (Sara's case), or when the child is too young to participate and provide their perspective on the situation (Malcolm's case). In this case, scientific predictions of the kind discussed above may be the only evidencing of needs for protection for social workers to draw on.

Listening to?
Above I identify seeing-believing and predicting-believing as two main modes of evidencing in the assessment reports under study. Against this background, however, the assessment of the 11-year-old Samir seems like a contrast. A crucial difference highlighted by this case is that it indicates a response to violence, rather than to children's symptoms and developmental damage. Another unique feature of this case is that the boy, because of his short stay in Sweden, lacks a biocartography (Ong, 2006) in the form of past-time professional documentation and written assessments. Hence, information surrounding education and health is scarce or altogether missing. The description of Samir's health status suggests that even in an assessment where the child's participation is high compared to other objectified children, determining health status as a domain is still restricted to adults, in this case shelter workers and his mother (see also Hultman and Cederborg, 2014). However, Samir's voiced concerns during conversations with social workers about physical, verbal and psychological 'gross and serious violence' are taken seriously and serve here as source of knowledge. They are discussed in the final pages of the assessment report. He occupies a position best equated with the position of a moral and trustworthy child with voice (Knezevic, 2017;Sundhall, 2008).
The investigators perceive Samir descriptions of dad's violence as credible and consistent.
The child's expressed experiences are sufficient as evidence -without any visible, measurable, and observable biological or legal evidence of harm, and without years passing under the Social Services' gaze. However, psychobiologism is still taking place. Here, however, it does not refer to the child but to the 'developmental disorder' and the medical condition of the parent. Elsewhere (Knezevic, 2020b), I address these assessments as examples of urgent removals due to assessments of parenting but also prevailing understandings of social problems. I argue that child welfare not only responds to development of children but also of parents and that those that are assessed as developmentally deviant -either chronically or over time -are assessed as problematic. Furthermore, I discuss this in relation to wider understandings of social problems as linked to individual pathologies rather than structural relations of power.
Reading the case of Michel and Samir, the violence is serious in both, and both are old enough to talk about it. Still, Michel becomes a child protection concern because of neglect whereas violence is a sufficient reason for removing Samir from the home. Taking a child's narrative of violence at home seriously can be, for instance, a matter of the child's ethnicity/race (e.g. Bruno, 2015;Knezevic, 2020a), but this can only be considered with caution in relation to the material under study. The striking resemblance between the assessments, nevertheless, is that both are informed by developmentalism and damaged bodies -either targeting the child (Michel) or the child's parent.
Similar psychobiologism informs the case report addressing the young child Audre. The young age of the child reduces the child's participation to child welfare professionals' observations of child-parent and parent-parent interactions. For professionals to assess needs for protection, a child -as the case of Audre clearly shows -is not required to speak. [Parent's] inability to mentalize and put themselves into the perspective of others indicates a very worrying immaturity that manifests itself in several areas, for example, in their difficulty to cooperate with each other.
[Mother] shows a high level of need for justice and tends to focus on how unfairly she is affected by the situation instead of maintaining focus on [Audre].
In these sections, I have discussed legal and scientific evidencing and primarily evidencing through observation-based documentations as well as professionals' expertise and observations in situ. These latter modes of evidencing are discussed elsewhere (Knezevic, 2020b), and referred to as 'temporal distancing' (Fabian, 2014) between (mature) observers, that is, health care professionals, teachers, social workers etc. and those viewed as immature. In the case of Audre's parents, their 'inability to mentalise' is viewed through what seems to be the professionals' objective all-encompassing outlook of possible perspectives and emotional states (Haraway, 1988). The parents, unlike the observers, show the 'inability [. . . to] put themselves into the perspective of others' and 'show empathy'. The possibility for a parent to resist these descriptions and voice feelings of unjust treatment seems restricted when 'need for justice ' (sv. rättvisebehov) in the quotation above is also seen through the developmentalist lens of 'immaturity'.

Discussion
The issues raised in this study are of concern for all fields in which EBP is advocated, in particular for child welfare social work that simultaneously has increased the emphasis on children's rights, including the right of participation. While not addressing EBP directly, I try nevertheless to show some broader epistemic implications of an epistemology rooted in positivism, visualism and prediction and how some articulations of it discern knowledge and knowers in distinct ways.
I have discussed psychobiological harm as the primary target of evidencing, whether the information stems from teachers, social workers, health professionals, medical experts, or scientific findings. In some cases, it is not until the presence of the visibly biologically and developmentally injured child body, as documented and observed by professionals, that the Social Services recommend interventions (Knezevic, 2020a;Leviner, 2014;Östberg, 2010). The biological body, to quote Fassin, is 'the site where truth is sought or denied' (Fassin, 2011: 284;Fassin and D'Halluin, 2005). In these contexts, what children say, witness or report on seems less relevant than what their bodies are evidencing. This suggests that the scope of children's own evidencing is one of the 'speaking' biological body (Fassin, 2011), or temporal categories subject to observations of development or scientific predictions (Fabian, 2014;Knezevic, 2020b). Ideals regarding children's participation rights become rather unintelligible within a knowledge culture underpinned by professionals' seeing-believing and scientific predicting-believing, rather than listening to and hearing.
Additional questions arise in relation to the status of the 'speaking' body when these children simultaneously may be assessed as immature, anxious or impaired cognitively or in speech. When viewed in this way, the focus is not only on the corporeal downplaying voice. It is also questionable that voices of those bodies that 'speak' -the damaged, impaired and immature -can be taken seriously. Developmental damage is a prerequisite for inclusion of a citizenry that faces exclusion on the same grounds that includes it -health/able-bodiedness (cf. Fassin, 2012;Knezevic, 2020a). From such a point of view, an analysis of modes of evidencing reframes the question from violence against children that might be harder/easier to evidence to how modes of evidencing make certain already disadvantaged children and families easier/harder evidencible (cf. Fassin and D'Halluin, 2005). Categories such as health, illness, disability, etc. are crucial to take into account together with age, class, gender and ethnicity/race when analysing these complex relations of power that serve to depict the deserving victims while simultaneously depriving them of participation rights.
This article poses critical epistemic questions regarding children's access to the domains that should be known, and thereby evidenced. The scientific evidencing that figures in assessments produces, in Fabian's terms (2014), a temporal distance between the 'speaking' body predicted or observed by professionals and the children's voicing of the present and the past (Knezevic, 2020b). Scientific prediction is beyond children's epistemic access and voicing. In addition, the psychobiologism discussed above is beyond children's 'expertise'. Whether the concern is future prediction, or children's health, children's access to and accounts of these domains are constrained, if at all existing.