Appreciative Inquiry and Lean
Implementation at DMS Production
Appreciative Inquiry and Lean
Implementation at DMS Production
Master Thesis by Camilla Damsgård Robertsen Cand.Merc. (Strategy, Organization, and Leadership) Supervised by Magnus Larsson Department of Organization Copenhagen Business School
Appreciative Inquiry and Lean
August 2010 Master Thesis by Camilla Damsgård Robertsen Cand.Merc. (Strategy, Organization, and Leadership) Supervised by Magnus Larsson Department of Organization enhagen Business School Characters: 181.987 Standard pages: 80,0
This thesis marks the end of my master degree at Copenhagen Business School and was written during the spring and summer of 2010.
I would like to thank the production director, team leaders, consultants, and employees at DMS Production for welcoming me to your workplace and providing me with an incredible interesting empirical foundation for my thesis.
I would also like to thank my advisor Magnus Larsson from Department of Organization for inspiration, ideas, and constructive comments.
At last, I would like to thank David and Henriette, my family, and friends who supported me through this process.
Copenhagen, September 2010.
Camilla Damsgård Robertsen
Formålet med dette speciale er at undersøge, hvordan en organisation, der allerede arbejder med Lean, kan implementere Appreciative Inquiry (AI) succesfuldt. Det besvares gennem en etnografisk analyse af DMS Production, en Novo Nordisk fabrik, som diskuteres ud fra aktør- netværksteori og skandinaviske neo-institutionalister. Ved at følge de forskellige grupper af aktører og deres kontroverser opnås indsigt i, hvilke udfordringer DMS Production oplever i forbindelse med implementeringen af AI og hermed nås frem til tre væsentlige kontroverser:
Kommunikationsproblemer mellem teamledere og Lean konsulenter, hyppige udskiftninger blandt teamlederne og bekymrede medarbejdere.
Specialet påpeger, at de tre kontroverser alle kan få gavn af en yderligere implementering af AI, men at det kræver, at implementeringen fokuseres på de ledelsesmæssige udfordringer AI ønskes anvendt til, et øget kendskab til AI samt at ledelsen inddrager alle de relevante aktører i den videre implementering. Det tydeliggøres desuden, hvordan implementeringen formentlig vil drage fordel af, at AI i højere grad oversættes til objekter og gentagne handlinger, der ifølge idé modellen vil bidrage til en større institutionalisering af AI.
På det teoretiske plan bliver specialets problemformulering hævet til at handle om, hvordan management idé nummer to introduceres i en organisation. Der konkluderes at dette kræver indsigt i, hvordan de to idéer adskiller sig fra hinanden særligt i forhold til den mængde af metode- og værktøjsmæssige elementer, idéerne indeholder. Samtidig kan eksisterende kontroverser være et godt udgangspunkt for fastsættelse af formålet med den nye implementering. Mere specifikt i forhold til AI og Lean diskuteres det, hvordan forskellighederne mellem de to idéer kan bidrage til organisationens identitetsarbejde i forbindelse med implementeringen. Der er tale om to ret forskellige management idéer, og udfordringen består i at få de to idéer til at støtte hinanden konstruktivt i stedet for at skabe en kamp mellem tilhængere af henholdsvis AI og Lean.
CHOICE OF THEORY 9
THE ETHNOGRAPHIC APPROACH 9
ANALYZING THE EMPIRICAL DATA 13
ORGANIZATION THEORY 15
ORGANIZATIONAL DEVELOPMENT 16
ACTOR-NETWORK THEORY 18
SCANDINAVIAN NEO-INSTITUTIONALISM 20
IMPLICATIONS FOR THIS STUDY 26
APPRECIATIVE INQUIRY AND LEAN 28
APPRECIATIVE INQUIRY 32
APPRECIATIVE INQUIRY AND LEAN 38
DATA ANALYSIS 39
CASE INTRODUCTION 39
OVERVIEW OF ACTORS AT DMSPRODUCTION 40
WHAT LEAN MEANS IN DMSPRODUCTION 45
WHAT AIMEANS IN DMSPRODUCTION 52
HOW AI AND LEAN ARE PUT TOGETHER 55
COMMUNICATION PROBLEMS 57
LEADER CHANGE 59
ACTOR-NETWORK THEORY 67
SCANDINAVIAN NEO-INSTITUTIONALISM 69
THE THEORETICAL IMPLICATIONS 74
THE EMPIRICAL IMPLICATIONS 77
RECOMMENDATIONS FOR DMSPRODUCTION 82
LIMITATIONS,WEAKNESSES, AND NEW PERSPECTIVES 83
HOW TO IMPLEMENT A SECOND MANAGEMENT IDEA 85
HOW TO SUCCESSFULLY IMPLEMENT AI IN A LEAN ORGANIZATION 86
FINAL CONCLUSION 86
This master thesis will deal with Appreciative Inquiry (AI) in a Danish context. AI was developed 20 years ago in Cleveland, Ohio. Today, Danish companies are starting to implement AI as a part of their leadership style or as a tool to develop their new strategy. One of these companies is Novo Nordisk. In DMS Production, a Novo Nordisk production facility in Hillerød, a Danish consultancy company (Resonans) started an AI process five years ago. This master thesis will analyze how the implementation of AI is going on. It will especially explore how the implementation of AI fits with the way the organization uses Lean. The thesis will analyze how the leaders of the facility today struggle with fitting the new ideas of AI with the existing Lean approach to the production. In a more general perspective, the thesis will then discuss how two very different management ideas such as AI and Lean are implemented in the same organization.
The research question is:
How can Appreciative Inquiry be successfully implemented in organization already working with Lean?
With implementation, I mean the process of integrating AI in the everyday life at DMS Production. The notion of implementation will be used in this broad manner through the thesis.
With a successful implementation, I mean, integrating AI so that it becomes useful for the organization. In the thesis I will use the notion “management ideas” to describe the category of fashionable ideas, such as AI or Lean, which leaders use as helpful guidance in their difficult task of leading an organization. In the thesis I will write Appreciative Inquiry (AI) and Lean with capital letters to clarify the use of the specific management ideas.
