The development of inclusive learning relationships in mainstream settings: A multimodal perspective

Abstract The debate regarding the inclusion of children with Special Educational Needs and Disabilities (SEND) in mainstream education in the UK partly revolves around what makes the classroom environment inclusive. Through the potential offered by the specific qualitative methodologies employed, this study aimed to explore the development of teachers’ pedagogical practices and learning relationships upon the inclusive education of children with special educational needs and disabilities in two primary school classes. The study considered the views and behaviours of primary school pupils with and without special educational needs, primary school teachers and teaching assistants (TAs) in one mainstream school. Drawing on a multimodal approach to discourse analysis to account for the complex relationships between symbolic and non-verbal modes of classroom signification, the study explored how meaning is produced in classrooms and children’s modes of communication, as well as in teachers’ practices. The two classes are compared on the basis of teaching observations, interviews, transcripts of dialogues, and analyses of classroom organisation and decoration. This paper suggests that the greatest influence on the educational and social outcomes of students with special educational needs is the behaviour and practices of the classroom teacher.

ABOUT THE AUTHOR Efthymia Efthymiou is an assistant professor of Education, at the College of Education, United Arab Emirates University. Her research focuses on inclusion, and educational environments which are conducive to supporting inclusive class teaching from the point of view of stakeholders, particularly children with SEND and how far the physical as well as psychological environments contribute to positioning and attitudes. She has experience in qualitative research techniques.
Alison Kington is a professor in Psychology of Education in the Institute of Education, University of Worcester, UK. Her focus is on "classroom life"; that is the experiences, relationships and perceptions of children and staff in schools, with an emphasis on the utilisation and application of research to improve professional practice. To that end, she has developed experience in mixed methods research design with specific expertise in classroom observation, critical event narrative and repertory grid techniques.

PUBLIC INTEREST STATEMENT
The debate regarding the inclusion of children with Special Educational Needs and Disabilities (SEND) in mainstream education in the UK partly revolves around what makes the classroom environment inclusive. Through the potential offered by the specific qualitative methodologies employed, this study aimed to explore the development of teachers' pedagogical practices and learning relationships upon the inclusive education of children with special educational needs and disabilities in two primary school classes in the UK. This paper suggests that the greatest influence on the educational and social outcomes of these students is the behaviour and practices of the classroom teacher. Future government reforms of education provision for students with special educational needs might consider a shift from a system-centred approach towards more personcentred practices that embrace support from multiple disciplines and plan learning according to the specific needs of each student.

Introduction
While the formal curriculum is the overt teaching of school subjects, the extra curriculum refers to teaching outside formal school hours-for example, extra maths or English classes, sometimes privately paid for-what the informal or hidden curriculum implicitly transmits are the values and rules, both academic and behavioural, of the school and society (Print, 2011). In this study, emphasis was placed on the hidden curriculum and the verbal and non-verbal means through which it is realised. For example, the content of visual displays, the arrangement of desks, the ability grouping of students, reward/punishment systems, the pacing, sequence and type of work (competitive/collective) in a classroom contributes to the values of a school and, possibly, how these values are communicated.
The ongoing debate regarding the inclusion 1 of children with special educational needs and disabilities (SEND) in mainstream education in the United Kingdom partly centres on the development of appropriate pedagogies and classroom relationships, and the ways in which these contribute to an inclusive learning environment. Social and local conceptualisations of government guidelines, as well as specific school agendas, currently influence a range of practices as the implementation of these guidelines can vary between settings. In the light of previous studies (Clark, Dyson, Millward, & Robson, 1999;Rose, Shevlin, Winter, & O'Raw, 2010;Tralli, Colombo, Deshler, & Schumaker, 1996) suggesting that mainstream settings adopting teacher-centred pedagogies may find it more difficult to support the psychological, social and educational needs of students with SEND due to reduced opportunity for dialogic interactions compared to settings with more child-centred practices, this study sought to explore this issue further (Goddard, 1995;MacInnis & Hemming, 1995;O'Neill & McMahon, 2005).
The research questions that framed the study were therefore: • What are the social pedagogic practices regarding child-child and adult-child interactions within this primary school context?
• What do teachers and children perceive as key features of, and influences on, these interpersonal relationships?
• To what extent do these interactions and relationships differ for SEND and non-SEND children?
Based on the perspective that inclusion, and SEND more specifically, is a socially constructed perspective, a qualitative case study of two classes within a primary school setting was conducted. The voices of students with special educational needs and their emic perspective of SEND 2 were prioritised, although the perceptions of teachers and peers were also garnered. In addition, semiotic evidence was collected in order to ascertain what, if any, influence these devices had on classroom communication. Finally, verbal and non-verbal interactions of children and adults were observed and recorded (i.e. gesture, gaze, body posture) in order to understand the kind of interactions that developed and the messages these interactions convey within the hidden curriculum (Harpur, Lawlor, & Fitzgerald, 2004). The next section gives a brief overview of the policy context within which inclusive practices are framed.

The legislative framework
In order to contextualise this study, it is necessary to briefly describe the key legislation relating to inclusive education. In Britain, the 1974 Committee of Enquiry, known as Warnock Committee, began a review of the educational provision for children with disabilities. The resulting Warnock report (Department of Education and Science, 1978) identified the importance of the inclusion of children with learning difficulties in mainstream schools, and argued in favour of reducing the number of schools that focused on the education of pupils with Special Educational Needs and Disabilities (SEND). The report also stressed the importance of individualisation and the evaluation of educational needs in relation to psychological and environmental factors which had previously not been accounted for. In addition, factors such as the severity or complexity of a child's learning needs, disruption to the education of other children, and the potential for academic failure within a mainstream classroom context were also addressed.
In 1981, an Act of Parliament was passed which resulted in two major changes: that Local Education Authorities (LEAs) 3 would have responsibility for the educational provision of children with SEN; and, that parents would have the right to actively participate in the assessment of their child's needs and decisions regarding their schooling (Lindsay, 2003). The Act was later modified in 1988 which saw the Department for Education (DfE) assume responsibility for the quality of education for children with SEN, as well as the allocation of financial resources which had previously been the remit of the Department of Health (Centre for Educational Research and Innovation, 1999). A second modification was implemented in 1993 which strongly suggested that children with SEN should, where appropriate, be placed in mainstream settings using suitable programmes of study, assisted by specialised teaching staff where necessary (Evans & Lunt, 2002).

