Publishing Disclaimer: In all of its publications and products, Military Review presents professional information. However, the views expressed therein are those of the authors and are not necessarily those of the Army University, the Department of the Army, or any other agency of the U.S. government.

---

 

Leading by Letting Go

Integrating Organizational Unlearning and Adaptive Leadership

 

Maj. Ben Garlick, U.S. Army

 

Download the PDF

 

There are over two thousand years of experience to tell us that the only thing harder than getting a new idea into the military mind is to get an old one out.

—B. H. Liddell Hart, Thoughts on War

 

In their article “Adaptive Leadership: The Leader’s Advantage,” Bill McCollum and Kevin Shea provide a compelling argument that “adaptive problems require leaders comfortable with leading and making decisions in highly complex environments” who can “explore new behaviors and lead differently, personally learn and develop learning organizations … and exercise patience when addressing complexity.”¹ While McCollum and Shea cite the central role of organizational learning to adaptive leadership, over the past several decades, an increasing number of researchers, academics, and practitioners across multiple fields have investigated organizational unlearning as an essential component of leadership.² If organizational learning is “the process of gaining collective knowledge,” organizational unlearning is the process of consciously shedding outdated and restrictive models or paradigms that inhibit the organization’s learning capability.³ Organizational unlearning is an essential but often overlooked component of adaptive leadership. Without the ability to deliberately shed ineffective mental models and routines, leaders and organizations struggle to fully adapt to complex challenges.

McCollum and Shea’s discussion of adaptive leadership relies on David Snowden and Mary Boone’s Cynefin sense-making and decision framework. This framework sorts problems into five contexts with corresponding appropriate leadership responses: simple, complicated, complex, chaotic, and disorder. Simple contexts feature clear cause and effect and require leaders to sense, categorize, and respond using established best practice.⁴ Complicated contexts still feature a clear relationship between cause and effect; however, it is not readily available to all observers and there may be multiple right answers. Leaders must sense, analyze, and respond, employing expert knowledge to sort through multiple “good practices” to find the optimal solution.⁵ Complex contexts, meanwhile, feature challenges outside past experience without a recognizable pattern. Faced with complexity, leaders probe first, sense the impact, and then respond based on an iteratively increasing understanding of emergent patterns. In a chaotic context, there is no discernable cause-and-effect relationship and high turbulence, requiring leaders to act first to establish order, sense the impact, and then respond to transform from chaos to complexity.⁶ Disorder occurs when it is unclear which of the four other contexts applies, and leaders cannot make sense of the events.⁷

McCollum and Shea combine the Cynefin framework with Ronald Heifetz, Alexander Grashow, and Marty Linsky’s bins of adaptive versus technical challenges to characterize the adaptive leadership environment. Adaptive challenges are those “where solutions lie outside the current way of operating, where the gap between the desired state and reality cannot be closed using existing approaches alone.”⁸ In contrast, technical challenges are “those where solutions lie within the current ways of operating, current expertise is sufficient, authoritative decision-making and standard operating procedures suffice, and culturally informed behaviors are not challenged.”⁹ Adaptive leadership is required to help organizations face challenges in the adaptive contexts of complexity and chaos, in contrast to the technical challenges of complicated and simple contexts.

McCollum and Shea identify six leader behaviors for addressing adaptive problems: get on the balcony (maintain big-picture awareness), identify adaptive challenges, regulate distress, maintain disciplined attention on the adaptive problem, give the work back to the people (empower subordinates to solve adaptive problems), and protect nontraditional voices from below.¹⁰ These behaviors enable the necessary space for organizational learning and experimentation to solve complex adaptive problems. Although McCollum and Shea note that hierarchical organizations like the military may be unreceptive to new learning due to well-entrenched cultures, their conception stops short of identifying the vital role of organizational unlearning in countering this tendency and enhancing an organization’s learning capability in complex contexts.¹¹

The earliest application of unlearning occurred in the field of education.¹² In one early conception, Alvin Toffler suggested that, given the acceleration of the pace of technology and mechanization, “knowledge will become increasingly perishable. Today’s ‘fact’ becomes tomorrow’s ‘misinformation’ … Students must learn how to discard old ideas, how and when to replace them.”¹³ The application of unlearning to organizations began in the mid-1970s with a study that developed a three-stage model of Swedish businesses struggling to avoid bankruptcy in the face of environmental changes. In the initial stage, past learning caused these organizations to incorrectly reinforce successful routines and models, consuming significant resources before managers realized the nature of the threat. Organizational unlearning occurred when these managers took drastic measures to challenge the status quo. If they overcame resistance, they entered a third stage of rediscovery and regeneration.¹⁴ From 1976 to 2019, leading management journals published 127 articles exploring organizational unlearning across various disciplines including crisis management, organizational change, technology implementation, and knowledge management.¹⁵ While there are numerous definitions and nuanced debates on the nature of organizational unlearning, a recent bibliometric analysis developed the following unified “connotation of organizational unlearning”: “An intentional process, whereby organizations question, identify and discard obsolete knowledge, routines, beliefs, or behaviors in order to acquire new knowledge and behaviors.”¹⁶

