Inclusion and participation in peer relations

The aim of the study was to compare interaction and relations between schoolchildren with intellectual disabilities and the same children’s interaction/ relations with non-disabled peers. Study group: 24 children, 7 16 years, with intellectual disabilities. The children spent significantly more time in active participation in horizontal relation to peers with similar functional prerequisites as compared to active participation in vertical relations to children without intellectual disabilities. Together with age-mates with similar functional prerequisites each one contributes to common activity to about the same extent, they have similar everyday experiences and create a peer culture based on acceptance, shared norms and interests. Together with children with different functional prerequisites their influence on interaction conditions becomes extremely limited, and marginalization increases. Peer relations are subjectively experienced as deeply problematic; self-assessments as to peer relations are low-rated. Social exchanges in their own peer culture seem to be increasingly important. Data are discussed in terms of school inclusion, participation in peer-culture and self-perception.


School inclusion and participation in peer relations
Children with disabilities have, as a matter of course, recognized fundamental human rights as formulated in international conventions and declarations. In Article 23 of the UN Convention on the Rights of the Child this is stated in the following way: 'A mentally or physically, disabled child should enjoy a full and decent life in conditions that ensure dignity, promote self-reliance and facilitate the child's active participation in the community'. Further, these children shall enjoy possibilities to 'learn life and social development skills to facilitate their full and equal participation in education and as members of the community' (Article 24, UN Convention on the Rights of Persons with Disabilities). The implication is that children with disabilities should be included in the community, and school inclusion is an important aspect of this principle. School inclusion is not to be conceived of as 'an aspect of education or a policy or set of policies for education but a principled way of viewing the development of education and society' (Ainscow, Booth, and Dyson 2006, 296).
Based on these principles, three different motives for inclusion could be identified: inclusion as a valued civil position per se, i.e., an aspect of citizenship; inclusion for promotion of learning and development; and inclusion for facilitation of full and equal participation.
Participation in school life Á and in the peer relations that develop in the social arenas connected to school Á is a multidimensional construct based on different elements and conceptualized in different ways by different authors (Molin 2004;Granlund and Bjö rk-Å kesson 2005;Janson 2007;Egilson and Traustadottir 2009). Individual as well as environmental factors have to be taken into consideration when assessing and evaluating the phenomenon. In the present article participation in peerinteraction will be discussed with concepts elaborated in a recent review of research on peer relations and functional diversity (Janson, Nordströ m, and Thunstam 2007). In this conception an objective and a subjective dimension is distinguished, and these two are further elaborated into three objective (affiliation, accessibility, interaction/ co-activity) and three subjective (acceptance, engagement, autonomy) aspects. Subjective aspects refer to experiences of the child in consideration (engagement, autonomy), for example a child with a disability label, but also to experiences of the peers which constitute this child's social environment (acceptance). An inclusive school, aiming at full participation for its students, has to take objective as well as subjective aspects into consideration.
Physical affiliation and physical presence is a basic prerequisite, a necessary but not sufficient condition for participation in peer relations. A further prerequisite is accessibility, which is a matter not only of physical, but also of socio-communicative (mastering of communicative techniques and adherence to norms of social conduct) and symbolic (meaning-related), properties of activity. Participation is a function of the degree to which an actor has access to, and masters, processes characteristic of these contexts (Janson 2001). Interactive co-activity, i.e., doing things together (Corsaro 1997), when not directed by adults, is a central pursuit in the peer group. Interaction is considered a series of interchanges over a limited span of time, observable and of relatively short duration between two or more persons whose actions are mutually dependent on each other (Hinde 1997). There is a significant relation between interaction and participation (Almqvist and Granlund 2005). An interactionistic perspective is also constitutive for a situational and relational concept of disability, dominating in Nordic discourses and policy-making (Gustavsson, Tøssebro, and Traustadottir 2005). While affiliation and accessibility are considered as prerequisites for participation, observable interactive co-activity indicates participation as a matter of fact. Still, however, it remains to understand the lived quality of such an observation. This is where the subjective aspects come into consideration.
Acceptance mirrors how the presence of the focused child is perceived in the peer group: with recognition and appreciation or as disturbing factor in creating an independent peer cultural arena (Corsaro 1997). Such acceptance is crucial for creating activity conditions which promote or hinder participation. In research such acceptance/non-acceptance is mainly assessed with sociometry or attitude measurement (Janson, Nordströ m, and Thunstam 2007). Engagement, in ICF-CY (World Health Organization 2004) considered the central aspect of participation, stands in the present discourse for the focused child's own subjective experiences of her/ his presence in the community of peers. Autonomy, finally, refers to the child's possibilities to act according to own choice. In sum, full peer participation is considered to be at hand when a child, formally affiliated to the institution in question, is free to enter into Á or sustain Á engaged and accepted co-activity on an arena, accessible in the extended sense discussed above (Janson, Nordströ m, Thunstam, 2007).