The impact of financial and personal cost on the effectiveness of inclusive education
The legislation outlined above was welcomed but brought with it issues of cost and effectiveness. Currently in the UK, the number of students with SEND taught in inclusive mainstream classes is currently disproportionate to the available funding for educational provision (Barton & Tomlinson, 2012). According to recent research, some school districts have incorporated students with SEND into regular classrooms to minimise the cost of the special education support that students with more severe learning disabilities need (Ryan & Cooper, 2012). This could be due in part to the fact that in England, the average cost of placing students with SEND in special schools is £57,150 a year, while the cost, per child, in mainstream primary education is around £9,000 per year (The Telegraph 27 September, 2008).
In reaction to this, the Office for Standards in Education (Ofsted, 2010) claimed that there were some schools that over-identified and over-labelled students as having SEND, resulting in lower expectations and subsequent underachievement. Their estimation was that around 457,000 children were incorrectly identified as having SEND in 2010. However, the system encourages this trend as consideration is given to schools with a high proportion of students with SEND in the league tables and in terms of increased funding from LAs. According to the national press, in England 1.7 million children were registered as having learning, behavioural or physical disabilities (The Telegraph 2010). However, less than three per cent of students have formal statements, while fewer than one fifth (18.2%) of students were reported to be on School Action 4 or School Action Plus 5 programmes, neither of which requires students to be formally assessed for disability. The Ofsted report (Ofsted, 2010) concluded that many students would not be identified as having SEND if schools were engaged in fully inclusive and effective teaching and learning strategies that corresponded to the needs of all children.
The different provision offered by Local Authorities (LAs), coupled with the differences in how schools and LAs identify children as having SEND, has been associated with a lack of resources, limited government funding, and the pressures of performing well in national league tables (NASUWT, 2008), all of which discourage schools from giving the support some SEND students in mainstream classes may need. In fact, the use of league tables as indicators of the effectiveness of schools in meeting the needs of diverse students has been reported to "mask considerable under-achievement or, alternatively, conceal genuine school effectiveness" (Mortimore, 1996, p. 29). The measures taken by schools to enhance and support the inclusion of students with SEND and their academic achievement are thus varied and external pressures can lead to attention being focused on students with high attainment levels (Gillborn & Youdell, 2000).

The effect of inclusion on students with special educational needs and disabilities
Inclusive policies relate to both the academic and social development of children with special educational needs (Palmer, Borthwick-Duffy, & Widaman, 1998). In very general terms, students with special educational needs and disabilities can have a positive or negative experience from both the educational and social aspects of schooling, often connected to the hidden curriculum or their positioning by others (Huber, Rosenfeld, & Fiorello, 2001;Wang & Reynolds, 1996).
In general, previous studies focused on aspects of inclusive teaching seem to be positive, especially when students with and students without special educational needs have similar abilities, interests, and concerns, and students with SEND can satisfactorily cope with the learning pace of the class (Karsten, Peetsma, Roeleveld, & Vergeer, 2001). Students with SEND can benefit academically and emotionally when supported by their peers, developing strong friendships. Previous research has suggested that there are social benefits to the inclusion of students with SEND within a mainstream 6 classroom, (e.g. possible elimination of stigmatisation and feelings of social isolation) when students when there is understanding and acceptance by teachers and peers (Banerji & Dailey, 1995;Van Houten, 2009). Furthermore, Downing, Spencer, andCavallaro (2004) and Buckley, Bird, Sacks and Archer (2002) found that the inclusion of children with special educational needs who had similar levels of academic ability benefitted both groups of children academically and emotionally.
Some studies suggest the academic position of students with special educational needs is established through individualised curricular tasks and education plans (Dyson, Farrell, Polat, Hutcheson, & Gallannaugh, 2004;Winters & O'Raw, 2010), the appropriateness of teachers' training and professional development in supporting students' needs (Campbell, Gilmore, & Cuskelly, 2003;Organisation for Economic Co-operation & Development, 2005;Ware et al., 2009), and the collaboration between mainstream and special teachers for adequate support (LoVette, 1996) as part of the Profile of Inclusive Teachers (European Agency for Development in Special Needs Education, 2012). However, other studies highlight the ineffectiveness of individualised instruction and special provision for students with special educational needs and disabilities (Hocutt, 1996) when the process of differentiation or specialisation results in intensification of disabilities (Lewis & Norwich, 2005) or when it is due to inadequate pre-service and in-service training (Cooper & Jacobs, 2011;Lackaye, Margalit, Ziv, & Ziman, 2006;Rix, Sheehy, Fletcher-Campbell, Crisp, & Harper, 2013).
A further reason suggested for the ineffectiveness of such an approach has been that some activities designed to promote social participation and peer interaction are neglected due to a substantial decrease in teaching assistant provision (Giangreco & Doyle, 2007;Vaughn, Elbaum, Schumm, & Hughes, 1998). This social positioning of students has been found to be important in the development of classroom dynamics and relationships. According to some researchers, the social position of students with SEND is said to be established through their self-esteem (Myklebust, 2007), friendships with peers (Wiener & Tardif, 2004), self-confidence (Bakker, Denessen, Bosman, Krijger, & Bouts, 2007), social skills (Nakken & Pijl, 2002), social participation (Booth, Ainscow, Black-Hawkins, Vaughan, & Shaw, 2000) and self-awareness (Cole, Waldron, & Majd, 2004). Conversely, other studies have identified that integrating SEND students into mainstream classes can lead to their poor performance, low self-esteem, low peer acceptance and loneliness (Dyson, 2003;Wiener & Tardif, 2004). Additionally, poor socialisation with teachers and peers can result in feelings of social rejection and encourage low aspirations (Bakker & Bosman, 2003;Lackaye et al., 2006).
These studies illustrate both the negative and positive aspects of inclusion regarding the academic positioning of students with SEND in mainstream classrooms and identify the impact of how practices are structured at the organisational level, with differing results according to their appropriateness or effectiveness for students with SEND.

