1. Introduction

Many scholars in the action research community argue that action researchers should be better at writing about their practice (Pålshaugen 1996; Gustavsen 2003). Indeed, according to Greenwood (2002), action researchers write too rarely and when they do write they are often sloppy with methods. This is a serious challenge since writing good scientific texts is of the main ways in which researchers learn about and improve their practice. As Pålshaugen (ibid., p. 152) puts it ‘the process of writing a scientific publication on the basis of one’s own practice in an action research project is an attempt to come to a better understanding of this practice (…) it is an attempt to interpret and, thereby, later improve it.’

Against this background, it is surprising that there is so little methodological guidance available describing productive ways in which action researchers can get on writing about their fieldwork. While there is an extensive literature advising how action research projects might be successfully organised (Emery and Purser 1996; Emery 2003; Pålshaugen 1998), authors in this field have so far abstained from recommending to researchers how they can tackle challenges in writing about such projects, leaving this latter type of knowledge largely tacit or unarticulated. This is hardly a good strategy for improving research practice, or for visualising the value of action research in society.

In the present paper, we address these challenges and introduce the development story; a ‘recipe’ for writing offering a straightforward way of portraying roles and functions which action researchers take on when operating in the field, and of visualising ways in which these roles and functions bring about positive changes among field members. In the context of the present special issue on the role of the engaged researcher, we believe the development story is particularly relevant as it aspires to help engaged researchers reflect deeply on their past field practice and, in that way, improve future practice.

We have structured the remainder of the paper in the following way: First, we briefly explain how challenges in writing about action research differ from those involved in more mainstream social scientific writing. Then, we present the development story and consider its usefulness by applying it to our own experiences from an organizational development project. Finally, we conclude by discussing how using the development story affects the engaged researcher’s field practice.

2. Challenges in writing about action research and the inadequacy of mainstream social scientific methods

The action research community has developed its methods mainly through field practice. Knowledge about analysis and writing in the action research community is largely tacit, but it exists nonetheless shaping the choices we make as we systematise and write about our field experiences. There is no doubt that mainstream social science has put more emphasis on producing methodological texts that newcomers can make use of in order to cope with uncertainty in their analysis and writing activities. But are these texts valuable to action researchers?

In mainstream social science, there is a vast methodological literature about qualitative data collection, analysis and structuring (Glaser and Strauss 1967; Strauss and Corbin 1994; Eisenhardt 1995), and there are also practical guides and procedures that researchers can use in order to write meaningfully about their experiences (Alloway 1977; Van de Ven et al. 1999). For instance, Van de Ven and Poole (1989) offer a step-by-step procedure for objective analysis and writing about complex innovation processes. A well known example of such writing-recipes in the field of organization theory is the learning history method, developed by George Roth and Art Kleiner at the MIT Learning Lab (Kleiner and Roth 1997; Roth and Kleiner 1998). A learning history is a written narrative of an organization’s recent learning processes such as a widespread managerial innovation or the penetration of a new market. The document ranges in length from 20 to 100 pages and each page is presented in two main columns (see Fig. 1).

Fig. 1
figure 1

The learning history (Kleiner et al. 2000, p. 5)

Full size image

In the first column, the learning processes are described by the field members who took part in them or were affected by them. The second column contains separate analysis and commentary by the learning historians, who are usually trained academics (Kleiner et al. 2000). The lower end of the page contains background information about key events and organizational context.

Although the learning history method facilitates rich description of organizational learning processes, the fact that it separates sharply the researchers’ narrative from that of the field members makes it inadequate for action research writing. In our experience, this distinction makes it difficult to show clearly how action researchers’ direct collaboration with organizational members at different intervals impacts on organizational learning and improvement. In fact, the inadequacy of the learning history as a support tool for action research writers vividly illustrates the gulf dividing action research and mainstream social science. While the mainstream social scientist requires methods to help her describe how detached she was from the field, the action researcher needs exactly the opposite; methods to help her portray how very closely involved she was. In the following section we introduce an alternative recipe for writing that sets collaboration between researchers and field members at centre stage.