Why Is This Interesting?
The research question is interesting because it focuses on how two different management ideas are brought together. Lean and AI differ in some of their most central principles. Most fundamentally AI is about elevating from successes while Lean is about solving problems.
Another important difference is that AI places human beings and their relationships in the
centre, while Lean focuses on work processes. This is done because Lean has as its goal to make the production process more effective while AI is created to provide enthusiasm, innovation and cooperation; involving individuals and their relationships instead of work processes.
Because AI and Lean are so different, this thesis will also be able to consider the more general question of how two very different management ideas can be implemented in one organization.
The research question is also interesting because it considers the specific situation of DMS Production. DMS Production and Novo Nordisk in general have a large focus on Lean production and the Japanese production philosophy, though the leaders at DMS Production has a burning wish to implement AI – also because they would like AI to supplement Lean.
Thirdly, the research question is interesting because few Danish studies of how AI and Lean can be implemented in the same organization exist. For a short description of the existing Danish literature on AI and Lean see the appendix. Even though the case of DMS Production seems unique, it will probably be far from the last Danish initiative to implement AI in an organization working with Lean.
To answer the research question, it will be necessary to explore theoretical approaches to the implementation process and how it can be analyzed as well as to understand what the specific management ideas, AI and Lean, include. The theory on implementation will include different perspectives, though focusing on which theories will be most useful in exploring and learning from the case study. Lean and AI will be approached to learn what DMS Production is actually implementing as well as what elements they have not been implementing.
A thorough analysis of the case study will be necessary in order to gain a solid understanding of the empirical issues and challenges involved in implementing AI in a Lean organization. I will analyze how different organizational members understand and use AI and Lean as well as what implications the implementation process has had for DMS Production. The empirical analysis of DMS Production will be based on an ethnographic approach. Through such an analysis, the thesis will aim at concluding not only how the situation looks like for DMS Production but also
what they should consider when continuing their implementation of AI. During the analysis, interviewees and other people who have provided insights into DMS Production will be kept as anonymous as possible. A few times this has not been possible. There is only one production director and making him anonymous would be impossible. All employees have been kept anonymous. When referring to observations or interviews, I will therefore only write “Field notes” or “Field interview”.
It has of course been necessary to consider the limitations of this thesis. Early in the process, I decided that this thesis will not aim at classifying AI’s effectiveness as a management tool – neither in general nor in DMS Production. In the case study, the decision to implement AI was already made and the interesting question is now how to fit this with an existing work with Lean. Therefore I will not either consider the motivation behind the decision of implementing AI. I also decided that the critique of respectively AI and Lean will not be a part of this thesis, though theory presents many concerns about the problems with AI and Lean. Instead the analysis will focus on what the actual, empirical implications of implementing the two management ideas seem to be.
To conclude this introduction I will shortly present the structure of the thesis. Following the introduction a chapter on Method will present my considerations about the theoretical choices as well as the empirical study. The next chapter on theory presents some theories on implementation. The goal of this section is to find the most useful theories for analyzing the implementation of AI and Lean. The following chapter explores the background of AI and Lean in order to gain a solid understanding of the two management ideas.
The empirical analysis in the next chapter develops the empirical data to a comprehensive understanding of DMS Production and their implementation of Lean and AI. The analysis has the empirical data in focus, only drawing on the theoretical consideration where this seems necessary.
After the empirical analysis, the discussion will combine theory and empirical analysis to answer the research question. It will be followed by a conclusion and an answer to the research question.
The following chapter will consider the most important methodological questions. It will present my considerations about method in regards to choice of theory, data collection, and data analysis.
Choice of Theory
The case study of this thesis has taken a large and important role. It has in many ways been my point of departure and therefore decisions about theory and method have also been made to fit the case study. The choice of theory is therefore closely connected with the goal of exploring my solid empirical data and learn as much as possible from it. The chosen theories offer an important background for the deep empirical analysis and discussion rather than providing specific guidelines to the case. They can be viewed as tools to explore the implementation of AI in DMS Production and learn as much as possible from this analysis.
The approach to theoretical choices also made it relevant to explore the theoretical background behind Appreciative Inquiry and Lean. While the theory in general explores the implementation process and the analysis of it as such, the chapter on Appreciative Inquiry and Lean offers a more specific approach to the case study.
I will argue further for these choices during the theory chapter.
The Ethnographic Approach
To analyze the organization of DMS Production, an ethnographic approach to the case study has been chosen (Hammersley and Atkinson 2007). By ethnographic approach I mean that the case has been explored through spending a longer amount of time in the organization observing organizational members’ actions and interactions as well as interviewing specific members. The ethnographic approach has made it possible to give a comprehensive picture of DMS Production. The approach also made it easier to deeply analyze how Lean affects different parts of DMS Production and how this influences the implementation of AI.
The ethnographic study has included spending 2-4 days a week at DMS Production over a period of one and a half month. During this time I spent two full days at the production with different employees. I did two interviews with employees and many more informal talks, three interviews with team leaders, two interviews with Lean consultants and one interview with the HR consultant. I participated in four leader group meetings, a strategy presentation for the factory, one meeting between leaders and union stewards, one meeting between leaders and Lean consultants, and a two-day seminar for the entire factory, including all meals and social activities (spare time and evening party). I had several meetings and informal talks with the production director and one interview with one of the AI consultants. In total this sums up to more than 10 formal interviews and 6 meetings besides the seminar. Interviews, participation, and informal conversations resulted in detailed empirical data and provided me with an intense understanding of DMS Production how Lean has affected DMS Production, how AI has been implemented so far and what happens empirically when Lean and AI are implemented in the same organization.
While the ethnographic method has been an important part of my empirical study, there are central ways in which this study differs from a traditional ethnographic study. First of all, while studying DMS Production for one and a half month is much for a master thesis, for ethnographers this is still a rather short period. Added to that, the topic of my study was to a large extent determined beforehand, because my access to the case very much depended on my research to become useful for them as well. This also made my study a bit different from a traditional ethnographic study.
