Unlearning for Democratic Organizing

Abstract

This article advances empirical and conceptual understanding of how unlearning can help to sustain democratic organizing. While learning to confront persistent tensions is widely recognized as a central element of democratic organizing, there is surprisingly little research on its practice. In exploring this gap, unlearning emerged as a distinct yet integral aspect of individual and collective learning in democratic organizing – and to sustaining democratic organizing across time and scale. Our abductive analysis of qualitative research across four established (>20 years) and scaled (>50 members) UK-based worker cooperatives uncovered the centrality of relational processes of unlearning habitual beliefs, values and patterns of behaviour in interaction with others, the organization and the environment. Empirically, we discern three relational dynamics that create a need for unlearning and three relational practices for engaging in unlearning. We offer conceptual depth by formulating three analytical propositions for future research and practice.

Introduction

Democratic organizing sits at the heart of the debate about the comparative advantages of capitalist firms (reduced free-riding, increased joint effort outcomes, stronger monitoring incentives) and cooperatives (resilience, gender balance, fairness, autonomy, collective agency), for organizational performance and socio-economic outcomes such as involuntary unemployment and unequal distribution of income (Berti & Pitelis, 2022). Democratic organizing refers to a practice of integrating collective decision-making with equity, individual autonomy and agency (Griffin et al., 2022; S. Parker & Parker, 2017). It is prefigurative, i.e. a present-time manifestation of an ideal future-time in which the default modes of relating, producing and consuming are markedly different from those of the present. Its prefigurative nature makes democratic organizing a messy and imperfect relational process of enacting values of an ideal society within the struggle for that society (Kioupkiolis, 2010; Maeckelbergh, 2009).

There is a growing sense that efforts to perpetuate and grow industrial democracy have ‘experienced concerted pushback from neoliberal, neo-Taylorist global capitalism and its political operators’ (Ravn et al., 2023, p. 3). Democratic organizing has been found to face pressures of ‘degeneration’ (including hierarchical norms, efficiency pressures and structural constraints) that, if left unaddressed, lead to a gradual decline of collectivist-democratic values or even make democratic organizing unsustainable (Diefenbach, 2019; Langmead, 2016). The continued survival and success of industrial democratic organizations evidences that democratic organizing is possible and necessary. By learning to better confront the relational dynamics that emerge from the ongoing tension between prefiguration and degeneration, democratic organizing can improve and sustain its daily struggle for a fair and sustainable political-economic system (Diefenbach, 2019; M. Griffin et al., 2022; Langmead, 2016).

This article emerged from our interest in understanding the conditions under which democratic organizing is sustained within worker cooperatives (WCs). What sets WCs apart from other forms of organization is that ownership rights are restricted to ‘worker-members’ who, often equally, distribute benefits and make decisions (Chaddad & Cook, 2004) and they adhere to the cooperative movement’s internationally recognized framework of values and principles (International Co-operative Alliance, 2019; T. Webb & Cheney, 2014). We were driven by the puzzle of how and why such organizations manage to uphold and expand their collectivist-democratic aspirations despite persistent structural, cognitive, emotional and behavioural pressures that may lead to degeneration (Powell, 2021).

The UK is an analytically significant setting. Its enabling ecosystem is far thinner than in Italy, France or Spain, where cooperatives benefit from constitutional mandates, fiscal incentives and sectoral legislation. Yet the UK has a historically influential cooperative tradition, exemplified by the Rochdale Pioneers. Although formal collectivist-democratic organizations remain relatively rare and small, several ‘exemplary cases’ have endured and flourished. While WCs are widely studied as exemplars of democratic organizing (Bretos & Errasti, 2017; M. Griffin et al., 2022; McMahon, 2023), there continues to be a need for ‘further research . . . on the resources, structures, and practices that contribute to the resilience of worker cooperatives’ (Cheney et al., 2014, p. 595) – i.e. for wider understanding of their ability to confront the ongoing tension between prefiguration and degeneration (Ravn et al., 2023).

Our article centres the role of unlearning in sustaining democratic organizing. While learning to confront persistent tensions is recognized as a central element of democratic organizing (Diefenbach, 2019; M. Griffin et al., 2022; Rothschild & Whitt, 1986), there is surprisingly little research on the practice of learning, with unlearning being even more understudied. We did not set out to explore unlearning, but it surfaced as a central theme in our abductive analysis (Tavory & Timmermans, 2014) of a broader project (Powell, 2021). Two emergent themes prompted the development of this article. First, we found that relational dynamics of confronting the challenges of democratic organizing created a need for unlearning. Second, we identified relational practices of unlearning that shaped abilities to effectively and democratically engage in these relational dynamics.

Based on our findings on unlearning in democratic organizing, this article aims to advance empirical and conceptual understanding of this understudied phenomenon. Unlearning has become a timely and significant concept in organizational studies (Klammer et al., 2025; Kluge, 2023). Going beyond the popular definition of unlearning as an intentional discarding process (Klammer et al., 2025), we define unlearning as an ongoing relational process of letting go of and overcoming engrained modes of organizing, being, and thinking. While not superseding other practices such as solidarity, mutual support or consensus-building, this article examines how unlearning can enable democratic organizing to better confront the relational dynamics through which such practices are enacted (Kluge, 2023; Pardim et al., 2024).

The article answers two research questions:

1. Why does democratic organizing have a need for unlearning?

2. How can democratic organizing engage in unlearning?

In response to RQ1, we explore three relational dynamics that create a need for unlearning: negotiating different models for ‘sharing the load’ of work and responsibilities, challenging ‘vested interests and sacred cows’, and upholding ‘a sense of exceptionalism’. These are not simply problems to be resolved but persistent contradictions that require continuous negotiation and adaptation. Relational dynamics are patterns of interaction, behaviour and change that emerge within and between individuals, groups and structures. Rather than discarding extant beliefs and behaviours to learn how to enact cooperative ideals, our findings demonstrate that democratic organizing involves an ongoing and integrative process of unlearning.

In response to RQ2, we identify three relational practices of unlearning. Relational practices are ordinary activities enacted through making sense of and shaping organizational processes and outcomes, which offer opportunities to both maintain and change the relational dynamics that shape them. ‘Integrating individuals’ refers to how unlearning addresses relational dynamics of sharing work and responsibilities by confronting beliefs and behaviours from conventional organizations with the ongoing challenge of maintaining cooperative practices among members. ‘Reviewing and renewing’ signifies how unlearning addresses relational dynamics around vested interests and sacred cows by regularly reflecting on their identity and purpose. ‘Turning outwards’ denotes how unlearning addresses relational dynamics that create a sense of exceptionalism through engagement with the wider cooperative movement.

When turning to the unlearning literature to conceptualize these findings, we found that it offered insufficient clarity to explain how unlearning can help sustain democratic organizing. Through an iterative back-and-forth between our findings and relevant literature, we developed three analytical propositions to add conceptual depth and guide future research and practice:

1. Democratic organizing is sustained through ongoing relational processes of unlearning.

2. Unlearning enables democratic organizing to work through defensiveness and overcome habitual relational patterns.

3. Creating systemic conditions for unlearning across levels of organizing enhances democratic resilience.

Our propositions offer a relational perspective on unlearning as an ongoing process of confronting the challenges of democratic organizing. We drew on Mary Parker Follett’s (Follett, 2003c; Stout & Love, 2015b, 2015a) relational process ontology to conceptualize our findings in these terms because it explains organizing as ongoing processes of reciprocal interactions between interdependent parts. We extended this conceptualization with the notions of transformative unlearning (MacDonald, 2002) and landscapes of practice (Wenger-Trayner & Wenger-Trayner, 2015) to explain our findings on the role of defensiveness and levels of organizing in unlearning from a relational perspective. We argue that unlearning can help to sustain democratic organizing by bringing people together across differences to develop new ways to adapt to the tension between prefiguration and degeneration.