3. Coping with the challenges: The promise of the development story

Before presenting our alternative recipe for writing, it is necessary to briefly describe our background, that is; to outline the institutional and theoretical foundations on which we build this alterative.

3.1. Institutional and theoretical background

The institutional framework of the action research programme in which we are currently engaged, Value Creation 2010 (VC2010), is an agreement dating from 1983, between the Norwegian labour market parties, to give both financial and research-based support to organizations who wish to improve their situation by means of extensive employee participation in development work. The establishment of this agreement was largely due to the Scandinavian culture of industrial relations, which stems from a long tradition of co-operation between the two parties of employers and employees (Pålshaugen 1996, p. 147). Presently, both the Work Research Institute and Western Norway Research Institute, where the authors of this paper are employed, participate in the VC2010 programme. A large part of our work consists of participating in developmental work at company, network and innovation system levels by interacting closely with broad sets of actors, helping them to organize their development work into broad development coalitions.

The theoretical foundation on which we build our recipe for writing is literature about communities of practice (Wenger 1998; Wenger and Snyder 2000). Communities of practice are groups of people informally bound together by shared enterprise and passion for a joint enterprise (Wenger and Snyder ibid, p. 139). Participants of such communities share their experiences and knowledge in free-flowing, creative ways fostering new approaches to problem solving. We believe that organizations normally consist of a constellation of different communities of practice, which frequently span the borders of the organization itself, and that much learning and improvement occurs at the interface between such communities (Hildrum 2007). To contribute to the development of an organization, it is thus necessary to facilitate dialogue within and across communities of practice.

Our approach to field work analysis and writing, and the basis on which we develop our recipe for writing, did not occur in a vacuum but is largely founded in our work experience from the VC2010 program and theories about communities of practice.

3.2. The development story

Our motivation for writing this paper stemmed from a set of difficulties we encountered when writing about our experiences from the VC2010 programme. While discussing possible strategies for dealing with these problems at a research workshop in May 2004, we developed the idea of the development story (see Fig. 2).

Fig. 2
figure 2

The development story

Full size image

The development story consists of three vertical columns crossed by four horizontal dimensions. The left hand column describes the field context at the time when the researchers made their first contact with the field members or alternatively at the initiation of a particular project or project phase when such contact had already been made. We assume that any action research field includes several communities of practice and that people within and across these communities are connected by a set of relationships, and sometimes a set of shared experiences. Field actors may also share arenas of dialogue on which they interact around a common set of boundary objects, such as mutual development goals and shared project plans. Boundary objects are physical artefacts, documents, terms and concepts around which diverse and dispersed communities negotiate the meaning of technical concepts, mutual goals and practices (Wenger 1998). To function as a boundary object an artefact must be agreed on and shared by the interacting agents while being flexible enough to allow for slight variations in meaning ascribed to it by different communities (Star and Griesemer 1989). The latter two dimensions are highly relevant in the sense that the goal of any action research project is to instigate and retain constructive broad dialogues around shared boundary objects.

The right hand column describes the field context at a moment of reporting or as the project is completed and the researchers depart from the field; it describes new communities of practice, new relationships, new arenas of dialogue, new or renegotiated boundary objects. To map out researchers’ contributions to these changes it is necessary to describe how they collaborated with the field-actors’ practices during different phases of the project. Hence the centre column describes the researchers’ interaction with the field members. It shows the relationship between emerging challenges that crop up at different intervals and the various and changing roles and functions that the researchers choose to play out in order to cope with the challenges. The shift of emphasis between one role-type and another occurs as the researchers face different dilemmas at various stages of the project, reflect on these and change their behaviour accordingly.

It is important to note that the development story is not a generic model designed to capture the essence of all action research projects. Rather, it should be viewed as an empty frame with a set of hooks on which writers might attach important topics and thus illuminate aspects of their behaviour that would otherwise have escaped their attention. In the following, we examine the practical value of the development story by using it to analyse and present an organizational development project in four industrial service firms.

