Abstract
Following recent calls to extend our understanding of organizational unlearning, we gain empirical insights into how the process unfolds in practice. Based on the findings of a study with change consultants in Europe, we conceptualize two forms of organizational unlearning. First, open-ended unlearning refers to instances where organizational knowledge is intentionally discarded, but the outcomes of the change process are not known. Second, goal-directed unlearning refers to situations where organizations implement specified knowledge structures that are incompatible with established ones. We also find that both forms of organizational unlearning necessitate preparatory work and interventions that increase their likelihood to succeed. Outlining the implications of the organizational unlearning concept in organizational practice contributes to a better understanding of existing research. We further discuss implications for theory and outline avenues for future research.
Introduction
Organizational unlearning is attracting increasing interest in management research. It refers to the idea that organizations and individuals intentionally discard established knowledge to integrate new knowledge, if any (Tsang and Zahra, 2008). Organizational unlearning has been explored in various contexts such as mergers and acquisitions (Tsang, 2016), innovation (Rezazade Mehrizi and Lashkarbolouki, 2016; Yang et al., 2014), environmental protection (Nygren et al., 2017), scholarship (Antonacopoulou, 2009), social service (Brook et al., 2016), and health care (Gupta et al., 2017).
However, while research on organizational unlearning is growing and developing (Klammer and Gueldenberg, 2019), a recent debate unveils fundamental questions concerning its foundation. Reviewing the historical roots of the term and early works on organizational unlearning (e.g. Hedberg, 1981), Howells and Scholderer (2016) conclude that the concept lacks justification and should be removed from the scholarly discourse. Recent articles consider this critique, with authors reaching different conclusions regarding the scope and usefulness of the concept. Some argue that it should not be overcomplicated (Starbuck, 2017; Tsang, 2017). Others introduce an interdisciplinary view and connect the concept to the field of psychology (Kluge and Gronau, 2018; Nygren et al., 2017). While scholars put considerable effort into clarifying the concept of organizational unlearning, the debate spins around in various academic circles where different positions are formed and reiterated (Becker, 2018a). Klammer and Gueldenberg (2019) reviewed the existing literature, indicating that part of the confusion stems from the lack of a broader empirical understanding of the concept.
In this article, we seek to understand how organizational unlearning can be described and characterized, how it occurs, and how it is used in organizational practice. We invite a practitioner’s voice to the conversation (Bartunek and Rynes, 2014). Following the observation that insights from practice are an important but underused source to develop theory in management and organizational studies (Sutton and Staw, 1995; Van de Ven and Johnson, 2006), we explore the concept of organizational unlearning through the perspective of those who regularly deal with it: change consultants. According to Clegg et al. (2004a), the role of consultants is defined through the disruption of an organization’s order by implementing interventions that question established knowledge and open up a space for finding new potential in the present and future. Since organizational unlearning is broadly understood as a process in which established knowledge is discarded to create new knowledge, if any (Tsang and Zahra, 2008), consultants may provide valuable insights into how organizations intentionally get rid of knowledge. Furthermore, change consultants have broad experience and expertise regarding projects and companies, and can provide examples of how organizations initiate unlearning. The question guiding this study is: How is organizational unlearning reflected in the practices of change consultants during organizational change processes?
We present the findings of a qualitative-inductive interview study conducted with 24 change consultants in Central Europe. We identify numerous aspects and components that change consultants associate with periods of organizational unlearning during change processes. We argue that these findings provide two contributions to the literature. First, we find that the term is associated with different instances in change processes. What we refer to as open-ended unlearning is an explorative process where knowledge is discarded but the outcome of new learning is unknown. Second, goal-directed unlearning targets at well-defined outcomes and is realized when organizations discard knowledge structures in order to implement new but already specified ones. By characterizing these two forms of organizational unlearning, we partially confirm and refine theoretical claims and highlight new aspects not discussed in the existing literature. Furthermore, we suggest implications and factors in organizational unlearning processes that have not yet been addressed and should be examined in future research.
Theoretical background
Research on organizational unlearning
Organizational unlearning is broadly defined as intentionally discarding organizational knowledge in the forms of beliefs or routines to create new knowledge (Klammer and Gueldenberg, 2019). It can be considered a counter-process to learning. Traditionally, most organizational research focused on learning and building knowledge, and little interest was directed to periods when established knowledge blocks the way for new learning (Nguyen, 2017). Following the general argument that knowledge can become obsolete or a hindrance for organizations and eventually needs to be discarded, the concept has been attracting increasing interest since the 1980s (Klammer and Gueldenberg, 2019). In contrast to organizational forgetting (Easterby-Smith and Lyles, 2011), unlearning is an intentional process. Brook et al. (2016) associate it with problem-triggered situations, especially when organizations face crises or environmental turbulence (Turc and Baumard, 2007). Such situations have been observed in different contexts. For example, Rezazade Mehrizi and Lashkarbolouki (2016) examined how organizations unlearn long-held business models because they can no longer be sustained. Martin de Holan and Phillips (2004) linked organizational unlearning with mergers and acquisitions wherein the routines and values of one organization are incompatible with those of the other. Similarly, Tsang (2008) observed organizational unlearning during organizational change processes. Examining Sino-foreign venturing processes, he observed that organizational unlearning was important at different stages to facilitate the acquisition of new knowledge.
Organizational knowledge can take various forms, and researchers have associated organizational unlearning with knowledge structures including beliefs, routines, assumptions, and values (Akgün et al., 2007; Fiol and O’Connor, 2017; Sinkula, 2002; Starbuck, 2017; Tsang and Zahra, 2008; Yildiz and Fey, 2010). Some authors distinguish organizational unlearning as cognitive and behavioral unlearning (Easterby-Smith and Lyles, 2011; Klammer and Gueldenberg, 2019; Yildiz and Fey, 2010). The former involves organizational values, mental models, and underlying assumptions. The latter refers to organizational routines and practices. It is important to note that many authors stress that organizations themselves cannot unlearn; the process is carried out by organizational actors (Brook et al., 2016; Grisold et al., 2017; Kluge et al., 2019). Zhao et al. (2013) further support this idea, claiming that organizational unlearning unfolds at the team level and is realized on the individual level.