While mobilizing familiar arguments for learning organizations (e.g. Argyris & Schön, 1978) and democratic practice in WCs (e.g. Greenwood, 1991), this article is the first to empirically examine and conceptualize unlearning in democratic organizing. It makes a significant conceptual contribution across the democratic organizing and unlearning literatures by extending the unlearning concept through a relational process approach that integrates both the emergence of the need for unlearning (relational dynamics) and the nurturing of ongoing purposive unlearning (relational practices) (Klammer et al., 2025). We leverage unlearning to advance the emergent relational process approach to the theory and practice of democratic organizing (M. Griffin et al., 2022; Hernandez, 2006; Langmead, 2016). Moreover, we speak to the wider organization and management literature on unlearning by extending its empirical and conceptual foundations and advancing arguments for ‘transformative’ unlearning across diverse organizational contexts (Klammer et al., 2025; Macdonald, 2002). Finally, we present unlearning as a meta-capability with relevance to addressing inherent tensions in a wide range of akin sub-fields, such as social innovation, participatory governance and the commons (Bartels, 2023).

We first discuss the tension between prefiguration and degeneration in democratic organizing to centre the relational perspective and role of (un)learning. Second, we explain how our analysis led us to identify unlearning and develop new empirical and conceptual understanding. Third, we explore three relational dynamics that evidence the need for unlearning (RQ1) and examine three practices of unlearning in democratic organizing (RQ2). Fourth, we outline three analytical propositions on how unlearning can sustain democratic organizing for future research and practice. Finally, we reflect on the significance of our research for advancing a fair and sustainable political economy.

Democratic Organizing, Prefiguration and Degeneration

Democratic organizing is a relational iteration of organizational democracy. Following Griffin et al. (2022), we understand ‘democratic organizations’ as integrating collective decision-making and governance with values of equity and individual autonomy. Our interest is particularly in WCs with highly autonomous and participatory collectivist-democratic organizing. In these ‘radical’ interpretations of cooperative prefiguration (Pateman, 1970; Rothschild, 2016; Rothschild-Whitt, 1979), power rests with the collective, and functional authority is derived and distributed from this collective power. The Mondragon Corporation and its constituent WCs are widely studied as prime examples of worker-owned democratic organizing at scale (Basterretxea & Albizu, 2010; Basterretxea et al., 2019; Flecha & Ngai, 2014). The John Lewis Partnership is another popular case (Cathcart, 2013a; Paranque & Willmott, 2014), with Cathcart’s (2013b) work notably addressing the ‘paradox’ of democratic participation in its highly ossified structure. Comparative analyses of efforts to manage and resist degeneration (Storey et al., 2014) highlight the need to strengthen understanding of how to handle the tensions of democratic organizing.

The potential of democratic organizing lies in its prefigurative nature. Prefiguration removes the temporal separation between the struggle of the present and the future as envisioned and desired by bringing the future into the present through action (Kokkinidis, 2012; Reedy et al., 2016). In contrast to a theory of social change that entails goal setting and planning, prefiguration theorizes through action; it is ‘something people do’ (Maeckelbergh, 2009, p. 68). Actors may not label their practice (e.g. ‘cooperation’), nor should we expect to find a consistency of interpretations of terms or what constitutes ‘good’ prefigurative practice (Griffin et al., 2022; Kioupkiolis, 2010; M. Parker et al., 2014). WCs arguably represent the pinnacle of cooperative prefiguration. They involve the highest degree of member participation of all forms of cooperative (Cheney et al., 2014), with minimal social and economic stratification and collectivist decision-making, minimizing opportunity for exploitation by secondary (managers) or tertiary (shareholders) actors.

The challenges of prefigurative practice are captured in the ‘degeneration thesis’: gradual slipping toward less collectivist-democratic practice or even wholesale failure of democratic organizing. The rich literature on multiple factors undermining democratic organizing (Berti & Pitelis, 2022) originates from Weber’s (1978) postulation of the inevitability of bureaucracy, along with Marxist and Socialist critiques of cooperatives as rearrangements of capitalist relationships (S. Webb & Webb, 1902). Degeneration frameworks include Cornforth’s (1995) categorization of constitutional, external or internal threats and Meister’s (1974) four-stage process of degeneration triggered by conflict between direct democracy and a ‘badly developed economic function’. While the quality and potential of their economic function is debated, the literature on the tensions associated with the dual nature of cooperatives as simultaneously capitalist enterprises and democratic member associations is well established (Novkovic et al., 2022).

We adopt the view that, although not a universal law (Michels, 1915), the ‘threat of oligarchy’ is always present due to, for example, limited leadership turnover, minority control of resources and low levels of participation (Diefenbach, 2019). Like open wounds, when left unattended, such issues ‘fester’ and undermine democratic practice (Chatterton & Pickerill, 2010, p. 481). Oligarchy is not a feature of structures or behaviours but a particular distribution of power that undermines democracy (Leach, 2005). Thus, degeneration produces moral injury in the form of distress arising when individuals witness, participate in or fail to prevent violations of deeply held moral commitments (B. J. Griffin et al., 2019). Degeneration matters not simply because formal democracy declines, but because members perceive a breach in core values.

A growing stream of research takes a relational approach, in which degeneration manifests as underlying forces clash (Hernandez, 2006; Langmead, 2016; Stryjan, 1994). Democratic organizations exist in constant tension with the isomorphic environment in response to which they were created (DiMaggio & Powell, 1983). Thompson (2015) suggested that the prevailing mode of organizing in business suppresses the ability to overcome the cooperation/coordination trade-off (i.e. non-hierarchical structure vs. efficiency of productivity) by privileging coordination. This must be qualified by recognition of variation in institutional pressures (e.g. the strength of neoliberal influence) and alternative non-hierarchical forms (open innovation, platform organizing). More importantly, democratic organizing does not resolve this paradox but embraces the interdependence of cooperation and coordination as a persistent paradox to navigate (M. Griffin et al., 2022).

Hence, the relational approach shifts the focus from democratic organizations as fixed entities to democratic organizing as an ongoing process. Practitioners of democratic organizing must attend to a range of both structural-relational and cultural-cognitive tensions in relation to their organization and environment (Thompson, 2015). For instance, both structuredness (formality, clarity, transparency, coherence) and structurelessness (flexibility, informality, creativity, improvisation) can be beneficial or detrimental to the way democratic organizing is practised (structural-relational) and understood (cultural-cognitive) (Freeman, 1972). Rather than settling on either state, democratic organizing is an ongoing relational process of adapting its component parts (Diefenbach, 2019; M. Griffin et al., 2022; Langmead, 2016).