Studies of peer relations in connection to functional diversity
To a certain extent, research demonstrates social and developmental gains for children with disabilities, not least more severe disabilities, in inclusive settings compared to more segregated settings. They engage in more peer interaction and more advanced play, gain more in social skills, are less lonely and depressed, receive more social support and develop more long-lasting relations with non-disabled peers than comparable children in self-contained specialized settings (Hunt et al. 1994;Kennedy, Shukla, and Fryxell 1997;Odom et al. 2004). Students with physical disabilities in regular classes tend to evaluate school in very positive terms, a fact which may depend on a very limited social network outside school (Hemmingsson 2002). Parents, in general very concerned about peer relations, judge inclusion to improve such relations (Odom et al. 2004), and to increase acceptance in so called 'typically developing children' (Diamond and Innes 2001).
However, inclusion is also confronted with challenges. Compared to non-disabled peers, children with diagnosed disabilities are less engaged in interaction, have lower sociometric status, and few, if any, close friends (Yude, Goodman, and McConachie 1998;Nakken and Pijl 2002;Diamond and Innes 2001;Nowicki and Sandieson 2002;Skär 2002;Odom et al. 2004). Irrespective of type, the frequency of interaction between functionally different children is very low (Kavale and Forness 1996;Vaughn, Elbaum, and Boardman 2001;Guralnick 2001;Kemp and Carter 2002), and Á though longitudinal studies are rare Á there are indications of increased social exclusion with increasing age (Ramsey 1995;Skär and Tamm 2002;Eriksson and Granlund 2004).
Attitudes towards children with disabilities tend to be less positive than towards other children, creating 'social barriers' important to influence in the inclusive school (Pivik, McComas, and Laflamme 2002). However, information and increased knowledge about disabilities does not necessarily enhance attitudes, and positive attitudes do not unequivocally predict increased interaction (Roberts and Smith 1999;Beck et al. 2002). Children with disabilities are usually accepted, but not in the way children without disabilities are, not as play-mates (Nowicki and Sandieson 2002;Nordströ m 2002;Skär 2002). They are to a certain degree met with respect and consideration, but experienced as persons in need of support, rather than equal peers.
Research on subjective experiences of participation, close to the engagement factor, is mostly reported either as self-assessments of social self-perception (selfconcept scales) or as interview-based descriptions of peer experiences. One review (Vaughan, Elbaum, and Boardman 2001) and one meta-analysis (Elbaum 2002) of self-assessments in students with learning difficulties demonstrate, with few exceptions, no significant differences between children in different educational forms. 'With regard to self-concept no placement is clearly preferable to another for all students with learning disabilities' (Elbaum 2002, 224). Scandinavian interview studies give detailed information about the problematic self-perception among children with disabilities when comparing themselves with non-disabled peers. As 'significant others' (Mead 1934) peers are important socialization agents in the child's social and identity development. Children with disabilities will be assigned different positions in horizontal relations to peers with similar prerequisites and in vertical relations to non-disabled peers, contributing to an identity formation and self-image as, respectively, 'normal ' and 'deviant'. Szö nyi (2005) found that students, perceiving special school affiliation as a threat to self-image, formed storylines in order to enhance self-esteem. Two approaches were identified: 1) normalization of the special school Á students describe the special school in a way which emphasizes the similarities with ordinary schools; and 2) normalization of self Á the student's personal difficulties are described in neutral terms, like 'a little slow' or 'in need of a little more support'. Affiliation to special schools was sometimes described as temporary, the interviewee not identifying her/himself with the collective 'special school pupils' (Szö nyi 2005).