Before approaching the process of organizational unlearning, it is helpful to differentiate several terms: single-loop learning, double-loop learning, unlearning, and forgetting. In single-loop learning, an actor evaluates a parameter based on a decision rule or standard and then makes a decision to produce a change in this parameter. The actor takes feedback from this result and compares it against the decision rule, making another decision on whether to change the system that restarts the loop.¹⁷ This same decision-result-feedback loop exists in double-loop learning; however, this time the feedback also enters another loop where the actor uses the information to assess the assumptions, norms, policies, and goals that feed the decision rule or standard itself.¹⁸ As a result, the actor revises these rules over time to move the system to a more optimal state. While forgetting is the unintentional loss of knowledge or behaviors, unlearning is an intentional, conscious, and selective process.¹⁹

Some writers have argued that organizational unlearning as a separate concept is unnecessary and duplicative of the second, deeper loop of evaluating assumptions, norms, and policies in double-loop learning.²⁰ Others have argued that organizational unlearning is a subsumable process under organizational learning or part of a cycle of learning—unlearning—and relearning. More recent scholarship, however, has demonstrated that organizational learning and unlearning are separate processes.²¹ Based on this argument, organizations can unlearn without the requirement to replace it with other knowledge through learning, along with the inverse that all learning does not require unlearning first. In this interpretation, unlearning is a separate but complementary process to learning, enabling organizations to overcome the inertia and resistance to change particularly inherent in hierarchical organizations. Through double-loop learning, an organization asks, “Why do we do things this way?”²² In contrast, during unlearning, an organization asks, “What do we need to get rid of? What is blocking us from succeeding in this environment?” Unlearning expands the decision space in which deeper levels of learning can occur by removing potential constraints. A rigorous approach to organizational unlearning helps an organization face the “element of emotional threat, stress, and anxiety” that comes with removing deeply held assumptions, norms, procedures, and paradigms.²³

The evolving nature of scholarly research on organizational unlearning and the broad spectrum of fields to which it has been applied means there is no universally agreed upon unlearning model in literature. While the debate on these models is outside the scope of this paper, several examples are illustrative. John-Arild Johannessen and Arnulf Hauan provide a four-step process grounded in change management. These steps include understanding undesirable behaviors under new circumstances, filtering out individuals or groups that advocate these unwanted behaviors, highlighting desired/undesired behavioral patterns depending on particular contexts, and expressing the desired behaviors by initiating communicative or symbolic acts.²⁴ Mohammad Rezazade Mehrizi and Mojtaba Lashkarbolouki offer their own four-step process in the context of struggling business models: realizing there is a problem with the business model, revitalizing the struggling business model, parallelizing a new model alongside the old, and finally marginalizing the failing model in favor of the new.²⁵ Finally, Mark Boncheck provides a simple three-step framework nesting organizational unlearning in the learning process: recognize the old model is no longer relevant or effective, find or create a new model, and ingrain the new mental habits.²⁶ The figure presents a proposed model for organizational unlearning based on the approaches above and the relationship between this model and the organizational learning process as defined by Trent Lythgoe.

In this framework, members of the organization experience or observe events in the environment. At either the individual or subgroup level, actors realize that the organization’s existing models, paradigms, frames, or processes may be outdated, ineffective, or interfere with the organization’s ability to successfully adapt to changes in the environment. This can occur as a bottom-up phenomenon where those closest to the problem identify these issues, or as a top-down activity where leaders observe system-wide challenges.²⁷ These actors then assess and evaluate the factors above to determine what existing knowledge is specifically limiting the organization. Next, actors take proactive steps to deinstitutionalize these models, paradigms, or processes. This can also occur through grassroots means, where those closest to the problem make changes and disseminate them outward, but this process may fail to reach the organizational level. In contrast, whether originally identified from the bottom or the top, organizational leaders can impose more drastic and sweeping deinstitutionalization measures across the organization. Finally, these actors decide whether new knowledge should take its place. If so, the process transitions to organizational learning. If not, the process returns to experience or observation. While this characterization of the unlearning process implies a linear and sequential relationship, elements of deinstitutionalization and subsequent learning will often occur simultaneously and iteratively, as they do in the model proposed by Rezazade and Lashkarbolouki.