Ongoing learning is essential to adapt democratic organizations in ways that prefigure their shared sense of direction (Hernandez, 2006; Stryjan, 1994; Varman & Chakrabarti, 2004). Such relational processes have been conceptualized as ‘spaces of possibility’ (Cornwell, 2012; Kokkinidis, 2015), ‘progressively creating slack’ (Varman & Chakrabarti, 2004, p. 204), a practice of ‘antiperfectionism’ (Kioupkiolis, 2010, p. 149) and ‘learn[ing] to live the paradox’ (M. Griffin et al., 2022). Unlearning could play a significant, yet understudied, role in dealing with the challenges democratic organizations constantly face with multiple factors, such as selection pressures, (in)efficiency and heterogeneous preferences regarding planning horizons, risk and investment limitations (Tortia, 2021).

Despite obvious links with the industrial democracy tradition and its focus on democracy as learning through difference (Ravn et al., 2023), there has been limited research directly focused on learning in democratic organizing or (worker) cooperatives more specifically. Canonical authors such as Rothschild-Whitt and Lindenfeld (1982) highlighted the lack of educational spaces where people might ‘learn democracy’, while Sauser (2009) emphasized the need to ‘foster democracy’ (p. 153). Thompson (2015) argued that ‘deep-level cooperation is achieved through an organizational culture, which enables solidaristic behaviour’ (p. 5) but did not explain how to develop this culture. Notwithstanding some nascent insight into cooperative learning (Hartley, 2014; Hartley & Johnson, 2013; Powell, 2016) and unlearning as a form of education and political emancipation (Chokr, 2009), we lack understanding of how unlearning influences democratic organizing. As the core principles of democratic organizing often clash with dominant cultural and structural norms, individuals and groups must unlearn and reconfigure habitual ways of thinking, acting and relating. This unlearning should avoid the utopian trap of new routines becoming fixed, instead creating space for alternative and evolving prefigurative practices.

Methodology

This article derives from a larger research project conducted between 2016 and 2021 (Powell, 2021). We designed an interpretivist multi-case study (Bartels & Wagenaar, 2025) that aimed to learn about members’ experiences of the tensions and practices of sustaining organizational democracy in established and scaled WCs in the UK. Case selection was guided by the principle of creating conditions for surprise by maximizing exposure to diverse experiences in comparable contexts: UK-based WCs with worker-member ownership, commitment to collectivist-democratic control, at least 20 years longevity, and between 50 and 200 members. Other factors such as growth, profitability, sector and governance mechanisms were considered but ultimately excluded, as they were either irrelevant or overly restrictive. The primary concern was to engage cases where age and scale were likely to place pressure on collectivist-democratic organizing.

We identified relevant cases and negotiated participation at the Worker Co-op Weekends in 2017 and 2018. All four case organizations (see Table 1) operated in the whole and organic food sector. Their pioneering position in these markets was a key factor in their survival and growth. Most were early movers in their locality with little competition, giving them space to experiment and innovate. These markets have since become mainstream and profitable, creating opportunities for growth but also greater competition, particularly in recent years. WC4 is notable for its two divisions: retail and wholesale (coded WC4R and WC4W). The cases were relatively homogeneous, offering scope to learn from differences in practices and outcomes.

Table 1.

Case studies.

Case	Region	Description	Age	Members	Interviews	Observation
WC1	Northern England	Wholefood wholesaler	43	~150	1	None
WC2	Northern England	Grocery	23	~70	8	2 days
WC3	Scotland	Wholefood wholesaler	40	~55	16	4 days
WC4R	Southern England	Grocery	47	~60	6	1 day
WC4W	Southern England	Wholefood wholesaler	33	~69	9	2 days

The project was guided by constructivist grounded theory (Charmaz, 2014). Constructivist grounded theory is a method of simultaneous data collection and analysis, constructing analytical codes and categories grounded in the data, and memo-writing for gradual theorization. Constructivist grounded theory enables building middle-range theory through proximity to the studied contexts and successive levels of data analysis and conceptualization (Charmaz, 2014).

In 2018, we began with exploratory interviews with five experienced ‘cooperative movement actors’ (CMAs) to refine the research focus and inform case selection and access strategies. We then conducted 40 interviews (20–60 minutes each) with members across four WCs. Using open interviewing (Bartels & Wagenaar, 2025), we aimed to capture concrete experiences of sustaining democratic organizing through participants’ emergent narratives. Alongside thematic guiding questions, we used targeted probes to direct conversations towards specific experiences and invited additional reflections to elicit unexpected insights.

Access varied across WCs, sometimes allowing us to observe meetings and everyday operations and review confidential internal documentation. Our involvement ranged from ‘passive’ to ‘moderate’ (Spradley, 1980). Ethical approval was obtained from our institution, and we adhered to requirements for informed consent and anonymity, while considering broader ethical implications (Social Research Association [SRA], 2003) related to our positionality as sympathetic yet external academics and our commitment to ‘giving back’ to participants.

Interviews were transcribed intelligent verbatim and analysed through iterative cycles of open coding, thematic grouping and memo-writing (Charmaz, 2014). Coding and analysis followed each set of expert interviews and fieldwork periods. Due to access constraints, we could not return to the field as typically encouraged in constructivist grounded theory research. Instead, we adopted a constant-comparison approach, involving batch-coding across cases (interviews, observation fieldnotes, and, where available, internal minutes or briefing) followed by memo-writing focused on emergent themes. Interview data were prioritized to foreground participants’ voices, while observational and documentary sources contextualized systems and historical trajectories.

We produced 20 memos (15–50 pages each): five from expert interviews, ten from case study interviews and fieldwork, and five integrative memos synthesizing earlier insights to develop the conceptual framework. The concept of unlearning first appeared in Memo 3, which examined challenges in ‘nurturing’ members into cooperators, and resurfaced in Memo 15, which explored maintaining a shared worldview among long-standing and new members.

During the final memo-writing phase, we identified relational process ontology, particularly Follett’s integrative process (Stout & Love, 2015a), and practice theory, notably landscapes of practice (Wenger-Trayner & Wenger-Trayner, 2015), as conceptual anchors. Some WCs were regularly ‘reviewing and renewing’ their organizational purpose, proving more effective when this occurred frequently and with openness to new members, ideas and practices. Integrative process explains how synthesizing diverse perspectives into a unified approach generates mutual benefit, while landscapes of practice illuminates how adaptation across organizational spaces reinforces coherence and resilience.

Unlearning was one of 13 primary themes in the original study. Its significance became clearer when leading scholars urged us to analyse it further. We adopted an abductive approach (Tavory & Timmermans, 2014) to revisit the data, identifying 22 relevant sub-themes and synthesizing them into six themes across two categories: why unlearning was needed (Sharing the load, Vested interests and sacred cows, a Sense of uniqueness and exceptionalism) and how it was practised (Integrating individuals, Reviewing and renewing, Turning outwards). We then engaged with the unlearning literature to conceptualize these categories in dialogue with the relational process framework that had emerged from the wider study.

Through this process, we established that unlearning is not a one-off, linear activity but an ongoing relational process responding to tensions across organizational spaces and time. The first category reflects relational dynamics (patterns of interaction) while the second comprises relational practices (enacted responses) that enable democratic engagement with these dynamics. The following sections address both categories in response to RQ1 and RQ2 and outline analytical propositions developed through iterative engagement between theory and empirical insights. The examples and quotes selected are reflective of the established sub-themes and codes or are highlighted as unique instances.

The following sections examine both categories in response to RQ1 and RQ2, providing an empirical foundation for understanding the need for and practice of unlearning in democratic organizing. We then present the analytical propositions developed through iterative engagement between theory and data, offering deeper conceptual insight into how unlearning sustains democratic organizing. In doing so, we address conceptual issues and debates that, without further development, risk undermining clarity and practical relevance. These propositions highlight the significant implications of our findings, adding conceptual depth and guiding future research and practice.