Peer culture and participation as social process
School, like any other social institution, is a cultural arena constituted by at least two differently structured main cultures Á an adult-directed teaching culture and a childdirected peer-culture. As contrasted to merely physical space, such a cultural arena is in the present study conceived of as a social system, founded in a collective consciousness of shared meaning and values, rooted in common experiences and expressed in shared routines and systems of communication. In line with this, children's peer culture is defined as 'a stable set of activities, or routines, artefacts, values, and concerns that children produce and share in interaction with peers' (Corsaro 1997, 95). While adult-child relations in school basically are vertically structured, with the adult in a dominant and the child in subordinated and dependent position, and oriented towards distant developmental and learning goals, child-child relations are characterized by relative horizontality and equality in social status, power and influence (Bukowski, Newcomb, and Hartup 1996) and oriented towards interaction, pleasure and togetherness per se. Also characteristic of children's peer culture is its striving for autonomy and independence (Corsaro 1997), important to note as a principal problem in educational attempts to influence and intervene in children's peer cultures (Janson 2008). While membership in adult governed cultures can be characterized as ascribed, membership in peer cultures must be achieved (Frønes 1995).
However, relations between children are not always horizontal. Frequently experienced instances of the contrary, as summarized in preceding paragraphs, is interaction between functionally different children, where so called typically developing children tend to adopt dominating caring positions, if they at all recognize the presence of the functionally disabled peer. Conditions for participation in joint activity is, as said, dependent on access to the activity arena, understanding of its underlying meaning structure and mastery of explicit and tacit rules for exchange and communication on that arena. Differences in these prerequisites will induce differences and tensions in social interaction and decreased collective ability to solve the social tasks set forth in activity. In children's peer interaction, such tasks Á like peer group entries, play negotiations and conflict solutions Á are oriented towards pleasure here-and-now. Disturbances will be experienced as barriers to such goals and legitimate marginalization of the actor who is perceived as the source of such disturbance. When, on the other hand, activity demands and interacting partners' expectations and functional abilities are in accordance, participation is promoted. For children with functional impairments Á cognitive, hearing, language, motor or visual Á activity demands may create specific barriers to such participation. For instance, having access to the physical arena without mastering the system of communication used on it, or having communicative and cognitive competence but restricted sensory information about what is going on, will easily force the child into a subordinated position (Janson 1996;Nordströ m 2002;Skär 2002;Brunnberg 2003). Participation is promoted or blocked in the encounter between individual prerequisites and environmental demands (Janson 2004;Egilson and Traustadottir 2009).

Aim and research questions
Functional differences obviously result in low and gradually decreased levels of peer interactions and relations. In activities with peers with similar disability the interacting partners act on an arena with similar access to necessary situational information and communicative strategies, and the interacting partners reciprocally contribute to ongoing activities. The focus of this study is interaction between schoolchildren with intellectual disabilities, as compared to these children's interaction with non-disabled peers. The aim is to contribute to understanding of functional similarity, or dissimilarity, as conditions in creating patterns of interaction. The following research questions will be addressed: . What characterizes interaction between schoolchildren with similar cognitive limitations, as compared to interaction between children with cognitive differences, in quantitative and qualitative terms: frequency, participation patterns and communicative processes? . What characterizes the social arenas on which, and the activities in which, these two types of encounters take place, in terms of accessibility demands? . How do children with intellectual disabilities experience interaction and participation in these two types of relations? . What may be concluded, in the case studied, concerning the normative base for inclusion, especially considering the aim of full and equal participation?

Participants and procedure
This study was conducted in a Swedish town of about 120,000 inhabitants. From two different schools, 24 children with intellectual disabilities and their teachers, 11 girls and 13 boys, parents and leisure home staff participated. Eighteen children at primary and lower secondary levels attended special classes at one school. Six youths at higher secondary level attended one of three special classes at the other school. The chronological age range of the study group was 7Á16 years. The developmental age span was 4Á12 years. The developmental age was earlier assessed with Wechsler Intelligence Scale for Children (WISC), third edition, Swedish version. Children with multiple disabilities such as intellectual disabilities and autism, hearing or sight impairments, deep or moderate intellectual disabilities were not included in the study sample. Two children did not accept to participate. The size of classes was five to nine children. The classes were located next to corresponding regular classes, and schoolyards were the same for all children.
After school, seven children individually attended different youth leisure homes located in the neighbourhood close to each child's home. Staffs as well as other children were well-known from preschool years. The leisure centres enrolled between 30 and 40 children, but the number present at the same time was normally 10 to 15. No one, except for the target children, had any known disabilities, i.e., there was no occasion for interaction with children with the same functional limitations.

