Exploring scientific, artistic, moral and technical reflection in teacher action research

Abstract

Reflection in action research is a complicated matter because of the many domains of reflection and most significantly, the lack of understanding of these domains of reflection in action research and how these are supported. In this paper, we propose a framework based on four domains of reflection, namely, scientific, artistic, moral and technical reflection. We describe an initial attempt to use this framework in relation to the actual practice of teacher reflection in action research and show that the framework allowed us to map the various domains of reflection that teachers use in relation to their action research. This helped us to gain insight into the differences and the course of reflection in action research. We discuss how the framework – through orientation, differentiation and deepening – might provide support for reflection in action research.

Keywords:

• Action research
• teacher reflection
• complexity

1. Introduction

Action research in education is meant to expand ‘the space of the possible’ (Davis and Sumara 2005, 458), in teaching as well as in reflecting upon teaching. According to some authors, reflection is the most important goal (Stenhouse 1975), the most important characteristic (Carr and Kemmis 1986) or the sine qua non (Editorial 2011) of action research.

Although reflection is deemed important in relation to action research, it is also considered to be a complex matter. Vetter and Russel (2011, 180) illustrate this complexity when they quote a teacher who describes action research as ‘taking a cross-country journey with a world map’. This teacher is referring to the experience of a continuous process of comparing, weighing and orchestrating a range of often contradictory beliefs without a frame of reference to hold on to. Other authors describe similar experiences in terms of a negotiation of ‘different voices’ (Bakhtin 1981) or as ‘borderline discourses’, or an ‘in-between ground of ambiguity’ (Alsup 2005).

Reflection in action research is complex, first of all because of the many beliefs among which you have to ‘choose’, but most of all because of the lack of a frame of reference that can give support when making such a choice. In this article, we propose a framework that offers an overview of different domains of reflection and their mutual relations. It is meant to serve as a map of reflection in action research. Just like a real map, it can help the user to determine where (s)he is and what the possible roads are.

In the literature concerning teacher reflection in general, and action research in particular, we see on the one hand an urge to broaden reflection, which entails using multiple forms of reflection and relating them to each other (Kelchtermans 2009; Leitch and Day 2000). This means not just more attention to scientific reflection that aims to bring theory and practice closer to each other (Kemmis 2009), but also more attention to moral and aesthetic reflection, in order to do more justice to the role of values in education (Carr 2006; Elliot 1987; Kim 2013). On the other hand, we see that broadening reflection raises a number of problems. For example, scientific reflection – which addresses the gap between theory and practice – is not about a gap between an educational theory and a complicated practice but, instead, is about a gap between a great variety of educational theories and a complex educational practice (for example, Agee 2004; Bernstein 1999; Luttenberg, Meijer, and Oolbekkink 2016; Rex and Nelson 2004; Wignell 2007). When moral and ethical reflections are considered, we see that, in comparison with technical instrumental reflection, they follow another route and with respect to content and meaning can lead to other conclusions (for example, Bridges 2001; Carr 2006; Elliot 1987; Kim 2013; McNiff 2011; Risser 1997), and as such contribute to the inherent complexity of reflection in action research.

In sum, developing reflection in action research is considered something worth striving for, but at the same time is also seen as a complicating issue because diverse domains of reflection, differing in character and outcome, have to be aligned and because a choice has to be made between diverse beliefs within the different forms of reflection.

There has been a growing tendency in recent years to understand this complexity from the framework of complexity science (Ahmadian and Tavakoli 2011; Davis and Sumara 2006; Phelps and Hase 2002; Sumara, Davis, and Carson 1997). Complexity, as a term, has been used in recent years in relation to action research, but reflection in action research was never explicitly elaborated as a complex matter. An example of this is the special issue of Educational Action Research in 2011 about reflection in action research (Editorial 2011). This special issue offers significant examples of the importance, diversity and complexity of reflection in action research but lacks descriptions and interpretations of this complexity. This might be developed by making an explicit connection with complexity science. Elsewhere, we elaborated in which sense reflection in action research can be considered a complex system (Luttenberg, Meijer, and Oolbekkink 2016).

