Resolving dilemmas: Swedish special educators and subject teachers’ perspectives on their enactment of inclusive education

ABSTRACT This article presents the findings from an interview study with 12 subject teachers and special educators in three Swedish lower secondary schools. The purpose of the study was to explore how school professionals perceive and respond to dilemmas related to inclusive education that require prioritizations between conflicting values, principles or goals. Through the use of conceptual tools and theoretical notions of policy enactment, the article sheds light on how such dilemmas can be played out in practice. Four different dilemmas were discerned in the analysis: special vs. general education settings, external control vs. professional freedom, curricular demands vs. students’ needs and individual vs. group. The analysis illustrates how actors’ perceptions and responses to dilemmas can be framed by policy demands in different ways, but also by contextual factors and the actors’ roles and positions in their schools. In this way, the article provides a partly new understanding of such processes in relation to previous research. Moreover, it is argued in the article that the notion of dilemmas provides an educational dimension that can be an important addition in analyses of policy enactment.


Introduction
Although inclusive education has long been a prominent goal in global educational policy, it has been difficult to realize in practice (Ainscow, Slee, and Best 2019).In this regard, several researchers have made efforts to understand difficulties in the realization of inclusive education by studying dilemmas related to diversity that schools and teachers face in their everyday work (e.g.Norwich 2008;Kerins 2014;Molbaek 2018).Dilemmas are generally understood as challenges that cannot be completely conquered due to their inherent conflicts of principles, values or goals.Thus, they can only be temporarily resolved, which require prioritizations where all available options carry negative consequences (Clark, Dyson, and Millward 1998;Miller 2018).In relation to inclusive education, such dilemmas can concern the identification of students in need of special support, or how teaching should be organized for groups of students with different needs (Norwich 1993(Norwich , 2002)).
Much of the theoretical work on dilemmas in education has focused on what constitutes a dilemma and the inherent conflicts of specific dilemmas (e.g.Berlak and Berlak 1981;Minow 1985;Norwich 1993;Clark, Dyson, and Millward 1998;Tetler 1998;Nilholm 2005).Moreover, empirical studies have illustrated how practitioners can experience such dilemmas, and how resolutions to dilemmas are framed by national and local conditions (e.g.Norwich 2008;Ineland 2015;Molbaek 2018).However, it has been argued that challenges in the realization of inclusive education need to be further understood in relation to the focus on standards, testing, competition and accountability that has increasingly come to characterize education policy and school practices, and which conflicts with the idea of inclusion (e.g.Slee 2019; Magnússon, Göransson, and Lindqvist 2019;Done and Andrews 2020).In this respect, research has illuminated how standards policies tend to be prioritized when inclusive education is recontextualized in schools (Dyson, Gallannaugh, and Millward 2003;Ainscow, Booth, and Dyson 2006;Smith 2018;Barow and Östlund 2018), and how inclusive education becomes reinterpreted as a concept that promotes neoliberal ideals of efficiency and individualism (Engsig and Johnstone 2015;Magnússon 2019).However, more research is needed on how these conflicts between different agendas take shape in practice, and how they frame dilemmas and their resolutions in schools (Hamre, Morin, and Ydesen 2018).
In this article, this is pursued through an empirical analysis of dilemmas in Swedish subject teachers and special educators' enactment of inclusive education alongside conflicting policy demands.Drawing on the theoretical work on policy enactment by Stephen Ball with colleagues (Ball et al. 2011b(Ball et al. , 2011a;;Ball, Maguire, and Braun 2012), the respondents' descriptions of dilemmas and their resolutions are set within a theoretical account of the school organization as an arena where local school actors interpret, negotiate and translate policy.Hence, the ambition with the article is to combine the theoretical notion of dilemma with Ball and colleagues' reasoning on policy enactment to provide knowledge relevant to both fields of research.In particular, the article aims to contribute with new understandings of how dilemmas can be played out in practice by studying how school professionals' perceptions and responses to dilemmas are framed in their enactment of conflicting policy demands to their local contexts.

Conceptualizations of inclusive education
Inclusive education can be described as an educational ideal that emphasizes acceptance for diversity and access, participation and achievement in education for all students (Artiles, Kozleski, Dorn & Christensen, 2006).Thus, it stands in contrast to traditional perspectives on special education where students' deficiencies are understood as the cause to their school difficulties, which legitimizes exclusionary practices (Clark, Dyson, and Millward 1998).However, inclusive education is a concept that has been interpreted and defined in many different ways (Amor et al. 2018).In a review of highly cited research on inclusive education, Göransson and Nilholm (2014) found four types of definitions of the concept.