Being in the Organization
When doing an ethnographic study, the basic foundation has been an understanding of the world as socially constructed. Therefore human actions and interactions become the main point of study. In the analysis of DMS Production, I will therefore explore the meaning behind certain human actions within DMS Production at the time of study (Hammersley and Atkinson 2007). I will explore how employees and leaders make sense of their actions and how working and acting in this culture become meaningful to them.
Being an ethnographer requires participant observation (Hammersley and Atkinson 2007).
When studying DMS Production, it has been my intention to let things happen without too much interruption from my side. Therefore listening, viewing, and feeling the atmosphere has been a main focus. On the other hand, I have needed to be an active participant to some extent. I have asked questions and done interviews in the situations where I felt this was a more efficient way for me to understand issues and to make employees and leaders reflect on matters they did not reflect on purely by themselves.
I could never have completely avoided affecting the empirical research. The very fact that I was present at the factory changed details in how everyday life was effectuated and especially how employees behaved. It has been a guiding point for me not to try and completely eliminate this, which had also been impossible, but instead to use it as a tool for better understanding the organization and its members (Hammersley and Atkinson 2007). Employees’ and leaders’
reactions towards my presence often told much more about their personal experience of being at DMS Production than I could else have understood. These reactions indicated how comfortable they felt within the organization, how they viewed the future and how much they trusted the intentions of their leader, who in some way had “sent” me there or at least
“allowed” me to enter the organization.
When understanding that my presence and the reactions towards it should be used as a tool for understanding the culture, how I was seen by employees and leaders became quite important.
In general I made an effort making people comfortable with my presence. This included putting myself in the role of a “marginal native” whether interacting with leaders or employees.
Hammersley and Atkinson describe this role of being marginal native as a “marginal position of being simultaneous insider-outsider” (Hammersley and Atkinson 2007). I was naturally an outsider because I could not participate actively in events (such as meetings or in the production) as the organizational members, though I maintained an insider position through dressing appropriately and talking with the groups in ways that integrated me, such as laughing with them and in general using their language. Some days I had planned to meet both employees and leaders, and sometimes I even had meetings with both leaders and employee
representatives. This was harder than expected. I easily ended up with a double role that required balancing between the culture of the employees and of the leader group. At the same time, it seemed important that I maintained an understanding of me as the expert. This role created the authority that I needed to be accepted as naturally as possible as a part of the factory.
Time, People, and Context
When studying DMS Production, it has been necessary to consider time, people, and context, as Hammersley and Atkinson (2007) also point out. In regards to time, an important decision was to gather the ethnographic study within one and a half months. This made it more natural for me to visit the factory; employees remembered me and slowly accepted my presence. At the same time, I needed at least one or two days a week to digest the impressions I received when being at DMS Production. By the beginning I had planned to solve this dilemma through taking half days at the factory. In reality it turned out to be better taking full days, where employees during the day began to relax in spite of my presence.
The people I interviewed were from the beginning primarily member-identified (Hammersley and Atkinson 2007). Along the way I found it easier to choose myself whom to talk to and especially to ask for an interview with someone who seemed interesting for my analysis.
Interviewing many different employees, leaders, and consultants has been a high priority.
In regards to the context, I was very aware to divide my time between everyday activities and extraordinary activities (Hammersley and Atkinson 2007). The field study has aimed at including both the routine of everyday production, meetings etc. and extraordinary experiences such as larger presentations and the two-day seminar. This provided a much deeper understanding of how people within DMS Production reacted towards each other.
Interviews, Recordings, and Note Taking
During interviews I made an effort to create both open and more reflexive questions. I saw myself as an active listener. While the interviewee did far most of the talking, I was still active in my responses to what was said. Because I used how the interviewee reacted towards me being present at the interview I sometimes ended up asking more reflexive questions than I had
planned. Hammersley and Atkinson (2007) point out how interviewees will react quite differently depending on whether they are used to and feel comfortable with the interview situation. This was very much my experience as well. While interviewing leaders went very well, interviewing employees became more awkward and during the field study I therefore decided to eliminate such interviews and instead be very investigative in getting to talk more informally with employees during breaks or when spending time at the production floor.
Interviews and meetings were mainly recorded and afterwards transcribed in such detail as seemed necessary. Observations were not always possible to record and therefore taking notes became a central point. During meetings I also took important notes about the experiences that the recorder would not catch (atmosphere, body language etc.). Field notes and transcribed interviews has been the basis for my empirical analysis.
Analyzing the Empirical Data
Analysis of the empirical data already started during the field study. As Hammersley and Atkinson points out: “Ideas are used to make sense of data, and data are used to change our ideas” (Hammersley and Atkinson 2007, 159). Because the field study has been quite an intense period, there was not much time left for writing during the field study. Instead I prioritized to write down more reflective notes along my field notes and to read those notes between visiting DMS Production. Such readings provided a better understanding of which topics to go deeper into and thereby whom to talk with and what questions to ask.
A consequence of analyzing data during the field study has been that the field study has been used not only for observing but also for testing explanations. Especially interviews turned out to be a great chance for testing ideas and explanations. I also participated actively in leader group meetings, where I had the chance to present my observations and ideas and receive feedback from the leaders.
After the field study, I started the analysis with repeated readings of the field notes as Hammersley and Atkinson (2007) and Emerson, Fretz, and Shaw (1995) also stress the importance of. The field notes were systematically coded in order to separate the main
categories. The coding was based on the empirical experiences, theoretical ideas and common- sense expectations (Hammersley and Atkinson 2007). I found it especially useful to study the empirical data looking for the things I was surprised with, reactions I found interesting and situations that puzzled me. Such situations and reactions became central, especially to the development of the three controversies presented in the empirical analysis. Being an outsider made it easier to be surprised with things the organizational members had begun to view as natural. This could for example be how team leaders addressed employees internally, the communication problems between team leaders and Lean consultants, or the constant worries employees expressed. These surprising elements became the main themes for the further analysis.
From these methodological considerations I will continue with the chapter on theory.
Answering how Appreciative Inquiry and Lean can be implemented in the same organization is interesting from an empirical perspective because some organizations today are struggling with exactly that question. However, AI and Lean can be seen as fashionable management ideas which can be out fashion any time. So one could ask what general and theoretical interest this topic is a part of.
