A Critical Analysis of Key Policies Shaping Services for Young Children with Disabilities in South Africa

There have been significant recent developments in the policy arena in South Africa in respect of disability and of early childhood development viz. the White Paper on the Rights of Persons with Disabilities and the National Integrated Early Childhood Development Policy. Using Bacchi’s analytical framework encapsulated in the question, What’s the problem represented to be?, these policies were subjected to scrutiny, analysing how disability and inclusion are represented within them, and how these representations shape the lives of young children with disabilities. Among the underlying assumptions contained in these policies is that barriers excluding persons with disabilities are socially constructed and their removal will result in inclusion. Further, the policies imply that marginalisation and vulnerability are inherent traits of children with disabilities. Inclusion is portrayed as an ideology rather than a practice. Amongst unproblematised elements of the polices is the supposition that children with disabilities are a homogenous group, and that there is a distinction between children with and without disabilities. The effects of these representations manifest in a focus on social barriers, which downplays the importance of habilitation and rehabilitation for the individual child and may result in children with disabilities being portrayed as passive recipients of services. It may also undermine choice and agency of children and their families. Further, the analysis indicates that defining the principle of inclusion too broadly makes it difficult to measure.


Introduction
By its very nature, social policy reflects a government's intent, outlining what needs to be done for whom, acting as "a form of prescriptive text" (Bacchi 2009, 34). It is concerned with the principles and practices of government in pursuit of social, political and economic outcomes (Goodwin 2011). Social policy is based on particular values and assumptions about human development, involving implicit and explicit assumptions about what constitutes the public good and what private goods individuals should or should not be able to pursue (IRIS 2012). Understanding policy as discourse 1 "start [s] from the assumption that all actions, objects, and practices are socially meaningful and that the interpretation of these meanings is shaped by the social and political struggles in specific socio-historical contexts" (Goodwin 2011, 170).
One of the consequences of political discourse and policies is the way in which they shape problematisations and categorisations of people, rendering some groups culturally, socially and biologically different. By producing narratives of what and who are in need of changing, they establish entry points for addressing the "otherisation" of already disenfranchised groups (Helberg-Proctor et al. 2017). Because the processes of policy making and policy analysis are located within particular paradigms or belief systems, they need to be subjected to constant scrutiny and revision (Thissen and Walker 2013). This is the challenge to "critically consider the categories, language and 'problem formulation' included in these policies to prevent (unintentionally) contributing to processes of exclusion and otherisation" (Helberg-Proctor et al. 2017, 9). It is therefore critical to subject to scrutiny policies that purport to protect and promote the development of young children with disabilities in South Africa.
As part of the reforms that took place in South Africa during the early years of the postapartheid era, a plethora of policies was developed. These were intended to reflect the values and ethos of the South African Constitution of 1996, and the commitment by the democratically elected government to redress the legacy of racial inequalities resulting from apartheid, and promote equity and non-discrimination, particularly for those who had experienced discrimination on various grounds, including disability.
Since then, there have also been significant developments in the legislative sphere, supported by South Africa's obligations under international law by virtue of its ratification of various treaties. For example, the Children's Act 38 of 2005 (amended 2007) was intended to give effect to the government's responsibilities under the UN Convention on the Rights of the Child (United Nations 1989). Active civil society 1 Discourse is understood as "socially produced forms of knowledge that constitute the 'real'" (Bacchi 2016, 8).
3 lobbying contributed to significant provisions for children with disabilities being included as part of this act (Jamieson and Proudlock 2009).
In 2007, the South African government signed and ratified the United Nations Convention on the Rights of Persons with Disabilities (UNCRPD), confirming its recognition that persons with disabilities are holders of rights that entitle them to be treated equally and without discrimination. This convention is considered to be a milestone in the struggle for disability rights, particularly because it provides an opportunity for signatory states to initiate policy reform and shape domestic law (Lord and Stein 2008). Indeed, in South Africa the UNCRPD was locally domesticated through the development and adoption of the White Paper on the Rights of Persons with Disabilities (hereinafter the Disability Policy) in 2015. Its intention is to expedite redress and transformation in respect of inclusion and equality for children and adults with disabilities in the country, towards achieving the mission of "inclusive and equitable socio-economic development" (RSA 2016, 42).
Fuelled by evidence from neuroscience, there has been growing recognition of the importance of early childhood development (ECD) as a foundation for learning and good health for every child. This culminated in the approval by Cabinet of the National Integrated Early Childhood Development Policy (hereinafter the ECD policy) at the end of 2015 (RSA 2015), with the vision that all young children and their families live in environments that are conducive to their development. Focusing on the period from conception until the year before children begin formal schooling (or for children with disabilities, until the year they turn seven), this policy aims to transform ECD service delivery in order to provide comprehensive, universally available and equitable ECD services. This is to be done through defining a comprehensive package of services for young children and establishing an enabling framework and co-ordinating structure within which these can be provided.