(a) Inclusive education as placement of students with special needs in mainstream classrooms (b) Inclusive education as a focus on meeting the social and academic needs of students with special needs (c) Inclusive education as a focus on meeting the social and academic needs of all students (d) Inclusive education through the creation of inclusive communities The two first definitions differ from the two latter since they specifically address the needs of a certain group of students.This highlights a tension within the concept related to the question of whether identification of the students that should be included is beneficial for those students, or if such labeling is a step towards exclusion (Magnússon 2019).Another tension regards the question of whether or not teaching should be differentiated to compensate low-achieving students (Nilholm 2005).Slee (2001) recognizes that many students might need additional resources in order to be able to fully participate and master classroom tasks, but stresses that this needs to be weighed against the risk of establishing new models of segregation.
On this point, Slee (2001) argues that we might need to ask 'inclusion into what?' since attempts to construct an education for all is also constrained by neoliberal standards policies, narrowing down the curriculum to facilitate testing and accountability structures.The focus on meeting academic performance targets thus risks classifying failing students as deviant and in need of special educational support (Slee 2019).Done and Andrews (2020) argue that the idea of full inclusion is incompatible with the pressure put on teachers to improve student performance in neoliberal education systems.From their perspective, attempts to define inclusive education without implying the need for fundamental changes to the educational system reconfigure inclusive education as surface adaptations and modifications of teaching and educational settings in order to meet students' academic needs.Thus, inclusive education becomes transformed into a matter of students' rights to gain access to mainstream classrooms and to be provided with support for reaching standardized targets of performance (Engsig and Johnstone 2015;Hamre, Morin, and Ydesen 2018).
The confrontation between inclusive education and the standards agenda is illuminated in some empirical studies that show how the former tend to be subordinated to the latter.Although many of the interviewed teachers in these studies were committed to inclusive education on a philosophical level, they did not tend to see it as an alternative.Instead, they made efforts to incorporate inclusive practices in teaching already shaped by standards policies (Dyson, Gallannaugh, and Millward 2003;Ainscow, Booth, and Dyson 2006;Molbaek 2018).This acceptance of the dominance of standards policies illustrates a movement between seemingly contradictory positions.Research has shown that such shifts between inclusive beliefs and pragmatic actions may cause ambivalence among teachers (e.g.Ineland 2015), but also that reflections upon such inconsistencies can be restrained by the heavy workload and constant confrontations with new problems and challenges that teachers face in their work (Ball, Maguire, and Braun 2012).

A dilemma perspective on inclusive education
A focus on dilemmas can account for complexities in the realization of inclusive education by illuminating goal conflicts and the different interests and contextual frames that influence how schools and teachers deal with such difficult decisions (Clark, Dyson, and Millward 1998;Dyson and Millward 2000).A dilemma can be distinguished from a problem in that dilemmas contain conflicts between different values, principles or goals.Whereas problems can be solved, decisions on how to handle a dilemma will not make the dilemma disappear.(Miller 2018).Any attempt to resolve a dilemma will lead to certain values, principles or goals being realized at the expense of others.Thus, it is only possible to find temporary resolutions to dilemmas (Clark, Dyson, and Millward 1998).
Researchers have described several different dilemmas in relation to inclusive education.Norwich (1993Norwich ( , 2002) ) suggest that these dilemmas specifically relate to three core questions within education, one of which is the issue of identification (also Minow 1985).This concerns questions about how schools can identify students in need of special support without stigmatizing them on that basis.One dilemma related to identification concerns the valuation of difference, and whether students' different needs should be understood in terms of differences or deficits (Nilholm 2005).The other two categories regard the organization of teaching and the questions of curriculum -what students should learn, and placement -where students should learn and with whom (Norwich 1993(Norwich , 2002)).For example, such dilemmas concern whether teaching should be organized around the principles of compensation or participation (Nilholm 2005), and how the needs of the individual should be balanced with the needs of the larger group, or with external perspectives on what students need to learn (Berlak and Berlak 1981;Tetler 1998;Molbaek 2018).
In their early empirical studies, Berlak and Berlak (1981) found that teachers' resolutions to dilemmas were sometimes consciously chosen, but sometimes seemed unreflective or shaped by external forces or internal needs of which they were not fully aware.In a similar way, Dyson and Millward (2000) emphasize that dilemmas can remain implicit and are thus not perceived as dilemmas by practitioners (e.g.Molbaek 2018).However, both unconscious and conscious responses to dilemmas are made in relation to contextual constraints and external factors such as policy demands and the perspectives of students, parents, colleagues and superiors (Dyson and Millward 2000).Thus, national or local policies and prioritizations can constrain and enable school actors' room for action.For example, systems of resource allocation based on category of disability will most likely affect how the dilemma of identification is resolved in schools (e.g.Kerins 2014).Moreover, a focus on standards, for instance in the form of high-stakes testing, has been argued to narrow down the curriculum and lead to compensatory teaching for lowattaining students (e.g. Kelly 2018;Smith 2018).In this way, dilemmas that have been resolved on higher levels might be less visible in the local school context.Although resolved dilemmas can seem like stable phenomena, they are products of multiple forces and processes that have found a temporary balance in a specific time and place (Clark, Dyson, and Millward 1998;Dyson and Millward 2000).Thus, it is important to study different interests, power relations and decision-making in relation to how dilemmas are perceived and resolved.

