Understanding the complexity of teacher reflection in action research

Abstract

Reflection in action research is a complex matter, as is action research itself. In recent years, complexity science has regularly been called upon in order to more thoroughly understand the complexity of action research. The present article investigates the benefits that complexity science may yield for reflection in action research. This article begins by explicating the sense in which the complexity of reflection in action research involves the role of values and existential knowledge in education. The gap between theory and practice is also explored. On the basis of a number of common features of complex systems (heterogeneous, open, dynamic, non-linear, adaptive, and co-adaptive), the sense in which reflection can be regarded as a complex system is discussed. To this end, the features of complex systems are translated into features of reflection in action research, which, in turn, are illustrated with examples from recent publications on reflection in action research. The aim of this analysis is to make reflection in action research more understandable and manageable. In line with this, it is argued that room for insecurity and unpredictability, combined with an explicit consideration of reflection as a complex system, contributes to the use of complexity as a stimulus for new learning.

Keywords:

• Reflection
• action research
• teachers
• complexity science

1. Introduction

The intrinsic connectedness of reflection and action research is emphasised in different ways in the literature. According to Carr and Kemmis (1986), action research is a form of self-reflexive research undertaken by participants in social situations, while Wallace (1991) views it as an expansion of teachers’ usual reflection, in the sense that it aims to be more rigorous with regards to theoretical enhancement and data collection. Stenhouse (1975) considers the development of solid reflection to be the primary objective of action research. Reflection is a characteristic of action research to the extent that this type of research can justly be characterised as reflective practice (Carr and Kemmis 1986).

Although reflection and action research are often mentioned in the same breath, when it comes to elaborations on the role of reflection there are differences in emphasis. On the one hand, reflection is an essential condition for action research. Reflection in action research makes the teacher aware of the various forms of knowledge and action in education, of the tensions and contradictions between the two, and of what is required to manage all of this. It is the ‘sine qua non’ for change in action research: no action research occurs without reflection (Editorial 2011). On the other hand, reflection is seen as a complicating factor, in the sense that it produces a number of problems which can interfere with the planned course of action research. One of the problems, for example, pertains to the role of moral values (Carr 2006; Kim 2013). Moral reflection may pose a problem, because reflection on moral aspects proceeds differently and may lead to different conclusions than reflection on technical–instrumental aspects.

The role of reflection in action research thus appears to be a paradoxical matter: indispensable on the one hand and a complicating factor on the other.

To further our understanding of this paradoxical situation we turn to complexity science, because in recent years action research scholars have paid an increasing amount of attention to this discipline (Sumara, Davis, and Carson 1997; Phelps and Hase 2002; Ahmadian and Tavakoli 2011). Here, action research is viewed as a complex system, the complexity of which is sought to be understood from the features of complex systems.

Reflection is often mentioned in the discussion on complexity science and action research, by, for example, Ahmadian and Tavakoli, who state that ‘the continuous interplays between doing something and revision of our thoughts’ (2011, 133) by means of observations, reflections, and so forth are in line with complexity theory. However, the extent to which the defining features that apply to action research as a complex system also apply to reflection in action research remains unclear. These defining features are, for example, heterogeneity, non-linearity, and adaptivity (Ahmadian and Tavakoli 2011). The present article investigates this issue further, and, more particularly, investigates the question of the sense in which reflection in action research can be characterised as a complex system and the implications this has for reflection in action research.

In terms of what follows, elaboration on the role of reflection in action research will be dealt with first, and the complicating factors will be addressed more extensively – more particularly the role of moral values and the gap between theory and practice. Subsequently, the sense in which reflection in action research can be regarded as a complex system is explored. To this end, first, a characterisation of complex systems will be given on the basis, among other things, of a number of common features of such systems. Each feature is described, elaborated towards reflection in action research, and is illustrated using examples of reflection in action research. Second, this investigation focuses on the implications for reflection in action research, and we claim that interpreting reflection as a complex system means accepting uncertainty and unpredictability. We state that this investigation is all about understanding this uncertainty, with the aim of managing it better.