Relational Dynamics That Create a Need for Unlearning

In answer to RQ1, we discuss three relational dynamics (emergent patterns of interaction) of confronting the ongoing tensions of democratic organizing that create the need for unlearning. We demonstrate that WCs can only develop temporary, imperfect solutions and constantly need to unlearn how they think about, enact and organize democratic practice.

Sharing the load

Unlearning proved crucial for relational dynamics of ‘sharing the load’ in decision-making and work. From viewing meetings as an inconvenience to not taking on ‘rotated’ responsibilities, members struggled to learn and maintain cooperative behaviour and had to confront others’ tendencies to disengage. This remained true even with membership policies and ‘member job descriptions’ in place. CMA4 argued that members espouse that ‘human beings are cooperators’ but could also fail to adequately engage in ‘cooperative culture’, leading members to question not only their understanding of what ‘being a member’ means, but also the cooperative itself, ‘in the worst-case people will end up thinking, well, what is all this? What’s all this about? Why can’t we just be a normal company?’ (CMA4).

You’ll find that there’s maybe a bit of a push and pull, where the business being very successful can tend to drag people towards, ‘Let’s just be a super successful business.’. . . And other people who would see it as the reason it’s a successful business is because of all of its parts, so that its ethical foundations are absolutely crucial, they can’t be allowed to drift. This shared responsibility, and willingness to accept other roles, again, has to be carried on because without it, the whole thing, not just the co-op but the business itself, would fall apart. (WCM39-WC4W)

A common diagnosis of this tension between prefiguration and degeneration across WCs was to highlight that members sometimes failed to connect with or became detached from the concept of ‘cooperation as a collective endeavour’ (WCM36-WC4R). By treating the WC like a conventional wage-labour relationship, ‘some people will gripe about, oh this that and the other. . . Well, you are part of the business so why not think about changing it?’ (WCM36-WC4R). Unlearning engrained attitudes and behaviours with which individuals ‘arrived’ at cooperation and through which they practised was presented as the solution. For instance, CMA1 argued that people must unlearn competitive mindsets, predispositions and ‘natural, normal states’, and move into a different state of prefigurative values and behaviour. However, this process can degenerate into imposing equally problematic views of the ‘ideal’ WC member. Embracing cooperative values and behaviours more fully might seem like a solution, but it overlooks the ongoing need to unlearn outdated or utopian cooperative beliefs in evolving situations.

The need for ongoing unlearning further manifested in relational dynamics of ‘sharing the load’ through multi-tasking and taking on responsibilities. Members who worked across the organization seemed inclined to assume more responsibilities, a ‘chicken or egg’ issue, with ‘extra-curricular’ activities common in the four WCs. Financially stable or democratically advanced WCs often relied on the individual goodwill of members. At WC2, members multi-skilled across at least two teams, integrating functional authority and maintaining a consistent ‘worker identity’. However, not all members were actively involved with all their teams, leading to a lack of engagement, additional tasks, specific roles and broadening responsibilities.

According to WCM04-WC2, this discrepancy arose in part from activities deemed rota-able versus those done in ‘spare time’, with issues or roles only gaining members’ interest when sufficiently serious as to be allocated time. However, complexities also emerged here by way of individual limitations and structural inequalities persisting from members’ lived experience and circumstance beyond the cooperative. Additional work could not be ‘expected’ of all members, and willingness varied, fostering self and mutual exploitation and informal hierarchies. ‘It’s so easy to martyr yourself in many different ways in a job like this. I think the key to remaining happy and al tempo here is not to get into a mindset where you think, “If I don’t do this, no one else will”. . .’ (WCM07-WC2). Likewise, the demands of efficiency and the risk of real or perceived inequitable workload distribution manifested as contentious and potentially degenerative tensions. ‘I think it [responsibility] should be shared for a number of different reasons. . . shared for somebody’s own personal growth, shared for just having a fresh set of eyes on things. . . shared for not having the same people having the same burden for years. . . or the same amount of voice. . .’ (WCM21-WC3). While efforts were made to address such disparities through flexible work schemes, adapting to member needs and more, it simply was not possible to mitigate or negate all socio-structural inequalities. Thus, WCs fell short of prefigurative aspirations to offer a better approach to human economic activity and worker wellbeing than the exploitative proclivity of profit-driven businesses.

In sum, our findings reveal that discarding conventional business practices to make room for learning cooperative practices is insufficient for dealing with the tensions of democratic organizing. Democratic organizing does not come down to imposing or moving towards an ideal form, but asks for an ongoing process of negotiating assumptions, values and practices.

Vested interests and sacred cows

‘Vested interests’ were a major issue in the relational dynamics of decision-making across our cases. Vested interests appeared in value-based decisions, such as allocating charitable funds or setting policies on vegetarianism/veganism, nutrition, changes to sickness policies, and operating procedures. But they became particularly noticeable with big decisions such as relocating the business (WC4W), adjusting pay scales (WC4R) and reducing wages or dividends (WC3). These decisions have significant ramifications for members’ economic and social positions, eliciting defensive responses that created power dynamics adverse to solidarity in decision-making and structuration as a potential solution. As WCM03-WC2 explained, in consensus decision-making, a minority could effectively block changes:

No one will necessarily know why something was set up or who did it or whatever and it would have made sense when it was done [. . .] That reason is kind of lost over time, but everyone has kind of accepted it as being truth. [. . .] [A decision that was] made fifteen years ago could have been completely correct but now the understanding has changed or society has changed [. . .] They are sort of sacred cows really [. . .] They can be challenged or asked about certainly, but sometimes people don’t even want to go there really because they know when they did, it didn’t lead to anything [. . .]. we are in a consensus environment, you don’t need many people to be wedded to an idea to make it pointless actually pushing.

Members often avoided challenging vested interests, treating them as ontological truths or ‘sacred cows’, and worked around them. CMA3 also suggested a propensity for members to avoid direct conflict, leaving problems unsaid and without resolution, offsetting immediate tensions at the cost of longer-term cohesion – highlighting a clear need for unlearning.

While differing worldviews are key to democratic decision-making and adaptability (Greenwood, 1991), the awareness of difference and potential for conflict were heightened in our WCs because values and personal beliefs were strongly asserted by some members.

Those sorts of people have very strong views on things and some of those things are unsaid in that sense. I’ve come across that quite a lot with co-ops where they are perfectly legitimate reasons for conflict because people just have very different worldviews on whatever issue. . . which again you might not necessarily have in a normal business situation because people don’t talk about their values so much in a normal business. (CMA3)

WCs explicitly asserted members’ right to hold differing values and beliefs and sought to acknowledge these in progressing toward resolutions. However, members would sometimes get very defensive, thus sustaining complex power dynamics across lengths of tenure, extent of involvement in specific roles or across the cooperative, formal or informal groups, and the individual–collective relationship.

At WC4W, it was suggested that part of the challenge was developing and implementing codified materials to address relational dynamics. Members would often struggle to recognize that policies and procedures could facilitate rather than constrain individual autonomy by increasing the amount of certainty.

WCM37-WC4W believed that reducing the ‘greyness’ and ‘nail[ing] things down’, be this with regard to working conditions, contracts or any other aspect of general policy, would create a more stable and functioning working environment, particularly for new members. The example below relates to the issue of establishing a standardized sickness policy.