4. A participative organization development project in Four industrial service firms

Between February 2004 and April 2005 researchers from the Western Norway Research Institute (WRNI) and Work Research Institute (WRI) took part in a participative organization development project in the municipality Flora on the Norwegian west coast. A participative development project aspires to bring together broad groups of actors both within and across organizations—such as owners, managers, employees and union representatives—for democratic dialogues about improvement. The project received funding from The Norwegian Confederation of Trade Unions (LO) and the Confederation of Norwegian Business and Industry (NHO) and was organized through the national research programme Value Creation 2010 (VC2010). The four firms which participated in the project are part of the INC group, a loosely integrated consortium of 18 industrial service firms linked together by shared ownership (see Table 1).

Table 1 The project partner firms
Full size table

The purpose of the project was to test participative organization-development methods as a means of improving the firms’ marketing capabilities and of enhancing organizational democracy. In the course of the project, the researchers organised five dialogue conferences, one in each company and one joint conference for all union leaders in the INC group. A dialogue conference consists of a series of dialogues between the participants, who iterate between small parallel discussion groups and short plenary sessions in which the main points from the group discussions are reported (Pålshaugen 1998).

Following the structure of the development story, we continue by describing the field context as the researchers entered and began interacting with the field members.

4.1. State of emergency: The field-context at the researchers’ entry

Covering 11,000 inhabitants, Flora is the largest municipality in the county of Sogn og Fjordane and the most important industrial branches are shipbuilding, oil, fish breeding and seafood processing. In late 2003 several local firms were downsizing and 1500 out of Floras 5000 jobs were threatened. The central branches of the labour market parties got involved and investigated ways of mitigating the crisis. One proposal was to organise participative development projects within and between local firms. This, it was hoped, would improve the firms’ capacity to respond flexibly to new challenges thereby making them safer and more competitive workplaces.

At this point the INC firms encompassed a number of idiosyncratic communities centred on different professions and practices such as welding, plumbing, electrical installation, engineering and business administration. Although the employees had extensive experience with flexible teamwork cutting across community- and firm-boundaries, there was little cooperation above the level of small teams and there was scarcely any direct interaction between managers and employees. Indeed, the only arena of regular dialogue between managers and union representatives was a yearly wage-negotiation meeting. The union representatives of the various INC firms did not have a tradition of cooperating across firm boundaries; each representative attended primarily to his or her own firm and had little knowledge of what the others were doing. Finally, there were no mutually approved strategies, development plans or projects that could function as boundary objects between the firms’ different communities of practice. A good indicator of this low level of internal communication is the fact that the employees regarded the local newspaper their main source for information about their own workplace.

4.2. Participative development in Flora: The researchers’ practices in the field

In late 2003, the regional branch of LO tried to initiate a participative development project but, as the region’s industry had no previous experience with this kind of development work, it was difficult to find companies willing to give it a try. To tackle this challenge LO arranged several meetings in which representatives of local firms and unions met researchers from WRNI and discussed possibilities for joint projects. During one of these meetings, in February 2004, the INC group agreed to carry out a pilot project in four of its companies. The basis for this decision was the agreement between the Norwegian labour market parties to support organizations willing to improve their situation by means of participative development work. This agreement formed a foundation from which the INC group’s managers and union representatives started discussing joint strategies for improving the workplace. A project planning group was set up encompassing managers, union representatives and researchers from the VC2010 program. The purpose was to gain broad acceptance for the project in the companies and to initiate the project as soon as possible.

4.2.1. Phase 1: Scepticism and trust-building

Two weeks after the February meeting, the INC group’s managers abruptly cancelled the project plans. The reason for this was that they were generally sceptical towards external consultants and felt uncomfortable with the researchers’ involvement. As one manager put it: ‘The researchers came across as project makers looking to make some easy money by conducting time-consuming and irrelevant research inside the INC group.’ However, after telling their side of the story and listening to the researchers’ response, the managers agreed to set up a new meeting. In a series of subsequent meetings in March, April and May the researchers built sufficient credibility for themselves to resume concrete preparations of a participatory development project.