2. The role of reflection in action research

In the literature on action research, reflection seems to be used in two ways. On the one hand, reflection is used in a contemplative sense, mainly as a form of awareness (for example, Phelps and Hase 2002), and, on the other, reflection is used in a more active sense, as a lever for change (for example, Ahmadian and Tavakoli 2011).

Reflection in action research, in the contemplative sense, is seen as minding, which means paying attention to and being aware of what takes place (Davis and Sumara 2005; Phelps and Hase 2002). It is a mental process whereby a teacher thinks about things by going back over them (Tripp 1998); it is a way of becoming aware of salient assumptions and ideas. In this instance, knowledge is conceived, in the Aristotelian meaning of phronesis, as a way of constant coping and as responsiveness with respect to what is appropriate here and now (Carr 2006; Elliot 1987). Thus, reflection is a way to trace the good from one’s own practice.

In the more active sense of action research, reflection is seen as something which has consequences for practice. It is then seen as, for instance, the manner in which theory and practice are brought together (Kim 2013). Such a more active interpretation of the role of reflection in action research can also be related to the role of values in education. In this case, reflection refers to the constant interaction between acting and contemplating which action is required (Ahmadian and Tavakoli 2011), wondering what truly matters, and how the distinction between ordinary work and meaningful work can be made (Editorial 2011).

This more active interpretation of reflection in action research is about investigating the rationality and fairness of one’s own educational practice (Carr and Kemmis 1986), about the realisation of teachers’ situational understanding and moral agency (Somekh and Zeichner 2009), and about revealing the different voices (Bakhtin 1981; Vetter and Russel 2011) which are of importance for education and the development of teachers’ professional identities. It clarifies the ‘living contradictions’ between someone’s thoughts, words, and actions (Whitehead 1989) and incites changes, such as reconsidering thoughts and rearranging the voices (Vetter and Russel 2011) in order to establish change and to realise a better alignment of the teacher’s inner processes and the context (Phelps and Hase 2002). Reflection then leads to change on a metacognitive level (Winkler 2001) – a commitment to critical reflection or to parrhesia: teachers’ courage, ability, and responsibility to speak for themselves (McNiff 2011).

The points mentioned demonstrate that reflection in action research can make the teacher aware of the different forms of knowledge and action in education, of the inherent tensions and contradictions, and of what is required to incite change on several levels. It is an awareness which all teachers must achieve for themselves (Editorial 2011). With this distinction between the two roles of reflection in action research, we now continue to indicate why reflection in action research is a complicating factor.

3. Reflection in action research as a complicating factor

In the Introduction we stated that reflection in action research produces a number of problems, such as the role of moral values. In what follows, a more systematic exploration of and further elaboration on these problems will be given based on the literature about reflection in action research.

One of the problems with reflection in action research, as said, concerns the role of moral values in reflection. It is generally agreed that education is not just a technical–instrumental matter, but a moral one as well (Kim 2013; Somekh and Zeichner 2009). The moral aspects must be taken into account if one is to become a good teacher. Technical–instrumental reflection is ‘mainly concerned with mastery and/or application of technical means for achieving given educational ends, and includes a simple description of observation or a focus on behaviors or skills from past experience’ (Taggart as quoted in Lee 2005, 703). Moral reflection, however, is seen to transform practice, because it reconstructs experience in light of the values one has in life, such as justice and equality (cf. Lee 2005). The role of moral values in reflection in action research makes reflection a complicating factor in several respects. It is complicated, because that which is effective and efficient from a technical–instrumental perspective may differ from what is morally right. In that case, if a teacher is to learn to act in a way that is morally right, divergent advice or guidelines must be reconciled (McNiff 2011).

Moral reflection is also a complicating factor because reflection on education, as a moral matter, is a personal form of reflection. This can be clarified utilising the Aristotelian distinction between poiësis and praxis (Carr 2006; Elliot 1987; Kim 2013). Poiësis applies to productive activities which are guided by instrumental end-means thought, named technè by Aristotle. In praxis, the goal is not to produce an object, but to realise the idea of the good that is fundamental for a life that is worthwhile from a moral viewpoint. Furthermore, knowledge of the goal can only partly, if at all, be theoretically specified in advance and has to be gained by understanding how the idea of the good must be interpreted and applied. With regard to the good, knowing what and knowing how are not separate matters, but two mutually supportive elements in a didactic process of practical action and reflection.

