From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden

This article argues that current iterations of solutions for preventing school segregation are constrained by an overreliance on particular representations of justice, in which the other is perceived as the responsible other. Studying the grounds for a decision to close a suburb school in Sweden, this article engages partly in an analysis on what implicit conception of justice that manifests itself, partly in exploring a conception of justice open towards a multiple and openended spatiality. It is argued that in order to imagine and construct such a spatiality, a majoritarian approach to justice must be abandoned in favour of a minoritarian one. Doing so, justice further needs to abandon a distribution of blame and responsibility and instead seek to pluralize forces flowing through different spatialities. A minoritarian approach to justice, I argue, can be envisioned by applying the concept of segmentarity as developed by Gilles Deleuze and Félix Guattari.


Introduction
When the term began for the pupils in the schools of the medium-sized Swedish city of Örebro late summer 2018, one thing in particular was different. The lower secondary school in the ethnically diverse and socioeconomically exposed suburb Vivalla, which for years had struggled with poor school results, had been shut down by the municipality. The decision to shut down the school had been preceded by a commissioned proposal from the Management of Preschool and School ('the Decision Basis') 2 in which it was reporting grounds for a possible solution that would resolve the problems at the school. The solution was to contribute to provide all pupils in the municipality with the same opportunities to attain the goals of schooling, and to increase the integration in the schools operated by the municipality. The proposal that came to be accepted suggested that the pupils in the seventh and eighth grades be moved to other schools in the municipality from autumn 2017, while pupils in the ninth grade would finish their last year at the school. 3 Hence, the lower secondary school (grades 7-9) in Vivalla was completely shut down in autumn 2018.
Solutions such as the above-mentioned to resolve problems connected to decreasing school results at certain schools, are today becoming increasingly accepted in political discourses in Sweden. Between 2006 and September 2017, at least 12 schools were shut down due to decreasing school results (Bering 2017), and in a survey on efforts by principals of schools to prevent school segregation made by the Swedish National Agency for Education (2018), this solution was one of the most commonly raised. The solution has also been commented in affirmative by leaders of political parties such as the Liberals and the Green Party (Lundberg Andersson 2018).
Shutting down a school in a city district forces pupils to leave their residential district; it compels them to cross boundaries, and it may have the result of connecting them with other parts of a city as well as with other schools. However, shutting down a school in an already exposed urban area is also restraining for the area and the inhabitants of that area. Furthermore, it is limiting and disrupting for the pupils changing school; they are not seldom confronted with a new social context in which they are defined as deviating or differing, forcing them to embody predominant images of their residential area with assigned identities as a result (Kallstenius 2010;Ambrose 2016). In the case of Vivalla, this is a confrontation the pupils at the closing school have not opted for.
The decision to close a school therefore actualizes questions such as the following: despite the negative consequences connected to shutting down a 2 Decision basis by the Management of Preschool and School in Örebro municipality (Förvaltningen förskola och skola i Örebro kommun) (2016), ref. no. 361/2016. 3 The decision to close the school was taken in the Minutes of the Primary School Board in Örebro municipality (Grundskolenämnden i Örebro kommun) (2016), ref. no. 758/2016. the Decision Basis 4 for closing the Vivalla school represents the Vivalla district and school, the dominant conception of justice in it, and the merger between representations of the Vivalla district and school and the conception of justice into the proposition of the Decision Basis. The third and final part explores what representations of the Vivalla district and school are left out from the Decision Basis, and what conception of justice would have been needed in order to include them. My aim is to contribute in the strivings for conceptualising a justice that manages to include just representations of space, groups, communities and societies. 5

Opening up a School Market: Proximity Principle and School Choice
During the past three decades, the Swedish school system has changed from being one of the most publicly dominated and unitary school systems in the world to a system of more or less total freedom of choice between public and independent schools, with both public and independent (privately run) schools being publicly financed (Blomqvist and Rothstein 2008;Trumberg 2011). This transition, described inter alia by Whitty, Power and Halpin (1998) and Trumberg (2011), was accomplished primarily via three major reforms, 6 which were driven by ideological ideas of human liberty and efficient public management (Neoliberalism and New Public Management). 7 The first reform entailed a devolution of financial and managerial control to local levels and a gradual change from management by rule to performance management. It also implied a transfer of economic responsibility from the state to local municipalities and their taxpayers. 8 The second entailed a consolidation of power over the shape and performance of the education 4 I.e. the Decision basis by the Management of Preschool and School in Örebro municipality (Förvaltningen förskola och skola i Örebro kommun) (2016), ref. no. 361/2016, referred to in note 2. 5 In this effort I am of course not alone. Despite the fact that any attempt of referencing the important contributions on this topic must be completely incomplete, I would, in the field of spatial justice, like to mention in particular Harvey (2009b), Lefebvre (1996), Mitchell (2003), Philippopoulos-Mihalopoulos (2015) and Soja (2010). 6 For which the overall features were set out in Swedish Government Official Reports (1988:20) and Government bill (1988/89:4). 7 Regarding the Swedish school system's openness towards neoliberal reforms' emphasis on an individualistic and marketized system, see Wedin (2017) who traces precursors to such reforms in post-war education policies' endeavours to democratize the education system. 8 Prepared in Government bill (1989/90:41) and Committee on Education (1989/90:UbU9).