Previous research has drawn a nuanced and developed understanding of unlearning on the individual level. Hislop et al. (2014) suggest that organizational actors engage in wiping or deep unlearning. Wiping refers to discarding behavioral patterns in continuous change processes, for example, to integrate new practices at the workplace. Deep unlearning refers to deeply held assumptions and beliefs. It occurs in episodic change processes when individuals come to doubt taken-for-granted views or positions. Studying learning and unlearning processes in social workers, Brook et al. (2016) find evidence for these two forms of individual unlearning. They add that unlearning may be followed by new learning, but can also imply a state characterized by deliberated non-action, that is, attending to a problem while being aware there is no solution (also see Grisold and Kaiser, 2017). Furthermore, Fiol and O’Connor (2017a) draw on findings from psychology, suggesting that while organizational actors gradually learn new knowledge, they may fall back on old knowledge.
Linking organizational unlearning and change
Most research is based on the implicit or explicit assumption that periods of organizational unlearning facilitate change (Brook et al., 2016; Klammer and Gueldenberg, 2019; Kluge and Gronau, 2018). The underlying rationale is that knowledge can eventually become outdated, 1 and established knowledge structures must be discarded in order to make room for the creation of new knowledge, if any (Klammer et al., 2019a). According to Becker (2010: 252), organizational change implies “relinquishing old ways prior to, or at least at the same time as, learning the new practices and procedures.” According to Tsang (2008), organizational unlearning and learning are two types of organizational change processes. Numerous studies examined organizational unlearning in relation to organizational change processes (Becker, 2010; Cegarra-Navarro and Wensley, 2019). Most report that organizational unlearning must occur prior to change. As such, Yang et al. (2014) and Leal-Rodríguez et al. (2015) confirm that unlearning in organizations increases their ability to innovate. Martin de Holan and Phillips (2004) observed that hotels in Cuba needed to discard their routines and practices to align with newly imposed standards and regulations. Similarly, studies found that new technologies afford new means of coordination and interactions and thus, their implementation is accompanied by organization-wide unlearning (Becker, 2010; Cegarra-Navarro and Cepeda Carrion, 2013). More recent assertions provide a better understanding of the interrelation between organizational unlearning and change. Three positions seem particularly relevant.
First, Tsang (2017) highlights that organizational unlearning might not be followed by learning new knowledge. Aligned with his previous work (Tsang and Zahra, 2008), he asserts that organizations can discard knowledge without aiming to integrate new knowledge.
Second, even if organizations aim to intentionally discard outdated knowledge structures, it does not mean that new knowledge is superior to outdated knowledge. While learning new knowledge is generally considered a “good thing” (Brook et al., 2016: 383), the value of knowledge can only be assessed once it is established in the organization. Thus, organizational unlearning does not afford value judgment (Tsang, 2017; Tsang and Zahra, 2008).
Third, organizational unlearning and learning processes do not follow a linear sequential order where one happens after the other. When organizations unlearn outdated knowledge, actors find themselves in a state of tension where they tend to draw on established knowledge structures while they should be enacting new knowledge (Yildiz and Fey, 2010). Hence, organizational unlearning occurs simultaneously with learning. Over time, outdated knowledge structures fade into the background and actors are more likely to use new knowledge (Fiol and O’Connor, 2017; Grisold et al., 2017; Kluge and Gronau, 2018).
Current debates and open questions
While organizational unlearning has received attention in the management and organization literature, the concept is not without criticism. For example, in a recent study, Howells and Scholderer (2016) review the conceptual grounds on which Hedberg (1981) and Nystrom and Starbuck (1984) positioned the idea of organizational unlearning. Assessing contemporary literature in psychology, they find no evidence that knowledge can be literally removed or eliminated. They consider the term misleading and see “good reasons to drop unlearning from the scholarly discourse” (Howells and Scholderer, 2016: 459). Furthermore, several authors affirm there is no coherent understanding of the term (Becker, 2018b; Grisold and Kaiser, 2017; Klammer and Gueldenberg, 2019; Kluge and Gronau, 2018). They emphasize that many definitions contain negative connotations that knowledge can be ‘eliminated,’ ‘erased’ or ‘extinguished’ while in reality, organizational unlearning implies that outdated knowledge may remain in the organization (Kluge and Gronau, 2018; Visser, 2017).
Several authors have attempted to clarify and adjust the concept. For example, from a learning-theoretical perspective, Visser (2017) places unlearning vis-à-vis single and double-loop learning and argues that it is best understood as a phase in which learning is interrupted. Drawing on the routine dynamics literature, Fiol and O’Connor (2017) conceptualize a model in which unlearning in organizations occurs when both the ostensive and performative aspects of an established routine are changed over time. Grisold and Kaiser (2017) and Kluge and Gronau (2018) apply different psychological theories, arguing that organizational unlearning happens when actors refrain from enacting outdated knowledge over time. However, as Cegarra-Navarro and Wensley (2019: 3) stress, the utility of the concept is still “insufficiently clear.” Klammer and Gueldenberg (2019) and Kluge et al. (2019) further highlight the lack of a common understanding, arguing that future research should clarify what organizational unlearning implies and how it can be measured.
While numerous theoretical articles explicitly call for more empirical research on organizational unlearning (e.g. Cegarra-Navarro and Wensley, 2019; Klammer and Gueldenberg, 2019; Kluge and Gronau, 2018; Visser, 2017), and despite some recent exceptions (e.g. Brook et al., 2016; Rezazade Mehrizi and Lashkarbolouki, 2016; Tsang, 2016), the field of organizational unlearning still appears “desk-based and speculative rather than empirical” (Brook et al., 2016: 374). Klammer and Gueldenberg (2019) highlight the need for more empirical research to make the organizational unlearning process more visible and comprehensible. Cegarra-Navarro and Wensley (2019) outline that future research should empirically attempt to understand how unlearning in organizations unfolds across stages and levels. Grisold and Kaiser (2017) encourage future research to look at the elements and stages that are involved in such processes. In other words, scholars are encouraged to gather empirical evidence and develop a practical understanding of the phenomenon to complement or juxtapose various existing theoretical stances.
