Volume 11, No. 3, Art. 14 – September 2010
Researching Practice and Practicing Research Reflexively. Conceptualizing the Relationship Between Research Partners and Researchers in Longitudinal Studies
Harald Tuckermann & Johannes Rüegg-Stürm
Abstract: Reflexivity has gained considerable significance in organization studies over the last decade. The discussion provides significant insight into interpreting one's findings and into presenting them to the academic community. But, the relationship between research partners in practice and researchers remains less in focus. Nevertheless, it provides the foundation from which data and interpretations emerge. To provide an orientation, this paper aims to develop a methodological heuristic inspired by social systems theory to conceptualize the relational foundation of longitudinal research. This "reflexive system theory heuristic" is illustrated empirically by reconstructing the research process of a longitudinal single case study on a change process in a hospital. The heuristic helps to observe and explain the dynamic relationship in the field and to explicitly acknowledge the status of empirical findings when understanding research as a generative activity. The heuristic conceptualizes the researched and research practices as communication systems, thus consistently developing a reflexive approach for field research. By taking into account the practice as well as the academic context, the heuristic can be extended to the insights drawn from the discussion on reflexivity. For social systems research on organizations, the heuristic provides a starting point to foster a methodological discussion.
Key words: organizational change; methodology; social systems theory: reflexivity; longitudinal studies; process perspective
Table of Contents
1. Introduction: Reflexivity in Organization Studies
2. Insights on Reflexivity and the Reflexive Systemic Research Model
2.1 Reflexivity in organizational research
2.2 A social system's perspective on research and practice
3. Illustration: Challenges in the Research System on Change in a Hospital
3.1 Merging two nursing departments
3.2 Challenges and dynamics shaping the research system
4. Discussion: Mutual Protecting of Research and Practice in Relation
5. Applicability, Limitations and Conclusion
1. Introduction: Reflexivity in Organization Studies
Reflexivity has gained considerable significance in organization studies over the last decade. It corresponds with the rising complexity and dynamics scholars perceive within the organizations they study. Reflexivity is integral to a process perspective that assumes social reality as becoming, rather than as existing objective entities (LANGLEY, 2009; TSOUKAS & CHIA, 2002). Such an epistemological stance assumes that insights are at "best partial and incomplete" and provide alternative interpretations, without claiming an objective truth independent of the observer (MORGAN, 1983, p.369). Reflexivity brings into view the assumptions, paradigms, perspectives that influence, if not shape, the results when we theorize from process data (LANGLEY, 1999) and present our findings to the academic community (HARDY & CLEGG, 1997). We become more conscious of our own assumptions and their influence on the findings. Reflexivity enhances our sensibility to the scope and the limitations of explaining organizational phenomena, which contributes to organization studies (WEICK, 1999). 
The relation of researchers and their practice partners within the organization provides the foundation of longitudinal empirical work (DUTTON & DUKERICH, 2006). How practitioners and researchers interact is particularly important in such studies that aim for direct observations (LANGLEY, 2009; MINTZBERG, 1979). Here, researchers engage with an organization over a certain period of time. A relationship evolves as both partners mutually interact with and observe each other. The researcher explores the organization for her or his interest. The organizational counterparts observe the researcher doing so. Particularly salient during periods of organizational change, practitioners face situations of high uncertainty, ambiguity and pressure to decisions. As in our own research (RÜEGG-STÜRM, 2002; TUCKERMANN, 2007) it is then likely that practitioners expect feedback, or even specific support from researchers. Such expectations emerge as part of the relationship between researchers and research partners. They are common to research processes which include a longitudinal and direct observation (e.g. BARLEY, 1990; IEDEMA, DEGELING, WHITE & BRAITHWAITE, 2004; VAN MAANEN, 1982). Handling practitioners' expectations marks a practical challenge for researchers. But handling such expectations tends to be undervalued in conventional academic practice which often calls for passive observing to avoid distortions by our own involvement (see ALVESSON & SKÖLDBERG, 2000). 
Despite its importance, the relationship between researcher and practitioners—who we call research partners in this paper—is hardly an issue in most methodological sections of academic papers (LANGLEY, 2009). Reflexivity in organization studies appears rather concerned with the researcher. Scholars deconstruct the research practice of others (ALVESSON & SKÖLDBERG, 2000), or they self-reflexively focus on the obtained material (ALVESSON et al., 2008; LEWIS & KELEMEN, 2002) and on presenting these findings to the academic community (GOLDEN-BIDDLE & LOCKE, 1993; HARDY & CLEGG, 1997). Self-reflections on conducting research in the field, on the dynamic relationship between research partners and researchers, are rarely reported and often anecdotal (BARLEY, 1990; IEDEMA et al., 2004; VAN MAANEN, 1982). There are few integrative frameworks that assist to systematically consider this relationship as a core component of conducting research. At the same time, prior research argues the need to include practitioners more actively in our field work (BALOGUN, HUFF & JOHNSON, 2003). Due to the importance of the relation in the field, its rare empirical as well as conceptual elaboration we pursue the question: "How can we conceptualize the relationship between research partners (practitioners) and researchers (academics)?" 
The purpose of this paper is to argue for a methodological framework, which extends the underlying assumptions regarding the organizational phenomena under study to researching them. We turn research on ourselves (WHITTINGTON et al., 2003), based on the assumption that research itself is a situated, social, communicative practice, just like the practice we explore (MORGAN, 1983). 
The heuristic we offer conceptualizes the relationship between research partners and researcher for conducting empirical work. Theoretically, it serves to systematically study organizational practice recursively interwoven with research practice. It includes both partners in their respective contexts. These components form what we call the "research system." The research system follows a communicative approach inspired by social systems theory (BAECKER, 2005) and is therefore termed "reflexive systemic research" (RSR) heuristic. It contains the involved research partners and researchers as a social dimension, the focus and topic of research (factual dimension) and the time dimension as an episode (HENDRY & SEIDL, 2003). The heuristic is reflexive in that the observers become visible in their specific context. Hence, the RSR-heuristic translates the metaphorical meaning of reflection—that is the view in the mirror through which the observer becomes visible in her or his specific context (LUHMANN, 1990)—into observable domains. Such domains can be distinguished analytically, despite and because they are interwoven in the real life of a research relationship. 
The RSR-heuristic can contribute a conceptual orientation for researchers in the field. Based on our own research of accompanying a change process in a hospital, we illustrate that and how the heuristic helps to observe and to explain the dynamic relationship between researchers and their partners in organizations. As in our case, it provided a basis to handling the challenges of the relationship with our partners productively. Furthermore, the RSR-heuristic acknowledges the status of empirical insights. Insights stem from descriptions which emerge from the research system, rather than from the organization. Insights are therefore propositions about the organization rather than direct truths, because both organizational and research practice operate within their own settings (KIESER & LEINER, 2009). In this view, research is not a discovery but a generative process. 
Specifically for research within social systems theory on organizations, the RSR-heuristic offers a starting point to foster a methodological discussion. So far, methodology was either rather a topic to reconstruct the societal system of science in general (LUHMANN, 1990). Or, field researchers have developed their own methodological approaches, including our own (JUNG, 2007; MINGERS, 1995; TUCKERMANN, 2007). They all share self-reflexivity, but system theoretical field research on organizations still entails deficits (PFEFFER, 2001). 
Our central argument builds on the importance of the relationship between research partners and researchers. It is vital for gaining and sustaining access to a research setting, and for the quality of data and interpretations. Gaining a more conceptual insight into the fragility, dynamic and complex encounters of temporarily engaging with an organization therefore not only provides a basis for scholarly quality. But it also strengthens the practical relevance of scientific work, particularly if we manage to engage practitioners in reconstructing the organizational reality they continuously construct (BALOGUN et al., 2003). In light of the pivotal importance of the fragile field relations, a more in depth understanding can assist to handle the double hurdle of practical relevance and scholarly quality (PETTIGREW, WOODMAN & CAMERON, 2001). In order to tap on this potential, we consistently conceptualize research as a situated social practice, thus similar to the practice we explore in organizations. To this end, social systems theory provides a promising foundation (LUHMANN, 2000). It explicitly avoids placing the observer outside the observed, and instead assumes relationship. Such a relationship emerges in communication, throughout which the involved actors mutually observe each other. Mutual observing involves expectations which emerge as structures that shape the research system. 
The remainder of the paper is structured in four sections: The next section contains a discussion of reflexivity in organization studies on the basis of which we suggest social system theory to develop the reflexive systemic research heuristic. The third section illustrates the heuristic empirically by reflecting on our own research process. In section four we summarize our learning from the empirical research and relate the heuristic to process oriented longitudinal studies on organizations. The final section five concludes the paper with its limitation and suggestions for further research. 
2. Insights on Reflexivity and the Reflexive Systemic Research Model
Reflexivity in organization studies has gained increasing importance. The social sciences in general and that of organizations in particular have experienced a crisis of representation which calls into question the role of the subject in research (HARDY & CLEGG, 1997). The domain of process studies questions the possibility for objective, neutral, all inclusive observing and analyzing. Accordingly, scholars become more aware of their practices of theorizing (WEICK, 1999). They interpret their own interpretations, in order to reconsider their relationship to the research process and to the generated insights. 