David Jivegård
From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden 5 system at the central governmental level. 9 The third gave parents a right to choose which comprehensive school for their children and introduced a school voucher system, resulting in that each pupil has 'a bag of money' that follows her when changing school. 10 As a consequence, the schools in Sweden today (with few exceptions) 11 are administrated by the municipality organiser and organisers of the independent schools in tandem. 12 The municipality organiser is responsible for deciding the size of the school voucher and to administrate it, to build and run municipal schools, and to make sure the municipal schools meet the requirements of the state. 13 The organisers of the independent schools are responsible for their schools meeting the requirements of the state. 14 For what is relevant here, the school voucher consists of a basic amount that covers teaching, teaching tools, pupil health service, food, administration and facility costs. 15 The facility costs can also be paid for by their actual costs by the municipal organiser. In addition to this, the municipal organiser has a responsibility to distribute resources in relation to the children's and pupils' different situations and needs. 16 Therefore, the basic amount is often divided into one part that is the same for each pupil and one part that differs in relation to the pupils' structural conditions. Moreover, the above-mentioned reforms entail that today there are two parallel basic rules regulating which municipal school a pupil will be assigned to: a proximity principle ensuring a right to a school placement near the pupil's place of residence and a rule -referred to as the 'school choice rule' -giving parents the right to choose which school their child will attend. 17 If the number of applying pupils to a particular school exceeds the number of places at the school, the municipal school organiser has the responsibility to seek to find the best solution for all involved pupils, considering the route to school and the best interest of the child. 18 Independent schools are exempted from the proximity principle. This means that they can only accept applying pupils. 19 If the number of applying pupils exceeds the number of places at the school, the independent school can choose between several selection criteria. 20 The most common selection criterion is queuing time, with sibling priority.
With an almost unrestricted possibility for school choice, Sweden stands out in a Nordic context. In Norway and Finland, implementation of school choice is optional for local authorities; however, in Norway, the legal right to attend the local school is left intact (Haugen 2020;Imsen, Blossing, and Moos 2017), and, in Finland, 97 percent of the schools are public and no school voucher system exist (Poikolainen 2012). In Denmark, an extensive possibility for school choice exists, but the state only covers 75-85 percent of the costs, the remainder is covered with tuition fees (Schindler Rangvid 2008). Hence, Sweden differs from the other Nordic countries by allowing nation-wide a possibility to choose a school other than the one closest to the place of residence, with the full costs covered by the municipality.
Research on the proximity principle's and the school choice rule's effect on school segregation in Sweden is ambiguous. Almgren and Lindbom (2007) e.g. seem to claim that school choice is one of the most promising integration instruments, whereas Bunar and Kallstenius (2007) and Francia (2011) present results pointing to a tendency of children with Swedish ethnicity to move to other schools when children with immigrant origin use their school choice to move to middle-class schools. For sure, studies show that schools located in areas that are characterised by unemployment, ethnic plurality and large inequality have problems with outflows of pupils (see e.g. Blomqvist and Rothstein 2008). Studies and reports further show a correlation between areas characterised by housing segregation and problems with school segregation in those areas. Difficulties often connected to such schools tend to lead to parents choosing for their children more popular schools, characterised by a local community with high education level and low unemployment, leading to a movement away from socio-economically poor areas of the city. 21 And due to the way in which the funding system is constructed -a school voucher connected to each pupil -an outflow of pupils without an equivalent inflow will in many cases lead to declining resources (Blomqvist and Rothstein 2008), which risks creating a vicious circle: an outflow of pupils creating a situation of declining resources, and declining resources creating a situation of further outflows of pupils. But whether it is the proximity principle or the school choice rule that give these effects is not ascertained.
In this article, I do not engage in a discussion on whether the school choice rule or the proximity principle is to blame for school segregation, rather the opposite: I argue that it is not possible to see the effects from the one regulation separated from the other. As I further discuss below, the school choice rule and the proximity principle are intricately connected, and the effects of this connectedness have large implications for the pupil compositions in the schools. This should not be understood such that the proximity principle, the school choice rule, and the 1990's reforms, are the only reasons for school segregation. However, the proximity principle and the school choice rule constitute a system that by acting close to pupil bodies decide the shape of the pupil compositions in schools. This system induces and augments other structuring and segregating forces in society, which means that they need to be understood, so to say, through this material system. It is this system and its induction that is of my primary interest. How justice is conceptualised has implications for how this system and its induction can be conceived and what solutions to school segregation are perceivable.
Before turning to these issues, some words need to be said about the method by which I approach them.