Although there has been some research conducted on policy analysis in the field of disability in South Africa (see Donohue and Bornman 2014;Storbeck and Moodley 2011), most studies have focused on policy implementation and have not engaged in indepth analysis of the problem representations within policy. The studies by van Rooyen, Le Grange and Newmark (2004) and van Rooyen and Le Grange (2003) reflect an indepth analysis of Education White Paper 6: Special Needs Education: Building an Inclusive Education and Training System (DoE 2001) from a poststructuralist framework using deconstruction as a strategy for reading policy. In their research, the scholars deconstruct interpretive discourses and radical humanist discourses by identifying how objects, agents, and actions are constituted, and examining voices on the margins. Our study contributes to this limited and emerging body of literature on disability policy analysis in the South African context. The interrogation of problem representations in respect of disability and inclusion signifies the unique contribution of this study. 4 Both the Disability Policy and the ECD policy purport to be inclusive of young children with disabilities. The provisions of the Disability Policy include ECD, and the ECD policy contains provisions for children with disabilities. The aim of this paper is to subject these policies to scrutiny, particularly analysing how disability and inclusion are represented, and how these representations shape the lives of young children with disabilities. The key research questions are the following: How is disability represented in the selected policies? What are the silences, and what is left unproblematic? What is the impact of these representations on the struggle for inclusion and equity for young children with disabilities?

Analytical Framework
Policy formulation is often viewed as a neutral, technical process, with policies developed by those considered experts in a particular field. Within the evidence-based paradigm, research and "scientific knowledge" are seen as providing rational grounds for policy development and analysis (Bacchi 2015;Goodwin 2011). Indeed, the ECD policy itself (RSA 2015, 18) refers to the "evidentiary foundation of ECD as a development imperative," indicating that it too is based on "evidence" of the benefits of ECD.
Typically, problems and problem identification (reflected in the "problem statement" of a policy) are assumed to be the starting point for determining what must be done, with policy seen as the means of addressing the problem. Implicit in the view of policy as a means of "problem solving" is the assumption that problems exist outside of the policy process and simply need to be identified and responded to. This suggests a linear process, moving from a problem towards a defined solution. Within the rationalist, critical and interpretivist approaches to policy, the primary focus is on finding solutions, which are often technical in nature (Bacchi 2016). These approaches view policy analysis as a tool to solve problems, underpinned by rationality and the ability to reach objective conclusions (Goodwin 2011). The underlying premise that there is something that is not working and needs to be fixed leads much work of policy to respond with a focus on best practice or what works. Bacchi (2015, 30) cautions of the dangers of relying on researchers to bring "solutions" and deliver knowledge on "what works" for the "problems" identified by others (such as government) because it masks the political contestation around competing representations of problems. In applying this approach to the disability sector internationally and in respect of disability mainstreaming and inclusive development, Marshall (2012, 58) notes a resultant "narrowing, totalising trend." This study draws on the work of Carol Bacchi as its analytical framework. Bacchi challenges the traditional and pervasive view that policies address problems. Instead, she argues that "policy is not a reaction to 'problems' that sit outside the process waiting to be addressed or 'solved.' Rather, policies produce or constitute 'problems' as particular types of problems" (Bacchi 2015, 4). In other words, policy formulation is 5 not a process of problem solving, but rather of problem construction. This approach provides a theoretical framework for policy analysis, encapsulating a shift in focus from the best way to solve social problems (a "problem-solving" paradigm), to interrogating the conceptual foundations of how these problems are represented (a "problem questioning" paradigm) (Bacchi 2009).
Bacchi's approach is summarised in the question, What's the problem represented to be? (WPR), and follows the simple logic that "what we propose to do about something indicates what we think needs to change and hence what we think is problematic-that is, what the 'problem' is represented or constituted to be" (Bacchi 2016, 8). It is based on the proposition that we are governed through problematisations, i.e. by the discourses that result in "problems" of a particular type within policies. Citizens are seen as individuals who are self-regulating, and little attention is paid to the circumstances that shape their lives (Bacchi 2015). Because governing is a "problematising activity" it is necessary to study problematisations rather than "problems." It follows that there is a need to interrogate the problematisations contained in policy through subjecting to scrutiny the premises and effects of the problem representations they contain (Bacchi 2009). By their very nature, problematisations within policies simplify issues and reduce complexities; they organise things in certain ways, which often has the effect of categorising and dividing populations into groups. Understanding how this is donewhat is left in and what is left out-constitutes the core of analysis.
Bacchi calls for the exploration of subjectification within policies, in order to identify the "way in which the problem representations within policies often set groups of people in opposition to one another-a dynamic Foucault calls 'dividing practices'" (Bacchi 2009, 16). Indeed, many public policies rest on binaries or dichotomies, implying mutual exclusivity. Because binaries serve to simplify complex relationships, the task in policy analysis is to identify where they appear and their influence in shaping an understanding of an issue. It is thus critical to subject to scrutiny the ways in which dominant frameworks of meaning shape policy proposals in respect of disability, if these are to result in real and meaningful change. As Marshall (2012, 66) observes, It is precisely the impoverishment, marginalisation and discrimination experienced by disabled people globally that demands that all those whose practices are implicated in the production of this "problem" adopt a mode of thinking that creates space for critical evaluation and incorporates reflexive practice.