The term "reflection" or "reflexivity" is vague and provides different meanings, while being regarded as a genuine human process (BRANNICK & COGHLAN, 2006; HOLLAND, 1999). For the scope of this paper that refers to interpretive work, we understand it as "the interpretation of interpretation" (ALVESSON & SKÖLDBERG, 2000, p.6). Research is understood as a generating endeavor, and not one of discovering what exists independent of the observer (WEICK, 1989). It involves careful interpretation of the data and reflection. Reflection acknowledges the constructive dimension of language, and the specifics of a particular context. In our understanding, such a context refers to the organization and to the researcher. The former is oriented to a contextualist understanding (PETTIGREW, 1997). The latter assumes that observing, interpreting or explaining depends on the perspective of the observer, who is oriented mainly towards an academic community. 
2.1 Reflexivity in organizational research
Reflexivity involves two general avenues: one reflects on the perspective of other perspectives, while the other is self-reflective. This latter pathway concerns handling reflexively the material obtained, the methods used to obtain them, and the relationship between researchers and research partners of the organization during the times in the field. 
Regarding the former, particularly critical studies deconstruct the research processes of others. Highlighting implicit assumptions, values, norms or power relations argue that research is dependent on the observer. Mats ALVESSON and Kaj SKÖLDBERG (2000) offer a detailed investigation into various methods to alert us of the otherwise hidden observer and the influence of his perspective on the results. Similarly, reflection on methodology within social systems theory is traditionally concerned with the use of methodology within the societal system of science (LUHMANN, 1990). Here, methodological issues are more a subject of reflection, than a practice explicitly applied to empirical research. 
The second avenue of reflexivity in organization studies is self-reflexive. Research increasingly takes account of itself, that is its influence on what is considered important throughout empirical research, how it is explained, and what kind of theories are generated from these insights (WEICK, 1999). 
Self-reflexivity concerns either conducting or publishing research (ALVESSON et al., 2008). The latter often takes on the relationship between the researcher and the material or data and the process of generating insights or theory. A wide range of insights, strategies, and practices have been elaborated on within this domain of relating one's insights to the academic community the researcher aims to address (e.g. GOLDEN-BIDDLE & LOCKE, 1993; HARDY & CLEGG, 1997). 
Conducting research concerns the relationship between research partners and researchers from which the insights and the data emerge in the first place, albeit recursively interwoven with publishing them. A main focus in conducting research is placed on the process of interpretation or analysis on the side of the researcher (ALVESSON, HARDY & HARLEY, 2008; LANGLEY, 1999). Here, the relation between the researcher and the empirical material is reflected upon by being explicit about one's own assumptions to demonstrate the scope and the limits of the generated insights. Alternatively, some even aim to minimize the researcher's influence on the material (MORGAN, 1983). Another pathway is to follow an alternative template strategy (LANGLEY, 1999) or to adopt a multi-paradigm approach (LEWIS & KELEMEN, 2002), respectively. The latter involves multi-paradigm reviews, research and theory building. The first strategy strengthens the researcher's consciousness of the merits and limits embedded within a certain paradigm. The second enables an immersion in different and alternative paradigms. The third strategy strives for developing "more accommodating understandings by juxtaposing and linking disparate paradigm representations" (LEWIS & KELEMEN, 2002, p.261). The scope of these strategies contains the researcher's insights or interpretations of the data already gathered in the field. 
Besides reflecting on how to interpret the empirical material, self-reflexivity on conducting research concerns the relation with research partners in the field. Approaches to participatory action research (SANKARAN, DICK; PASSFIELD & SWEPSON, 2001) argue for an intense collaboration between the involved partners. A similar claim is made as organizational research takes a practice turn exploring more micro aspects of strategizing (BALOGUN et al., 2003). Practitioners are to be involved more actively in research activities, and are invited to reflect on their own practice in order to allowing researchers to adapt to the increased complexity within organizations. Including the practitioners is aimed to enhance more relevant research, apart from enhancing access to the organization. Anne LANGLEY (2009) notes, it is particularly in process studies that the relationship between researcher and the organization requires or involves a degree of reciprocity. On the one hand, data is collected in situ. On the other hand, such encounters take place over a prolonged time period. The relationship between organization and researchers evolves as both interact with each other. 
Such interactions involve systematic challenges regarding roles, expectations, and differing interests. In empirical detail, confessional tales in ethnography provide often anecdotal, inspiring and informative insights into the relationship between the researchers and their partners in the field. John VAN MAANEN (1982) highlights his involvement in police work and the dilemmas faced during the field phase. Stephen BARLEY (1990) adds the focus on the iterations of observing and analyzing during his study of hospitals. From a discursive perspective, Rick IEDEMA and colleagues (2004) reflect on their feeding back research results in a hospital. These accounts demonstrate the fragile, dynamic, and complex dimension of longitudinal research in practice and the significance of specific local contexts in which they take place. 
At the core of such challenges lies the relational foundation of a research project. Jane DUTTON and Janet DUKERICH (2006) argue for so-called high quality connections between researchers and their practice partners as a foundation to handle the unavoidable dynamics within the field, to allow for mutual learning, and to sustain momentum within the research context. The relational foundation is the keystone underlying good quality in research projects. This insight is shared by works of social constructionism, where relations are assumed as the basis for constructing realities (VAN DER HAAR & HOSKING, 2004). This epistemological assumption can be transferred to the field in terms of appreciative inquiry, for instance, viewing research as an inter-active process of generating insights regarding a social phenomenon. Constructing reality is embedded in the multiple realities of the different perspectives involved. 
The discussion of reflexivity enhances organization studies to access the complexity of the social by means of generating alternative perceptions and by giving voice to alternative interpretations, while raising awareness of the situated, observer-dependent scope and limits of research in organizations. As such, reflexivity does contribute enormously to enrich our understanding of organizations within interpretive, process oriented research. 
In summary, the above discussion of reflexivity in organization studies bears insights on four domains: The first domain concerns the research interest and focus. Researchers intend to explore the organization. Our brief discussion on "reflecting on others" emphasizes the observer dependence of seemingly objective insights. The second domain regards the relation between the researcher and the obtained material. It is highlighted that the analytical use of different paradigms impacts theorizing in this part of conducting research. As a third domain, we mentioned briefly reflections on the writing process which focus on the researcher relating the generated insights to her or his academic context. The fourth domain of conducting research emphasized the relationship between research partners and researchers. It appears rarely reported, despite being considered the foundation for the quality in research. 
Following the insight of the relational foundation of research, we focus mainly on the fourth domain. The first three domains have contributed to studying organizations by emphasizing the observer-dependence of findings and how to handle it in terms of theorizing. But it is within the fourth domain, from which the material, data, interpretations emerge in the first place. The anecdotal and rarely reported empirical insights express the dynamic, complex, fragile and challenging processes of engaging in the field. They demonstrate the challenges in gaining as well as sustaining access (BARLEY, 1990), the dilemmas of the researcher's role during the field phase (VAN MAANEN, 1982), or the fragility of providing feedback during member validation (IEDEMA et al., 2004). Overall, they demonstrate the importance of the relational foundation as the basis for field work (DUTTON & DUKERICH, 2006). Building on these insights, and inspired by our own reflective anecdotes, we suggest to further conceptualize them. 
Accordingly, we pursue the question: How can we conceptualize a heuristic for self-reflexive research that considers the four domains? The purpose of this paper is to suggest a methodological systematization of the four identified domains of reflexivity in terms of what we call the reflexive systemic research (RSR) heuristic. We focus primarily on the relationship between researchers and research partners. It is the locus from which data and insights emerge. Therefore, the relationship between researchers and practitioners lies at the core of longitudinal empirical research, although within the context of the other three domains. 
We build on the above insights by considering research and practice within the same theoretical foundation, in order to turn research on ourselves (WHITTINGTON et al., 2003). The relationship between research partners and researchers are placed in the center. This resonates with the need to include practitioners more actively in research in a symmetrical relationship (BALOGUN et al., 2003; VAN DER HAAR & HOSKING, 2004). As a consequence, research is considered context specific (PETTIGREW et al., 2001). These contexts refer to the research partner and the researcher who needs to be explicitly visible in such a heuristic. It allows to observing him or her within the specific context. For that matter, we suggest to re-conceive the way we conduct research using a social system theory perspective. 
2.2 A social system's perspective on research and practice
Social system theory (LUHMANN, 1984) provides a promising foundation for the proposed heuristic. It contains an explicit communication and organization theory for the purpose of conceptualizing the research process (LUHMANN, 2000). It is increasingly applied to organizations (HENDRY & SEIDL, 2003). Moreover, social system theory allows us to consider the above insights of reflexivity. The first one—conceptualizing research and management practice within a unified framework—is addressed in this section. 