As a first step, the researchers made arrangements to expand the meetings to encompass a more diverse set of actors from the INC group. The purpose was to develop an operational project plan and to anchor it firmly through all levels of the firms. The meetings proved necessary as there were many unresolved issues that had to be dealt with before a mutually accepted project plan could be put in place. The managers insisted that the plan had to complement the INC group’s main business strategy. The union representatives partly agreed to that, but pointed out that the disintegrated structure of the INC group caused problems for the unions and that the project should also deal with this issue. For instance, one representative noted, if the INC group had been more strongly integrated the unions could have claimed a full-paid representative on the consortium level. Some of the union representatives looked at organization development projects with profound suspicion and argued that similar projects in the past had led to nothing. In the words of one representative: ‘Where is the result of the time studies from last year?’

In the course of these expanded meetings, the researchers listened to the participants, contributed to the discussions, summarized and reported back. Finally, after a new series of meetings the managers and the unions reached an agreement about a preliminary project plan. The goal was to test participative development methods and to achieve a culture-shift in the group by moving from a strong production orientation towards a stronger market orientation. To reach that goal, the project planned to organise a sequence of broad dialogue conferences in which employees from all four INC firms as well as representatives from key customers and suppliers would participate.

In this phase of the project, the researchers’ main task was to build confidence in the project and to earn the trust of the managers, union representatives and employees. The key to this was to listen to everybody’s opinions, and to make sure that all parties’ needs were reflected in propositions about how a future project would be conducted. Having gained a certain degree of credibility as independent participants, the researchers could take the role as intermediaries between managers and union representatives helping these groups to reach a compromise. This initial trust-building turned out crucial for the researchers’ capacity to keep the dialogue going in subsequent phases of the project.

4.2.2. Phase 2: Divergent project intentions and brokering

After the initial project planning meetings in the spring of 2004, the researchers contacted the central branches of LO and NHO to present the preliminary project plans. This was necessary as these organizations were to fund the project, and thus had to formally accept the plans before the project could be initiated. The central branch of LO responded very negatively and threatened to withdraw funding. Arranging inter-firm dialogue conferences with the predefined goal of promoting market orientation would, one LO representative argued, would make it difficult to achieve the goals of broad participation and democratic dialogues. According to the LO representative, the INC group had moved too fast. In his view, it would have been better to organise smaller internal conferences in each of the four firms, so that the managers and the employees could get familiar with working together and start searching for mutual instead of predefined development goals.

At this juncture, the project preparations came to a halt once again. The researchers found themselves facing two important actor-groups with divergent project intentions. To solve this problem, the researchers started brokering between the INC management and LO by way of frequent e-mails and telephone conversations. In this process, the project plan functioned as a boundary object between INC and LO in the sense that the researchers sent it back and forth between the two groups and adjusted it gradually according to comments until a compromise in the shape of a mutually accepted plan had been achieved. The compromise ended up close to the original intentions of LO. Four separate search conferences were to be organised, one in each of the firms, but the conferences would still be organised according to the explicit purpose of strengthening the firms’ internal market orientation.

Thus, facing a new challenge, the researchers changed their roles from that of the trust-builder towards that of the broker, assisting the field actors in negotiating around a shared boundary object. A key lesson learned from this phase was that in project preparation processes, researchers must make sure that all relevant actors are kept involved. Although it is tempting to move rapidly, it is important to take the time to ensure that all parties are satisfied asking questions such as: Are you still with us? Have we run too fast? Do you and your colleagues agree?

4.2.3. Phase 3: Communication problems and translation

The conferences, one in each of the four firms and one for the union representatives, were organised between November 2004 and April 2005 and were jointly coordinated by the researchers, union representatives, managers and a representative from the central branch of LO. Each conference was organised as a sequence alternating between small group work-sessions and plenary meetings. While the purpose of the small group sessions was to discuss important challenges and opportunities facing the firms, the plenary sessions aimed at visualising the diversity of challenges and finding ways of arriving at mutually accepted development strategies.