Aristotle calls this form of reflection phronesis, which leads to the development of the capacity to sensibly and wisely assess what the appropriate expression of the good is in a certain situation. The outcome of the consideration is binding, to a greater or lesser extent, on the one making the consideration (Habermas 1991), including a willingness to accept the outcome of that reflection as something that demands consequences – a moral disposition to ‘do the right thing in the right place at the right time in the right way’ (MacIntyre 1984, 141).

Finally, moral reflection can become a complicating factor because its scope differs from technical–instrumental reflection (Habermas 1991). The outcome of reflection, from the view of education as a technical–instrumental matter, in a sense, is non-committal. In reflection on education from a moral matter perspective, the conclusion is binding, in principle, on the teacher and on all of those involved.

In other words, reflection on the moral aspects in education involves a different kind of consideration, follows another direction, and can lead to different conclusions, in terms of contents and scope, than reflection on the technical–instrumental aspects. This can also lead to problems when the differing conclusions prove hard to reconcile or are even incompatible.

Another problem with reflection in action research concerns the role of the teacher’s personal integrity. To become a better teacher through action research, it is essential for that teacher to reflect on his/her personal development, who he/she is, and who he/she desires to be (Kim 2013). This personal or artistic reflection is concerned with the teachers’ personal preferences and experiences and seeks answers to the question ‘what is good’ (Coldron and Smith 1999). The nature of the problem, in part, corresponds with the problem concerning reflection on the moral aspects of education. There is a kind of consideration in which personal experience and preferences play a prominent role, have a personal direction, move outwards from within, and draw conclusions that have a personal nature in terms of their contents and scope. It is about the question of what teachers themselves consider to be the overall and long-term right course of action. Here, too, problems may arise when the outcome proves hard to reconcile or is incompatible with the outcomes of other forms of reflection on action research.

The final problem we discuss in this article regarding reflection in action research is the commonly referred to gap between didactics as a social science and education as teachers’ practice. Although this gap is recognised and discussed in relation to, for example, teacher education (Korthagen and Kessels 1999) and teacher action research in general (Kemmis 2009), we discuss the problem here in relation to teachers’ reflection in action research. We aim to illustrate how the problem with this gap adds to the complexity of reflection in action research. The usual interpretation of this problem is that educational science does insufficient justice to the complexity of educational practice, remaining empty in that sense, and conversely that educational practice makes insufficient use of educational science, remaining blind in that sense (Bulterman-Bos 2008).

In this analysis of the gap between theory and practice, emphasis is often placed on the complexity of educational practice, in comparison with which educational science is presented as an area with a more or less coherent set of knowledge and research methods. However, educational science, as a social science, is a field of knowledge with a horizontal knowledge structure (Bernstein 1999; Wignell 2007). This means that it consists of a collection of equivalent theories that are partly isolated, partly overlapping, and each entitled to validity in its own way. An empirical phenomenon can be linked to various theories in which the interpretation depends on the applied theory. As a result, concepts can have different meanings, and one phenomenon can be characterised in several ways. In other words, the gap between theory and practice is a gap between a varied collection of educational theories and complicated educational practice.

There is a second reason why the usual analysis of the gap between theory and practice is incomplete. Generally speaking, theory only refers to educational science as a scientific discipline, but education is also related to ethics and philosophy as theoretical disciplines. In ethics and philosophy, too, a gap lies between theory and practice, in the sense that practice always proves to be more complicated than theory. Moreover, ethics and philosophy have a horizontal knowledge structure as well; both involve sets of equivalent theories, partly isolated and partly overlapping, with each entitled to validity in its own way.

The gap between theory and practice is a gap between a complex network of educational, ethical, and philosophical theories and concepts, and complex educational practice. This adds to the complexity of the situation when linking theory and practice in reflection. The complexity of educational practice itself is juxtaposed with the complexity of educational, ethical, and philosophical theories.