The basic issue of implementing AI in a Lean Organization is how to implement two different management ideas in the same organization. Implementing one management idea might be a complex task but implementing two ideas which are not made to fit together can only be more complex. One goal of this thesis is therefore to explore how to deal with implementation of two different management ideas in the same organization.
I will first of all explore this through a short presentation of how organization theory and Organizational Development deals with implementation. From this I will look closer at actor- network theory and Scandinavian neo-institutionalism for a theoretical approach which can provide a foundation for my empirical analysis and discussion. I will end this chapter discussing in more detail how and why I will use the different theories.
A traditional approach to implementation is based on organizational change which is a part of organization theory. Theory about organizational change within organization theory marks a shift from looking at stability to looking at the need for change within organizational structures (Hatch and Conliffe 2006).
Organizational change can be divided into two types of change, one where the source of change is external and one with an internal source (Goodman 1984). A change made because of external forces is called adaptation which means “the modification of the organization or its components to fit or to be adjusted to its environment” (Goodman 1984, 4). A change made
because of internal forces is called planned change. Instead of considering why change happens, planned change considers how to create change, the methods and techniques.
Organizational Development is a research tradition which considers the theoretical and practical aspects of planned change. Organizational Development includes specific values and assumptions. Such values “tend to be humanistic, optimistic, and democratic” (French and Bell 1999, 62). Because of these values, certain intervention methods are also viewed as a part of Organizational Development. These methods all focus on solving the human problems in organizations through looking at the organizational processes (French and Bell 1999).
Organizational processes are how things are done in organizations, as opposed to what is done.
Such processes “include communication, allocation of rewards, human resource practices, strategic management, exercise of authority, and self-renewal or continuous learning.” (French and Bell 1999, 4) This separation of task and process is specific for Organizational Development.
The Action Research Model
Because of the many different intervention methods which are a part of Organizational Development, choosing the right method has been an important concern. French and Bell suggest organizations to ask themselves some key questions: “What are we trying to accomplish? What activities/interventions will help us get there? What is the proper timing and sequencing of the interventions?” (French and Bell 1999, 146). A way of approaching the choice of intervention method is the action research model. The action research model, as taken from Kurt Lewin, presents how an early diagnosis is followed by a decision about which actions to take (Mills, Dye, and Mills 2009). According to this method, organizations should always consider the goal of an implementation before taking action. After this action but before taking any further action, the organization should evaluate what happened and again set up new goals for which specific actions should be chosen. French and Bell describes it this way:
“OD is thus an iterative process of diagnosing, taking action, diagnosing, and taking action. All organizational improvement programs are complex processes of goals → actions → redefined goals → new actions.” (French and Bell 1999, 3).
The action research model shows how OD systematically approaches the change process assuming that setting the right goals helps choosing the right intervention method and executing it the most efficient way.
Social Relations, Organization as a Whole, and Social Constructions
The different intervention methods within Organizational Development consider the practical part of organizational change from very different perspectives. From the beginning, Organizational Development has focused on social relations. Researchers such as Schein (1992) and Kotter and Heskett (1992) emphasized how change in organizational culture is an important part of organizational change. Organizational Development is also based on a total systems perspective (French and Bell 1999) and thereby views organizations as a whole where all parts of the system affect each other. Lean presents a way of changing the organization through this whole system approach. Another trend in Organizational Development is to look at the organizational reality as socially constructed. One example of this trend is Appreciative Inquiry (French and Bell 1999).
Other Ways of Looking at Planned Change
That the two management ideas to be considered in this thesis can be viewed as Organizational Development interventions does however not mean that the theory of Organizational Development is the right way to analyze the case study of the thesis. Organizational Development has a very rational approach to organizational change. For example does Organizational Development not see the right choice of intervention method as impossible, but just as something which should be carefully considered through analysis and determination of goals. A less rational view of the change process is provided by actor-network theory and Scandinavian institutionalism which I will use to analyze the implementation process. Another reason for this choice is that Organizational Development has a very normative and prescriptive approach to implementations (Mills et al. 2009). In this thesis the case study takes up a central role. In order to learn from the case study it will be necessary to use a theoretical background which opens up the case study instead of just providing the solutions.
Actor-network theory (ANT) is developed by Callon and Latour (1981) to describe how ideas are translated in a network of actors. It is not developed to be used for analyzing implementation but it provides a useful framework for the analysis. I will therefore now describe ANT through three important concepts: Translation, Actors, and Controversies.
ANT describes the concept of translation in different ways. It can be described as a process where one actor is placing himself as a spokesman for other actors, translating their will into one’s own language (Callon 1986). It can also be described as the network building process an actor takes part in to be able to speak on behalf of others (Scheuer 2006). Callon and Latour (1981) focus more on the negotiation which the process of translation includes:
“By translation we understand all the negotiations, intrigues, calculations, acts of persuasion and violence, thanks to which an actor or force takes, or causes to be conferred on itself, authority to speak or act on behalf of another actor or force” (Callon and Latour 1981, 279).
They analyze contract theory from Hobbes with ANT, describing the formation of a contract as an example of a translation process. The fundamental idea is that translation happens when an actor forces others to accept him as speaking and acting on behalf of them. While the case of the Leviathan is dominated by physical violence, the translation process might just as well be dominated by some form of persuasion (Callon and Latour 1981).
According to Callon, the translation process includes 4 steps (Callon 1986). First of all one or several human actors need to problematize an issue and create interest for a solution. Second, these actors need to find partners and thirdly, incorporate them in the project through giving them new roles. Fourth, the original actors need to mobilize these new partners to make sure these partners upholds a role of representing others in order for the translation process to continue to new parts of the growing network. This focus on the strategic translators; the original actor, makes it interesting to look closer at personal goals, motivations, and intentions.
The translation metaphor makes it possible for the researcher to study how the idea meets practice on the premises of the meeting (Scheuer 2003). This means that ANT has a basic
language useful for the specific empirical analysis (Latour 1996) but also that the basic concepts leaves room for the empirical setting to be the main influence on the analysis. Therefore the meeting between idea and practice can be studied in a rather simple way. This is exactly why I will use ANT for the analysis of the case study in this thesis.