The WPR Approach
The starting point of the WPR approach is a particular policy-what Goodwin refers to as the "policy text." Then one works backwards (from what is proposed) to identify the problem representation. In this way, analysis begins with what is proposed as "solutions," seeking to identify and subject to scrutiny the problem representations 6 which are implicit within them (Bacchi 2012). This requires critically examining the assumptions, omissions and effects of the problem representation.
The WPR approach proposes a set of questions to probe how "problems" are represented in policies, thus providing a clear methodology that can be applied across various policies (Bacchi 2009;. These questions are not envisaged as a formula, but rather are seen as facilitating "open-ended critical engagement" (Bacchi 2012, 23) towards developing increased awareness of the forms of power that shape how problems are represented. The goal is to identify the impact that a particular policy has on the daily lives of those it purports to serve, in order to "be able to say which aspects of a problem representation have deleterious effects for which groups, and hence may need to be rethought" (Bacchi 2009, 18). The questions posed by Bacchi have been adapted and clustered according to elements proposed by Helberg-Proctor et al. (2017), as contained in Table 1 below. Genealogy: developments, decisions, processes, events in the discourse on disability and inclusion What is left unproblematic in this "problem" representation? Where are the silences?
Silencing: criticism, gaps, limitations, simplifications in the discourse on disability and inclusion Prognosis What are the implications of this "problem" representation?
What effects are produced by this representation of the "problem"?
Effects: perspectives, opinions on disability and inclusion, and implications for young children with disabilities and the society they live in Because of the recognition that the ways in which "problems" are identified "elicit particular forms of subjectivity, influencing how we see ourselves and others" (Bacchi 2012, 22), WPR requires posing these question to one's own proposals, subjecting oneself as researcher to analysis, and one's own conceptual and structural positionality to scrutiny (Marshall 2012). Thus, an essential part of the analysis is selfproblematisation or "reflexivity." Reflexivity is central to the WPR approach, and the researcher is a reflexive subject, having to concede that they are an integral part of the problem representations that they seek to analyse, and cannot be positioned outside discourse (Lindsköld 2010). 8

What Assumptions Underlie the Representation of Disability?
The disability discourse in South Africa has its roots in the disability rights movement, which was strongly aligned with the anti-apartheid struggle and the call for equal rights for all (Rowland 2004). The understanding of disability as reflected in South African disability discourse has been shaped by many other factors, including conceptualisations reflected in policies, academic studies, and civil society both nationally and internationally, culminating in the signing of the UNCRPD in 2007. Historically, the main shifts have been from the medical model to the social model of disability and from a charity to a rights-based approach.
The Disability and ECD policies draw from the UNCRPD (UN 2006, Preamble [e]) in their understanding of disability, which is described as an "evolving concept." It is seen to comprise two key elements: having a long-term (physical, mental, intellectual or sensory) impairment, and experiencing barriers that prevent participation in society on an equal basis with others. Disability is represented as being imposed by society, which erects attitudinal and structural barriers that prevent people with impairments from participating. In its definition of disability, the Disability Policy categorises these barriers as social, psychological and structural (RSA 2016, 7). It also points out that "persons with disabilities should be defined within the context of defining the beneficiary group for purposes such as … protection against discrimination, service delivery, reasonable accommodation support measures, social security, etc" (RSA 2016, 18). The policy thus argues that it is important to look at beneficiary groups-such as young children-to see how barriers are experienced. Based on this understanding of disability, the strategic approach of the Disability Policy includes "mainstreaming," which is seen to occur on two levels: "One is ensuring that the disability element is inherent in a programme or project and persons with disabilities are included as one of the beneficiaries or target groups. The other is ensuring that budget allocations provide for any reasonable accommodation measures that may be required to provide universal access to services" (RSA 2016, 45).
In the ECD policy, the definition of disability is based on that of the UNCRPD, viz. young children with disabilities are those with various impairments, which in interaction with different barriers, may hinder their participation in society (RSA 2015, 11). The emphasis is on factors which may limit or undermine participation. In looking at the availability of ECD services, the ECD policy (2015, 23) states that government must ensure that "cost and other barriers such as disability" should not prevent people from making use of services. In this example, disability itself is seen as a barrier.