Exploring organizational unlearning in practice: insights from change consultancy
Drawing on these recent claims and to gather more empirical insights into organizational unlearning, we adopt the idea of engaged scholarship (Barge and Shockley-Zalabak, 2008; Van de Ven and Johnson, 2006). Thereby, we aim to increase scientific knowledge by investigating how a phenomenon or concept is perceived in practice to revisit and develop theory (Sutton and Staw, 1995). Engaged scholarship considers “the advice and perspectives of key stakeholders (researchers, users, clients, sponsors, and practitioners) to understand a complex problem or phenomenon” (Van de Ven, 2018: 37). Practitioners face phenomena in their everyday activities and possess knowledge that lies ‘in the action’. Hence, they can provide rich insights into how phenomena evolve and explain their underlying characteristics (Schoen, 1983; Van de Ven and Johnson, 2006). Practitioners may be able to report on features or dimensions of a phenomenon that seem important while dealing with it, which can help researchers re-examine their assumptions and develop a deeper understanding of that phenomenon (McKelvey, 2006).
This is not to say that practitioners’ insights provide a holistic understanding of a phenomenon. They pursue their own agenda and gain biased perceptions (McKelvey, 2006). In fact, one prevalent critique on engaged scholarship is that practitioners contribute their own ‘flavor’ when describing a phenomenon. While a valid concern, which we will discuss in the implications section of this article, we argue that engaged scholarship contributes to theory because the combination of theoretical and practical knowledge “is more penetrating and insightful than knowledge produced when scholars or practitioners work on a problem alone” (Van de Ven and Johnson, 2006: 817). Van de Ven and Johnson (2006) remind us that any way of seeing is also a way of not seeing (Poggi, 1965), and any form of inquiry leads to partial knowledge. Thus, by turning to practitioners who have encountered organizational unlearning processes, we aim to extend our knowledge of organizational unlearning in practice, how it unfolds in organizations, and how it is related to other factors and phenomena in organizational contexts. This may clarify existing confusion regarding the concept or elucidate new aspects.
Exploring organizational unlearning through the perspective of change consultants
We draw on the experiences and perceptions of practitioners who frequently deal with unlearning periods in organizations, namely change consultants. Generally speaking, the role of change consultants is characterized by accompanying organizations through change periods and imposing initiatives and projects. However, Clegg et al. (2004a: 36) provide a more detailed view. According to them, consultants disrupt dominant orders in organizations and “seek to disturb existing patterns and structures that have become an obstacle to tomorrow’s excellence.” By being “noisy and uncomfortable,” consultants create disorder and chaos to find new perspectives and open up “departures for action” (Hicks et al., 2009: 301). Consultancy involves questioning established knowledge structures in organizations, allowing “new world views, new ways of world-making and to encourage people to disrupt established ways of thinking” (Clegg et al., 2004a: 46). Therefore, perceiving organizational unlearning as a process in which established knowledge is being questioned and discarded, and creating new space for actions (Brook et al., 2016; Nygren et al., 2017) means consultants can be considered initiators, facilitators, enablers, or accompaniments of organizational unlearning processes.
Furthermore, two characteristics define change consultants. First, they are organizational outsiders (Czarniawska and Mazza, 2003) who access many levels of an organization (e.g. top management, regular employees) and units (e.g. product development, accounting), which enables them to obtain a holistic organizational view. This might help us understand how unlearning unfolds across organizational levels (Cegarra-Navarro and Wensley, 2019). Second, top management tends to call on consultants to identify and solve problems (Czarniawska and Mazza, 2003). Over time, consultants may gain experience from many organizations and problem contexts (Hislop, 2008; Kakabadse et al., 2006). They may possess experience and expertise of cases in which outdated knowledge had to be dealt with, and thus, can report on how these processes unfold across organizations.
Method and procedure
To explore how organizational unlearning is reflected in the practices of change consultants during change processes, we conducted a qualitative-inductive interview study. With the goal to understand how/if organizational unlearning is reflected in their practices, we approached respondents and asked them to think of examples and narratives of instances in which organizations identified, dealt with, and discarded outdated knowledge structures. Some consultants were also experienced in coaching and individual consulting (e.g. leaders and managers). In these cases, we highlighted our interest in organizational unlearning, that is, in instances were knowledge was intentionally discarded across an organization. Furthermore, when they provided accounts of such instances, we asked additional questions to understand how they perceived it. The questions focused on the characteristics of such processes, and their barriers and enablers.
We adopted principles from grounded theory to analyze how change consultants relate these processes to their own work experience and draw theoretical findings from the interview data (Charmaz and Belgrave, 2012). Traditional interpretations of grounded theory require researchers to be naive about the phenomenon investigated and enter the field of inquiry with as few predetermined thoughts and pre-knowledge as possible (Glaser, 1978, 1992). We did not adopt this classical approach (Ward et al., 2015) and followed the epistemological background of constructivist grounded theory (Charmaz, 2006). This more recent approach to grounded theory (Mills et al., 2006) rejects the notion of researchers as “distant experts” (Charmaz, 2000: 513). Charmaz (2006) acknowledges the influence of researchers’ knowledge on acquiring and interpreting the data. Consequently, a rigorous analysis must intentionally deal with different perspectives and “the tension that exists between developing a conceptual analysis of participants’ stories and still creating a sense of their [the researchers’] presence in the final text” (Mills et al., 2006: 32). We engaged in memo writing throughout the research process. This allowed us to analyze the data, develop ideas and concepts, and refine questions, and prompted us to reflect on our role in the research process. We collected 15 pages of memos taken during the research process, which included ideas and emerging themes. In addition, one of this paper’s authors was not familiar with the discourse on organizational unlearning. This researcher questioned emerging themes and ensured that emanating categories were not biased through preconceived notions and existing claims.
To analyze and code the interview transcripts, we adopted a hierarchical coding strategy. First, in the initial coding phase, we identified narratives related to periods of organizational unlearning. Second, we developed common categories among the narratives (focused coding), and finally, we employed axial coding to specify the relationships between categories (Charmaz, 2006). In all stages, we continuously reflected on the emerging categories by going back and forth between the interview transcripts and emerging coding structure (Miles and Hubermann, 1994; Strauss and Corbin, 1990). We used the qualitative data analysis software NVivo to code data, develop theoretical assumptions, analyze them, and document the research progress.