As HR it’s really important to just kind of nail things down. . . like, what do we want to do with sickness? And [other] policies. . . because people thought policies was like this real kind of like top-down structure and it’s actually not. . . it’s more democratic because you are all deciding what we do as opposed to HR which used to be really grey and it’s just like well, what did we used to do with this? Oh. . . what seems fair? And then you can’t help but bring in your own bias [. . .] it used to be just really really chaotic when you used to start, you didn’t have a clue what was going on. . . (WCM37-WC4W)

While this example highlights how the ‘tyranny of structurelessness’ (Freeman, 1972) might ring truer in established and scaled WCs than one might assume, it also reveals the need for unlearning engrained beliefs about democratic organizing. WCM37-WC4W suggested that this resistance came from a vocal minority of libertarian, laissez-faire members who interpret cooperatives as being about ‘being your own boss’.

Hence, without unlearning, WCs struggled to address relational dynamics of decision-making and avoid degenerating into vested interests and defensiveness.

A sense of exceptionalism

We discerned another form of an ontological truth and defensiveness that highlighted the need for unlearning: a sense of uniqueness turning from a ‘bulwark’ (CMA1) for prefigurative cohesion into a restrictive and isolating sense of exceptionalism. As purposive organizations, WCs are bound to the place, time, people and purpose they are created in, by and for. This foundational worldview took on a special role in the WCs because their prefigurative nature created challenges uncommon beyond democratic organizing.

I think it [a sense of uniqueness] is a strength in as much as to have that, particularly if you’re small, you’re isolated, you’re on your own, you’re up against it. . . then having that kind of very strong identity is a kind of bulwark against that. It maintains morale, it supports people’s engagement. (CMA1)

The recurrent belief among members that their organization was somehow special and uniquely different served to sustain WCs through their foundation and early years. It also shaped their ‘brand personality’ in the capitalist marketplace. For instance, WCM37-WC4W highlighted the importance of ‘holding on to “that thing” and not losing what you are’ to differentiate from competitors.

Such ‘holding on’ becomes problematic when it limits thinking, responsiveness to change and cooperative development. While WCM07-WC2 felt evangelical about spreading the cooperative mode of working and organizing, WCM34-WC4W mentioned struggling to maintain a clear shared understanding of values to be able to communicate to members and others. WCM10-WC3 highlighted the importance of autonomy, self-motivation and working for the collective good at WC3 but noted that this ‘sense of uniqueness’ was hard to ‘label’ and articulate. These insights suggest that WCs’ ‘sense of uniqueness’ is a loose set of ideas requiring sustained time and effort to comprehend, reassess and adapt to prevent becoming cult-like, driven by charismatic leadership with members sharing a ‘pious belief’ in its purpose and values (WCM07-WC2).

Interestingly, this ‘sense of uniqueness’ distinguished WCs not only from conventional businesses but also from other WCs. While not expressing difficulty accepting the ‘worker cooperative badge’ (CMA3), members could slip into a detrimental ‘sense of exceptionalism’:

Do you know how many times I’ve been told, when I’ve raised issues about what this co-op are doing and what that co-op are doing, [. . .] is really good, constantly, ‘Yeah, but they’ve got a different model to us’? Obviously, I’m constantly going back to, ‘Yeah, a model that works.’ (WCM24-WC3)

This internal notion of ‘model’ remained abstract and unclear, making it difficult to interact with, change or sustain. The very notion of being a WC can serve as a limiting conceptual boundary (CMA4) that turns a sense of exceptionalism into an ontological truth. This is illustrated by the limited participation of many potential WCs in the national cooperative movement due to political disagreement or resistance to identification (CMA3).

Amplifying a sense of exceptionalism can lead to isolation, exacerbated by not engaging with the movement or understanding cooperation. Even at WC2, the most progressive of the cooperatives studied, members developed such attachments, inhibiting their ability to balance a concrete (clear, coherent and accessible) or abstract (fluid, loose and changeable) shared understanding. Therefore, ongoing unlearning is needed to prevent a sense of uniqueness degenerating into a sense of exceptionalism as an ontological truth.

Relational Practices of Unlearning

Each of the three relational dynamics we identified revolves around structural-relational and cultural-cognitive tensions in democratic organizing. We demonstrated that unlearning is about more than letting go of incongruent beliefs and behaviours in favour of more cooperative ones. In answer to RQ2, the following analysis of three relational practices (enacted responses) further highlights that unlearning is an ongoing process of working through extant beliefs and behaviours in relation to others, the organization and the environment.

Integrating individuals

Most organizations and businesses need to integrate new people, and unlearning plays a key role in this process (Becker & Bish, 2021). For WCs, ‘integrating individuals’ revolves distinctly around the ongoing need to confront relational dynamics of ‘sharing the load’. CMA3 suggested that while WCs might struggle with conventional business activities, they excel in finding the ‘right kind of member’ and improving cooperative practice and relationships among all members. Integrating individuals involves fostering diverse views on what it means to be a member to break down preexisting assumptions and cooperative responses.

We found that established members ‘making space’ for newer members ‘to take what they discover in terms of culture and practice, own it, and bring it up to date’ (CMA1) was crucial for integration.

So, what you don’t want to do is have members that are just there for ever and almost sort of disempower the new people. . . so actually how alive those founders are to that potential is often quite important. . . [it works] where they are trying to give away power but in a way that they are empowering people as well. (CMA5)

Established members created structural space by leaving the WC or stepping away from roles, created cognitive space by mentoring new members, and cultural space by enabling them to challenge vested interests. However, while some members independently ‘evolved’, it required ongoing facilitation through induction training, probationary mentoring, appraisals and mutual support. Most WCs used these methods, but only WC2 fully engaged members in ongoing ‘training’ that challenges and reflects on working practices. Without mechanisms to reflect on and challenge accumulated positions and interests, unlearning is unlikely, leading to resentment and detachment.

WCs were also constantly unlearning how to handle tensions between the ‘worker-member’ identity and the balance between multi-skilling and specialization. Specialization was increasingly being adopted to address skill loss and complexity but posed a challenge for maintaining a sense of collective relevance. While the integrative potential of multi-skilling and task or role rotation was widely acknowledged across the WCs, only at WC2 was it effectively still practised. Instead, a negotiated balance was struck involving multi-skilling or rotation as part of onboarding, rotation being encouraged for positions of authority, and the presence of mechanisms for members to transition across tasks and roles. Member job descriptions and policies increasingly enabled specialized recruitment for high-skill functions, while still emphasizing cooperative practice. Ongoing unlearning and adaptation were necessary, as highlighted by WC1 shifting to specialist recruitment under pressure to compete for skills.

Recently, we have changed our recruitment practices very, very slightly to allow us to recruit for more specific roles [. . .] we then decided to recruit externally for design skills, but we didn’t just go out looking for designers, we went out looking for people who had design skills but also membership qualities. . . (WCM01-WC1)

Our analysis shows that unlearning supports WCs to integrate individuals by valuing diverse views on how to share the load over institutionalizing a fixed idea of the ideal member.