In this section, we elaborated on a number of problems with reflection in action research, related to the role of moral values, the personal integrity of the teacher, and the complicated gap between research and practice. These problems lead us to consider complexity science as a way of gaining understanding, not only in action research as such, but also in the way reflection is used in action research. We will elaborate on complexity science in the following section.

4. Action research and complexity

There has certainly been a growing interest regarding action research and particularly in understanding its complexity using complexity science (Davis and Sumara 2005; Phelps and Graham 2010). This interest is the result of the recognition that human behaviour is too complex and intricate to be studied only from the classic, analytical research perspective (Ahmadian and Tavakoli 2011). The aim of complexity science is not to reach a comprehensive explanation of complex phenomena, but rather to act as an umbrella notion to notice similarities and relations among often seemingly disparate phenomena. In addition to this, complexity science endeavours to deliver an understanding of the common features of complex phenomena.

Phelps and Hase (2002) investigated the applicability of complexity science on action research in the social sciences and reached the conclusion that the methodological connections between action research and complexity theory can contribute to a better understanding of what is happening in action research. Davis and Sumara (2006) elaborate on this, applying it to learning and knowledge development, which are regarded as the unfolding structure of an ongoing process of adjusting a complex system. Based on a number of common features of complex systems, Ahmadian and Tavakoli (2011) show the sense in which educational situations can be regarded as complex systems and the sense in which its complexity can be understood from the features of complex systems. They claim such systems are:

•	heterogeneous, consisting of different, relatively independent components;
•	open, existing by the grace of exchange with the context;
•	dynamic, changing over the course of time;
•	non-linear, lacking proportion in terms of influence;
•	adaptive, involving influence and adjustment between system and context; and
•	co-adaptive, involving influence and adjustment between components.

The question here is how reflection in action research can also be regarded as a complex system and the sense in which the complexity of reflection can be made understandable through the features of such complex systems. In what follows, we investigate this further on the basis of the features of complex systems described by Ahmadian and Tavakoli (2011).

A brief characterisation of complex systems will be given first, followed by a description of each feature and illustrated using examples of reflection in action research drawn from publications by Winkler (2001), Jove (2011), McNiff (2011), and Vetter and Russel (2011). The first example is always taken from the action research of a teacher named Holly, as described by Vetter and Russel (2011). This teacher’s action research covered a two-year period. During the first year, she explored the contribution of diaries to students’ metacognitive awareness and reading skills. During the second year, she investigated student involvement. The examples are relevant in this matter because they were explicitly intended to serve as records of the reflection process in action research.

5. Reflection in action research as a complex system

A complex system can be seen as a coherent whole, comprising different parts, which is able to maintain itself as a relatively autonomous entity in the face of changing circumstances. The parts themselves can also be seen as coherent wholes that are relatively independent of each other and of the system. 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 and cannot be understood through knowledge of the composing parts. Complexity science tries to understand how a coherent and purposive whole emerges from the interactions of simple and sometimes non-purposive components. Therefore, complexity science focuses on the interaction; that is to say, tries to understand the whole not through an understanding of the parts, but by an understanding of the interaction of the parts. Thus, it is not the parts themselves that are of interest, but the relations between them (Davis and Sumara 1997).

Although complex systems can differ greatly, they have a number of common features (Ahmadian and Tavakoli 2011).

Complex systems are heterogeneous, thus meaning that they consist of different elements which are relatively autonomous and independent. The elaboration in Sections 2 and 3 shows that, with regard to reflection in action research, different elements can be distinguished. This involves professional expertise or knowledge derived from practice and more scientific knowledge to serve practice. This also involves knowledge that is objective, neutral, and not binding in principle, and values which are binding on those directly involved or on anyone.

In Vetter and Russel (2011), the sense in which there are relatively different elements of reflection in Holly is evident. Part of that reflection is more scientific in nature and pertains to the doing of action research – familiarising oneself with quantitative and qualitative research skills and learning to cope with the uncertainty that qualitative analysis may involve. Her reflection on Slechty’s theory on student involvement, propagated by the school board, is of a similar nature.

Furthermore, personal experiences play an important role in Holly’s reflection. She uses her past experiences as a learner as the starting point of her research and relies on the importance of a good bond with her students as a decisive factor in student involvement. Indeed, this is something that she integrates into lesson plans as a personal touch.