Resolutions to dilemmas as policy enactment
The theoretical work on policy enactment by Stephen Ball and colleagues (e.g.Ball, Maguire, and Braun 2012) provides a framework for analyzing actors' perceptions and responses to dilemmas in relation to policy demands, organizational structures and contextual frames.The concept of policy enactment rather than implementation illustrates a complex process of recontextualization where school actors make sense of policies and prioritize between conflicting demands in relation to local circumstances.However, Ball, Maguire, and Braun (2012) stress that this work is conditioned by material and personal factors and structures in the local school organization.Moreover, Ball, Maguire, and Braun (2012) argue that policies exercise discursive power by creating certain meanings that regulate the ways to think, talk and act.In this respect, they stress that different kinds of policy entail different conditions for enactment.Imperative policies produce reactive and passive policy subjects by creating formal and narrow frames for the work of local school actors.For example, standards policies tend to create teacher practices influenced by performance delivery, which leaves little room for reflective judgement.Educational ideals such as inclusive education are instead often promoted by exhortative policies, which leave more room for creativity and reflection.Thus, these policies can more easily be ignored, modified, or repackaged in creative ways.Ball, Maguire, and Braun (2012) also emphasize that the enactment of policies in schools is a collective process that involves a heterogeneous group of actors with different capabilities and inclinations to enact different policies in different ways.These differences are highlighted through the concepts of interpretations and translations of policy (See also Ball et al. 2011b).Interpretations have a strategic character and involve prioritizations between different alternatives in relation to the strength of specific policies and local contextual limitations.Translations are more practical recodings of policy that are shaped by the logic of classroom practice.Although translations can be highly creative, they are associated with a higher degree of compliance.Based on the interpretation and translation processes, Ball, Maguire, and Braun (2012) have constructed a typology of different policy actors within the local school organization (See also Ball et al. 2011a).This account of the policy enactment by policy narrators, policy transactors, policy translators, policy critics, policy receivers, and so on illustrates how different actors' responses to policy are conditioned by their position in the school organization.For example, Ball, Maguire, and Braun (2012) describe how members of the school management engage in initial interpretations of policies in order to integrate them into an institutional narrative -a story about the school and 'what we want to be' -through which policies are articulated to staff.These interpretations thus shape the ways policies are perceived by teachers (e.g.Shieh 2022).Whereas some members of staff are responsible for interpretations of certain policies or engage in collective translations and enforcement of policy demands, others are consumers of such translation work.To these, often newly qualified teachers, policies seem distant from their immediate concerns and can be difficult to separate from advices from colleagues, although policies bear heavily upon their practice.However, Ball, Maguire, and Braun (2012) stress that actors often move between interpretations and translations and respond differently to different policies because their authority to make decisions as well as their knowledge, experience, and commitment may shift between different aspects of their work.

The Swedish context
Sweden has traditionally had a reputation of being an inclusive educational system.However, in recent years Sweden has lost ground in international comparisons of equity in school systems (OECD 2018).It has been argued that an increased focus on educational standards in Sweden has promoted traditional special education as an answer to failures in attaining required levels of knowledge (Isaksson and Lindqvist 2015).This development can also be seen in the light of the marketization of the Swedish school system during the last decades, where the introduction of free school choice based on a voucher system has been followed by school competition and increased social segregation (e.Magnússon 2020).Furthermore, empirical studies have illuminated tensions between Swedish general education teachers and special educators concerning their views on inclusive education.In this regard, special educators have advocated adjustments of teaching and environmental aspects, and have tended to be more skeptical towards students being taught outside of the mainstream classroom (e.Lindqvist et al. 2011).General education teachers on the other hand have expressed a need for more time and more support from special educators in order to fulfill their tasks of teaching students with different needs in their classrooms (e.g.Sandström, Klang, and Lindqvist 2019).Although these findings might not be unique for the Swedish case, they illustrate some of the complexity that permeates the enactment of inclusive education.
National regulations of special support in the Swedish educational system are primarily found in the Education Act (SFS 2010, 800).Within mainstream education, these regulations regard two different levels of support.If a student risks not achieving the minimum demands of the curriculum, the general education teacher is obligated to provide extra adaptations for that student.These can consist of modifications and adaptations in the student's learning environment within the frames of general classroom teaching.No formal decision is needed for this level of support (Swedish National Agency for Education 2022).If the extra adaptations are not successful, an investigation into the need for special support shall be initiated.Special support shall be documented in an individual education plan, since it involves more comprehensive measures, for example, interventions from a special educator.However, it is stated in the Education Act that permanent teaching outside of the frames of mainstream classroom teaching should be avoided if possible (SFS 2010, 800).