The hidden curriculum and attitudes to SEND students
The inclusion of SEND students in a mainstream educational environment requires effective delivery of the curriculum, as well as aspects of the hidden curriculum, which some research reports can be influenced by the teacher's perceptions of children with SEND (Bartolomé, 1994). Likewise, the support in mainstream classrooms for students with SEND (e.g. the deployment of teaching assistants 7 ), can lead to teachers developing differing relationships with certain groups as they might have more contact with non-SEND students than with students with SEND (Blatchford, Bassett, Brown, & Webster, 2009). This is reflected in the findings of Miles and Singal (2008) who discuss the disproportionate amount of time spent on some students with diverse needs. Here, the teacher intentionally or unintentionally relies on a teaching assistant to manage disruptive behaviour or to instruct students with challenging performance, positioning them through this aspect of the hidden curriculum. However, students with SEND learn to work with a particular adult and teaching style and may find it difficult to cope with change in this area.

What are the conditions required for inclusion?
If inclusion is to remain a legitimate aim, then understanding and actualising what pupils find enjoyable and productive in the classroom needs to be harnessed, enabling them to progress and experience a positive learning environment. These conditions may depend on teachers' organisational skills and motivation to work within an inclusive classroom, the level of teacher talk, students' desire to engage in challenging tasks, and their active participation in different collaborative activities (Hopkins, 2008). Hocutt (1996) suggests that individual study plans can be successful in this aim as they involve giving priority to each individual student needs and then allowing policy-makers (and other stakeholders like teachers, parents, children) to decide on the type of curriculum considered most useful and constructive for each student. A range of possibilities could provide equal educational opportunities to students with and without SEND. Inclusive education is the basis for the elimination of educational and social exclusion and is a medium for the abolition of discrimination against people who are different, to allow them to achieve social inclusion in the broadest sense. As Ranganathan (2000) states, The social development of the child involves different aspects of her social nature, like the social milieu that produces her -home, school, neighbourhood and community -and the influence each has on her attitudes and social values, interpersonal relationships and social behaviour … Then, there is her habitual or typical response to others … to determine what kind of social structure she, and each such child belongs to. (Ranganathan, 2000, p. 57) In support of this, researchers (e.g. Vaughn, Elbaum, & Schumm, 1996;Vaughn & Schumm, 1995;Vaughn et al., 1998) have identified a number of factors that are said to be required for effective inclusive education: • Appropriate training and education for teachers in mainstream education to enable them to consistently respond to the needs of all students.
• Establishment of the right of mainstream teachers to choose whether or not to teach inclusive classes.
• Encouragement of teachers to develop proposals for the implementation of inclusion.
• Promotion of collaborative relationships between teachers of mainstream and special education.
• A focus on the needs of students, not on ideology.
• A tailor-made curriculum to meet the needs of all students.
• Evaluation and monitoring of the progress of children and the support services.
• Involvement and participation of parents in the educational process and decision-making.
• A guarantee of equal social and academic potential for all children.
So, as outlined, a plethora of studies, mainly in the United States, have investigated the educational and social aspects of students' inclusion in mainstream educational contexts. In order for this study to explore the development of inclusive learning relationships, two theoretical frameworks were adopted which are now discussed.

Theoretical framework
Key to this research was the study of classroom discourse between the teacher and pupils, and the pupils themselves. Moreover, this study investigated how classroom discourse influences the interactions of children with SEN and their peers, but what constitutes classroom discourse? It could be claimed it is constituted by social processes which affect the local interaction of children in a classroom and the ways they interact. These social processes can be expressed in explicit and implicit ways. The local interaction of children may be directly observed when classroom discourse is expressed explicitly but not when classroom discourse works in implicit ways. In order to focus on this discourse, two theoretical frameworks were adopted; multimodal theory and linguistic ethnography.

Multimodal theory
The analysis of communication was aided by the framework of multimodal theory (Kress, 2001;Kress et al., 2004), within a socio-cultural perspective of communication and multimodality (Ivarsson, Linderoth, & Säljö, 2009). The modes under investigation were the non-material (such as gesture, images, speech, writing), and the material (such as paint, wood, paper) (Kress & Van Leeuwen, 2001, p. 22) which together compose a multimodal perspective. This enabled the exploration of how children participate in semiotically mediated activities and how they interact and represent their meanings through different semiotic resources available in the learning context. For example, information displayed in the form of pictures can shape the ways in which children communicate in the classroom, as well as the kind of roles children take in their communication. So, the rich and in-depth analysis of classroom interaction can be studied through a multimodal lens, where talk is one of multiple modes of communication along with gesture, gaze, posture, photographs and spatial arrangements (Jewitt, 2006).
These interactions with various forms of communication, the teaching and learning approaches, and the classroom relationships were examined alongside the use and organisation of space. For example, the organisation of furniture places some children in closer proximity to the teacher which can, in turn, impact upon the social and learning dynamics in the classroom. A teacher, for instance, might place all children with SEN on a separate table away from mainstream peers in order to scaffold the teaching by ability group. In this case, different modes could produce different perceptions in children regarding their peers with SEN in the classroom. This theoretical lens enabled the study to also focus on the media and modes in each classroom, taking into account how teachers and students interacted and interpreted them.