Sampling
Explorative and grounded theory-driven studies require rich data and an appropriate sampling strategy. Charmaz and Belgrave (2012: 351) emphasize that a rich data set decreases the likelihood of making “misleading claims or writing a superficial analysis.” In order to ensure that different perspectives of practitioners are covered in the study, we followed a purposive sampling strategy that was defined prior to the study (Mason, 2002). Drawing on principles from grounded theory, we also adopted an iterative theoretical sampling procedure and searched for suitable participants based on emerging themes (Strauss, 1987). In our initial sampling phase, we conducted online research to identify change consultants in Central Europe suitable for our study. Two considerations guided the selection of our sample. First, since change consultancy is a broad field (Kakabadse et al., 2006), we aimed to include a wide range of change consultants (and their various fields of expertise). This familiarized us with different foci and helped us understand where and how organizational unlearning might occur. Second, Werr et al. (1997) noted that large consultancy firms often have pre-defined schemes and tools their consultants use. We did not want to draw on reports and insights reflecting the standardized and pre-defined strategies of these large consultancies. Thus, we explicitly targeted freelancers and smaller consultancies.
We searched for consultants who speak either English or German (the languages spoken by the authors of this paper). We included the United Kingdom, Germany, Austria, and Switzerland in our search frame. We also ensured that the consultants had broad and possibly international consulting experience. Following our search criteria, we initially contacted 25 potential respondents via e-mail explaining the overall idea of the study and providing a broad definition of organizational unlearning (defined as ‘an intentional effort to get rid of knowledge structures in organizations’), and asked if they were willing to participate. Approximately half replied, and after clarifying additional questions, 10 were willing to participate in our study. All interviews were conducted either face-to-face or via Skype.
During the interviews, we took memos and reflected on emerging themes and topics, which guided the selection of further participants. For example, the initial interviews revealed that instability, which derives from processes in which knowledge is being questioned, requires special forms of communication. Subsequently, we contacted change consultants who specialized in communication. Furthermore, we followed a snowballing approach (Robinson, 2014) and asked all interviewees for recommendations on other consultants or general fields of consultancy expertise (e.g. change during crisis) that could provide further insights for the study. During this second phase of our study, we conducted interviews with 14 additional change consultants. Ultimately, we interviewed 24 change consultants with different foci and expertise (two were based in the United Kingdom and the interviews were held in English), who came from 12 companies. The average interview length was 63 minutes and the median was 53 minutes (the shortest interview lasted 25 minutes, while the longest interview lasted 101 minutes). Our twofold sampling strategy, that is, iterative selection of interview partners based on emerging themes and topics, and asking for recommendations for additional respondents by previous interviewees (i.e. snowballing), was intended to counter referral chains that are caused by snowballing approaches and may likely lead to a selection bias. While iteratively moving between theory, sampling, data collection, and analysis, we failed to identify new themes or categories with interviewee number 19. We conducted additional interviews to confirm that new themes ceased to emerge (Breckenridge and Jones, 2009).
Findings
We identify two forms of organizational unlearning, namely (1) open-ended unlearning and (2) goal-directed unlearning. We also find that consultants prepare actors to intentionally discard outdated knowledge structures. Since this preparatory phase takes place prior to the actual unlearning processes, we will present this finding first.
Providing a rationale for organizational unlearning
Our findings suggest that before initiating organizational unlearning, consultants aim to establish a rationale for discarding obsolete knowledge in organizations. They first aim to understand the role of knowledge, express appreciation for outdated ways of doing things, and prepare actors for unlearning.
Understanding the embeddedness of outdated knowledge
Respondents highlighted the relevance of understanding how knowledge is embedded in organizational structures. Actors, who rely on existing knowledge structures, face difficulties when trying to change their mindsets and behaviors: It starts with realizing what [the organization] can let go of—what [it] should let go. Many people refuse to do so. […] That is the first step to embrace the new. (I 20)
Thus, consultants first seek to understand how outdated knowledge is embedded in organizations to subsequently help discard it: Sometimes, the new cannot unfold because the old is still present somewhere. We experience that during structural changes or changes in leadership. There are barriers for change because old elements still have an impact. (I 7)
We found that outdated knowledge structures are reinforced in organizations in different ways. Commonalities are that they make actors recall outdated behaviors or structures, or reminisce about existing assumptions or beliefs. Consultants described numerous triggers associated with outdated knowledge, ranging from architectural features (e.g. the design of seminar rooms) or cultural artifacts (e.g. the way people dress) to emotional associations with existing knowledge that hinder organizational unlearning processes. One respondent provided an example in which an organization underwent changes in management and planned to modernize existing business models. While actors were struggling to engage in the change process, the consultant soon realized why they were reluctant to adapt. Employees had problems letting go of outdated knowledge created when the founder was actively in charge of the company: At one point, I see this old guy walking around in the offices, which I think is weird because employees cannot be that old—that is impossible. At first, I just thought this man could be a senior manager. Later, I was in a meeting and suddenly he popped into the room. He just walked in […] and says, “Oh, sorry, you’re in a meeting,” and then he leaves. It was the founder and former CEO. (I 23)
However, we also found that outdated knowledge structures can be essential for an organization’s stability. Failing to account for the integral role of specific artifacts, in that discarding it is recklessly enforced, may result in resistance to change. Hence, respondents indicated the importance of understanding the embeddedness of outdated knowledge in terms of making room to create new knowledge and the delicate balance between knowledge structures that are more or less important for organizations. One consultant gave an example: [As part of a change process] one banking institute decided that the branch manager should no longer be called ‘director’. This was terrible for many people. Why? In this region, there is something like a triumvirate including the town’s mayor, local professor, and director of the local bank. Suddenly, the position of the director ceased to exist and he was now only a branch manager. […] People think about such connotations and artefacts, and this leads to resistance. (I 3)
Appreciating outdated knowledge
Related to the previous point, some consultants mentioned that established knowledge must be appreciated. They stressed the importance of consciously addressing and appreciating outdated knowledge before discarding it, because organizational actors usually invested significant time and effort into establishing it. Therefore, letting go of knowledge also entails emotional characteristics. Regarding outdated knowledge: We consciously dismiss old knowledge […] when we want to develop a new self-conception [as an organization]. It is important to appreciate and celebrate knowledge from previous years showing all employees that the past is not being devalued. I believe this is an important precondition—a ‘completion’ in a way—to open up for new ideas, assumptions, and routines. (I 9)
Respondents describe a tension between the value of established knowledge and letting go of that knowledge: Established knowledge provides safety and structure in an organization. However, it can also be a hindrance […] Because employees feel safe in these established structures, they hold on to them for too long. (I 5)
Consultants mention different methods to appreciate outdated knowledge before letting it go. For example, one respondent mentioned a case in which the organization staged a funeral wherein outdated knowledge was symbolically buried and thus, expelled (I 21).
