CHAPTER 3
The role of action research in environmental management

[Chapter 3 in: Allen, W.J. (2001) Working together for environmental management: the role of information sharing and collaborative learning. PhD (Development Studies), Massey University.]

As indicated in Chapter 2, we can look towards the body of knowledge that has been generated through action research for guidance in developing frameworks for the new approaches that seek to emphasise sustainable policy orientations and people-centred research and development . Accordingly, this chapter begins by outlining the underlying concepts of action research in more detail. Some differences between action research and mainstream science are then explained, particularly to justify its use as an appropriate methodology to the research and development challenges outlined in earlier sections of this thesis. Some more practical details of practising action research are then discussed. Finally the process of critical reflection in action research is highlighted, and an illustration of how it's use in practice can help in getting people to think more deeply about the use of environmental practices is outlined.

"... if one wants to find out about the plant nutrient which is limiting growth to such and extent that there is no obvious pathology in its absence then the research needs to conduct experiments under rigorously controlled environmental conditions. The experimenter cannot participate with the nutrients in their 'dance in plant nutrition, nor is it sensible to examine the effects on the 'dance' of a multitude of factors working at once. The experiment must be conducted in a reduced and highly controlled world observed by afar by the observer! If, on the other hand one wants to actively explore with rural communities how they might design their own, more sustainable futures, then the method of enquiry needs to be participant-observer and the complexity of the situation must be embraced. There is no other sensible way to proceed." (Bawden 1991 p.33)

Action research (AR) comprises a family of research methodologies which aim to pursue action and research outcomes at the same time. It therefore has some components which resemble consultancy or change agency, and some which resemble field research. The focus is action to improve a situation and the research is the conscious effort, as part of the process, to formulate public knowledge that adds to theories of action that promote or inhibit learning in behavioural systems. One of the key characteristics of this approach is collaboration, which enables mutual understanding and consensus, democratic decision making and common action (Oja & Smulyan 1989 p.12).

In this sense the action researcher is a practitioner, an interventionist seeking to help improve client systems. "This help takes the form of creating conditions in the behavioural world of the client system that are conducive to inquiry and learning. Lasting improvement requires that the participatory action researcher help clients to change themselves so that their interactions will create these conditions for inquiry and learning" (Argyris et al. 1985 p.137). Hence to the aims of contributing to the practical improvement of problem situations and to the goals of developing public knowledge we can add a third aim of action research, to develop the self-help competencies of people facing problems.

Within this broad definition there are four basic themes: i) collaboration through participation; ii) acquisition of knowledge; iii) social change; and iv) empowerment of participants. The process that the researcher uses to guide those involved can be seen as a spiral of action research cycles consisting of phases of planning, acting, observing and reflecting (Masters 1995). As Oja and Smulyan (1989) point out, the underlying assumption of this approach -- which can be traced back to Lewin's writing in 1948 -- is that effective social change depends on the commitment and understanding of those involved in the change process (p.14). In other words, if people work together on a common problem "clarifying and negotiating ideas and concerns, they will be more likely to change their minds if research indicates such change is necessary. Also, it is suggested that collaboration can provide people with the time and support necessary to make fundamental changes in their practice which endure beyond the research process (Oja & Smulyan 1989 p.14-15).

Thus the role of the action researcher is identical to that proposed for contemporary facilitators in helping communities identify and adopt more sustainable natural resource management practices (eg. Pretty & Chambers 1993, Pretty 1998).

These facilitators may come from the community or they may be research or agency staff. However, their most effective role will be to involve the wider community to develop participatory attitudes, excitement and commitment to work together on jointly negotiated courses of action to bring about improvements and innovation for individual and community benefit. While this role is similar to much of consultancy, action research provides a means by which is more rigorous, and which allows for the development of public knowledge to advance the field.