Linguistic ethnography
Linguistic ethnography (LE) is a theoretical framework that describes the dynamic aspects of communication and that can represent how meaning making is constructed through children's talk in the classroom and playground. Drawing on this framework, the study explored the synthesis of official and unofficial social and learning activities within the classroom, and focused on the identification of diverse kinds of inter-student communication, relationships and identities. Bakhtin's concept of evaluation is fundamental to the study of how children communicate their position through talk, as they evaluate their social experiences and become aware of their positioning in the world, developing their self-perception. Children's social backgrounds are part of their particular evaluations about how to act in the world and this approach to classroom communication and discourse analysis, based on Bakhtin's theories of communication, affiliates broadly to the study of language in a particular context (Maybin & Tusting, 2011).

Research design
Qualitative methods were adopted for this research in order to explore the influences of teachers' pedagogical practices upon the inclusive education of children with special educational needs and disabilities. It was also critical for the voices of the children to emerge as social actors capable of influencing the structures that exist in their context (James & Prout, 1990). Children's communication as a reflection of their inclusion was explored through their personal accounts of their experiences in focus groups and informal discussions in the classroom and playground. Furthermore, reflections on inclusion involved non-verbal communication and contexts which mediated verbal and non-verbal communication, so participant observation, field note, and in-depth interview data were also collected for analysis. Combining different methods allowed data triangulation and increased the authenticity (Barbour, 2001;Messiou, 2002) and trustworthiness of the study. The impact of context on the communication of children was also explored so that the meaning-making could be understood in relation to the context in which it was produced (Maykut & Morehouse, 1994).

The research site and sample
Sunny Hill School (pseudonym) is a voluntary controlled, mixed primary school nestled in a small village in the Midlands of England. There were 122 children (aged 5 to 11 years old), on the school register; 17 were supported by School Action Plus and seven children with special educational needs were supported by School Action. The selection of teachers, Danny and Bam, was to fit the purpose of this study (Miles & Huberman, 1984). Danny was in his thirties, a recently qualified teacher, and had been teaching year 4/5 at Sunny Hill for two years. Danny had 22 students in his class; two had been identified as having special educational needs and disabilities. Bam was in her fifties, and had been teaching year 6, and working at the school for eight years. In Bam's class, three of the 25 students had statements of special educational needs and disabilities.
A total of four students with mild to moderate disability, aged 11-12 years old, participated in the study; two were in Danny's (Year 4/5) 8 mainstream class and a further two were in Bam's mainstream class (Year 6) 9 but had support from a teaching assistant and were withdrawn for Literacy and Numeracy. The students were chosen by the teachers as they needed adequate cognitive and language capabilities to express their attitudes to specific issues during interviews, to remember past incidents, and to converse with others in the same group. The parents of the children gave informed consent prior to data collection and the children assented to participating in the study (BERA, 2011). The data collection phase of the research took three months.

The instruments
Qualitative methods such as interviews, focus groups, and observations were employed to enable a better understanding of school practices and the interactions involved in the inclusion of students with special needs at the social, organisational, and interactional levels in mainstream classrooms. The multi-method approach adopted for this study meant that the research was reshaped as different methods were employed. On this basis, it was essential to reflect at every stage of the research process and to be flexible according to the needs of the study in order to construct a coherent research design (Avramidis & Wilde, 2009). The process of observing students and studying the nonverbal stimuli in the school environment supports a "multiperspectival" (Cohen, Manion, & Morrison, 2007), methodological approach due to the complexity of the process of inclusion and of the positioning of students with special needs in the classroom. The employment of multiple methods enabled a holistic view of inclusion examining multiple causalities and perspectives and taking into account the different views of participants (Table 1).
Specifically, 12 participant observations in Year 4/5 and 11 in Year 6 during lessons and breaktime, were supplemented by 3 semi-structured qualitative interviews with teachers and teaching assistants, and 3 focus groups with students with and without special educational needs from years 4/5 and 6 ( Table 1).

The observations
Throughout the study, written descriptions in the form of field notes were made during and immediately after the observations (Lindlof & Taylor, 2010). In ethnographic research, it is accepted that the researcher spends considerable time involved in the lives of the participants in order to describe the observed phenomena and develop an understanding of the cultural aspects of their lives, attitudes and motivations (Murchison, 2010).
Twenty-three observations were conducted in lessons (approximately 40 min each) where children worked individually or in mixed ability groups in order to observe any similarities or differences in the ways the lessons were facilitated for children with (and without) special educational needs and disabilities and how inclusion was achieved. For the purposes of this study, a single researcher was introduced gradually into the daily routines of the children and, after initial interactions had been made, fully participated in classroom activities (Adler & Adler, 1994). This approach facilitated detailed observations of children in individual and group activities, and enabled verbal and nonverbal interactions, as well as developing teacher-child and child-child relationships, to be documented. Participating in activities with children and teachers in this way provided an effective way to establish trust and rapport which, in turn, aided the data generation process (Bogdan & Biklen, 2007).
As part of the fieldnotes, another set of notes, memos (Glaser & Strauss, 1967), further comments, speculations and experiences of the data collected were also added. Comments were also made on efficiency of the methods employed and potential solutions and issues of rapport with the participants. This highlighted the reaction of participants to a researcher presence in the field (Hammersley & Atkinson, 1983).

The interviews
Qualitative interviewing facilitated the interaction with the social reality under investigation (Gubrium & Holstein, 2002). The relationship between the interviews and observations was reciprocal, as information elicited through observations was then extended through interviewing (and vice versa). A semi-structured interview technique was adopted for the collection of data from teaching staff. Interviews with teachers and teaching assistants took approximately 30-40 min each and questions focused on specific practices used for supporting the academic and emotional needs of children with special educational needs, the communication of children with SEND with their peers, and how classroom tasks facilitated their interactions ( Table 2). The teachers and teaching assistants provided insights into specific incidents regarding the classroom interactions and reconstructed specific events which corroborated observations, or students' focus groups.