Preparing for organizational unlearning
The consultants highlighted numerous interventions to prepare organizational actors for discarding outdated knowledge. According to our respondents, an organization’s history and culture strongly influence employees’ perception of change. From the consultants’ perspective, understanding the underlying values and beliefs comprising organizational culture is crucial when selecting an appropriate approach to initiate and facilitate organizational unlearning: We must invest much preparatory work in advance [before organizations let go of old knowledge]—preparatory mental work and heavy thinking actually—with all members of the organization. (I 19)
Convincing employees of the benefits of new ways of thinking and behaving is vital according to our interviewees. Consultants use various interventions to increase actors’ awareness of existing assumptions, beliefs, or routines, such as conceptualizing visions or providing one-on-one coaching. An important factor portrays the need to involve the entire organization and highlight the benefits of discarding previously established knowledge. Once actors understand why they should refrain from behaving in a certain way, they are more likely to discard this behavior: I am convinced that employees need to see the sense and value added when we ask them to get rid of existing knowledge structures. (I 5)
Another consultant mentioned how they confronted actors with their own behavior to make them aware of the benefits of discarding it: The question was, “how do employees understand that they need to unlearn?” […] We once videotaped them [and after showing it to them], and invited them to think of ways in which they can improve. […] They realized they needed to let go of their old behavior. (I 8)
Open-ended unlearning
Our findings reveal that consultants distinguish between two types of organizational unlearning. The first type is explorative, as the new knowledge structures to be learned have not yet been defined. In such scenarios, organizational actors actively question underlying assumptions and beliefs. As they become aware that they cannot further rely on how things have been done, but do not yet clearly see how things should be done in the future, we refer to this type as open-ended unlearning. Consultants highlight a number of means to initiate and facilitate the scrutinization of existing assumptions and beliefs. Interventions include breaking patterns, creating space for experimentation, ensuring (transparent) communication, providing time, and encouraging failure.
Breaking patterns
Consultants initiate open-ended unlearning by surprising organizational actors with unfamiliar interventions. The underlying rationale is that actors are driven by implicit assumptions and beliefs and in doing so, consultants hope to break with actors’ expectations of what would usually happen: We always try to break up established structures by changing the rules of the game. For example, we modify seating arrangements in the offices […], and the new positions affect employees. (I 9)
Other means include changing the regular organizational context by holding meetings in different settings and locations or finding and removing common elements of their daily jobs. In one project, a consultant removed all laptops and made employees visualize their thoughts on posters (I 20). Another way to break patterns involves bringing in people with other perspectives. When confronted with new or differing opinions, actors question their own underlying assumptions: When we manage to break up patterns and established assumptions, we immediately have completely different spheres of action, which we or organizations have not yet thought about. (I 19)
Creating space for experimentation
Consultants mentioned the importance of providing space where actors can explore new possibilities. These safe zones allow employees to experiment and create new ideas. According to respondents, two aspects are important. First, actors need freedom to develop and explore ideas. Second, they should not feel pressured by daily business: It is important to create space and provide employees with the opportunity to consciously experiment with different things. They should also be allowed to make mistakes. (I 7)
We found that such spaces do not need to be architectural. One consultant discussed an organization that aimed to develop more flexible structures. The organization invited actors to become part of special initiatives (e.g. talent development programs) where they were encouraged to think outside the box (I 15). Within this space employees were encouraged to ideate possible future scenarios for the organization. This type of experimentation yielded new insights while helping the organization initiate an organization-wide change process.
Ensuring (transparent) communication
The interviewees emphasized that open-ended unlearning evokes uncertainty. Actors cannot predict whether new ways of thinking and behaving will be appropriate. Therefore, consultants highlighted the role of transparency and communication.
We need transparency. […] Whoever is good at doing that—aligning different expectations, clarifying roles, and creating transparency—will be likely to succeed in the process. (I 12)
Transparency ensures that actors receive feedback about how progress unfolds. Our respondents suggested that fostering transparent communication is vital in conveying actors’ new roles and new ways of thinking and doing: Approving new ‘doings’ needs a lot of communication. It requires a mix of personal, oral, and written communication and necessitates new communication paths in the organization. (I 10)
Transparency such as open communication or regular meetings allows individuals to understand how changes will benefit them.
Providing time
Actors invest much time and effort in establishing knowledge. Therefore, as our respondents pointed out, employees need time to question existing knowledge and reflect on new, emerging ideas: I find that generally, unlearning is extremely time-consuming. […] Many people think we will take ten simple steps and then see first success. […] This success takes much longer and is actually an on-going story that comes with failure and crisis. (I 2)
To separate actors from established assumptions and beliefs, consultants implement interventions. For example, one consultant referred to a change project in which workshops were conducted every month for two and a half years (lasting two to three days each) (I 19). In the time between the workshops, actors worked on tasks given by the consultants, and collectively reflected on the outcomes. Over time, actors of this organization were able to get rid of outdated assumptions and develop new thinking patterns.
Encouraging failure
The respondents emphasized that open-ended unlearning occurs when actors have unfamiliar experiences. Consultants want organizational actors to refrain from drawing on established knowledge. Employees are likely to experience failure when trying new ways of thinking and behaving: One common way for how such processes fail is that the new yields greater uncertainty than the old. […] Then, people quickly fall back into the old ways of believing and doing things. (I 15)
Some respondents indicate that organizational actors tend to search for perfectly correct solutions. However, our respondents noted that they rarely exist in organizational life, and organizations should encourage or at least not sanction failure as a means and driver for finding new perspectives: You would need top managers, of course, who protect these new ways of doing things and provide teams with the freedom to continuously use and follow through with the new. I think this is the challenge when trying to unlearn the old and come up with something new. […] That employees are still allowed to do [the new], even if things do not always run smoothly. (I 12)
Goal-directed unlearning
The second type of organizational unlearning refers to situations in which organizations implement new and pre-defined practices or routines, which interfere with established ones. This type of unlearning is goal-directed, as there are defined knowledge structures that must be incorporated in the organization. During this type of unlearning, actors reflect on how they currently do things compared to how they should be done. This enables adjusting their own behavioral patterns to assimilate targeted behavior. Organizational actors have less freedom to explore and experiment. This type of organizational unlearning aims to realize a desired end-state by institutionalizing new practices or routines. Our findings reveal that this form of organizational unlearning is reflected in the following themes.