In the present article, we propose a framework that helps to better understand this complexity. This framework is based on literature about domains of reflection which are useful in the context of education. Also, we elaborate on the conditions that influence the course of reflection in action research (diversity, redundancy, interaction, decentralized control, improvisation/conversation and minding). The outcome is a heuristic framework in which four forms of reflection and the conditions for possible courses of reflection are central.

In what follows, we describe an initial attempt to validate and illustrate this framework by examining the applicability of the framework to the actual practice of teacher reflection in action research. It appears that the framework allowed us to map the different forms of reflection that teachers use in relation to their action research and has provided insight into a possible course of reflection in action research.

2. Theoretical background

2.1. Framework of reflection as complex system

Reflection in general is considered important for teachers to help them deepen their professional development and thus enhance their competences, to help them cope with difficult aspects of the profession and to allow them to find solutions in their own practice to problems which experts cannot solve. Luttenberg and Bergen (2008) show that various views on reflection can be distinguished which either focus on the content of reflection or on the nature of reflection. They indicate the importance of taking both content and nature of reflection into account because these are intricately intertwined and cannot be separated; we therefore focus on domains of reflection in this article.

In this study, we distinguish four domains of reflection – scientific, technical, artistic and moral reflection. This distinction is based on the scientific, technical, moral and artistic traditions as described by Coldron and Smith (1999). These four traditions represent the functional problems that reflection in action research is dealing with constantly. They are about finding answers to questions such as what is true (scientific reflection), what is effective and efficient (technical reflection), what is good (artistic reflection) and what is just (moral reflection) (see Coldron and Smith 1999). These questions relate to different content. Scientific reflection is about generalizable insights that are the result of scientific research activities; technical reflection is about education as a whole of means and skills to achieve certain goals. Artistic reflection is about the personal signification of the teacher in a real situation of his or her practice; for example, in everyday classroom interaction. Moral reflection is about general values that apply equally to everyone in every situation.

These four distinct domains of reflection have a solid foundation in western thinking and western culture (Habermas 1991) and play an important role in education (Coultner 2001; Oser 1994). They are closely related to the core components of professionalism, identified for teachers in the work by Furlong et al. (2000), which include knowledge, autonomy and responsibility. Knowledge relates closely to scientific and technical reflection; autonomy to aesthetic reflection; and responsibility to moral reflection. As Furlong et al. (2000) indicate, autonomy involves the personal judgement of a professional, and responsibility involves the judgement a professional has to make based on both personal values and other people’s values. Finally, knowledge involves the body of specialized knowledge that professionals have and use. The four domains of reflection also seem to cover the domains of reflection that teachers use for making difficult decisions in education (for example, Luttenberg and Bergen 2008; Luttenberg, Imants, and van Veen 2013).

For the framework of reflection in action research (see Figure 1) we identified two dimensions underlying the four domains of reflection. These dimensions are based on two distinctions relevant for action research and for education in general: justification–application and knowledge–value. The first dimension (justification–application) represents the distinction between more general theoretical and more concrete personal and practical considerations that play a role in teacher reflection. This dimension refers to the gap between theory and practice often referred to in action research, and the gap between the roles of theorist and practitioner (see Kemmis 2009). This distinction not only refers to the distinction between educational sciences as an academic discipline and educational practice, but also to the distinction between the moral and ethical theories important for education and the application of these theories in education. In relation to the domains of reflection, both scientific and moral reflection focus on justification based on more general theoretical insights. On the other hand, technical and artistic reflection are more concerned with personal and practical considerations.

Figure 1. Dimensions and domains of reflection in action research.

The second dimension (knowledge–value) represents the distinction between value-free knowledge – which is, in principle, objective and neutral – and value-related knowledge – which is subjective in nature, relates to social values and is biased. This dimension refers to the distinction between education as a technical–instrumental matter versus a moral–ethical matter (Bridges 2001; Carr 2006; Kim 2013). Education as a technical–instrumental matter stresses the importance of evidence and so-called objective guidelines, while the moral–ethical aspect of this dimension acknowledges the value-related aspect of education. In the first case, it is about insights that in principle do not oblige to anything, and are, in that sense, non-committal. With respect to the second case, it is about insights that do oblige and thus have to be respected.

