Why Organizations Struggle to Learn (Part I): Insights from Argyris and Schön

In today’s world, almost every organization claims to value learning. And to be fair, most of them probably do recognize its importance since adapting to rapidly changing conditions and demands of today’s environment has become an absolute necessity in a competitive market. 

In line with this, we see more and more companies investing in training programs, running feedback cycles, conducting performance reviews, and constantly talking about continuous improvement.

But the real question is this: Do all these efforts actually make the organization more innovative, more resilient, more capable of reaching its strategic goals? Or do similar problems quietly resurface, while the same strategic miscalculations are repeated in slightly different forms?

More broadly, what does it actually mean for an organization to learn?

Even at the individual level, identifying the signs of learning and measuring change is already complex. When we shift to the collective level, to something as layered and dynamic as an organization, that complexity increases even further. 

In this article, we will try to unpack this question. As a starting point, we will go back several decades and revisit the influential theory of Organizational Learning developed by Argyris and Schön (1)

Organizational Learning: Argyris and Schön

In much of the literature on learning in organizations, the core challenge has been this: how do we translate something we usually understand as an individual process into a collective, organizational one?

Learning is typically seen as something that happens within a person. We think of individuals constructing knowledge, testing assumptions, and revising their understanding. Argyris and Schön tried to extend this logic to organizations. Drawing on what they call a “theory of action” perspective, they argued that both individuals and organizations operate based on underlying action logics that guide how knowledge is constructed, tested, and restructured.

In doing so, they distinguish between two types:

Espoused theories: what individuals or organizations say they believe.

Theories-in-use: the assumptions and norms that actually shape behavior.

However, as with people, the two do not always align.

For instance, an organization may declare that it encourages open dialogue. That is the espoused theory. But if employees who challenge decisions are subtly discouraged, the theory-in-use instead is: avoid questioning authority.

This gap is not necessarily intentional. Often, organizations genuinely believe in their stated values. But when routines, incentives, and informal reactions send a different message, theories-in-use tend to dominate. And it is within this gap that learning, or the lack of it, becomes visible.

Learning Through Error

According to Argyris and Schön, learning begins with error detection. An error signals a discrepancy between expected outcomes and actual results and members choose to investigate that gap rather than ignore it.

An error, in this sense, is not simply a mistake. It is a signal. It points to a mismatch between a strategy and its outcome. The crucial question then becomes: At what level do members of the organization respond to this mismatch?

Single-Loop Learning

Single-loop learning focuses on adjusting actions without questioning the underlying assumptions. The guiding question is: How can we fix this?

If a project fails, timelines are tightened. If performance drops, new metrics are introduced. The strategy is modified, but the logic behind it stays the same.

Double-Loop Learning

Double-loop learning goes further. 

Instead of asking only how to fix the action, the organization asks: Why were we doing it this way in the first place? This involves questioning underlying norms, values, and mental models. It is not just behavioral correction, but cognitive reconsideration. Because it challenges foundational assumptions, double-loop learning has the potential to create more fundamental and longer-lasting change.

Deutero-Learning

Finally, deutero-learning refers to “learning to learn.” It involves reflecting on how the organization typically responds to problems. Does it allow questioning? Does it encourage reflection? Or does it default to defensiveness?

Deutero-learning strengthens the skills and mindset required for ongoing self-directed learning, making both single- and double-loop learning more likely in the future.

This level strengthens the organization’s capacity for ongoing adaptation.

In short, according to Agrys and Schön, organizational learning is not simply about adding new training programs or refining procedures. It is about how organizations respond to discrepancies, whether they adjust surface-level strategies or confront the deeper assumptions that shape them.

When Individual Insight Is Not Enough

Argyris and Schön ground organizational learning in the learning processes of individual members. 

However, they are also clear about one important point: organizational learning is not simply the sum of individual learning processes.

An organization does not automatically learn just because its individuals do.