Parsing into little steps
As the consultants indicated, organizational actors may feel overwhelmed and struggle when too much change occurs at once. One way to facilitate the discarding of outdated knowledge includes splitting the process into smaller steps while providing actors with a clear idea and guidelines of what they should unlearn: [Unlearning] is not a one-time thing, but a tiresome process. One has to proceed empathetically. You need to understand the organization’s situation. What is happening there, what is going on in the heads of the people, how can they loosen up? This is where one can intervene […] We do this step-by-step. (I 19)
Providing guidance in this process decreases anxiety. Furthermore, consultants encourage management to ensure continuity, meaning that new steps are constantly added. Otherwise, actors might not take the process seriously. As one respondent noted, actors are more likely to unlearn when they are aware of how specific practices should be replaced with new ones: I cannot simply erase old knowledge by pushing a button. However, I believe it is about management making offers. For every step where people should unlearn, they need an alternative for what they should do instead. (I 5)
Reducing the influence of outdated knowledge
The consultants mentioned various interventions to reduce the influence of outdated knowledge on actors’ behaviors. Examples include changing the language used in the organization, because established and engrained terms or expressions may cause actors to retrieve outdated behavior. Furthermore, organizations foster unlearning when eliminating space or time to lament outdated knowledge: When employees are directly affected by restructuring processes, I sense they are busy and thus, forget the old structures. (I 4)
Additional means to prevent the recall of outdated knowledge involve changing the architectural space so that actors can no longer enact outdated behaviors.
Embedding the new
Consultants further reported that organizations enable this type of unlearning through structural changes accompanied by clear guidelines and rules. Organizations facilitate goal-directed unlearning when directing attention to new knowledge structures to gradually replace established ones. Actors tend to fall back into previously established habits when they fail to understand the application of new knowledge: [Organizational unlearning] seems to work when organizations give in to it and are backed by top management. It should be a relatively stable process. […] Regarding this ‘constantly falling back’ thing … I think that once you have taken the first step, you will never fall back to where you started. (I 20)
Identifying organizational actors who promote new ways of thinking and behaving positively supports the unlearning process in organizations. Here, ‘unlearning cells’ which comprise motivated actors moved to vital sections of the organizations, facilitate the discarding of knowledge. They positively support change and provide feedback: [In doing so] sub-cultures emerge when motivated employees gradually infiltrate their own units. (I 19)
Associating new knowledge with new artifacts or symbols, which allows actors to establish connections with newly introduced changes, can be another crucial factor in embedding fresh perspectives. Artifacts or symbols include new roles, metaphors, and analogies (e.g. illustrating how actors should behave; I 10), or codes of conduct (e.g. how people should dress; I 21).
Feedback structures
Our findings show that clear feedback structures accompany goal-directed unlearning. Consultants suggested that actors are more likely to enact new behavior when having positive experiences. Conversely, when organizational actors realize that outdated behavior is being rewarded, they stop engaging in new behavioral patterns. This is captured in the following quote: For example, [employees] observe how a colleague is being punished even though he tried to conform to new ways of working. Of course, the others will stop enacting new knowledge, and fall back on old knowledge. (I 23)
Comparing and contrasting new behavioral patterns with outdated ones seems essential when organizations intentionally discard knowledge. This helps increase awareness of the progress of organizational unlearning.
Furthermore, interviewees outlined the importance of communicating the organization’s commitment. Organizations that foster unlearning, as consultants suggested, take actors’ feedback seriously and constantly communicate that the process is being maintained. As one respondent noted, “clarity and honesty is good for employees” (I 10). When actors see and feel commitment, they are more motivated to continue with new knowledge structures and stop drawing on outdated knowledge.
Discussion
We explored organizational unlearning from the perspective of change consultants. The interviews revealed that the consultants could relate with the concept of organizational unlearning, and they provided numerous accounts in which organizations discarded existing knowledge structures. This led to conceptualizing two major forms of organizational unlearning. Furthermore, we identified a number of interventions and factors associated with organizational unlearning during change. Next, we discuss the main findings of this study.
Different forms of organizational unlearning
The interviews revealed that organizational unlearning can occur in two forms. On the one hand, the process of open-ended unlearning implies that organizational actors disassociate from established beliefs and routines to develop and experiment with new perspectives. On the other hand, goal-directed unlearning means that actors are required to discard existing knowledge in order to implement new and specified knowledge structures into the organization. To some extent, this differentiation overlaps with the conceptual distinction between cognitive and behavioral unlearning (Easterby-Smith and Lyles, 2011; Hislop et al., 2014; Klammer and Gueldenberg, 2019). While cognitive unlearning refers to a process in which organizational actors discard existing beliefs and assumptions, behavioral unlearning describes a process in which organizations discard established practices and routines. However, open-ended and goal-directed unlearning extend existing definitions in two central ways.
First, we assert a crucial difference between our proposed forms of organizational unlearning in terms of pre-defined goals. During open-ended unlearning, organizations cannot set clear and explicit outcomes or goals for the process. The process is open-ended in the sense that organizations aim to release outdated knowledge while exploring new opportunities, regardless of how realistic or desirable they are. While organizations engage in open-ended unlearning, they are searching for new opportunities that may arise for the organization. Thereby, the conceptualization of open-ended unlearning captures recent positions in the literature that implicitly or explicitly address the open-endedness of the process. For example, Nygren et al. (2017: 473) argue that unlearning can “expand the problem space so a wider range of options for action emerges,” even though it is not clear which actions will ultimately be chosen. Peschl (2019) argues that dynamic and uncertain futures enable organizations to utilize organizational unlearning as an open-ended strategy to identify emerging purposes and goals. For Brook et al. (2016), unlearning embraces a state of “not knowing” and “non-action.” In contrast, goal-directed unlearning presupposes a desired end-state or defined knowledge structures that should be implemented, but are incompatible with established ones. Organizations usually desire outcomes that should be accomplished in a certain period. This is aligned with more dominant approaches in the literature that relate the discarding of outdated knowledge to already defined and specified new knowledge structures (Klammer and Gueldenberg, 2019). This can involve integrating new routines and practices during mergers and acquisitions (Martin de Holan and Phillips, 2004; Tsang, 2008; Yildiz and Fey, 2010) or implementing new technologies into processes (Becker, 2010; Dumas et al., 2018). According to the consultants, the distinction between open-ended and goal-directed unlearning processes in organizations is important, as they lead to different outcomes and require different interventions. However, this distinction has not been explicitly discussed in the literature.