In turn, by establishing conditions for the development of others, the action researcher acquires increasing skills in such things as the ability to build shared vision, to bring to the surface and challenge prevailing mental models, and to foster more systemic patterns of thinking. To paraphrase Senge (1990) action researchers are responsible for building frameworks and networks through which people are continuously expanding their capabilities to shape their future. That is, action researchers are responsible for developing a learning environment which challenges the status quo and generating liberating alternatives (Argyris et al. 1985 p.xi). Accordingly, the general aims of AR are frequently expressed in terms of orienting process criteria (e.g. participation, emancipation) and it seems worthwhile to continue to stress these characteristics to differentiate AR from other approaches to social change (Altrichter et al. 1991). These characteristics are well captured by Zuber-Skerritt's (1992 p.15) CRASP definition of action research as: Critical collaborative enquiry by Reflective practitioners, who are Accountable in making the results of their enquiry public, Self-evaluative of their practice, and engaged in Participative problem solving and continuing professional development.

This broad outline of action research sketched above is capable of encompassing and learning from a variety of research and intervention methods in a number of fields. Today we can identify clear applications of AR in a number of fields including organisational management, community development, education, agriculture and participatory evaluation (Deshler & Ewert 1995). The term 'action research' itself can be regarded as an umbrella term that includes several traditions of theory and practice. It is broad enough to include, for instance, Soft Systems Methodology (Checkland 1981) and Guba and Lincoln's (1989) Fourth-generation evaluation. Other terms including participatory research, action learning, praxis research, participatory inquiry, collaborative inquiry, action inquiry, and cooperative inquiry are also used in the literature (e.g. Whyte 1991).

As indicated in the previous chapter (Figure 2.2), although research approaches for addressing 'soft system' problem situations such as action research should be seen as complementary to other science approaches, there are some significant differences between action research and more mainstream science approaches. As the name implies, action research represents a form of inquiry into how human beings design and implement action in relation to one another. Hence, it is a science of practice - a concept which contrasts strongly with the mainstream science tradition. "We are accustomed to distinguishing between theory and practice, between thought and action, between science and common sense" (Argyris et al. 1985 p.1). Accordingly, while researchers attempt to bridge these conceptual chasms, the debate over whether or not action research is a science, or whether it could or should aspire to scientific status continues (e.g. Susman & Evered 1978, Checkland 1981, Argyris et al. 1985). While, as Checkland (1990 p.4) observes, these problems have not been too inhibiting to practitioners in the field, a comparison of some of the main points of difference between action research and mainstream science are useful particularly in justifying its use as an appropriate methodology to the research and development challenges outlined in the previous chapter.

For more than one hundred years the positivist conception of science has dominated the practice of physical, biological and social sciences. The underlying basis for this mainstream approach is the consideration of scientific knowledge to be obtainable only from sense data that can be directly experienced and verified between independent observors (Susman & Evered 1978 p.583). While this epistemology was designed with the natural sciences in mind (particularly physics) proponents argue that it characterises all sciences insofar as they are scientific; and this has also been the predominant opinion among the social sciences (Argyris et al. 1985 p.12).

But, to use Nelson's moon-ghetto metaphor; while science has enabled us to control the soft landings of space craft on distant planets, it has not helped us solve the 'lesser' problems associated with urban slums (Rosenhead 1989 p.4). In particular, positivist science has proved to have some deficiencies when it has been removed from the closely defined laboratory setting and asked to cope with the kind of organised complexity facing humanity and the life sciences in the 'real' world (for a more complete discussion of this topic see Checkland 1981). In fact Lewin's concern that mainstream science was not helping in the resolution of critical social problems was the driving force beyond his development of action research (Susman & Evered 1978). In mainstream social science implementation has been seen as a problem of application, of practice, perhaps of politics -- but not of theoretical science (Argyris et al. 1985 p.19). From the perspective of action research, however, implementation is not separable from crucial theoretical issues.