2.2. Conditions for the development of reflection as a system

Davis and Sumara (1997) describe a complex system as a coherent whole of parts that is able to maintain itself as a relative autonomous entity in the face of changing circumstances. To survive as a system, a constant interaction and exchange of matter, energy or information between the parts and between the parts and the context must occur. Because of these interactions and changes, the whole is more than its composing parts. According to Davis and Sumara, complexity science tries to understand how a coherent and purposive whole emerges from the interactions of simple and sometimes non-purposive components.

Although complex systems are often too complex to understand completely from the knowledge of the composing parts or the interactions and exchanges between them, it is often possible to point at conditions that contribute to the development of the system (cf. Davis, Sumara, and Luce-Kapler 2000; Davis and Sumara 2005). Davis and Sumara (2005) mention a number of conditions that apply to action research, which can in turn be applied to the framework depicted in Figure 1, as follows:

•	Sufficient diversity in relation to the issue of reflection on action research and the means to express that diversity. Here, we indicate that there has to be sufficient variation in questions and solutions to make an exchange between domains useful.
•	Adequate redundancy between the domains of reflection in relation to the issue of action research. We mean that the different considerations teachers have in relation to their action research should be dealt with in such a way that these considerations relate to each other and that alignment is possible.
•	Sufficient interaction between domains of reflection and sufficient mechanisms to prompt these forms of reflection to interact. In order to be of added value for the system as a whole, exchange between the forms of reflections should not be a single action, but instead a continuous process of exchange of questions and answers. Moreover, there should be an incentive to progress through that exchange with regard to possibilities and answers.
•	Decentralized control, which implies that the content, course and outcome of the exchange between the domains of reflection should be sufficiently free in order to be unexpected or unplanned. Learning in a complex system is theoretically a matter of all subsystems. The focus should not be on just one of the subsystems, but on the possibilities of all subsystems together. Decentralized control is necessary to make sure all subsystems play a role in the exchange.
•	Improvisation/conversation – reflection should be a case of ongoing engaged attention and responsibility of all domains of reflection for a joint project. It should be a process in which unexpected, unforeseen possibilities can arise through insights from various forms of reflection together. The exchange between these forms should, as in conversations in the traditional meaning of the word, in principle be open and with mutual care.
•	Minding, referring to a conscious and attentive way of following the process of exchange of different forms of reflection.

3. The framework at work: a comparative case study

The framework of reflection depicted in Figure 1 and the conditions described in the previous section imply that the development of reflection involves an ongoing exchange between domains of reflection that simultaneously, mutually and continuously affect one another. This means that the exchange between the different elements of the proposed framework is not sequential in nature. The framework can help to provide insight into the functioning of the different domains of reflection in conjunction with one another and, this being the case, can help to better understand the relation between the domains of reflection as a system in the context of action research.

In this section, we illustrate the course of reflection in action research and the relations between this course and the conditions for the development of reflection as a complex system. We performed a comparative case study of two teachers doing action research in their own schools. Both teachers had the opportunity to do action research on their own practice, together with a group of their colleagues. The two teachers on which this study was based researched an aspect of their own practice as a teacher. A supervisor from the university was involved in the supervision of all teachers in both schools. Workshops were organized in one school for all teacher researchers around different aspects of action research, which involved dialogue between the teacher researchers around their action research projects. The first author was involved in the supervision and organization of the workshops in one of these schools. This study is based on three interviews held throughout the teachers’ one-year action research project. The two teachers were selected from a larger group of teachers, because they appeared to differ in their breadth and variation of reflection. These criteria seemed relevant because we aimed to better understand how diversity of forms of reflection works as an important condition for the development of reflection as a complex system. The analysis of this comparative case study followed the method of the case-ordered descriptive meta-matrix (Miles and Huberman 1994). We used a mixed strategy of both a case-oriented approach and a variable-oriented approach (Ragin 1987) and also employed contrast tables as an exploratory device (Miles and Huberman 1994).

For both teachers, a synchronous reflection profile was constructed which involved a schematic overview per phase of the action research process and the forms of reflection that were used in this phase. To be able to do so, we selected for each fragment the form of reflection in relation to the phase of the research process. The results were ordered in a matrix that offered for both teachers an overview of all forms of reflection they used in relation to action research at the beginning, middle and end of their action research process. Based on this overview, we further reduced the data for each teacher by clustering similar forms of reflection in each phase. This resulted in a new matrix in which, for each teacher and for each phase, a profile was made of relevant forms of reflection. This profile can be interpreted as an indication for reflection in relation to action research in each phase.