Finally, elements of reflection can be identified in which moral considerations are predominant. This is the case, for instance, with regard to her consideration leading to the conclusion that a research question relevant to herself, her students, and her colleagues is more important than a scientifically interesting one. It is also evident in her acceptance of the consequences of that conclusion: if it is more difficult and entails more work, so be it.

Another example of heterogeneity in reflection is found in McNiff (2011). In her article, McNiff describes her work with teachers in Qatar. In this article, which is intended as a report of reflection in action research, she gives a description of the important considerations on which she bases her work. Some of these considerations are more general and neutral, such as the principle that people have a natural ability to think and make choices independently, and have reservoirs of tacit knowledge which they can translate into statements about what they think and expect. Other considerations are more value laden, such as the conviction that people have the duty and responsibility to speak for themselves, and have the right to claim their capacity to act in the world. She also mentions considerations which concern her own actions as a person and as a teacher. With regard to her actions as a person, she continually questions whether these actions are both formally and materially just. In her actions as a teacher, she is concerned with establishing personal and social progress for those involved.

These elements of reflection (the more general ones and the more personal ones, the more neutral ones and the more value-related ones) can be viewed as different, relatively autonomous forms of reflection with their own identities.

Complex systems are open, which means they exist as a result of a constant exchange of energy, matter, or information with the context. An external influence disrupts the existing stability, but is also necessary in order to correct existing imbalances within the system. This results in a state of dynamic stability by which means the complex system is capable of maintaining itself under variable conditions.

In particular, reflection in action research involves the partly deliberate searching for and allowing of external information. This includes, for example, the scientific insights that teachers use to bring theory and practice more closely together. However, the new experiences which they gain from the students in the context of the research, or from conversations with colleagues, can also serve as external sources of information for reflection.

In Holly’s story, the openness of reflection is mainly expressed through the external information that is allowed in the several forms of reflection. First, the information she allows pertains to the carrying out of action research itself: several research methods, supervision of the research, and consultation among the research group. Other examples include the school policy of raising students’ levels of performance and the feedback she receives from students.

Another illustration of the openness of reflection is found in Jove (2011), who takes a closer look at her work as a teacher-educator through action research. The method she uses is self-reflection. In her records, Jove accurately describes the outside information that affects her reflection and the manner in which it does so. She is inspired by the conceptual framework of Deleuze and Guattari (1987), which relates to notions such as striated spaces, smooth spaces, and lines of flight, as well as by Whitehead’s (1989) notion of ‘living contradictions’. Taking inspiration from these scholars, she aims to be more open to the contradictions in her theoretical framework, teaching, and action research. Through contact with the Centre of Contemporary Art, La Panera, in Lleida, she is introduced to the principles of teaching and learning processes through art. These turn out to be more suited to her vision than the education within her institute. For that reason, she revises her teaching and organises it so that her students can also learn through art. At a later stage, she studies her students’ assignments to explore what they can teach her about her different approach. This yields useful and, in part, expected information, but does not satisfy her completely. After a short conversation with a colleague, who asks about her expectations, she realises that she was mainly looking for what she had expected, and thus was not seeing the other information that her students’ assignments offered.

Indeed, her reflection is influenced by external sources of information which disturb the existing balance, although their simultaneous occurrence is necessary to correct existing imbalances. They ensure that reflection adapts and changes, while maintaining itself as a coherent complex system.

Complex systems are dynamic; that is to say, they are constantly in motion and change over the course of time. In a sense, one can say that change, in terms of reflection, is crucial in action research. This change can take on different forms. Teachers may become aware of the forms of knowledge gained without reflection that are embedded in the practice of their actions. Through theory, they can enhance their knowledge, come to a recalibration of their values, or realise a rearrangement of their voices. Each is an example of changes in elements which, in time, change reflection as a whole.