The Cartography: Large-Scale Segmentations, Small-Scale Movements
My overall argument in the article is that conceptions of justice and representations of space cannot be understood in isolation from each other. Instead, they are highly connected. Consider e.g. the prevalent principle for distributing tenancy apartments in Sweden, which is queuing time. At first glance it might seem objectively fair or just, but when adding a spatial layer 21 Swedish National Agency for Education 1996Education , 2004 and investigating how this principle structures and segments bodies in certain, very specific, ways in the city, by forming socioeconomically demarcated concentric circles from the city-centre to the outskirts of the city, the principle does not necessarily feel that fair anymore. The same argument can be made with housing cooperative units, 22 although the primary structuring component now being income or capital.
If now considering the entanglement of these and other systems with pupil compositions at schools, we soon realise that questions of justness or fairness in the systems have far-reaching consequences for the ordering of pupil compositions at schools, and of urban space in general. So how we conceptualise justice affects how space is structured and ordered. But the opposite is also true: how we perceive or represent space affects how justice can be conceptualised. And moreover, how we conceptualise justice affects how we are able to represent space. Thus, by making visible what perceptions or representations of space are prevalent in certain decision processes, I claim, it is possible to draw conclusions on what implicit conception of justice is present.
Using the case of the closing of a suburb school actualises ethical questions. Is it e.g. legitimate to use a problematic and serious situation in an excluded community for a theoretical discussion on justice? Although the question cannot be answered with full certainty, my weighted assessment on the question is that it is, and also that it in some cases is important that we as critical scholars do so. The reasons for such legitimacy in the present case can be summarised as follows. The material I use does not conform to the Vivalla school, the pupils at the school or the population in the Vivalla district; it conforms to how the Vivalla school, the pupils at the school and the Vivalla population are represented as subjects and entities. This means that I analyse how the case of the Vivalla school come to be represented by prevailing discourses. In the present case, by discourses engaged in the questions of segregation and school segregation. These discourses are not produced by the people in Vivalla, but by the majoritarian (I conceptualise the majoritarian below). My overall claim in the article is that a majoritarian justice entails a certain production of representations of space, groups, and communities. Given this overall claim, it is paramount that a majoritarian justice is exposed to and intervened and confronted by the representations itself produces, especially such representations that are not part of the lived reality of a socalled majority in society (since that is where the problematic effects of a majoritarian justice will be most present). My analysis can thus be said to constitute a critical perspective on the majoritarian production of the excluded other, which is the same as imposing the majoritarian responsibility for its own production. With regards to how the material is treated, this means that when I, e.g., present statistics on 'groups' and city districts, it is not to be regarded as objective facts about the state of things, but as material for analysing the production of representations of the excluded other.
To structure the material and to enable a translatability between representations of space and conceptualisations of justice, I employ a deleuzeoguattarian cartography. The first to be said about my use of cartography as a method is that a map never is a mirror of what it portrays; it is not an adequate imitation nor a transparent reflection of a stable territory (Bosteels 1998, 147). This may seem obvious, but it has far-reaching consequences for the relation between 'reality' and the representation of 'reality'. If the map does not mirror reality it needs to be recognized that it rather produces territories, borders, and boundaries; it produces reality. If now turning to cognitive maps (i.e. mental representations of reality), also these do not represent reality but rather produce it by marking out territories, borders, and boundaries. From a cartographic point of view, therefore, concepts are not objects, but territories (Deleuze and Guattari 1994, 101); they produce and mark out territories by which we live our lives. Cognitive mapping, therefore, is not second to the territories it brings forth; it is productive, it makes and unmakes the environment by setting up territories.
A cartographic method is therefore of assistance in answering the question why certain problems become identified and why certain solutions (and not others) attach themselves to them. Moreover, it politicises the question of which representations become prevailing, i.e. the question of just representations. Finally, it helps answering the question of the relation between concepts and representations; as for the present case if the way justice is (implicitly) conceptualised makes it necessary to represent certain territories (such as Vivalla and the Vivalla subject) as essentialising certain problems.
In the chapter '1933: Micropolitics and Segmentarity' in A Thousand Plateaus, Deleuze and Guattari (2005) explores a cartography for how a relational space can be conceived. The cartography makes a distinction between large-scale segments and small-scale movements and can be described as follows.
We are constantly segmented in different ways. It could be that we are segmented in dual oppositions such as man/woman, rich/poor, etc., which Deleuze and Guattari call binary segmentation. It could be that we are territorially segmented, such as household, neighbourhood, municipality, nation-state, etc., which Deleuze and Guattari call circular segmentation. It could be that we are segmented linearly: first child, then grown up, then old; or, in the family, in school, on the job (Deleuze and Guattari 2005, 208-9).
Depending upon the circumstances, these segmentations take different shapes; they can be more or less rigid, more or less supple (Deleuze and Guattari 2005, 210-2). But regardless of what shape they take the segmentations are not pre-given, instead they are produced in space and time. To conceptualise this changing character of segmentations, Deleuze and Guattari use the concepts of deterritorialisation and reterritorialisation. Deterritorialisation designates small-scale movements that escape segmentation, these are lines of flight, moments of destabilisation and change. Reterritorialisation designates the opposite movement, towards large-scale segmentation; it is the tendency towards ordering and stratification (Deleuze and Guattari 2005, 10;Bruncevic 2018, 24;Lalor 2015).
The cartography so far can thus be said to consist of segmentations, lines of flight that escape segmentation (deterritorialisation) and movements toward segmentation (reterritorialisation). At least two perspectives can here be seen. We could either say that already-rigid segments come first, and then seek to understand how lines of flight constantly escape them. Or we could say that lines of flight are primary, and then seek to understand how they become segmented. I use both these perspectives in the article. In the Decision Basis, it is clear that the first perspective is predominant. This perspective follows therefore, from my material. The second perspective is David Jivegård From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden used in the final, explorative, part of the article. I use it to break with the predominant way of representing space in the Decision Basis. However, it is important to note that regardless of what perspective we choose, we need to acknowledge that both segmentation and deterritorialisation occur. Thus, oscillating between lines of flight and already-rigid segments are supple segmentations that ties the social world together (Deleuze and Guattari 2005, 222). In the Decision Basis, the deterritorialisations are not accounted for; thus, it is unable to acknowledge supple segmentation.
Finally, Deleuze and Guattari (2005, 223) identifies a strategic dimension in which lines of flight are either coded and segmented, or decoded. Coding of lines of flight is actualised by an abstract overcoding machine that defines rigid segmentations. This abstract machine reproduces segments by opposing them two by two (binary segmentation), making all the power centres resonate (circular segmentation), and laying out a divisible, homogeneous space (linear segmentation). Decoding is actualised by an abstract machine of mutation, which decodes rigid segmentation and produce deterritorialising lines of flight. Thus, if rigid segments block lines of flight, the mutation machine makes them flow between the rigid segments. Hence, the two abstract machines constitute two different poles; one that is operating with large-scale phenomena and coding into segments, one that is operating with small-scale phenomena and decoding lines of flight.
Between these poles, a domain of negotiations exists. In this domain, large-scale segments are, at times, undermined by fissures and cracks, at other times, lines of flight are already drawn towards rigid segmentations (Deleuze and Guattari 2005, 223-4). The existence of a negotiation domain is why Deleuze and Guattari can claim that it is wrongly said that a society is defined by its contradictions: That is true only on the larger scale of things. From the viewpoint of [the smaller scale], a society is defined by its lines of flight […]. There is always something that flows or flees, that escapes the binary organisations, the resonance apparatus, and the overcoding machine. (Deleuze and Guattari 2005, 216) This is where we can start talking about the powers of the schools. If flows of pupils involve the mass of movements of pupils between areas of the city, what the schools govern is the conversion of these flows into segmented pupil compositions of the schools. By using already-rigid segments of the city they can position themselves in favourable flows and attain profitable pupil compositions. But by doing so, they also further rigidify the segmentations. In this sense, schools have, through the system for regulating the constructing of pupil compositions, become power centres that control the points where flows are converted into segments (see Deleuze and Guattari 2005, 226). But it is also important to note that they cannot control the flows as such. The lines of flight therefore define the power centres impotence field. For, power centres can only operate in the conversion point between lines of flight and segments (Deleuze and Guattari 2005, 217); thus, a social field is always animated by all kinds of movements of decoding and deterritorialisation affecting segments. These lines of flight are not contradicting the segments but escaping them (Deleuze and Guattari 2005, 220).
To conceptualise the negotiation field in which rigid segmentations and lines of flight meet, Deleuze and Guattari use the concept Body without Organs (BwO). It denotes the non-stratified or destratified plane on which lines of force and power operate directly on bodies. Legal theorist Merima Bruncevic (2018, 33) defines a BwO as that which is 'continually and constantly dismantling the totality, the assemblage, or the appearance of the totality.' On a BwO forces flow freely and thereby enable movements and flows. Segments as well as lines of flight appear on the BwO. Hence, the concept of BwO manages to explain how fluid segmentation appear in the oscillation between rigid segmentation and lines of flight (Deleuze and Guattari 2005, 149-66;2009, 4, 72, 181, 322-40;Colebrook 2002, 100-5;Braidotti 2011, 14). A BwO does not appear in isolation from everything else, instead it converges with other BwOs and lines of flight on a plane of consistency (Deleuze and Guattari 2005, 4). In this meaning, a plane of consistency is a kind of relational space, drawn up by the points and lines that it consists of.
For what is relevant for this article, the cartography can be summarised as follows. On a plane of consistency, BwOs appear as destratified planes on which both lines of flight and segmentations appear. Lines of flight are actualised by an abstract machine of mutation (deterritorialisation) and segmentations appear when lines are being reterritorialised by an overcoding abstract machine. The abstract machine of mutation operates David Jivegård From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden 13 with small-scale movements, whereas the overcoding abstract machine operates with large-scale segmentations.

Large-Scale Segmentations in the Decision Basis
The Decision Basis does not say much about the information it uses to come to its proposition. However, it says quite a bit about the perspective from which Vivalla is being evaluated. Hereby, it is possible to draw conclusions on the geometry it is using, the axioms that are needed for this geometry and, by extension, the way it represents space. This way it is also possible to draw some, at least preliminary, conclusions on how groups, communities and society-at-large are represented.
So how then are the spatial aspects of pupil compositions represented in the Decision Basis? First of all, it recognises that a system that allows a school choice risks resulting in highly segregated pupil compositions and that they, essentially, are governed by two factors: the composition of the residential area and the ability to make a 'good' choice between schools when navigating on the school market. 23 This recognition, as I discuss in the next part, is one key for including just representations in a solution, and it could have opened up a possibility in the Decision Basis for understanding how the problem is produced on systemic levels and how it is affected by a plurality of coding and segmenting forces operating in society. However, in the Decision Basis this recognition is far too abstract to be acted upon.
Instead, two assumption chains are introduced. The first assumption chain says that a segregated school market leads to poor equality, that poor equality is negative for school results, and that poor school results lead to poor opportunities in life. The second assumption chain says, in the opposite direction, that mixed pupil compositions lead to increased equality, that increased equality leads to improved school results, and that this, in the long run, leads to increased social cohesion. 24 The term 'equality' should here not be understood as alike, i.e. in the sense that equal schooling means same schooling. In Swedish the quote uses the term 'likvärdighet' which rather means equivalence. In the preparatory works it is distinguished from 'likformighet' (equability). 25 And by the Swedish National Agency for Education (2012,8) it is defined as that which requires that 'the quality of education [is] of such standards that the established goals can be achieved, regardless of where in the country the education is carried out'. In the Decision Basis the two assumption chains are followed by two statements. First, that the development for the strongest performing pupils in Sweden are stable more or less over time, whereas the development for the weakest performing pupils is deteriorating. Second, that statistically there are four distinguishable groups among the weakest performing pupils: 1. pupils with low-educated parents, 2. newly arrived migrated pupils, 3. pupils that are both living and going to school in areas with poor structural conditions, and, 4. boys. Moreover, these groups are said to tend to coincide. 26 Admittedly, the Decision Basis does not say much about the distribution of the four groups between the schools of Örebro. However, since they are part of how it represents the situation on an abstract level, a reasonable assumption is that the concrete level is understood in the same manner. Therefore, further down, I include statistics on some of these groups, in order to concretise and criticise the way in which the Decision Basis approaches the situation in Vivalla, i.e., how it represents the situation. But before doing that I need to know what are the spatial representations that are compared.