General education teachers are thus the professional group with the main responsibility to teach students with different needs.In this work, they are supported by special education teachers and special educational needs coordinators (SENCOs).Both of these professions are intended to work with pedagogical investigations, writing individual education plans, and engaging in preventive work to remove obstacles in the learning environment.Both professions are also supposed to engage in different forms of support for general education teachers as well as direct teaching of students in need of special support, although the work of SENCOs is more directed towards the organizational level and to school development (SFS 2007, 638).In this article, the term 'special educator' will be used when referring to SENCOs and special education teachers together as a broader professional category.

Materials and methods
The empirical material in the current study consists of interviews with eight subject teachers and four special educators at three lower secondary schools in two Swedish municipalities.In the Swedish K-12 school system, the lower secondary school is the final stage of compulsory education, encompassing the school years 7-9.While grades are introduced one year prior to this level of education, the transition to the lower secondary school also means that the class teacher is replaced with several subject teachers.Thus, these school years were considered a suitable context for studying conflicts and prioritizations between different values, principles and goals in schools´ enactments of inclusive education.
In the selection of schools and participants, it was considered desirable to focus on a few school cases that might offer insights into the particularities of each case, but also provide variation in terms of respondents and the contextual conditions in their schools.In the selection process, statistics from the Swedish national agency for education was used to search for lower secondary schools that varied in terms of contextual factors that were important to the study.As can be seen in Table 1 below, the three schools that participated in the study were all public schools but differed in terms of school size and location.In school 2 and 3, initial contact was made with the principal who then forwarded the request to the staff.In school 1, an initial interview was conducted with a special educator who then helped to establish contact with new participants.The interviewed teachers taught in various school subjects such as mathematics, science, the Swedish, English, and German languages, social studies, physical education, art, and crafts.
The data collection was carried out between October 2021 and March 2022.In accordance with the Swedish Ethical Review Act (SFS 2003, 460) ethical approval was not considered to be required for this study.No sensitive personal data were collected, as specified by the EU general data protection regulation (European Commission 2016).Ethical considerations were made with respect to the Swedish Research Council's research ethics principles (Swedish Research Council 2002, 2017).All respondents who accepted to participate in the study signed a consent form with information about the study and ethical considerations, such as their right to terminate their participation in the study at any time and that the empirical material would be anonymized and used only for research purposes.
Each interview lasted about 45-60 minutes and was audio-recorded with the permission of the participants.The interviews were designed to explore the respondents' School 1 (a small, public school in a rural municipality) 1 special educational needs coordinator 1 special education teacher 2 subject teachers School 2 (a medium-sized public school in an urban area) 1 special educational needs coordinator 3 subject teachers School 3 (a large public school in an urban area) 1 special education teacher 3 subject teachers experiences of teaching students with different needs and how this work is organized at their schools.By offering the respondents a space to talk about these issues, the interviews aimed to provide insights into how dilemmas related to inclusive education take shape in the school context and how local school actors resolve them.This entailed an interest in how the respondents explicitly described dilemmas that they face in their everyday work, but also in tensions that reveal implicit dilemmas, for example such that have already been resolved by other school actors (Dyson and Millward 2000).The open-ended questions asked in the interviews were therefore followed up by requests for clarifications or elaborations of descriptions related to such tensions.Shortly after each interview, the recording was transcribed and any information that could be used to identify individuals, schools, and/or locations was anonymized.The audio recordings were then stored on an encrypted USB drive in a locked space inaccessible to unauthorized persons.The interviews were analyzed by thematic analysis consisting of the following steps inspired by Braun and Clarke (2006): 1. Becoming familiarized with the data: The analysis of each interview started with a thorough reading of the interview transcript and a summarization of the preliminary interpretations of meanings and patterns in the interview.
2. Initial coding: In this phase, text extracts were coded into different categories in order to organize the data into meaningful groups.These categories reflected topics that emerged in the interviews, for example, communication between school actors, local routines and regulations, responsibilities of different actors, classroom priorities, and contextual factors.
3. Searching for dilemmas: The focus was then directed towards the respondents' orientations towards conflicts of values, principles or goals that they face in their everyday work.In this effort, dilemmas described in previous research were used to compare with the respondents' descriptions.However, this step also involved a responsiveness to other tensions and conflicts or aspects described by the respondents as problematic or difficult in order to identify new or implicit dilemmas 4. Reviewing and refining: This phase included an examination of whether the candidate dilemmas that had been identified were coherent and distinct and whether they had sufficient support in the data set as a whole.In this respect, dilemmas were considered to be sufficiently supported if they were explicitly or implicitly described by respondents in all three schools.