In traditional research, the researcher �makes every effort to remain objectively remote from the system being studied� (Bawden 1991 p.37). He or she is separated from the system being studied by a �hard� boundary and the system is reduced to one, or only a few parts, with the rest of the system assumed to be held constant. This research is appropriate in many circumstances, particularly in the bio-physical sciences. On the other hand, action research involves taking action in social systems of which the researcher is unavoidably a part. �Indeed, it is the activity of the (researcher)-observor joining with other participant-observors, that enables the system to become a researching system in the first place!� (Bawden 1991 p.37). These involve the study of �soft� systems without clearly defined boundaries between the researcher and the system.

Because the research involves complex and dynamic problems, exploring the social process of learning about situations is inextricably linked with the acts of changing those situations. In these systems the researcher must actively participate with others in the critical exploration of complex and dynamic issues of implementation which relate to the relationships between individuals, groups and their physical and socio-cultural environments. Furthermore, success in social change is not achieved simply by making the right decision at a particular time, but rather through developing a social process that facilitates ongoing learning (e.g. Korten 1980, Whyte 1989).

Thus, while as Argyris et al. (1985 p.18) remind us that there are continuities in the core features of mainstream science and action research including hard data and public testing, there are crucial differences as well. For one, action research sits squarely within the tradition of qualitative research methodology, rather than the more mainstream quantitative research paradigm. As Bunning (1995) points out, one reason for this is that action researchers seek to influence the phenomena being studied during the action research process itself, in the belief that the true nature of social systems become most evident when you seek to make changes to them. Because of this interventionist approach, the experimental standardisation of positivistic research is neither possible or desirable. Similarly, because action research thus addresses whole system issues which are invariably multi-variate (and somewhat indeterminate!) these are best approached within a qualitative and holistic framework, rather than a reductionist, and quantitative framework.

Another contrast between action research and mainstream science is that action research is focussed on what could be, rather than what is. "New thinking in action research seems to take the social construction of reality seriously. The emphasis is on possibility rather than prediction. From a constructivist perspective (action research) can contribute to people realising their values -- envisaging a preferred future and organizing effectively to achieve it" (Elden & Chisholm 1993 p.127). As these authors go on to point out, this highlights how action researchers are not 'value neutral', but rather concerned with selecting problems to solve that would both contribute to general knowledge and practice solutions concerning democratic, humanistic values. In this way, action research is change oriented and seeks to bring about change that has positive social value (e.g. healthy communities, environmentally sound management, etc.).

These points and others which contrast the differences between mainstream science and action research are outlined in Table 3.1.

Another point of distinction concerns the issue of participation in the research process. It is already clear from the above discussion that action research is by definition participatory, however, the implications of this -- particularly in the way that research is written up -- reveal clear differences in the relationship of the researcher and the researched within different research paradigms. Moreover, this distinction enables us not only to see the difference between mainstream positivist science and action research, but also clear differences between action research and more mainstream qualitative and interpretivist social science approaches. These differences are well discussed by Kemmis (1991 pp.58-60), and are summarised here (Box 3.1) from this account.

Fundamental, then, to action research is the concept of 'learning by doing' in which learning is perceived as experiential and reflexive. It recognises that people learn through the active adaptation of their existing knowledge in response to their experiences with other people and their environment. As the dynamics of a social system are often more apparent in times of change, learning and change can enhance each other.

However, while the above discussion of action research has concentrated on aim, there is also a need to specify the approaches and processes that the action researcher -- as a 'change agent' -- uses to achieve these aims in practice. Clearly, the present which is already determined by its own past is hard to change. However, as Dick (1996??) points out, the one exception to this is the change agent's own behaviour. "By act of will you can change your own behaviour. If you change your own behaviour in interaction with others, you can then change the relationships and the processes and actions that characterise it" (Dick 1996). In short, the action researcher has little option but to work with processes and relationships. That is all that is available. But through them the mechanisms for participation, more democratic and transparent decision making processes, and the prevailing culture, can be influenced.