Concentricities and Binarisations in the Decision Basis
In the Decision Basis, it is possible to see a concentricity of the representations of spatial territories and a binarisation of the effects measured. I begin with the concentricity. Two circular segmentations are the focus of attention in the Decision Basis. The first could be denoted a Vivalla-segment, the other is a segment that corresponds to the rest of the municipal schools in the municipality (I will call this segment 'municipalitysegment'). Although the two segments belong to different scales, they are used for comparisons. The Vivalla-segment belongs to a city district scale, whereas the municipality-segment belongs to the municipality scale.

David Jivegård
From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden Therefore, when they are compared to one another a concentric relationship between the two are produced, a relationship between part and whole. With this I mean that the Vivalla-segment is perceived to be part of the municipality, but at the same time it is measured against the municipality-segment. This concentricity effectively creates a separation of forces by making visible only effects of the two segments, and making invisible relations, forces, and movements between and within them. This prepares the ground for binary segmentations.
In the Decision Basis, binary segmentations are visible e.g. when comparing the share of pupils in the Vivalla School-segment that has reached the goals of schooling in all subjects with the share of pupils in the municipality-segment that has reached these goals. In the Decision Basis this is presented in a graph (shown in Figure 1 below). The practice of using the normal, overall curve, for setting up the 'normal expectation' in the population, and against which different curves belonging to parts of the overall population can be measured and compared, has by Foucault (2007) been called security apparatuses. In the Decision Basis, this practice is performed by using the overall, pupil population, curve of reaching the goals of schooling as a norm for defining deviations in a specific part of the population (Vivalla). Hereby, the two segments are filled with properties constructed out of mutually constitutive differences (Deleuze and Guattari 2005, 210).
By representing the situation in Vivalla this way, the Decision Basis elaborates with a temporality in which the spatiality is closed. In other words, the dual opposition has already, via the two constructed representations of circular segments, been done. Therefore, we only see how the movements as oppositions repeat themselves through time. As a result, the graph in Figure 1 is only able to show how the problem-event manifests itself over time. Representing space in such a binary fashion creates an impression of a big gap between grades belonging to pupils in the Vivalla-segment and grades belonging to pupils in the municipality-segment. Moreover, it hinders us from seeing that the same mechanisms that produce poor grades in certain parts of a city also might be producing acceptable grades in other parts: e.g., uneven distribution of wealth, standard of living, etc. It hereby prevents an understanding of how the event is produced on systemic levels in space and time simultaneously, by forces affecting the event in the multiplicity with which it is entangled. This is due to the Decision Basis representing space from the perspective of large-scale segmentations; a perspective from which binary and circular segmentations are perceived to operate in tandem in a temporality in which space is closed and always-already divided into places and groups (segments). The spatial division and opposition have already been done and therefore we only see how these divisions and oppositions repeat themselves through time (the Vivalla-segment and the municipalitysegment, poor grades and acceptable grades).
The spatial opposition between a Vivalla-segment and a municipalitysegment is further visible in the other statistics presented in the Decision Basis: e.g. statistics showing how many pupils are actually attending the Vivalla School, how many of the pupils estimated to attend the Vivalla School have chosen to attend other schools, what schools they have chosen to attend, how many pupils have chosen the Vivalla School in relation to how many were projected to attend the school (measured by residential proximity to the school). All these, are ways of depicting the problem in which either space or time is held still, whereas the other (space or time) is understood as 'passing through' a stationary zone of representations (cf. Deleuze and Guattari 2005, 219).
To conclude, in the Decision Basis the Vivalla School is understood as a circular segment; but due to it also being understood as a binary segment, it appears as if it was necessarily separated from, yet enclosed by, the municipality-segment. Space is hereby represented in such a way that the Vivalla School is viewed on one side, and the rest of the schools in the municipality on the other, creating a situation where the different but interconnected entities that the schools constitute are regarded as two separate forms, as two rigid circular segments, through which binary segmentations pass. Therefore, in conjunction with the binary segmentation, the circular segmentation makes the interconnection between the schools -or rather, the intensive movements within the whole school segment -invisible; the flows of pupils between different schools as well as the forces affecting these flows are disguised by the way the problem is represented. The only flows visible are those going away from the Vivalla School and the almost non-existing flows going into the school. The Vivallasegment is hereby perceived as its own cause for attaining poor school results, and the municipality-segment is perceived as its own cause for attaining acceptable school results.

Properties of the Segments
The two circular segmentations, thus, define the spatial representation, whereas the binary segmentations, in the present case, define the temporal representation. Together, they disconnect the understanding from causal flows, processes and movements creating the situation, and instead constrain it to correlative representations of effects (how many have passing grades, what is an acceptable share of passing grades, how many choose this school, how many choose other schools, etc.). Rather than understood as subjected to affections, the segments are represented as filled with properties. Thus, the effects of various processes in society are represented as properties belonging to each segment.
Provisionally accepting the assumptions and statements of the Decision Basis, it is possible to concretise the representations that appear. Important to note, however, is that this does not mean that the concretised representations are statements on the state of things, nor that they are actual causes to the problems at the Vivalla School, only that they are represented as such. The most influential 'properties' having an impact on the situation in Vivalla are, in the Decision Basis, said to be the four groups (i) pupils with low-educated parents, (ii) newly arrived migrated pupils, (iii) pupils that are both living and going to school in areas with poor structural conditions, (iv) and boys. The concentration of 'properties' visible in the first three statements are produced by excluding and segregating processes, as well as the level of inequality, in society, which affects both the Vivallasegment and the municipality-segment (if nothing else: the higher the concentration of these 'properties' in the Vivalla-segment, all else unchanged, the lower the concentration in the municipality-segment).
Looking at the structural conditions of the Vivalla-segment, the Vivalla district is by the Swedish Police defined as a particularly exposed area, 27 statistics present it as a segregated area (Statistics Sweden 2018), and the income levels of the population are low, in relative terms. 28 Located in the midst of the Vivalla district, the Vivalla School had 220 pupils (2016) spread equally over all grades. Out of these pupils, 95 percent were either born abroad or had both parents born abroad, 10 percent had two parents with a post-secondary education, 34 percent of the ninth graders had passing grades in all subjects, and 16 percent were newly arrived migrants. 29 Comparing this to the municipality-segment, essential differences are 27 This means that the Police claim the area to be characterized by social problems and displaying a presence of criminality, which, allegedly, have led to a widespread aversion towards participating in legal proceedings and difficulties for the Police to fulfil its mission (National Operative Department of the Police 2017). A critical remark on this claim by the Police is that it is judging the results from its own activities. The report is therefore open for critique. It could e.g. be claimed to be attributing its own failures to causes beyond its control in order to escape responsibility (particularly exposed areas, not poor police work). 28 Vivalla has an average income of SEK 139,000/year. This can be compared with the population of the municipality (Vivalla included), which has an average income of SEK 281,000/year (Statistics gathered by email correspondence with the municipality of Örebro). 29 When talking about newly arrived migrants in the school context in Sweden, one usually refers to the definition in the Swedish Education Act, Chap. 3, Art. 12a. It is this definition of newly arrived that I use, i.e., someone that due to migration has begun his or her schooling after the age of seven and whose schooling in Sweden has not lasted more than four years.