Actors and Their Power
The second concept, actors, is a central idea in ANT. Actors or actants are not only human actors but also non-human actors. ANT therefore replaces the idea of knowledge as a sociopolitical matter with an idea of knowledge as a socio-technical matter (Scheuer 2003). In this way, objects are seen as creating the social world along with human actors.
When actors are involved in the translation process, the development of power becomes an important element. Callon and Latour describe that:
“An actor, as we have seen, becomes stronger to the extent that he or she can firmly associate a large number of elements […] Strength thus resides in the power to break off and to bind together” (Callon and Latour 1981, 292).
When translating an idea, the human actor will need to gather other actors in a network and this ability creates the level of power that the actor will later need when the idea shall be translated and become a part of the organizational practice.
Callon and Latour though also stress that “no actor is so powerful that its decisions and associations as a whole will be finally and definitely considered as technical reality” (Callon and Latour 1981, 298). This point is interesting because it emphasize that the translation process should be viewed as a continuous process. Instead of ever reaching a point where the idea can be viewed as completely translated, Callon and Latour (1981) point out that each time the actor successfully translates the idea as a part of the other actors’ interests, desires, or forces, this actor will increase his power.
Power is not only related to the actor introducing the idea, it is also very much related with the idea itself. The translation activity also makes an idea more powerful when this process of translation creates chains of actors in the network who all take action based on the idea
(Scheuer 2003). But it is important to remember, as Callon and Latour (1981) also point out, how this process is an unstable situation requiring constant convincement of all involved actors.
Follow the Controversies
When studying the process, ANT suggests that researchers “follow the actors” or “follow the controversies” (Scheuer 2003, 26). Different ANT researchers have focused on one of these two approaches. Following the actors will reveal how each of them experience the situation, while the idea behind following the controversies is that knowledge among the actors is created through how such controversies are solved.
During the analysis of the case study in this thesis, following actors and following controversies has been a main focus when considering how to approach the empirical data. I will follow the actors in the first part of the analysis, exploring how to understand the situation of DMS Production including the implementation of AI and Lean. In the second part I will follow the controversies to learn more about the actual translation process of AI and Lean.
During the 1990’ies Scandinavian Neo-institutionalists began using the notion of translation, taken from ANT. They created a theory of organizational change which includes both stability and change (Czarniawska and Sevón 1996). Combining the translation term of ANT with this new approach to change, a different understanding of the implementation process was created. I will now present how the Scandinavian neo-institutionalists replace the idea of implementation with the concept of translation, how they base this on an analysis of management fashions, and what new understandings of the translation concept they arrive at.
After this I will present two different models of translation from Scandinavian neo- institutionalists.
From Implementation to Translation
In the introduction I explained how I use the notion of implementation in a broad manner to describe the processes of introducing Appreciative Inquiry in an organization. I therefore do not attach implementation with a specific understanding of the process. On the other hand, the
Scandinavian neo-institutionalists often place translation in opposition to other concepts such as diffusion (Czarniawska and Sevón 1996) or implementation (Scheuer and Scheuer 2008).
Using the translation metaphor from ANT, the Scandinavian neo-institutionalists see themselves as able to leave earlier understandings of the implementation process.
Scheuer and Scheuer (2008) offer an analysis of how the translation metaphor can be used instead of the concepts of intervention or implementation to explain organizational change. I will shortly outline the analysis to show why the Scandinavian neo-institutionalists find the translation metaphor more useful than earlier concepts.
According to Scheuer and Scheuer (2008) the notion “intervention” has been used by researcher of healthcare interventions, focusing on evidence-based knowledge which is spread from decision centers in the organization to local knowledge users. Scheuer and Scheuer (2008) point out that this metaphor does not concretize how the process of implementation should happen and therefore ideas might very well never make the way to the local knowledge users.
The second metaphor Scheuer and Scheuer (2008) present is “implementation”. The metaphor of implementation is used by researchers within public administration and public policy (Goodman 1984, Scheuer and Scheuer 2008). This idea of implementation is based on a political perspective, where implementation of an idea is a political decision. The challenge is then for the political decision to manifest in real change. Research has then tried stating what the crucial variable for real change is. However focusing on the crucial variables, the policy goals of the implementation are easily forgotten and thereby the result is that policy outcomes are not in accordance with policy goals (Sheuer and Scheuer 2008).
The translation metaphor therefore remains as an alternative way of viewing organizational change. Using the translation metaphor should be a way of avoiding that ideas are never used by practitioners or that the outcomes of the change are not in accordance with the goals of the organization. Translating the idea to the organization makes it possible for practitioners to find the idea useful and for the organization to adjust the idea so it will fit organizational goals.
Scandinavian neo-institutionalists prefer the notion of translation for one other important reason. When ANT uses the concept of translation it is applied to a wide range of social ideas.
The Scandinavian neo-institutionalists use it more specifically to explain how organization should handle management ideas. Management ideas have been analyzed as management fashions. According to Abrahamson (1996) management fashions are shaped by socio- psychological, technical, and economic forces together. These forces create specific fashions to travel around the world, from one organization to another. Based on this idea of traveling management fashions, Scandinavian neo-institutionalists argue that a translation process, where the idea is adjusted to the organization, is important. Czarniawska and Sevón explain that:
“Organizational actors, like a collective ant-eater, catch many, spit out most, and savor some, presumably on the grounds of relevance to some organizational problem.”
(Czarniawska and Sevón 1996, 25).
If it was true that leaders in organizations would simply chose the management fashions which could solve their specific organizational problem, which is how Organizational Development approached, one should think the translation process was irrelevant, since the idea or fashion had been chosen with such a specific goal in mind, but Czarniawska and Sevón argue further:
“But the match does not lie in the attributes of an idea or in the characteristics of the problem […] The perceived characteristics of a problem and the match between them are all created, negotiated or imposed during the collective translation process […] With some exaggeration, one can claim that most ideas can be proven to fit most problems, assuming good will, creativity and a tendency to consensus. It is therefore the process of translation that should become our main concern.” (Czarniawska and Sevón 1996, 25) If any idea can be negotiated into fitting any problem the translation process suddenly becomes very important. One could even say that translation is a necessary consequence of the traveling management fashions. Viewing the process of the idea being introduced to the organization as a social, collective process also means that one cannot any longer talk about an implementation process but should instead view the process as translation of the idea.