Data collection
Data were collected through observations and interviews. Observations at schoolyards were carried out during nine months from September to May in breaks when most children were present in the schoolyard. Also observations at youth leisure homes were carried out during free activities, mostly outdoor activities. Each child was observed on three to four occasions. Two independent observers focused on the same child at the same time and observations were continuously recorded on portable recorders. As guiding principles for the observations, the social behaviour scale ISB, Individual Social Behavior (White and Watts 1973;Guralnick and Groom 1988) was used in a modified version (Janson 1996). The original scale consists of 12 categories. In the revised version seven of these were used together with two supplemented categories, relevant for this study: active peer-related participation, passive peerrelated participation, activities alone, adult-related activity, leads positively/neutral/ negatively, follows verbal lead of peer, follows peer's activity (joins), refuses peer's lead, expresses positive/negative effect.
Observations in schoolyards of the six youths at upper secondary level were not carried out. Data regarding this group are based on interviews.
Semi-structured interviews with the children were mostly carried out immediately after finished observations, and mostly before transcription and categorization of the observations. They were interviewed on one or two occasions. Issues for child interviews were the peer situation at home, at school and, if relevant, at the leisure centre; also, with whom they play, how often, type of activities, initiative, best friend. For parent interviews, mother or father or both could participate. During staff interviews, the teacher or teacher together with class assistant participated. All interviews except two were recorded and transcribed. From these two interviews notes were written. Issues of adult interviews were with whom the child plays, number of peers and their age, how often they meet, character of interaction and relation, initiative, influence over activities, the child's own reactions and experiences of the peer situation, ease or difficulties in establishing and maintaining peer relations.

Data analyses
Observations were qualitatively and quantitatively analyzed. After transcription they were categorized with respect to social character and social direction using the categories in ISB, Individual Social Behavior (White and Watts 1973). Statistical correlation between the assessments of the two observers was not calculated. Any difference was discussed to consensus. The observation schemes were summarized for each child with regard to time spent in active peer-related activities, passive peerrelated activities, adult-related activities and activities alone, i.e., four of the nine categories. The schemes were also summarized for two age groups: pupils at primary level and secondary level respectively. The remaining five categories were not quantitatively analyzed but included in the qualitative analysis. For the testing of differences between age-groups and conditions the Wilcoxon Signed Ranks Test and Wann Whitney Test (SPSS for Windows, version 15.0) were used.
Qualitative data from interviews and observations were analyzed through content analysis (Lincoln and Guba 1985). They give complementary and more detailed information to the quantitative analysis of observations.

Ethical consideration
The Swedish Research Council's research ethics principles in the humanistic-social scientific field were followed. To parents and teachers participation was clearly formulated as being voluntary, with the right to withdraw at any time without giving reason. They were informed as to how data were to be handled and reported and that confidentiality would be preserved. Children's consent was handled by their parents. The teacher at first approached the parents in order to let them decide about participation for their children and themselves, and to ask whether the researcher could contact them.

Result
Participation in peer activities of children with disabilities varies both quantitatively and qualitatively in the two types of group (Table 1).
The children spend significantly more time in active participation with peers with same functional prerequisites as compared to active participation with children without disabilities. Interaction with functionally different peers was focused also on schoolyard conditions but was so infrequent that percent of time was zero or close to zero. Passive participation takes up the same amount of time (10%) in schoolyards and in playgrounds at youth leisure centres but may be differently characterized. Together with peers with the same disability, passive participation is often a prelude to active participation. After a period of observation the child enters into a state of reciprocal exchange. Together with non-disabled children this practically never happens.
Research results demonstrate that solitary activity and social withdrawal increase with increasing age for children with disabilities in their relations to children without disabilities. From the result in this study, Table 2 shows, however, that the disabled children's solitary and adult-related activities decreases when the play-arena affords interaction possibilities with children with similar functional prerequisites. And the active participation in peer activities increases significantly.