David Jivegård
From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden visible: on an aggregated level, 29 percent of the pupils in the municipalitysegment (Vivalla included) were either born abroad or had both parents born abroad, 55 percent had two parents with a post-secondary education, 85 percent of the ninth graders had passing grades in all subjects, and (Vivalla excluded) 5 percent of the pupils were newly arrived migrants (Swedish National Agency for Education 2016/17a, 2016/17b, 2016/17c).
All of the characteristics that the statistics represent are of course produced by causal small-scale movements in society; but due to how space is represented in the Decision Basis, these movements are not included in the understanding of the situation. Instead, the result of the movements become attached to the segments as such in the form of measured effects, they become represented as properties belonging to each segment. Therefore, the Vivalla School becomes represented as an existing corporeal totality -with its building, its staff, and its pupils -consisting of a half full school and pupils attaining poor school results. The statistics hereby operates with closed spatialities, which represent the Vivalla district and school as causes in themselves. Such a view closes the school in on itself, it understands the Vivalla School and district as necessarily carrying certain properties, which restrain them from going into transformation processes.
The temporalisations and spatialisations effectuated in the Decision Basis are hereby intricately connected and dependent of each other. The temporalisation is dependent on the spatialisation carving out a territory in order to insert its causation (circular segmentation), but the spatialisation is just as dependent on the temporalisation inserting a beginning and end to the causation in order to carve out a territory (binary segmentation).
The result is a discursive understanding revolving around two forms, a Vivalla-segment and a municipality-segment, each consisting of perceived homogeneous space which creates the appearance of a necessary separateness between them. The binarisation between poor and acceptable school results hereby becomes attached to the two mutually constitutive circular segments. In the Decision Basis, there are no signs of a concrete understanding of the poor school results and the poor reputation of the school being affected by forces creating certain movements of pupils between different schools in patterns regulated by e.g. housing system, labour market, identity-norms, meaning-production, sense of belonging and processes of racialisation and exclusion (such as, e.g., the practical impossibility for newly arrived persons to be accepted at popular schools that use queuing time as selection criterion). Nor are there any signs of an understanding of these movements of pupils further creating movements of teachers, inhabitants in the city areas, etc. Hereby, the Decision Basis represents the problems at the Vivalla School as nothing but necessary causes of their own manifestations. This way, any understanding of a communicability between different segments are cut off and a solution inserted into the 'failing' segment from above becomes conceived as the only solution.

Majoritarian Justice
When the problem has been defined through an analysis that understands reality as fixed and rigid -consisting of perceived territories of homogeneous space -the burden that comes with the enforcement of a solution becomes easy to position: it can 'rightfully' be located in the segment where blame was put. Essential for this conception of justice is therefore a distribution of guilt: the Vivalla-segment has performed poorly and therefore has to take the consequences of its performance. For instance, when the problem gets defined like above, it would appear absurd to close a well-functioning school and redistribute its pupils over a couple of non-well-functioning schools, although the same objective might be reached. For that reason, this conception of justice is performed through three parallel processes: 1. the problem-description is concentrated through the construction of rigid circular segments (giving the problem a certain form, a kind of fictitious problem-bearer), 2. differences are converted into opposites (disconnecting them from the forces by which they are constituted), and, 3. a solution is constructed, for which the burden is carried by the same segment that analytically has been understood as the problembearer.
This conception of justice locks the ways it is possible to think of, and indeed understand, the problem. The representational practice that it needs suppose a replacement of space with conceptual segments. These segments are, further, perceived as predetermined in the sense that they evolve as a David Jivegård From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden 21 product of their solution, e.g. the Vivalla-segment evolve as the result of a binary opposition to a municipality-segment, where the latter define acceptable school results, and the former, as a consequence, come to embody poor results. But since the poor school results appear as a result of the municipality-segment defining acceptable school results, the representational thinking neither assumes correspondence between the municipality-segment and the Vivalla-segment, nor with the surrounding space. This means that the conception of justice replaces an understanding of the school results as affected by material forces in society with an understanding of the Vivalla-segment as carrying them as properties. Hereby, justice operates as a linear segmentation that underscores, rectifies, and homogenises the Vivalla-segment and the municipalitysegment as representations (cf. Deleuze and Guattari 2005, 211). In short, it subordinates the definition of the problem-event to the conception of justice (Deleuze and Guattari 2005, 212, 362). It is a conception of justice that partly makes segments their own unit of measure, partly it makes possible a translatability between them (correlative comparisons).
This conception of justice could also be seen in the way the varying part of the basic amount is calculated, which operates with propertied markers. Thus, it mitigates the effects of rigid segmentation, but it does not affect the processes and movements causing the segmentation as such. At best, it compensates for these processes; at worst, it justifies them. In addition to this, since it is attached to each pupil (the school voucher) and fail to make visible the processes causing the segmentation, high compensation tends to be perceived as unfair.
This conception of justice submits differences to 'pre-existing images'such as the Vivalla-segment and the municipality-segment. It, therefore, implies a standard for measurements; a standard that at the same time makes it possible to represent segments as their own unit of measure, and enables a translatability between them. I argue that this standard adheres to justice's need for distributing guilt. Guilt is the variable that decides how the segments are constructed and measured. In this regard, justice is, what Deleuze and Guattari call, majoritarian. When stating that justice is majoritarian, we should not understand majoritarian as some thing's or phenomena's quantity. For sure, majority and minority can be defining something's quantitative state, but in order to do that something must enable the quantitative measurement. This qualitative, defining, thing is what is meant by majoritarian. It is a qualitative definition enabling a quantitative measurement: the Vivalla-segment deviates because it is defined by the majority -i.e. the average pupil in the municipality-segment (Deleuze and Guattari 2005, 105;Deleuze 1995, 173). Only by defining it in relation to a majority can it be measured, translated into a denumerable set (Deleuze and Guattari 2005, 470). Majoritarian is that which locks the Vivalla-segment into a defining relationship with the municipality-segment. Thus, the municipality-segment is majority because it appears twice, once in the constant and again in the variable from which the constant is extracted (Deleuze and Guattari 2005, 105). Majoritarian justice is therefore that which appears twice, once as a variable in the abstract overcoding machine (majoritarian) and again as a constant extracted from this variable (majority/minority).
Guilt is what serves as this standard measure. The abstract scales of justice are built around guilt as a variable (conception of justice) and what they are measuring is guilt as a constant (represented space). In order to conform to this standard measure, a majoritarian justice needs to overcode space as rigid and fixed. A majoritarian standard, therefore, extorts representations of rigid binary and circular segments in order to create forms needed for attributing guilt: someone has to be represented as the responsible-other. Thus, majoritarian justice depends on a representation of space that is contained within an overcoding abstract machine. In order to make guilt translatable across different segments, majoritarian justice has to be, or be part of, an overcoding abstract machine. With this I mean that a majoritarian justice locks the representations of the segments to an overcoding abstract machine.
This means that if the problem were to be represented as open towards multiple and open-ended spacetimes, the causality needed for distributing guilt would not be possible to construct. What characterises a majoritarian justice is therefore a standard measure that restrains the possible representations of the problem-event. This standard measure disregards each segment's openness towards small-scale movements, and treats them instead as fixed essences, as large-scale segments; it understands segments as having properties rather than as affecting and being affected by their spatiotemporal surroundings. This is what I call a majoritarian justice: a justice that represents segments as always-already constituted and predetermined rather than as in continuous progress. Therefore, although segments depend on society also being characterised by small-scale movements (Deleuze and Guattari 2005, 213), a majoritarian justice overcodes space and turns the representation of it into rigid places and segments. A majoritarian justice is hereby dependent on making fixed and rigid what is necessarily also vague and fluent (cf. Deleuze and Guattari 2005, 212, 367).
This creates a situation where the Vivalla-segment and the people 'populating it' is foreclosed from acting as active agents. By forcing them into a majoritarian relation to space, they are conceptually denied communicability with 'society' -i.e., their multiple entanglement with a surrounding. The people 'populating' the Vivalla-segment are therefore assigned to the field of reactions; they are defined first after the municipality-segment as majority has been defined. The fictitious subject (the Vivalla-segment) that hereby is constructed, is the form which subordinates the pupils of the Vivalla School under the 'justness' of the decision. Hence, they are constituted as reactive objects, rather than as active agents. They have been constructed as a blameworthy fictitious subject, which through the insertion of justice needs to be made responsible -the 'just' decision to close the school is in this regard nothing but a mere consequence.
As I see it, this is othering in the extreme: the denial of any worthwhile agency. For the Vivalla-segment -or rather: the pupils 'inhabiting it' -is already from the moment a binary process of converting differences into opposites was actualised in the Decision Basis, assumed blameworthy and responsible. Against this blameworthy segment stands the municipalitysegment: the norm-producing, 'normal', majority. And since this segment is constituted in the same majoritarian process as the Vivalla-segment, it will have to turn the blame back on itself if it cannot attribute it to some other segment (cf. Deleuze 1986, 127-9). Thus, in order for the 'good' municipality-segment to remain, a blameworthy segment must also remain (cf. Bottomley 2004;Richardson 2005), which attributes a majoritarian justice with a concealed injustice. Therefore, it is not necessarily the decision in itself, but the process of majoritarian standardisation by which it is being produced, that attributes this conception of justice with a concealed injustice. This, by Richardson (2014) called 'epistemological injustice', by which propertied segments are represented as both their own cause and effect, is produced by an axiomatic that is dependent on representing space as consisting of rigid circular segments coinciding with (or collapsing into) rigid binary segments, cut-off from the forces constituting them. It is this epistemological injustice that enables the attribution of guilt and responsibility. I argue that this is the condition of a majoritarian justice: justice as the scalpel, cutting up homogenised space; justice as the separating of forces, representing affections as properties; justice as moralisation, attributing blame where 'wrong' forces are manifested. It is well concluded by Deleuze in Nietzsche and Philosophy (1986, 128): I who accuse you, it is for your own good; I love you in order that you will join me, until you are joined with me, until you yourself become a painful, sick, reactive being, a good being.