As a second step, we made a description for each teacher of the continuities and discontinuities between the consecutive profiles. We refer to this as the diachronic reflection profile. Therefore, we made for each teacher an overview of the similarities and differences per profile for the three phases and noted the most important changes in those profiles. For both teachers, this resulted in an overview of the course of reflection throughout the action research process.

As a third step, the developments described in the diachronic reflection profile are related to the conditions that played a role. For each teacher, this resulted in an overview of the role of the conditions in the development of reflection in relation to action research.

4. Course of reflection in action research

For both teachers, we characterize their synchronous reflection profiles, as illustrated by representative quotations from the interviews with them. These quotations are grouped in tables that refer to the proposed reflection framework. But first, we briefly characterize the two teachers in our study, Mary and Mark:

Table

Mary has worked as a geography teacher for five years. Before that, she worked for a number of years as a development worker in Third World countries. In her action research she focuses on developing constructivist teaching methods in relation to student outcomes. She has been interested in this for a long time but felt she needed more experience as a teacher to be able to do this.

Mark has worked as a chemistry teacher for three years. Before that, he was a researcher and developer in a business context. Initially, he found the transition to teaching difficult, but now he feels he has everything under control didactically. His action research focuses on ‘technasium’ students: a group of students who follow a curriculum with special attention to research and technology. He aims to investigate how to select students for this special track in school in order to improve students’ arguments for choosing the technasium track.

4.1. Synchronous reflection profile of Mary

Characteristic of Mary’s reflection profile, in the first phase of her action research, is the variation in domains of reflection. This is apparent in the argumentation for her choice of action research topic which is motivated more or less equally by all four forms of reflection, as becomes evident in Table 1. Her action research topic started from her desire to involve students more actively in her lessons and to involve them more in learning.

Table 1. Representative quotations from synchronic reflection profiles in Mary’s beginning phase.

Scientific	Moral
I want to see if students can be involved in another way with what they do, and experiment with different pedagogies	What I like very much is to make children conscious of the world in which they live and that is really something that comes from my subject, geography
Technical	Artistic
I think it is important that you have interaction because then you can help children to get further with their thoughts	I really like that, if in a particular way of working (through interaction), suddenly they understand and that they are happy. These are the things that motivate me

In the middle phase of the action research process, Mary’s reflection profile can again be characterized as varied, mainly in relation to the process of action research. Different forms of reflection go hand in hand with different steps in the development of her action research. Starting from technical and scientific forms of reflection, she tries different things in her action research. Her reflection profile in this phase shows several forms of reflection that are related to action research in general; for example, constantly looking for what is possible in action research and seeing at the same time that there are limits to what you can do in class. It appears that Mary’s reflection profile consists of a core of technical and scientific forms of reflection which are related to the course of the action research process itself, along with moral and artistic forms of reflection which are related to action research and education in general. She stresses the opportunity to explore for teachers and the ‘fun’ of being engaged in new things as a teacher (see Table 2).

Table 2. Representative quotations from synchronic reflection profiles in Mary’s middle phase.

Scientific	Moral
But that is what I learn from this – constantly looking for what can you do and what you can’t do	I think you should give teachers the opportunity to fully explore a specific topic. That is of the utmost importance
Technical	Artistic
We [i.e. my colleagues and I] came to the conclusion that many of the assignments we give are too extensive for one 50-minute lesson. There is a lack of time to discuss the assignments with the students	I think for myself that it is fun to be engaged with new things and to look at myself critically

In the final phase of the action research process, Mary appears to combine different forms of reflection; she makes statements in the interview that fit in more than one domain of reflection. But the combinations she makes vary. For example, acquiring knowledge and concepts as a way to broaden your worldview (scientific reflection) serves the ideal of a better world (moral reflection). Also, as illustrated in Table 3, certain teaching methods that seem to have an effect on students (technical reflection) are connected with the teacher’s satisfaction when a student feels happy in class (artistic reflection). Furthermore, the idea that a moral obligation of school is to provide time for action research is tied to scientific output and the importance of quality of the research work (scientific reflection). To summarize, we can conclude that Mary’s reflection profile in this final phase is characterized by a series of combinations of different types of reflection (see Table 3).