Returning to adaptive leadership, the unlearning process is an essential component of leading in complex adaptive environments. In simple and complicated contexts, the clear cause-and-effect relationships make it comparatively easy to determine when outdated knowledge is present and unlearning is necessary. In the complex context, however, this is never clear. Sometimes the solution is additive, building upon existing knowledge through learning. Other times, the solution requires subtraction in the form of unlearning. In many cases, a combination of the two is necessary. As McCollum and Shea point out, “Adaptive leadership, by necessity, requires a tolerance to mistakes.”²⁸ Leaders and organizations must be allowed to “get it wrong” as part of the probe, sense, respond process. There are other significant barriers to the organizational unlearning process in an adaptive environment: past successes and competency traps, hierarchical structures, biases and the incompatibility between new and existing knowledge, defensive routines, and security from psychological threats.²⁹ Of these, Rod Thornton identifies a “zero-defects” mentality driven by faith in past successes and increasing technological supremacy as a major barrier for the Army.³⁰ He also notes that “military organizations are not like other organizations” when it comes to unlearning, and unlearning in the military context is more difficult due to the intermittent pressure exerted by periodic battlefield experiences instead of the continuous pressure for survival facing businesses.³¹ Thornton’s perspective implies that the Army’s risk-averse, zero-defects mentality in peacetime undercuts its ability to adapt and evolve through organizational unlearning, wasting the opportunity to proactively challenge past knowledge until the realities of failure or setbacks become apparent on the battlefield.

To operationalize organizational unlearning as a component of adaptive leadership and overcome these barriers, McCollum and Shea’s list of adaptive behaviors and activities must be expanded. In addition to the current six, leaders must promote a climate of inquiry and disruption. This goes beyond the current prescription to “protect leadership voices from below” to the creation of an environment where challenging the status quo is both valued and acted upon. Leaders must establish the psychological safety necessary for challenging existing models, paradigms, and processes. A critical component of this is recognizing the value of different perspectives, backgrounds, and worldviews for challenging assumptions and breaking out of the sense of security and comfort provided by past precedent. Leaders must establish dedicated time and space in the organization’s processes to foster the reflection and experimentation necessary to accomplish this goal. By doing so, leaders can create an environment where unlearning is not just a reactive response to crisis but is instead a proactive and ongoing process of adapting to changing environments.

A simple example demonstrates the significant challenges inherent in organizational unlearning of embedded processes in a hierarchical organization. In 2018, then–Secretary of the Army Mark Esper eliminated several Army online training requirements and associated administrative procedures.³² While the Army’s intent was to not replace these requirements, soldiers soon reported unit-level substitutes including slides, typed forms, or handwritten coversheets involving the same information. The Army and the Department of Defense have since announced multiple initiatives aimed at eliminating similar administrative requirements.³³ Despite identifying these processes as outdated and ineffective, they persist in the Army’s institutional memory.

Compared to the simple example above, organizational unlearning of deeper paradigms and models represents an increase in orders of magnitude of difficulty. At the start of the British Army’s thirty-eight-year counterinsurgency effort in Northern Ireland, many units failed to unlearn the models they employed in colonial counterinsurgency campaigns in Malaya, Aden, and Kenya.³⁴ The operational context of these interventions permitted methods that were ineffective for operations inside the United Kingdom. This would lead to the single most damaging mistake of the campaign, which became known as “Bloody Sunday.” 1st Battalion, Parachute Regiment (1 PARA), known for their aggression and liberal use of force against protestors, operated under a model one writer characterized as “positive action and maintaining momentum into a crowd.”³⁵ Implicit in this model was the paradigm that 1 PARA were “warfighters” not peacekeepers or riot police. On 13 January 1972, 1 PARA’s “positive action” would carry them to the point of no return, precipitating a clash that left fourteen people dead under controversial circumstances and produced immense blowback against the British among the population and the international community.³⁶

While some commentators fault the decision to use the PARAs for counterinsurgency operations for the incident, the leadership of 1 PARA failed to adapt to their changing operational environment once that decision occurred. Specifically, the organization failed to unlearn its contextually ineffective models of counterinsurgency or challenge the “warfighter” paradigm. This limited the learning capability of the organization and shrunk the decision space available for innovation and adaptation, as old knowledge effectively blocked the adoption of other solutions. Many U.S. Army units faced similar challenges in shedding outdated models and paradigms during counterinsurgency operations in Iraq and Afghanistan.³⁷