Small-Scale Movements the Decision Basis Fails to Consider
How then do we resist majoritarian justice in such a way that other representations of time and space -and, as a consequence, of human life, community, and society -become possible? Instead of overcoding time and space, how do we formulate conceptions of justice that operate with spacetimes? These are questions for this tentative part of the article, in which I conduct an analysis on the relationship between the proximity principle and the school choice rule. My aim is to outline some elements of a minoritarian justice -i.e., a conception of justice that opens up the representation of space and acknowledges its relationality.
A good starting point is to recognise that the analysis cannot begin in a spatial understanding where segments are always-already constructed, instead it is necessary to start in the process of becoming, in the lines of flight and the constant production of segmentations. Thus, in order to reach beyond representations of a rigid, always-already segmented space, we need to add a further spatial layer, a layer that is able to conceptually include small-scale movements, i.e., lines of flight. Deleuze and Guattari's concept of BwO provides the tool for doing just that, for locating the analysis to the intersections of lines of force and power operating directly on bodies (Deleuze and Guattari 2005;Deleuze 1997;Bogard 1998;Lalor 2015).

The School-BwO
From the viewpoint of lines of flight, the proximity principle and the school choice rule have fundamentally different potential effects with regards to residence. The proximity principle implies a concentric relationship to the home, which reterritorialise pupils in relation to their residences. It thereby contributes in reinforcing circular residential segmentations. However, in relation to binary segmentations the proximity principle may either be deterritorialising or reterritorialising. Whether it takes the one or the other expression depends, inter alia, on how relations between residential areas and catchment areas are constituted. If e.g. a catchment area of a school cuts through several differently segmented residential areas, encounters that transcend binary segments along class or ethnic lines of differentiation may be effectuated, leading to deterritorialisations. However, by the school choice possibly effectuating deterritorialisations of residential areas by enabling reterritorialisation to class-or ethnic/race-segments, this effect might be taken away. With this I mean that a residential area may consist of a mix of social 'groups' (e.g. different classes and ethnicities); with only the proximity principle this mixture would propagate into the pupil composition of the school in the residential area. However, when also adding the school choice rule to the analysis, it is obvious that the mixture risks being erased by social 'groups' using the school choice in conform ways. Thus, the effects of the proximity principle are virtual in the sense that they will come about under certain conditions (cf. Deleuze and Guattari 2009, 358f;Deleuze 1997). However, whether one impact or the other will have the power to express itself as an actuality depends on, inter alia, the state of affairs between the proximity principle and the school choice rule in their practice as well as between small-scale movements and large-scale segmentations in general. Hence, if encounters between different segmentations will have the power to express themselves as actualities depend on factors that transcend those segmentations.
The school choice rule has a deterritorialising effect on the pupil composition from the perspective of circular residential segments. In this sense, the school choice rule is more complex than the proximity principle. It connects each school with every other school by functioning as an indifferent inductor enabling both flows of desire and corporeal flows of pupils between schools. The enabling of flows of desire can be understood through Hirschman's (1970) distinction between voice and exit. Voice denotes organisation members' attempts of repairing or improving already existing relationships, whereas exit denotes their withdrawal from the relationship. When connecting schools and thereby enabling the possibility of changing social setting depending on the circumstances, the possibility for exit increases. Taking into consideration that a disgruntled individual primarily has to choose between exit and voice, we could also add that when exit increases, voice decreases. Hence, the introduction of a school choice enables flows of desire in the sense that it directs the attention towards exit rather than voice. 30 The school choice rule thus makes of schools, their composition of the pupils and the movements between different schools, a BwO on which these flows can effuse. Segmentations with regards to ethnicity, culture, class, religion, gender, etc., are induced as organisers on this school-BwO. So are state views on expressions connected to such segmentations and thereby also career paths and future prospects laid out differently for different persons. Other BwOs and systems are also connected to this BwO.
Hereby, the relation and interplay between segments and lines of flight consist of a mixture of two different processes. Taken separately the two processes form two extreme poles. In the one pole only reterritorialisations on already existing segments (such as class-segments or ethnic segments) are produced, which would entail a solely reactive force since the lines of flight are fully coded by, and thereby act only in relation to, the segmentations and the overcoding abstract machine actualising them. In the other pole only deterritorialisations that disrupt and complicate the way resonance is formed by segmented parts of a city in conjunction are produced; this would entail an active force since the lines of flight act solely in their own right, as decoded lines of flight. The relation should of course not be understood according to the one or the other extreme pole, but rather as a mixed relation. It is thus more accurate to describe the relationship as movements towards the one or the other pole, and in questions of degree. Movements towards the former pole could be called segmentational assimilation and increases when the reactive forces and the influence of the overcoding abstract machine increase. Movements toward 30 On this production, although they do not conceptualise it in the same manner, see Ambrose (2016) andBunar (2010).