Focus groups
Focus groups were appropriate for studying children as social actors in the schools, giving them voice, and allowing the interactions of the participants to develop through discussion (Kitzinger, 1995). Two focus groups (4 students in each) were conducted with each of the following groups; non-SEND pupils in years 4/5 and 6, SEND pupils in years 4/5 and 6. The focus groups, which lasted approximately 20 min each, engaged the children in open discussions regarding their experience of interactions with children with special educational needs and how they experience school. The children exchanged ideas and challenged each other's perspectives.
The questions were in an open-ended format to give them the opportunity to interact and exchange ideas on issues they found important. The year 6 students (aged 10-11 years old) were capable of identifying the views of others around them and of communicating their own thoughts and evaluations with insight and sophistication (Maybin, 2006). This was very important for the research objectives, as was needed to delve into how they understand and evaluate their social world with emphasis on their interactions with school peers. Body language and facial expressions were also recorded as data during conversation for use in interpretation and analysis. A topic guide (Table 3) was used to specify the themes while new and unexpected themes also emerged (Auberbach & Silverstein, 2003).
The questions included the characteristics of a good student, views on individual and group oriented lessons, and interactions and friendships with peers.

Research credibility: Trustworthiness and authenticity
Interactions with children in school and involvement in the various classroom activities enabled a process of verification to take place in order to ensure that interpretations of the analysis of data   2. How easy would it be for someone new in the classroom to identify the students with special educational needs?
Probe: why is that the case?
3. Do students with special educational needs receive additional support and assistance from internally funded sources (i.e. teaching assistant, extra help from the teacher)?
Probe: If yes, how is that allocated?
Probe: If no, why is this?
4. Are any inclusive education programmes used in your classroom?
Probe: If yes, how would you describe these? What are the perceived benefits of these programmes for students with SEN?
Probe: If no, why is this?
were authentic (Crozier & Tracey, 2000). The use of constant comparison (Glaser & Strauss, 1967;Strauss, 1987) enhanced the internal and external trustworthiness of the interpretations by (i) selecting participants who were as representative (as possible) of the phenomenon under study and, (ii) identifying the commonalities and differences in behaviour, perspectives, and incidents that existed and were observed within that phenomenon (Strauss & Corbin, 1990).
As the research involved children (who are seen as a vulnerable participant group), the need to protect the authenticity of the context within which their perspectives were observed was essential. Therefore, the design of this study adopted a multi-method approach for the collection of data in order that the perceptions and views of the children could be presented as closely as possible to the original communicative act, taking into account factors that might have influenced their responses. For instance, the impact of the presence of teachers and other children in the setting during the collection of data, noises inside and outside the classroom, the time of the school day were all factors considered in relation to possible impact on authenticity (Messiou, 2002).
The triangulation of data from focus groups, observations, and interviews with teachers increased the validity of research and the trustworthiness of the data for analysis (Barbour, 2001;Messiou, 2002) providing a critical perspective of the data collected (Lewis, 2002). For instance, the teacher's of year 6 face-to-face interview revealed that her instructional approach tended to be non-interactive/non-dialogic, which limited students with special needs' social interactions with teacher and peers. This finding was triangulated with class observations, which showed that the teacher's main objective was supporting more the academic, and less the social needs of her students. Additionally, the tape-recorded, class conversations between the teacher and her students (Excerpt 6) further suggested that she preferred a more teacher-centred pedagogic model with social implications for students with special needs.

Data analysis
Data were coded, categorised and transferred into analytical matrices (Miles & Huberman, 1994) which were then used to refine emergent themes and identify patterns. Subsequently all of the qualitative data, including observational and interview data, were subjected to a thematic analysis using NVivo and used to generate case studies relating to the social pedagogic practices, child-child and adult-child interactions, and key features of these interpersonal relationships for SEND and non-SEND pupils (Lawrence-Lightfoot & Hoffmann Davis, 2002).
NVivo was also used to synthesise and integrate the various strands of data through the use of key classroom attributes. These qualitative attributes were initially developed from the observations of classroom practice but also included characteristics of each class, as well as pupil age and ability level. Following analysis of observation and interview and focus group data, a number of additional attributes were added. The ongoing integration and triangulation of the various forms of data provided various levels of rich analysis and problem refocusing.

Ethical considerations
Ethical approval was gained from the institutional research ethics committee prior to commencement of the research. All necessary steps were taken to ensure the participants understood the research requirements and that the key principles of informed participant consent, confidentiality, right to withdraw, anonymity and avoidance of harm were upheld (BERA, 2011).

Findings
The education experience of the students with SEND differed depending on year group (Year 4/5 compared to Year 6) since each class was a unique educational microenvironment with different sociocultural parameters. Although the data collected shed light on a number of issues, this paper focuses on those relating to the framing of diversity, and discusses the similarities and differences identified between the two classes in Sunny Hill School. http://dx.doi.org/10.1080/2331186X.2017.1304015