5. Final analysis: In this step, theoretical notions and concepts from Ball, Maguire, and Braun (2012) were used to discuss resolutions to dilemmas as enactment of inclusive education.This part of the analysis aimed to shed light on how different actors engage in the work of resolving dilemmas, and how their perceptions and responses to dilemmas are framed in their enactment of conflicting policy demands to their local contexts.

Findings
Four different dilemmas were identified in the analysis: • Special vs. general education settings • External control vs. professional freedom • Curricular demands vs. students' needs • Individual vs. group In the following, the four dilemmas are presented and analyzed in relation to the respondents' descriptions of their everyday work and the organization of teaching in their schools.

Special vs. general education settings
A commonly brought up issue in the interviews was whether students benefit from receiving special educational support outside the framework of regular teaching or whether this alienates them from the classroom community.Although many respondents expressed their own views on this issue rather than talking about it as a complex dilemma, their discussions reflected a conflict that corresponds with the placement dilemma described by Norwich (1993Norwich ( , 2002Norwich ( , 2008)).Swedish policy regulations of special educational activities in schools state that decisions on whether students should receive teaching outside of the mainstream classroom may not be delegated by the principal to anyone else (SFS 2010, 800).Thus, resolutions to this dilemma were shaped at the schoolwide level, and functioned as local policies enacted by teachers and special educators.However, there were clear differences between how the two professional groups engaged in this enactment.The subject teachers often described resolutions as institutionalized structures and procedures in their schools, thus positioning themselves as receivers of policy (Ball, Maguire, and Braun 2012).The special educators on the other hand often used a strategic language in their descriptions of these local policies.Since the special educators are important actors in the organization of special support in their schools, they can be described as policy transactors, responsible for the facilitation and enforcement of certain policies (Ball, Maguire, and Braun 2012).While all special educators stressed that the issue of placement has no universal solution, they tended to problematize the use of special support groups.
I think that teachers' knowledge about neurodevelopmental disorders and things like that needs to increase.To get away from the idea that it is a certain group of students you teach at a certain level.That you dare to take care of students in the classroom instead of sending them away.(SENCO, School 2).
Nevertheless, all three schools had established small teaching groups, for example aimed towards students with neuropsychiatric diagnoses.Respondents in school 2 and 3 also described that they had flexible small groups where students could receive special support during a few lessons each week.In school 1 there were currently no such groups.However, respondents in all three schools emphasized that the different needs of students were primarily expected to be met through subject teachers' adaptations of mainstream classroom teaching.The subject teachers varied in their views of special support in smaller groups, but they tended to discuss these issues in relation to their everyday work rather than discussing inclusive principles.
I have so many students who appreciate being able to leave the classroom and work in peace and quiet.Maybe initially they say "no, I don't want to".But then they notice that this is exactly what they need.You need a combination.Sometimes support in the classroom and sometimes you need to go somewhere else and work with another teacher.I don't think it seems to be that stigmatizing.(Art and Swedish language teacher, School 2).
In general, the interviewed subject teachers more often reflected upon consequences that followed school-wide level resolutions to the dilemma than they discussed the dilemma itself.In this regard, teachers in all three schools expressed that they needed more support and resources in order to be able to teach all students in their classrooms (cf.Sandström, Klang, and Lindqvist 2019).Although they were recommended to turn to special educators for guidance when they found that their adaptations of teaching were insufficient, several teachers mentioned that they instead often sought initial support from teacher colleagues.This was partly because these colleagues were closer at hand, but also because the opportunities to receive instant support from special educators were perceived as limited due to the lack of time and resources.
They can give me consultations.That's about it.We can arrange student health team meetings where we can discuss how to proceed.However, I usually contact the class mentor instead to brainstorm ideas.If extra adaptations do not work, you can bring in the student health team.But they have such a large group of students to take care of.Some students I may not take to the student health team because it can take such a long time before I get help from them.(Mathematics and PE teacher, School 3).
Although there were some differences between the interviewed subject teachers and special educators regarding their views on the dilemma and their responses to schoolwide level resolutions, disagreements were subtly expressed in the interviews.In many cases, respondents also stressed an understanding of other perspectives than their own, thus showing recognition of the dilemmatic character of these issues.

External control vs. professional freedom
It can be argued that teachers need a certain degree of autonomy when adapting teaching to students' different needs.However, it is not certain that autonomous teachers can or choose to teach in inclusive ways.This conflict was surfaced by the respondents in different ways in the interviews, and reflected a dilemma between external control of teaching and professional freedom while trying to establish more inclusive practices.Some of the interviewed special educators discussed this dilemma from a school-level perspective in relation to their own efforts to develop accessible learning environments in their schools.