The ECD policy holds as its foundation evidence from neuroscience indicating the sensitivity of the brain in the earliest period of life, which renders the young child "developmentally vulnerable to biological, social and environmental risk factors" (RSA 9 2015, 19). Various risk factors seen as being predictive of poor early childhood experiences are identified, including disability: "Approximately 23 per cent of children between the ages of birth to 9 years in low and middle income countries are at risk for disabilities. While this is indicative of their compromised development, children with disabilities are, in addition, at risk of low access to early childhood development services and at an increased risk of poor quality care" (RSA 2015, 20). Disability is thus seen as a risk factor associated with poor access to services and poor early childhood outcomes. In its situation analysis and problem statement (s3) the ECD policy identifies factors that impact on the availability of services for children with disabilities, which include the current funding model that does not provide additional funding for programmes catering for children with disabilities, the curricula for ECD personnel which does not take into account the learning and care requirements of children with disabilities, as well as stigma and beliefs about disability and inclusion (RSA 2015, 44).
In response to the dire situation of young children in the country, the ECD policy sets out its strategy and implementation plan (RSA 2015, s5), which includes a set of programmatic priorities, one of which is the "inclusion and support for children with disabilities within all early childhood development programmes" (2015, 66). Another element of the strategy is to "prioritise vulnerable children to ensure equitable access" (2015, 68) specifically designed to address the barriers that exclude vulnerable children, including those with disabilities.
The Disability Policy sees children with disabilities as being subject to compounded marginalisation. Pillar 2 of the policy, "Protecting the rights of persons at risk of compounded marginalisation," makes reference to national legislation (the Children's Act 38 of 2005) and the UNCRPD which contain provisions that recognise specific agerelated situations and protect the rights of children with disabilities (RSA 2016, 61). Thus both policies acknowledge that there are various factors contributing to the exclusion of children with disabilities.
An analysis of the representations of disability using the WPR reveals a number of underlying assumptions. First, because barriers experienced by individuals with disabilities are seen to be socially constructed, the responsibility for addressing or eliminating disability falls on society. For example, the Disability Policy states that "the 'Universal Accessibility' approach places the responsibility on society to adjust environments, products and systems to accommodate the individual rather than the individual working around these environments" (RSA 2016, 57). Implicit in the logic of this policy is that the removal of disabling barriers and the provision of support will result in the inclusion of persons with disabilities. The ECD policy cautions however that universal access does not imply universal coverage, but rather that "Government must ensure that quality services are available to all children whose parents wish them to use such services; and that cost and other barriers such as disability should not prevent those who wish to make use of these services from doing so" (RSA 2015, 32).
Second, the assumption in the policies is that by virtue of their disability, certain children are "marginalised" and "vulnerable." For example, the introduction to the ECD policy, which gives the rationale underpinning it, states that "[s]pecific attention needs to be given to the vulnerability of children with disabilities to abuse, neglect and exploitation" (RSA 2015, 27). Further, as part of its strategy and implementation directive, the ECD policy contains the injunctive to "prioritise vulnerable children to ensure equitable access" (2015, 68), which includes children with disabilities. Pinto et al. (2012, 8) have argued that labels such as "marginalised" and "vulnerable" are "valueladen" and reinforce a deficit paradigm, suggesting that the term "priority populations" reflects greater neutrality in naming specific actions for which the state is responsible. A review of the ECD policy suggests a link between these, in which the extent of children's vulnerability becomes the basis on which priorities are established.
Further, the notion of "disabled" (or "with disabilities") as a descriptor of particular children in the policies may be challenged, with this term used not as an adjective but as a verb, thereby conveying "disabling" as a process of attribution. As Bacchi and Eveline point out (2010, 336), It follows that, since "differences" are attributions, attention shifts from those deemed to be "different," and from the characteristics identified as the basis of "difference," to the dynamics of the processes of declaring or claiming difference/s. Who is doing the designating of difference? What kinds of power do they exercise? What are the effects that accompany particular kinds of "differencing" practice?

What Assumptions Underlie the Representation of Inclusion?
Writing to define the concept of inclusion as a key element of the Sustainable Development Goals, Silver (2015) argues that social inclusion is a multifaceted and subjective construct that is shaped by particular contexts. Often contrasted with "exclusion," social inclusion has been described as "the process of improving the terms for individuals and groups to take part in society [and] … the process of improving the ability, opportunity, and dignity of people, disadvantaged on the basis of their identity, to take place in society" (World Bank 2013, 3).
Historically the concept of inclusion in the context of disability emerged mainly from the education sector (as a special education/disability issue) as an attitude and approach that embraces diversity in the learner population and promotes equal opportunities for all learners, with literature indicating a trend of using the terms "social inclusion" and "inclusion" interchangeably. For example, Banks and Keogh (2016) hold that inclusion can only be achieved for people with disabilities through the removal of barriers that exclude them, while Koller, Pouesard and Rummens (2018, 1) note that "for children with disabilities, social inclusion requires overcoming significant social, economic and political barriers to achieve meaningful involvement in society. Social inclusion … establishes a basis for overall well-being and is a critical component of becoming a 11 valued and contributing member of the community." Further Cobigo et al. (2012) argue that there are many dimensions to social inclusion, including access to community resources and activities, relationships with family, friends and others, and having a sense of belonging. Thus its core elements are "participation and engagement in … mainstream society" (Cobigo et al. 2012, 76). In both of the policy documents discussed here, reference is primarily made to "inclusion" and not "social inclusion." For example, considering provisions relating to children, the Disability Policy mentions social inclusion as a goal of community-based rehabilitation (RSA 2015, 4), and (in describing the pillar to support and promote empowerment for persons with disabilities) cites an article of the UNCRPD relating to access to health care and rehabilitation services, which are to be "directed towards social inclusion" (2015, 80). The use of "social inclusion" in relation to health and rehabilitation services is significant, indicating that reference is being made to the broader understanding of social inclusion and addressing the factors that lead to the exclusion of particular groups, and not viewing rehabilitation only as an adaptation of the individual to their impairment.