Interaction between children with similar functional prerequisites
The social character of the active interaction between children with similar functional prerequisites is for the most part an interaction on equal conditions. For instance, a child takes the initiative and settles play conditions, next time another child does. However, some children who have better prerequisites for organizing and directing activities occasionally get a leading position just as among children without disabilities. There are also examples of developed best-friend relations. Equality and reciprocity are also reflected in the directness and immediateness of the children's communication with laughter, chatter and smiling, an obvious pleasure in acting together. This equality and reciprocity also promotes active social exchange. Children with intellectual disabilities have characteristics, qualities and skills which influence the child's interaction with other children. The study demonstrates that their interaction is characterized by short dialogues, or monologues, often without any answer being expected, and weak directness in activities as regards initiatives. One aspect of the community in their own group is the way they manage everyday problems in the group. A central concept for characterizing equal and horizontal relations is 'negotiations' (Bukowski, Newcomb, and Hartup 1996;Frønes 1995;Corsaro 1997). Having similar access to socio-communicative contexts (Janson 2001) they negotiate on their own developmental level what to do and how. Sometimes no verbal negation takes place. Somebody initiates, and the rest follows as long as they like. Conflict solving is problematic. They handle it in their own way. When they cannot get help from an adult they solve the problem, for instance through going separate ways and after a while returning to each other. Another jointly-formed condition is taking part in the activities based on one's own prerequisites. This means, for instance, that anybody, if tired or uninterested, can leave a joint activity such as playing football, and after a while be welcomed back. This openness in the activities, the breaks in interaction, is characteristic and mostly accepted. These internal norms for their community become more and more observable as active participation increases from primary level to lower secondary level.

Interaction between functionally different children
In relation to children without disability, a vertical unequal interaction pattern develops. The child with disability enters the interaction in a subordinated dependent position. He or she does not understand the function of rules and routines, communication is too abstract, speed too high or the number of children to interact with at the same time too great. The children with prerequisites to direct and activity set up the activity conditions as well, what to do and how. The less influential child can join in as long as he/she adjusts to conditions, as long as there is no-one else to act with, or as long as the child manages demands and expectations. But the participation can also have other more concrete prices Á offering sweets or other attractive things or playing the buffoon. In the vertical relations, functional differences may lead to two results: Non-disabled children exclude the child with disability from participation, not necessarily an active rejection, but as a result of disinterest. Or the disabled child chooses a marginalized position as consequence of too-big demands from the environment. The big share of acting-alone time, adultrelated time and passive participation is an illustrative consequence of this.
Interviews with the disabled children themselves indicate explicitly that the number of non-disabled peers is extremely limited at home, at school and at leisure time irrespective of age. 'No, they are together with each other, only' or 'Nobody to play with at home' were typical comments. Some of them avoided continuing the dialogue, or got extremely silent, thoughtful or embarrassed: 'Earlier it was easier'. At home they stay indoors looking at TV. The requirements for age-mates are small: 'I want just one friend, not many'. Usually this 'only one friend' is some years younger and meetings are irregular and sporadic.
The lack of interaction between the functionally different children was confirmed by parents, teachers and assistants. They mean that the child with disability 'is accepted as he or she is, but not as a peer per se'. In the same way as children without disabilities, disabled children want to spend more and more time together with agemates, at the same time as they encounter increasing difficulties both to maintaining established peer relations and to establishing new ones. Their experience is that nondisabled children prefer peers without disabilities. Some parents notice two dividing points, the first at the age of four to five years, next at the age of 11Á12 years. Gradually the functional differences become so big that no common activity takes place, especially not in team sports and in activities with strong communicative elements.