In the course of the conferences, the employees found it difficult to talk about developing their own workplace, especially when the boss was present. Many found it awkward to discuss these matters with people they did not know personally and chose to remain silent in the first sessions. To deal with this challenge, the researchers intervened directly in the small group discussions and tried to motivate the employees to talk together. The researchers also held plenary pep-talks where they tried to visualise how the conferences could benefit the employees. These talks seemed to help as more and more people chose to enter the dialogues. But as the dialogues gained momentum, so did new kinds of challenges. One was that people from different parts of the organization attached different meaning to the same concepts, such as ‘development,’ ‘production orientation’ and ‘market orientation,’ and thus continuously misinterpreted one another. In response to this, the researchers engaged more actively in the dialogues rephrasing and translating between people who had trouble understanding one another. Resembling the case with the project plan, these concepts functioned as boundary objects between different communities of employees in the sense that different interpretations were discussed until a mutual understanding had been achieved. Yet another challenge was that some participants, notably managers, took very dominant roles in the dialogues preventing others from making their voices heard. To mitigate this problem, the researchers intervened and tried to neutralise the dominant person by encouraging silent participants to voice their opinions.

Hence, as with the two previous project phases, the third phase involved a specific set of challenges and required a specific set of roles and functions to mitigate these. Since trust had already been established at this stage, the field members perceived the researchers as independent parties who intervened in the groups with the sole intention of improving the process.

4.2.4. Phase 4: Complexity and analysis

Following the last conference in April 2005, the researchers faced the task of converting a huge pile of handwritten notes and large paper sheets with pullet points into a collectively approved project report. The paper sheets from the plenary sessions were of limited value as they concealed a large amount of suggestions and anecdotes that had emerged in the group sessions. The notes from the group sessions, on the other hand, were overly complex and partly incomprehensible. The role of the researchers thus shifted towards that of the analyst. All written material from the conferences was transcribed and systematised into conference reports. Reports were then submitted to the conference participants for comments. After the researchers had received a broad set of comments, they updated the reports accordingly, resubmitted and requested approval. On the basis of the collectively approved conference reports, the participants made a binding commitment to follow up on their mutual development goals.

Taken together, the researchers’ main task in the project was to promote progress in the dialogue by motivating the participants to identify and discuss important matters related to organization development. In performing that task, the researchers experienced a variety of different challenges and solved these by taking on different roles and functions as necessary. The first challenge was the INC managers’ strong scepticism towards external researchers and consultants. After the researchers had gained the trust of the INC group through a long sequence of meetings, the project’s fund-raisers objected to the project plan. The researchers then started operating as brokers between LO and INC, using the project plan as a boundary object between the two. Then, during the conferences, the participants had trouble communicating. The researchers responded by instigating constructive dialogue and clearing up misunderstandings. Finally, facing the complexity of the written conference results, the researchers started operating as analysts converting the pile of jumbled notes into comprehensible reports, submitting these to the field actors and gaining general approval. It is necessary to note that the different role types and functions did not replace another. Rather, the researchers took on several roles and functions simultaneously and shifted emphasis between these as new challenges required.

4.3. Small steps forward: The field-context at the researchers’ departure

We can now proceed to describing how the field context looked when the project ended and the researchers departed from the INC group. Although the project lasted for only one year, it is possible to identify some changes within the INC group. First, both employees and managers argued that their relationship was closer and less hierarchical at project completion, than it was at the start of the project. The employees now have experiences with liaising directly with the managers, as well as with voicing their opinions about challenges and opportunities in their day-to-day work. There is no guarantee that this dialogue will continue, but the field members have established new arenas of dialogue to pursue the project-results further. For instance, in one of the firms the manager now has meetings with the employees every Friday to discuss matters related to the workplace. Some of the employees we talked to argue that there is a stronger intra-firm ‘community-feeling’ than before. The community-feeling also extends across firm-boundaries. For instance, union representatives from the different firms within the INC group have launched a monthly forum in which they discuss mutual challenges. It can also be argued that the INC managers’ long term goal of improving marketing capabilities through continuous participative organization development is now shared by a larger segment of the employees. Finally, on a more general level, the employees have acquired new skills in undertaking participative development projects; this form of work is no longer entirely new to them and similar future projects might turn out easier to undertake.