2.2.1 Sketching social systems theory
As to the conceptualizing of organizational and research practice from the same theoretical foundation, social systems theory focuses on communicating as organizing practice. Practices denote actions that need to be understood, emerging from the communicative interplay of those involved (LUHMANN, 2000). Communication is the fundamental element of social systems. It is "the synthesis of utterance (including physical movements as well as speech or writing), information and understanding" (HENDRY & SEIDL, 2003, p.179). Thus, communication is not a transfer of objective information between a sender and a receiver but a socially accomplished generation of meaning. 
Social system theory analytically distinguishes individuals, and social systems, like organizations (LUHMANN, 2000). Individuals are so-called psychological systems. The basic element of their operations is thoughts, which can be in the medium of words, images, feelings, sensations, or the like. Processing thoughts indicates that one piece of thought connects to another. The distinguishing feature of individuals is their consciousness, which means the awareness of themselves as processing thought. An outside observer is unable to enter the thoughts of an individual, what is observable though are the actions, or utterances of an individual. As individuals are capable of perception, they provide a relevant environment for social systems, without which interactions or organizations could not operate (SIMON, 2007). Such perceptions can contribute in terms of an utterance to a social or communicative system. What happens next with such utterances is subject to the structure and processes of the social system at a given point in time. 
Such social systems are for example interaction systems and organizations. According to Dirk BAECKER (2005), interaction systems, like conversations or meetings, are communicative systems that emerge as individuals perceive their mutual perceiving. What defines an interaction system is therefore the physical presence of individuals (see WATZLAWICK, BEAVIN & JACKSON, 2003). Any action, behavior, utterance or the lack of them can be regarded as a communication. As communications connect or are followed by other such elements, communication forms as a process that organizes itself from understanding (LUHMANN, 1986, p.95). It unfolds its dynamics because the understanding oscillates between the actors, and their respective utterances are to be understood. 
Structures emerge from expectations the participants form with regards to what and how they understand each other's utterances. Mutually held or stabilized expectations are structures of communication. Structures are repetitive and actualized in the processes of emerging events. Such communicative events disappear in the moment they emerge, so that processes are understood as events unfolding over time. The relation between structure and process is mutually constitutive in that processes are sequences of occurring events. Structures in turn influence the probability which elements rather occur than others. 
Interaction systems are temporary and can form limited structures (LUHMANN, 1984). Mainly, they concentrate on a topic and often form around different roles ascribed to the participants. They also draw on certain rules, like the one that one person talks at a time. Interaction systems can therefore adapt flexibly to changing participants, topics or other issues. But, due to the same reason of comparatively limited structures, interaction systems are much more fragile in terms of time. Once the conversation or the meeting ends, the interaction system ceases to exist, even though it may be continued at some other point in time. 
Like interaction systems, organizations are also social systems. In comparison, they can form a more differentiated set of structures that allows them to continue operating, irrespective of the physical individuals' presence. Structures like routines, rules, positions, decision-making procedures may be formal or informal and are actualized in the daily processes of organizing (LUHMANN, 2000). The basic communicative elements of organizations are decisions. Accordingly, organizations are understood as interrelated networks of decisions (HENDRY & SEIDL, 2003). They are dynamic and action generating. Whereas interaction systems draw their boundary on who is present, organizations mark their limit according to membership. In this sense, organizations provide an environment for interaction systems. Vice versa, interaction systems serve as a locus to handle such incidents, for which the organization has not yet developed respective structures and processes. In other words, interaction systems provide necessary environments for the organization to continue functioning, while the organization provides the environment for these interaction systems to occur. 
2.2.2 Research as second-order observation
In terms of social systems theory, researching organizations can be understood in terms of a factual, a social and a time dimension. The former concerns the focus and content of the research as so-called second-order observation. The latter two regard how research takes place, namely as a temporary interaction system between research partners and researchers. This interaction unfolds over time generating its respective structures which enables and restricts the research process. 
With organizations as networks of interwoven decisions, research can be understood as searching for how and why that takes place. Both of these levels refer to the concept of observation or observing. Formally, observing is defined as an operation of distinguishing and marking one side of this distinction(LUHMANN, 1984). Handling distinctions can be assumed within individuals processing thought, as well as within social systems processing communication. Interactions, like conversations, or organizations as decision-systems, are observing systems. Research means to explore the observing of organizations, that is: how and why does an organization mark a side of a distinction and what distinctions does it use? Whereas the organizational observing is on a first-order level, research is referred to as second-order observation in that it focuses on the observations being processed within the organization. Organizational research takes on the distinctions with which a social system operates. Second-order observation in organizations is concerned with premises of decisions. Such decision-premises form part of the structure. They are the mutually held expectations which guide decisions. These decision-premises provide coherence for the participants so that the process of communication continues (WILLKE, 1994). 
Distinguishing between first and second order observation allows us to consider organizational and research practice reflexively within the same terms. One important implication of the concept of observing is that observations are partially blind. Observation is the use of a distinction with which we separate perception in two domains and mark one of them. Using other distinctions, or observing differently, gives rise to different domains and marks. Therefore, and first, observation cannot be all inclusive, but is selective. Second, the operation of observation is blind to the distinction it employs in that moment. Strictly speaking, it is impossible to observe an organization and simultaneously observe this observing, because it would imply an observation with two distinctions at the same time. In the words of Niklas LUHMANN and Peter FUCHS (1989, p.10, our translation): "A system can only observe what it observes. It cannot observe, what it does not observe. Furthermore, it cannot observe that it does not observe, what it does not observe." The blindness applies to both, the organizations and their researchers. 
By means of social systems theory, research and organization can be conceptualized within the same theoretical terms. Research is a specific kind of human practice (MORGAN, 1983). In terms of conducting research, it can be conceived of as an interaction system in order to explore, and in relation with the organization. Research focuses on a second-order observation to explain how and why the researched social system operates, while observing is blind to its own operating requiring self-reflexivity. 
2.2.3 The reflexive systems research model
The reflexive systems research heuristic attempts to address the insights we draw from the above discussion on reflexivity in organization studies. The RSR-heuristic aims at allowing the researching perspective to be observable itself (reflexivity), at highlighting the relationship between researcher and practitioner (involving practitioners), acknowledging the specific context of organizations and research (context specificity), and at starting from a common understanding of research practices and researched practices (turning research on ourselves). The model conceptualizes research as a communicative social practice. In theoretical terms of social systems theory and as researchers become involved with the research partner, their communicative interacting unfolds over time. We call this interaction system with its unfolding process and its dualistically emerging structures the research system:
Figure 1: Reflexive systems research heuristic 
The research system contains the following structural components: Alongside the social dimension, we find at least two observers involved, researcher and research partner, both of which can of course be more differentiated. On a factual dimension, the research system focuses on a specific topic and a certain unit-of-analysis within the organization to be investigated. In our case, these were change initiatives within a hospital. With regards to the time dimension, research can be understood as episodes, somewhat similar to strategic episodes (HENDRY & SEIDL, 2003). Broadly speaking, episodes contain a beginning, a conducting and an end. 
The components of the research system along the factual, social and time dimension mutually constitute each other. They are enacted as the research process unfolds. The process' elements are communicative acts that take place between the researcher and the research partner who mutually observe each other. Such mutual observing allows for structures to emerge, which are understood as mutually held expectations. Structures in the social, factual and timely dimension stabilize and enable the process, while simultaneously constraining it. 
Taking social system theory to develop the model allows to understand theory as a part of its topic and the observer as a part of or in relation with the observed (LUHMANN, 1990). Despite these similarities, this theoretical perspective acknowledges the differences in that practitioners and academics refer and orient themselves to their respective environments, marked by the "distance" in the figure above. Practitioners are oriented towards the functioning of their organization and its future prospects. Researchers are interested in generating insights about such a practical phenomena, and are oriented towards their scientific community. In short, practitioners are in need to act or to decide, respectively, whereas researchers are in need of reconstructing such evolving series of decisions. Finally, both operate within their respective spheres: practitioners are coupled to the organization and researchers to academia. Not only do they pursue different outcomes—viable decisions for the organizations, and viable publications within science, for instance. But also, the respective sub-systems in which they participate operate according to their own structures (KIESER & LEINER, 2009). In the context of these differences, practitioners and researchers engage within the research system as an expression of their temporary relationship (or coupling). Thus, both share a double belonging to their respective contexts organization and science as well as to the research system. 
It is mainly because of these different horizons of reference between research partners and researchers that inconsistencies, misunderstanding, conflict, contradictions or paradoxes can emerge throughout the research process. They are observed by those involved as disappointments to their expectations, may they be experienced positively or negatively. Disappointments mark the dynamic in their relationship. Such a dynamic is the norm, rather than the exception, despite being often omitted in academic writing. In our cases the dynamic was integral to our research. The practitioners faced uncertainty, ambiguity and pressure to decisions in conducting their change initiative. 
Strictly speaking, organizational research as proposed here implies that research takes place within the research system. Even though it may be physically within the organization, research as an interaction system is not. It is a distinct temporary social system that is designed to generate insights about one of its relevant environments, the organization. Due to the operational closure of each system, research results are not direct truths, but propositions of the researcher. Such interpretations are subject to the academic system. 