Discussion and conclusions
Three normative motives for inclusion were identified in the background to this article: inclusion as a matter of citizenship, a democratic right for its own sake, inclusion for promotion of learning and development, and inclusion as a prerequisite for full and equal participation. The aim of the present study was to elucidate the third of these motives in the case of school inclusion and with special focus on participation in peer relations, recognized as 'the most challenges to student participation' (Egilson and Traustadottir 2009, 56). Inclusion in the participatory sense will be discussed in relation to the other two motives, citizenship and promotion of learning and development.
As introduced in the background, central aspects of participation are accessibility, co-activity, engagement and acceptance. These aspects are studied and compared on two different arenas of participation for children with intellectual disabilities, aged 7Á16 years: activities in the free time (breaks and meals) during school days in regular schools with integrated special classes, and activities in the afternoon in leisure homes, where all other participants are without diagnosed or recognized disabilities. What characterizes participation on these contrasting arenas and how is it experienced by children with intellectual disabilities?
The result reveals that in schoolyards, where it is possible to relate to peers with as well as without disabilities, children with intellectual disabilities interact only with each other. Half of the observed time they are actively engaged in such interaction, participation on equal terms, marked by horizontality, reciprocity and joy. Their similar functional prerequisites give similar access to physical, socio-communicative and symbolic contexts (Janson 2001), which promotes the co-construction of a joint field of action and eases reciprocity in contributing to what is going on in this field. The play process with entries and play contributions, approaches, invitations and responses, as well as other everyday social tasks, such as problem solving and conflict management, are handled in their own way Á they co-construct a peer culture on their own, with its shared norms, routines and communication system (Corsaro 1997).
At leisure centres and at home there are no peers, age-mates, with similar functional prerequisites. In these contacts with fellows without disabilities, the interaction pattern is mostly characterized by verticality and subordinated positions for the child with disability, or passive participation, like being a spectator or a listener. Their influence on interaction conditions is very limited, resulting in marginalization, direct exclusion or withdrawal. Interaction sessions are extremely few and limited in scope and time. This marginalization may be understood as a result of the surrounding children's dissatisfaction and non-acceptance of the disabled child's involvement, level of performance and contribution to activity. For functionally different children, with their activity limitations, the 'activities, routines, artefacts, values and concerns' (Corsaro 1997) of typically developing children's peer culture may create barriers for participation, to difficult to pass. Activity disturbances like disruption, rule breaking or deviation from expected role behaviour, rather than individual characteristics, is what usually give rise to negative peer reactions (Coie, Dodge, and Kupfersmidt 1990;Ytterhus 2000;Dyson 2005), implying that level of performance, and functional and social competence are the conditions for acceptance or rejection (Dyson 2005). An alternative to active exclusion seems to be vertically subordinated participation.
Children with disabilities are more likely to be accepted if they do not demand active influence in activities but accept passive or subordinate positions, 'act friendlysubmissive' (Siperstein and Leffert 1997, 397), just as an 11-year-old boy in the study who acted as ball boy in other children's football-playing. This boy developed his special strategy to master two critical factors in social participation: entry and maintenance (Garvey 1984;Corsaro 1997). He observed what was going on, created an independent contribution, entered the activity field and was accepted Á though in a very marginal role. The boy's description of the situation was, however: 'I like playing football with them'. Seen from one angle, this partial participation seems to give the boy satisfaction and contribute to his self-esteem. From another, this satisfaction might be seen as an indication of a very low level of aspiration as to participation in activities together with non-disabled peers, a situation also described by Skär (2002). Children with disabilities tend to describe themselves as active participators even if observations reveal that they are hardly more than spectators.
Problematic peer-situations for children with intellectual disabilities occur in both inclusive and specialized self-contained school-forms. The issue of school inclusion has to consider children's social participation on two quite different school arenas, the peer-directed peer culture and the adult-directed teaching culture respectively. In teaching and other adult-directed situations adult authority has the mandate to adapt requirements and other situational factors to match the varying capabilities in a functionally and developmentally diverse group of children, thereby promoting social participation for all. In the classroom, values like equity, equal rights and accepting attitudes as norms for human relations are jointly explored by students and teachers. However, on arenas where children's peer culture rules, matters are more complicated and functional diversity tends to lead to marginalization and exclusion. When it concerns acceptance of diversity, attitudes may predict intention to act in a certain direction, but not actual social behaviour (Roberts and Smith 1999;Dyson 2005). There seems to be a contradiction Á though not necessarily experienced as such by the individual child Á between the rhetoric of equity and accepting attitudes in the classroom and relative rejection on the schoolyard. Adult interventions in order to support interaction in the latter case may prove counter-productive, threatening the peer culture's strive for autonomy in relation to adult authority. Studies of peer cultures demonstrate how defensive processes may counteract such attempts (Corsaro 1997).
