Double Loop Learning

Action and learning.

Action is fundamental to individual and organizational lives. People act to produce intended consequences, and they typically express their actions in conversations. Invariably, however, actions produce unintended consequences, particularly in difficult situations. Most learning occurs when people detect and correct mismatches or gaps between the intended and unintended consequences of their actions. This simple process, although prevalent, produces learning that is typically flawed because of hidden designs people hold without being mindful of their limiting effects.

The designs beneath.

People act with two types of theories in their minds. The first, called “espoused theory,” helps them proclaim to the world what they ought to be saying, believing, or espousing. The second, called “theory in use,” is more influential because it informs what people actually do, regardless of their external claims. From examining over 10,000 individual and organizational cases, Argyris has found that the theory in use carries the same basic design across different situations, cultures, races, genders, ages, social statuses, and so on, although manifestations may vary. Notice the emerging promise here: By uncovering the principal structure of the theory in use, we can suggest changes to make learning and action more productive.

The overall structure of the hidden design goes as follows. A set of well-entrenched governing values informs the theory in use, which influences the action strategies people use to conduct their lives, and most learning occurs from detecting and correcting the gap between the intended and unintended consequences of these actions.

Components of the theory in use that describe how action is actually implemented are called Model I, which is associated with single loop learning (SLL). Components of the theory in use that prescribe how action should be implemented are called Model II, which promotes DLL. Clients should learn to surface and be aware of their use of Model I and its SLL before they are coached toward implementing Model II and its productive DLL. The broader DLL theory, therefore, is both descriptive and prescriptive.

...