Reviewing and renewing

The main reason why co-ops fail and the issue with co-ops is conflict. . . it’s conflict mostly coming from a different worldview or not being clear on the worldview of what the co-op’s for, where it’s going, and so when conflict happens it’s because they never really dealt with that underlying, ‘Why are we here?’ sort of problem. (CMA3)

‘Reviewing and renewing’ involves regularly reflecting on and, if necessary, redefining what the organization is ‘about’, comparing espoused versus actual values. WCs engaged in collective interpretation and search for ‘what creates the unity’ (CMA1). On the one hand, this meant addressing a structural tension between clarity and consistency of purpose and identity versus being adaptive, responsive and inclusive entities. On the other hand, it was a developmental process particular to any WC. While renewing the purpose of a business is common in organizational development, doing so in a way that instils democratic values and capacities is a hallmark of WCs (Greenwood, 1991) and requires ongoing unlearning of what it means to be cooperative in relation to each other and the organization through processes of individual–collective alignment (Langmead, 2016).

In CMA1’s ‘design co-op’, the founding context had become obsolete, shifting from a specific political movement in the 1970s to a general set of values binding the membership. In contrast, WC1 maintained its purpose and core values over 40+ years despite significant growth. By ‘continually redefining what we’re about’ (CMA1), WCs could unlearn outdated core values and practices, avoiding vested interests and exceptionalism that undermine prefiguration.

Two crucial questions for ‘reviewing and renewing’ are: Who are we? How do we interact with one another? The second question translates abstract notions of identity and purpose into concrete behaviour. For example, a department at WC4R overcame difficulties by working with an external counsellor.

When I came here, it was just men. Really, people would scrabble for top dog and you were, like, ‘What?’ When I went on my first Co-op Weekend and people talked about hidden hierarchies, [. . .] it was a real eye-opener. But now we’ve got this guy coming in and [. . .] we have found [. . .] a collective vision, whereas in the past it’s been a lot of individuals just trying to get on. Yeah, what it’s created is the new dynamic where everybody is locked into it. What this counsellor has done is just open us up to potential possibilities, and looked at the things that don’t work, and the things that do work, and the vision for the future of it working. Whereas we’ve never really had that, we’ve never had a lot of space to discuss. What he’s given us are actually tools to work out how we can actually be more creative. If you’ve got certain targets to hit each day, sometimes ideas and creativity go out of the window. . . (WCM32-WC4R)

This story highlights key elements of unlearning: acknowledging the problem and voluntarily deciding to seek help, re-examining identity and purpose, finding solutions based on a renewed shared understanding, and enabling members to sustain their collective vision and self-manage their learning process. However, defensiveness emerged when expanding the programme organization-wide, with fears of behaviour change and financial impact on bonuses.

While ‘reviewing and renewing’ might seem like ‘quite a cultural shift’, CMA2 suggested that it’s not necessarily complicated. It’s about embedding joint reflection on espoused values and actual practices across the organization to unlearn undemocratic practices and ossified structures.

Turning outwards

While most organizations can benefit from actively engaging with their ecosystem, WCM32-’WC4R’s narrative highlighted the distinct value of ‘turning outwards’, looking beyond organizational boundaries, for WCs to unlearn their framing of relational dynamics and engrained practice. This was evident from attending the Worker Co-op Weekend, external cooperative training being a mandatory induction component at CMA1’s Design Co-op, and WC4R working with an external facilitator to resolve internal conflicts. CMA1 and CMA3 emphasized that the three externally facing cooperative principles (education, training and information; cooperation among cooperatives; and concern for community) are intrinsic to cooperation and help WCs evolve beyond their founding myths.

I think that the co-ops that transmit culture most effectively are also those co-ops that are part of a wider movement. . . it’s incredibly difficult to do it in one business. . . because it appears bizarre. . . for people who come from different walks of life into a worker co-op, it’s like a shock. . . it’s like a culture shock. . . nobody’s ever told them anything about this. . . but if they don’t know that this is actually part of something bigger, it will always remain slightly weird to them. . . and they will always find that they are kind of probably defending it. . . and on the back foot themselves about it. . . (CMA1)

‘Turning outwards’ helps overcome a sense of exceptionalism, addresses isolation tendencies and makes cooperation feel less ‘weird’ by connecting members to a larger movement. It provides tools for improving cooperation and business practices. CMA1 described this as a ‘kind of cooperative pedagogy’ giving members the tools to discover and ‘start to own’ cooperative practice for themselves rather than ‘force-feeding’ structures and practices. This underscores how unlearning transforms understandings of membership and cooperative skills.

So whether it’s through the formal training Co-ops UK or other people put on, the Worker Co-op Weekend, or local networking events. . . when we’re at a local networking event, people are having a beer, and they’re talking business. . . they’re talking about, ‘Oh how do you deal with this problem?’ and actually [. . .] that learning about how to run your worker co-op is really passed on from one generation to another, from one worker cooperative to another [. . .] in some sectors you do a qualification. . . if you’re an accountant or a whatever, you go to university, you do your qualification, you know how to do that job. . . whereas in a worker co-op there isn’t that. . . there isn’t that formal education. . . the education in being a worker co-op is much more informal. (CMA3)

CMA3 framed cooperative practice as an ongoing and evolving process of unlearning through sharing experience. While formal training and events provide a frame, the real value of ‘turning outwards’ lies in the conversations that occur around these activities, fostering a multi-generational community of practice. ‘Turning outwards’ is crucial for ongoing unlearning what a WC is and how to enact it.

Unlearning in Democratic Organizing: Three Propositions

Our empirical findings demonstrate that unlearning plays a significant role in confronting structural-relational and cognitive-cultural tensions between prefiguration and degeneration. WC members are constantly engaged in relational dynamics around sharing the load, vested interests and sacred cows, and a sense of exceptionalism. A common response was to stress the need to learn and embrace cooperative practice. However, institutionalizing normative ideals in structures, beliefs and practices detracted from integrating individuals, adapting organizational purpose and conduct, and connecting with the cooperative movement. Instead, through iterating between empirical insights and theory, our abductive analysis surfaced the insight that unlearning enables relational dynamics that prioritize diverse views and confront extant beliefs and behaviours in relation to others, organization and environment. Beyond discarding old and learning new values and behaviours, democratic organizing can be sustained through an ongoing, purposive process of unlearning in which members let go of engrained ways of thinking, acting and relating in response to emerging challenges (Becker, 2018; Hislop et al., 2014; Tsang, 2017).

When we started to conceptualize our findings in terms of unlearning, we found that it offered lots of potential but also limited clarity. Unlearning is commonly defined as an intentional discarding process of eliminating practices, knowledge and routines to make space for potential new ones (Klammer et al., 2025). The widely accepted view is that unlearning involves moving away from (or in some instances abandoning entirely) values, norms, assumptions, dominant logics, behaviours, practices, procedures or structures. ¹ This perspective resonates with our findings insofar as it claims that timely and effective unlearning may strengthen organizational abilities to adapt to internal tensions (Matsuo, 2019) or strategic threats (Morais-Storz & Nguyen, 2017) and integrate new individuals (Becker & Bish, 2021).

However, unlearning is widely debated due to the risk for ‘conceptual confusion and terminological ambiguity, hindering both empirical research and more practical applications within organizations’ (Visser, 2017, p. 54; see also Howells & Scholderer, 2016; Tsang, 2017). Three key topics of debate are the extent of intentionality versus the unintentionality of forgetting, value judgements made regarding discarded routines (beyond the obsolete and inferior, or improvements in organizational performance) and the relationship between unlearning and learning (they are not necessarily bound together) (Hislop et al., 2014; Tsang & Zahra, 2008). Developments toward a relational understanding of unlearning as a deep or transformative process have the potential to clarify these issues but lack conceptualization and empirical grounding (Becker, 2019; Klammer et al., 2025).