We have had such discussions about fixed placements and so on.And the same beginning and end of lessons, etc.And I don´t really think that they want it.They have probably found their way.I don´t think all teachers feel that way.But I think that if our principal were to decide on this, they probably wouldn´t follow it.(SENCO, School 1).
As illustrated in the quote, attempts to formalize teaching seemed to be difficult in school 1.Respondents in school 2 and 3 on the other hand reported that a formalized lesson structure had been introduced in their schools in order to create continuity and clarity and to reduce the need for individual adaptations.In general, the interviewed teachers and special educators emphasized positive aspects of the formalized lesson structures, although some teachers called attention to the risk that teaching will become too static and impersonal.In the interviews with subject teachers, the dilemma was generally not discussed per se.Instead, they reflected upon the consequences of its resolutions.In particular, the dilemma of external control vs professional freedom became visible in teachers' descriptions of extra adaptations for particular students that were expected to be implemented across school subjects by all of the student's teachers.These extra adaptations were designed in dialogue with students and parents, generally by the teacher assigned as mentor for the specific class or by a special educator.Whereas some of the special educators and teachers described how they enforced the implementation of extra adaptations for specific students, other teachers described how they responded to these demands in their everyday work.Some of the more experienced teachers emphasized that they did not always think that formalized extra adaptations fit well with their school subjects.
The special educators go through extra adaptations in this folder: that this student does not like to sit there and so on.I attend these meetings because I teach the classes, but I rarely put that much energy into it.I try to be neutral, because it is often not true, or in here it's not true.(Crafts teacher, School 3).
Based on these responses, the formalized extra adaptations can be described as what Ball, Maguire, and Braun (2012) call exhortative policies, which leave room for critical interpretation and agency in their enactment by teachers.However, not all teachers seemed to utilize this space for action.Instead, some teachers described how they struggled to incorporate extra adaptations into their teaching, here illustrated by a recently graduated teacher.Some students have documented extra adaptations: this student should sit at the front of the classroom and be shielded from visual distractions.I can´t keep everything in my head . . .I find that difficult.That's a limitation for me -to know how to adapt teaching.What is good and bad teaching for this specific student?I don´t know.When I introduce a new subject area, it's the first time for me.Then I'm busy figuring out what to do.(Swedish and German language teacher, School 1) Overall, many of the interviewed teachers seemed to accept constraints of their room for action in exchange for guidelines on how to adapt their teaching to specific students' needs.Thus, the dilemma of external control vs professional freedom became visible in the teachers' responses as a tension between their need for creativity and their need for guidance (cf.Ball, Maguire, and Braun 2012, 65).Whereas the perceived need for guidance in relation to extra adaptations call for special education expertise on how to teach certain students, the formalized lesson structures were described as based on another form of authoritative knowledge, rooted in the idea that 'good teaching is good teaching for all'.In both these cases, the responses to the dilemma and its resolutions varied between teachers, especially regarding their orientation towards creativity or guidance and their resistance towards local policies.

Curricular demands vs. students' needs
This dilemma emerges from some of the complexities that arise in the teaching of a common curriculum to a heterogeneous group of students (Norwich 1993(Norwich , 2002)).As this dilemma took shape in the respondents' descriptions, it highlighted a conflict between students' individual needs and the demands of the curriculum (cf.Berlak and Berlak 1981;Molbaek 2018).National policies directing the focus towards knowledge goals had a strong, although often implicit, presence in the talk of respondents from all three schools.However, some respondents explicitly described a pressure to cover the teaching content defined in the syllabus and to reach certain results, which could constrain their explorations of new and creative ways of teaching and their attentiveness to students' broader needs and interests.
So many of us are tied up with grading matrices -the amount of knowledge goals that must be achieved in three years.We dare not try -because what if time passes without it producing results.And it doesn't have to be a difficult class for one to think that way.It is easier not to be brave because then you only lose two or three students.It might sound crass, but I think that is what happens.(Swedish and English language teacher, School 1).
Many of the interviewed teachers gave accounts of a stressful job where they needed to teach much content in a short amount of time.Moreover, they expressed general concerns about unmotivated students, or students who cannot handle the level of difficulty in the teaching.However, most of them did not problematize standards policies, but talked about them as given frames for their work.Even when teachers identified standards policies as conflicting with individual students' needs, they did not seem to have the authority to resist them.Thus, these imperative policies seemed to produce narrow responses to this dilemma (cf.Ball, Maguire, and Braun 2012).In this respect, some teachers described how they tried to find space for alternative ways of teaching within a standards framework, for example by engaging in informal formative assessment (cf.Dyson, Gallannaugh, and Millward 2003;Ainscow, Booth, and Dyson 2006;Molbaek 2018).In general, the interviewed special educators were more directly critical towards standards policies.Since special educators have very specific roles in schools as experts and facilitators in relation to the teaching of students in school difficulties, they embody certain policies, but might not be as affected by others (Ball, Maguire, and Braun 2012).Although the special educators could not escape the standards agenda, they could thus talk about this dilemma from a somewhat outside perspective.