An initial reaction to specialized educational settings among children in need of support, coming from marginalizing settings, may be relief and experiences of enhanced social participation. But the other side of this coin is marginalization from the more general peer community. School attendance in a special group implies difficulties in maintaining potential relations, horizontal or vertical, to children in the wider school context, in the neighbourhood and at home. The special school is a school of both participation and exclusion (Molin 2004;Szönyi 2005).
Considering the aim of promoting social participation, educational inclusion of children with disabilities seems to face a normative dilemma: the value of the presence in the greater general peer community, with potential, though not actualized, full membership in its culture versus the value of a smaller, in certain ways segregated, peer community with full participation. Tetler (2000) differentiates between two states of tensions. On one hand genuine incompatibilities (dichotomies) excluding productive syntheses and demanding valuation and choice between either/ or, on the other hand contradictions, described as less rigid and representing competing rather than incompatible alternatives. Contradiction is not to be solved but 'might be a constructive instrument for reflecting and acting' (Tetler 2000, 260, author's translation) and reorientation. The above-mentioned dilemma might be looked upon as such a contradiction, to be handled not as a choice between either/or but as an instance of co-existence Á which also may demand different plans of action in different situations. '[F]lexibility may well be an important common characteristic of high-quality inclusive programs' (Schwartz et al. 2002, 14).
Another issue, which could be seen as a parallel dilemma, constantly present for children and youth with disabilities in school, is the issue of normality versus deviance, being like others or being special. Self-image is a central theme in disabled children's storytelling (Tetler 2000;Brunnberg 2003;Szö nyi 2005). Special classes or school affiliation may be a threat to self-image, but being excluded by fellow children in a regular school class is no less so. The experience of having functional resources which correspond to expectancies in the social surrounding, of being accepted as an interacting partner, and of joint activity on equal conditions, all this contributes to a special but shared perspective (Gustavsson 1998) and to a self-image of competence and sense of belongingness.
Referring to the three normative aspects of inclusion it could be said that the first Á the citizenship right of access to public arenas Á favours inclusive placement, as at least formally providing opportunities for building a community between all children irrespective of differences. The third aspect Á full and equal participation Á is not realized for children with disabilities in the culture of non-disabled peers. Differences in accessibility to situational information and in communication strategies necessary for performing actual activity result in partial or passive participation in a marginalized position as spectator or listener, or no interaction at all. In interaction with functionally equal peers, however, conditions for full, active and equal participation are more favourable. The second aspect Á learning and development Á is promoted in both inclusive and more segregated placements but in different ways. Children with disabilities might in the vertical relations benefit from socially and communicatively competent models among typically developing children. The interaction in horizontal relations with functionally similar peers promotes development on another basis, learning through meeting and exchange of perspective (Frønes 1995). Here the basis for developmental gain is the children's equality in relation to each other, an equality in power and influence. They have about the same influence over interaction terms. To establish and maintain important peer relations they have to reciprocally take into consideration the wishes and opinions of each other. On their own developmental level they have to learn to negotiate common ideas about what to do and how. These demands, for establishing and maintaining important peer relations, promote the development of social and communicative competence Á take another's perspective, interpret social cues, express oneself.
In any case, social interaction is a fundamental condition for learning and development. As marginalization in, or even withdrawal from, relations to nondisabled children seems to increase with age, social exchange within peer cultures, created by children with activity limitations themselves, seems to be increasingly important. However, proper value and esteem is not always attached to membership in such cultures by the surrounding society. A leading policy document concerning children with disabilities from the Swedish National Agency for Education in 2005 said: 'Identification with other [children] who in specific respects show similar difficulties . . . seems to be a driving force that partially counteracts the striving for inclusion . . . [risking] that students give up and accept an exclusion which may characterize their future ambitions and thereby their basis for participation in society' (cited in Janson, Nordströ m, and Thunstam 2007, 56, author's translation). The consequences of fellowship between children and youths with disabilities are obviously considered a problem, in contrast to their own, parents' and teachers' perspectives, where these relations are considered important and highly valued.
In the light of this, what should be strived for in a general school culture with the ambition to combine the value of access to the 'general' peer culture (if it is at all possible to talk about such a thing) with the value of belongingness to a specific community with optimal conditions for membership? What is needed is an allembracing school culture which recognizes the membership and celebrates the presence of all, but which, in doing this, creates room for a diversity of cultures, mirroring a diversity of functional and social prerequisites and experiences. What should be strived for is not primarily interaction/co-activity in every situation but respect and dialogue between such cultures. And that, of course, also implies belongingness and access to the school context in general, its physical environment, its learning opportunities and caring responsibilities, in short to all cultural resources and social standards of this most powerful social agency. Flexibility, as referred to above, in combination with individualization, is a cardinal characteristic for administrative proceedings in a school, acting accordingly. Part of this, and in line with the UN Convention on the Rights of the Child, are provisions which respect the child's right to be heard in all matters that concerns her or his being and existence. Choice of school placement is definitively such an issue.