If the insights emerge within the research system, any insights can be related to the organization and/or the research system when making sense of a piece of data. Because of the researcher's ignorance during initial stages, distinguishing the two domains can be challenging. Vice versa, initializing the research system with practitioners bears potential insights on their organizing. Research system and organization are temporarily coupled by means of the involved people. Their interacting with researchers may therefore demonstrate organizational practices and perspectives. In this respect, the research system may provide a source of plausibility in its resonating organizational practice and dynamic. 
A second implication relates to the individuals a researcher is engaging with. Distinguishing research as an interaction system from the organization can assist the researcher in interpreting practitioners' communications that are perceived as different in encounters with the researcher and within the organization. To this end, it is helpful to distinguish between individuals as psychological system and as persons. The former is an operationally closed system that processes thought. Psychological systems are not elements of a social systems, but relevant environments. As human beings we couple ourselves to communications through the communicative structure of the "person." The person is not the psychic system, but the bundle of expectations regarding this psychic system. Such expectations stem from the psychic system and from the communication system (SIMON, 2007). 
The RSR-heuristic resonates with the four domains pointed out in the discussion of reflexivity in organization studies: Research and organization practice draw on the same theoretical foundation. At its core lies the domain of the relationship between research partners and researchers. It therefore shifts the focus from the researcher towards the ongoing interaction patterns with the practitioners, without disregarding their respective, specific contexts. The research system shapes and is shaped by the employed methods for data gathering and analyzing, which in part applies to the relationship between researcher and the obtained material. Interpreting one's generated insights and presenting them relates the researcher to the academic community, specifically to peers and reviewers in terms of publishing and presenting the findings. Finally, the research interest and focus of the researcher himself points towards observer dependence, particularly when conceptualizing research as a communicative system within the RSR-heuristic. 
The RSR-heuristic allows placing the researcher inside the research process. Research becomes observable itself, enabling a second order observation on research. Furthermore, the RSR-heuristic assumes the process to be dynamic. Distinguishing analytically between the research system and the organization, between the actors involved and the different contexts to which they refer as well as their distinct perspectives, suggests that surprises, misunderstandings, contradictions are to be expected. They are an integral part of conducting research in the field, expressing its fragility and complexity, while its participants face uncertainty and ambiguity. Therefore, stabilizing the research system over time is a core challenge, and one on which we will focus in the following empirical illustration. 
3. Illustration: Challenges in the Research System on Change in a Hospital
Studying organizational change directly involves systematic challenges. Often they crystallize during the field phase with regards to the role of the researcher (VAN MAANEN, 1982). Taking the researcher's perspective, the challenge of role highlights the relationship between practitioners and researchers. This relationship is illustrated with an example of our own research on change processes in a hospital. 
The following description draws on a single case study (TUCKERMANN, 2007) and follows broadly the episodic distinctions used in confessional tales: accessing the research context and establishing relations; conducting research; and terminating the field stay. The time ranges from 2003 until 2007, with a focus on the events that shaped the research system, and particularly our relation with the project team. 
3.1 Merging two nursing departments
The following illustration reflects on the research process that evolved as we investigated the change process of a nursing department within the context of a hospital merger. In 2003, the Kanton's healthcare system was restructured to form hospital regions, one of which consisted of a large hospital (Laho) and a regional hospital (Reho). The Laho is located in the state capital with ca. 70.000 citizens and Reho is situated 20 KM apart in a rural city of 9.000 inhabitants. The hospital region is one of the ten largest providers in Switzerland. In 2005, 445 medical doctors, 964 nurses and 884 support staff members treated ca. 60.000 patients in residence, and 70.000 in day care using 753 beds. Budget-wise, Laho is about nine times as large as Reho. It is considered a pure acute hospital, whereas Laho also offers special services in treatment, research and education, similar to a university hospital. The services are defined by the government, for which the hospital region receives a budget in return. 
The hospital region is led by the Kanton's health department, which presides over the board of directors consisting of external experts. The executive board of the hospital region consists of the rotating heads of clinical departments and the director of nursing as well as administrative departments and is lead by a CEO. The CEO of Reho is a member of this board.
Figure 2: Structure of the hospitals and the hospital region 
Part of the official task of the board of directors was to strengthen the management perspective within the hospital region and to conduct a strategy development process. In fall 2003, the head of this board of directors approached the second author asking for support in his endeavor to strengthen the management perspective in the newly founded hospital region. 
In March, 2004, the authors were invited by the executive board to participate in their regular meeting, after prior conversations expressing the mutual interest for collaboration. We as researchers and the hospital were interested in establishing a research partnership. The practitioners were looking for an outside view on various change initiatives. We presented ourselves and were understood as novices to the healthcare domain, but as experts in management and organization studies. Based on a list of such initiatives that were considered important and at stake by the organization, we decided to accompany the merger of the nursing departments of Reho and Laho. The nursing departments are important to implement the hospital region, because nursing operates in a matrix with all clinics and most support functions of a hospital. Nursing therefore provides an important condition for the hospital region to function in day-to-day practice. 
In June, 2004, a first meeting took place between the researchers (the first author and his partner, see acknowledgments) as well as the key actors in the nursing merger. These actors were: first, the head of Laho's nursing department who is also the head of nursing of the hospital region and a member of the executive board. Second, the head of Reho's nursing department; and, third, a change agent who was appointed by Laho's nursing director to conduct the merger in Reho and to adapt it to Laho's organization, leadership and professional standards. Our objective in this meeting was to clarify expectations regarding the research process on both sides, including continuous field access and feedback sessions as part of the partnership. One of the two researchers asked the question: "Are you the project team for the merger?" "Ah, I guess so," says Laho's head of the nursing department extracting her agenda. "Then we should meet regularly, shouldn't we?" The three of them decided on dates to meet at a six-week interval. 
From July, 2004, onwards, we started our qualitative data gathering interviewing nurses at Reho and at Laho, reading written material and observing meetings as well as work routines on the wards. Early in July, we were invited to an important meeting by the head of the region's nursing department to be held at Reho. At that meeting, the change agent was officially presented to the leading ward nurses at Reho for the second time. Also the merger project was presented in terms of the intended results. 
During fall, 2004, and after several feedback workshops with the project team, we as researchers began to realize that the boundary between feeding back observations and being drawn into actively supporting the merger project was becoming blurry. Increasingly, we assumed the responsibility for facilitating the process between the three participants; this provided the opportunity to frame their perceptions of the merger process. 
We sensed that our gate keeper—the head of nursing of the hospital region—might intend to use our insights for a crucial decision. In November, 2004, we were asked to provide slides of our insights and hypotheses of the management challenges at Reho's nursing department for her. She intended to use it at a workshop of Reho's executive board. At the same time, the decision emerged to replace Reho's head of nursing with the change agent. This decision aroused a wide array of interconnected topics about personal careers among the members of the project team, of the relationship between the two hospitals, and of the relation between nursing and the clinics, among others. In order to avoid a source of scientific legitimization of such a decision within a context of those topics we designed our slides as questions for further development. 
Within the context of this decision, the feedback sessions became more difficult to conduct. Open reflection within and between the project team and ourselves was inhibited. Therefore, we entered our observations rather as questions, and by subtly engaging within the discussions. Meanwhile, cooperation between Reho's head of the nursing department and the change agent became increasingly difficult, an observation about which both agreed. 
While preparing the feedback workshop to be scheduled on December 1st 2004, the disappointment on behalf of the region's nursing head became obvious. Her critique included the limited definite results from the research process relative to the effort she had invested in the partnership. After several telephone conversations and email correspondence, we arranged the feedback workshop in two parts: During the first part, the two authors met with her to reflect on the wider merger context. This included the precarious relationship between the two hospitals, the instability of the newly founded executive board, and the resulting challenges to Reho's management in general, and that of the nursing department in particular. The second part of the workshop was conducted with the project team. Here, we presented an explanation of the leadership and management challenges within Reho's nursing department that focused not on individuals, but on the communicative interplay between actors. Based on this insight, the project team envisioned possibilities to interrupt this dynamic in order to develop the organization. During this process, the participants realized the amount of projects and initiatives that Reho's nursing department was supposed to handle during 2005. The group decided that the researchers would support them in the development of a project map. This explicit support took place during spring, 2005, in cooperation with the project team and in the vicinity of the nursing department at Laho. 
We were also allowed to continue follow the change initiative at Reho and the field phase ended in December, 2006. We decided to end the field phase as we increasingly sensed that Reho's organization had become stabilized and had adapted to the standards of the hospital region. The first author analyzed all data until February, 2007 when the final version of the case study was discussed with the project team. Each member found it a valid representation of their experience. A member of the project team summarizes: "I read all 120 pages in one weekend. It was like reading a good novel." 
3.2 Challenges and dynamics shaping the research system
In the following subsections, we reflect on challenges that shaped the research system in our view. They emerge from the communicative interplay between us as researchers and our research partners. These challenges can be understood as critical moments for the mutual expectations between the involved actors, i.e. the emergence of structures of the research system. They culminate to view research as a paradoxical process of change. 