Inclusion of persons with disabilities is named in the UNCRPD (UN 2006) as a general principle (Article 3), as a general obligation (Article 4) and as a right (Articles 29 and 30
). It is also the vision of the Disability Policy, which is "South Africa-a free and just society inclusive of all persons with disabilities as equal citizens" (RSA 2016, 42). The Disability Policy represents inclusion as a vision or goal, and as "the ultimate objective of mainstreaming" (RSA 2016, 9). Key elements of inclusion in the policy are that it is a universal human right, which recognises the diversity of people in respect of many factors, including race, gender and disability, and that inclusion is seen to encompass a sense of belonging: feeling respected, valued for who you are; and feeling supported so that every child can fully participate in society with no restrictions. Further, inclusion requires that the system (and not the individual) changes in order to accommodate diversity and the range of individual differences and needs (RSA 2016).
Building on the definition of disability presented in the previous section, the Disability Policy views mainstreaming (and ultimately inclusion) as something that can be achieved through the removal of barriers, in what is referred to as "universal access." This is defined as the "removal of cultural, physical, social and other barriers that prevent people with disabilities from entering or using or benefitting from the various systems of society that are available to other citizens or residents" (RSA 2016, 15). Underpinning this representation of inclusion is the concept of universal design, which calls for all things to be designed to enable use by a wide range of people including various categories of persons with disabilities, without having to be specially adapted for a particular individual. The fundamental premise of universal design is the recognition of human diversity as opposed to the concept of the "average person" (RSA 2016, 57). The Disability Policy (RSA 2016, 81) envisages inclusion cutting across all aspects of ECD, citing three directives as to how this is to be achieved: 1. Children with disabilities must have equitable access to all ECD programmes and facilities. This requires that mainstream ECD programmes and facilities are made accessible for them. Thus building plans, playgrounds, equipment, toys and ECD practitioner training must comply with universal design norms and standards.
2. Develop disability specific intervention and support programmes.
3. Develop a national integrated referral and tracking system.
The ECD policy (RSA 2015, 13) defines inclusion as the process by which all children are actively welcomed and supported so that they can participate and benefit from early learning and development opportunities that are conducive to the learning of all children and through which all children receive the necessary support to enable them to participate on an equal level with others.
The question arises as to why inclusion is limited to early learning whereas comprehensive ECD programmes are much broader, encompassing also health and nutrition, caregiver support and social protection (RSA 2015). The ECD policy (in the rationale underpinning it), identifies ECD services that must be provided to the public, which include "inclusive and specialised services" (RSA 2015, 27). The policy calls for "clear guidance and norms on which infants and young children should be able to access mainstreamed early childhood development services … and which early childhood development services should provide specialist support through programmes targeted specially at children with disabilities" (RSA 2015, 28). Use of the terms "inclusive/mainstream" and "specialised" as mutually exclusive suggests that some children require inclusive services while others need more targeted, specialised services, as if these are two separate, distinct strands, rather than a continuum of support to be provided to children with varying needs.
Using the WPR approach, there is a need to subject to scrutiny the assumptions underlying this representation of inclusion. Within the Disability and ECD policies, the concept of inclusion is posited as an antidote to exclusion, with the logic being that children who are excluded from services and benefits need to be targeted so that they can be included. Although its definition of inclusion refers to "all children," the proposals in the ECD policy imply that inclusion applies only to children with disabilities, i.e. that inclusion applies only to those who are currently excluded. For example, "Securing the right of children with disabilities to ECD requires … [t]he development of management capacity across all government departments and spheres of government to plan, design, co-ordinate and monitor implementation of inclusive ECD services for children with disabilities" (RSA 2015, 28). Also, gaps in the current ECD system include "inclusive and/or specialized services for children with disabilities" (2015, 45).

13
The ECD policy provides for "inclusion and support for children with disabilities within all ECD programmes" (RSA 2015, 66). It calls for guidelines to be provided "for the design of ECD services to achieve quality outcomes, goals and targets for children with disabilities" (2015, 66) and the need to "[s]ecure a sufficient number of ECD practitioners to provide quality, inclusive and appropriate ECD services to children with disabilities and their families" (2015, 66).