For learning to become organizational, the knowledge gained through error correction must move beyond the person who discovered it. It needs to travel. It must move beyond the individual and become part of how the organization as a whole approaches future problems. Only then can the organization adapt more effectively to change and sustain learning over time.

Otherwise, insights remain isolated within teams or individuals and disappear when roles change or people leave. 

Illustrative Example ▾

Imagine a company where different managers give slightly different instructions. Priorities shift quickly. When something goes wrong, the response is often: “You should have clarified.”

Over time, employees adapt. One starts documenting everything in writing. Another avoids taking initiative unless instructions are extremely explicit. Someone else lowers their involvement to reduce risk.

Learning is happening here. But it is happening at the individual level. Employees adjust their behavior based on experience. They become more cautious, more strategic, more self-protective.

As a result, new employees enter the system and encounter the same uncertainty. The same errors reappear, and the same adaptations repeat themselves.

Reconsidering the Issue What could be done differently? ▾

From an Argyris and Schön perspective, the core problem in this scenario is not that employees are careless. The issue lies in the level at which the error is being addressed.

In this case, the organization is likely operating within single-loop learning.

An error occurs. The response is: “Next time, clarify better.” Behavior is adjusted, but the system that produced the confusion remains untouched.

According to Organizational Learning theory, the situation should have been treated as a structural signal rather than an individual shortcoming.

The question should not be “Who failed to clarify?” but “Why does this ambiguity keep recurring?” That is where double-loop learning begins.

The organization would need to examine underlying assumptions and norms:

– Are decision rights clearly defined?

– Do managers share a common communication standard?

– How are shifting priorities formally handled?

– Why do employees feel the need to protect themselves instead of acting confidently?

This shifts the focus from correcting behavior to questioning the mental models and structures that shape behavior.

At an even deeper level, deutero-learning would require the organization to reflect on how it typically responds to errors in general:

– Do we create space for open reflection?

– Or do we personalize mistakes and move on?

For Argyris and Schön, real organizational learning is not about producing more cautious employees. It is about making underlying assumptions visible and revisable.

For this to happen, certain conditions matter. Inquiry must be encouraged. Transparency, trust, and open communication need to exist not only as espoused values but as theories-in-use. Because only then can ideas circulate freely and assumptions be challenged without fear. Without these conditions, learning may still occur at the individual level, but it struggles to become collective.

Limitations and Critiques

Argyris and Schön’s work has been highly influential in the field of organizational learning. At the same time, it has not been without criticism.

One important critique is that their model may underestimate how much learning is shaped by broader organizational realities(2). Power dynamics, hierarchy, culture, and informal norms actively shape what is allowed to be questioned and what is not.

For example, some scholars argue that it is often managers, not employees, who determine what counts as “learning” in the first place(3). They decide which problems deserve attention, which assumptions are open to discussion, and which topics remain untouched. In that sense, learning is not always an open, collective process. 

Others point out that the distinction between espoused theories and theories-in-use, while useful, may oversimplify real-world situations(4). Not all behavior reflects deeply embedded assumptions; sometimes it reflects uncertainty or lack of experience.

These critiques do not necessarily invalidate Argyris and Schön’s contribution. On the contrary, their work was highly influential precisely because it provided a structured way to think about learning beyond training programs and surface-level adjustments and opened an important conversation.

At the same time, the critiques remind us that learning in organizations is rarely a clean or linear process. It is shaped by context, authority, culture, and complexity. And the organizational learning is both a cognitive process and a political one. 

References

(1) Argyris, C., & Schön, D. A. (1978). Organizational Learning: A Theory of Action Perspective. Addison-Wesley.

(2) Greenwood, J. (1993). Reflective practice: a critique of the work of Argyris and Schön. Journal of Advanced Nursing, 18(8), 1183–1187.

(3) Örtenblad, A. (2002). Organizational Learning: A Radical Perspective. Social Science Research Network.

(4) Sun, P. Y., & Scott, J. L. (2003). Exploring the divide – organizational learning and learning organization. The Learning Organization, 10(4), 202–215.