David Jivegård
From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden 27 the latter pole may be effectuated by deterritorialisations creating new segments or by them enabling passages to the limits of already existing ones (cf. Deleuze and Guattari 2005, 4).
On the school-BwO, zones of intensities connected to the residential, segmented, composition of the city are augmented by the proximity principle. Thus, the schools, through the proximity principle, induce their milieu as organisers, which gives rise to segmentations and reterritorialisations. The effect of this co-production by the school choice rule and the proximity principle, is intensive corporeal flows of pupils around the city that deterritorialise and reterritorialise segments of different kinds. The school-BwO does hereby not form a totalisable whole. Instead, disparate segments transform themselves into, cross over and affect one another. Looking e.g. at the Decision Basis, in 2016, it is possible to see a strong flow away from the Vivalla School, bifurcating to at least 14 other schools. Moreover, it is possible to see that more than 50 percent of the pupils that due to the proximity principle were expected to attend the school chose other schools. 31 And from the adjacent districts only four -of the projected 103 -pupils chose the Vivalla School. Hence, in certain parts of a school-BwO the movements, flows, and thus ravages, will be more intense than in other parts. Vivalla seems, of what can be told from the Decision Basis, to be such a part.
However, by only looking at the proximity principle and the school choice rule, it is not possible to see how flows move, and why they move the way they do, nor for that matter by only looking at the effect of the proximity principle and the school choice rule or at the pupils in a certain area of a city. This is due to the school-BwO being open towards and communicating with other segments and BwOs on a plane of consistency (cf. Deleuze and Guattari 2005, 156). Instead, therefore, we need to investigate how the proximity principle, in the present case, is connected to different segments and BwOs, and how certain movements, through the school choice, are 31 The school year preceding the closing decision, the number of pupils attending the school was down to 219, whereas the projected number was set to 518. By dividing the number of opt-out pupils (518-219 = 299) with the number of projected pupils the result well exceeds 50 percent (≈ 58 percent). The calculation however relies on the assumption that the projected number of pupils equals the number of pupils that would have attended the school according to the proximity principle. However, any decrease in the number of projected pupils adhering to the proximity principle would increase the share of opt-out pupils that otherwise would have attended the Vivalla school according to the proximity principle. produced within these connections. We need to look at what is being induced through the proximity principle and the school choice rule.
Here an examination of other segmentations and lines of flight would be needed. There is unfortunately not room for that here, but I will give some brief examples. The first example is the Swedish apartment stock. This consists, by and large, of two types of apartments: tenancy apartments and units in housing cooperatives. The rent for tenancy apartments is set by the use-value for the tenant, whereas the price for buying a share in a housing cooperative is set to the market value. These two different modes for regulating housing costs contribute in segmenting the demography in a city by constituting a housing-BwO on which lines of flight are produced, organised and coded. The school-BwO converges, inter alia, with the housing-BwO which therefore should be part of an analysis of a school-BwO. Another example is the curriculums in the various school subjects, gearing the teaching in certain directions thus deciding which pupils and what kinships will be rewarded. Further examples are the racialisation of the labour-market, giving rise to segmentations both of families and of individual pupils by deciding who will be affected by inertness and who will move more freely on the labour-market; as well as the level of welfare commitment, which will distribute material standards in different ways (Who has a room of one's own to study in? Who has parents with time and ability to help out with homework?). I give these examples only to point at the importance of keeping in mind that the school-BwO is not a closed totality but always open towards that which has the power to affect it.
For that reason, the Vivalla School should not be regarded only as a closed totality, it should also be recognised that this 'totality' is caught up in states of becomings. The continuously changing composition of pupils and teachers, the increasing and decreasing amounts of pupils, the sealing and opening of rooms in the school building depending on these amounts, the buses with pupils moving around the city, the kinships between children that are formed in relation to the choices of schools, all these are examples of becoming actualised by the intensive, small-scale movements on the school-BwO. These becomings correspond to lines of flight that effuses on the school-BwO, designating regions, thresholds and zones of intensity all over it. In order to understand this openness, this multiplicity, time and space must be conceived as spacetimes: it has to be recognised that causality David Jivegård From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden spreads out as effects in these varying spacetimes, which prevent causality from being located to specific segments (see e.g. Massey 1994, 5). The pupil composition at the Vivalla School is, thus, caught up in states of becomings that depend on the expression of the school-BwO. And the school-BwO, in turn, is dependent on a plane of consistency on which it, together with other BwOs, appear. It would therefore be incorrect to regard the Vivalla School as the cause of its own problems. Instead, the 'cause' of the problems needs to be regarded as spread out, displaced, decentralised: the effects that manifest themselves as problems need to be understood as effects of other effects -there is no original cause that can be localised to a certain place or segment.
Hence, when pupils use their school choice they pass over thresholds, they pass from one segment to another -in the case of Vivalla, most of them by taking buses to schools located in the central/southern, wealthier, parts of the city, between 5 and 12 kilometres away from their homes -(likely) in order to become something else, to become someone else. Whether this appears as deterritorialisations or reterritorialisations depends on how rigid the city's segmentations are. If uses of school choice appear as reterritorialisations, they are reactive to rigid segmentations and an overcoding abstract machine, it is then possible to talk of segmentational assimilation; if they appear as deterritorialisations, they are active forces working directly as decoded lines of flight.
But the crossing of segmentation thresholds not only produce something in relation to the new school, by deterritorialising the Vivalla-segment they also change the productions and affections in the Vivalla-segment (cf. Deleuze and Guattari 2005, 3). Hence, lines of flight produce ravages on at least two places at once on a BwO: the place they deterritorialise (e.g. the Vivalla-segment) and the place they reterritorialise (e.g. some other schoolsegment). By doing this they effectuate new lines of flight (Deleuze and Guattari 2009, 85). When patterns start to evolve on who will use the school choice to attend what school, small-scale movements of lines of flight begin forming mass phenomena, which means that school-segments take on more rigid, large-scale, characters. This way continuous flows of pupils convert into segments of the schools, and in this point of conversion works as the power centres of rigid circular segmentations (Deleuze and Guattari 2005, 226). This affects what encounters with other school-segments will be possible and what kinships or peer networks will have the power to express themselves within a certain school. Thus, in highly stratified, claustrophobic, school-BwOs the ability to perform as individual pupil differ considerably. 32 Put in other terms and returning to the alleged aims of the Decision Basis, whether the proximity principle and the school choice rule are successful integration tools can analytically be described as whether they, to a lesser or larger extent, constitute a BwO that gives rise to active forces, to encounters, to lines of flight that do not reterritorialise on already existing segments. Hence, if the school-BwO gives rise to deterritorialisations that push existing segments to their limits; if pupils are continuously creating new segments in the multiplicity of the city; if regions on the BwO are constantly being transgressed in such a way that encounters are made possible, enforcing dynamic transformations of the school-segments, hereby challenging the static structure of rigid segments that attribute different school-segments with different expressions with regards to school results and reputation; then the proximity principle and the school choice rule can be considered successful integration tools. Of course, and again, this is not an either/or scenario but a continuum on which the school-BwO can be considered more or less integrating, more or less segmentationally assimilating.
Having analysed the case of the Vivalla School, it is clear that the proximity principle and the school choice rule contribute in creating reterritorialisations on already existing segments, differentiating social classes and ethnicity according to the scheme inertness and deceleration for the excluded, speed and acceleration for the included. Thus, although the Vivalla-segment may be said to not be performing in satisfying ways, it is a segment that is constituted by facing a school-BwO. On this school-BwO it forms a segment that contribute in producing lines of flight which give rise to certain reterritorialisations. This conjoins the circular segment of the Vivalla School with binary segments of class and ethnicity, implementing, with regards to school results, desire for certain movements between school segments, which is manifested as a kind of cream-skimming of the school, a 32 Studies e.g. show that peer networks have gained importance for the school performance of individual pupils (Bergsten 2010; Swedish National Agency for Education 2009).