The ECD policy (2015, s1.3.3) recommends that specific centres be identified as inclusive or specialised. It raises the question, are universal design standards to be applied only to identified inclusive and/or specialised centres? And is curriculum development for all ECD practitioners, or only those working in specific (specialised) services?
Marshall (2012) notes that inclusion has historically been central to the disability rights movement, which frames disability as a form of exclusion. Use of the term "inclusion" within government policies raises the question of what happens when such a concept "travels"-i.e. is used by the very structure that is seen as primarily responsible for exclusion. Children with disabilities who participated in a situation analysis in South Africa conducted by the Department of Social Development, Department of Women, Children and People with Disabilities and UNICEF (2012) related experiences of deep hurt as a result of the hostile and prejudicial attitudes that they encounter which result in their exclusion from activities and services of mainstream society, including health and education. While they affirmed the importance of removing physical barriers as necessary for improving access to services, addressing the long-term effects of rejection and being treated without dignity requires a much more deep-rooted and radical transformation of society's responses to diversity. In essence, theirs is "a call to be valued and to belong" (DSD, DWCPD, and UNICEF 2012, 33). Indeed, the building of social relationships is a key element of inclusion (Cobigo et al 2012;Šiška and Habib 2013).
The Disability Policy provides an example of how inclusion is embedded in its vision of "South Africa [as] a free and just society inclusive of all persons with disabilities as equal citizens" (RSA 2016, 14), but this carries with it the associated danger that it "represent[s] an ideology rather than actual practice" (Brandt 2015, 37). Hansen (2012) cautions against a view of inclusion that is "limitless." However, many policies do not provide clear guidance on how to identify priority populations and to provide targeted and universal interventions (Pinto et al. 2012). Tools such as disability and inclusion based policy analysis are useful in "the process of bringing into view the sources of disadvantage that result in unequal opportunities and outcomes for people with disabilities, and using the tools of public policy to overcome these disadvantages" (IRIS 2012, 3).
14 Within the policies the assumption is made that children with disabilities and their families aspire to be part of "mainstream" society, and that this society is good and positive and enables them to thrive. This ignores all that is problematic in mainstream society-such as high rates of competitiveness, violence, and inequality (Silver 2015). Inclusion in this sense is not defined by choice, but as forced participation in "a hegemony dominated by the 'Old Masters' of societal norms, sensibilities and the marketisation of education" (Hodkinson 2012, 7). This reflects a broader trend of depoliticising or de-problematising "development" within disability policy. Marshall (2012, 58) explains, Demands for inclusion of disability on the mainstream development agenda overlook the nature of that agenda itself, which numerous other groups are campaigning to fundamentally transform. Development's role in producing disability is … largely unconsidered, with inclusion into the existing development agenda potentially leaving its ableist presuppositions and conceptual underpinning unchallenged.
In their representation of inclusion, the policies imply that the state has a major role to play in providing inclusive services. However, they fail to acknowledge that "[i]nclusion … invests power in those who are able to include and emasculates those who are included" (Hodkinson 2012, 6). The authors of this article argue that, used as a singular noun, inclusion is a control mechanism whereby (often unqualified) personnel decide who can or cannot be included. "Children … were reconceptualised as product, commoditised and scrutinised and held up to the professional's notion of societal good taste and educational performativity" (Hodkinson 2012, 7). Hodkinson suggests that inclusion can also be seen as "the noun of the Many," characterised by the mantra "all can be included." The ideology of inclusion is that structures and services should change to accommodate all children's needs, but in practice it is about children and families fitting into institutional procedures, in what Hodkinson (2012, 8) describes as "suffocating conformity."

Silences: What Is Left Unproblematic in these Representations of Disability and Inclusion?
Within the Disability and ECD policies, there is theoretical recognition of the rights of children with disabilities, and the social model of disability with its focus on the removal of barriers. Indeed, the Disability Policy states categorically that "the social model of disability remains the only policy stance of government" (RSA 2016, 45). However, many have been critical of the social model (see, for example, Haegele and Hodge 2016;Samaha 2007), arguing that it does not sufficiently take into account personal and unique experiences of disability. It tends to downplay the impact of the impairment on the individual child, reinforcing the perception that interventions should focus on societal aspects rather than individual requirements (Šiška and Habib 2013). Major limitations of the social model are not only that the personal experiences of a child (and their parent) in relation to their impairment are silenced and remain unacknowledged, it also does not take cognisance of differences between individual children with disabilities (Haegele and Hodge 2016). The social model thus needs to be complemented by an approach that takes these factors into account. For example, because negative attitudes often stem from a lack of knowledge about disability, it is important that parents have specific information about their own child's impairment before or as soon as possible after their birth. The ECD policy (RSA 2015, 45) acknowledges that such information is often not available to parents, while the Disability Policy advocates for removing the barriers to information and communication (RSA 2016, 55) without specifying how this applies to parents of young children.