Therefore, we developed three analytical propositions to conceptually clarify how unlearning can sustain democratic organizing. We turned to Follett’s relational process ontology to conceptualize our overall finding that unlearning is best understood as an ongoing process of reciprocal interaction between interdependent parts (Stout & Love, 2015b, p. 464) and extended it with insights on defensiveness and organizational landscapes that helped to explain these sub-themes from a relational perspective. Follett’s work has been recognized as a valuable yet under-acknowledged philosophical grounding and substantive theory for organization studies (Weick, 1995), industrial democracy (Chen, 2016) and prefiguration (Maeckelbergh, 2009). Follett (2003f) directs our attention to the ‘integrative process’ through which we co-create each other and ‘the situation’. The situation is a dynamic and ever-evolving ‘whole a-making’ constituted by all interrelating actors and factors (Follett, 2003a, p. 67). Integration occurs when all those involved observe ‘the law of the situation’ and co-create a new outcome that is qualitatively better than what they all did or desired separately. This, Follett insists, will not happen based on good intentions or intellectual principles; ‘experience and our learning from it should be equally continuous matters’ (Follett, 2003d, p. 108). Accordingly, we present the following propositions.

Democratic organizing is sustained through ongoing relational processes of unlearning

Our first proposition is that unlearning can sustain democratic organizing when understood as an ongoing relational process. Drawing on Weick and Quinn’s (1999) distinction between continuous and episodic change and noting parallels with single- and double-loop learning (Argyris & Schön, 1978, p. 3), unlearning is typically understood as a binary of ongoing, gradual unlearning of practices, processes and behaviours, versus a more intentional or sudden process of deeper unlearning that affects values, norms and cognitive frameworks (Hislop et al., 2014; Rushmer & Davies, 2004; Tsang & Zahra, 2008). The latter is usually called ‘deep’ unlearning and framed as the result of a sudden, unexpected and potentially painful event or experience rupturing habitual patterns and engrained beliefs. However, the unlearning literature has moved toward understanding deep unlearning as a more gradual, developmental realization of significant changes to beliefs and routines in interaction with others (Akgün et al., 2007). Drawing upon transformative education theory, Macdonald’s (2002) notion of ‘transformative unlearning’ conceptualizes it as a relational process of becoming receptive to the possibility of alternative views, recognizing their value and the potential weaknesses of one’s own, and working through the discomfort that comes with letting go of engrained habits.

Our findings show that democratic organizing can be sustained through deep unlearning that is not episodic but ongoing and transformative. We saw how uncooperative attitudes and behaviours were not simply unlearned when new members joined a WC but spurred an ongoing process of confronting what Follett (2003e) calls ‘the habit-patterns’ (p. 25) formed through experience, training and socialization originating both outside and within the WCs. This turned out to be not a matter of mere experience with WCs or simply embracing cooperative values as unshakable normative ideals. Across all cases, the need for more established members to unlearn was insufficiently acknowledged – a failure to engage in ‘progressively creating slack’ (Varman & Chakrabarti, 2004, p. 204). Established members could disengage from ‘sharing the load’ of joint decision-making, uphold vested interests that became ontological truths, and refuse to ‘make space’ for new members – stopping to create ‘spaces of possibility’ for democratic organizing (Cornwell, 2012; Kioupkiolis, 2010; Kokkinidis, 2015). Breaking with habits, confronting conflicting interests and values, and changing modes of relating are all part of the ongoing relational process of unlearning that democratic organizing requires (Follett, 1998, p. 119).

Unlearning is a sensitive process of identity regulation that can lead to either emancipation or oppression of members (Alvesson & Willmott, 2002). Our findings on ‘reviewing and renewing’ showed that it is necessary to nurture a cooperative disposition in members and to evolve the organization over time to reflect and remain consistent with its membership. Similarly, our findings on ‘integrating individuals’ demonstrated that it is necessary to pursue cooperative practices of multi-skilling and collective decision-making, as well as recognize the benefits of bureaucratic practices such as clearly defined responsibilities and accountability (du Gay, 2000; du Gay & Vikkelsø, 2017). We echo relational arguments for individual–collective alignment in democratic organizing (Langmead, 2016). As Follett emphasized (Stout & Love, 2015a), integrating is a balanced, pragmatic and ongoing process.

Our findings under RQ2 highlighted that unlearning needs to be actively nurtured and supported. The WCs could not settle on what constitutes the right member, how to best integrate them into the organization, how to resolve tensions between specialization and multi-skilling, what the organization is about, how to enact their espoused values, and how to position themselves within the WC movement. Even where certain values or practices seemed congruent with cooperative principles, we found that WCs were more successful when they did not treat these as set in stone but recognized how changed circumstances and persistent challenges required them to unlearn their habits of mind and behaviour. For instance, responding to increasing membership diversity by ‘going back to what creates the unity’ (CMA1) to find renewed integration. Rather than turning situations into an either-or situation, such unlearning is most fruitful when seeing all parts in relation to each other and integrating the values of both sides into a new, emergent whole (Follett, 2003f).

Unlearning enables democratic organizing to work through defensiveness and overcome habitual relational patterns

Our second proposition is that unlearning sustains democratic organizing by enabling ongoing processes of working through defensiveness and overcoming habitual relational patterns. While Follett offered a ‘method of integration’ to facilitate such relational processes, she did not pay much attention to the defensiveness involved. However, the organizational learning literature has established that organizations often uphold a defensive mode of action and learning that suppresses openly sharing thoughts and feelings, confronting tensions and conflicts, and joint learning and change (Argyris & Schön, 1978). This is especially problematic when deep unlearning is restricted to episodic events, as individuals are likely to act defensively when faced with the psychological impact of an intense break with underlying assumptions or values. But even when it is more gradual, transformative unlearning is deeply challenging for those experiencing it, involving ‘the loss of prior ways of seeing – the loss of fundamental assumptions which until now had brought certainty and security’ (Macdonald, 2002, p. 174). Unlearning is an ongoing effort that is more than simply a cognitive process or delineated activity. It requires constant adjustments as well as systematic collective effort. Awareness of tensions and paradoxes is essential for navigating them – effectively developing relational practices and generating artefacts to manage complexity (M. Griffin et al., 2022).

Our findings highlighted the adverse impact of defensiveness and habitual relational patterns on democratic organizing by discussing the ‘sacred cows’ that were so hard to challenge that it was felt not to be worth the hassle of engaging in relational dynamics around vested interests. Macdonald (2002) argues that ‘transformative unlearning’ is a process of working through intense beliefs to change ‘the whole person’ in relation to others and their environment. Changing an entire ecological system of feelings, thoughts and behaviours – i.e. ‘the situation’ (Follett, 2003b, p. 21) – is a deeply challenging relational process of becoming receptive to alternative worldviews, recognizing the need to change and grieving the loss of engrained habits and beliefs. For instance, our findings on ‘sharing the load’ showed that some WCs had to come to terms with the uncooperative way multi-skilling was enacted, which challenged deeply engrained beliefs about its superiority over specialization (B. J. Griffin et al., 2019).

Unlearning makes it possible to work through defensiveness by creating conditions for mutual appreciation, positive regard and opening to the vulnerability of change. We identified relational practices of ‘making space’ for new members and collectively ‘reviewing and renewing’ organizational values and identity. The counselling story of WC4R illustrates how teams can become receptive to the need for unlearning counter-cooperative habit-patterns (e.g. hidden hierarchies) and transform their shared ways of relating, thinking and acting. It also illustrates the need to embed unlearning within and across democratic organizations to work through defensiveness and overcome habitual relational patterns engrained in the wider WC membership.