Table 3. Representative quotations from synchronic reflection profiles in Mary’s final phase.

Scientific	Moral
You have to get extra time to develop and research things. But I think that it is good to see if that resulted in something, for yourself in the first place. But, as a school, you have the right also to evaluate, develop criteria and make demands	You have to get extra time to develop and research things. But I think that it is good to see if that resulted in something, for yourself in the first place. But, as a school, you have the right…. also to evaluate, develop criteria and make demands
Technical	Artistic
Through these teaching methods, I saw her do this, someone who would never respond, but does so now. The satisfaction is mainly that you make and see someone be happier in class, or more attentive. That’s what I enjoy and that’s where I get my energy from	Through these teaching methods, I saw her do this, someone who would never respond, but does so now. The satisfaction is mainly that you make and see someone be happier in class, or more attentive. That’s what I enjoy and that’s where I get my energy from

4.2. Diachronic reflection profile of Mary

Mary’s reflection profile from the first phase of her research shows a variation in domains of reflection which relate to action research, research as such, her chosen topic of study and education in general. In comparing her reflection profiles from the first and middle phases, we see a division in reflection lines. One line is mainly related to action research itself, to the process of doing research, and uses mainly technical and scientific domains of reflection without moral domains. The other reflection line is mainly related to action research in general and uses all domains of reflection. It is as if the broader context of the action research is put between brackets, which then creates two layers of reflection – one layer related to action research itself, and one layer related to the context of that research. In comparison with this division, the final phase of her action research brings together the different domains of reflection.

4.3. Synchronic reflection profile of Mark

Mark’s profile illustrates an emphasis on other domains of reflection. Although all domains of reflection are present, the division over the four domains (technical, artistic, scientific and moral) shows that mainly technical and artistic domains are well represented and scientific and moral domains of reflection are barely represented (see Table 4). The only remark that was labelled as moral is that Mark indicated there are a number of agreed-upon rules in his school which ‘just have to be complied with, there is nothing behind that’. There does not seem to be an overarching general value or ideal: ‘But other ideals? No’.

Table 4. Representative quotations from synchronic reflection profiles in Mark’s beginning phase.

Scientific	Moral
–	–
Technical	Artistic
Make sure I have the right target group. If there is a lot of rubbish [i.e. unmotivated or incapable students], it will be a disaster	I would like to have the right students in my class, which makes my task as a teacher, in any case, easier and more fun

In relation to scientific reflection, one remark was made – that as a teacher, Mark tries to improve his teaching by experimenting and researching. But he indicates that this is not related to ‘real’ research because it is a bit ‘trial and error’ and ‘quick and dirty’. In relation to his research project, technical and artistic domains of reflection are used as a motive. Mark’s motivation to do research is related to his specific educational practice. The research question is important for him because he wants more students to opt for his technasium class. In summary, one can state that Mark, at this phase of his action research process, takes a pragmatic perspective and uses action research as a means to solve a concrete problem.

During the middle phase of his action research project, Mark is mainly focused on answering his research question step by step following a classical empirical approach in which mainly scientific ways of reflection are found. Together they form a report of an empirical research project carried out in a technical manner. Furthermore, Mark focuses on direct implications of his action research for his school in terms of making ‘better choices’, having ‘arguments for conclusions’ and so forth. We did not find artistic or moral domains of reflection in relation to his research or action research in general. Mark’s reflection profile in this phase appears to consist mainly of technical domains of reflection that are concerned either with the technical part of his action research or with the pedagogical implications of the results of that research, as illustrated in Table 5.

Table 5. Representative quotations from synchronic reflection profiles in Mark’s middle phase.

Scientific	Moral
Well, you think something like, what should a student have to fit in with the technasium	–
Technical	Artistic
I interviewed a colleague to find out what he considers important and what a student should be able to do to make him fit in with the technasium	–

Mark’s profile in the final phase of action research can be characterized by an evaluation of the outcomes of his research (see Table 6). On the one hand, he refers to the outcomes of the action research, which in his case are insights into the necessary qualities of students and criteria to achieve a good mixture in the group of students for the technasium track. On the other hand, Mark describes the restrictions in relation to the goals of action research to learn inquiry as a stance and to transfer that stance to your students. The reflection profile of Mark in this final phase therefore boils down to an evaluation of the outcome and the value of his research in terms of scientific and technical reflection. He wonders whether his research is valid on the one hand, but on the other also values that he has some criteria to ‘lean on’.