For Holly, the dynamics of reflection as a complex system find expression in the changes which the forms of reflection undergo, and thus establish that the said reflection is continuously in flux and changes over time. Her learning process is illustrative in this respect. Her reflection on the common educational practice at her school, affected by her own experiences (struggling reader), culminates in developing, trying, and researching an alternative approach for her class. Her reflection on the school policy concerning student involvement changes, because of reflection based on her own experiences as a student and through her own practical knowledge regarding involvement, to an approach implicitly questioning school policy. Her reflection concerning the research question that she should investigate changes because of her consideration that her research must be valuable to herself, her students, and her colleagues. Finally, her reflection on students’ learning processes changes as a result of her reflection, which is fuelled by the information resulting from her research regarding the students’ actual learning needs.

The afore-mentioned example of Jove (2011) shows that her reflection, at a certain stage, proves to be qualitatively different from the ‘same’ reflection before or after that, because new external insights are involved, thus giving rise to adjustments. First, through a comparison of her actual actions and her ideal vision, she comes to realise that she must try a different approach, and finds this in the ideology of the Centre of Contemporary Art, La Panera, in Lleida. Next, by critically comparing her expectations and goals with her students’ output, she reaches the conclusion that the said output does not correspond with the effect she had envisioned her teaching to have. Seeking a way out, she consults others and thus comes to realise that the problem does not lie in her students’ output, but in her perceptions and assessments thereof.

Both examples involve changes which jointly mark a development from a teacher dissatisfied with her teaching style to a teacher who, by tapping into new sources of knowledge and information, and through varying forms of reflection, gradually develops a personal style.

Complex systems are non-linear; that is to say, the increase or decrease of one element or feature does not automatically involve a proportional increase or decrease in a related element or feature. With regards to the development of reflection in action research, this means that a change in one of the elements of reflection does not automatically involve a proportional change in other related elements of reflection.

A phenomenon typical in action research is that an increase in theoretical knowledge does not automatically lead to the enhancement of practical knowledge. Holly shows that Slechty’s model, however theoretical-sounding it may seem, is eventually disregarded, because it is inconsistent with the outcome of her own reflection. On the other hand, seemingly simple questions or incidents related to one element of reflection under certain conditions turn out to establish a significant change within or based on another element. When Holly is feeling frustrated with the perceived futility of qualitative analysis, a suggestion from the research group to broaden her research question causes her to radically modify her research question. In another instance, when Holly claims to already know the answer to her research question, a colleague contradicts her, causing Holly to view the data completely differently. In this context, Holly uses the round-stone metaphor. They can be placed on top of each other, and as a result seem to teeter. The smallest touch can shake them and collapse the whole.

Other examples of this feature can be found in Winkler (2001) and Jove (2011). The former shows that an extensive focus on teachers’ educational experience, combined with relevant theoretical concepts, proves insufficient in establishing the expected expertise or enhancement. They remain separate forms of knowledge which, although both capable of increasing, do not cause a confrontation or qualitative change in the other per se. Conversely, Jove’s example illustrates the manner in which a colleague’s seemingly simple question – ‘what is few for you, in your view?’ (2011, 271; emphasis added) – in circumstances in which several principles are already being questioned, has far-reaching consequences in terms of the manner in which she views her students’ products.

The non-linearity is systematic and orderly, thus meaning that the change in order is not imposed externally, but emerges out of system interactions. Because of the dynamics of the system and the continuous concomitant interactions and adjustments among the constituent elements, new properties and behaviours emerge that are not contained in the essence of the constituent elements and that cannot be predicted from knowledge of the initial conditions (Mason 2008).

Complex systems are adaptive; that is to say, through an exchange of energy, matter, or information with the context, they constantly affect that environment and are, in turn, affected by that context (Davis and Sumara 2006). The context can strongly affect the development of the system as a whole and the development of individual elements. Concerning reflection in action research, this means that the progress of action research as the context of that reflection, and more generally education, as a ‘field of action’ of action research, can significantly affect the development of reflection.

In Holly’s case, the adaptivity of reflection is expressed in the reciprocal changes between the elements of her reflection and her educational context. She conforms, in part, to the school policy of enhancing students’ performance levels, starts a learning process as a researcher, and starts working with Slechty’s model. She is advised and directed by mentors and colleagues. She adapts her own teaching based on the outcome of her research and her students’ feedback. Conversely, Holly establishes changes in the educational context because of her behaviour, which is changing through reflection.