The example of the evolution of Task Force 714’s counterterrorism operations in Iraq demonstrates an exception to this pattern. Then–Maj. Gen. Stanley McChrystal and his unit incorporated organizational unlearning in response to the complex adaptive problem of dismantling al-Qaida in Iraq (AQI). When McChrystal assumed command in October 2003, he inherited a task force extremely proficient at conducting raids but ill-suited for the challenge of defeating a networked enemy.³⁸ The task force possessed a siloed and hierarchical structure, with teams of operators at outstations charged with executing raids separated from the intelligence analysts, interrogators, and intelligence, surveillance, and reconnaissance assets charged with developing targets. The dominant paradigm was that the “finish” operations reigned supreme over the “find” and “fix” phases of the targeting cycle, and each specialized component of the task force worked in isolation to accomplish their tasks.³⁹ McChrystal, observing this disconnect, became determined to overhaul the task force’s operations through the alternate paradigm “it takes a network to defeat a network.”⁴⁰

Unlearning the old paradigm required unlearning several related operational models and procedures. The task force shed its previous “need-to-know” model of intelligence sharing, making room for a “need-to-share” model that prioritized flat communication across the organization. The organization eliminated its previous hierarchical and sequential “find–fix–finish” targeting cycle and replaced it with a “find–fix–finish–exploit–analyze” cycle.⁴¹ The exploit and analyze components became the main effort, encouraging coordination between teams and enablers to act rapidly on real-time intelligence. Operators unlearned their singular role as the finish force, enabling them to integrate across the targeting cycle by directing intelligence, surveillance, and reconnaissance feeds side by side with analysts, debriefing target information alongside interrogation teams, and rapidly delivering captured enemy material for exploitation.⁴² McChrystal and his staff employed a key deinstitutionalization method to extinguish the old model through a daily operations-and-intelligence briefing, bringing together forward teams, the headquarters and enablers, and interagency partners. This venue allowed McChrystal to synchronize efforts and priorities, provide guidance, and rapidly approve proposals, ensuring the task force adhered to the new ways of doing business while encouraging “initiative and free thinking, liberating subordinates to act without hesitation.”⁴³ This interplay between effective organizational unlearning and subsequent organizational learning allowed McChrystal and Task Force 714 to transform from conducting ten operations a month to ten operations a night by the end of 2005, eliminate the AQI leader Abu Musab al-Zarqawi in 2006, and fracture AQI as an organization by 2010.⁴⁴

As today’s Army adapts to a dynamic environment requiring proficiency across the continuum of competition, crisis, and large-scale combat operations, an intentional and proactive approach to organizational unlearning is more important than ever. In this environment, units may be called upon to shift rapidly between different complex contexts and manage the integration of old and new knowledge, procedures, models, and paradigms. The old refrain “if it ain’t broke, don’t fix it” is antithetical to success in this environment. To remain adaptive, leaders must intentionally and thoughtfully remove impediments to organizational learning capacity through organizational unlearning and the leader behavior of promoting a climate of inquiry and disruption. Fostering this behavior will allow organizations to proactively adapt to complexity through the probe, sense, and respond approach, while neglecting this behavior will result in stagnation, inefficiency, or failure.

 

---

Notes

• Epigraph. B. H. Liddel Hart, Thoughts on War (Faber & Faber, 1944), 115.