While the ECD policy in its entirety applies to children below schoolgoing age, the Disability Policy refers to "persons with disabilities." Theoretically this includes people of all ages-children, young people, adults and the elderly-but in reality, unless there is a definite reference to other groups, there is the tendency for bias in favour of adults. This may result in limited attention being paid to the specific needs of children in comparison to those of adults.
As part of the recognition of the "compounded marginalisation" of persons with disabilities, the Disability Policy states that "persons with disabilities do not constitute a homogeneous group" (RSA 2016, 60). Although it therefore theoretically takes into account diversities between children with disabilities, it provides no strategies as to how these are to be (differently) addressed. Indeed, there is a trend in the ECD policy to refer to children with disabilities as a homogenous group, with similar needs, and a lack of acknowledgement of the impact of context in determining levels of disability. The simplification of disability glosses over the far-reaching implications of poverty and associated exclusion from opportunities and access to quality ECD services. The ECD policy in particular does not consider explicitly the complexity of disability, and is mainly silent in respect of socio-economic and cultural diversities, resulting in significant omissions in action taken for disability. As Brandt (2015, 25) observes, Disability can be of different nature-temporary or permanently, physically or mentally-and thus, the feasibility to remove barriers for PWD can be differently challenging.
There is an implicit assumption in the policies that surveillance systems and the identification of children with disabilities, while necessary (Philpott and McKenzie 2017), will lead to action to provide appropriate and accessible services. Unless plans, with associated budgets and resource allocations, as well as accountability, monitoring and evaluation systems are in place, the rollout of inclusive services will not be realised.
Within the policies, a binary is created between children with and without disabilities. Children with disabilities are implied to be a group distinct from those without disabilities, whereas in reality there is a continuum of developmental delays, impairments and contextual factors that determine the extent to which all children are 16 able to participate. While the Disability Policy is unequivocal in the call for all ECD services to be available for children with disabilities, the ECD policy defines inclusion narrowly, applying only to certain children (those with disabilities) and centres. It refers to the fact that a "sufficient number of ECD practitioners" (RSA 2015, 66) need to be trained for inclusion. This raises the question in respect of the ECD policy: is inclusion about all services, or only those used by or targeting children with disabilities? If the former, then all ECD practitioners need to be trained to cater for diversity. And what does it mean to design ECD services not just for children with disabilities, or without disabilities, but for all children? The policy is silent on this.
In these policies, exclusion and inclusion are portrayed as a binary, and perceived to have a mutually exclusive relationship, i.e. if you are not excluded, then you are included (Cobigo et al. 2012). They are silent on what is in between these poles, thus categorising children into one group or the other. In this view, "exclusion is the otherness of inclusion which means that inclusion presupposes exclusion. If we didn't exclude we didn't need to include … Inclusion and exclusion are two connected and interdependent processes" (Hansen 2012, 94).

What Effects Are Produced by These Representations of Disability and Inclusion?
Conceptualising inclusion as the opposite of exclusion leads the policies to primarily focus on the removal of barriers to inclusion, instead of on the factors that contribute to inclusion. The emphasis is on what must not be there, instead of what must be there.
One of the dangers of the focus on barriers is that it may downplay the importance of critical support services required by children with disabilities, such as habilitation and rehabilitation services, aimed at increasing their independence. The ECD policy must be commended on citing the need for community based rehabilitation (CBR) as part of ECD services towards inclusion and support for children with disabilities (RSA 2015, 67), but there is no reference to CBR workers in respect of training and they are not named as personnel providing ECD services. This feeds into the perception that children with disabilities need separate "special" services, which are currently in very scarce supply.
Within the policies, the focus on barriers and changes that need to be made to the environment may result in disabled children and their families being portrayed as passive recipients of services and there is a danger in this that they may "fade into the background" (Brandt 2015, 26). Brandt goes so far as to say that a focus on barriers may lead to divisions in society-as those who are not disabled are accused of creating disability, for example through the erection of buildings without ramps or lifts. This is an example of what Bacchi refers to as "dividing practices" (Bacchi 2009, 16). The effect of this is to undermine opportunities for participation and a sense of belonging of children with disabilities and their families, which are key elements of inclusion.
Use of the concept of inclusion to imply that those on the margins must be brought into the "mainstream" conveys "the acceptance and achievement of the dominant societal values and lifestyle which may lead to moralistic judgments if people reject or cannot achieve the dominant norms" (Cobigo et al. 2012, 79). Instead, there needs to be a greater emphasis on social capital, with recognition of and value placed on "reciprocity and trust among members of a group … to conceptualise social inclusion as a mutual exchange" (Cobigo et al. 2012, 79). Such an approach acknowledges and is respectful of the expectations, choices and needs of individual children and their parents.