David Jivegård
From Exploitation Through Justice Towards Exploiting Justice: Conceptions of Justice in the Closing of a Suburb School in Sweden kind of segmentational assimilation. 33 Due to this, the problem should be viewed as a common, decentralised yet central, problem that is most effectively addressed on the level of the BwO.

Towards a Conception of a Minoritarian Justice
What I have tried to show in the above is that movements of pupils appear as lines of flight that, on the one hand, can be coded by an overcoding abstract machine in relation to certain segmentations (e.g. institutionally stratified divisions of class and race/ethnicity), which means that the movements are controlled by reactive forces (in the former part I described how a justice that conforms to such reactive forces contributes in placing pupils in reactive relations, how they become disconnected from smallscale movements affecting them), but also, on the other hand, how movements of pupils just as well can appear as lines of flight that are decoded by an abstract mutation machine and that have the capacity to evolve through specific encounters. When the latter is the case, the movements constitute active forces, with the power to constantly push various segments to their limits, enabling small-scale movements that refuse to be coded by segmentations' overcoding abstract machine. The question here is, therefore, how to envision a justice that is able to conform to active forces and that thereby manages to turn pupils into active agents?
First of all, it is important to note that there are not only actual effects operating on the BwO, but also plenty of virtual effects that will come about under certain conditions. This means that the school choice rule and the proximity principle have the power to express themselves in such a way that various segmentations of the schools in the municipality of Örebro would increase their connections and their suppleness. Because what effects are actual and what effects stays virtual is a question of how the plane to which they belong is constituted. Following Deleuze and Guattari, the suppleness of the segments and the intensive movements on the BwO, will increase if the degree of active forces increases. A justice that conforms to a majoritarian standardisation restrains the possibility for such active forces, since it connects to the overcoding abstract machine. Justice would, instead, have to release itself from this majoritarian standardisation, it would need to be a becoming, a process (cf. Van Marle 2012). This conforms to Deleuze and Guattari's definition of the minoritarian. Minoritarian is everything that deviates from the majoritarian model (Deleuze and Guattari 2005, 105). A minoritarian justice engages itself, therefore, with precisely this, with transforming reactive forces into active forces and thereby bringing agency back to the pupils 'populating' the school-BwO, by deviating from the majoritarian model -it is, as Purdom (2000, 223) puts it, what makes 'the standard tremble'. The primary question for a minoritarian justice is, therefore, this: how could lines of flight be released from the grip rigid segmentations have on them?
Since lines of flight are processes of increasing expansion, connection and creation, the following kinds of further questions would be posed applying a minoritarian justice: how can the intensive movements on the school-BwO be increased, how can the connections with the Vivalla district be increased, how can the connections with other schools on the BwO, with other parts of the city, with various career paths, with universities, religious institutions, sports clubs, and so on, be increased? How can other schools' connections be increased in the same way? And how can this be done without overcoding the segments involved? Or, to summarise it all in one question: how could pluralities of lines of flight constantly be produced and connected with each other in constantly varying ways?
A minoritarian justice in the case of the Vivalla School would, therefore, be to turn reactive forces into active forces, it would be the responsibility to make possible lines of flight in the system regulating pupil composition. Justice would be to prevent rigid segmentation, and thus the overcoding abstract machine from coding the production in the system. The system for deciding pupil compositions would then be unjust if the power expressing itself in the system constitute reactive forces, if the forces are reactionary in relation to an already segmented city and governed by an overcoding abstract machine.
In Anti-Oedipus, Deleuze and Guattari states that art often uses lines of flight 'to short-circuit social production, and to interfere with the reproductive function of technical machines by introducing an element of dysfunction' (Deleuze and Guattari 2009, 31). Could this be the opening of justice towards active forces, justice as the short-circuiting of the coding into segments, justice as the introduction of an element of dysfunction? Perhaps reforming the grading system from absolute grades in relation to learning outcomes, to relative grades given proportionally in each class, would be enough of a dysfunction -perhaps on some school-BwOs? In any case, a minoritarian justice would be a justice that engages itself directly with the actual BwO and the lines of flight effusing on it. It would be a justice that promotes an understanding of the multiple. Such a justice would, paraphrasing Deleuze (1986, 50), need to change itself with the conditioned and determine itself in each case along with what it determines, it would need to be inseparable from the affections befalling it. In this regard, a minoritarian justice does not offer alternative sites of opposition. Instead, it is a reformulation of difference expressed not in opposition to a polarised other, but rather as a differentiated relationality to the whole.

Conclusions
The decision to close the Vivalla School is constrained by a conception of a majoritarian justice. Majoritarian justice can be defined as a conception of justice that represents space as closed and rigid. It depends on a majoritarian standardisation against which large-scale segmentations are measured. This dependence makes representations of small-scale movements difficult, if not impossible, to include in decision-making. It therefore represents the pupils as part of reactive forces. Majoritarian justice is performed when everything is already in place, when everything is already there.
A possibility for including small-scale movements in decision-making is feasible by applying a minoritarian justice. A minoritarian justice can be defined as a conception of justice that varies with the problem at hand, it operates on the level of the problem, adjusting itself in relation to how the problem unfolds; it resists the essentialisation of guilt to a certain represented segment or subject, and instead acknowledges the lines of flight that always permeates society. It thus recognises that we are all responsible for the actualisation of society. A minoritarian justice is therefore not about fitting space and time into its framework, but rather to reframe itself in relation to the spacetimes at hand. It is not a means for exploitation, but a means for exploiting justice.

Notes on contributor
David Jivegård is a PhD-candidate in legal theory. He works primarily with questions regarding the relations between legal concepts and equality, exclusion and oppression. His ongoing PhD-project is on the relation between conceptions of legal subjectivity and prohibitions on vagrancy and begging.