3.2.1 Initializing research: The dilemma of gaining access, and self-positioning as organization experts
The mutual interest in cooperating had been established when the hospital region was asked by its public owner to enhance the managerial perspective within the organization. Within the executive board, we perceived a strong interest for us to accompany different change initiatives. The selection criteria were our research interest, and that of the practitioners on initiatives facing profound challenges. We expected a high level of interest from the practitioners. Accordingly, the executive board expected some kind of outcome from our research for conducting these change initiatives. We acknowledged this expectation by assuring continuous feedback of our findings. This so-called learning partnership aimed at providing a reflexive space for our research partners to elaborate on their unfolding initiative. The learning partnership aimed at avoiding expert driven consultancy, while addressing the expectations of the practitioners. We offered an alternative account on the change process, as a source for different insights into what the practitioners were doing. 
This selection contained the dilemma of gaining access: Hot projects assure profound practitioners' and researchers' interest. At the same time, expectations for a more consultative support can emerge more easily. 
Our interpretation draws on the insight that change processes of and within pluralistic organizations confront practitioners with times of high uncertainty and ambiguity (LANGLEY & DENIS, 2006). Expectations to actively support the initiative become more likely and appear legitimate. Setting up the learning partnership was our way to address these expectations. As the unfolding story shows, the research partnership with its focus on conducting reflexive communicative settings led to a profound disappointment. We will return to this disappointment at the end of this section. 
During the initial period of the collaboration, we explicitly presented ourselves as organization and management scholars from a university with a high reputation in this field. The advantage of this positioning was that it did not involve the core of medical and nursing activities per se, but how they are organized and interweave. With respect to the assumed practitioners' uncertainty, the positioning secured both partners as experts in their respective field, following our insights of hospitals as expert organizations. Nevertheless, this positioning triggered respective images as business consultants representing an economic logic which was considered as a potential threat to the medical and nursing rationalities. Apart from expressing our explorative approach during personal encounters throughout observations or interviews, it was helpful during the field phase, that we were not financed by the hospital region. 
3.2.2 Conducting the field research: The intervening effects of observing
A core theme for us researchers regarding our relation was to handle observing and intervening. In the first meeting, we posed the naive question: "Are you the project team?" As a reaction the present actors considered themselves as such. This incident illustrates that observing can bear a considerable intervening effect. 
We interpreted this incident of unintentionally participating in constituting the project team in two ways: Regarding the organization, it demonstrated that the initial set-up of the change initiative appeared rather emergent or intuitive, and less formally designed. One hypothesis for further exploration was that more attention is paid to the factual than to the social dimension of organizing. Another was that actors pragmatically exploit emerging opportunities, rather than engage in detailed planning. The other interpretation referred to the research system: The fact that the reaction occurred swiftly could have been associated to our partners' image of us as management researchers. Immediately forming the project team could have been to demonstrate to us their handling such topics. 
In terms of social systems theory, the intervening effect of observing can be explained by the operational closure of a system. A communication is considered by the observing counterpart who decides on how to react to the communication according to his own structures. The contributor of that communication does not decide on it, despite developing expectations on probable reactions. In more practical terms, observing can become intervening, because research partners observe the researchers. The incident at the beginning of this section illustrates this core assumption of our heuristic that has also been noted by others (e.g. BARLEY, 1990, p.241). Researchers are observed during their field visits, when interviewing, informally conversing, observing or taking notes. That may raise anxiety among practitioners, which we observed particularly in the beginning of our work, so that we let them see our diaries on request or left them sometimes lying open on the table when leaving the room. We also restrained from note taking during seemingly critical incidents. With regards to recording, we explained how the recorder functioned and explicitly invited interviewees to shut them off whenever they liked. Doing so was intended to show transparency, and respect for our partners' perspectives. 
In the line of the argument for the intervening effects of observing, we do not follow the implication to minimize the researchers involvement in order to access the natural state of affairs as much as possible (see MORGAN, 1983). Rather, our approach was to become aware of such effects, develop more respective expectations and use them as a source to generate insights into how organizing change was conducted within the nursing department. In detail, we learned to ask more open questions, for instance "How do you understand yourselves as a group?" could have been more appropriate. Later on, the advantage of lengthy stays is that organizational practice tends to resume normality, as practitioners continue their daily work (VAN MAANEN, 1982). In regular meetings of the Reho's head nurses we almost became normal members in that it was rather commented on if we did not attend. Similarly, our stays on wards did not change their daily routines, particularly during critical incidents. We just stepped out of the way or assisted with running errands, or the like (see also BARLEY, 1990). Vice versa, lengthy stays replaced our initial ignorance with an awareness of the organizing practice and the practitioners' understandings of that practice. We became more conscious of our own distinctions in relation to those of the practitioners. Thus, our awareness increased by accidentally touching issues that would have otherwise left aside by the practitioners. 
In sum, and in terms of the reflexive systemic research (RSR) heuristic, observing can be assumed to bear intervening potential, because researchers are observed when observing practitioners. But, whether an observation bears an observable effect as an intervention depends on the system, rather than on the researchers. In that respect, reactions to observations provide interesting insights and a source of validation or data. 
3.2.3 Participants in the research system: The complexity of "the" research partner
The RSR-heuristic entails an over-simplification of the research partner. We encountered different actors relevant to our research with their different interests, ambitions and positions. Apart from the executive board that decided the overall research cooperation we built relationships with the project team and members of the clinics, the nursing and the administration involved in the merger. 
In particular the project team provided a challenging heterogeneity for the research process: The head of the nursing department of the hospital region and member of the executive board was interested in a successful merger which meant that Reho adapted to the nursing standards, organizational structures, and leadership practices of Laho. With regards to ourselves, we served as a sounding board, if not a source for legitimizing precarious decisions on conducting the change process. 
The head of the nursing department of Reho, who was later replaced by the change agent of Laho, was interested in preserving Reho's identity as he called it and to use us to convey their perspective to Laho. 
The change agent, who was employed by the head of the region's nursing department, was also interested in a successful change process. With regards to us, she expected companions and empathic listeners in her uncertain endeavor. 
In view of the heterogeneity of "the" research partner, we suggest to differentiate their respective roles and expectations systematically. Inspired by Edgar SCHEIN's (1997) concept of client, the following distinctions became helpful for us:
Primary Research partners: all those who decide on the access and its termination to the research site. In our case, these were the executive board in general and the head of the nursing department of the region in particular.
Core gate keeper of the actual site: The change agent was our core gate keeper and also a core informant, with whom we kept a close relationship and even friendship.
Core informants: all those providing data, of which particularly those in lower hierarchy are in need of protection. That is why we fed back findings up the hierarchy with explicit authorization for the next level.
Peripheral informants: These informants provide relevant context information.
Those potentially affected by research outcomes: The head of the Reho's nursing department was not only a core informant, but more so one who was potentially affected by our findings. Even though we regard the respective decision to replace him as inevitable, we were conscious of his precarious position within the initiative. 
For us, it became important to acknowledge the respective roles and expectations during the field phase. Our heuristic proved helpful in this regard, particularly by distinguishing the research system from the organization in which it was embedded. In detail, this meant the need to be aware of a potential role conflict of our research partners: They were representatives of their organization on the one hand, and participants of the research on the other hand. As our relations became more personal and moved towards friendship, this could provide a dilemma, also highlighted by Jane DUTTON and Janet DUKERICH (2006, p.23). The respective practitioners became somewhat loyal to us and to their employer. 
In light of the complexity of our research partner, we did not aim to be neutral, but rather to adopt what we call an allparteiliche position. Whereas neutrality has been considered a social role in itself (see also VAN MAANEN, 1982, p.115), this German phrase is close to taking into account multiple voices as indicated in appreciative inquiry (VAN DER HAAR & HOSKING, 2004). It signifies that we took all accounts at face value and acknowledged the different perspectives in our analysis. Pragmatically, we assumed that individual actors were well intended and that any decision, activity, communication made sense. Part of our analysis was to make this sense explicit within the specific context and to focus on the interplay of these voices to explain the course of events. An exploratory research approach was helpful in this respect. Nevertheless, researchers tend to be ascribed towards one group or actor rather than the other (BARLEY, 1990). In our case, we faced this challenge particularly during the initial stages because we entered Reho through the mandate of the region's head of the nursing department. Later on, the challenge shifted in that we did not comply with her expectations to legitimize her decisions. Hence, claiming and sustaining an allparteiliche position provides a constant balancing between the interests of different organizational members. In doing so, it allows the researcher the opportunity to keep the distance necessary to research. Furthermore, it may also help to avoid the betrayal of subjects, discussed in some ethnographic work (BARLEY, 1990; VAN MAANEN, 1982), whereas others highlight the collaborative relation (BALOGUN et al., 2003). 