The Neo-institutional Understanding of Translation
After presenting why the Scandinavian neo-institutionalists find the translation metaphor more useful than earlier concepts, I will continue with a deeper presentation of how they define translation. According to Scheuer:
“The term translation is connected with the network building activity and the specific job that an actant does, when he/she/it speaks or tries to be able to speak on behalf of a heterogenic actor-network” (Scheuer 2006, 15).
This understanding might be closely related to Callon and Latour’s definition (1981), though it minimizes the violent part which took up much space in the article on Leviathan. Instead Scheuer (2006) stresses how actors cannot only be forced but also convinced into doing what is needed for the idea to be introduced.
Czarniawska (2008) presents a slightly different understanding of the translation metaphor.
Differentiating between globalization and localization, she describes how a way of acting is being globalized by being translated into an abstract concept which is later localized through being translated into concrete action another place or time. In other words, her definition of translation is based on the process of going from concept to action.
Czarniawska (2008) stresses how translation of an idea includes forming this idea to one’s own needs. She describes this process through the gardening metaphor. Just as when a gardener moves a plant from one garden to another, the organizational idea must be freed from soil from the old place and carefully adjusted to fit the new place. In the situation of dealing with organizational ideas, soil is the elements of an implemented idea which are specific to its place of implementation.
Referring to a working paper by Latour, Czarniawska and Joerges (1995) point out how translation means displacing, drifting, inventing, mediating and modifying the idea. During this process, they supplements, not only the idea will change; the actor translating the idea will change as well. Czarniawska and Joerges develop this understanding to include a model over how the translation process takes place when the idea is successfully converted into an institution; the idea model.
The Translation Models
During the theoretical development of the translation metaphor among Scandinavian Neo- institutionalists, different models of translation have been created. These models include the idea model, the editing Model, the imitation model, translation as diffusion of recipes, and translation as association (Scheuer 2006). I will now present the idea model and the imitation model.
The Idea Model
The idea model is developed by Czarniawska and Joerges (1995). It describes the process of how ideas are translated in an organization, thereby creating organizational change. Simply put, the idea model explains how the translation process includes an idea which is turned into an object, which is turned into action, which is repeated and stabilized until it has become institutionalized in the organization. This model is very interesting when considering the translation of Lean and Appreciative Inquiry at DMS Production because objects and actions are usually important parts of how such management ideas are implemented.
Objectifying ideas can mean that ideas are made into linguistic artefacts or into actual designs (Czarniawska and Sevón 1996). This is an important part of the translation process because it integrates the idea as a part of the existing pattern of actions. The new objects and the linguistic artefacts become powerful and therefore this process might also very well change the idea itself. The process of making the idea and its objects into action happens in the individual organization but it is necessary that the action evokes positive emotions which create both excitement and a promising feeling (Czarniawska and Sevón 1996). In other words, members of the organization must see the action as potentially useful before being willing to engage in the particular action.
Czarniawska (2008) emphasizes that the institutionalization requires that the translation process includes the steps of idea, object, action, and repetition in the mentioned order.
Scheuer (2003) criticizes that this model does not describe the complexity of the translation process. He also argues that Czarniawska seems to focus on the idea and its travel instead of what happens in the organization. These are important critiques, though I still think the idea
model is interesting because it highlights some elements of the translation process which I have found in the empirical analysis.
An important difference between the idea model and this thesis is the claim of “original ideas”.
According to Czarniawska (2008), the translation model, as opposed to the diffusion model, requires one to accept that original ideas do not exist and that all ideas are local translations. As Scheuer (2006) points out, a consequence of this understanding is that one cannot talk about good or bad replications of an idea. Since this thesis considers how Appreciative Inquiry can successfully be implemented, viewing replications as good or bad seems almost necessary to grasp the fundamental question. However, I will also consider whether translations have coursed the idea to be well institutionalized in the organization. The institutionalization therefore also becomes an important part of my understanding of successful implementation.
The Imitation Model
The imitation model is developed by Sevón (Czarniawska and Sevón 1996). The imitation model is interesting because it combines the perspective of the idea model with a social logic of appropriateness. This creates an identity searching process where the translation of an idea is in focus. On the other hand, how the idea will be translated therefore also depends on how the organizational identity is formed along the way.
The model suggests that organizations imitate other organizations in order to learn from their experiences. Through this learning process, the organization develops its organizational identity. It does not mean that it copies the other organization’s identity but it finds inspiration from the other organization. Through the process of developing the organizational identity, the organization considers self-identification: “What am I like?” as well as the desired identification:
“What would I like to be?” (Czarniawska and Sevón 1996, 57-59). The distinction between those two different questions is important, though many organizations might not be very well aware of which question they are asking themselves or which question they are answering.
The idea of desired identification is also combined with a logic of appropriateness where organizations are expected to consider what they would like to be along with what is appropriate or suitable in the specific situation (Czarniawska and Sevón 1996). These
considerations about appropriateness make the choice of an idea as well as how it is formed and translated important aspects of not only the internal organizational identity but also the external image.
It adds even more to the complexity that: “organizations may have more than one identity that controls their attention” (Czarniawska and Sevón 1996, 57). This idea is especially interesting when considering the case of this thesis where two management ideas creates two translation processes which potentially both affects the identity work of the organization.
Implications for this study
I have now presented how implementation can be viewed as a part of organizational change which is studied by organization theory and especially through the study of planned change in Organizational Development. While Appreciative Inquiry and Lean can be placed within the tradition of Organizational Development I have already pointed out how I think ANT and Scandinavian neo-institutionalism will be more useful for the case study and analysis of this thesis. The argument was that understanding the implementation process in the case requires one also to be open for the less rational actions involved in the process. I would like to use ANT and Scandinavian neo-institutionalism to help reveal what is going on in the case, how actors understand their situation, and how interaction between actors makes sense rather than providing normative solutions for how the implementation process should take place.