Although it is possible to detect some changes in the firms between the researchers’ entry and departure, the question of whether the researchers had anything to do with these remains to be answered. Would the process have been noticeably different if the firms had chosen to organise the project by themselves? To answer that question it is necessary to revisit the development story.

4.4. Development story revisited: Relating changes in the field to the researchers’ practices

We have now arrived at what is probably the most difficult task of action research writers, namely to demonstrate relationships between their own actions and changes in the field. Let us begin by inserting our main findings in the development story framework (see Table 2).

Table 2 Development story revisited
Full size table

Before commenting on Table 2, it is important to note that we are not trying to demonstrate causal relationships between the researchers’ practices and changes in the field. We do however, believe that the conferences contributed to altering many relationships within the INC group and that the project preparation process and the carrying out of the dialogue conferences would have broken down at several instances without the researchers’ interventions. For instance, it is doubtful whether a participative development project would have been organised in Flora at all without the researchers’ effort to gain the trust of the INC managers in the project preparation phase. Furthermore, if the researchers had not served as trusted brokers between the INC managers and LO, the project funding would probably have been withdrawn. Finally, the conference group discussions, which led to more trustful relationships between managers and employees, would almost certainly have been much less lively and fruitful without the involvement of the researchers. The potential for improvement and organizational democracy in the INC group is presently far from fulfilled. Whether the researchers’ actions will induce really profound effects depends on the extent to which the INC group is successful in pursuing the development process further.

5. Concluding discussion

In this paper we have presented a new recipe for writing, the development story. It is now time to discuss how the development story has affected our subsequent field practice. Are we now better action researchers?

Put simply the development story helped us identify important field experiences and organise events, allowing us to get an overview of the entire INC project from start to end. In turn, this made it easier to see the effects what we did in the field. While writing the development story we found that the field actors initiated the project on the basis of a small patch of common ground; the 1983 agreement between the labour market parties provided a starting point for collaboration. In addition, all actor groups were eager to turn the firms into safer and more competitive workplaces. Our main role as engaged researchers in the project was to encourage a broad dialogue towards achieving this mutual goal and subsequently negotiate strategies to safeguard that the dialogue did not break down. This involved organising discussions over different means of communication as well as saying the right things to field actors in critical moments during the preparation and execution of the dialogues.

Writing the development story forced us to reflect deeply on these roles and actions, and how they impacted on the dialogue. It also assisted us in identifying and discussing alternative courses of actions and alterative role types. This, we believe, has helped us avoid some stumbling blocks in our subsequent fieldwork. For instance, the development story shed light on the implications of our failure to involve the INC project’s funding agency in the initial planning phase. As a result, we presently put more emphasis on involving all relevant actor groups from project start and onwards. The development story also made it clear to us that the project plan functioned as a boundary object between field actors with divergent viewpoints. As a consequence, we are more aware of the potential of project plans and other documents as vessels of negotiation, to be used actively to help field actors reach compromises. It is of course possible that we would have made these learning experiences in any case, but we believe the development story helped us see more clearly relationships between broad involvement, boundary objects and researchers’ ability to facilitate constructive organizational dialogues.

This paper shows that the development story has some value for action researchers, but it also has limitations. For instance, it is difficult to write a development story about an ongoing project. The project must be completed and all relevant field experiences must be available before it is possible to create a really comprehensive overview of researchers’ collaboration with the field. For this reason, the development story works better as a support-tool for post-project reflection and learning, than as a tool for within project decision support. Moreover, although the development story may be useful it is not sufficient to solve all difficulties involved in analysis and writing. While the development story can help us identify and link together important field events it does not help us mark out which ones, out of a potentially huge amount of important events, we should write about in order to address a particular audience or complement a particular strand of existing research.

The most important point that emerges from this paper, however, is that action researchers should direct more attention to difficulties involved in analysis and writing. There is a great need for more methodological literature about writing that is adapted to the specific challenges of action research. The recipe that we have presented in this paper is but one out of many possible ways of trying to deal with those challenges. If experienced action researchers really want to publicise what engaged researchers do, they should do more to share their skills in analysis and writing with neophyte colleagues who also have valuable research experiences, but find it difficult to write meaningfully about these.