Forms of grouping and academic/social development
Forms of grouping were observed in year 4/5 and 6 classes. The participant observations and the interviews with the teachers and the TAs, and the focus groups of students with and without special educational needs and disabilities showed that, at an organisational level, grouping was instantiated as part of the school practices as students were categorised by ability and behaviour/temperament.
We have to make sure that it's at their level so they can succeed. So we differentiate it that way, and then, with the top end of the classroom, with those working at a higher level, then we give them extra challenges, then we give them different texts or different tasks to do. (Bam, year 6) The positioning and identification of all students and thus, their degree of inclusion in each classroom was important in terms of their academic and social development.
Observations identified that, in general, when students with SEND were in low ability groups, both their performance and social interaction with their peers were reduced and there was an increase in their need to work with the TA. The interviews with both teachers revealed the same points.
We do take people with special needs out for literacy and numeracy to get that extra support and it takes away from kind of the social side … they are not involved in a large class environment. (Danny, year 4/5) Obviously, they just couldn't access the curriculum at the level that the year 6's are working at really … they all access the whole curriculum at a level that they can mostly with the TA … (Bam, year 6) Teachers were seen to use grouping by ability as a way to monitor performance and behaviour, according to focus group discussions. These groups assigned specific academic and social identities to the students with SEND, either as less able or naughty students (Excerpt 1). Excerpt 1. Students with SEND (year 4/5 and 6).Harris: No. Like we can when he says "you can sit anywhere you want" but she* picks the morning places and stuff like that.Barry: Because all the time she wants us to sit with some people that will help, that are intelligent, so she puts like the clever people with the not so clever people so …Mary: They're probably the people you're not going to actually work with but … *she: the teacher Likewise, both teachers stated that placing students with SEND in low ability groups did not facilitate their educational progress, nor did it facilitate the inclusion of SEND children in full class teaching.
In this school we've been encouraged to kind of always plan. So taking, looking at the National Curriculum plan for your top learners or your high ability and then from there work down. So, start at the top and … scaffold down which is not always effective for students with special needs. (Danny, year 4/5) However, the teachers did suggest that the educational needs of children with SEND were best addressed through additional support from the TAs. In the year 4/5 and 6 classes, respectively, grouping by ability determined the implementation of specific educational practices, including the deployment of the TA to the less able groups, which often created additional needs to be addressed. I think you've got to balance the two. I mean obviously the teachers, they've got to get the results and they've got to be seen to be moving the children on you know sub levels … something more on the social side … (Bam, year 6) These TA-led groups were seen to isolate some children as they were unable to engage in peer learning or exchange ideas with their peers.
Get them more socially interactive … So citizenship side of things … Occasionally we have you know friendship issues, collaboration … or perhaps if we'd got something a bit more about their social awareness … (Danny, year 4/5) Furthermore, this practice promoted a negative academic identity for these students as the SEND group which only supported their assimilation with other students in low ability groups. This had implications for the academic and social needs of these children.
Within this context, the core issue was diversity, and how it was framed within the classrooms. Focus group data (Excerpt 2) suggested that the provision of curriculum differentiation was seen to encourage the development of cognitive and behavioural skills and the mastery of practical tasks by children with special educational needs and disabilities. Stephen: And they'll also be ready for things to come like secondary school and stuff.
Lisa: And university.
Stephen: Yes … if they want to go.

Lisa:
If they want to go. 15 Stephen: I think it is very, I think it is quite important and it's nice to see them getting on really as well.

Lisa:
Yes getting on with their life, not worrying over their special need and that.

Stephen: Yes it's quite nice because I think it's pretty good that they can work at their level as well which helps because … 21
However, the differentiation of resources affected the social relations of the students with special educational needs with their peers (Excerpt 3).

Tom:
But we still accept them as people in our class and we talk to them and everything.

Lisa:
Yes; you don't like to single them out because they're already out and that's because they've got special needs. But we don't want to make them feel that way because it's not really fair on them.
Stephen: They just feel part of the class really.

Lisa:
They are all part of the class, it's just they've got learning disabilities and they're different.

Tom:
And there's nothing wrong with that because everyone's different.

Seating arrangements and opportunities for interaction
As with the data relating to grouping above, the data indicated that when the seating for students with SEND was organised in ability groups, their academic involvement in whole class learning was minimal as they interacted only with those of the same ability and thus, could not challenge their capabilities with higher ability peers.
Well, obviously, it's quite hard because by the time they get to year 6 of course you really have got the spectrum of ability in there really. So with the very special needs for maths and English we withdraw them from the classroom and they work in a small group working at their level. Obviously, they just couldn't access the curriculum at the level that the year 6's is working at really. (Bam, year 6) Interactions with the class teacher and their peers were also constrained because the activities and seating arrangements facilitated more individual and less collaborative work. The findings indicate that, in Year 4/5 and 6 classes at Sunny Hill School, the seating arrangements for students were based on mixed ability groups of 3-6 students, whereas students with SEND were seated individually. Moreover, the single seating of the students with special educational needs in Year 6 was in close proximity to the TA who supported them in their studies. Danny (year 4/5) explained, Obviously we do take people with special needs out for literacy and numeracy to get that extra support and it takes away from kind of the social side of the morning part, they are not involved in a large class environment.
However, even the collaborative tasks in mixed ability groups sometimes included individual work as the pace of the task was beyond the capability of the students with special educational needs, inhibiting their interaction with peers and positioning them as passive learners not involved in the production of knowledge. Bam pointed out, Obviously, they just couldn't access the curriculum at the level that the year 6's is working at really.
For students with SEND in year 4/5, the constant presence of the TA made the need for additional support more visible to other students, and minimised their interaction with the class teacher. Socially, the students with SEND in the low ability group interacted only with peers at the same table, which marginalised them from the rest of the groups (Excerpt 4). Researcher: I see, okay.

Mary:
Because I go out with the woman who helps people with problems, when everyone in my class has got a problem.
During an interview, Danny (Year 4/5 teacher) identified the need for the students with special educational needs to be socially involved with their peers and he sought to do this by placing more emphasis on group work and less on individual learning. However, in practice, he reported that his teaching style was more suited to competitive rather than collective tasks, requiring the students to strive for superiority, involving winners and losers.
You've got those children that are on your less able table, they are aspiring to move up … like winners. Okay, they've got a lot harder task to do to move up but they are still aspiring to move up … not to be losers … the children are very, very astute they are more astute than adults are, and they very, very quickly work out where the pecking order is and the hierarchy.
(Danny, year 4/5) He was also aware that this could present an issue for the children with SEND special educational needs as they could not work at the same pace as their peers in order to reach the targets of their teams.