3.2.4 Dynamic evolving of expectations over time
During the beginning of our involvement, our primary research partner—the region's head of nursing—invited us to attend an "important" meeting at Reho. At this meeting the change agent was presented to Reho for a second time, because she had not been well received during her first week of work at Reho. In a similar way, our primary research partner asked us to provide her with slides for a presentation she held a Reho's leadership workshop. The workshop coincided with the looming decision of replacing Reho's head of nursing with the change agent. While we attended the "important" meeting at Reho, we provided her with very general slides focusing on our activities and suggesting open questions for Reho's leadership about potential topics to address for a successful merger. In both cases, we felt to be used for our partners' purposes. Whereas we added to the weight of Laho's crew at the initial workshop, we were inclined to think that our insights on Reho's challenges in organizing nursing could be used as further arguments to replace its head of nursing with the change agent. Particularly, in this second incident, our challenge was to be helpful without adding a seemingly objective, scientific expert opinion on the problems within the nursing department to be used for his replacement. In our view, and as we presented later, these problems could not be solely ascribed to the person holding the position as head of nursing. 
In more general terms, expectations on behalf of researchers and practitioners evolve over time. In this sense, "field work is always emerging" (VAN MAANEN, 1982, p.138). This is particularly the case during fundamental change initiatives in which neither of the participants knows how it will turn out. In situ, practitioners face uncertainty and ambiguity under which they are called to decide. Therefore, using the presence of seemingly neutral expert scientists for one's interests appears to us to be a systematic feature in such a setting. Particularly, in the second incident we are sure to have disappointed our primary research partner, thus our own reaction contributed to the crisis of the research system. This is reflected upon in the following section. 
3.2.5 Double exposure during member validation
Member validation is a common way in interpretive studies to validate findings, but a challenging one, as the reflexive account of Rick IEDEMA and colleagues (2004) demonstrates. In our case, the project team provided a nucleus of the two organizations and their merger, represented by its different team members. To this end, the project team enabled an alternative view on our insights gathered within clinics and the nursing departments. At the same time, however, feeding back findings meant for us to minimize any account that could be exploited by the members of the project team for their respective purposes, thus drawing us into the dynamic of their change process. 
Due to this worry and due to the heterogeneity of the project team our feedback was explicitly subtle. We usually translated our emerging insights into questions or developed communicative designs through which we could jointly work on the interpretations with the project team. We aimed to participate rather as a discussion partner and less with the overall authority of the scientists. 
This subtlety contributed to the disappointment of expectations that culminated in the crisis of the research system around December 1st 2004. Our primary research partner, the head of the region's nursing department, phrased her disappointment to argue an end to the research process. She perceived a mismatch between the practitioners' contribution to our research and what we provided in return. In this respect, the subtlety of our feedback contributed to her disappointment and the feedback session expressed the central turning point of the research process. 
This factual subtlety involved a time dimension that appeared crucial to us. Although preliminary insights about the current dynamics within the nursing department emerged early in the research process, we decided not to feed them back directly to the practitioners. We worried about the potential criticism of a lack of in depth field understanding, and that of exploiting our interpretations for the looming decision on personnel. 
Reflecting on the crisis of the research system resulted in the following interpretations: First, the disappointment could have resulted from our avoiding attempts to exploit findings. Particularly the requested presentation of our analysis by the region's head of nursing may have proved unhelpful to pursue her interests. Second, the crisis of the research system could well reflect the looming decision to replace Reho's head of nursing with the change agent. It was associated with other unresolved and entangled topics regarding the relationship of the two hospitals, the relation between nursing and the clinics, personal careers and so forth. The dynamics involved with this decision may well have transferred to the research system, and became attributed towards us. Third, and more profoundly, the disappointment could be related to the organizing in hospitals which we understood as being focused mainly on interactions with patients and on material outcomes. Our careful and subtle provision of feedback by means of questions and of facilitating conversations may well have remained unnoticed. In our view, providing, designing and conducting a communicative setting for the involved practitioners to reflect on how the merger of the nursing department unfolded, was what we could contribute as researchers in terms of a fair give and take. Our research partners, and particularly our primary research partner, may have been unaware of this process dimension. Therefore, they did not consider our contribution on facilitating the process which could explain their disappointment. In short, we were faced with our own findings. 
In other words, the research system was faced by the crisis in part because of what it allowed us to observe. We aimed to handle this challenge by focusing on the contents of our insights at that moment of time. These included the fragility of the newly founded executive board, which in turn helped to explain plausibly why certain issues regarding the overall relationship between both hospitals remained unresolved so far. These topics were discussed with our primary research partner alone as they involved the management of the hospital region. With the entire project team, we reflected on the current challenges of organizing and leading Reho's nursing department. Our explanation contained a communicative dynamic of different practices found at different levels of the nursing hierarchy. Within the context of the hospital merger and a lack of resources, their interplay contributed to the evaporation of leadership and formal structures, upheld by increasing improvisation. Our explanation therefore avoided a personal attribution, while suggesting that working on the communicative practices within the nursing department may be promising. At the end of the meeting, we agreed to continue our research, while explicitly supporting the project team in a separate initiative of mapping all other projects and their interdependencies within the nursing department. 
An important and more general insight is that member validation involves a mutual exposure why it is precarious to the relationship of researchers and practitioners: Practitioners are exposed by the observations, anecdotes and quotes used. Within the internal public of the organization, they may breach the confidentiality usually promised by researchers. Rick IEDEMA and colleagues (IEDEMA et al., 2004, p.19) reflected in their research that such exposure also extends to the analytic systematization which tends to objectify research subjects raising defensiveness towards the results. Furthermore, particularly interpretive approaches investigating the micro processes of organizing bring to the foreground what tends not to be observed by practitioners. While allowing for reflection and development, explicating what is taken for granted poses risks for practitioners to change, or for attributing responsibility of what may be regarded as critical. Not only are the practitioners exposed particularly in feedback sessions, but also the researchers. Data collection activities by observing, interviewing, reading and analyzing are a more private or secure affair. But presenting one's insights to the research partners puts them to the test of their perspective. Depending on how this test turns out, research can be continued or be canceled, results can be authorized for further usage or not, or changes demanded. Apart from this potential threat, member validations bear the opportunity of joint reflection and dialogue to further advance interpretations for practitioners and researchers alike (IEDEMA et al., 2004; VAN DER HAAR & HOSKING, 2004). Overall, member validations are precarious for both sides. 
Apart from mutual exposure, member validation inhibits the systematic challenge of connectivity. Researchers hold a different perspective than practitioners, which may be a core value for the latter and means avoiding "going native" for the former. In terms of social systems theory, researchers are ascribed the second-order observation on the first-order observation of the system investigated. Member validation means connecting the insights of second- to first-order observation (which could be seen as a re-entry operation). This is precarious on the social and time dimension as pointed out above. But also on the factual dimension of the insights themselves it is a challenge (VON WEIZSÄCKER, 1986): At one extreme, results can be totally new and are therefore not understood. At the other extreme, they can be known without providing anything new. In both cases, the contents are of no informational value for practitioners. They do not make a difference that makes a difference (BATESON, 1983). 
4. Discussion: Mutual Protecting of Research and Practice in Relation
The above challenges of gaining access, initializing the research partnership, the intervening effects of observing, the complexity of the research partner, the dynamic evolving of relations and the double exposure in member validation shed light on the complexity and the dynamics of field research. They highlight the fragility of conducting longitudinal research in organizational change. In order to explore, how and why the merger of the nursing department unfolded involved handling distinct challenges during our stay in the field. 
An important lesson for us was to acknowledge that both the change initiative and the research system to explore it require protecting. We call this insight mutual protection. Of the above challenges, the intervening effect of organization and member validation express this insight. Research systems need to be protected to sustain field access while not compromising the conditions that allows the relating of empirical findings back to the academic community. Simultaneously, change initiatives within their organizational context require protection from researchers. 
Protecting the research system often crystallizes at the researchers, when handling evolving expectations, attempts of exploitation, or during member validation. Vice versa, the organization and the change initiative require protecting from such a reflexive platform, because observations entail a potential to serve as interventions. In our case that involved a wide array of aspects. They ranged from avoiding taking notes, of excluding potentially precarious incidents from analysis, being aware of protecting individuals from exposing themselves when feeding back findings, to refraining from involving ourselves in such a way that undermines the practitioners' responsibility. 
We found the RSR-heuristic helpful to handle this mutual protection. The heuristic allows for an analytical distinction of the practically interwoven structures or practices to initiate, to sustain and to terminate the research system from those activities that concern data gathering and interpreting. The research system can then function as a temporary parallel structure within the organization. As an interaction system, it provides a reflexive platform for practitioners to observe their initiative unfolding. Simultaneously, their interpretations and making sense of the process provides alternative and enriching insights for researchers. 
Organizing mutual protection involves communicative structures on two levels. On a micro-level, such structuring refers to the single research encounter, as an interview, as observing meetings or daily organizing, or as a feedback workshop. The macro-level of structuring communication is focused on connecting the single communicative encounters. It encompasses the time dimension of conducting research, the different social contexts in which it takes place and the topics that emerge. Structuring communication follows BARRETT, THOMAS and HOCEVAR (1995) who argue for a recursive relationship between meaning, language, and action. This notion applies to organizational as well as research practice. 