Because inclusion is so broadly defined in the Disability Policy, it is difficult to measure whether service providers are successful in achieving it. Defining inclusion in a way that is "limitless" has deleterious effects because it is "at risk of being mere ideological rhetoric, an aspiration that lacks targets and required definitions and thus fail[s] to develop effective strategies" (Cobigo et al. 2012, 78). For this reason, an understanding of what inclusive practices actually mean is critical in guiding policymakers and service providers in the design of efficient services as well as in measuring outcomes of such efforts.
It needs to be borne in mind that the intention of the WPR methodology is not to uncover the "truth" behind policy making and policy development but to interrogate and make visible the assumptions or presuppositions and silences embedded in the policy text (Lindsköld 2010). WPR draws our attention to the view that policies are never neutral and value-free. Thus, norms, values and taken-for-granted assumptions need to be questioned and challenged in reflexive ways by policy researchers and social actors at various levels.

Recommendations for Policy Implementation
Having effective policy provisions for disability and inclusion is only half the story, of course, whether in South Africa or elsewhere. The key element of any policy's success is dependent upon its implementation.
In the World Report on Disability (WHO and World Bank Group 2011), the International Classification of Functioning, Disability and Health (ICF) is used to integrate the biomedical and social paradigms into the concept of disability, thereby combining personal and social aspects as part of the support provided to adults and children with disabilities. Barriers are seen as being avoidable, with the intention being to change the situation for children with disabilities through their removal. Thus, "It is not about putting an end to disability per se, but rather to reduce or remove the disadvantages associated with it" (Brandt 2015, 20). If disability is defined in line with the bio-psychosocial model, then rehabilitation for individuals becomes important. If we are to tackle barriers and inequalities within society, it is necessary to address both aspects, viz. the social determinants of disability and the individual circumstances of disabled children. 18 There is a pressing need to move beyond theoretical discourse to identify tools to improve social inclusion. Such tools are well documented in the literature; for example, according to Philpott and McKenzie (2017), there are three broad interlinking dimensions to an inclusive society for children with disabilities, viz. awareness and visibility, reaching out and welcoming children with disabilities and their caregivers, and provision of specific interventions and supports. It is critical that inclusion be based on a developmental approach, in which it increases with expanded opportunities to interact with others and participate in activities.
There needs to be a major shift as part of policy implementation towards developing a vision and understanding of disability inclusion that is based on the recognition of individual identity, such that disabled children and their families have the choice of how, where and when inclusion happens, and whether they want to be included at all. This requires that they are an integral part of shaping what inclusion means in practice.

Conclusion
The WPR approach to critical policy analysis employed in this study exemplifies "problem questioning" as a mode of thinking about policy. The questions were the following: What are the embedded presuppositions or assumptions in the representation of the "policy problem" by policymakers with respect to the Disability Policy and the ECD policy? How have particular representations come about? What are the silences in these representations and what is left unproblematic? What are the consequences of particular representations? And can we think about the problem in different ways?
The analysis of the two key policies shaping services for young children with disabilities in South Africa challenges views that policy formulation is neutral, value-free and technical in nature, and opens up the debate for alternate ways of thinking about policy as political, as argued by Marshall (2012). It is hoped that insights into the WPR approach provided in this article challenge other researchers engaged with different policy texts and contexts to question the "problems" embedded in policy proposals and to dig deeper in their meaning-making. Critical to this process is the practice of reflexivity, which necessitates that researchers interrogate their own positionalities and context, taken-for-granted assumptions and agendas inherent in policy proposals, no matter how enlightened they may appear at surface level. The WPR approach is valuable not only to researchers but also to individuals involved in policy implementation and development to ensure that they are not complicit in entrenching oppressive, dominant discourses. Bletsas (2012) suggests that policy problems and their representations are never fixed and stable and must be studied as they are being constituted in context, as there may even be competing problem representations in a particular context. Further, government and its policymakers are not the only players who should be part of the policy analysis process. Bacchi (2009) asserts that there are other actors who may include professionals, researchers, donor organisations, non-government organisations (NGOs) and civil society. The focus of the present study has been on two policies, but 19 one cannot be sure of how particular representations in policy and various policy actors at different levels have actually produced and shaped practice in the ECD sector and the lives of young children with disabilities and their families in varied contexts. The WPR methodology has the potential to uncover power relations that may play out in context. This is an important area for further research.
There is often the assumption that disability inclusion is a good and necessary responsibility of government. However, subjecting to scrutiny representations of disability and inclusion in the ECD and Disability policies in South Africa is based on the notion that "no concept or category is value-free or uncontested" (Bacchi 2009, 32), providing an opportunity for questioning, probing and problematising these familiar concepts. The WPR approach enabled the interrogation of two social policies, and raised critical questions with which the state and civil society alike must grapple, such as the following: Do we see inclusion as a means of accommodating certain children who are different from the norm or as the transformation of the ECD system so that all children are welcomed and accommodated? How do we take into account the agency of children and parents in the inclusive project? In line with Bacchi's stance, the intention here is not to provide answers and solve the "problems" of disability and inclusion, but to stimulate the development of platforms of struggle on which engagement can take place.