Curriculum, pedagogy and classroom relationships
A further area of interest that emerged from the data was how curriculum and pedagogy influenced the developing relationships in the classroom. The two teachers and teaching assistant all agreed that the organisation of the classroom was at least partly associated with how they could incorporate the official curriculum into their personal teaching style and philosophy. During interviews with teachers, they reported that, when pedagogy was focused on both the educational and social aspect of learning, lessons were more dialogic and interactive, the students with special educational needs had more visual and verbal interaction with the teacher, and there was a greater level of communication with peers.
I'd certainly I would do … more group work, more interaction with me, social interaction between children, working together, team work … more of the drama, and the art and the creative side … (Danny, year 4/5) Some activities give them the opportunity to put it right without being on top of them all the time. It's important to let them take responsibility for their own actions … to work with others and have more interaction and talking with the teacher. (Bam,year 6) In addition, students with SEND were more likely to be active participants, speakers (as well as listeners) in the process of learning. Conversely, when pedagogy was focused only on the educational achievement, a monological mode of communication was more apparent, students with SEND interacted less frequently with the teacher and peers, and were passive learners.
Based on the observations in year 4/5, although the students with special educational needs sometimes participated in group activities, they worked with the same ability students at the same tables, and communication with students at other tables was restricted. The nature of the activities was educational rather than social, thus, did not enhance mixed ability interaction. As Danny explained, I prefer them to work individually, particularly, in the core lessons of the curriculum such as in Mathematics, English and Science … they need to focus on their work … make sure they will complete their work … not wasting their time. (Year 4/5) Conse.quently, they interacted less with the teacher and more with the TA who supported them through more systematic and individualised work. Occasionally, in subjects like, Art, the students brought materials from home for a specific task, combining school knowledge with out-of-school knowledge. This situation promoted more interaction between the students although mainly within the same ability group (Excerpt 5). In year 6, Bam also reported not utilising interactive modes for teaching curriculum subjects. She stated that she was more comfortable with an authoritative pedagogic model to instruct the students, but confirmed that this approach was not aimed at the students with special educational needs who would have support from the TA. Socially, the students with SEND did not interact with their teacher or with their peers as instruction was mainly non-interactive/non-dialogic. Observations suggested that Bam's main aim was to support the academic, not the social aspect of the students' learning (Excerpt 6).
Excerpt 6. Conversation between teacher and students in year 6. Overall, the data indicated that students with SEND tended to engage less than their peers in either verbal or non-verbal interaction with teachers or peers academically and/or socially when activities were competitive or teacher-centred. By contrast, students were involved in more verbal and nonverbal interaction when activities fostered both their academic and social collaboration and were more student-centred.

Discussion
The findings from this study suggest that the teacher is critical to interactional and collaborative educational practices, not only to emphasise and prioritise the voices of the students at local level, but to bring out those voices and allow them to emerge and grow as autonomous entities. The study attempted to identify how the concept of inclusion in the two classrooms (years 4/5 and 6) of Sunny Hill School was realised, and how pedagogical discourses were produced. This section returns to the key issues identified in the initial research questions and examines them in the light of the findings and previous literature in this area.

Social pedagogic practices regarding classroom interactions
The engagement of the teacher and the students in pedagogic practices with monomodal and monologic rather than multimodal and dialogic characteristics, and the transmission of the curriculum through a mainly monomodal communicative medium was reported to inhibit the acquisition of knowledge of students with special educational needs. When the curriculum was not recontextualised through practices for the transmission and acquisition of knowledge that incorporated mental, emotional, physical and spiritual properties, this limited the emphasis on the development of intellectual and physical skills (Steiner, 1996). However, when students were involved in cross-curricular tasks enabled them to develop both their intellectual, emotional and physical skills.
Monologic discourse promoted through educational activities was reported to reduce the active participation of the students with SEND and positioned them as passive learners. The emphasis on the accomplishment of specific academic targets implied the teacher's symbolic control of the ways of distributing knowledge. Their positioning as passive learners in the learning process, lack of communication from the teacher, as well as their limited peer collaboration could lead them to perceive their poor performance as lack of cognitive skills and academic failure (Licht & Kistner, 1986).

Key features of, and influences on, interpersonal relationships
Teacher-centred approaches and hierarchically based pedagogic practices emphasised the academic rather than the social elements of pedagogy as grouping was often done by ability, thereby emphasising individualised and competitive activities rather than collective ones with a common target. The grouping of students with special educational needs by ability, coupled with their seating arrangements, was seen to inhibit interaction with the teacher (Blatchford et al., 2009). This physical marginalisation of students with SEND was seen as an inhibitor of social interactions and informal communication, which could imply that the various pedagogic discourses distributed specific social relations and consciousness as they integrated specific student identities (Hampton & Mason, 2003). This conclusion supports previous studies that have shown that social status and self-esteem of students with SEND are interrelated (Koster et al., 2009;Woolfolk, 1995). Thus, students with SEND experience lower peer acceptance compared to their peers and develop fewer friendships (Koster, Pijl, Nakken, & Van Houten, 2010) because of their poor academic attainment (Vaughn et al., 1998) and disruptive behaviour (Roberts & Zubrick, 1992).
Multimodal data suggested that placing students with SEND in the low ability group in both year 4/5 and 6 classes did not seem to facilitate the educational progress and inclusion of the SEN children in full class teaching as their educational needs, as perceived by the teachers, were mainly addressed through additional support from the TAs. From the interviews with the teachers of year 4/5 and 6 it seems the teachers required their support, as much as the children with SEND, in order to lighten their work load. Based on the students' with SEND verbal interactions, grouping by ability meant that the students with SEND were isolated, could not engage in peer learning or exchange ideas with their peers most of the time, as they interacted mainly with the TA. Furthermore, in-class observations indicated that this practice maintained the identity of these students as the students with SEND and promoted their assimilation only with other students in low ability groups. Consequently, their social needs were not met either. It seemed that grouping by ability supported the needs of the teacher as much as, or possibly more than, the children with SEND. It seems that the students with SEND in these classes were not the only or even main target of the TA's intervention. It may even be that class teachers saw the TAs role as being to support them.
Previous studies of diverse views of the inclusion of students with special educational needs indicate that placing them in mainstream settings with additional support and/or separate tasks might obstruct their inclusion (Ainscow, 2000) if they do not have equal access to the same educational material (Murray & Lawson, 2007). As Dyson (2001, p. 25) noted "the more their educational responses emphasise what learners have in common, the more they tend to overlook what separates them". This study revealed that the material production of the semiotic resources (Kress & Van Leeuwen, 2001) identified principles of power and symbolic control as they represented pedagogic structures that marginalised and positioned the students with SEND differentially. This practice identified some as 'the students with SEND' and promoted their assimilation only with others in low ability groups. This finding supports previous research which has shown that students with SEND can develop feelings of failure (Dyson, 2003), poor socialisation with peers and teachers (Estell et al., 2008) and low aspirations for their future (Lackaye et al., 2006).