Another example is found in McNiff (2011), in her report of her work with teachers in Qatar. She advocates critical reflection as one of the basic principles of human dignity. Simultaneously, she realises that critical reflection can prove problematic in a culture where directly expressing critical comments is not customary. She wishes to respect that culture, not from a kind of distanced reverence, but from the sincere conviction that all humans have the right to claim their capacity to think and act for themselves in the world. The question is how both principles can be realised in the situation at hand. She describes three subsequent initiatives in which a form of sincere critical reflection fails to be realised. It is only after the third initiative, partly owing to two of the teachers, that an atmosphere of like-mindedness arises, in which one of the other teachers, through an educational conversation, causes McNiff to reflect critically on her own performance. Affected by the context, it becomes evident to her that this different perception of critical reflection as a nested skill is the appropriate expression of what she deems morally right in that situation.

The continuous exchange between complex systems and their context means that changes in the said context necessitate adjustments in order to comply with the new demands. Change in the system as a result of adjustments among individual elements will, in turn, lead to change in the context. This interconnectedness of system and context is therefore an important feature of complex systems.

Complex systems are co-adaptive; that is to say, through interaction the constituent parts of the system constantly adapt to and affect one another. Each of the various elements of reflection raises specific questions and demands specific solutions which, within the system as a whole, are not isolated, but depend on and have implications for the questions and solutions within the other elements of the system.

In Holly’s case, co-adaptivity is expressed in the consecutive adjustments to her position as a teacher and researcher. First, through a balancing of her own educational experiences and the educational practice of her school, she comes to the realisation that she needs to try a different approach. Next, a comparison of theoretical insights with her own experience as a teacher causes her to question those theoretical insights. Experience with action research teaches her to handle uncertainty better, to be more open to the data, and to allow those data to speak out more. This tolerance of uncertainty and openness to what takes place allows her to let go of her own assumptions about learning and pay more attention to the students’ needs.

Another example of co-adaptation is found in Jove (2011) in her description of handling the ‘living contradictions’ which mark her development as a teacher and researcher. Her reflection on the educational practice common at her institute is affected by her views on learning through art and her experiences with that. This reflection results in her adoption of a different approach to her teaching. Her assessment of this different approach is affected by her valuation of the students’ assignments, making her aware of her ineffective approach and therefore able to again adjust her method. Investigating the returns of this revised approach makes her uncomfortable, as she is forced to view the contrast between her expectations and her actions differently. This, in turn, leads to the discovery that she expects students to do things she never asked of them.

These are examples of interaction among several elements of reflection in which information from one element of reflection affects other elements, as a result of which these other elements adapt.

6. Implications for reflection in action research

In the previous section, we showed that features of complex systems translate to reflection in action research. We gave several examples from the action research literature which illustrate the heterogeneity, openness, dynamic nature, non-linearity, adaptivity, and co-adaptivity of reflection. We thus conclude that reflection in action research can be perceived as a complex system. In this section, on the one hand we show the implications that this has for understanding reflection in action research, and on the other we elaborate on conditions that can encourage teacher reflection in action research.

The different roles and significance of these notions show that action research sets in motion a process whose progress and outcome are uncertain and unpredictable, and that this also applies to reflection in action research. Within a complex system, the input of a new idea or a new action can lead to several unpredictable outcomes.

We suggest that interpreting reflection in action research as a complex system means accepting that this uncertainty and unpredictability applies to that reflection as well. In this case, reflection does not serve only as an instance to direct our thought and action, but also as an instance which carefully monitors the process and comments, orients, and questions what is appropriate here and now (Davis and Sumara 2005): not checking the uncertainty, but understanding that it is key, with the aim of managing it better. In other words, uncertainty and unpredictability do not automatically imply unintelligibility or inaccessibility to understanding. They involve a form of understanding in which unpredictability becomes a source of new knowledge. Deliberately introducing a new element to the situation establishes changes within reflection as a complex system that are unpredictable and can therefore generate knowledge and insight which could not have been gained without that unpredictability.

The fact remains that the uncertainty and unpredictability inherent in reflection as a complex system in action research may involve tensions which, however productive they may be in principle, in and of themselves, are perceived as negative and uncomfortable, and as such may have a frustrating, blocking, or counterproductive effect. In order to prevent this, we propose the ensurance of certain conditions.