Table 6. Representative quotations from synchronic reflection profiles in Mark’s final phase.

Scientific	Moral
I had a different expectation there. Even my own action research did not convince me that it’s not the case […]. I am still questioning myself, how valid is my action research?	–
Technical	Artistic
I have something to lean on, a few criteria to keep in mind … [e.g. to see if students ‘fit’ in the technasium track] Just somewhat less quick and dirty …	–

4.4. Diachronic reflection profile of Mark

In the case of Mark, we see in the beginning phase mainly technical and artistic domains of reflection. He has a concrete problem and seeks a solution for this problem through action research. The reflection profile in the middle phase consists mainly of technical domains of reflection that give a description of the stepwise answering of the research question. The line of reflection shifts from the problem as a didactic one to the method used in his action research to answer his problem. The reflection profile in the final phase consists mainly of an evaluation, in technical and scientific domains of reflection, of the outcome of his study. This outcome is mainly about the usefulness of the solution found and the validity of the research. The reflection line returns to the didactic problem which was the starting point of his action research, whereby the circle is closed.

4.5. Conditions for the development of reflection

In this section, we elaborate on the relation between the course of reflection and the conditions for development of reflection as a complex system as described by Davis and Sumara (2005). In what follows, we describe the conditions that played a role for both teachers in the way their reflection unfolded during their action research trajectory. Table 7 characterizes the role these conditions played in the development of their reflection.

Table 7. Conditions for the development of teacher reflection, and how they related to the two teachers’ development of reflection during the one-year period of action research.

Condition	Description	Mary	Mark
Diversity	Variation in domains of reflection	Variation	Focus
Redundancy	Alignment between domains of reflection	All domains of reflection	Knowledge of domains of reflection
Interaction	(Mechanisms that prompt) Interaction between domains of reflection	Focus on product and process	Focus on product
Decentralized control	Free exchange between the domains of reflection	Decentralized and centralized	Centralized
Improvisation/conversation	Attention and responsibility of all domains of reflection	Diverging and converging	Converging
Minding	Following the development of reflection attentively	Research and reflection	Research

Diversity is about sufficient variation in domains of reflection to come to a meaningful exchange between these domains. Mary’s case shows a broad range of variation in her reflection: in the beginning phase, in both reflection domains and in the way she motivates her action research project. In the middle phase, her reflection refers to the action research process (technical and scientific) and to action research and education (moral and artistic). In the final phase, she combines different domains of reflection in one statement; for example, both technical and artistic. In Mark’s case variation is less in all phases, in the sense that it is restricted mainly to technical and scientific reflection, and furthermore that scientific reflection is principally in the service of technical reflection.

Redundancy is about domains of reflection related to each other and to the situation so that a reasonable alignment is possible. In both Mary’s and Mark’s cases we see that a reasonable alignment is possible in the domains of reflection they use. The difference is mainly in the fact that for Mary the domains of reflection used are more numerous and varied than in Mark’s case. For Mark, it is chiefly about an alignment between scientific and technical domains of reflection, and so between domains on the knowledge side of the system and the justification and application side of it.

Interaction between domains of reflection refers to mechanisms that prompt these domains of reflection to interact and to be tied into more sophisticated possibilities (for the system of reflection as a whole). In both cases we see sufficient interaction between the domains of reflection to come to more sophisticated possibilities. The difference is that the added value in possibilities in Mark’s case is especially about the product of reflection, and in Mary’s case it is both about process and product. The interaction in Mary’s case has, more than in Mark’s case, the character of a condition that contributes to the development of reflection as a system.

In its development and result, reflection can end up other than planned or expected, as reflected in the condition of decentralized control. In Mary’s case, we see a sequence of changes in the course of reflection during the action research process that is expressed by a change in the research question, for example. The ‘control’ is not so much in the research question but is directed by what seems reasonable in the course of the action research process. Thus, control is not so much absent but is made dependent on the insights that arose in the course of the action research process. In Mark’s case, the control is the hands of Mark himself, who wants a clear and definite solution for a concrete problem of practice.