In this sense the action research project begins with a process of communication and agreement between people who want to change something together. In terms of the aims of action research outlined earlier, this joint and bounded undertaking aims to build-up the participating actor's capacity to act, and support them in improving their problem situations in a self-reliant and empowering manner. As Schwedersky & Karkoschka (1996 p.35) point out, as we think in these terms, the notion of the project as a mechanistic operation designed to reach a preconceived 'end' or 'solution' is transformed into a concept of collaboration as a 'process'. Together those involved cover a certain amount of ground, and as the actors come to a cross-roads in the process they think together about which way they might go next.

However, some people are more suited to, and interested in, participating in an action research change inquiry than others. As Bunning (1995) points out, the reality of that because of downsizing, reduction in organisational levels and increased accountability, there are higher levels of stress and pressure around than ever before. While it is precisely those symptoms that indicate that change and development is needed, if people are not provided with the capacity to participate successful change is unlikely to be developed. Thus more will be learnt by a few genuinely committed co-researchers dedicated to exploring change within a smaller case study approach, than may be gained by engaging with a larger number of less willing participants in a bigger inquiry. Bunning (1995) suggests the following profiles (Table 3.2) provide a guide to selecting co-researchers for effective participation in the action research group:

Thankfully, for the action researcher, the idea of learning collaboratively is not new -- although as pointed out above some people are more effective than others. "Most of us, if we wish to learn a new skill or broaden our perspectives on an issue, will seek out some collaborative learning environment such as a club or training programme. Similarly, talking an issue through is a natural process for many people. We gain new insights as we express our own views and we subsequently modify our views as other people provide us with new ways of looking at the issue at hand (Kilvington et al. 1999 p.14). However, as these authors observe well-functioning groups do not happen by accident, and skills in managing group dynamics to keep the group moving in a positive direction are therefore central to the successful practice of action research. Awareness of what is happening to a group and access to the skills necessary to address this are crucial to the long-term viability of groups and their success in achieving their goals.

Similarly, the process of learning by building on experience is a natural one for most people and action research provides a framework for formalising and making this process more effective. "In brief, it consists of an iterative and cyclic approach of action and research with four major phases: plan, act, observe and reflect" (Zuber-Skerritt 1991 p.xiii). The basic underlying assumption which underpins theory and practice is the existence of an experiential-based learning cycle (from Kolb et al. 1979) that people can learn and create knowledge: i) on the basis of their concrete experience; ii) through observing and reflecting on that experience; iii) by forming abstract concepts and generalisations about what to do next; and iv) by testing the implications of these concepts in new situations -- which will lead to new concrete experiences, and hence the beginning of a new cycle. As a number of reviewers point out, this model is similar to other conceptions of basic adaptive processes, or problem solving, creativity, and decision making (e.g. Bawden et al..1984, Ison & Ampt 1992) . A more comprehensive form of the action research cycle from Susman & Evered (1978) is shown in Figure 3.1.

Figure 3.1. Phases within an action research cycle (adapted from Susman & Evered 1978)

While Susman & Evered (1978 p.588) consider all five phases to be necessary for a comprehensive definition of action research, they do acknowledge that action research projects may differ in the number of phases carried out in collaboration between the action researcher and the client system. In particular they point to the case where the researcher may only be involved in collecting data for diagnosis and feeding this back to the client system. Another example involves the researcher evaluating the actions undertaken by the client system and feeding data back to it. Also different schools of action research describe this cyclical process using lesser or greater number of steps. For example, Zuber-Skerritt refers to four phases (see above), while Checkland's (1981) Soft Systems Methodology outlines seven steps or phases.

In addition to the difference in the number of phases within each cycle, contemporary applications of action research also enable the use of different techniques for data and information collecting especially in the diagnosing and evaluating phases. These may include the use of questionnaires, semi-structured interviews or focus groups, with the choice often largely dependent on the researcher's skills and backgrounds. Literature reviews as well as records, memos and reports from the client system will also be commonly used.