We offer this model as a contribution particularly to longitudinal research that observes organizations directly (MINTZBERG, 1979). As researchers and their partners engage over a certain period of time, they relate and enact the research system. In line with others, the RSR-heuristic places the relationship between research partners and researchers at the center. It provides the foundation of the possibility and of the quality of conducting research in the field (DUTTON & DUKERICH, 2006). 
The RSR-heuristic focuses on the relationship by conceptually understanding it as a temporary interaction system in which roles, topics, and other structures evolve, recursively interwoven with the process of unfolding events over time. It focuses on how representatives of academia and the organization interact to generate insights, without disregarding their respective contexts. These contexts help to explain the dynamic, complex and fragile nature of conducting research in the field. While practitioners are more oriented to the notion that their organization continues operating, academics are concerned with exploring this organizing in order to generate findings to be related to their academic context. That is why, differences, misunderstandings, contradictions or even paradoxes can be considered the norm, rather than the exception. 
In order to successfully handle the challenges as part of the research system, we follow the call to include researcher partners more actively in research (BALOGUN et al., 2003). A sustainable relationship not only allows continued access as we showed in the illustration above. Practitioners can also assist the researcher in interpreting their own interpretations. By participating in the researchers' reflection, research partners can contribute to the quality and richness when generating insights. In other words, practitioners can not only be conceived of as a data source, but as a partner in interpreting such data within a symmetrical relation. Because such encounters involve the mutual exposure illustrated above, both partners rest on a sustainable relationship. In our case, this relational quality developed throughout the initial months of the research, without which the crisis we faced with the research partner could not have been handled. 
The importance and fragility of the relationship has been acknowledged in the literature on reflexivity in organization studies. We see the particular value of the RSR-heuristic in that it allows for an analytical distinction between the different domains of practices, process patterns or structures that evolve throughout field research:
the organization's that are to be investigated;
those to establish and develop the research process, like clarifying expectations regarding content, process, respective roles, outcomes as well as (on-going) feedback (BARLEY, 1990; WIMMER, 1992);
the communicative practices and structures that relate to the research process as it occurs, involving data gathering, interpreting or member validation (ALVESSON & SKÖLDBERG, 2000); and
the relating of the researcher to academia, including data analysis presentation or writing (ALVESSON et al., 2008; GOLDEN-BIDDLE & LOCKE, 1993; LANGLEY, 1999). 
In their practice, researchers consider these domains, sometimes more, sometimes less explicitly. The above heuristic is therefore intended as a guide for organizing our awareness throughout our research practice in reference to these domains and their interrelations. 
For empirical research on organizations within social systems theory, the model can serve to motivate a methodological discussion. Despite a variety of approaches of theoreticians (LUHMANN, 2000), reflecting consultants (WIMMER, 2004) and dual approaches of consultancy and research (BUCHER & RÜEGG-STÜRM, 2008), other organizational scholars, including ourselves, (JUNG, 2007; MINGERS, 1995) develop their respective methodological approaches. As Thomas PFEFFER (2001) argued, social systems theory lacks an overall methodological core with regards to longitudinal research in organizations. We believe that the proposed model can serve as a starting point for its development which could aid to further advance organizational research within social systems theory. The RSR-heuristic exemplifies a re-entry of the researcher within research. It provides orientation for such a re-entry by placing the researcher in a temporary coupling relation with the organization. Doing so, includes but goes beyond reflecting on one's epistemological and explanatory assumptions and views. By means of the research system and the analytically distinguished domains, the model can support a systematized discussion on conducting empirical research in organizations. 
5. Applicability, Limitations and Conclusion
The RSR-heuristic fosters discussions on integral challenges in longitudinal research that take a process perspective (LANGLEY, 1999). These challenges regard for instance: the role of the researcher as passive observers or active participants (BARLEY, 1990; VAN MAANEN, 1982); the tension of non-understanding and triviality when feeding back findings during member validation (IEDEMA et al., 2004); the tension between observing and intervening in the organization under study through research; and the fragility of the research process on the one hand and that of organizational daily practice on the other hand. This latter aspect points towards the need of mutually protecting both, while they potentially produce irritations for one another. Handling such challenges involves a reflexive approach to conducting field research. For this matter, we suggest the RSR-heuristic as a conceptual guide to systematically explore what is going on during research. 
Although the RSR-heuristic is generic in the sense that it aims at offering orientation rather than specific pieces of advice, it faces three limitations: First, it emerged as part of our longitudinal research on organizational change, despite useful insights drawn from the literature. Therefore, it mainly applies to longitudinal research projects that directly observe organizations in situ. Further applicability in other areas of research lies outside the scope of this paper. Second, and in terms of Thorngate's (WEICK, 1999) magical triangle of simplicity, accuracy and generalizability, the heuristic falls short in accuracy. We aimed to generalize the research system from our research with support of respective literature. Therefore, the RSR-heuristic requires researchers to re-contextualize it to their own contexts and that of their partners. Third, we considered quality criteria for reflexive research rather implicitly within this paper. Instead of applying the conventional criteria—rigor, validity and replicability—the heuristic aims for a symmetrical appreciation of practitioners and researchers throughout the field phase by being reflexive on the research process. Therefore quality criteria should be more process oriented in explicating how the research system was initiated and how it evolved over time in relation to the generated insights. Due to the context specificity and observer dependence of insights, generalizability is limited. Here, we associated rigor with making explicit one's own underlying assumptions, demonstrating the reflexive path of analyzing by interpreting the interpretations. Furthermore, we suggest taking the dynamic unfolding of the relationship between researchers and research partners into account. Doing so encompasses member validation and considers the research process over time. For the future, we suggest the need to elaborate on the quality criteria in process oriented studies. 
There has been a long history in organization studies about the tension between scientific rigor and practical relevance of our work. Andrew PETTIGREW and colleagues (2001, p.697) claimed that "theories of change in the fields of management and organization must face the double hurdle of scholarly quality and practical relevance." Recently, the topic has been taken up again. Alfred KIESER and Lars LEINER (2009) argue that the "rigor-relevance-gap" cannot be bridged. Taking a system theory perspective, science and organizational practice function differently and are operationally closed. Even if insights, ideas, suggestions of one system cannot directly be transferred into the other, both can irritate and inspire each other. The research system suggested in this paper provides a view on the coupling mechanism to do so. It acknowledges that any communicative act that is observed by researchers or that is offered by them to practitioners is an irritation at the most. Whether at all and which meaning is generated from such irritations depends on the respective system. Hence, bridging the two worlds may be impossible. But mutual observing and structural coupling are possible providing the possibility for meaning generation. Empirical research does take place and it sometimes irritates the practice it investigates. Thus, even conducting social system theoretical research can be practically relevant, and not only of scholarly quality. But in doing so, we need to be modest: Our research partners in particular and practitioners in general decide on the relevance of our work with regards to their specific context. 
The RSR-heuristic can be seen as a skeleton to assist researchers during their longitudinal field visits. The heuristic in itself certainly lacks the inspiring details and stories from which it emerged. They express the richness and ongoing movement, whereas our heuristic could provide some orientation. It is rather the map, but not the enchanting landscape we explore while collaboratively shaping it with our research partners to learn from their insights. 
Foremost, we thank Mrs. Silke BUCHER for the joint empirical research with the first author under supervision of the second author. Furthermore, we are grateful to the editors of the special issue, particularly Ms. Patricia WOLF, and to two anonymous reviewers for the precise and helpful comments and suggestions in the past months.
Alvesson, Mats & Sköldberg, Kaj (2000). Reflexive methodology. New vistas for qualitative research. London: Sage.
Alvesson, Mats; Hardy, Cynthia & Harley, Bill (2008). Reflecting on reflexivity: Reflexive textual practices in organization and management theory. Journal of Management Studies, 45(3), 480-501.
Baecker, Dirk (2005). Form und Formen der Kommunikation. Frankfurt/M.: Suhrkamp.
Balogun, Julia; Huff, Anne Sigismund & Johnson, Phyl (2003). Three responses to the methodological challenges of studying strategizing. Journal of Management Studies, 40, 197-224.
Barley, Stephen (1990). Images of imaging: Notes on doing longitudinal field work. Organization Science, 1(3), 220-247.
Barrett, Frank; Thomas, Gail Fann & Hocevar, Susan (1995). The central role of discourse in large-scale change: A social constructionist perspective. Journal of Applied Behavioural Science, 31(3), 352-372.
Bateson, Gregory (1983). Die Ökologie des Geistes. Frankfurt/M.: Suhrkamp.
Brannick, Teresa & Coghlan, David (2006). Reflexivity in management and business research: What do we mean? Irish Journal of Management, 27(2), 143-160.
Bucher, Silke & Rüegg-Stürm, Johannes (2008). Strategizing as protective interrupting: Preliminary insights from two case studies on strategic change within hospitals. Conference Paper, EGOS, Amsterdam, 1-40.
Dutton, Jane E. & Dukerich, Janet M. (2006). The relational foundation of research: An underappreciated dimension of interesting research. Academy of Management Journal, 49(1), 21-36.
Golden-Biddle, Karen & Locke, Karen (1993). Appealing work: An investigation of how ethnographic texts convince. Organization Science, 4(4), 595-615.