All the time it is like, "I have to go through this subject area, and this and that must be tested".And then you rush forward and don´t have time to adapt the teaching to those students who don´t understand.I think it's a bit like that . . .For some, not all teachers.In order for these tests to be taken, the teaching must be at a certain level.And it should, but . . .Some tasks that students get . . .They come to me with their tasks and they don´t understand what to do, and the tasks can be at university level sometimes.(SENCO, School 2).
While most teachers tended not to emphasize the conflict between standards policies and individual students' needs, many of them highlighted that contextual constraints complicated their adaptations of teaching.In particular, large class sizes were mentioned as a major obstacle.
In my second subject, the German language, where I have about 12 students in ninth grade, there are much greater opportunities.In English, we work in a quite standardized way: everyone works with this part, we study for this test.In the German language class maybe we can watch a series, we can explore historical characters, we can write dramas . . .A much more stimulating teaching.(English and German language teacher, School 2).This quote highlights another pattern found in the interviews.Teachers of subjects such as crafts, art, and modern languages did not seem to be as pressured by imperative standards policies as teachers in mathematics, science, Swedish, and English.While not being 'in the direct gaze of policy' (Ball, Maguire, and Braun 2012, 69), these teachers, particularly those who taught in practical-aesthetic subjects, also perceived individual adaptations as easier to integrate in their teaching because of the creative process that characterizes their subjects.

Individual vs. group
Subject teachers in all three schools reported that their working conditions constrained their opportunities to adapt their teaching to all of their students' different needs.This caused a dilemma where they constantly had to prioritize between different students.In particular, this conflict concerned the attention teachers gave to individual students with the most difficulties in relation to the whole group of students (cf.Tetler 1998;Molbaek 2018).Despite not being pronounced in the interviews with special educators, this was the dilemma most often brought up explicitly by the subject teachers.They were also the actors resolving the dilemma in their classrooms, although many of them stressed that contextual factors such as large class sizes and an increased teaching load made this work a great challenge.
It is very difficult in a class with 30 students.A student can focus on the task when you stand next to him and tell him what to do."You can write this" . . .and then "look at the next task, I'll be back".But nothing happens.When I help other students, that student gets nothing done.(Mathematics and science teacher, School 2).
The need to create a functioning classroom environment where students can work without being disturbed by their classmates meant that the teachers often felt that they needed to prioritize the students who were acting out in order to put them 'on task'.Moreover, many teachers perceived that they were obliged to focus on low-attaining students in order to help them reach the expected knowledge goals.Thus, the attention the teachers felt they needed to give these students in the classroom was directed towards the aims of standards policies.These prioritizations had the consequence that the students who were able to manage on their own were often left to do so.This reluctant neglect of high-attaining students was something that several teachers spontaneously raised in the interviews as a problem.
Last week there was a girl working on the more difficult tasks for an upcoming exam, and then I had another student struggling to get an E. Who should I prioritize then?I know she will get an A anyway.But she really wanted to discuss these assignments.(Mathematics and PE teacher, School 3).
In the interviews, the teachers described several different ways of resolving the dilemma in their everyday work.Whereas some teachers used in-class ability grouping, others adapted their whole-class teaching to the students in greatest need of support through repetitions, visualizations, and examples or by adding short activities in their lectures to capture the attention of students with concentration difficulties.These descriptions highlight examples of deliberately, rather than unconsciously, chosen resolutions to the dilemma, where the teachers reflected upon potential negative consequences of their prioritizations.
There is a risk of lowering the level.If you are going to read a text, you may know that some will never be able to get through it by themselves.That you have to read it aloud because otherwise you have lost some of them.And then you read it aloud to everyone.It´s an adaptation for a few students, but for the others it will be -It´s not certain that it´s bad for them, but they never get to engage in difficult texts.(Swedish and German language teacher, School 3).
As illustrated above, this dilemma was in focus for many teachers while discussing their teaching of students with different needs.It can be argued that the pressure to prioritize passing grades for all students is primarily directed at the school level, and then transmitted to teachers through authoritative interpretations reflecting the institutional priorities of the school (Ball, Maguire, and Braun 2012).Many teachers did describe a pressure to act in certain ways when resolving the dilemma in their classrooms, but they were the ones who made the decisions.In comparison with the other dilemmas, the teachers thus responded to this dilemma more actively.By reflecting upon what they needed to do and the consequences of their choices, they engaged in critical interpretations that included not only 'recoding' but also 'decoding' of policy (Ball, Maguire, and Braun 2012, 43-45).