The reason for the flexibility in method design is because action research is designed to deal with and respond to 'real-world' situations, unlike mainstream research where you can -- and should -- start with a very precise research question. Given a precise research question a study can then be designed to answer it, also with precision. However, given the nature of the social systems, action research design cannot be fully detailed in advance and then rigorously and inflexibly implemented. Rather the research design is emergent, meaning it develops progressively, influenced by the events that take place during the project and by the progressive analyses that are made. In action research the use of the elements that bring rigour into mainstream research (control, standardisation, etc.) would defeat the purpose. "The virtue of action research is its responsiveness. It is what allows you to turn uncompromising beginnings into effective endings. It is what allows you to improve both action and research outcomes through a process of iteration" (Dick 1993). As in many mainstream science procedures, the use of repeated cycles enable the action researcher and his/her colleagues to converge on an appropriate conclusion (Figure 3.2).

Figure 3.2. The iterative nature of action research (Source: Damme 1998)

It is by being deliberate and intentional about this process that you can maximise your learning. The rigour in action learning lies in the quality of the data and the interpretations of this to help people think about -- reflect on -- how they can improve the situation in question. "At each of the steps you learn something. Sometimes you are recalling what you think you already understand. At other steps you are either confirming your previous learning or deciding from experience that your previous learning was inadequate. This is equivalent to what Gummesson (1991) calls the 'hermeneutic spiral', where each turn of the spiral builds on the understanding at the previous turn. It is these - the responsiveness to the situation, and the striving after real understanding - which define action research as a viable research strategy" (Dick 1993).

Thus, in some sense of the terms, action research tends to be cyclic, participative, qualitative and critically reflective. All of these features (except the last) can be seen as choices to be made by the researcher in the context of the problem being studied (Dick 1993). And it is this process of critical reflection that distinguishes action research from everyday inquiry (Dick 1996, Wortley 1996, Bunning 1995) and also makes it a particularly suitable approach with which to help develop the change needed for areas such as environmental management and sustainable development. Indeed, in the sense that action research seeks alternatives to the status quo that will both illuminate what exists and inform fundamental change, it is a form of critical theory and seeks to stimulate critical reflection among human agents so that they may more freely choose whether and how to transform their world (Argyris et al. 1985 pp.70-71).

As Kemmis and McTaggert observe, to do action research one must plan, act, observe and reflect "more carefully, more systematically, and more rigorously than one does in everyday life: and to use the relationships between those moments in the process as a source of both improvement and knowledge" (1988 p.10). It is the process of reflection in this process, on one's own views as well as those of others, that provides the basis for learning -- enabling all those involved to develop a more holistic perspective of any given situation, within which they can best make their particular contribution.

The challenge for the action researcher lies in the fact that learning can be difficult, even at an individual level. Accepting new information that challenges the way we think and the things we do is, even with the best of will, difficult to undertake, to accomplish, and to sustain (Michael 1995). Finding out about problems also implies that we may have to act to correct them. What often stops us doing this is an anxiety, or the feeling that if we allow ourselves to enter a learning or change process, if we admit to ourselves and others that something is wrong or not right, we will lose our effectiveness, our esteem, and maybe even our identity. Most of us need to assume we are doing our best at all times, and it may prove a real loss of face to accept and even "embrace" errors. Adapting poorly, or failing to realise our creative potential may be more desirable than risking failure and loss of esteem during the learning process (Allen & Kilvington 1999).