We highlight a second important distinction in terms of open-ended and goal-directed unlearning, and existing concepts, namely cognitive and behavioral unlearning. The terms cognitive and behavioral unlearning, and their characterizations, seems to imply that one process involves cognitive effort while the other does not. Our interviews, however, indicate that both forms of organizational unlearning processes require considerable reflection and cognitive effort. Even if organizational actors incorporate new and well-defined routines, it is important that they can reflect on how and to what extent they differ from established ones (e.g. as I 8 reported, this can be enabled by videotaping behaviors and screening them to those involved). Thus, understanding and sense-making are essential in both forms of organizational unlearning. However, building on this observation, a key difference between both processes is the availability of feedback. Whereas open-ended unlearning processes might not yield immediate feedback because there is no reference model or endpoint (regarding the desirability and appropriateness of new knowledge), goal-directed unlearning involves a clear goal; thus, it is characterized by providing and receiving immediate feedback. Some interviews also indicated that in some cases, open-ended unlearning precedes goal-directed unlearning. These findings are important means by which to better understand different forms of organizational unlearning, an effort predominantly taken for individual unlearning (e.g. Brook et al., 2016; Hislop et al., 2014).
Organizational unlearning in the context of organizational change
Many consultants suggested that the term organizational unlearning focuses on periods in which organizations need to question, identify, and disassociate from outdated knowledge. As recently suggested (e.g. Fiol and O’Connor, 2017; Grisold et al., 2017; Rezazade Mehrizi and Lashkarbolouki, 2016), the term refers to a process involving trial and error, which takes time, and often occurs simultaneously with learning. Notably, some organizational unlearning interventions are similar to previously discussed interventions and approaches described in the organizational change literature. For example, the notion of ‘unlearning cells’ (I 19) can be associated with the notion of change agents (Hartley et al., 1997). Similarly, the role of transparency and communication in the context of change has been stressed (Mazutis and Slawinski, 2008). However, assuming a different perspective, our respondents focused on the role of outdated knowledge during organizational change and respective interventions to deal with it. In numerous accounts, consultants emphasized the need to focus on how outdated knowledge is embedded in the organization (e.g. artifacts, managers, etc.). These findings stress the role of contextual factors that may facilitate the recall of outdated knowledge and are thus detrimental to discarding knowledge. This has received limited attention in previous research. Grisold et al. (2017) and Kluge and Gronau (2018) argue that organizations should manipulate the environment to impede actors from being reminded of outdated knowledge. Contributing to these approaches, our study elicits numerous environmental features or ‘retrieval cues’ (Kluge and Gronau, 2018) that can reduce the influence of outdated knowledge structures and thereby facilitate organizational unlearning. As the interviews indicated, these cues can be leadership behaviors (e.g. by punishing the enactment of new behaviors; I 23), language (e.g. by re-naming certain positions in an organization; I 3), symbolic artifacts (e.g. artifacts used to execute work; I 20), or the enactment of outdated knowledge structures in the form of previous leaders who are still present in the organization (I 23). Thus, based on our findings, we believe that contextual and environmental aspects are highly relevant for the organizational unlearning process. Furthermore, some interviewees stressed that knowledge, even if outdated and should be subject to change, may play a key role in the organization’s identity (Fahrenbach and Kragulj, 2019). This relates to Tsang (2008), who observed that a routine can be more or less socially embedded. If it is more embedded, it is harder for organizational actors to discard the routine. However, while the organizational unlearning literature implicitly holds the premise that outdated knowledge no longer provides value to the organization, our findings suggest it is not automatically a ‘bad thing’. It may be important to acknowledge and appreciate it before letting it go. This dual role of outdated knowledge structures in organizational unlearning processes has not been discussed in the existing literature.
Organizational unlearning interventions
Our findings indicate interventions and enablers that support organizational unlearning. This is unsurprising, as change consultancy involves entering organizations and implementing interventions to disrupt the existing order (Clegg et al., 2004a, 2004b). However, interventions play a negligible role in the existing literature on organizational unlearning. Researchers agree that unlearning in organizations requires attention and conscious effort (Martin de Holan, 2011; Tsang and Zahra, 2008), while intention per se might not suffice because organizational actors face uncertainty and anxiety. This, in turn, might cause them to draw on outdated knowledge structures (Blackman et al., 2012; Brook et al., 2016; Hislop et al., 2014; Visser, 2017). Kluge and Gronau (2018) and Grisold et al. (2017) outline interventions to modify an organization’s environmental context to inhibit actors from drawing on outdated knowledge. Nygren et al. (2017) explain how a dialogue setting can help researchers reframe and unlearn their established assumptions. Becker’s (2008) study on technology adoption reveals factors that enable organization-wide unlearning. However, the question of how to realize organizational unlearning has received scant attention. Paraphrasing Rupčić (2017) in “How to unlearn and change—that is the question!,” we highlight numerous interventions that can be applied in different stages of the organizational unlearning process. Specifically, our findings indicate cues and triggers that might be associated with the recall of established knowledge structures, including the role of transparency and communication, or coherence of the feedback provided. Our findings also reveal the importance of preparatory action prior to the organizational unlearning process. Specifically, consultants emphasized that the role of established knowledge should be acknowledged and understood, explaining how they provide a rationale for organizational unlearning in their work.
Interdisciplinary and multi-level perspectives on organizational unlearning
Aligned with previous claims, our findings indicate that organizational unlearning refers to different knowledge structures, namely routines, values, and beliefs (Easterby-Smith and Lyles, 2011; Klammer and Gueldenberg, 2019). Regarding how organizational unlearning unfolds, we found empirical evidence for Zhao et al.’s (2013) theoretical assumption that unlearning processes unfold across all levels in the organization: the organizational, team, and individual levels. For each level, the respondents provide examples and interventions; for instance, unlearning on the organizational level is facilitated by clear communication (I 10) and transparency (I 12), but the process is also realized by groups or teams when providing space for experimentation (I 7). Furthermore, consultants approach individuals directly, such as by videotaping and screening the recordings to them (I 8).