Creating systemic conditions for unlearning across levels of organizing enhances democratic resilience

Our third proposition is that the resilience of democratic organizing is enhanced by creating systemic conditions for unlearning across levels of organizing. The organizational response to the counselling experience in WC4R demonstrates that unlearning needs to happens across ‘levels’ of organizing that can’t be addressed in isolation – what Follet (2003e, p. 34) calls ‘the whole situation’. It requires a supportive environment that links individual, group, organization and external levels, through dedicated space and resources for joint reflection and training. While primarily focusing on organizations, unlearning literature has increasingly explored individual, group and organizational levels (Klammer & Gueldenberg, 2019). Unlearning is enabled by opportunities for individual reflection, group habit changes, and consolidating new understandings across the organization (Cegarra-Navarro & Wensley, 2019; Cegarra-Navarro et al., 2014).

Unlearning can be facilitated within a ‘community of practice,’ like a team, but is harder to spread across a ‘landscape of practice’ (Wenger-Trayner & Wenger-Trayner, 2015), which comprises multiple communities. We conceive of each WC as a landscape comprised of communities including departments, teams, working groups and coordinators, as well as informal groups with social and/or political affiliations. Such boundaries engender relational dynamics of both cohesion and tension by ‘scaffold[ing] our mindsets, emotions, and behaviours as we cope with paradoxes’ (Lewis & Smith, 2022, p. 534). Members move around the landscape to experience different communities and avoid over-specialization in recruitment and membership policies. However, WCs also operate within a wider landscape of practice – that of worker cooperation, which is in turn made up of further communities. Echoing Paton (1989) and Kleinman (1996) regarding the role of external resources in mitigating counter-democratic forces, by ‘turning outwards’ to the wider cooperative movement, WCs can check and improve their cooperative practice and reduce their sense of isolation or exceptionalism while maintaining their all-important sense of uniqueness or ‘institutional distinctiveness’ (Soetens et al., 2023, p. 363).

Conclusion

Our analysis of how democratic organizing is sustained in worker cooperatives (WCs) offers new empirical and conceptual insight into the role of unlearning. The article addresses a gap in understanding how prefigurative collectivist-democratic practices endure despite pressures of degeneration, particularly in larger, long-standing WCs. Our findings show that unlearning is a significant yet underacknowledged aspect of daily practice and essential for confronting tensions in democratic organizing.

While WCs have scope for nurturing prefigurative practices, they face constant threats of degeneration as they balance being both businesses and membership organizations. Research on learning in WCs and democratic organizing remains limited, which is problematic because, as our analysis demonstrates, sustaining democracy involves not only learning new ways of thinking, acting and relating, but also unlearning habitual beliefs and behaviours that become embedded as ontological truths. Doing things differently requires letting go of what came before.

We identified three relational dynamics – patterns of interaction that create the need for unlearning – and three relational practices – enacted responses to this need. In answer to RQ1, democratic organizing requires deep, ongoing unlearning because members struggle with tensions between cooperative ideals and tendencies toward disengagement, vested interests and exceptionalism. In answer to RQ2, these issues can be addressed by working through defensiveness about membership, organizational purpose and movement positioning.

Through abductive analysis, iterating between theory and data, we developed three analytical propositions. These posit that unlearning sustains democratic organizing when members create conditions for shared unlearning of habitual patterns and beliefs incongruent with democratic practice, supported by organizational resources. Across landscapes of practice (Wenger-Trayner & Wenger-Trayner, 2015), organizations must prioritize diverse views and integrate them into adaptive responses. By regularly revising artefacts and structures, democratic organizations can continue to unlearn and improve prefigurative practice.

This article is the first to examine how unlearning sustains democratic organizing, creating new conceptual and empirical synergies between literatures on unlearning and democratic organizing. Our conceptualization resonates with relational studies on prefigurative practice in ‘spaces of possibility’ (Cornwell, 2012; Kokkinidis, 2015), individual–collective alignment (Langmead, 2016), ‘antiperfectionism’ (Kioupkiolis, 2010) and ‘learning to live the paradox’ (M. Griffin et al., 2022). We advance the relational approach by explaining how unlearning enables democratic organizing to engage more sustainably with these processes.

We extend the relational approach to democratic organizing by acknowledging unlearning as crucial in adapting to emerging challenges and ongoing tensions (Byrne, 2022; Hernandez, 2006; Stryjan, 1994; Varman & Chakrabarti, 2004). As Ravn et al. (2023, p. 4) note, ‘There is no simple or single recipe for efforts to develop industrial democracy.’ Democratic organizing is a process of learning from difference that pivots on shared abilities for unlearning. While developed in the context of WCs, their commitment to collectivist-democratic principles, size and age has highlighted the significant role of unlearning in addressing the tensions of democratic organizing.

Our propositions also clarify and extend unlearning theory (Klammer & Gueldenberg, 2019). We conceptualize unlearning as intentional, ongoing and transformative, aligning with established views of organizational learning (Argyris & Schön, 1978; Klammer et al., 2025; Macdonald, 2002). Drawing on relational process ontology (Follett, 1998; Stout & Love, 2015b), we argue that replacing beliefs and behaviours is insufficient; sustaining democracy requires continuous adaptation of individuals, organizations and environments. We expand unlearning beyond routines and performance to include reshaping identities to uphold prefigurative ideals of fairness and democracy.

Our propositions have wider relevance to akin sub-fields, including employee-owned businesses, social enterprises, social innovations, social movements, the commons and participatory initiatives. Unlearning can be seen a meta-capability for confronting their inherent tensions, defensive behaviour and habitual relational patterns by facilitating the integration of difference in relation to others, organizations and environments. For instance, for social innovations and commons, unlearning offers a way to work through entrenched identities and governance routines, enabling adaptive responses to shifting political, ecological and cultural conditions (Bartels, 2023).

Our findings and propositions have clear practical implications. Democratic organizations typically have learning resources such as induction programmes, training, appraisals and mutual support. However, these initiatives often face issues such as de-prioritization, low member interest and high transaction costs (Cornforth et al., 1988; Ng & Ng, 2009; Rothschild & Whitt, 1986). Indeed, only at WC2 was a collective training programme successfully integrated. Our research suggests that unlearning should be prioritized to sustain democratic organizing but also appreciated as a process for (re)engaging members in learning and practising democracy. For unlearning to become a staple of democratic organizing and enable participants to adapt habit patterns and integrate diverse views into new prefigurative practices, we recommend that practitioners:

1. Stimulate and support ongoing unlearning to sustain democratic organizing,

2. Train and facilitate members to work through defensiveness and overcome habitual relational patterns, and

3. Embed unlearning across levels of democratic organizing.

Although our propositions may be less directly applicable to conventional businesses, lessons may be learned from democratic organizing for sustained organizational performance and survival (Berti & Pitelis, 2022).

We acknowledge limitations in time horizon, case number, access and UK context. A more explicit focus on unlearning from the outset may well have added further depth to our findings. Our propositions open avenues for deductive inquiry along with further qualitative exploration into the dynamics, practices and conditions that sustain democratic organizing, and call for cross-contextual research investigating how unlearning unfolds at varying scales and over time.