The basic ideas from ANT about studying actors and their controversies in order to understand the process of translation will be used from the very beginning of the empirical analysis, as a tool to grasp the situation of the case. It is important to realize that different actors have different views on and understandings of the organization as well as of Lean and Appreciative Inquiry. When taking their controversies seriously, it becomes clear where the translation or implementation process has somehow encountered problems. When realizing which problems the actors actually face, though they might not be aware of them, the frames for a further implementation of AI are set.
Along with the concepts of actors and controversies, the idea of a translation process as it is described by both ANT and the Scandinavian Neo-institutionalists will be used for analyzing the case. The two different models from Scandinavian neo-institutionalism, the idea model and the imitation model, will also be used for analyzing the case and discussing what they can change to improve their translation process.
From this presentation of the theory on implementation, I will continue with a presentation of Appreciative Inquiry and Lean.
Appreciative Inquiry and Lean
After having considered the theories on implementation, it will be necessary to gain a deeper understanding of the two management ideas which this thesis focuses on. While the Neo- institutionalists think that we cannot talk about an “original idea” or an “original movement”
(Czarniawska 2008), it does not change the fact that each of these management ideas, AI and Lean, has written and quite specific roots from which most implementation processes are inspired. In the empirical analysis I will both look at the level of institutionalization but also compare the implementation of AI and Lean with the literature on AI and Lean. I will therefore now present how these “original ideas” are described in the literature on AI and Lean.
This section will aim at explaining the core principles behind Lean and the specific methods and tools that are used in Lean.
Toyota Production System and the History of Lean
In the beginning of the 20th century Toyota started producing vehicles in a way we today know as “Lean”. Monden (1994) presents Toyota Production System (TPS) which Womack and Jones (2003) later present as Lean Production or Lean Thinking. The basic idea behind Lean is that cleaning up the production will optimize the activities and secure a more effective value creation. In the beginning creation of value was understood as merely cost reduction, though with Womack and Jones’s book on Lean Thinking (2003) the concept of value also included enhancing the by costumers perceived value (Hines et al. 2004).
During the 1990’ies Lean became increasingly popular in the Western World. In Denmark production facilities started adopting Lean and today organizations and companies from completely different sectors are implementing Lean.
The principles of Lean
The fundamental principle of Lean is to optimize production through eliminating waste. Waste includes many different types: excessive production resources (workforce, facilities, or
inventory), overproduction, excessive inventory, and unnecessary capital investment (Monden 1994).
To eliminate waste Monden (1994) suggests following the 5S principles. The 5S principles in Japanese are: Seiri, Seiton, Seison, Seiketsu og Shitsuke. They order the organization to identify and separate necessary and unnecessary parts, maintain a clean work place, create a good place for work tools or parts in order to ease identification, and make sure the principles are adopted by workers and continuously followed. When 5S is continuously followed, workers create what is called Kaizen. Kaizen is ideas for activities that can improve the production efficiency.
While the 5S principles create a more lean production environment, Monden (1994) stresses that the goal on the human side is to create creative and flexible employees. Employees become creative and flexible through being involved in the process of 5S and constant improvement. Monden (1994) also points out that when waste is eliminated, workers’ jobs become more meaningful which keeps moral high and shows respect for humanity.
A central tool in creating a Lean production is systematic problem solving. Problem solving includes the following steps (Monden 1994): 1. Define the problem, 2. Find the reasons behind the problem, 3. Generate ideas for solving the problem, 4. Select the best idea, and 5. Present the best idea for management.
The more modern version of Lean problem solving includes the 5-Why method (Womack, Jones, and Roos 2007). The technique is that each problem should be solved through asking
“Why” 5 times till the most fundamental reason for the problem becomes visible. The general idea behind problem solving in Lean is that it integrates workers in the process of solving problems, generating ideas, and executing the chosen ideas.
“The truly lean plant has two key organizational features: It transfers the maximum number of tasks and responsibilities to those workers actually adding values to the car on the line, and it has in place a system for detecting defects that quickly traces every problem, once discovered, to its ultimate cause.” (Womack et al. 2007, 99)
This transformation of responsibility and detection of defects and problems are some of the main characteristics of Lean.
Lean Culture and Lean Thinking
Lean can be seen both as a way to organize work and as a leadership philosophy (Edwards, Bojesen, and Nielsen 2010). The philosophy of Lean is what Womack and Jones (2003) present as Lean Thinking. They explain that:
“Lean thinking is lean because it provides a way to do more with less and less – less human effort, less equipment, less time, and less space – while coming closer and closer to providing customers exactly what they want” (Womack and Jones 2003, 15).
The organization should become leaner while costumers’ wishes are kept as the main focus point. Womack and Jones (2003) describe how implementing Lean requires that managers take value as their starting point. Value is understood as value to customers. When value has been defined, managers should identify the value stream, create a production flow (instead of thinking production as batches) and thereby let the customers pull production (ibid). When managers install this as a continuous process involving employees and striving for perfection, the organization is on the way to creating a Lean culture.
Tools and Methods in Lean
While the philosophy behind Lean seems as an important way to understand the core of Lean, what is mostly implemented as the first thing in organizations are the tools and methods, Lean suggests. The original methods include:
Kanban (small tags telling how much should be produced) Just-in-time (producing what is needed when it is needed)
Smoothening production (maintaining a constant production pace - flow)
Shortening setup time (when starting a new production type on the same machine) Multi-function workers (every worker controls more than one machine at a time) Kaizen (continuous improvement activities)
Design of machine layout
Visible control system (boards showing what is going on right now etc.)
Autonomation (autonomous check of abnormalities)
When implementing Lean today, 5S, flow, Kanban, and Kaizen are still central tools. Shortening setup time is often talked about as SMED. OEE is another new term, a tool working with eliminating bottle necks in production. A very important method in modern Lean is value stream analysis. While Womack and Jones (2003) introduced this, Rother and Shook (2003) further explored the analysis and presented it in “Learning to See” and “Seeing the Whole”.
