Theories of Action

The term theory of action figures prominently in several literatures, including the philosophy of action and practical reason, sociology, artificial intelligence and policy analysis. This entry focuses on the theory of action approach developed by Chris Argyris and Donald Schön, which has made seminal contributions to professional education, organizational learning and action research. These contributions include the concepts of double-loop learning, theories-in-use, organizational defensive routines and Action Science. They also include practices for helping individuals and organizations improve their ability to double-loop learn and for creating knowledge that can be used for this purpose.

The theory of action approach begins with the premise that human beings design action to achieve intended consequences. These designs can be seen as theories of action of the form ‘In situation S to achieve consequence C, do A’. Theories of action include the assumptions under which the actor believes the causal connection to hold and the values that make the intended consequence desirable. A theory of action consists of a complex set of interrelated propositions, a kind of master programme for producing action.

Theories of action are of two kinds. Espoused theories are those that individuals believe they follow and are able to state. Theories-in-use are those that can be inferred from actual behaviour. For example, an individual's espoused theory for handling a disagreement might be ‘Get all the issues on the table, and talk it through’. Observing what that individual actually does might lead to inferring the theory-in-use: ‘Emphasize facts that support my position, and downplay facts that support the other's position, while presenting myself as an even-handed seeker of truth’. Individuals are usually unaware of discrepancies between their espoused theories and their theories-in-use.

Seeing our behaviour as determined by theories of action directs attention to the knowledge we hold about people, situations and what causes what and also to the reasoning by which we bring our knowledge to bear as we design action in particular situations. But the theory of action approach does not presume that we are consciously aware of all this. Rather, in Schön's phrase, when we act intelligently, the knowing is in the action. We rely on tacit knowledge, much as native speakers utter sentences that are grammatically correct without thinking about or even being able to state the rules of grammar that govern their speech. Or, to switch analogies, we are like someone who knows how to ride a bicycle. We can maintain balance, make turns and dodge obstacles without thinking about how we are doing it. What we are not doing is thinking step by step through a set of rules of the kind that would be necessary if we were to programme a robot to ride the bike. If we tried to do that, we would fall.

What, then, do we gain by understanding behaviour in terms of theories of action? One answer is that it offers a way to reflect on our behaviour in order to become more effective. The idea that there is a design, a theory-in-use, that is governing our behaviour, that it may differ from our espoused theory and that we are probably unaware of discrepancies between the two provides a template for productive reflection. It tells us that we have to begin with the data of our actual behaviour and infer the theory-in-use rather than only introspect on what we were trying to do or what we think we did. Then, we can reconstruct and critically reflect on the assumptions and reasoning embedded in the theory-in-use that we have discovered in our action. We become researchers into our own practice.

...