A first condition is giving teachers the freedom to allow this uncertainty and unpredictability (Alsup 2005; Vetter and Russel 2011). This freedom is essential if teachers are to become aware of the prejudices and presuppositions they face. It also allows for interactions between components and context and among components, allows them to observe this process, and leaves them open to possible outcomes. As we saw in the illustrations we gave of reflection as a complex system; for example, the features of adaptivity and non-linearity. Indeed, teachers can play with these outcomes, testing hypotheses and drawing possible conclusions as a result.

A second condition is that the process of reflection in action research must be facilitated and supported in a broad sense. In action research, using several reflection practices to enhance reflection is common (e.g. annotations, diaries, learning communities, metaphorical reflection, memos, and papers) (Vetter and Russel 2011). This is useful in itself, but provides little support and creates no order with regard to the different elements of reflection as a complex system; for example, the heterogeneity. In order to give teachers that support, they must be introduced to, familiarise themselves with, and learn to handle the main features of reflection as a complex system (e.g. co-adaptivity, non-linearity, etc.). Sufficient support requires an explicit focus on the several forms of reflection which, as elements, constitute reflection as a complex system. This involves a focus on both the existence of different elements of reflection and a division of use in action research of these elements.

An example of such a division based on Habermas (1991), tailored to education by Oser (1994) and Coultner (2001), is into scientific, pragmatic, aesthetic, and moral reflection. The division broadly corresponds with the traditional division of views on education (Coldron and Smith 1999) and appears to largely cover the forms of reflection which play a role during the reflection of teachers in important educational situations (Luttenberg, Imants and van Veen 2013).

Such an explicit focus on reflection as a complex system, including a focus on the different elements of reflection, will not eliminate the uncertainty and unpredictability. However, it can raise metacognitive awareness, which furthers teachers’ understanding of what they are doing in terms of reflection at a certain moment.

7. Concluding remarks

The starting point of this article was the observation that reflection in action research, on the one hand, is an essential condition for change in action research and, on the other, is a complicating factor because of a number of problems; for example, the divide between didactics as a social science and within teachers’ practice. The question at hand was whether reflection in action research, as action research, can be regarded as a complex system. On the basis of a discussion of the features of complex systems in which we translated these features towards reflection and provided examples drawn from instances of reflection in action research, the sense in which this is the case has been demonstrated. This perception of reflection as a complex system offers an explanation of a number of important features of reflection in action research. Moreover, it offers guidelines to be taken into consideration if one is to do justice to the complexity of reflection in action research.

In this article, we elaborated mainly on reflection at the teacher’s level, as the teacher’s cognitive activity. According to Davis and Sumara (2005), complexity science is about studying learning systems in the broadest sense, from a micro to a macro level, from a sub-human to a supra-cultural level. The power of complexity science lies in its ability to provide an insight into the exchange and cohesion of knowledge generation among the different levels. Regarding action research, they elaborate on this idea based on the connection between the individual teacher’s learning process and more comprehensive social and cultural forms of knowledge generation. Social forms of knowledge generation involve larger collective wholes, such as schools. They claim that the aim of action research, from the complexity science perspective, is to accommodate the relative self-interest of the individual parts of a complex system in the system as a more comprehensive whole. At the individual teacher’s level, it is about incorporating the various separate elements of reflection into reflection as a system. At the school level, it is about accommodating a relatively autonomously acting teacher’s self-interest in the school as a more comprehensive form of knowledge generation. The merit of complexity science is that it can demonstrate the conditions that these transformations require.

For example, this means that complexity science must be able to demonstrate how and under what circumstances reflection, at the level of individual elements, can contribute to reflection as a system. Complexity science should be able to demonstrate how and under what conditions the individual teacher’s learning process, in the context of action research, can contribute to the learning processes of colleagues or the school as a larger collective whole. Because this connection in knowledge development often remains unclear in action research, we suggest that investigating this claim is a valuable next step in examining the relationship between complexity science, reflection, and action research.

Disclosure statement

No potential conflict of interest was reported by the authors.