The Lean Organization
No matter how the organization translates the Lean philosophy and which tools managers decide to use, implementing Lean in a modern organization has consequences. One consequence is employees’ reaction. Edwards et al. (2010) describe from their thorough case studies that:
“In many cases Lean is received with passive acceptance and nothing that looks like enthusiasm. If the lean process stopped at a low development step, it is our judgment that employees would not require more development of lean or more involvement in the process.” (Edwards et al. 2010, 126)
This estimation supports the understanding that Lean is difficult to maintain as a central part of the culture even after it has been translated and implemented to the organization.
Another difficulty which often arises in Lean organizations is conflicts between operational management and top management or consultants (Edwards et al. 2010).
“Some projects are from the beginning very oriented towards consultants – perhaps supported by top management, where operational management is often neglected or even put in opposition to the lean initiative. A tension rises between lean development and operations, which often affects employees but also weakens operational management” (Edwards et al. 2010, 48-49).
Different researchers all suggest that trust is an important part of a successful Lean culture (Edwards et al. 2010, Melander et al. 2009, Holbeche 1998). While it is concluded that Lean is a continuous process that constantly risk lack of employee support or management conflicts, trust seems to be a way to secure a more successful Lean culture. While Melander et al. (2009) describe trust as top management committing to agreements that makes employees’ fear irrelevant, Edwards et al. (2010) develop the term further as a part of their notion of “social
capital”. They describe trust as being reached through integrity, authenticity and involvement of employees.
To sum up, modern Lean organization might use the original tools or the ones later developed and it might include a Lean philosophy as a part of the organizational culture. To be a successful Lean organization it must though also actively work on creating a trustful atmosphere where employees’ involvement is maintained while management conflicts are avoided.
This section will explore the theoretical background and core principles of Appreciative Inquiry.
The Background for Appreciative Inquiry
In 1987 Cooperrider and Srivastva published the article “Appreciative Inquiry in Organizational Life” (Cooperrider and Srivastva 1987) based on the work from Cooperrider’s PhD. This article marks the formal beginning of Appreciative Inquiry. It presents AI as a theory that will enhance social systems effectiveness and integrity. AI is an Organizational Development theory based on social constructionism. In psychology the development of positive psychology has been closely connected to the work of AI.
Social constructionism is an important foundation for AI. Gergen and Gergen (2003) present social constructionism as having three main roots. The first root is the assumption that knowledge always has communal origins, explaining how knowledge is created through social interaction. The second root is the centrality of language, which is based on the work by Wittgenstein explaining that the way we understand the world is created through the
“language games” we engage in (Gergen and Gergen 2003, 4). The third root is the saturation of knowledge, extending the understanding that traditional ideas cannot be viewed as objective truth to include that also no particular groups can claim authority of the truth.
Social constructionism stresses that there exist many subjective truths which are created in relations and dialogue. One implication of this is that all social actions are open to multiple interpretations and therefore narratives in form of stories and interpretations takes up a
central role. Another implication is that meaning as well as meaningfulness is believed to be created through dialogue and thereby, knowledge and meaningfulness are also closely connected.
The Principles of Appreciative Inquiry
AI has a series of principles. The practical principles are described by Cooperrider, Whitney and Stavros (2008). The first principle is the constructionist principle. It points out that language constructs the way we understand our organization and how we act within it. It also stresses the importance of viewing organizations as living systems that continually evolve.
The principle of simultaneity describes how inquiry and change take place simultaneous.
Changing an organization through AI (whether that is strategic development or more daily leadership) requires a process where inquiry is used positively as a part of the change process.
The poetic principle explains how poetry is a way of interpreting the things that are experienced. This happens in organization every day when stories about the organization are formed and continuously developed by its members. AI stresses that the organizational members should pay more attention to how they form these stories, what they focus on, and how they tell them to each other.
The fourth principle, the anticipatory principle, stresses the constructionistic idea that how we view the future controls our actions and thereby forms the future itself. Therefore our view of the future should not be neglected and most importantly, AI again stresses that it is up to the organizational members to create the view (and thereby the future) that they really want for their organization.
The positive principle explains how thinking, acting, and talking positively is a way of creating a generative environment in the organization. It is a basic assumption behind AI that human beings perform better when affected by positive social attitudes. The positive principle also reminds practitioners that problems can be understood as frustrated dreams of something better.
These principles are accompanied by some believed effects which are used as important arguments in AI theory. A central one is the Pygmalion effect which is based on how the Rosentahl studies showed that when teachers had been given specific expectations for different students, the expectations resulted in the children, whom teachers had low expectations to, soon started performing worse than children, whom teachers had higher expectations to. The Pygmalion effect tells us that expectations easily become reality, so if we expect a good future there is a better chance we will actually get it (Cooperrider et al. 2008).
Recognition, Appreciation and the Generative
Appreciative Inquiry can easily be misunderstood as being all about appreciation. In Danish it is even more difficult to avoid misunderstandings because we only have one word to use for both
“recognition” and “appreciation”. Appreciate means valuing, recognizing the best in others or increasing in value (Cooperrider and Whitney 2005). On the other hand, recognizing only means seeing others and letting others know you have seen them. Understanding appreciation as the same as recognizing would therefore mean losing an important part of what AI stands for. AI is both about praising and developing or increasing in value. The increase in value happens through the inquiring approach. Inquiry means exploring and discovering through asking questions (ibid). According to AI, these questions should be asked in an appreciative way and thereby being able to begin the increase in value that AI is searching for. In this understanding, the theory of AI focuses on unleashing information and commitment at the same time (Whitney and Trosten-Bloom 2003).
Another important term in AI is the generative. The theory of AI discusses how to create generative meetings, generative conversations, generative ideas, and generative interventions.
That any of these situations becomes generative depends on their ability to generate further discussions, ideas, and motivation among the involved parties. It is also described as a situation that causes people involved to change how they think and become open to possibilities. For an idea to become generative it is necessary that the idea seems compelling to people. When people are compelled by a generative idea, the successful result, according to AI, will be that 1.
The idea leads to new ideas or 2. The idea motivates people to take action (Bushe 2007).