1. Bill McCollum and Kevin Shea, “Adaptive Leadership: The Leader’s Advantage,” InterAgency Journal 9, no. 1 (2018): 108, https://thesimonscenter.org/featured-articles/featured-article-adaptive-leadership/.
2. Shubham Sharma and Usha Lenka, “On the Shoulders of Giants: Uncovering Key Themes of Organizational Unlearning Research in Mainstream Management Journals,” Review of Managerial Science 16, no. 6 (2022): 1600, https://doi.org/10.1007/s11846-021-00492-7.
3. Definition of organizational learning from Trent J. Lythgoe, “Future Proof: How to Build a Learning Organization,” in L100: Developing Organizations and Leaders (U.S. Army Command and General Staff College, 2024) 2; author’s definition of organizational unlearning adapted from Mark Bonchek, “Why the Problem with Learning Is Unlearning,” Harvard Business Review, 3 November 2016, 3.
4. David J. Snowden and Mary E. Boone, “A Leader’s Framework for Decision Making,” Harvard Business Review 85, no. 11 (November 2007): 70.
5. Snowden and Boone, “A Leader’s Framework for Decision Making,” 71.
6. Snowden and Boone, “A Leader’s Framework for Decision Making,” 74.
7. McCollum and Shea, “Adaptive Leadership,” 102.
8. McCollum and Shea, “Adaptive Leadership,” 100.
9. McCollum and Shea, “Adaptive Leadership,” 100.
10. McCollum and Shea, “Adaptive Leadership,” 106–8.
11. McCollum and Shea, “Adaptive Leadership,” 103.
12. Sharma and Lenka, “On the Shoulders of Giants,” 1603.
13. Alvin Toffler, Future Shock (Bantam Books, 1990), 414.
14. Sharma and Lenka, “On the Shoulders of Giants,” 1604.
15. Sharma and Lenka, “On the Shoulders of Giants,” 1618.
16. Sharma and Lenka, “On the Shoulders of Giants,” 1604.
17. Lythgoe, “Future Proof,” 3.
18. Lythgoe, “Future Proof,” 4.
19. Lythgoe, “Future Proof,” 4.
20. Max Visser, “Learning and Unlearning: A Conceptual Note,” The Learning Organization 24, no. 1 (2017): 50, https://doi.org/10.1108/TLO-10-2016-0070.
21. Sharma and Lenka, “On the Shoulders of Giants,” 1603.
22. Lythgoe, “Future Proof,” 4.
23. Visser, “Learning and Unlearning,” 53.
24. Adrian Klammer and Stefan Gueldenberg, “Unlearning and Forgetting in Organizations: A Systematic Review of Literature,” Journal of Knowledge Management 23, no. 5 (2019): 868, https://doi.org/10.1108/JKM-05-2018-0277.
25. Mohammad Rezazade Mehrizi and Mojtaba Lashkarbolouki, “Unlearning Troubled Business Models: From Realization to Marginalization,” Long Range Planning 49, no. 3 (2016): 298–323, https://doi.org/10.1016/j.lrp.2015.12.005.
26. Boncheck, “Why the Problem with Learning Is Unlearning,” 4.
27. Klammer and Gueldenberg, “Unlearning and Forgetting in Organizations,” 866.
28. McCollum and Shea, “Adaptive Leadership,” 104.
29. Sharma and Lenka, “On the Shoulders of Giants,” 1667.
30. Rod Thornton, “Cultural Barriers to Organisational Unlearning: The US Army, the ‘Zero‐defects’ Culture and Operations in the Post‐cold War World,” Small Wars & Insurgencies 11, no. 3 (2000): 139–40, 143–44, https://doi.org/10.1080/09592310008423292.
31. Thornton, “Cultural Barriers to Organisational Unlearning,” 148–49.
32. Meghann Myers, “The Army Just Dumped a Bunch of Mandatory Training to Free Up Soldiers’ Time,” Army Times, 14 April 2018, https://www.armytimes.com/news/your-army/2018/04/24/the-army-just-dumped-a-bunch-of-mandatory-training-to-free-up-soldiers-time/.
33. Todd South, “Army Cuts Hundreds of Hours of Redundant Online Training Requirements,” Army Times, 15 May 2024, https://www.armytimes.com/news/your-army/2024/05/15/army-cuts-hundreds-of-hours-of-redundant-online-training-requirements/; Patty Nieberg, “Here Is the Training That the Army Says Is No Longer Mandatory,” Task and Purpose, 1 April 2025, https://taskandpurpose.com/news/army-training-changes-optional/.
34. Chief of the General Staff, Operation Banner: An Analysis of Military Operations in Northern Ireland (UK Ministry of Defence, July 2006), 3-1.
35. Rod Thornton, “Getting It Wrong: The Crucial Mistakes Made in the Early Stages of the British Army’s Deployment to Northern Ireland (August 1969 to March 1972),” Journal of Strategic Studies 30, no. 1 (2007): 98–99, https://doi.org/10.1080/01402390701210848.
36. Thornton, “Getting It Wrong,” 99–100.
37. Raffi Khatchadourian, “The Kill Company,” The New Yorker, 6 July 2009, 59, https://www.newyorker.com/magazine/2009/07/06/the-kill-company.
38. Liam Collins, “Dismantling Al-Qaida in Iraq,” in Routledge Handbook of U.S. Counterterrorism and Irregular Warfare Operations, ed. Michael A. Sheehan et al. (Routledge, 2021), 267.
39. Collins, “Dismantling Al-Qaida in Iraq,” 267.
40. Collins, “Dismantling Al-Qaida in Iraq,” 267.
41. Collins, “Dismantling Al-Qaida in Iraq,” 269.
42. Collins, “Dismantling Al-Qaida in Iraq,” 267.
43. Collins, “Dismantling Al-Qaida in Iraq,” 270.
44. Collins, “Dismantling Al-Qaida in Iraq,” 277.

 

 

 

Back to Top

January-February 2026