Improvisation/conversation is about the open and mutually engaged attention and responsibility of all domains of reflection for the development of reflection as a joint project. In Mary’s case we see a variation in both diverging and converging directions in the development of reflection. On the one hand, there are new options that make reflection head in different directions. On the other hand, these options and directions are related to one another, through which cohesion is maintained. In Mark’s case, the research is a converging opportunity and reflection has the character of a goal-oriented conversation.

Minding is about following the development of reflection attentively. Both teachers work on their action research with attention. The difference is that in Mark’s case the focus of attention is the scientific method of the research, whereas in Mary’s case there is also attention to the development of reflection as a research-oriented attitude. In the first case, the development is more or less ‘outside’ the teacher; while in the latter case, the development is also ‘inside’ the teacher.

5. Conclusion and discussion

The starting point of this study was the assertion that action research is meant to expand the space of the possible, not only in relation to teachers’ classroom practices but also in relation to reflection. The problem with reflection is that with this ‘expansion’ the complexity also increases because of the many domains of reflection and the differences among them. We claimed that this is a problem mainly due to a lack of a framework for understanding the complexity of reflection in action research. In this article, we aimed to develop such a framework in order to grasp this complexity. We proposed a heuristic framework with four basic domains of reflection and with conditions for the development of this reflection. Using this framework, we illustrated the course of reflection in action research of two teachers, Mary and Mark, including the relations between this course and the conditions for the development of reflection as a complex system.

We showed that the development of reflection in Mary’s case goes from variation in the beginning phase, to differentiation in the middle phase, towards synthesis in the final phase. Her trajectory is multiform and process oriented. The development of Mark’s reflection can be described as a development that through mainly technical and scientific domains of reflection goes to a concrete problem, via a scientific method to a practical solution. The trajectory is rather straightforward and from the start is product oriented. In conclusion, these courses of reflections can be characterized as extending (Mary) and restricting (Mark). The extending course of Mary leaves sufficient space for extension of the possibilities of reflection, whereas the restricting course of Mark leaves few possibilities. In the first case, reflection is broad, diverging and process oriented; and in the second case, it is focused, converging and product oriented.

These cases illustrate that the conditions for the development of reflection, as mentioned by Davis and Sumara, all play a role, although these conditions are used more in the case of Mary than in the case of Mark. Together these cases illustrate for Mary a combination of both using the conditions and an extending course of reflection, and in Mark’s case illustrate a limited use of conditions and a more focused course of reflection. In the cases discussed, the course of reflection and the conditions for the development of reflection may be influenced also by these teachers’ diverse backgrounds, and the focus of their action research project.

With respect to the conclusions we draw, we have to make a few important notes. To start with we performed a comparative case study of only two teachers we selected based on the differences in variation of reflection and teachers who are beginners in action research. If teachers would have been selected based on other characteristics, the outcomes of the course of reflection and the use of conditions may very well have been different.

Also, the distinctions made should not be interpreted as absolute or mutually exclusive. We want to indicate gradual differences that vary in the phase of action research, the subject of action research or the situation. In addition, these differences should not be interpreted as characteristics of teachers. We aim to describe characteristics of the different ways in which teachers’ reflection on action research and conditions can evolve. Moreover, we do not mean to give any preference in making this distinction.

We intend to stress also that the framework we developed is not intended to give a stepwise description of the course of reflection of teachers in action research. It offers the possibility to take ‘snapshots’ of the structure of reflection in a certain phase and it can help to describe the shifts between phases. It does not offer an explanation, for example, for the differences in the course of reflection. It does not explain either why a teacher chooses certain domains of reflection in a certain phase of the action research process or chooses certain conditions.

With these restrictions in mind, the conclusions that the framework covers the reflection which teachers use in relation to their action research and that the framework can offer insight into the development of reflection as a complex system seem justified.

The first conclusion seems justified because our results confirm previously found research results. These results show that the domains of reflection distinguished are used in relation to education in general (Coldron and Smith 1999), in relation to important phases in the professional development of teachers (Luttenberg, Imants, and van Veen 2013) and in relation to difficult situations which require the decision of teachers (Luttenberg, Hermans, and Bergen 2004; Luttenberg and Bergen 2008). In fact, the results of these studies are similar to those of this study in which we focus on teachers doing action research.