While existing research investigated distinct levels in isolation (Hislop et al., 2014; Klammer and Gueldenberg, 2019), we see potential for future research to explore how organizational unlearning unfolds across levels. This can be illustrated by I 23’s account of the founder and former CEO who was still physically present in the organization and therefore reinforced outdated beliefs and routines. With reference to existing research (Hislop et al., 2014), we can explore such cases by means of individual unlearning since the manager would have been required to fundamentally unlearn his established beliefs about his own role in the company. At the same time, we can take into account that the manager’s ability to unlearn was connected to how well the collective could unlearn as his presence triggered the enactment of old knowledge structures (Kluge and Gronau, 2018). Looking at it from this perspective, organizational unlearning yields interesting insights in terms of recursive relationships and feedback loops on different levels which influence the success of such processes (Farjoun et al., 2015).
Exploring how these levels are connected might help in painting a more detailed picture of organizational unlearning. This, in turn, can be enabled by interdisciplinary research approaches (Rousseau, 2011). Recent attempts to introduce psychological perspectives into the organizational unlearning research (Grisold et al., 2017; Klammer et al., 2019b; Kluge et al., 2019) highlight new research methods applicable to study knowledge loss in organizations. For example, Kluge et al. (2019) suggest examining organizational unlearning processes through an experience sampling method wherein organizational actors report their thoughts, emotions, and perceived progress of the organizational unlearning process. Using such methodologies can potentially contribute to a better understanding of theoretical claims about cognitive processes of individual actors during the enactment of organizational unlearning, for example, how they cope with cognitive dissonance (Grisold and Mendling, 2017; Yildiz and Fey, 2010).
New avenues for research on organizational unlearning
Our findings indicate aspects related to the unlearning process in organizations not yet addressed in existing studies. We note avenues for future empirical research (Cegarra-Navarro and Wensley, 2019; Klammer and Gueldenberg, 2019; Kluge and Gronau, 2018; Visser, 2017).
First, respondents emphasize the role of culture in organizational unlearning processes. In the interviews, consultants reported that the history, values, and atmosphere in the organization strongly influence how well organizations intentionally discard established knowledge. While culture has been examined in relation to organizational change (e.g. Cameron and Quinn, 2011), it has not been discussed in relation to organizational unlearning. We suggest that future research does so.
Furthermore, several change consultants indicated that the discarding of knowledge becomes important when organizations aim to develop more agile work practices. This observation is not surprising, given that research observes that the implementation of agility forces organizational actors to rethink established routines and develop new skills and competences (Teece et al., 2016). Accordingly, contexts where agility is implemented in organizations might provide opportunities to explore the role of organizational unlearning before, during, and after organizational change processes.
Several consultants noted that changes in (top) management provide interesting insights into the dynamics of organizational unlearning. For example, they can facilitate (I 5) or impede (I 23) the organizational unlearning process. Our respondents suggested the importance of unlearning during changes in family businesses. Here, change occurs on many levels, for example, when the organization is taken over by a younger successor replacing a leader that had been in charge for years or decades (Dyer, 2003). Such instances may elucidate the relation between organizational unlearning and culture, strategy, or routines. As such, it also seems promising to explore how distinct organizational levels are involved in unlearning processes, or how different groups of employees (e.g. ‘old employees’ who have been working in an organization for a long time and ‘new employees’ who recently joined the organization) would engage in the process.
Limitations
Our study has several limitations. First, engaged scholarship is not without criticism (Bansal et al., 2012; Kieser and Leiner, 2009; McKelvey, 2006). Practitioners interact with the world through their own agendas. For example, consultants are paid to help organizations in some way and thus, they are interested in bringing about desired success. McKelvey (2006) notes that practitioners are not independent observers, but have biases, assumptions, and expectations. Furthermore, research and practice have their own methods, language, and criteria for evaluating the validity of knowledge (Kieser and Leiner, 2009). Therefore, it is challenging to transfer experiences gained in practice into research, or vice versa. We concur with both assertions. However, engaged scholarship can account for perspectives and experiences gained ‘in the wild’. Kieser and Leiner (2009) note that research and practice can contribute to each other in the sense that one provokes the other, such as by illuminating previously unknown aspects of a phenomenon or the related dynamics. Accordingly, the implications of this study highlight many aspects and blind spots not yet considered in established research on organizational unlearning. Thus, we believe that engaged scholarship adds insights that, when approached from more scientific perspectives, further advance our understanding of organizational unlearning.
A second limitation results from the role of consultants in change processes. Consultants might only acquire a limited picture of intra-organizational dynamics. For example, they may only be able to provide limited insights into cultural aspects or the atmosphere during organizational unlearning processes. Moreover, consultants are called into an organization (mostly to address existing problems or crises), and it is unclear how pre-existing dynamics affect their work. For example, when an organization calls in a consultant after ‘in-house’ solutions have failed, actors may be more willing to unlearn (or not).
Finally, consultants with broad experience reported on organizational unlearning processes. Thus, we could theorize about different forms of the process and underlying characteristics. However, the consultants could not provide detailed insights into the process. In-depth single case studies are needed to elicit how organizational unlearning unfolds over time and across organizational levels (Rezazade Mehrizi and Lashkarbolouki, 2016).
Conclusion
We explored organizational unlearning from the perspective of practitioners who regularly deal with the process, namely change consultants. We examined what organizational unlearning looks like in practice, how it unfolds, and how it can be brought about through various interventions. We identified two different forms of organizational unlearning. Open-ended unlearning is characterized by intentionally disassociating from established knowledge structures to explore new opportunities while the outcome of the process is not known. Goal-directed unlearning involves giving up established ways of doing things to integrate new, already specified knowledge structures. Our results also reveal different interventions and enabling factors that support the respective organizational unlearning process. We also found that prior to discarding established knowledge structures, organizations need to provide a rationale for unlearning. During this stage, it is important to understand how knowledge is embedded in the organization and create awareness that obsolete knowledge needs to be discarded. We argued that more empirical research on organizational unlearning is needed. Here, we provided specific contexts, aspects, and factors that can be examined in future research.
Footnotes
Acknowledgements
We would like to thank the three anonymous reviewers for their constructive feedback and comments. We also want to thank Alexia Panayiotou, serving as the associate editor, for her guidance and valuable suggestions during the revision process.
Authors’ Note
Thomas Grisold is now affiliated with the University of Liechtenstein, Liechtenstein.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by EU H2020 program under the MSCA-RISE agreement 645751 (RISE BPM) and Research Fund of the University of Liechtenstein (Grant Number ent_18_6).