Hardy, Cynthia & Clegg, Stewart (1997). Relativity without relativism: Reflexivity in post-paradigm organization studies. British Journal of Management, 8(2), S5-S17.
Hendry, John, & Seidl, David (2003). The structure and significance of strategic episodes: Social systems theory and the routine practices of strategic change. Journal of Management Studies, 40(1), 175-196.
Holland, Ray (1999). Reflexivity. Human Relations, 52(4), 463-484.
Iedema, Rick; Degeling, Pieter; White, Les & Braithwaite, Jeffrey (2004). Analysing discourse practices in organisations. Qualitative Research Journal, 4(1), 5-25.
Jung, Stefan (2007). Die Form der Reform. Witten: Universität Witten/Herdecke.
Kieser, Alfred & Leiner, Lars (2009). Why the rigour-relevance gap in management research is unbridgeable. Journal of Management Studies, 46(3), 516-533.
Langley, Ann (1999). Strategies for theorizing from process data. Academy of Management Review, 24(4), 691-710.
Langley, Ann (2009). Studying processes in and around organizations. In David A. Buchanan & A. Bryman (Eds.), The Sage handbook of organizational research methods (pp. 409-429). London: Sage.
Langley, Ann & Denis, Jean-Louis (2006). Neglected dimensions of organizational change: Towards a situated view. In Rune Lines, Inger G. Stensaker & Ann Langley (Eds.), New perspectives on organizational change and learning (pp.136-159). Bergen: Fagbokforlaget Vimostad & Bjorke.
Lewis, Marianne W. & Kelemen, Mihaela L. (2002). Multiparadigm inquiry: Exploring organizational pluralism and paradox. Human Relations, 55(2), 251-275.
Luhmann, Niklas (1984). Soziale Systeme: Grundriß einer allgemeinen Theorie. Frankfurt/M.: Suhrkamp.
Luhmann, Niklas (1986). Systeme verstehen Systeme. In Niklas Luhmann & Karl Eberhard Schorr (Eds.), Zwischen Intransparenz und Verstehen. Fragen an die Pädagogik (pp.72-117). Frankfurt/M.: Suhrkamp.
Luhmann, Niklas (1990). Die Wissenschaft der Gesellschaft. Frankfurt/M.: Suhrkamp.
Luhmann, Niklas (2000). Organisation und Entscheidung. Wiesbaden: Opladen.
Luhmann, Niklas & Fuchs, Peter (1989). Reden und Schweigen. Frankfurt/M.: Suhrkamp.
Mingers, Susanne (1995). Systemische Organisationsberatung: Eine Konfrontation von Theorie und Praxis. Frankfurt/M.: Campus.
Mintzberg, Henry (1979). An emerging strategy of "direct" research. Administrative Science Quarterly, 24, 582-589.
Morgan, Gareth (1983). Toward a more reflective social science. In Gareth Morgan (Ed.), Beyond method. Strategies for social research (pp. 368-376). London: Sage.
Pettigrew, Andrew (1997). What is processual analysis? Scandinavian Journal of Management, 13(4), 337-348.
Pettigrew, Andrew; Woodman, Richard W. & Cameron, Kim S. (2001). Studying organizational change and development: Challenges for future research. Academy of Management Journal, 44(4), 697-713.
Pfeffer, Thomas (2001). Das "zirkulaere Fragen" als Forschungsmethode zur Luhmannschen Systemtheorie. Heidelberg: Carl-Auer.
Rüegg-Stürm, Johannes (2002). Dynamisierung von Führung und Organisation: Eine Einzelfallstudie der Unternehmensentwicklung von Ciba-Geigy 1987-1996. Bern: Haupt.
Sankaran, Shankar; Dick, Bob; Passfield, Ron & Swepson, Pam (Eds.) (2001). Action learning and action research. Concepts, frameworks, processes, applications. Lismore: Southern Cross University Press.
Schein, Edgar H. (1997). The concept of "client" from a process consultation perspective. Journal of Organizational Change Management, 10(3), 202-216.
Simon, Fritz B. (2007). Einführung in die systemische Organisationstheorie. Heidelberg: Carl-Auer.
Tsoukas, Haridimos & Chia, Robert (2002). On organizational becoming: Rethinking organizational change. Organization Science, 13(5), 567-583.
Tuckermann, Harald (2007). Organisationaler Wandel als Entfaltung von Paradoxien – systemtheoretische Rekonstruktion einer Krankenhausfusion. PhD thesis, St. Gallen University.
van der Haar, Dorieke & Hosking, Diane-Marie (2004). Evaluating appreciative inquiry: A relational constructivist perspective. Human Relations, 57(8), 1017-1036.
van Maanen, John (1982). Fieldwork on the beat. In John van Maanen, James Dabbs; Robert Faulkner, (Eds.), Varieties of qualitative research (pp.103-151). Beverly Hills: Sage.
von Weizsäcker, Ernst Ulrich (1986). Erstmaligkeit und Bestätigung als Komponenten der Pragmatischen Information. In Ernst Ulrich von Weizsäcker (Ed.), Offene Systeme I. Beiträge zur Zeitstruktur von Information, Entropie und Evolution (pp.82-113). Stuttgart: Klett-Cotta.
Watzlawick, Paul; Beavin, Janet H. & Jackson, Don D. (2003). Menschliche Kommunikation. Formen Störungen Paradoxien. Bern: Hans Huber.
Weick, Karl E. (1989). Theory construction as disciplined imagination. Academy of Management Review, 14(4), 516-531.
Weick, Karl E. (1999). Theory construction as disciplined reflexivity: Tradeoffs in the 90s. Academy of Management Review, 24(4), 797-806.
Whittington, Richard; Jarzabkowski, Paula; Mayer, Michael; Mounoud, EIEonore; Nahapiet, Janine, & Rouleau, Linda (2003). Taking strategy seriously. Responsibility and reform for an important social practice. Journal of Management Inquiry, 12(4), 396-402.
Willke, Helmut (1994). Systemtheoretische Strategien des Erkennens. In Klaus Götz (Ed.), Theoretische Zumutungen. Vom Nutzen der systemischen Theorie für die Managementpraxis (pp.97-116). Heidelberg: Carl-Auer.
Wimmer, Rudolf (1992). Was kann Beratung leisten? Zum Interventionsrepertoire und Interventionsverständnis der systemischen Organisationsberatung. In Rudolf Wimmer (Ed.), Organisationsberatung: Neue Wege und Konzepte (pp.59-111). Wiesbaden: Gabler.
Wimmer, Rudolf (Ed.) (2004). Organisation und Beratung. Heidelberg: Carl-Auer.
Harald TUCKERMANN is a lecturer and co-head of a management program for healthcare professionals at the University of St. Gallen. He is in charge of the research program "HealthCare Excellence" (HCE) that explores management and organizational change in hospitals. His research focuses on organizational change from a social systems perspective and on strategizing in pluralistic organizations, including the handling of paradoxes. His academic experience includes a visiting research position at Southern Cross University, Lismore, Australia, after finishing his Masters Degree in Economics and Business Administration at Witten/Herdecke University, Witten, Germany. His practical experience ranges from strategy consultancy, process consultancy, to a bank apprenticeship.
Department of Organization Studies, Institute of Public Management and Tourism
University of St. Gallen
Dufourstrasse 40a, 9000 St. Gallen
Tel.: +41 71 224 25 08
Fax: +41 71 224 25 36
Johannes RUEGG-STÜRM holds the chair for Organizational Studies at the Institute of Public Management and Tourism of the University of St. Gallen HSG. His habilitation in 1998 focused on corporate transformation and organizational change, including a stay as Visiting Research Fellow at the Centre for Corporate Strategy and Change (Professor Andrew PETTIGREW), Warwick Business School, University of Warwick, Coventry, UK from 1995 to 1996. Since 1993, he is Lecturer for Business Administration at the Institute of Management (IfB) at the HSG and headed various research projects at the IfB; teaching activities in the context of IfB Management Seminars on Management of Change, Organization Design, Process-based Organization, Financial Controlling and Management Accounting, Management. His experience in industry from 1989 to 1992, ranges from the fields of Management Accounting, Reporting, Operational and Strategic Controlling, with Ciba-Geigy AG, Basel, Switzerland. His award winning Doctoral thesis, "Corporate Development and Ethics" was concluded in 1989, and succeeded his studies in Business Administration, with a major in finance and accounting from 1986 to 1989 at the University of St. Gallen. Current research interests: systemic-constructionist approaches to management, leadership and organization; management of strategic change and renewal; management of healthcare organizations; strategy-as-practice
Department of Organization Studies, Institute of Public Management and Tourism
University of St. Gallen
Dufourstrasse 40a, 9000 St. Gallen
Tel.: +41 71 224 23 23
Fax: +41 71 224 25 36
Tuckermann, Harald & Rüegg-Stürm, Johannes (2010). Researching Practice and Practicing Research Reflexively. Conceptualizing the Relationship Between Research Partners and Researchers in Longitudinal Studies [100 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 11(3), Art. 14, http://nbn-resolving.de/urn:nbn:de:0114-fqs1003147.