Because of this, "learning, which mostly upsets beliefs and habits in individuals and organizations, is hardly likely to be embraced easily and enthusiastically, even though there is a growing, and sometimes powerful, recognition of the need for change" (Michael 1995 p.470). Indeed, as Argyris et al. (1985 ch.3) point out, individuals and organisations have a number of defensive reactions that resist change -- or learning -- by preventing open dialogue and the integration of new information which may challenge their existing worldviews (values, assumptions, paradigms, etc.). These defenses include making some subjects 'undiscussable' (Argyris et al. 1985 p.87), or an unawareness that their 'espoused theory -- the world view and values people believe their behaviour is based on -- is different to their 'theory in use' -- the worldviews and values implied by their behaviour (Argyris et al. p.82).

Accordingly, as Aryris et al.. (1985 pp.84-85) suggest that the first match to any inquiry into a mismatch betwen intention and outcome is likely to search for another strategy that will satisfy the 'governing variables', the belief systems and values which the individual or organisation is trying to maintain. For example if a land manager views his/her enterprise solely in terms of sheep production and notes that the vegetation condition of the land is deteriorating, the action strategy will likely be to try a different grazing regime. In such a case when new strategies are used to support the same governing variable (i.e. the land as a sheep production system) this is called single loop learning (Figure 3.3). A similar science example might arise in response to funder requirements for a scientist to be more participative. The response might be to find a 'friendly' group of people to work with that are happy to acknowledge the scientist as the 'unquestioned expert' - the governing variable.

However, another possibility is to change the governing variables themselves. For example rather than try a new grazing strategy, the land manager may choose to initiate a more open form of enquiry. The associated action strategy might then be to look at how the enterprise could function as a tourism, or forestry, system for example. The scientist may choose to involve appropriate stakeholder groups in a more collaborative approach, changing the role of science to one of a co-researcher and recognizing that the role of 'expert' is more a matter of perspective. These cases are called double-loop learning, and involve more fundamental shifts in people's belief systems and values. In this way they can often minimise the gap between espoused and theory-in-use.

Figure 3.3. Single and double-loop learning (Adapted from: Argyris et al. 1985)

Accordingly, Meziro (1991, quoted in Bunning 1995) draws attention to the need to address three elements through the reflective process: i) content, the substantive issues involved; ii) process, how such issues were raised and addressed; and iii) premises, which are the values, assumptions, paradigms and whole framework of individual and collective mindsets, which inevitably influenced what was attended to and what was not, and other issues such as goals, process and interpretation.

Developing double-loop problem solving approaches is thus a critical part of changing people's actions in respect to the environment. However, it also requires the action researcher to deal with the defenses of individuals and organisations -- which is no small undertaking! In many cases this will mean having to address situations in which participants may feel embarrassed or threatened. However, as Grudens-Schuck (1998 p.61) points out, unless research and education programs build specific processes for confronting people about unworkable theories and organizational defenses, the use of local knowledge and interpretations of events cannot be a sound foundation for collaborative learning and positive change.

The growing use of action research within environmental research and development initiatives explicitly recognise that natural resource management issues (such as biodiversity protection and enhancement) are not characterised so much by problems for which an answer must be found, but rather by issues which need to be resolved and will inevitably require one or more of the parties to change their views. The underlying assumption of these approaches is that effective social change depends on the commitment and understanding of those involved in the change process. In other words, if people work together on a common problem �clarifying and negotiating� ideas and concerns, they will be more likely to change their minds if their �joint research� indicates such change is necessary. Also, it is suggested that collaboration can provide people with the interactions and support necessary to make fundamental changes in their practice which endure beyond the research process.

Similarly, exploring the social process of learning about situations is inextricably linked with the acts of changing those situations. �Certainly surveys and other social research results are useful, but so is information on why different people see things as they do, and the political relationships between stakeholders. It is by bringing these aspects into the open, and stimulating debate between the different groups through action research approaches that the social parameters � so neglected in most analyses � are automatically brought into the process� (Bosch et al. 1999). Thus, the action research approach seeks to influence the phenomena being studied during the action research process itself, in the belief that the true nature of social systems (social, cultural and institutional considerations) become most evident when you seek to make changes to them.