The impact of individual differences on interactions and relationships
The teachers' interviews, class observations and students' with and without SEND verbal interactions revealed that, when there was a lack of cultural awareness or respect for diversity through the nonprioritisation of cross-cultural activities, this resulted in a less flexible pedagogic practice, with emphasis on the academic attainments of the students rather than social awareness of diversity (Murray & Lawson, 2007). Pedagogic discourse, which comprises the principles of recontextualisation, constitutes the rules of the specialised communication through which pedagogic subjects are selectively formed. Bernstein's (1990) pedagogic discourse is incorporated into his theory of codes and cultural reproduction: These codes of discourse, ways of relating, thinking, and feeling, specialize and distribute forms of consciousness, social relations, and dispositions. (Bernstein, 1990, p. 135) As this paper indicated, primary students develop different identities within their educational and social environment. This environment positions them according to their needs and selects specific ways to cater for them through labelling and a homogeneous approach which aims to "measure" their potential and to decide their position in a hierarchy. However, as Gill and Thompson (2012, p. 7) observed, the idea of implementing alternative approaches to the integration of students with SEND needs to "serve certain ends of society" which incorporate democratic thinking, cultural awareness, equality in education, and individual development; in this way, a "human-centred" (Gill & Thompson, 2012, p. 2) education is fostered that supports the academic, social and personal advancement of students.
In year 4/5, and 6 classes, grouping by ability determined the implementation of specific educational practices and created more needs which the teachers had to face. Within this context, the core issue was diversity, and how it was framed within the specific classrooms. Based on the teachers' interviews, the provision of curriculum differentiation, special resources, i.e. books, pictures, special equipment, could encourage the development of cognitive and behavioural skills and the mastery of practical tasks by children with SEND. However, it seemed that this grouping reinforced the separate identification of the students with SEND in the classroom and their positioning in lowability groups as students' with SEND interviews suggested. The differentiation of resources affected the social relations between the students with SEND and their peers, making any weaknesses more visible and assigning them specific identity positions. Their peers in the focus groups expressed either sympathy to the problems of SEND children or they labelled them as less clever.
Some studies highlight the ineffectiveness of individualised instruction and provision for students with SEND (Hocutt, 1996) due to the need for increased training (Cooper & Jacobs, 2011;Lackaye et al., 2006;Rix et al., 2013), especially with regard to activities that promote social participation and peer interaction (Power & Hyde, 2002;Vaughn et al., 1998). Ainscow (2007) went on to emphasise the need for teachers to be given additional training in how inclusive teaching is implemented and how differentiates from current practices in the classroom. Furthermore, some studies (Edwards & Kuhlman, 2007;Huefner, 1998;Jones & Fuller, 2003) supporting inclusion, identify the need to train teachers in special education to acquire the knowledge and skills to meet the needs of all students in mainstream classrooms (Miles & Singal, 2008) and ensure successful inclusion with emphasis on a responsive and interactive pedagogy (Edwards & Protheroe, 2003) that caters for the diverse needs and potential of different ability students (Slee, 2010).

Conclusion
Based on the contribution of the specific research methodologies to the investigation of teachers' pedagogical practices and their effects on the inclusive education of children with special educational needs and disabilities, this paper suggests that the greatest influence on the educational and social outcomes of these students is the behaviour and practices of the classroom teacher. This conclusion is the result of (a) the social pedagogic practices employed; (b) the key features of, and influences on classroom interpersonal relationships; and, (c) teacher awareness of students' individual differences.
The identification of the influences that included or excluded students with SEND in these mainstream classrooms aimed to generate a more detailed awareness of these issues and the extent to which positive pedagogical features contribute to inclusion. Future government reforms of education provision for students with special educational needs might consider a shift from a systemcentred approach that mediates a negative hidden curriculum, highlighting the authority of the teacher and the integration of students with labels in predetermined practices, towards more person-centred practices that embrace support from multiple disciplines and plan learning according to the specific needs of each student.

Limitations of the study
The findings of this small-scale study provide evidence about the cultures of two particular classes and the impact on the academic and social inclusion of students with SEND. As with all research, there were limitations to the study. Firstly, the multimodal perspective which focused on the emerging voices of students sought to gain an in-depth understanding of their views and experiences. However, this focus could have been further enhanced and supported with data relating to teachers' perceptions, views and experiences of pedagogy which were not studied to the extent that might have more fully delineated the quality of their impact and influence over students. This further exploration of the teachers' ideological positions might have produced additional information regarding ways in which teaching practices could become more effective.
Another limitation of this study is associated with the number of students with SEND in each of the classrooms. The total number of participant children was relatively small and, consequently, resulted in limited evidence of the dynamics of teacher/student and student/student interactions. Furthermore, the video recording of classroom activity could possibly have added semiotic insight into the meaning making through non-verbal behaviour of both students and teachers, as well as into the attitudes of teachers towards students with SEND. A final limitation is that the research design did not include an element specifically focused on practical implications of the evidence which would have facilitated possible developments in the educational provision of these schools. http://dx.doi.org/10.1080/2331186X.2017.1304015 and/or concentrated support under School Action" (Farrell, 2005, p. 46). 6. Mainstreaming : "the integration of children with disabilities with their peers in general education based on individual assessment" (Hocutt, 1996, p. 79). 7. Classroom assistant, learning support assistant. 8. Year 4/5: 8 to 10 years old; KS2. 9. Year 6: 10 to 11 years old; KS2; National tests and teacher assessments in English and maths, and teacher assessments in science.