The second conclusion seems justified because the framework offers an insight into at least two important characteristics of reflection as a complex system. According to Davis and Sumara (2005), complex systems can be characterized both as determined by a structure and as emergent. Structure-determined means that complex systems, to a certain extent, are self-determining, which means that they are capable of adapting themselves from and via their structure and maintain cohesion amidst changing circumstances. The framework we developed shows in which sense different domains of reflection are not an arbitrary collection. As a complex system they have a structure with similarities and differences related to orientation regarding knowledge and value or justification and application. Moreover, the course of that reflection can be understood as something which emerges from and is related to the structure of reflection of teachers in a given situation. Reflection as a complex system appears to have a structure that enables the system to adapt to changing circumstances and maintain cohesion.

Emergent means that complex systems are, to a certain extent, self-organizing. They are capable of creating a ranking of possibilities in a certain situation which is different from that expected or is different from the possibilities until then. The dynamics of the system, the continuous interaction between domains of reflection and the interferences between structure and conditions make it clear in which sense the course of reflection in action research should be understood as something unique and unpredictable. The combinations that are possible in relation to interaction and interference are different every time and can in that sense be an occasion for another course of reflection. At the same time they are so abundant it is impossible to predict how that course will go.

These two characteristics of the framework we developed make it possible to see the similarities and differences in the courses of reflection both between teachers and for individual teachers, in order to understand from one and the same framework a way that shows the similarities while doing justice to the differences and the uniqueness of their individual developments. The domains of reflection, their mutual exchange and the use of conditions lead to differences in the course of reflection in action research. The framework shows that the similarities and differences of that course can be understood as the combined result of the reflection structure and conditions used by the teacher. Structure and conditions are not static entities, but possibilities that can be activated by a teacher in line with the progress of his/her research process. Furthermore, they are mutually determining in the sense that both influence each other constantly. That influence is not a mechanical process but part of an ongoing effort on the part of the teacher to keep a workable balance in the exchange between different domains of reflection.

6. Implications

The framework we developed in this article can offer a support for reflection in action research in different ways. It is first of all possible to distinguish the different domains of reflection a teacher uses when conducting action research by means of relevant functions and considerations for each domain of reflection. It is possible to identify differences and to show how distinguished domains of reflection can lead to different interpretations of the research situation and the research problem, which can then lead to a consideration of different research alternatives.

These insights from the framework can be used to make more explicit choices in relation to the course of the action research project, such as giving more or less space to scientific domains of reflection or considering the role of reflection in action research as a means or as a goal, for example. Moreover, the conditions that are important for the development of reflection can explicitly be taken into account. Just as in the case of the research alternative, a teacher can decide in which way he wants to use which conditions.

In the previous paragraphs we indicated how the framework can offer a support with respect to the different choices a teacher has to make in action research. At the same time, it can also be a useful tool in the context of the supervision of action research. Supervisors can use both instruction and practice to teach teachers how to use the different domains of reflection. These reflection domains can be linked to the considerations made in the actual phase of the action research process of a teacher. Implicit considerations can be made more explicit by helping teachers recognize them as domains of reflection which can then lead to additional consideration. Because the insights provided by one domain can have consequences for the insights provided by other domains, a teacher should be taught to make flexible transitions between the different domains of reflection. This means that (s)he must learn to consider the different domains of reflection in conjunction with each other and not separately in order to gain a comprehensive perspective of reflection as a system. Also, conditions play an important role in the functioning of reflection as a complex system. For example, by creating opportunities to use certain conditions in research supervision, possibilities for reflection can be improved.

The final choice for the focus of the reflection is the teacher’s. He or she has to answer questions about what is important, in which direction to go and what this means for questions he or she has to answer for himself/herself. In that sense, this framework will not make it easier. Just like a map, it can indicate where he or she stands, what type of reflection plays a role and what the different possibilities and conditions are. However, it does not decide in what direction he or she should go; the responsibility for these choices or the urgency lies with the teacher.

In conclusion, the present framework seems to provide a useful tool for identification of the domains of reflection which teachers use in action research and for enhancing the possibilities of reflection in action research. It can illuminate how teachers reflect in relation to action research and suggest possible constructive alternatives. In this way, the framework can contribute to expanding ‘the space of the possible’.

Disclosure statement

No potential conflict of interest was reported by the authors.