Discussion
This article set out to explore how Swedish special educators and subject teachers perceive and respond to dilemmas in their enactment of inclusive education alongside conflicting policy demands.By combining theoretical notions of dilemmas and policy enactment in the analysis, the aim of the article was to contribute with knowledge relevant to both fields of research.
To some part, the findings reflected previous research that show how teachers' work is framed by standards policies, which constrain inclusive ambitions (e.g.Ainscow, Booth, and Dyson 2006;Molbaek 2018).However, the use of policy enactment as a theoretical lens enabled an illustration of a complex process where different actors' perceptions and responses to dilemmas and their resolutions are influenced by their organizational position, contextual frames and policy pressure in different ways.Thus, the conceptual tools provided by Ball, Maguire, and Braun (2012) allow for a partly new understanding of such processes in relation to previous research.
An important finding in this respect was that dilemmas that were resolved on different levels in the local school organization seemed to be perceived and responded to in different ways by different actors.The first two dilemmas -special vs. general education settings and external control vs. professional freedom -were primarily resolved at the school-wide level and transformed into local policies or institutionalized procedures.Whereas the interviewed special educators often reflected upon the conflicts of values, principles or goals in these two dilemmas and sometimes engaged in strategic discussions from a school-level perspective, teachers discussed established resolutions to these dilemmas and the consequences that followed for their everyday work.The teachers' responses illustrated both compliance and resistance in their enactment of these local policies.
The two other dilemmas -curricular demands vs. students' needs and individual vs. group -were resolved by teachers within the frames of classroom practice.Whereas there were some differences in how the two dilemmas on the school-wide level were resolved in the three schools, no clear differences were identified across schools with regard to teachers' responses to the two classroom dilemmas.It can be argued that the external context, described by Ball, Maguire, and Braun (2012, 36) as 'pressure and expectations generated by wider local and national policy frameworks', was very similar at the three schools.Some respondents explicitly highlighted standards policies as constraining their resolutions to dilemmas.However, most of the interviewed subject teachers seemed to comply with these imperative policies.Although some teachers found standards policies problematic and reflected upon alternatives, they did not seem to have the authority to resist them.Thus, their responses included elements of both critical interpretations and compliant translations.
While policy enactment can be understood as a process of recontextualization to practice alongside conflicting policies and in relation to contextual frames and local concerns, dilemmas highlight the educational matters at the center of this process.Thus, the combination of notions of dilemmas and policy enactment in the analysis also showed to be an addition to the theoretical reasoning of Ball, Maguire, and Braun (2012).In particular, concepts such as interpretations/translations, policy actors and imperative/exhortative policies gained an educational dimension by being related to central conflicts of values, principles and goals in schools.In that way, a more complex illustration of how school actors enact policy in relation to important areas of their work was enabled.
For example, Ball, Maguire, and Braun (2012) emphasize that the notion of different policy actors in schools does not mean that policy actors are static roles that are attached to certain individuals.Nevertheless, the focus on dilemmas in this article shed more light on how actors move between different roles and different kinds of enactments in relation to the educational matters and tensions with which policies intervene.In this regard, the impact of standards policies seemed to differ between different dilemmas, and standards policies seemed to influence perceptions and responses to different dilemmas in different ways.For example, teachers reflected upon the constraints imposed by policies and contextual factors on their resolutions to the dilemma of individual vs group, as well as the consequences of their resolutions to this dilemma.In comparison, the dilemma of identification was not identified in the interviews although being considered a central dilemma in research literature (e.g. Minow 1985;Norwich 1993Norwich , 2002;;Nilholm 2005).This can be understood as a consequence of standards policies that connect the concept of special needs to the attainment of predefined knowledge goals (cf.Isaksson and Lindqvist 2015;Slee 2019), and which restrain not only actions but also reflections on whether identification of these students is desirable or not.In this way, the absence of the identification dilemma can be argued to reflect a deficit perspective, where school difficulties are tied to individual shortcomings, which would be a major barrier to the realization of inclusive education in schools (Done and Andrews 2020).
It should be pointed out that the main aim of this article has not been to identify an inventory of dilemmas but rather to argue for the importance of studying how they are played out in practice in order to gain a more nuanced understanding of the conditions for realization of inclusive education.Thus, studies in other contexts may find other dilemmas or other perceptions and responses to the dilemmas identified here.The findings from this study suggest that such research on dilemmas needs to take into account the external demands, organizational structures and contextual constraints that frame how dilemmas are resolved in schools, and how they are perceived and responded to by different school actors.In a similar vein, the study suggests that research on policy enactment in schools benefit from a focus on dilemmas in order to more explicitly address the work done in schools and the conflicts of values, principles or goals that permeate school actors' recontextualizations of policies.

Table 1 .
Participating schools and respondents.